THE CAMBRIDGE

ENCYCLOPEDIA OF
A

Borders
Book Shot
DAVID CRYSTA
$5.98
THE CAMBRIDGE
ENCYCLOPEDIA OF
LANGUAGE
DAVID CRYSTAL
PROFESSORIAL FELLOW,
UNIVERSITY COLLEGE OF NORTH WALES,
BANGOR

The right nf the

Cambridge
to print and trll
all manrit'
H0( gruntnl h\

Mfnry i III .n 1534

The Intwrsm hay printed
and published continuously
smct- 1584

CAMBRIDGE UNIVERSITY PRESS

CAMBRIDGE
NEW YORK PORT CHESTER MELBOURNE SYDNEY
Published by the Press Syndicate of the University of Cambridge
The Pitt Building, Trumpington Street, Cambridge CB2 1RP
40 West 20th Street, New York. NY 10011, USA
10 Stamford Road, Oaklcigh, Melbourne 3166, Australia

© Cambridge University Press 1987

Printed in the United States of America

British Library (dialoguing in publication data

Crystal, David
The Cambridge encyclopedia of language.
1. Language and languages
I. Title
400 P121

Library of Congress cataloging in publication data

Crystal, David 1941-

The Cambridge encyclopedia of language.
Bibliography.
Includes indexes.
1. Language and languages - Dictionaries.
2. Linguistics - Dictionaries. I. Title.
P29 064 1987 403 86-32637

ISBN 0-521-42443-7 paperback

 Contents

The Cambridge Encyclopedia of Language is 10 Social identity 38

organized in 11 Parts, comprising 65 thematic Language and social stratification, class,

sections. Each section is a self-contained presen- status, role, solidarity, and distance; the
problem of sexism.
tation of a major theme in language study, with
cross-references included to related sections 11 Contextual identity 48
and topics. Situationally determined varieties of
speech and writing; restricted and secret
language; verbal play and art; word
Editorial advisors vi games.
Preface vii
12 Stylistic identity and literature 66
The concept of style; authorship identity
I Popular ideas about language 1 and forensic linguistics; literary language
Widely held linguistic beliefs and in poetry,drama, and prose.
attitudes, and the basic functions of
language.
Ill The structure of language 81
1 The prescriptive tradition 2
Popular notions of linguistic authority The dimensions of language analysis that

and correctness; purism and language underlie all forms of language, whether
change; the role of linguistic description. spoken, written, or signed.

2 The equality of languages 6 13 Linguistic levels 82

Myths about primitive languages and The relationship between the main
language superiority. components of language analysis; models
of linguistic structure.
3 The magic of language 8
and verbal
Linguistic superstitions
14 Typology and universals 84
taboos; the mystical power of proper Analysing the structural similarities and
names. differences among the languages of the
4 The functions of language 10 world.
The many cultural, social, and personal
roles which language performs. 15 The statistical structure of language 86
The study of the statistical regularities
5 Language and thought 14 found in language; the frequency of
The complex relationship between sounds, letters, and words.
language and thinking; the notion of
language relativity. 16 Grammar 88
Syntax and morphology; the structure of
words, phrases, clauses, and sentences.
II Language and identity 17
The many ways in which language 17 Semantics 100
expresses a person's individuality or social The study of meaning in language; the
identity. semantic analysis of words and sentences.

6 Physical identity 18 18 Dictionaries 108

The relationship between language and The use and evaluation of dictionaries;
age, sex, physical type, and physical the past, present, and future of
condition; voiceprints; male vs female lexicography.
speech.
19 Names 112
7 Psychological identity 22
Patterns and trends in the use of personal
The relationship between language and
names; place names and their history.
personality, intelligence, and other
psychological factors. 20 Discourse and text 116
8 Geographical identity 24 The study of stretches of spoken and
The regional background of a speaker; written language above the sentence; the
accents, dialects, linguistic areas, and the nature of conversation; analysing textual
study of dialectology. structure.

9 Ethnic and national identity 34 21 Pragmatics 120

Language, ethnicity, and nationalism; the The factors that govern our choice of
problem of minority languages and language in social interaction; speech acts
dialects. and their analysis.

CONTENTS -hi
 IV The medium of language: VI The medium of language:
speaking and listening 123 signing and seeing 219
The study of the auditory-vocal channel The development and use of deaf sign
of communication; the production, languages.
transmission, and reception of speech.
35 Sign language 220
22 The anatomy and physiology of speech 124 Popular fallacies about sign language; the
The vocal tract and vocal organs; the development and use of signs by the deaf.
nature of articulation.
36 Sign language structure 222
23 The acoustics of speech 132 The way signs are used to convey
The nature of sound waves and the way grammatical contrasts; American Sign
they transmit speech; the sound spectro- Language.
graph and its use in speech sound analysis.
37 Types of sign language 224
24 The instrumental analysis of speech 138 The range of contrived sign languages;
Some of the techniques used in the analysis finger spelling, cued speech, and other
of speech acoustics and physiology. svstems.

25 Speech reception 142

The ear, and the process of hearing;
speech perception and its investigation. VII Child language acquisition 227
26 Speech interaction with machines 149 The study of the way children learn to
The and practice of automatic
principles understand and speak their mother
speech recognition and speech synthesis. tongue- methods, theories, and findings;
later language learning in school.
27 The sounds of speech 152
Phonetics; the description of vowels and 38 Investigating children's language 228
consonants; kinds of phonetic Techniques for finding out about child
transcription. language; speech production and
comprehension; theories of language
28 The linguistic use of sound 160
acquisition.
Phonology; phonemes, distinctive
features, and other models; comparing 39 The first year 236
the sound systems of languages. The development of infant vocalization;
early speech perception and interaction.
29 Suprasegmentals 169
The prosody of speech; the structure of 40 Phonological development 240
intonation; tone languages; the The sound system; the
acquisition of the
relationship between speech and music. learning of vowels, consonants, and
intonation.
30 Sound symbolism 174
The relationship between sounds and 41 Grammatical development 242
meaning; the role of onomatopoeia. The acquisition of grammar; growth in
sentence length and complexity.

42 Semantic development 244

The acquisition of vocabulary; first words
V The medium of language: and their content; distinguishing the
writing and reading 177 meanings of words.
The study of the development and functions 43 Pragmatic development 246
of written language, in all its forms. The acquisition of conversational skills;
the language of twins.
31 Written and spoken language 178
The relationship between speech and 44 Language development in school 248
writing; how sound is portrayed in The study of language in school; later
written language. oral development; learning to read and
write
32 Graphic expression 182
The physical substance of written
language; types of graphic expression; hand-
and electronic forms.
writing, print, typing, VIII Language, brain, and
33 Graphology 194
handicap 257
The writing system of a language; the The neurological basis of language, and
history of writing; the alphabet; spelling, the range of physical or psychological
punctuation, and other contrasts; systems problems that can give rise to disabilities
of shorthand. in spoken, written, or signed language.
34 The process of reading and writing 208 45 Language and the brain 258
Psychological accounts of the process of Brain structure and function; hemispheric
reading, writing, and spelling; spelling dominance and localization; slips of the
regularity and spelling reform. tongue and critical periods.

iv • CONTENTS
46 Language handicap 264 58 Artificial languages 352
Incidence, causation, and classification; The history of artificial languages, and
deafness, aphasia, dyslexia, dysgraphia; the present-day position; Esperanto,
disorders of voice, articulation, and Basic English, and other systems.
fluency; language delay; alternative
communication systems and
59 World languages 357
aids.
The international use of languages;
official languages; World English and its

varieties.
IX The languages of the world 283
The range of languages in past or present 60 Multilingualism 360
use- numbers, speakers, sources; Causes and extent of bilingual attitudes
and practice; language maintenance and
identifying and explaining linguistic
shift; language switching.
change.

47 How many languages? 284 61 Language planning 364

Government policies about language
Identifying, counting, and classifying the
languages of the world.
and use; bilingual educational
selection
programmes.
48 How many speakers? 286
Determining how many people speak a 62 Foreign language learning and
language; the world's most widely used teaching 368
languages and families. The role and status of foreign languages
in school and society; theories of
49 The origins of language 288 language learning, and methods of
Myths and experiments about the origins language teaching; language materials
of language; wolf children; humans and
and laboratories.
primates; the evidence of palaeontology.
63 Language for special purposes 378
50 Families of languages 292 The development of special varieties of
Discovering the history of languages; language in science, medicine, religion,
comparative philology; the language the law, the press, advertising, and
families of the world.
broadcasting; the related problems of
51 The Indo-European family 296 intelligibility and change.
The history of Indo-European languages,
where they are spoken, and how they are XI Language and
classified.
communication 395
52 Other families 304 The relationship between language and
The distribution, family grouping, and
other systems of human and non-human
use of the world's languages (other than
communication, and the scientific study of
Indo-European).
language.
53 Language isolates 326 64 Language and other communication
Languages which cannot be related to any
of the major families.
systems 396
Language defined; chimpanzee
54 Language change 328 communication; semiotics;
The identification of change in sounds, communication by non-linguistic sound,
grammar, and vocabulary; face, gesture, and touch.
glottochronology; explanations for
language change. 65 Linguistics 404
The history of ideas language study;
in
55 Pidgins and Creoles 334 domains and personalities in 20th-
The origins, distribution, and present-day century linguistics; linguisticjnethods.
use of the world's pidgins and Creoles.

Appendices
X Language in the world 341
I Glossary. 414
The problems of communication posed by
II Special symbols and abbreviations
the diversity of the world's languages and
varieties, and the search for solutions. used in the encyclopedia. 434

56 The language barrier 342 III Table of the world's languages. 436
The problems caused by foreign rV Further reading. 444
languages in the field of international
communication; language and the V References. 448
business world. VI Index of languages, families,
57 Translating and interpreting 344 dialects, and scripts. 452
The and practice of translating
principles
VII Index of authors and personalities. 456
and interpreting; the role of machine
translation. VIII Index of topics. 458

CONTENTS. v
 EDITORIAL ADVISORS
Professor Charles Ferguson
Department of Linguistics, Stanford University
Professor Victoria A. Fromkin
Graduate Division, University of California
Professor Shirley Brice Heath
School of Education, Stanford University

Professor Dell Hymes

Department of Anthropology, University of Virginia
Dr Stephen Levinson
Department of Linguistics, University of Cambridge
Dr John Marshall
Neuropsychology Unit,The Radcliffe Infirmary, Oxford
Professor Wilga Rivers
Department of Romance Languages and Literatures, Harvard University
Professor Sheldon Rosenberg
Department of Psychology, University of Illinois
Professor Klaus Scherer
Department of Psychology, University of Geneva
Professor Roland Sussex
Department of Russian, University of Melbourne
Professor Jan Svartvik
Department of English, Lund University
Professor Michael Twyman
Department of Typography and Graphic Communication,
University of Reading

The late Professor C. F. and Professor F. M. Voegelin

Department of Anthropology, Indiana University
Preface
My purpose in writing this book is to celebrate
the existence of human language, and to provide
a tribute to those who engage in its study. Its aim
isto illustrate the enormous diversity of the world's
languages, and the great range, complexity, and
beauty of expression that can be encountered in
any of them, whether spoken by millions or by
hundreds - from the most polished formulations
of respected literature to the most routine utter-
ances of everyday conversation. At the same time,
I want to convey something of the fascination and
value of linguistic research, which has led to in-
numerable general findings about language struc-
ture, development, and use, and which has
prompted so many important applications in rela-
tion to the problems of the individual and society.
The book therefore operates on two levels. It
reflects the kind of interest in language history and
behaviour that we encounter daily as we argue over
the history of a word's meaning or listen in fascina-
tion to a young child's early attempts to talk. At
the same time, it reflects a deeper level of interest,
arising out of our attempt to make sense of what
we observe, and to find patterns and principles in
it - an interest that can lead to a professional career
in linguistic research or in one of the language-
related professions, such as language teaching or
therapy.
have certain practical aims also. I hope the book
I optimistic outlook — thanks largely to the backing
will help promote an informed awareness of the
complexity of human language, draw attention to
the range of human problems that have a linguistic
cause or solution, and emphasize the fact that
people have language rights which should not be
neglected. Earlier this year, in fact, I received a copy
of a plea for a 'Declaration of Individual Linguistic
Rights', sponsored by Francisco Gomes de Matos
of the Federal University of Pernambuco, Recife,
Brazil. The plea points to the widespread occur-
rence of linguistic prejudice and discrimination
around the world, and to the problems people face
when they wish to receive special help in language
learning and use. All people have the right to use
their mother tongue, to learn a second language,
to receive special treatment when suffering from
a language handicap but in many parts of the
. . . staff of the Press, for their invaluable advice during
world, these rights are absent or inadequately pro-
visioned. Only concentrated public attention on the
issues will promote the recognition of such rights,
and it is my hope that this encyclopedia will play
itspart in helping to develop a climate where people
will sense the importance of language in the indivi-
dual and in society, and act accordingly.
I have used the term 'encyclopedia', but not with-
out misgivings: if there were a term for 'embryo
encyclopedia', it would be better. The subject of
language is truly vast, and it is possible only to
make a start in 480 pages. In particular, because
my background is in linguistics, I am conscious
of paying insufficient attention to other traditions
of thinking and research, such as in philosophy,
psychology, and artificial intelligence. Also,
although I write from a linguistic point of view,
this book is not an introduction to linguistics: I
have stopped short of a discussion of the many
approaches to the analysis of language that linguis-
tics provides, and I give few technical details about
theoretical differences, hoping that my references
will provide sources for those who wish to enquire
into these matters further.
This is just one of many apologies scattered
throughout the book. Facts about the use of lan-
guage are extremely difficult to come by, and, when
obtained, fall quickly out of date. Language
changes rapidly, as do the techniques and theories
that scholars devise to study it. On the other hand,
few books can have been written with such an
and enthusiasm of the team of editorial advisors
appointed by Cambridge University Press. To
know that one's plans and material will be scruti-
nized by scholars of such eminence is immensely
reassuring, and I have benefitted immeasurably
from their advice while the book was being written.
I am therefore delighted to acknowledge my debt
of gratitude to these advisors: it has been a privilege
to have their support, and I hope the result does
them no disservice. Needless to say, the responsi-
bility for what remains is mine alone.
Finally, it is my pleasant duty to thank members
of the Department of Linguistic Science, University
of Reading, and of the Centre for Information on
Language Teaching, London, for help in research-
ing aspects of the work; the editorial and design
the period of this book's preparation; and, above
all, the support and assistance of my wife, Hilary,
in helping this project come to fruition.
DAVID CRYSTAL
PREFACE • vii
PARTI
Popular ideas about language

Why does language provide such a fascinating
object of study? Perhaps because of its unique role
in capturing the breadth of human thought and
endeavour. We look around us, and are awed by
the variety of several thousand languages and dia-
lects, expressing a multiplicity of world views, liter-
atures, and ways of life. We look back at the
thoughts of our predecessors, and find we can see
only as far as language lets us see. We look forward
in time, and find we can plan only through lan-
guage. We look outward in space, and send sym-
bols of communication along with our spacecraft,
who we are,
to explain in case there
who wants to know.
Alongside this, there is the importance
to language, as a means of understanding ourselves
and our society, and of resolving some of the prob-
lems and tensions that arise from human inter-
action. No sector of society is unaffected, and all
can benefit from the study of the linguistic factors
that constitute a barrier, as well as a
munication. But linguistic problems rarely admit
simple solutions, and it is this elementary observa-
tion that has led to the present work.
The main aim of this encyclopedia is to provide
information about all aspects of language structure
and use, so that the complex forces which act upon
language, and upon the people who use it, will be
more readily understood. The work is founded on
the belief that the systematic analysis and discus-
sion of language in an objective way is an essential
step forward towards any world in
respect and tolerance is a reality. 'They don't speak
like us; therefore they aren't like us; therefore they
don't like us.' This is the kind of logic that the
information in this book seeks to deny.
But such a world is way off. The world
a long
we currently see displays many signs of linguistic
intolerance and tension. They appear most notice-
ably in the language riots of India or Belgium, and
in the disfiguredroad signs of Wales or northern
Spain; but they are present in more subtle ways,
in the unmotivated preservation of traditional pur-
istlinguistic practices in many schools, and in the
regular flow of complaints on the world's radio
is anyone there channels and in the press about other people's
usage.
we attach opening part of this book, therefore, we
In the
look the most important ideas that have
at
influenced the nature of popular opinion about lan-
guage, in both 'civilized' and 'uncivilized' societies.
We begin with the idea of correctness, and the his-
torical development of prescriptive attitudes to lan-
means of com- guage. We look at the desire to keep language
'pure', as encountered in the movements in support
of language academies, and the general concern
over linguistic change. We address the proposition
that all languages are equal, in the face of the wide-
spread view that some are more equal than others.
This is followed by a discussion of popular beliefs
about the magical and mystical power of language,
and a general investigation of the wide range of
functions that language performs in everyday life.
Part I then concludes by considering the intriguing
which mutual but intricate question of the relationship between
language and thought.

The cultural diversity of language, as reflected in a disputa-

tionbetween three medieval doctors (an engraving by
Marcantonio Raimondi), a ritual debate among Rotinese
elders, and a confrontation between human beings and their
computer database.
1 The prescriptive tradition

At the beginning of any book on language, readers

Prescriptivism
have a distinct advantage over the author. More
than in most areas of enquiry, they already 'know' In most general sense, prescriptivism is the view
its

the subject, in the sense that they already speak
and read a language. Moreover, because in modern
societies linguistic skills are highly valued,
readers will have definite views about the nature
of language and how it should function. This is
not the usual state of mind of someone who opens
an encyclopedia on, say, astronomy, Roman myth-
ology, or physics.
We must therefore begin our investigation by
looking at the main opinions and beliefs people
already hold about language as a result of the nor-
that one variety of language has an inherently
higher value than others, and that this ought to
many be imposed on the whole of the speech community.
The view is propounded especially in relation to
grammar and vocabulary, and frequently with
reference to pronunciation. The variety which is
favoured, in this account, is usually a version of
the 'standard' written language, especially as
encountered in literature, or in the formal spoken
language which most closely reflects this style. Ad- George Orwell (1903-50)
herents to this variety are said to speak or write In Politics and the English

mal processes of education and social development. 'correctly'; deviations from it are said to be 'incor- Language (1 947), Orwell
lists six rules 'that one can
These views will provide a frame of reference fami- rect'.
rely on when instinct fails'.
liar to many readers, and they will also act as a All the main European languages have been stu- These rules were not written
point of departure for the detailed, systematic, and died prescriptively, especially in the 18th century with literary or scientific
objective study of the subject in the following approach to the writing of grammars and dictio- language in mind, but with
the everyday need to foster
pages. naries. The aims of these early grammarians were
language as an instrument
threefold: (a) they wanted to codify the principles forexpressing and not for
AN EMOTIONAL SUBJECT of their languages, to show that there was a system concealing or preventing
It isnot easy to be systematic and objective about beneath the apparent chaos of usage, (b) they thought'. In this way, Orwell
hoped, it would be possible
language study. Popular linguistic debate regularly wanted a means of settling disputes over usage,
to halt the decline in the
deteriorates into invective and polemic. Language (c) they wanted to point out what they felt to be
language, which he saw as
belongs to everyone; so most people feel they have common errors, in order to 'improve' the language. intimately connected with the
a right to hold an opinion about it. And when opi- The authoritarian nature of the approach is best political chaos' of the time.
nions differ, emotions can run high. Arguments can characterized by its reliance on 'rules' of grammar. 1 Never use a metaphor, si-
flare as easily over minor points of usage as over Some usages are 'prescribed', to be learnt and fol- mile or other figure of speech
major policies of linguistic planning and education lowed accurately; others are 'proscribed', to be which you are used to seeing
in print.
(§61). avoided. In this early period, there were no half-
2 Never use a long word
Language, moreover, is a very public behaviour, measures: usage was either right or wrong, and when a short one will do.
so that it is easy for different usages to be noted it was the task of the grammarian not simply to 3 If it is possible to cut a
and criticized. No part of society or social be- record alternatives, but to pronounce judgment word out, always cut it out.
4 Never use the passive
haviour is exempt: linguistic factors influence our upon them.
where you can use the ac-
judgments of personality, intelligence, social status, These attitudes are still with us, and they moti-
tive.
educational standards, job aptitude, and many vate a widespread concern that linguistic standards 5 Never use a foreign
other areas of identity and social survival. As a should be maintained. Nevertheless, there is an phrase, a scientific word or a
result, it is easy to hurt, and to be hurt, when lan- alternative point of view that is concerned less with jargon word if you can think

of an everyday English equi-

guage use is unfeelingly attacked.
The American linguist Leonard Bloomfield
(1887-1949) discussed this situation in terms of
three levels of response people give to language.
'standards' than with the facts of linguistic usage.
valent.
This approach is summarized in the statement that 6 Break any of these rules
it is the task of the grammarian to describe, not sooner than say anything
prescribe — to record the facts of linguistic diversity, outright barbarous.

The 'primary response' is actual usage. 'Secondary and not to attempt the impossible tasks of evaluat- (See further, p. 378.)

responses' are the views we have about language,
often expressed in some kind of terminology. 'Ter-
tiary responses' are the feelings which flare up when
anyone dares to question these views. Bloomfield
tells the story of visiting a doctor who was quite
firm in his view that the Amerindian language
Chippewa had only a few hundred words (p. 6).
When Bloomfield attempted to dispute the point,
the doctor turned away and refused to listen. Irra-
tional responses of this kind are unfortunately all
too common; but everyone is prone to them —
linguist and non-linguist alike.
ing language variation or halting language change.
In the second half of the 18th century, we already
find advocates of this view, such as Joseph Priestley,
whose Rudiments of English Grammar (1761)
insists that 'the custom of speaking is the original
and only just standard of any language'. Linguistic
issues, it is argued, cannot be solved by logic and
legislation. And this view has become the tenet of
the modern linguistic approach to grammatical
analysis.
In our own time, the opposition between 'des-
criptivists' and 'prescriptivists' has often become

2 •
I POPULAR IDEAS ABOUT LANGUAGE
extreme, with both sides painting unreal pictures
of the other. Descriptive grammarians have been
presented as people who do not care about stan- Where traditional grammatical rules come
dards, because of the way they see all forms of
usage as equally valid. Prescriptive grammarians from
have been presented as blind adherents to a histori-
Example of a
cal tradition. The opposition has even been pre- prescriptive rule Descriptive comment
sented in quasi-political terms — of radical
Latin and Greek
liberalism vs elitist conservatism.
The unchanging form of these You should say or write The Latin rule is not universal. In
If these stereotypes abandoned, we can see
ar< languages, the high prestige they It is I and not me, It is Arabic, for example, be is

that both approaches are important, and have more held in European education, and the because the verb be is followed bv the accusative. In
undisputed brilliance of classical followed by the English, me is the educated
in common than is often realized — involving a literature led to their adoption as nominative case in informal norm; / is felt to be very
mutual interest in such matters as acceptability, models of linguistic excellence by Latin, not the formal. In French, only moi is
grammarians of other languages. accusative. possible (c'est moi, etc.)
ambiguity, and intelligibility. The descriptive
approach is essential because it is the only way The written language
Writing is more careful, prestigious You should say and Whom is common in writing, and
in which the competing claims of different stan- whom and not
and permanent than speech, write in formal styles of speech; but
dards can be reconciled: when we know the facts especially in the context of who, in such sentences who is more acceptable in

of language use, we are in a better position to avoid literature.People are therefore often as — did you speak to? informal speech. The rules which
told to speak as they would write. govern acceptable speech and
the idiosyncrasies of private opinions, and to make writing are often very different.
realisticrecommendations about teaching or style. Logic
The prescriptive approach provides a focus for the Many people feel that grammar You shouldn't say / Here, two negatives do not make
should be judged insofar as it haven't done nothing a positive, but a more emphatic
sense of linguistic values which everyone possesses,
follows the principles of logic. because two negatives negative— a construction which is
and which ultimately forms part of our view of Mathematics, from this viewpoint, make a positive. found in many languages (e.g.
social structure, and of our own place within it. is the ideal use of language. French, Russian). The example is

not acceptable in standard

After 200 years of perhaps sanguine
dispute, it is
English, but this is the result of
to expect any immediate rapport to be achieved, social factors, not the dictates of

but there are some grounds for optimism, now that logic.

sociolinguists (p. 410) are beginning to look more

seriously at prescriptivism in the context of
explaining linguistic attitudes, uses, and beliefs.

Murray's Grammar Left: Lindley Murray (1745-1826)

One of the most influential and artificial, Latinate analy-

grammars of the 1 8th cen- sis which was to fuel two
was Robert Lowth's
tury centuries of argument. In
ENGLISH GRAMMAR,
Short Introduction to English Rule 1 6, for example, we find
Grammar (1 762). This was the negation principle illus- ADAPTED TO THE
the inspiration for Lindley trated: Two negatives, in

Murray's widely used English English, destroy one

Grammar^ 794). Both gram- another, or are equivalent to DIFFERENT CLASSES OF LEARNERS.
mars went through over 20 an affirmative.'
editions in the decades fol- Murray's rules were widely
lowing publication. taught, and formed the basis
Murray's book had an for much of the linguistic pur-
enormous influence on ism still encountered today.
AN APPENDIX,
school practice and popular However, they were also fier- CONTAINING
attitudes, especially in the cely attacked. One writer in
tULt! AtB OlSMrjTIOXS,
USA. His alliterative axiom the American Journal of Edu-
contains several watchwords cation (in 1 826) compares rot ASSISTING THC MOtl ADVASCID tTVDCXr:
of prescriptivism: 'Perspi- the grammar to a foreign to iriiTE trim ntincurrr jsd jccvticr.
cuity requires the qualities of rack on which our simple lan-
purity, propriety and preci- guage has been stretched'.
sion'. Another 833) insists that
(in 1

Some of Murray's general grammarians should dis-

linguistic principles were un-
exceptionable, such as
Keep clear of double mean-
ing or ambiguity' and 'Avoid
unintelligible words or
phrases.' But most of his
analyses, and the detailed
principles of his Appendix,
cover'and not 'invent' rules. Br LINDLEY MURRAY.
Long before the advent of
modern linguistics, the battle THE NINTH EDITION,
lines of both descriptivism WITH COKRICTIONI AND ADDITION}.
and prescriptivism had been
clearly established.
gojfc:
fntwed bf T. Witea nd r. Speace, Hjr>-Ou«en«, .D©" 1 ty 5*»*ee*

'Rules and observations for *0* LONGMAN AND HIS, t ATCINOSTtft.ftO W DAtTOV ;

AMD HAIVIT, GB ACI.CHDSCH JTKIIT, LONDON

promoting perspicuity in j

AND WILSON AND 1PINCK, YORK.

speaking and writing', con- Right: Murray's English 1804.
tain the kind of arbitrary rule Grammar Priee bouod, Amy, 5*. 64. ... h> eim bout*, twfw/m r^f, 5$.

1 THE PRESCRIPTIVE TRADITION •

3
 the reputation of the members of this academy
The academies
would be enough to make them the allowed judges of
Some felt that the best way to look
countries have style and language; and no author would have the
after a language is to place it in the care of an impudence to coin without their authority There . . .

academy. In Italy, the Accademia della Crusca was should be no more occasion to search for derivations
founded as early as 1582, with the object of purify- and constructions, and it would be as criminal then to
ing the Italian language. In France, in 1635, Cardi- coin words as money.

nal Richelieu established the Academie francaise, In 1712, Jonathan Swift presented his Proposal for
which set the pattern for many subsequent bodies. Correcting, Improving and Ascertaining the Eng-
The statutes of the Academie define as its principal lish Tongue, in which he complains to the Lord
function: Treasurer of England, the Earl of Oxford, that
to labour with all possible care and diligence to give our language extremely imperfect; that its daily
is
our language, and to render it pure,
definite rules to improvements are by no means in proportion to its daily Daniel Defoe
eloquent, and capable of treating the arts and sciences. corruptions; that the pretenders to polish and refine it (16607-1731)
have chiefly multiplied abuses and absurdities; and that
The 40 academicians were drawn from the ranks
in many instances it offends against every part of
of the church, nobility, and military - a bias which grammar.
continues to the present day. The Academie's first
dictionary appeared in 1694. His academy would 'fix our language for ever', for,

Several other academies were founded in the I am of the opinion, it is better a language should not
1 8th and 1 9th centuries. The Spanish Academy was be wholly perfect, than it should be perpetually
founded in 1713 by Philip V, and within 200 years changing.
corresponding bodies had been set up in most
The idea received a great deal of support at the
South American Spanish countries. The Swedish was done. And in due course,
time, but nothing
Academy was founded in 1786; the Hungarian in opposition to the notion grew. It became evident
1830. There are three Arabic academies, in Syria,
that the French and Italian academies had been
Iraq, and Egypt. The Hebrew Language Academy
unsuccessful in stopping the course of language
was set up more recently, in 1953. change. Dr Johnson, in the Preface to his Diction- Jonathan Swift (1667-1745)
In England, a proposal for an academy was made
ary, is under no illusion about the futility of an
in the 17th century, with the support of such men
academy, especially in England, where he finds 'the
as John Dryden and Daniel Defoe. In Defoe's view,
spirit of English liberty' contrary to the whole idea:

When we see men grow old and die at a certain time

one after another, century after century, we laugh at
the elixir that promises to prolong life to a thousand
years; and with equal justice may the lexicographer be
derided, who being able to produce no example of a
nation that has preserved their words and phrases from
mutability, shall imagine that his dictionary can embalm
his language, and secure it from corruption, and decay,
that it is in his power to change sublunary nature, or
clear the world at once from folly, vanity, and
affectation.

From time to time, the idea of an English Academy

continues to be voiced, but the response has never

CRQ< been enthusiastic. A similar proposal in the USA

Kippers sur toast? Adver-
tisementslike this could be
found, with the appropriate
language change, in almost
any European city. They
illustrate the way English has
permeated public life,
despite the efforts of many
countries to stop it. The Ger-
man post example,
office, for
insisted for many years that
Fernsprecher should be
used on phone booths,
though Telefon was far more
common in speech; but in
1 981 they made the change.
In 1977, the French went so
far as to pass a law banning
the use of English loan
words in official contexts, if
an equivalent word exists in
French - but it is a law ho-
noured more in the breach
than in the observance.
Whether one approves or
not, theacademies seem to
be no match for Franglais,
Angleutsch, Swedlish,
Spanglish, and all the other
hybrids which have become
so noticeable in recent years
(§§55,61).
was also rejected. By contrast, since the 18th cen-
tury, there has been an increasing flow of individual
grammars, dictionaries, and manuals of style in all
parts of the English-speaking world.
Language change
The phenomenon of language change probably
attracts more public notice and criticism than any
other linguistic issue. There is a widely held belief
that change must mean deterioration and decay.
Older people observe the casual speech of the
young, and conclude that standards have fallen
markedly. They place the blame in various quarters
- most often in the schools, where patterns of lan-
guage education have changed a great deal in recent

4 •
I POPULAR IDEAS ABOUT LANGUAGE
years (§44), but also in state public broadcasting
institutions, where any deviations from traditional
norms provide an immediate focus of attack by
conservative, linguistically sensitive listeners. The
concern can even reach national proportions, as
in the widespread reaction in Europe against what
is thought of as the 'American' English invasion.
UNFOUNDED PESSIMISM
It isunderstandable that many people dislike
change, but most of the criticism of linguistic
change is misconceived. It is widely felt that die
contemporary language illustrates the problem at
its worst, but this belief is shared by every gener-
ation. Moreover, many of the usage issues recur
across generations: several of the English contro-
versies which are the focus of current attention can
be found in the books and magazines of the 18th
and 19th centuries - the debate over it's me and
very unique, for example. In The Queen's English
(1863), Henry Alford, the Dean of Canterbury, lists
a large number of usage issues which worried his
contemporaries, and gave them cause to think that
the language was rapidly decaying. Most are still
with us, with the language not obviously affected.
In the mid- 19th century, it was predicted that
British and American English would be mutually
unintelligible within 100 years!
There are indeed cases where linguistic change
can lead to problems of unintelligibility, ambiguity,
and social division. If change is too rapid, there
can be major communication problems, as in con-
temporary Papua New Guinea - a point which
needs to be considered in connection with the field
of language planning (§§55, 61). But as a rule, the
parts of language which are changing at any given
time are tiny, in comparison to the vast, unchang-
ing areas of language. Indeed, it is because change
is so infrequent that it is so distinctive and notice-
able. Some degree of caution and concern is there-
fore always desirable, in the interests of
maintaining precise and efficient communication;
but there are no grounds for the extreme pessimism
and conservatism which is so often encountered
- and which in English is often summed up in such
slogans as 'Let us preserve the tongue that Shake-
speare spoke.'
THE INEVITABILITY OF CHANGE
For the most part, language changes because
society changes (§10). To stop or control the one
requires that we stop or control the other
which can succeed to only a very limited extent.
Language change is inevitable and rarely predic-
table, and those who try to plan a language's future
waste their time if they think otherwise - time
which would be better spent in devising fresh ways
of enabling society to cope with the new linguistic
forms that accompany each generation. These
days, there is in fact a growing recognition of the
need to develop a greater linguistic awareness and
tolerance of change, especially in a multi-ethnic
William Caxton
One of the earliest English
voices to complain about the
problems of linguistic change
was William Caxton (1422?-
91 He was writing at a time
).
when English had under-
gone its greatest period of
change, which had resulted
in a major shift in pronuncia-
tion, the almost total loss of
Anglo-Saxon inflections, and
an enormous influx of new
vocabulary, mainly from
French:
And certaynly our language
now used varyeth ferre from
that whiche was used and
spoken whan was borne
I . .
And that comyn Englysshe
that is spoken in one shyre
varyeth from a nother. In so
moche that in my dayes hap-
pened that certayn mar-
chauntes were in a shippe in
Tamyse [Thames] for to
have sayled over the see into
Zelande, and for lacke of
wynde thei taryed atte for-
lond, and wente
to lande for
society. This requires, among other things, that to refreshethem. And one of
schools have the knowledge and resources to teach theym named Sheffelde, a
a common standard, while recognizing the exis- mercer, cam in to an hows
tence and value of linguistic diversity. Such policies and axed for mete, and spe-
cyally he axyd after eggys'.
provide a constructive alternative to the emotional
And good wyf answerde
the
attacks which are so commonly made against the that she coude speke no
development of new words, meanings, pronuncia- Frenshe. And the marchaunt
tions, and grammatical constructions. But before was angry, for he also coude
speke no Frenshe, but wold
these policies can be implemented, it is necessary
have hadde egges, and she
to develop a proper understanding of the inevitabi-
understode hym not. And
lity and consequences of linguistic change (§54). thenne at last a nother sayd
Some people go a stage further, and see change that he wolde have 'eyren'.
in language as a progression from a simple to a Then the good wyf sayd that
she understod hym wel. Loo!
complex state — a view which was common as a
What sholde a man in thyse
consequence of 19th-century evolutionary think- dayes now wryte, egges' or
ing. But there is no evidence for this view. Lan- 'eyren? Certaynly, is harde it
guages do not develop, progress, decay, evolve, or to playse every man by
cause of dyversite &
act according to any of the metaphors which imply
chaunge of langage.
a specific endpoint and level of excellence. They
simply change, as society changes. If a language (Preface to Eneydos, 1490;
modernized punctuation)
dies out, it does so because its status alters in
Caxton's plaint echoes
socLty, as other cultures and languages take over
through the ages, though
its role: it does not die because it has 'got too old',
problems of linguistic change
or 'become too complicated', as is sometimes main- have never been so serious
tained. Nor, when languages change, do they move since, with the subsequent
in a predetermined direction. Some are losing standardization of English,
and the spread of the written
- a task inflections; some are gaining them. Some are mov-
language.
ing to an order where the verb precedes the object;
others to an order where the object precedes the
verb. Some languages are losing vowels and gaining
consonants; others are doing the opposite. If meta-
phors must be used to talk about language change,
one of the best is that of a system holding itself
in a state of equilibrium, while changes take place
within it; another is that of the tide, which always
and inevitably changes, but never progresses, while
it ebbs and flows.
I THE PRESCRIPTIVE TRADITION 5
 2 The equality of languages

It comes near to stating the obvious that all lan- plex grammar: there may be relative simplicity in The Roman goddess
guages have developed to express the needs of their one respect (e.g. no word-endings), but there seems Fort una, holding a cornuco-

always to be relative complexify in another (e.g. piaand a rudder - an appro-

users, and that in a sense all languages are equal.
But this tenet of modern linguistics has often been
denied, and still needs to be defended. Part of the
problem is that the word 'equal' needs to be used
very carefully. We do not know how to quantify
language, so as to be able to say whether all lan-
guages have the same 'amounts' of grammar, pho-
nology, or semantic structure (§§16, 17, 28). There
may indeed be important differences in the struc-
tural complexity of language, and this possibility
needs to be investigated. But all languages are
arguably equal in the sense that there is nothing
intrinsically limiting, demeaning, or handicapping
about any of them. All languages meet the social
and psychological needs of their speakers, are
equally deserving of scientific study, and can pro-
vide us with valuable information about human
nature and society'. This view is the foundation on
which the whole of the present book is based.
priate deity to associate with
word-position). People sometimes think of lan- the uncertain destinies of
guages such as English as 'having little grammar', languages.
because there are few word-endings. But this is
once again (§1) the unfortunate influence of Latin,
which makes us think of complexity in terms of
the inflectional system of that language.
Simplicity and regularity are usually thought to
be desirable features of language; but no natural
language is simple or wholly regular. All languages
have intricate grammatical rules, and all have
exceptions to those rules. The nearest we come to
real simplicity with natural languages is in the case
of pidgin languages (§55); and the desire for regu-

'Primitive' languages
There are, however, several widely held misconcep-
tions about languages which stem from a failure
to recognize this view. The most important of these
is the idea that there are such things as 'primitive'
languages - languages with a simple grammar, a
few sounds, and a vocabulary of only a few
hundred words, whose speakers have to compen-
sate for their language's deficiencies through ges-
tures. Speakers of 'primitive' languages have often Navaho Indian Chief Manulito
been thought to exist, and there has been a great
deal of speculation about where they might live,
and what their problems might be. If they relied Simple savages?
on gestures, how would they be able to communi- Edward Sapir was one of will-give to thee he-or-they- level of abstraction

cate at night? Without abstract terms, how could the first linguists to attack in-future introduced by some
the myth that primitive languages (expressed by
they possibly develop moral or religious beliefs? Southern Palute
peoples spoke primitive round thing and visible) -
In the 19th century, such questions were common, ma^a-vaania-aka-arja-'mi
languages. In one study, quite contrary to the claim
give will visible-thing visible-
and it was widely thought that it was only a matter he compared the gramma- that primitive peoples could
creature thee
of time before explorers would discover a genuinely tical equivalents of the sen- only talk about concrete
tence he will give it (a Yana objects.
primitive language.
stone) to you in six Amer- ba-ja-ma-si-wa-'numa Sapir also gave part of the
The fact of the matter is that every culture which indian languages. (Hy- round-thing away to does-or- fullTakelma verb paradigm:
has been investigated, no matter how 'primitive' phens separate the parts of will done-unto thou-in-future
'okuspi gives gave it to
it may be in cultural terms, turns out to have a the Indian sentences, and
Nootka you
in the literal translations
fully developed language, with a complexity com- o'-yi-'aqx-'at-e'ic 'bspink will give to you
that follow they join words that give will done-unto thou- v 6spi
parable to those of the so-called 'civilized' nations. can give to you
that are equivalent to a
Anthropologically speaking, the human race can art 'bspik evidently gave to
single Indian form. For pho-
Navaho you
be said to have evolved from primitive to civilized netic symbols, see p. 442.)
states, but there is no sign of language having gone n-a-yi-diho-'a'l
Wishram He points out the similarity to
thee to transitive-marker will
through the same kind of evolution (§48). There a-c-i-m-l-ud-a
the way the verb varies in
round-thing-in-future - a comparison which
Latin
are no 'bronze age' or 'stone age' languages, nor will he him thee to give will
Among many fascinating many traditional scholars
have any language types been discovered which Takelma would have considered
features of these complex to
correlate with recognized anthropological groups 'ok-t-xpi-nk grammatical forms, note the verge on blasphemy!
(pastoral, nomadic, etc.). All languages have a com-

6 •
I POPULAR IDEAS ABOUT LANGUAGE
 major motivation
-

larity is a for the development Nationalism In the 1 8th and

of auxiliary languages (§58). But these are the only 19th centuries, language
evaluations were often tied
exceptions. Similarly, there is no evidence to sug-
to questions of national iden-
gest that some languages are in the long term 'easier tity (§9), especially in
for children to learn' than others - though in the

^
Germany, in a school of

short term some linguistic features may be learned thought which can be traced
at different rates by the children of speakers of dif-
back to the view of Johann
Herder: 'Has a nation any-
ferent languages (Part vm). thing more precious than the
None of this is to deny the possibility of linguistic language of its fathers?'
differences which correlate with cultural or social Johann Gottlieb Fichte
features (such as the extent of technological devel- (1762-1814) praised the
opment), but these have not been found; and there M German language, and
missed others, in his
dis-

is no evidence to suggest that primitive peoples are

Addresses to the German
in any sense 'handicapped' by their language when
they are using it within their own community. ^^A Nation (1 807), even to the
extent of claiming that the
native German speaker 'can
always be superior to the

Languages of excellence foreigner and understand

At the other end of the from so-called 'primi-

scale
tive' languages are opinions about the 'natural
r^x him fully, even
the foreigner understands
himself. But comparable
better than

claims were made for French

superiority' of certain languages. Latin and Greek
and Spanish; and English
were for centuries viewed as models of excellence was similarly lauded by
in western Europe because of the literature and Thomas Macaulay (1800-
thought which these languages expressed; and the 59): in his Minute on Educa-
tion (1 835), referring to the
study of modern languages is still influenced by
languages of India, he wrote
the practices of generations of classical linguistic Johann Herder (1744-1803) that English 'stands pre-
scholars (p. 374). eminent even among the
The idea that one's own
language is superior to attitudes in presenting the view that in Paradise languages of the West ... It
others is widespread, but the reasons given for the Adam spoke Danish, God spoke Swedish, and the may safely be said that the
literature now extant in that
superiority vary greatly. A language might be serpent spoke French. language is of greater value
viewed as the oldest, or the most logical, or the than all the literature which
language of gods, or simply the easiest to pro- A LINGUISTIC MYTH three hundred years ago was
extant in all the languages of
nounce or the best for singing. Arabic speakers, A belief that some languages are intrinsically su-
the world together.'
for example, feel that their classical language is perior to others widespread, but it has no basis
is

the most beautiful and logical, with an incompar- in linguistic fact. Some languages are of course
able grammatical symmetry and lexical richness. more useful or prestigious than others, at a given
Classical Arabic is strongly identified with religion period of history, but this is due to the preeminence
(p. 384), as the language of the Qur'an is held to of the speakers at that time, and not to any inherent
provide miraculous evidence of the truth of Islam. linguistic characteristics. The view of modern
From this viewpoint, it would be self-evident that, linguistics is that a language should not be valued
as God chose Arabic as the vehicle of his revelation on the economic influence
basis of the political or
to his Prophet, this must be the language used in of its speakers. were otherwise, we would have
If it
heaven, and thus must be superior to all others. to rate the Spanish and Portuguese spoken in the
However, a similar argument has been applied 16th century as somehow 'better' than they are
to several other languages, such as Sanskrit and today, and modern American English would be
Classical Hebrew, especially in relation to claims 'better' than British English. Yet when we make
about which language is the oldest (§49). For such comparisons, we find only a small range of
example, J. G. Becanus (1518-72) argued that Ger- linguistic differences, and nothing to warrant such
man was other languages. It was
superior to all sweeping conclusions.
the language Adam spoke in Eden, but it was not At present, it is not possible to rate the excellence
affected in the Babel event, because the early Ger- of languages in linguistic terms. And it is no less
mans (the Cimbrians) did not assist in the construc- difficult to arrive at an evaluation in aesthetic,
tion of the tower. God later caused the Old philosophical, literary, religious, or cultural terms.
Testament to be translated from the original Ger- How, ultimately, could we compare the merits of
man (no longer extant) into Hebrew. Latin and Greek with the proverbial wisdom of
There have been many other spurious linguistic Chinese, the extensive oral literature of the Polyne-
evaluations, reflecting the sociopolitical situation sian islands, or the depth of scientific knowledge
of the time. Charles V of Germany (who ruled from which has been expressed in English? Perhaps one
1519 to 1558) is said to have
spoken French to day some kind of objective linguistic evaluation
men,Italian to women, Spanish to God, and Ger- measure will be devised; but until then, the thesis
man to horses! The Swedish writer, Andreas that some languages are intrinsically better than
Kempe (1622-89), satirized contemporary clerical others has to be denied.

2 THE EQUALITY OF LANGUAGES •

7
3 The magic of language
The magical influence of language is a theme which
reverberates throughout the literatures and legends
of the world. Language, especially in its written
form, is thought to contain special powers, which
only the initiated are allowed to understand or con-
trol. The beliefs are often linked to a myth about
the divine origins of language (§49), but they
extend beyond this, to influence religious activities
of all kinds, and to reflect a widespread primitive
superstition about objects and events which have
a symbolic meaning and use.
The belief that words control objects, people,
and spirits can be seen in the use of magical formu-
lae, and many other
incantations, litanies of names,
rites in black and white magic and in organized
religion. The language is thought to be able to cure
sickness, keep evil away, bring good to oneself and
harm to an enemy. Such language usually has to
be used with great exactitude, if an effect is to be
obtained: meticulous attention is paid to pronun-
ciation, phraseology, and verbal tradition (a factor
which appears, most notably, in the history of
Sanskrit and Massoretic Hebrew). There often has
to be a great deal of repetition, in order to intensify
the power of the words. The language, however,
does not have to be intelligible to have its effect:
many magical formulae are meaningless to those
who use them, but there is still great belief in their
efficacy (p. 11).
Cases of abound. To
linguistic superstition
primitive peoples, the written language must
appear to be omniscient, when encountered for the
first time. Several stories
of illiterate people
tell
stealing an object from a parcel, and being found
out when they delivered the message which accom-
panied it. The
writing, it would seem, had a voice
of its own -
or perhaps a god lived in the letters.
Such ideas are found throughout history. The
search for mystical meaning in alphabetic script
can be seen in the use of runic charms, or in the
systems, still in use, which relate letters to numbers,
such as gematria (p. 61).
At another level, the mystique of language is
something which we encounter throughout
modern society, especially in the field of advertising
(pp.386— 9). Conquerors, too, well
power that exists in words. Napoleon,
preferred newspapers to battalions. And what bet-
ter way is there to remove a nation's influence than
to burn its writings? Cortez did this to the Aztecs
in 1520; and the Nazis and Allies did it to each
other in World War II.
VERBAL TABOOS
The word taboo has been borrowed from Tongan.
where it means 'holv' or 'untouchable'. Taboos
8 •
I POPULAR IDEAS ABOUT LANGUAGE
exist in all known cultures, referring to certain acts,
objects, or relationships which society wishes to
avoid — and thus to the language used to talk about
them. Verbal taboos are generally related to sex,
the supernatural, excretion, and death, but quite
often they extend to other aspects of domestic and
social life. For example, certain animals may be
considered taboo: the Zuni of New Mexico prohi-
bit the use of the word takka ('frogs') during cere-
monies; until recently, many southern Americans
avoided the word bull in polite speech, replacing
it by a euphemism, such as he-cow or male beast;
in Lappish and Yakuts, the original name for bear
is replaced by such phrases as our lord or good
father; and wolves, weasels, rats, lice, snakes, and
many other animals have been given name-taboos
by various cultures. Even people can be affected:
certain members of the family are considered taboo
among Australian aborigines; either a special
language has to be used to them, or they are not A Jewish boy wearing
phylacteries (Hebrew
directly addressed at all (§10).
tefillin) These are a pair of
The use of a taboo word can lead to a variety small leather boxes contain-
of sayings, practices and responses. The mention ing scriptural passages, tra-
of a devil or unclean spirit can evoke a verbal or ditionallyworn by male Jews
over 1 3 years of age, as a re-
physical reaction, such as a divine invocation, or
minder of God's Law. They
the sign of the cross. An obscenity can be the cause are worn on the left arm fac-
of shocked recrimination ('go and wash your ing the heart, and on the
mouth out'), physical violence (especially if 'ladies' forehead during morning
are present), or legal action (as in the trial over weekday prayers. The bands
of the phylacteries are knot-
the publication of the unexpurgated D. H. Law-
ted so as to form the Hebrew
rence novel, Lady Chatterley's Lover (p. 61)). The letters daleth, yod and shin,
influence of taboo words can even extend across which form the divine name
language boundaries. It has been noted that Creek Shaddai
Indians avoid their native words for 'earth' and
'meat' {fdkki and apiswa respectively) because of
their phonetic resemblance to English taboo words,
which is the dominant language around them. A
similar phenomenon has been recorded with Thai
learners of English, where English yet closely
resembles Thai jed (an impolite word for 'to have
intercourse'). And Chinese people called Li (a com-
mon family name) can find their name a source
of embarrassment in Rangoon, in view of the Bur-
mese word li ('phallus').
The usual way of coping with taboo words and
know the notions is to develop euphemisms and circumlocu-
it is said, tions. Hundreds of words and phrases have
emerged to express basic biological functions, and
talk about death has its own linguistic world, with
its morticians, caskets, and innumerable ways of
dying. English examples include to pass on, pass
over, make one's bow, kick the bucket, snuff the
candle, go aloft, and cut the painter. French has
fermer son parapluie ('to close one's umbrella'),
the indescribably final n 'avoir plus mal aux dents
('to have no more toothache'), and many more.
PROPER NAMES
The use of words as personal labels
of particular significance - a fact which is early
learned by children, who are often anxious to con-
ceal their own names, and who so easily hurt, and
Death can lead to major taboo effects on the
is a matter use of names. Often, the names of the dead are
not to be uttered - though this may well be out
of fear rather than respect: while a name endures,
it is believed, the dead person does also, and those

are hurt, by name-calling. Many primitive people who utter the name bring the evil of death upon
do not like to hear their name used, especially in themselves. In some cultures (such as the Polyne-
unfavourable circumstances, for they believe that sian), therefore,when a person dies, other people The name of God
the whole of their being resides in it, and they may of the same name have to be renamed, or, if the The true name God, or of
of
thereby fall under the influence of others. The name happens to correspond to a word in the lan- individual gods, is a closely
danger is even greater in tribes (in Australia and guage, that word would have to be changed. By guarded secret in many cul-
New Zealand, for example), where people are contrast, some cultures (such as the Greenlandic) tures, if indeed it is known at
all. The real names of Allah
given two names - a 'public' name, for general place great store by the name of the dead person,
and Confucius are secret, as
use, and a 'secret' name, which is known only to who isthought to be unable to rest in peace, unless were the names of many
God, or to the closest members of their group. To a child has been named after him. In yet others, Egyptian deities.
get to know a secret name is to have total power if a child dies, the next by the same mother will Observant Jews do not
pronounce the divine name
over its owner. be called by some evil name, to show the death
as occurred in the Hebrew
it

The Todas of southern India dislike uttering their spirit that the child is not worth bothering about.
of the Old Testament. It was
own names, if they are asked
to the extent that, Sophisticated societies have had their supersti- written with four consonants,
for their ask someone else to give
name, they will tions too. In the Roman levies, the authorities took YHWH (the tetragramma-
it. The Sakalavas of Madagascar do not communi- good care to enrol those men who had auspi-
first ton), vowel points not being
written in pre-Massoretic He-
cate their own name, or the name of their village, cious names, such as Victor and Felix. The names
brew (p. 202). In reading
to strangers, in case mischievous use should be of Greek gods were carved on stone and sunk in aloud, the forms 'Adonai or
made of it. In folklore, there are many examples the sea, to guard against profanation. In Plato's 'Elohim are substituted. The
of forbidden names which, when discovered, break Cratylus, the debaters worry about using the names form Yahweh is a scholarly
attempt at reconstruction, in-
the evil power of their owners - Tom-tit-tot, Varga- of gods as etymological examples (p. 404), and in
terpreting its meaning as part
luska, Rumpelstiltskin. the Christian era there are long-standing prohibi-
of the verb to be', to give the
The process of personal naming can even affect tions over taking the name of the Lord 'in vain' the One who Is'. The
title

thewhole of a language. Stories are common of (p. 61). Older Hebrew names usually had mean- name Jehovah has been
tribal chiefs who change their name when they take ings, such as Nathaniah ('Yahweh has given') or traced back only to the 1 4th
century: it is reached by in-
office, as a resultof which any everyday words Azzan ('Strong'). When Adrian VI became pope,
serting the vowels of Adonai
which resemble that name have to be replaced, so he was advised not to retain his own name on under the tetragrammaton,
that the name will not be used in inauspicious cir- the grounds that all popes who had done so had and arose from a misreading
cumstances. It is when
reported, for example, that died in the first year of their reign. People in the by Christian scholars of the
two sources as one word. It
Queen Rasoherina of the Anemerina tribe in Mada- 20th century may find it easy to dismiss such atti-
is thus not of Scriptural ori-
gascar came to the throne, the word sopherina ('silk tudes, but things have not greatly changed. It is
gin, and the true pronuncia-
worm') was forbidden, and replaced by zana dandy unlikely that popular opinion would ever allow a tion of YHWH is now quite
('silk's child'). new ship to be named Titanic. lost.

Out with the old, in with the new

The mystique words canof (1980), with its capital city
affect place names too, as Salisbury renamed Harare
a country searches to re- (1982); Dahomey has be-
place forms which have un- come Benin (1975), French
happy associations. In Sudan has become Mali
1868, Edo was renamed (1960), and Gold Coast
Tokyo ('eastern resi- has become Ghana (1 957).
dence'), symbolizing a new
period in Japanese history.
St Petersburg became
Petrograd and then Len-
ingrad; Christiania became
Oslo. It is common practice
for new nations to change
their names, or the names
of their major cities, to sym-
bolize their independence
and freedom from imperia-
list influence. Thus in re-

cent times in Africa, for ex-

ample, we have seen
Upper Volta change its The old and new Japan:
name to Burkina Faso two contrasting towers in
was
(1985); Rhodesia re- Kyoto. In the foreground is
named as Zimbabwe the Toji temple.

3 THE MAGIC OF LANGUAGE •

9
4 The functions of language

The question 'Why do we use language?' seems many literary devices of grammar and vocabulary
hardly to require an answer. But, as is often the which convey the writer's feelings (§12). However,
way with linguistic questions, our everyday fam- in these more complex cases it becomes difficult
iliarity with speech and writing can make it difficult to distinguish the emotional function of language
to appreciate the complexity of the skills we have from the 'ideational' function described above.
learned. This is particularly so when we try to
define the range of functions to which language
Social interaction
can be put.
'To communicate our ideas' is the usual answer Mrs P sneezes violently. Mrs Q says 'Bless you!'
to the question — and, indeed, this must surely be Mrs P says 'Thank you.' Again, this hardly seems
the most widely recognized function of language. to be a case of language being used to communicate
Whenever we tell people about ourselves or our ideas, but rather tomaintain a comfortable relation-
circumstances, or ask for information about other ship between people. Its sole function is to provide
selves and circumstances, we are using language a means of avoiding a situation which both parties
in order to exchange facts and opinions. The use might otherwise find embarrassing. No factual con-
of language is often called 'referential', 'prop- tent is involved. Similarly, the use of such phrases
ositionaP, or 'ideational'. It is the kind of language as Good morning or Pleased to meet you, and ritual
which will be found throughout this encyclopedia exchanges about health or the weather, do not
— and in any spoken or written interaction where 'communicate ideas' in the usual sense.
people wish to learn from each other. But it would Sentences of this kind are usually automatically
be wrong to think of it as the only way in which produced, and stereotyped in structure. They often
we use language. Language scholars have identified state the obvious (e.g. Lovely day) or have no con-
several other functions where the communication tent at all (e.g. Hello). They certainly require a
of ideas is a marginal or irrelevant consideration. special kind of explanation, and this is found in
the idea that languageis here being used for the

purpose of maintaining rapport between people.

The anthropologist Bronistaw Malinowski
Emotional expression (1884—1942) coined the phrase 'phatic commu-
Mr X carefully leans his walking stick against a nion' to refer to this social function of language,
wall, but it falls over. He tries again, and it falls which arises out of the basic human need to signal
a second time. Mr X roundly curses the walking friendship — or, at least, lack of enmity. For
stick. How should we classify this function of lan- someone to withhold these sentences when they
guage? cannot be 'communication of ideas', for
It are expected, by staying silent, is a sure sign of
there no-one else in the room.
is distance, alienation, even danger.
Here we have one of the commonest uses of lan- These illustrations apply to English and to many
guage — a means of getting rid of our nervous European languages. But cultures vary greatly in
energy when we are under stress. It is the clearest the topics which they permit as phatic communion.
case of what is often called an 'emotive' or 'expres- The weather is not as universal a conversation-filler
sive' function of language. Emotive language can as the English might
like to think! For example,
be used whether or not we are alone. Swear words Rundi women Burundi, Central Africa), upon
(in
and obscenities are probably the commonest sig- taking leave, are quite often heard to say, routinely
nals to be used in this way, especially when we and politely, 'I must go home now, or my husband
are in an angry or frustrated state (p. 61). But there
are also many emotive utterances of a positive kind,
such as our involuntary verbal reactions to beauti-
Sneezing in Tonga
ful art or scenery, our expression of fear and affec-
When someone sneezes, is kuma (be well'); and in sneeze should be thinking
the English stock response Malagasy, velona
it is about nothing, instead of
tion, and the emotional outpourings of certain
is Bless you. But there is no ('alive'). In Tonga, a about the one who has
kinds of poetry. equivalent to such forms in sneeze is often taken to be sneezed. A major differ-
The most common linguistic expressions of emo- many languages, and any a sign that your loved one ence with English is that
tion consist of conventional words or phrases (such remarks which might be is missing you. It is quite the person who has
made can have a totally common for someone to sneezed may utter the
as Gosh, My, Darn it, and What a sight) and the different meaning and func- say jokingly, after a phrase - a kind of Bless
semi-linguistic noises often called interjections tion. In German, one says sneeze, Ikai ke nofo noa me!
(such as Tut-tut, Ugh, Wow, Ow, and Ouch). Also, Gesundheit ('health'); in mua!- literally, Not to be
an important function of the prosody of language Mende (Sierra Leone), the nothing, alas." The sense
word to use is biseh (thank intended is that the loved
(§29) is to provide an outlet for our attitudes while
you'); in Bembe (Congo), it one who has caused' the
we speak. At a more sophisticated level, there are

10 1 POPULAR IDEAS ABOUT LANGUAGE

will beat me.' Moreover, phatic communion itself are the lyrics of popular songs and the range of
is far from universal: some cultures say little, and phonetic effects which can be encountered in
prefer silence, as in the case of the Paliyans of south- poetry. Unintelligible words and phrases are com-
ern India, or the Aritama of Colombia. monplace in the oral poetry of many languages,
and can be explained only by a universal desire
to exploit the sonic potential of language.
The power of sound
In 1952, children skipping in a school playground Speaking in tongues
were heard to chant: 'Shirley Oneple, Shirley Two- paka bante rine sokuntare like ordinary language: the (or its written equivalent,
pie, Shirley Threeple .'
and so on up to 'Shirley
. .
mare paka tore moti sha- sounds are simpler and glossographia) must be
Tenple'. The instance was recorded by Iona and lara tamere pakashara more repetitive; there are classed along with other
Peter Opie in The Lore and Language of School- merime . . few predictable structural cases of functional pseudo-
units; and there is no sys- linguistic behaviour, such
children (1959), and it clearly illustrates the 'pho-
Thisis part of an utterance tematic word- or sentence- as jazz scat' singing.
netic' character of children's rhymes and games. which occurred sponta- meaning. When asked, Glossolalia needs to be
It is largely nonsense, and yet it performs an impor- neously at a religious ser- glossolalists are usually distinguished from cases
tant function: the repetitive rhythms help to control vice. It displays the repeti- unable to repeat utterances of xenoglossia, where
tive,reduced range of exactly, or give a detailed people miraculously speak
the game, and the children plainly take great delight
syllabic and rhythmic pat- account of their meaning. a language they have not
in it.
terns typical of tongue- Glossolalic speech is in- previously learned or
There are many situations where the only appar- speaking, or glossolalia-a terpreted in a general way. heard. Claims for such
ent reason for a use of language is the effect the widespread phenomenon To speak in tongues is tak- cases are rare, difficult to
sounds have on the users or listeners. We can group within the Pentecostal tra- en as a sign of the sincerity prove (e.g. to rule out the
dition of Protestantism and ofa person's belief, or as possibility that the speaker
together here such different cases as the rhythmical
charismatic Roman Catho- evidence of conversion. heard the language as a
litanies of religious groups, the persuasive cadences licism. The speakers treat it as a child), and usually turn out
of political speechmaking, the dialogue chants used Though many glossola- highly significant, emo- to involve chance effects -

by prisoners or slaves as they work, the various believe they are

lists tional event,which reflects as when a few syllables
speaking a real but un- their new-found sense of happen to resemble a se-
kinds of language games played by children and
known language, the utter- the presence of God. In this quence in some language.
adults (p. 59), and the voices of individuals singing ance patterns are quite un- respect, the phenomenon
in the kitchen or the bath. Perhaps the clearest cases

Graphic power Icelandic names

The names of dwarfs in the
La mandoline 13th-century Icelandic Edda
Vaillet et le bambou are like a painting in sound.
A few of the names resemble
words in the language, but
most have no meaning. (8
and b are the th sound in this
and thin respectively.)

Nyi ok Nidi,
NorOri, Su5ri,
Austri, Vestri,
Alpjofr, Dvalinn,
Nar ok Nainn,
Nipingr, Dainn,
Bifurr, Bofurr,
Bgmburr, N6ri,
Ori, Onarr,
Oinn, MjoSvitnir,
Viggr ok Gandalfr,
Vindaifr, porinn,
Fili, Kili,

Fundinn, Vali,
br6r, prbinn,
pekkr, Litr ok Vitr,
Nyr, Nyradr,
Rekkr, RadsiviOr.

Writing and print can exercise a purely visual effect

upon the reader, over and above the linguistic content
of the words (§32). This is best illustrated in poetry
where the shape of the poem reflects its subject matter
- asin this poem from Guillaume Apollinaire's Calli-

grammes (1918), showing a mandolin, a bamboo stick,

'I like coffee, I like tea, I like radio, and TV a typical and a flower.
ball-bouncing monologue

4 THE FUNCTIONS OF LANGUAGE •

1
The control of reality Recording the facts
In the northern borderland of Nigeria, an Igbo man A solicitorpreparing a case for a client. He pulls
is

invokes the spirit powers in his ancestral prayers,
using a formulaic curse: Kwo, unu, kwosi okiro!
('Wash, all of you, wash down upon all of our
enemies!'). In an English church, a priest holds a
baby over a font, and pours water on its head,
saying I baptize you
down an old book of judgments from his shelf,
and reads a report of a case which took place 25
years ago. What use of language is this? At first
sight, it would appear to be 'ideational'; but the
situation in which the communication takes place
is quite different in several respects.
When information is stored for future use, it is
impossible to predict who is likely to use it - indeed,
much of the material may never be referred to
again. There is therefore no 'dialogue' element in
the communication. The information has to be as
self-contained as possible, for it is impossible to
predict the demands which may one day be made
upon it, and in most cases there is no way in which
the user can respond so as to influence the writer.
Accordingly, when language is used for the pur-
poses of recording facts, it is very different from
that used in everyday conversation - in particular,
it displays a much greater degree of organization,

impersonality, and explicitness.

This function of language is represented by all
kinds of record-keeping, such as historical records,
geographical surveys, business accounts, scientific
reports, parliamentary acts, and public data banks.
It is an essential domain of language use, for the

availability of this material guarantees the know-

ledge-base of subsequent generations, which is a
prerequisite of social development.

Devil dancer' performing a healing ritual in Sri Lanka. The Domesday Book

All forms of supernatural belief involve the use

of language as a means of
controlling the forces
which the believers feel affect their lives. The var-
ious prayers and formulae which are directed at
God, gods, devils, spirits, objects, and other physi-
cal forces are always highly distinctive forms of
language (p. 384). In some cases, the language
might be regarded as a form of ideational commu-
nication, with a supernatural being as the recipient
— but if so, it is a somewhat abnormal type of com-
munication, for the response is usually appreciated
only in the mind or behaviour of the speaker, and
there may be no evident response at all.

In other cases, the function of the language is

to control matter, or the reality which the matter

issupposed to represent. For example, the garden-
ing ritual of the Trobriand Islanders involves a
formulae which 'charm' the axes, making
series of
them At a Roman Catholic Mass,
effective tools.
The two volumes which language
about Essex, Norfolk, and tion of this kind of
the speaking of the words This is my body is comprise Domesday Book Suffolk; the larger volume is well symbolized by the
believed to identify the moment when the commu- This was the summarizing contains the abbreviated popular label for the books,
nion bread is changed into the body of Christ. record of William I's survey account of all other coun- Domesday', which came to
of England, which was car- ties surveyed (the whole of be widely used by the 1 2th
Several other situations, apart from the magical
ried out in 1086. The England except some of century. From this record
and the religious, illustrate this 'performative' func- smaller volume contains all the most northerly areas). there would be no appeal!
tion of language — such as the words which name the information returned The once-and-for-all' func-
a ship at a launching ceremony.

12 i POPULAR IDEAS ABOUT LANGUAGE

 These signals enter into the whole of our linguis-
The instrument of thought ticbehaviour, so much so that it is often a problem
A man sits alone at a workbench, staring at a piece distinguishing the identifying function of language
of equipment with a puzzled frown. He says: 'So from that used for the communication of ideas.
if I put red four there, and link it to blue three, In a public meeting, for instance, Mr A may make
that'll leave blue six free. Then
can use that for
I a speech in support of Mr B, and it may be difficult
green four. Right.' He sets to work. to decide whether the reason for his speech is to
People often feel the need to speak their thoughts make a fresh point, or simply to demonstrate to
aloud. If asked why they do it, they reply that it all concerned that A
on B's side. The arena of
is

helps their concentration. Authors often make political debate is of such manoeuvrings, as
full

similar remarks about the need to get a first draft individuals strive to express their solidarity with
down on paper, in order to see whether what they (or distance from) each other.
have written corresponds to what they had in mind.
The French thinker, Joseph Joubert (1754-1824), There are two kinds of way pronounce them, with
I

once said: 'We only know just what we meant to mental calculating prodi- the sound of my own voice,
say after we have said it.' gies: those who 'hear' and this interior audition
numbers and those who stays with me a good part
Perhaps the most common use of language as
'see' them. Both rely on of the day.' In observing
an instrument of thought is found when people some kind of 'inner' lan- him perform on stage, he
perform mathematical calculations 'in their head'. guage, especially when was usually seen to move
Very often, this supposedly 'mental' act is accom- faced with a complex prob- and he
his lips or mutter,

panied by a verbal commentary. However, it is not lem. Inaudi was one of the oftenaccompanied this by
great 'auditory' calculators. exaggerated gestures and
essential that language used in this way should
Though he did not learn to pacing.
always be spoken aloud or written down. Often, read or write he wasuntil An interesting parallel is

people can be seen to move their lips while they 20, by the age of 7 he was sometimes drawn between
are thinking, but no actual sound emerges. Lan- able to multiply two 5-digit prodigious calculating abili-

numbers in his head. ties and language. Are

guage is evidently present, but in a 'sub-vocal'
When he was studied by these mental feats very far
form. the psychologist Alfred removed from our impres-
Several theories have been proposed concerning Binetin 1894, Inaudi's sive everyday generative
the role of language as the instrument of thought Jacques Inaudi (1867- auditory techniques clearly ability (§16) to manipulate

- notably that of the Russian psychologist, Lev

1950) emerged - in his own the complex structure of a
words, 'I hear numbers . vast range of novel sen-
Semenovich Vygotsky (1896—1934), who argued
.

resound in my ear, in the tences?

for a concept of 'inner speech', a mental use of
words to evoke a sequence of thoughts. Does all
thought, then, require language? This complex
question will be reviewed in §5. Graphic identity
The characteristic typefaces
news-
of several British
The expression of identity papers provide an illustration
of identity using the graphic
The crowds attending President Reagan's pre- medium (p. 185). These ex-
election meetings in 1984 repeatedly shouted in amples are taken from the
all

Guardian's spoof edition of 1

unison 'Four more years!' What kind of language
April 1978, in which news
is this? from the fictitious island of
Such language is hardly informative to those who San Serriffe was presented
use it, but it plainly has an important role in foster- in a series of typical formats

- and language styles lam-

ing a sense of identity in this case, among those S»« SERRIFFE S BIGGEST Dlllt SUE
pooning actual British news-
who share the same political views. Many social
papers of the time. The joke
situations display language which unites rather relies totally on the reader
than informs — the chanting of a crowd at a football being able to identify these
match, the shouting of names or slogans at public formats immediately, using a
mixture of typographic and
THE SS TIMES
meetings, the stage-managed audience reactions to
linguistic cues.
television game shows, or the shouts of affirmation
at some religious meetings.
Our use of language can tell our listener or reader
a great deal about ourselves — in particular, about
our regional origins, social background, level of
education, occupation, age, sex, and personality.
The way language is used to express these variables
is so complex that it requires separate discussion

(§§6—12), but the general point can be made here,

that a major function of language is the expression
of personal identity — the signalling of who we are
SS GUARDIAN
Printed in Metro and sometimes Bodoni Saturday April 1 1978

and where we 'belong'.

4 THE FUNCTIONS OF LANGUAGE •

13

5 Language and thought
seems evident that there is the closest of relation-
It
ships between language and thought: even-day
experience suggests that much of our thinking is
by language (p. 13). But is there identity
facilitated
between the two? Is it possible to think without
language? Or does our language dictate the ways
in which we are able to think? Such matters have
exercised generations of philosophers, psycholo-
gists, and linguists, who have uncovered layers of
complexity in apparently straightfonvard
these
questions. A simple answer is certainly not poss-
ible; but at least we can be clear about the main
factors which give rise to the complications.
KINDS OF THINKING
Many kinds of behaviour have been referred to
as 'thinking', but not all of them require us to posit
a relationship with language. Most obviously, there
is no suggestion that language is involved in our
emotional response to some object or event, such
as when we react to a beautiful painting or an
unpleasant incident: we may use language to
explain our reaction to others, but the emotion
itself is 'beyond words'. Nor do people engaged
in the creative arts find it essential to think using
language: composers, for example, often report
that they 'hear' the music they wish to write. Also,
our everyday fantasies, day-dreams, and other free
associations can all proceed without language.
The thinking which seems to involve language
is of a different kind: this is the reasoned thinking
which takes place as we work out problems, tell
stories, plan strategies, and so on. It has been called
'rational', 'directed', 'logical', or 'propositional'
thinking. It involves elements that are both deduc-
tive (when we solve problems by using a given set
of rules, as in an arithmetical task] and inductive
(when we solve problems on the basis of data
placed before us, as in working out a travel route).
Language seems to be very important for this kind
of thinking. The formal properties of language,
such as word order and sentence sequencing,
constitute the medium in which our connected
thoughts can be presented and organized.
INDEPENDENCE OR IDENTITY?
But how close is this relationship between language
and thought? It is usual to see this question in terms
of two extremes. First, there is the hypothesis that
language and thought are totally separate entities,
with one being dependent on the other. At the
opposite extreme, there is the hypothesis that lan-
guage and thought are identical - that it is not
possible to engage in any rational thinking without
using language. The truth seems to lie somewhere
between these two positions.
14 •
I POPULAR IDEAS ABOUT LANGUAGE
.
Within the first position, there are plainly two
Non-verbal and
possibilities: language might be dependent upon
verbal thought
thought, or thought might be dependent upon lan-
The two dimensions to
guage. The traditional view, which is widely held - linguistic
rational thinking
at a popular level, adopts the first of these: people and non-linguistic - can be
have thoughts, and then they put these thoughts discovered in a simple ex-
into words. It is summarized in such metaphorical periment, which anyone can
perform.
views of language as the 'dress' or 'tool' of thought.
The view is well represented in the field of child 1 Think of where you work.
language acquisition (§38), where children are seen Now visualize the route you
follow,as if you were driving
to develop a range of cognitive abilities which pre- along in a car, as you pro-
cede the learning of language. ceed from work to your
The second possibility- has also been widely held: home. The sequence of vis-
ual images which you bring
the way people use language dictates the lines along
to mind will be largely inde-
which they can think. An expressive summary of
pendent of language.
this is Shelley's 'He gave men speech, and speech 2. Now imagine you have to
created thought, /Which is the measure of the explain to a visitor how to
universe' (Prometheus Unbound). This view is also reach your house from work.
Think out the steps of your
represented in the language acquisition field, in the
explanation, as you would
argument that the child's earliest encounters with present them, without saying
language are the main influence on the way con- anything aloud. The se-
cepts are learned. The most influential expression quence of ideas will be ex-
of this position, however, is found in the Sapir— pressed internally using lan-
guage.
Whorf hypothesis (see facing page).
A third possibility, which is also widely held
these days, is that language and thought are inter-
dependent - but this is not to say that they are
identical. The identity view (for example, that
thought is no more than an internalized vocaliza-
tion) is no longer common. There are too many
exceptions for such a strong position to be main-
tained: we need think only of the various kinds
of mental operations which we can perform with-
out language, such as recalling a sequence of move-
ments in a game or sport, or visualizing the route
from home to work. It is also widely recognized
that pictorial images and physical models are help-
ful in problem-solving, and may at times be more
efficient than purely verbal representations of a
problem.
On the other hand, these cases are far outnum-
bered by those where language does seem to be
the main means whereby successful thinking can
proceed. To see language and thought as inter-
dependent, then, is to recognize that language is
a regular part of the process of thinking, at the
same time recognizing that we have to think in
order to understand language. It is not a question
of one notion taking precedence over the other,
but of both notions being essential, if we are to
explain behaviour. Once again, people have
searched for metaphors to express their views. Lan-
guage has been likened to the arch of a tunnel;
thought, to the tunnel itself. But the complex struc-
ture and function of language defies such simple
analogies.
.
The Sapir-Whorf hypothesis Having a word for it

The romantic idealism of the late 18th century, There is nothing in every- of vehicles- car, lorry, bus, autoist autonaut
as encountered in the views of Johann Herder day English to correspond moped, truck,
tractor, taxi, roadist vehiclist
to the many Arabic words and so on - and might have chassimover murderist
(1744_1803) and Wilhelm von Humboldt
for horse or camel, the Es- just one word for all of mobilist roadent
(1762-1835), placed great value on the diversity kimo words for snow, or the these. wheelist vehicuwary
of the world's languages and cultures. The tradition Australian languages' There is in fact no single doice (Driver Of Internal
was taken up by the American linguist and anthro- words for hole or sand. word in English for the Combustion Engine)
pologist Edward Sapir (1884-1939) and his pupil Speakers of English have driver of all kinds of motor pupamotor (Person Using
to resort to circumlocutions vehicles - motorist being Power-Assisted Means
Benjamin Lee Whorf (1897-1941), and resulted ifthey want to draw the dis- restricted to private cars, of Travel On Roads)
in a view about the relation between language and tinctions which these lan- and driver being unaccep- licentiat (Licensed Internal
thought which was widely influential in the middle guages convey by separ- table for motorcycles -a Combustion Engine
decades of this century. ate words - such as the lexical gap which greatly Navigator Trained in
size, breed, function, and worried the British Automo- Automobile Tactics)
The 'Sapir-Whorf hypothesis', as it came to be
condition of a camel. On bile in 1961 It
Association .

called, combines two principles. The first is known the other hand, several lan- was felt such a word
that
However, none of these
ingenious ideas has sur-
as linguistic determinism: it states that language guages cannot match the would be useful, and they
vived.
determines the way we think. The second follows many words English has therefore asked for sug-
available to identify differ- gestions. Among the 500
from this, and is known as linguistic relativity: it
ent sizes, types, and uses they received were:
encoded in one language
states that the distinctions
are not found in any other language. In a much-
quoted paragraph, Whorf propounds the view as
follows:
Examples such as these made the Sapir-Whorf Words for hole in
We dissect nature along lines laid down by our native hypothesis very plausible; but in its strongest form Pintupi
languages. The categories and types that we isolate from it is unlikely to have any adherents now. The fact
takes between three and
It
the world of phenomena we do not find there because that successful translations between languages can 14 English words to dis-
they stare every observer in the face; on the contrary,
be made is a major argument against it, as is the tinguish the various senses
the world is presented in a kaleidoscopic flux of of hole in this Australian
fact that the conceptual uniqueness of a language
impressions which has to be organized by our minds aboriginal language, but the
- and this means largely by the linguistic systems in our such as Hopi can nonetheless be explained using
can nonetheless
distinctions
English. That there are some conceptual differences
minds. We cut nature up, organize it into concepts, and be conveyed.
ascribe significances as we do, largely because we are between cultures due to language is undeniable,
yarla a hole in an object
parties to an agreement to organize it in this way - an but this is not to say that the differences are so
pirti a hole in the ground
agreement that holds throughout our speech community great that mutual comprehension is impossible.

and is codified in the patterns of our language. The One language may take many words to say what pirnki a hole formed by a
agreement is, of course, an implicit and unstated one, rock shelf
another language says in a single word, but in the
but its terms are absolutely obligatory; we cannot talk end the circumlocution can make the point. kartalpa a small hole in the
at all except by subscribing to the organization and ground
Similarly, it does not follow that, because a lan-
classification of data which the agreement decrees. yulpilpa a shallow hole
guage lacks a word, its speakers therefore cannot in

which ants live

Whorf view by taking examples
illustrated his grasp the concept. Several languages have few
from several languages, and in particular from words for numerals: Australian aboriginal lan- mutara a special hole in a
spear
Hopi, an Amerindian language. In Hopi, there is guages, for example, are often restricted to a few
one word (masa'ytaka) for everything that flies general words (such as 'all', 'many', 'few'), 'one' nyarrkalpa a burrow for small
animals
except birds - which would include insects, aero- and 'two'. In such cases, it is sometimes said that
planes and pilots. This seems alien to someone used the people lack the concept of number - that abori- pulpa a rabbit burrow
to thinking in English, but, Whorf argues, it is no gines 'haven't the intelligence to count', as it was makarnpa a goanna burrow
stranger than English-speakers having one word once put. But not so, as is shown when these
this is katarta the hole left by a

for many kinds of snow, in contrast to Eskimo, speakers learn English as a second language: their goanna when it has broken
where there are different words for falling snow, ability to count and calculate is quite comparable the surface after hibernation

snow on the ground, snow packed hard like ice, to that of English native speakers.
slushy snow (cf. English slush), and so on. In Aztec, However, a weaker version of the Sapir-Whorf
a single word (with different endings) covers an hypothesis is generally accepted. Language may not
even greater range of English notions - snow, cold determine the way we think, but it does influence
and ice. When more abstract notions are consi- the way we perceive and remember, and it affects
dered (such as time, duration, velocity), the differ- the ease with which we perform mental tasks.
ences become yet more complex: Hopi, for Several experiments have shown that people recall
instance, lacks a concept of time seen as a dimen- things more easily if the things correspond to
sion; there are no forms corresponding to English readily available words or phrases. And people cer-
tenses, but there are a series of forms which make tainly find it easier to make a conceptual distinction
it possible to talk about various durations, from neatly corresponds to words available in their
if it

the speaker's point of view. It would be very diffi- language. Some salvation for the Sapir-Whorf
cult, Whorf argues, for a Hopi and an English phy- hypothesis can therefore be found in these studies,
sicist to understand each other's thinking, given which are carried out within the developing field
the major differences between the languages. of psycholinguistics (p. 412).

5 LANGUAGE AND THOUGHT •

15
 PART II
Language and identity

'Who are you? How old are you? Where are you that all have an influence on the way
in which lan-
from? What do you do? What are you doing now? guage is and that language, in turn, exercises
used,
. .
.'
We would only have to speak, to provide our a dominant influence on our perception of social
interrogator with innumerable clues about our per- structure, whatever our mother tongue.
sonal history and social identity. The linguistic sig- An even wider range of linguistic variation is

nalswe unwittingly transmit about ourselves every
moment of our waking day are highly distinctive
and discriminating. More than anything else, lan-
4 guage shows we 'belong', providing the most
natural badge, or symbol, of public and private
identity. The reports and discussion in this part
of the encyclopedia plainly demonstrate this fact
and illustrate how our perception of our own and
others' language can become, in varying degrees,
a source of pleasure, pride, anxiety, offence, anger,
and even violence.
The various sections of Part
lationship between language and the many
of our identity aswe interact with others. We begin
with the relatively permanent features of language
that express aspects of a person's physical or
psychological identity - factors such as age, sex,
body type, personality, and intelligence. Next, we
look at the linguistic facts and issues surrounding
the notion of geographical background, and the
way this is manifested in regional accent and dia-
lect. This leads, in particular, to a consideration
of the world of dialectology, with its atlases and
questionnaires, which has attracted widespread
interest.
The following sections review the complex set
/ of factors that enter into the definition of ethnic
subsumed under the heading of contextual identity.
Here we examine how the immediate situation in
which people communicate can influence the kind
of language they use. Three main features of con-
text are distinguished - the setting, the participants,
and the type of activity in which they engage. This
leads us to consider such divergent topics as greet-
ings, news-readings, speech making, everyday con-
versation, proverbial expressions, and slang. In
addition, there are separate sections devoted to
visual varieties, restricted languages, hidden and
n explore this re- secret languages, word games, humour, and the
'faces' many forms of verbal art.
These last topics lead naturally to the final sub-
ject of Part II: personal linguistic identity, with its
reliance We begin by iden-
on the concept of 'style'.
tifying different kinds of approach to stylistic study,
and look in detail at one of them, stylostatistics,
where we encounter linguistic detective work in
areas as far apart as literary authorship and the
investigation of murder. The concept of stylistic
distinctiveness then leads us to examine the re-
lationship between literary and non-literary uses
of language, with particular reference to the tra-
ditional study of rhetoric and to each of the major
literary genres - poetry, drama, and the novel. Part
ii then concludes with a summary of recent trends

and social identity: racism and nationalism, strati- in literary theory that have focused on the role of
fication into classes and castes, status and role, soli- language in the interpretation of texts: the ante-
darity and distance, social stereotypes - it emerges cedents and consequences of structuralism.

The linguistic reflection of cultural identity, here seen in Mon-

treal in 1980 during the campaign leading to thereferendum
for an independent Quebec.
 6 Physical identity

Several factors define a person's physical identity, from 73-123 mm); Melanesians had a mean of Bodies, minds, and
the most obvious being age, sex, physical type 84 mm (variation from 70-110 mm); and the Japa- voices
and so
(height, build, facial features, type of hair, nese had a mean of 73 mm (variation from
The German psychiatrist,
on),and physical condition. These factors, supple- 55—90 mm). The relative shortness of the Japanese Ernst Kretschmer (1 888-
mented by the criteria made available through tongue is noteworthy, and people have speculated 1 964) proposed a threefold
modern genetic techniques, are also taken into ac- whether this could be a factor that would contri- classification ofbody types,
count when identifying the broad, biologically claiming that these corre-
bute to spoken language learning difficulties. But
lated with certain mental
defined groups of human beings known as 'races'. it is not possible to reach a firm conclusion — es-
conditions. The pyknic type
Such considerations naturally lead to several ques- pecially as only very small samples of speakers have (thick trunk, short limbs)
tions. Are there any correlations between language so far been used. were thought to be more
and the physical characteristics of an individual It is difficult to be sure what effect this kind of prone manic-depressive
to
psychosis; the leptosomic
or race? Can any of the differences between lan- genetically determined difference could have on a
type (thin trunk, long legs)
guages, or the variations within a language, be language. It might have no effect at all. People more prone to schizo-
explained by referring directly to the physical con- might compensate for the 'lack' of one anatomical phrenia; and the athletic type
stitution of the users? feature by making greater use of some other fea- (broad shoulders, thin hips)
prone to neither.
ture. Unfortunately, the information is not avail-
While these distinctions
able, as detailed comparative studies of racial vocal
proved in due course to be
anatomy are lacking. Certainly, everyday experi- too simplified for practical
Physical type ence suggests that the effects are minimal. One indi- application, certain general

There seems to be little clear relationship between cation of this is the language-learning ability of correlations between body
type and voice have been
speech and such physical characteristics as height, second-generation immigrants, whose accents may
observed. In one study,
weight, head size, and shape. That there is some be indistinguishable on tape from those of the indi- which matched voices with
correlation is evident from our surprise when we genous population. A commonly reported exper- photographs of these types,
hear a large, fat person come out with a thin, high- ience is for a London bus passenger to hear behind the pyknic type were

him matched most accurately,

pitched voice. There is a general expectation that a perfectly articulated Cockney Any more fares
followed by the leptosomic,
size relates to loudness and pitch depth. However, please?, only to find a conductor who is plainly with the athletic type least
there is no conclusive way of predicting from physi- West Indian or African. well predicted. Other studies
cal appearance alone whether a person's vocal Despite the superficial differences, it has gener- have also shown significant
range is going to be soprano, contralto, tenor, or ally been concluded that vocal tracts the world over correlationsbetween voices
and photographs of the
bass. are sufficiently similar that we can regard them as
speakers, doubtless be-
There is little in the anatomy of the human vocal variants of a single, universal type. Work in phone- cause of the general physical
tract to account for the linguistic differences tics (§27) proceeds on this assumption. But we still relationship between a per-
between people and groups. The proportions of know very little about the general potential for son's size and physique, and
the dimensions of the larynx
the various vocal organs (§22) seem to be very simi- sound production in human beings - a subject that
and vocal tract. There are,
lar in all human beings. Individual variations do the Polish linguist Jan Baudouin de Courtenay however, wide margins of
exist in size and shape: for example, the height (1845-1929) christened anthropophonics. error in these studies.
of the palate varies a great deal, as does the length
and flexibility of the tongue. Some people can make 1st octave
the tip of their tongue touch their uvula; others Counter Large Small above 2nd 3rd 4th
octave octave octave Middle C octave octave octave

H
can hardly make their tongue touch their hard
palate. More men than women can make the edges C, E, G, B, D F Ac e g b d
1

f
2
g
2
2
b* d> P
J 4
e
4
g
4

4
b
4

D, F, A, C E G B d I a c' c' g' b

1
d f' a c
!
c' g b' d f* a
of their tongue curl upwards. But, pathological
cases aside (§46), these differences do not seem to Bass
add up to much, as far as spoken languagecon- is Baritone
cerned. There is no evidence to suggest that ana- r H Tenor
tomical variations have any effect on the ability Contralto
of a person to learn or use speech. Mezzo-soprano
We have to reach a similar conclusion when we Soprano
consider the kinds of anatomical variation that dis- Voice types Probably the around two octaves, with shows the average ranges
tinguish the world's racial groups. Certainly several most common way of talk- good singers achieving for each type of voice (after
differences could be relevant for speech - for exam- ing about human voices is three octaves or more. Ex- M. Nadoleczny, 1923).
the classification into six ceptional cases have been Speaking level tends to be
ple, the considerable variation in the length of the
basic types: soprano, recorded, with coloratura towards the bottom of the
tongue. In one study (F. Brosnahan, 1961), the mezzo-soprano, contralto, sopranos reaching as high singing range, as shown by
tongues of Japanese, Melanesians, and blacks were tenor, baritone, and bass. as e 4 (2637 Hz), and cross bars.
measured: blacks had the longest tongues on aver- Generally, the individual basses as low as contra'
1

age (a mean of 97 mm, with individual variation range of each type is F (44 Hz). The diagram

18 • II LANGUAGE AND IDENTITY

Physical condition The voice 'breaks'
That there must be some kind of relationship Cutting across the distinc- the upper limit by only a tional change of the former
tions between age, sex, few tones. usually takesbetween 3
between physical condition and language is plain and 6 months, whereas the
and physical type is the As well as the pitch
from the way language can be affected in cases of phenomenon of voice change, certain other vocal latter may take much
physical handicap. Several disorders of constitu- mutation, which accom- features usually mark the longer. For this reason, it is

tional origin have a direct effect on a person's abi- panies the development of onset of puberty. The voice generally felt to be wise to
secondary sex charac- is often husky and weak, delay adult singing instruc-
lity to use language, variously affecting the ability
teristics during puberty. At with poorly controlled vocal tion until well after the
to comprehend and produce speech, read, and this time, the child voice fold vibration (§22). Subse- change in speaking voice
write. Temporary handicaps may have minor but differentiates into male and quently, males, the voice
in has taken place, to avoid
quite noticeable effects — such as the change in voice female types, due mainly to depth is the most notice- the risk of vocal strain
the rapid growth of the able feature; in females, (p. 276).
quality that accompanies a cold or a sore throat,
larynx. the voice becomes louder, There is no predictable
or the alterations in pronunciation that may follow
The development is far and it changes in timbre - rule relating adult singing
a visit to the dentist. At a much more serious level, more noticeable in boys: the thin childlike voice be- registers to child voices.
there are such cases as the child with cleft palate, male vocal folds become comes fuller and more Whether a boy is soprano
or the adult with myaesthenia gravis, where speech about 1 cm longer, where- vibrant. or alto, he will develop a
as with girls the increase is The term 'break' is not bass or baritone singing
can be fundamentally and dramatically affected.
only around 3—4 mm. As a always an accurate way of voice in about two-thirds of
Here, it is often possible to make deductions about consequence, in boys, the describing the changes cases - a phenomenon
the nature of the person's handicap solely from entire vocal range is both that take place. The that accounts for the com-
a tape recording. Voice quality, individual sounds, broadened and lowered by change from infantile to mon complaint among
grammar, vocabulary, and other features of lan- about one octave. In girls, adult voice is often a choirmasters about the
there is no such octave gradual transition, rather shortage of tenors, and the
guage can all be affected. It is thus a complex field
shift', and the increase in than a sudden shift, especi- fact that operatic tenors re-
of study, which needs a separate review to do it voice range is much less ally in females. Moreover, ceive the higher salaries!
justice (Part vin). marked: the lower limit of the speaking voice and the Similarly, sopranos are far
their range extends by only singing voice may be differ- more common than other
one-third of an octave, and ently affected. The muta- female voices.

and changes in the facial skeleton, especially

Age around the mouth and jaw.
What can be said of the normal process of aging, There are other, more general signs of age.
from a linguistic point of view? In general terms, Speech rate slows, and fluency may be more erratic.
there is a clear and unmistakeable relationship: no- Hearing deteriorates, especially after the early fif-
one would have much difficulty identifying a baby, ties. Weakening faculties of memory and attention
a young child, a teenager, a middle-aged person, may affect the ability to comprehend complex
or a very old person from a tape recording. With speech patterns. But it is not all bad news: vocabu-
children,it is possible for specialists in language lary awareness may continue to grow, as may stylis-
development, and people experienced in child care, tic ability - skills in narration, for example. And
to make very detailed predictions about how lan- grammatical ability seems to be little affected.
guage correlates with age in the early years - a
research field treated separately in Part vn.
Little is known about the patterns of linguistic
change that affect older people.
It is plain that our

voice quality, vocabulary, and style alter as we

grow older, but research into the nature of these
changes is in its earliest stages. However, a certain
amount of information is available about the pro-
duction and comprehension of spoken language by
very old people, especially regarding the phonetic
changes that take place.
Speech is likely to be affected by reductions in
the efficiency of the vocal organs (§22). The mus-
cles of the chest weaken, the lungs become less elas- 30 50
tic, the ribs less mobile: as a result, respiratory Age (years)

75 is only about half that at age

efficiency at age
30, and this has consequences for the ability to
speak loudly, rhythmically, and with good tone.
The cartilages, joints, muscles, and tissues of the
larynx also deteriorate, especially in men; and this
affects the range and quality of voice produced by
the vocal folds, which is often rougher, breathier
and characterized by tremor. In addition, speech
is affected by poorer movement of the soft palate
The aging voice The fundamental frequency of the
voice (§23) changes quite markedly with age. The
graph shows the change that takes place in males: the
level drops sharply at adolescence, continues to
decrease until middle age, and then increases into
senescence. The data points are a composite of
averages taken from various published studies. For
females, the level is stable during middle age,
decreasing later. (After R. D. Kent & R. Burkard, 1981 .)

6 PHYSICAL IDENTITY •
19
VOICEPRINTS
The traditional method of identifying a person is
through fingerprint patterns, which seem to be
unique to each person. In recent years, several
attempts have been made to provide an analogous
technique using the voice. One approach, which
received widespread publicity in the 1960s, was
developed by an American acoustic scientist, Law-
rence Kersta (1907 -).
'Voiceprints' are made from an acoustic analysis
of speech by a sound spectrograph (p. 136). It is
assumed that no two people will have identical
vocal tracts, and therefore the patterns of sound
vibration they produce when they speak will be
different. Kersta claimed it was possible to tell
people apart by analysing the visual patterns shown
on the spectrograms of ten common words (cf. the
ten fingerprints). The patterns were displayed both
as bar voiceprints and as contours.
The approach attracted considerable interest
among law enforcement agencies, who saw its
potential value in crime detection, and voiceprints
were soon used as evidence in US courts. In 1965,
for example, a youth boasted on a television pro-
gramme of having set fire to several shops in Los
Angeles. His face was concealed, and the television
company exercised its legal right not to say who
he was. Using other clues in the broadcast, detec-
tives were able to trace a youth, who was brought
to trial. Voiceprint evidence established that the
voice of the youth in court and that of the youth
in the programme were the same. Despite an attack
by defence lawyers on the voiceprint evidence, the
youth was found guilty.
Critical reactions
After several cases, criticism of the technique began
to grow. In 1976, a special committee of the Acous-
tical Society of America expressed its concern that
voiceprints were being admitted as legal evidence
when there had been insufficient scientific evalua-
tion of the technique.
The main thrust of the criticism was directed
20 • I LANGUAGE AND IDENTITY
at the fingerprint analogy. Fingerprint patterns are A contour representation
established in the foetus; they change in size, as of bar voiceprint (a)
(below).
people grow, but not in form. Voices, however,
are partly the result of learning, and they vary:
speakers can utter sounds in different ways on dif-
ferent occasions. Also, speakers might produce
voiceprints that would not be distinguishable (at
the level of detail shown on a spectrogram).
Professional impersonators were invoked. In Bri-
programme showed a spectro-
tain, a television
gram of John Bird impersonating the Prime
Minister, Harold Wilson, and compared it with
one of Wilson's own voice. The similarities were
thought to be much greater than those seen on
spectrograms both of Bird's impersonation com-
pared with his normal voice, and of Wilson's voice
on that occasion compared with other occasions.
On the other hand, in the US, people who could
not hear the difference between President Kennedy
and Elliot Reid's impersonation were able to see
from the voiceprints that the two voices were not
the same.
The technique remains controversial. On the
positive side, it is accepted that some features of
the voice are indicative of speakers rather than of
languages (e.g. the higher formants, §23). Also,
more refined techniques in spectrography and more
sophisticated methods of pattern recognition using
computers could circumvent some of the first criti-
cisms. Interest in the possibilities remains high,
because the potential value of the approach is very
great - not only in forensic science, but in such
fields as commerce (identifying people over the
phone), medicine (distinguishing abnormal body-
noises), and engineering (identifying abnormal
moving parts in machinery). On the negative side,
the error rate among analysts is still unacceptably
Voiceprints. The bar voice-
prints of four male speakers
high (as much as 1 in 20, according to some critics), uttering you, taken from Ker-
and many people doubt whether the properties of sta's paper in Nature, 1 962.
the vocal tract are in principle capable of making One speaker has uttered the
the discriminations required by the theory. As a word a second time. Can you
tell which two voiceprints are
result, speech scientists have been extremely cau-
from the same person?
tious about making claims for voiceprinting proce- ('See foot of facing page for
dures. answer.)
ic>
•*-
\
Sex and the honorific (p. 99) varieties of speech, but Some markers of
females use them much more commonly and in Japanese female
Phonetic differences (p. 19) are the most obvious a wider range of situations. For example, a man style
measures of sexual identity; but languages provide might use a certain honorific form only in talking
• Use of atashi (
instead
many instances of males and females learning dif- to a superior, whereas a woman might use it for
'I'),

of watakushi.
ferent styles of speech — as in Japanese, Thai, Carib, a social equal as well. The polite forms of nouns, • Sentence particle wa
Chukchi, and Yana. Pronunciation, grammar, verbs, and adjectives are also used more frequently used at the end of sentences
vocabulary, and context of use can all be affected. by women. with rising intonation, instead
of with falling intonation.
In Koasati, certain verb forms differ according to There have been fewer studies of male speech • Interjections of surprise,
the sex of the speaker. If the form ends in a vowel style, reflecting a tradition that sees female speech such as ara, ma, uwa.
or /tf/, there is no difference; but in cases where as the 'special' variety; and a separate label for • Less frequent use of such
the woman's form ends in a nasalized vowel, or the male style not often used. But the style can
is
interjections as a or e.

in certain consonants, the man's form substitutes • Use of sentence particle

be clearly defined, and is heard in contexts where
yo following a noun, instead
/s/. This can be seen in the following examples traditional notions of masculinity are to be found of da (male) or desu yo (sex-
(after M. R.Haas, 1944): (assertiveness, toughness, etc.). By no means all neutral).
male language is distinctive, however: as in the case • Use of no ('matter') at the

Female Male of women, sex-neutral speech will often be used, end of statements, instead of
n da (male) or n desu (sex-
isko isko he drank and on occasion there may be the use of feminine neutral), e.g. Dekinaino (It's
lakdwc lakdwc you are lifting it features, as signals of gentleness or consideration. [a matter of being] impos-
kq kd's he saying
is sible') vs Dekinai n da.
molhil molhis we are peeling it ENGLISH MALE AND FEMALE SPEECH • Use of polite forms of
nouns, such as osakana
tacwan tacilwas don't sing! In English, the situation is less clear. There are no
('fish') for sakana (sex-
grammatical forms, lexical items, or patterns of neutral).
When Haas carried out her study, only middle- pronunciation that are used exclusively by one sex, • More frequent use of
aged and elderly women used the female forms; but there are several differences in frequency. For particle ne ('right?', 'okay') at
younger women were beginning to use the forms example, among the words and phrases that the end of sentences.
• Less frequent use of the
typical of male speech. But members of each sex women are supposed to use more often are such assertive particles ze and
were quite familiar with both speech styles, and emotive adjectives as super and lovely, exclama- zo at the end of statements.
could use either upon occasion. If a man were tell- tions such as Goodness me and Oh dear, and
ing a story involving a female character, he would intensifies such as so or such (e.g. It was so busy).
use women's forms when quoting her speech. This use of intensifies has been noted in several
languages, including German, French, and Rus-
JAPANESE MALE AND FEMALE sian.
SPEECH More important are the strategies adopted by
A clear case of linguistic sex-differentiation is Japa- the two sexes in cross-sex conversation. Women
nese, where well-defined styles of speech have been have been found to ask more questions, make more
known since the early 11th century. Females have use of positive and encouraging 'noises' (such as
used a style known as joseigo or onnakotoba, mhm), use a wider intonational range and a more
which evolved among upper-class women as a sign marked rhythmical stress, and make greater use
of their special position in society. Books on femi- of the pronouns you and we. By contrast, men are
nine etiquette fostered the use of special vocabulary much more likely to interrupt (more than three
and grammar, alongside norms of gentle and times as much, in some studies), to dispute what
submissive behaviour. This traditional view is has been said, to ignore or respond poorly to what
undergoing considerable change today; but clearly has been said, to introduce more new topics into
defined sexual roles still predominate, and distinct the conversation, and to make more declarations
linguistic forms are widely encountered. of fact or opinion.
Japanese female speech is a style over which Most interpretations of these differences refer to
women have conscious control. It is used when the contrasting social roles of the sexes in modern
women wish to emphasize their femininity; on society. Men are seen to reflect in their conversa-
other occasions, they adopt a sexually neutral style. tional dominance the power they have traditionally
Thus a woman mayuse feminine style in talking received from society; women, likewise, exercise
to her friends about her children but use neutral the supporting role that they have been taught to
style when talking to business colleagues. It is also adopt — in this case, helping the conversation along
possible for women to use the masculine speech and providing men with opportunities to express
style if they wish to express themselves in an asser- this dominance. The situation is undoubtedly more
way - and this is often done these days by
tive complex than this, as neither sex is linguistically
many who are concerned to promote notions of homogeneous, and considerable variation exists
sexual equality. A particular example is the increas- when real contexts of use are studied. The danger,
ing use of boku (T) among schoolgirls - tradition- as some commentators have pointed out, is that
allyused only by males. in the process of criticizing old sexual stereotypes,
There are also frequency differences in the use researchers are in danger of creating new ones (p.
of forms. Both males and females use the formal 46). (a) and (e)

6 PHYSICAL IDENTITY 21
7 Psychological identity

It is common practice to identify individuals, or LANGUAGE AND INTELLIGENCE

groups of people, in terms of their psychological Decades of controversy over the nature and assess-
attributes - whether they have high intelligence, ment of intelligence preclude any straightforward
good concentration, an aggressive personality, a statement about its relationship to language. It is
Verbal vs non-
poor memory, and so on. We generally make these evident that people are judged as more or less intel-
judgments on the basis of the non-linguistic way ligent, based on how they behave in certain situ- verbal IQ
in which people behave when they carry out tasks ations, and in response to certain tasks. There is Some intelligence tests do
not contain any tasks that re-
and interact in specific situations. For example, we a long tradition of intelligence testing, in which
quire a knowledge of lan-
do not need to refer to language in order to see sets of tasks are presented in order to ascertain
guage in order to solve them.
whether someone can pay attention, remember levels of achievement, and to demonstrate indi- A person is asked to carry

which route to take, fix a piece of equipment, or vidual differences; the scores that result are widely out such activities as building
behave in a friendly manner. But very often we used in educational, clinical, and other contexts. an object, matching shapes,
Most research has been carried out in relation finding a way through a
do rely on language in order to evaluate such mat-
maze, detecting picture simi-
ters, and this forms an important part of the study to the development of children's intellectual pro- larities and differences, or
of identity. cesses, as they learn about the world, react to situ- deciding which entities go
Any of the fields of academic psychology can ations, solve problems, and carry out all kinds of together'. These 'non-verbal'

prompt a linguistic enquiry of this kind. We might tasks. Several theoretical positions exist, which tests contrast with 'verbal'
tests, which rely on a prior
investigate whether there is a relationship between are reviewed in Part vn. Studies with mentally awareness of language com-
language structures or skills and such notions as handicapped children have shown that a certain prehension or production -
memory, attention, perception, personality, intelli- minimum level of intelligence, as measured on for example, tests of general
gence, learning, or any other recognised psycho- conventional tests, is a prerequisite for language knowledge, memory for di-
gits, arithmetic, vocabulary
logical domain. These studies have both theoretical development. However, this need not be very high,
comprehension, and similari-
and practical implications. They suggest ways of and there is no clear relationship between intelli- ties between words. Several
constructing models of our mental processes - a gence and the ability to use particular language kinds of material have been
major preoccupation of the field of psycho- structures. Attempts have been made to relate intel- devised to help promote non-

linguistics (p. 412). And they relate to several issues ligence to quantity of infant babbling, amount of verbal skills. The picture be-
low shows preschool chil-
of language learning - both normal (in such con- vocabulary, grammatical complexity, the prosodic
dren using equipment which
texts as mother-tongue education and foreign lan- features of speech, the use of figurative expressions, helps to train movement, ma-
guage learning) and pathological (in such contexts and other variables. In no case is there a neat corre- nipulation, and perception.
as speech and hearing disorders). The main findings lation, though stereotypes of performance The children thread wooden
blocks along an increasingly
are thus more appropriately reviewed in other sec- undoubtedly exist, and here the psycholinguistic
difficult series of wires, from
tions (§§25, 34, 38, 45). Furthermore, any linguis- study of intelligence overlaps with that of person- a very basic loop to a multi-
tic medium (speech, writing, signing) can be the ality. dimensional maze.
focus of enquiry, though only spoken language
characteristics are considered here (for hand-
writing and signing, see §§32, 35).
This cluster of cross-references shows how the
topic of psycholinguistic identity extends well
beyond the subject matter of the present section.
It is also somewhat arbitrary dealing with it next

to the section on physical identity instead of later,

as part of the section that deals with the distinctive
features of 'style' (§12). However, this decision
should not be construed as taking sides in the con-
troversies that have raged over the role of 'nature'
and 'nurture' in the formation of such attributes
as personality and intelligence. From a linguistic
point of view, it is simply to recognize the fact
that, once adulthood is achieved, any features of
language that can be related to psychological attri-
butes seem to be relatively permanent, and thus
have more in common with the long-term charac-
teristics of physical constitution, than with the tem-
porary and consciously controllable features that
form the basis of stylistic study.

22 • II LANGUAGE AND IDENTITY

PERSONALITY Vocal stereotypes
Psychopathology
This complex field deals with the characteristics Listeners are very ready to make stereotyped judg- Voice characteristics are
that enable us to distinguish between people, and to ments about personality: comments such as You also an important diagnostic
make predictions about their behaviour - charac- can tell he's anxious fromhis voice or She sounds feature of abnormal person-

teristics generally classified as personality traits and very strong minded are often to be heard. System- alities. Patients suffering
from schizophrenia, depres-
types. Traits are styles of behaviour that an indivi- atic information has been obtained in social psy-
sion, and other such condi-
dual displays, whatever the stimulus, in many chology experiments since the 1930s, when tions often speak with voices
different circumstances. Types involve the identifi- researchers began to use the new broadcasting that are monotonous, weak,
cation of a salient feature that is then used as a medium to get large-scale listener judgments of hesitant, slow,and deviant in
one study (T. H. Pear, 1931), timbre.Abnormal intonation,
label for the whole personality. In one study (G. different voices. In
loudness, rhythm, and timbre
W. Allport &H. S. Odbert, 1936), nearly 18,000 4,000 listener judgments were obtained about nine can also be heard in the
trait labels were found to be available in English speakers played over the air. Age and sex proved voices of many autistic child-
to distinguish one person's behaviour from another easiest to identify, and among the vocations repre- ren. Some clinicians have
(honest, tidy, shy, thoughtful, stupid but of sented, actors and clergymen were most frequently maintained that psychopath-
...),
ological syndromes can be
course many of these overlap in meaning, and most recognized. But even when the listeners made the
detected on the basis of the
studies classify traits into much smaller sets of basic wrong decisions about vocation, they were extre- voice alone; but this seems
dimensions, such as dominance, extraversion, or mely consistent in their errors. People who sound exaggerated. Even exper-
likeability. like clergymen, it appears, will be rated as if they ienced clinicians find it diffi-
Several interesting inferences have been drawn are clergymen, whether they are or not. cult to make predictions
about some types of mental
about the relationship between personality traits Stereotypes of this kind markedly colour inter-
state solely from audio re-
or types and aspects of speech, especially in personal and intergroup relationships. They are cordings.
'matched-guise' experiments. The first of these widespread, with similar results being found in
studies (W. E. Lambert et al., 1960) aimed to show several other areas where accent, dialect, or lan-
how English- and French-speaking Canadians guage conflicts exist, such as between speakers of
viewed each other. English-speaking college stu- white and black American English, Canadian and
dents in Montreal were asked to listen to recordings European French, Hebrew and Arabic in Israel, and
of a passage being read aloud in English and in urban and rural accents of British English. They
French, and to mark on a checklist what the person- are also seen in social and occupational contexts
ality traits of the speakers were. They were told (e.g. affecting the way in which a jury judges the

to disregard language, and concentrate solely on
voice and personality. However, the students were
not told that the voices were in fact those of per-
fectly bilingual speakers, each of whom read the
passage both in an English and a French 'guise'.
The results were illuminating. The English guises
of the speakers were evaluated much more favour-
ably than were the French guises: for example, they
were thought to be better looking, more intelligent,
kinder, and more ambitious. But in a second part
to the study, there was an even more interesting
finding. When French-speaking Canadians were
given the same test, they too rated the English
guises as higher, in almost all respects, indicating
the low esteem in which the French language was
held at that time.
There is of course no correlation between such
attributes as intelligence or attractiveness and the
credibility of witnesses in court, p. 387), and in
education (where teachers' evaluations of a pupil's
capabilities can be more influenced by speech style
than by written composition, artistic work, or per-
sonal appearance). Our impressions of a person's
guilt, innocence, intelligence, or stupidity are, it
seems, much affected by phonetic and linguistic fac-
tors.A more informed popular awareness of the
dangers of vocal stereotyping is thus an important
aim of this branch of sociopsychological research.
Personality traits and voice stereotypes
This graphic presentation
of personality traits was de-
vised by the British psycho-
Hans Eysenck
logist
(1916- ). The inner ring

speaking of English or French. But it is a fact that shows the four ancient
Greek temperaments,
people do form such stereotyped impressions on
based on the predomi-
the basis of linguistic features (especially prosody, nance of one of the four
§29). Moreover, all accents, dialects, and lan- 'body fluids'. The outer ring
guages are affected by evaluations of this kind. If represents the location of
speakers use a standard accent, speak quickly and different traits, grouped on
a statistical basis, and re-
fluently, and use few hesitations, they are likely
lated to two principal di-
to be rated as more competent, dominant, and mensions: instability/
dynamic. The use of regional, ethnic, or lower-class stability and extraversion/

varieties, on the other hand, is associated with introversion. But would it al-

greater speaker integrity and attractiveness. Even ways follow that someone
if

sounds reliable (sober, etc.)

national personalities can be perceived: British
then he/she is reliable
speakers rate French as a more romantic language, (sober, etc.)?
it seems, and German as a more businesslike one.

7 PSYCHOLOGICAL IDENTITY 23
8 Geographical identity

The most widely recognized features of linguistic a dialect — whether urban or rural, standard or
identity are those that point to the geographical non-standard, upper class or lower class. And no
origins of the speakers - features of regional dia- dialect is thought of as 'superior' to any other, in
lect, which prompt us to ask the question 'Where terms of linguistic structure — though several are
are they from?' But there are several levels of res- considered prestigious from a social point of view.
ponse to this question. We might have a single per-
son in mind, yet all of the following answers would
be correct: 'America', 'The United States', 'East
Coast', 'New York', 'Brooklyn'. People belong to
regional communities of varying extent, and the
Where are you from?
dialect they speak changes its name as we 'place'
How easy is it to tell where Shaw's Henry Higgins: '/ merge, with a consequent
someone from? A few
is can place any man within blurring of speech patterns.
them in relation to these communities. years ago, it would have six miles. can place him
I And nowadays, through
Languages, as well as dialects, can convey geo- been relatively straight- within two miles in London. radio and television, there
graphical information about their speakers, but forward for a specialist to Sometimes within two is much more exposure to
this information varies greatly, depending on the work out from a sample of streets' (Pygmalion, Act 1 ). a wide range of dialects,
speech the features that These days, dialect iden- which can influence the
language of which we are thinking. The variation
identified someone's tification has become much speech of listeners or
can be seen if we complete a test sentence using regional background. more difficult, mainly be- viewers even within their
different language names: 'If they speak they — , Some dialect experts have cause of increased social own homes. A radio dialect
must be from .'—If the first blank is filled by been known to run radio mobility. In many countries, show would be much less
'Swedish', the second blank will almost certainly shows in which they were it is becoming less com- impressive today. On the
able to identify the general mon for people to live their other hand, meticulous
be filled by 'Sweden'. But 'Portuguese' would not regional background of analysis can bring results,
whole lives in one place,
inevitably lead to 'Portugal': the second blank members of their audience and mixed' dialects are and there have been
could be filled by 'Brazil', 'Angola', 'Mozambique', with considerable success. more the norm. Also, as several notable successes
and several other countries. 'French' would give But it is doubtful whether towns and cities grow, in the field of forensic
anyone has ever de- once-distinct communities linguistics (p. 69).
us the choice of about 40 countries, and 'English'
veloped the abilities of
well over 50. 'Dialect', by contrast with 'language',
is a much more specific geographical term.
Dialect or accent?
POPULAR NOTIONS OF DIALECT Itis important to keep these cause a grammatical differ- §28).
It sometimes thought that only a few people
is terms apart, when discuss- ence is involved. Similarly, Usually, speakers of dif-
ing someone's linguistic the choice between wee ferent dialects have differ-
speak regional dialects. Many restrict the term to
origins. Accent refers only bairn and small child is dia- ent accents; but speakers
rural forms of speech - as when they say that 'dia- to distinctive pronunciation, lectal, because this is a of the same dialect may
lects are dying out these days'. They have noticed whereas dialect refers to contrast in vocabulary. But have different accents too.
that country dialects are not as widespread as they grammar and vocabulary the difference between The dialect known as stan-

once were, but they have failed to notice that urban as well. If we heard one bath with a short a' [a] and dard English' is used
person say He done it and bath with a long a' [a:] is throughout the world, but it
dialects are now on the increase (p. 32). Another
another say He did it, we a matter of accent, as this is spoken in a vast range of
view is to see dialects as sub-standard varieties of would refer to them as is solely a matter of pro- regional accents.
a language, spoken only by low-status groups — using different dialects, be- nunciation (or phonology,
implicit in such comments as 'He speaks correct
English, without a trace of dialect.' Comments of
this kind fail to recognize that standard English
Dialect, idiolect, and lect
is as much a dialect as any other variety - though
Probably no two people are vestigate a language, we a term for any variety of a
a dialect of a rather special kind (p. 39). Or again, identical in the way they have no alternative but to language which can be
languages in isolated parts of the world, which may use language or react to begin with the speech identified in a speech com-
not have been written down, are sometimes the usage of others. Minor habits of individual munity - whether this be on
differences in phonology, speakers: idiolects are the
referred to pejoratively as dialects, as when personal, regional, social,
grammar, and vocabulary first objects of study. Dia- occupational, or other
someone talks of a tribe speaking 'a primitive kind
are normal, so that every- lects can thus be seen as grounds. The term variety
of dialect'. But this fails to recognize the true com- one has, to a limited extent, an abstraction, deriving is itself often used for this
plexity and range of all the world's languages a personal dialect'. It is from an analysis of a purpose; but in recent
(§47). often useful to talk about number of idiolects; and years, many sociolinguists
the linguistic system as languages, in turn, are an (p. 41 2) have begun to use
In this encyclopedia, as is standard practice in
found in a single speaker, abstraction deriving from a lect as a general term in
linguistics, dialects are seen as applicable to all lan- and known as an
this is number of dialects. this way.
guages and all speakers. In this view, all languages idiolect. In fact, when we in- It is also useful to have
are analysed into a range of dialects, which reflect
the regional and social background of their
speakers. The view maintains that everyone speaks

24 LANGUAGE AND IDENTITY

 point in the chain can we say that one language A schematic dialect conti-
Language vs dialect
ends and the next begins? On what basis can we nuum between dialects A
and G. The possible degrees
One of the most difficult theoretical issues in draw boundary lines between Portuguese, Spanish, of mutual intelligibility are re-
linguistics is how to draw a satisfactory distinction French, and so on? We are used to thinking of presented by different shad-
between language and dialect. The importance of these languages as quite different from each other, ing,from maximum (red) to
this matter will be repeatedly referred to in Part but this is only because we are usually exposed zero (white).

ix, where we have to make judgments about the to their standard varieties, which are not mutually
number of languages in the world and how they intelligible. At the local level, it is not possible to
are best classified. make a clear decision on linguistic grounds.
At first sight, there may appear to be no problem. But decisions are of course made on other
If two people speak differently, then, it might be grounds. As one crosses a well-established national
thought, there are really only two possibilities. boundary, the variety of speech will change its
Either they are not able to understand each other, name: 'Dutch' will become 'German', 'Spanish'
in which case they can be said to speak different will become 'Portuguese', 'Swedish' will become
languages; or they do understand each other, in 'Norwegian'. It is important to appreciate that the
which case they must be speaking different dialects reasons are political and historical, not linguistic
of the same language. This criterion of mutual intel- (§47). Arguments over language names often
ligibility works much of the time; but, unfortu- reduce to arguments of a political nature, especially
nately, matters are not always so simple. when there is a dispute over national boundaries.
For example, in the South Slavic continuum, var-
MUTUAL INTELLIGIBILITY ieties spoken on the Yugoslavian side of the border
One common problem with this criterion is that between Yugoslavia and Bulgaria are called dia-
dialects belonging to the same language are not lects of Macedonian by the former country, but
always mutually intelligible in their spoken form. dialects of Bulgarian by the latter - as part of a
It can be very difficult for someone from the south claim to the territory. However, because there is
of England to understand some of the regional dia- a dialect chain in the area, linguistic criteria will
lects of Scotland or Northern Ireland, for instance; never be able to solve conflicts of this kind.
and the degree of intelligibility can be even worse
when people attempt to communicate with English
speakers from other countries. However, at least Dialect continua in Europe
all of these speakers have one thing in common:

they share a common written language. On this

count, the varieties they speak could justly be called
dialects of the same language.
A rather more serious problem arises in cases
where there is a geographical dialect continuum.
There is often a 'chain' of dialects spoken through-
out an area. At any point in the chain, speakers
of a dialect can understand the speakers of other
dialects who live in adjacent areas to them; but
they find it difficult to understand people who live
further along the chain; and they may find the peo-
I pie who live furthest away completely unintelli-
i gible. The speakers of the dialects at the two ends
of the chain will not understand each other; but
they are nonetheless linked by a chain of mutual
intelligibility.
This kind of situation is very common. An exten-
sive continuum links all the dialects of the lan-
guages known as German, Dutch, and Flemish.
Speakers in eastern Switzerland cannot understand
speakers in eastern Belgium; but they are linked
by a chain of mutually intelligible dialects through-
out the Netherlands, Germany, and Austria. Other
chains in Europe include the Scandinavian conti-
nuum, which links dialects of Norwegian, Swedish,
and Danish; the West Romance continuum, which
links rural dialects of Portuguese, Spanish, Catalan,
French, and Italian; and the North Slavic conti-
nuum, which links Slovak, Czech, Ukrainian,
Polish, and Russian.
The theoretical problem should be clear. At what

8 GEOGRAPHICAL IDENTITY -25

 naires to all the school districts in the German The earliest use of
Dialectology Empire. took him ten years to contact nearly
It
dialectology?
The systematic study of regional dialects is known 50,000 local teachers, who were asked to provide
40 sentences Then Gilead cut Ephraim
variously as dialectology, dialect geography, or equivalents for in the local dialect. off
from the fords of the Jordan,
linguistic geography; but these terms are not exact An enormous amount of data was received, and and whenever an Ephraimite
equivalents. In particular, the latter terms suggest this led to the publication in 1881 of the first fugitive said 'Let me cross',
a much wider regional scope for the subject. A dia- Deutschen Reichs.
linguistic atlas, Sprachatlas des the men of Gilead
asked him,
lect specialist who spends his life researching the A larger series of works, based on Wenker's files, Are you an Ephraimite?'. If
he answered 'No', they said,
local usage of a single Yorkshire village can hardly appeared between 1926 and 1956; but even today,
Then say "Shibboleth".' He
be called a 'linguistic geographer', though he is cer- much of the original material has not been pub- would say 'Sibboleth', since
tainly a By contrast, the 'geo-
'dialectologist'. lished. he could not pronounce the
grapher' designation would be quite appropriate The postal questionnaire method enables a large word correctly. Thereupon
amount of data to be accumulated in a relatively they seized and slaughtered
foranyone involved in plotting the distribution of
him by the fords of the Jor-
forms over a large area, such as Scotland, or the short time, but it has several limitations - chiefly
dan.
eastern United States. that dialect pronunciations cannot be accurately (Judges XII, 4-6)
There is another difference between these terms. recorded. The alternative, to send out trained field
Traditionally, dialectology has been the study of workers to observe and record the dialect forms, The Ephraimites were be-
trayed by their regional pro-
regional dialects, and for many people that is still was used in the linguistic survey of France,
first
nunciation. As a result of this
its main focus. But in recent years, dialectologists which began in 1896. The director, Jules Gillieron story, shibboleth, which then
have been paying more attention to social as well (1854-1926), appointed Edmond Edmont (1849- meant 'ear ofcom' or 'flow-
as geographical space, in order to explain the extent 1926) — a grocer with a very sharp ear for phonetic ing stream', has in modern

of language variation (§§9-10). Factors such as
age, sex, social class, and ethnic group are now
seen as critical, alongside factors of a purely re-
gional kind.
But whatever the approach, the contemporary
fascination with dialects seems no less than that
shown by previous generations. Radio pro-
grammes on dialect variations are popular in
several countries, and compilations of dialect data
continue to be produced in the form of grammars,
dictionaries, folk-lore collections, and guides to
usage. Local dialect societies thrive in many parts
of the world. Dialects continue to be seen as a major
source of information about contemporary popu-
lar culture and its historical background; and dia-
lect variation forms part of the study of change
(§54).
Probably the most important application of dia-
lectology these days education, where the de-
is in
velopment of dialect 'awareness' in children is
widely recognized as a way of getting them to see
the heterogeneity of contemporary society, and
their place within (§§44, 61). Teachers are often
it
faced with a conflict between the child's spontan-
eous use of dialect forms and the need to instil
a command of the standard language, especially
in writing. The conflict can be resolved only by
developing in children a sense of the relationships
between the two kinds of language, so that the
value of both can be better appreciated. There
needs to be an awareness of the history, structure,
and function of present-day dialects - and this is
what dialectology can provide.
THE HISTORY OF REGIONAL
DIALECTOLOGY
While there has been sporadic interest in regional
dialects for centuries, the first large-scale systematic
Germany and France, did not take place
studies, in
end of the 19th century. In 1876, Georg
until the
Wenker (1852-1911) began sending out question-
differences — to do the field work. For four years, use come to mean 'dis-
tinguishing mark' or
Edmont went around France on a bicycle, conduct- criterion'.
ing interviews with 700 informants using a
specially devised questionnaire of nearly 2,000
items. The Atlas linguistique de la France was sub-
sequently published in 13 volumes between 1902
and 1910. It stands as the most influential work
in the history of dialectology.
In the first half of this century, major projects
were initiated in many parts of Europe, such as
Romania, Italy, Holland, Spain, and Denmark, and
there have been several impressive publications. In
due course the large-scale dialect surveys of the
United States and England began (p. 30). A great
deal of dialect work has also been undertaken in
Japan and China, as well as in parts of Africa, Aus-
tralia, and South America. In some countries, even,
surveys leading to a 'second generation' of linguis-
tic atlases have begun. Direct interviewing and pos-
talquestionnaires continue to be used today, as
does the tradition of presenting the linguistic
material in the form of maps; and in recent years,
dialectology has benefited enormously from the
development of techniques using tape recorders.
The field is also now being influenced by the elec-
tronic revolution, with computers helping to
'crunch' the data provided by questionnaires, and
making more
large data-bases of regional variants
available, accessible, and analysable - and even
more visible, using computer graphic techniques.
However, nowadays there are fewer big regional
dialect projects, and some of those that have begun
may never be completed. This is mainly because
of the large costs involved in collecting, analysing,
and publishing dialect data; but it is also partly
because of the new direction dialect studies have
taken. Younger scholars are
these days more likely
to be attracted by the sociolinguistically inspired
approaches that developed in the 1970s, with their
focus on social factors, and on urban rather than
on rural dialects (p. 32).

26 •
II LANGUAGE AND IDENTITY
 Questionnaires
THE FARM COW-HOUSE In
there
a large dialect survey,
will be many infor-
THE FARMSTEAD mants and several investiga-
Q. What do you call the place where you keep tors. One way of ensuring
Show an
surrounding
aerial photograph of a farmstead and
fields . —
your cows? April 1953, the animals that that the results of all the in-

give you milk replaced your cows. terviews will be comparable,

while also saving a great
1 ... these? Fields.
Rr. BEEF-HOUSE (COW-)BYRE, COW- deal of time, is through the
2 ... this? Farmstead. use of questionnaires. On
HOLE/HOUSE/HULL/SHADE/SHED,
3 ...this? Farmyard. LATHE, MISTALL, SHIPPON the other hand, unless the
4 ... this? Stackyard. questions are particularly in-
the various buildings? genious, the responses will
. . .

1 Nb 1 baio" 2 baio" [barman 1 byre-man lack the spontaneity of infor-

If necessary, ask the relevant question below. (= cowman) 1.2.3] 3 ku:baio B mal speech. Results thus
4-5 baio" 6 baia" 7 baio", have to be interpreted with
5 ... where you keep pigs? Pigsty.
the place
^bao'z 1
8 baio" 9 bats caution.
April 1953, the animals that go {i. grunting) Opposite is an extract from
replaced pigs. Cu the questionnaire used in the
2 1 baiar 2 baiar 3 baia, ku:as
... the place where you keep hens? Hen-
6
4 bats, OQ bai.az 5 ku:bauaj
1
6 baia,
English Dialect Survey (p.
house. —
April 1953, the birds that lay eggs
k 3 u:as ["old name"] 30).The dots at the begin-
for you replaced hens. ning of each line stand for
7 ...the place where you keep pigeons? Dove- What do you call . .
.'; /' =
3Du ku:bat.3
K
°ku: Jujcj 3 2 beta" 3
cote. —April 1953, the birds that go (i.
1

batar 4-5 baia

,

6bai3°D bat,3z 1
The second
imitate.
illustrates the depth of
extract

cooing) replaced pigeons.

phonetic detail recorded by
8 ... the place where you keep your cows? We baiaj, Obaia
4 1
2-3 4 JOpm the field workers. Abbrevia-
—
Cow-house. April 1953, the animals that
1 bai.3
tions after each number
give you milk replaced your cows.
5 La 1-3 j"Qpm 4 Jipn, ^Opn 1 stand for the different nor-
9 ... the yard in which cattle are kept, thern counties of England.
5 Jipn 6 Jipm, JOpm ["older"], ^ipan 1

especially during the winter, for fattening,

III.11.3,
oa j\pmz 7 Jipn, 1

and for producing dung? Straw-yard.

^Jipanz 1
8-9 Jipn 10 Jipan
(Verify the kind of cattle and the purpose). 1 1 JippOn 12JipOn, C'jQpQn 1

10 ... the small enclosed piece of pasture near 1 3 Jipan, °Jipm

2
1 4 JipQn
the farmhouse, the place where you might
put a cow or a pony that's none too well?
Paddock.
1 What's the barn for and where is it?

PAUSY, adj. n.Lin.' [po/zi.] Slightly intoxicated.

From Strine to Scouse
Slightly the worse for drink; said of persons who combine an
amiable desire to impart information with an incapacity to call to The contrast between re- used by engineers for dis- parentheses):
mind all the necessary words. Drunk naw he was n't whit '
!

you'd call drunk, nobbud he was pausy like.' and standard

gional dialect covering Kew brutes and
PAUT, v. and sb. Sc. Nhb. Dur. Lakel. Yks. Lan. Chi English usage has been a for making other calcula- Ullo dur\ ('Greetings; I am
Der. Not. Lin. Wor. Suf. Also written pawt Sc. Lakef' source of humour the world tions. pleased to make your ac-
Cum.' 4 n.Yks.* e.Yks. 1 m.Yks. 1 w.Yks. ne.Lan. 1 Der.' over. In Let Stalk Strine Tiger: Imperative mood of quaintance.)
Not. * n.Lin. 1 sw.Lin. 1
1
pawte w.Yks. port w.Yks. ; ;
(1965), Afferbeck Lauder the verb to take. As in: Gisalite ('Could you oblige
Not.*; and in forms paat Cai. 1 Nhb.' Cum. 14 paoot
se.Wor. 1 ; pout Sc. Ham.) N.Cy. 1 s.Wor. ;
;

powt Sc
(said to be Professor of 'Tiger look at this, Reg . me with a match, please?')
(Jam.) Bnff. 1 n.Cy. Suf. r [p§t, poat, pat.] 1. v. To poke Strine Studies at the Uni- X: The twenty-fourth letter Ay-ay ('I say!')
or push with the hand or a stick; to stir up; to paw versity of Sinny) uses stan- of the Strine alphabet; also La ('I say, young man.)
handle, or finger things. Cf. pote. dard spellings to represent plural of egg; also a tool for Ere, tatty-head\ (I say.
Sc. To search with a rod or stick in water, or in a dark
confined place. To make a noise when searching or poking
or the popular impression of chopping wood. young woman!)
in
water (Jam). n.Cy. Grose (1790). Whb. 1 Divent paat on wi'd, an Australian accent, with
or ye'll spoil'd. Com. Children pawt when they make repeated' bizarre results: Some of the colloquial In the Appendix to this
attempts to get things with their hands (E.W.P.) Cum. 4 A dog ;
Egg Nishner. A mechanical pronunciations here are work, selected verses from
pawls at the door when it wants to get in, and children pawt when
they make repeated attempts to get hold of things with their hands, device for cooling and puri- found in many dialects. For The RubAiyat of Oma'r
n Yks. n.Yka. 2 Kneading with the fingers into a soft mass.
1
; fying the air of a room. example, Gissa (Please Khayyim are translated
n.Lin. Sutton Wds. (1881); n.Lin. 1 I wish we hed n't noi cats give me
Jezz: Articles of furniture. .') is a feature of
. . into Scouse by Stan Kelly:
really, thaay're alus paw tin' atone, when one's gettin' one's melt
aw.Lin. 1 Some lasses are always pawting things about they've no
Asin:'Setthetible, love, Strine, but it is also well
business with. a. Wor. To beat down apples, Porson Quaint Wds and get a coupler jezz.' known in Liverpool, as can Gerrup dere La! De
(1875) '5- Money: The day following be seen from the section knocker-up sleeps light;
Hence Pouting, vbl. sb. the practice of spearing
(1) Sunny. (Sunny, Money, on 'Forms of Address' in Dawn taps yer winder,
salmon ; also used attrib. (a) Pout-net, sb. a net fastened
; Chewsdy, Wensdy, Lem Yerself Scouse ends anudder night;
An extract from the English Dialect Dictionary Thursdy, Fridy, Sairdy.) (1966), by Frank Shaw, And Lo! de dog-eared mog-
Joseph Wright (1855-1930), published this dictionary Scone: A metereological Fritz Spiegl, and Stan Kelly gies from next-door
in sixvolumes between 1 898 and 1 905; contained term. As in: 'Scona rine.' (whose standard English Tear up de jigger fer an
it

1 00,000 entries. Wright was largely self-taught,

Sly Drool: An instrument translations are given in early fight.
and did
not learn to read until he was a teenager - a fact that
may have been an advantage to him in his later
studies, as his early awareness of dialect differences
would not have been influenced by the forms of the
standard written language.

8 GEOGRAPHICAL IDENTITY • 27
 Isoglosses The map illustrates
Lines on maps isoglosses marking the parts of
England and Wales that pro-
Once the speech of dialect informants has been col- nounce the fxj in such words as
lected, it is analysed, and the important features car- the rhotic areas. The main
are marked on a map of the area in which the infor- boundary line runs southwards
from the west of Birmingham to
mants live. When several points on the map have
the east of Oxford, skirts the west
been located, it is then possible to see whether there of London, and ends on the Kent
is way these features are used. The
a pattern in the coast. Some relic areas in the
usual way
of identifying dialect patterns is to draw north of England are also to be
lines around the places where the people use a seen. The information is based on
the relatively conservative speech
linguistic feature in thesame way. These boundary
of rural people, as collected by the
lines are known one
as isoglosses. For example, English Dialect Survey (p. 30).
famous isogloss runs across England, from the
Severn to the Wash: it distinguishes northern
speakers who pronounce a rounded w/u/in words
like cup from southern speakers who keep the
vowel open and unrounded, /a/. A series of lexical
isoglosses, identifying various words for snack, is
illustrated on p. 30.
When isoglosses were first introduced (in 1892),
it was expected that they would provide a clear

method for identifying dialect areas. Because peo-

ple from a particular part of a country 'speak in
the same way', it was assumed that the isoglosses
for many linguistic features would coincide, and
form a neat 'bundle', demarcating one dialect from
another. However, early dialectology studies soon
discovered that the reality was very different. Iso-
glosses criss-crossed maps in all directions, and
very few actually coincided. There seemed to be
no clear dialect boundaries at all — a finding which
made some scholars go so far as to argue that the
whole idea of a dialect was meaningless.
In due course, however, supplementary' notions
were developed to make sense of the data. It was rhotic areas

noted that, while isoglosses rarely coincided, they

did often run in the same general direction. Some
areas, called focal areas, were seen to be relatively The main kinds of isogloss
homogeneous, containing few isoglosses. Where Term Separates Examples
focal areas merged, there was a great deal of isolex lexical items nuncb vs nuncheon (p. 30)
linguistic variation, with many isoglosses present: isomorph morphological features dived vs dove
these became known as transition areas. Often, a isophone phonological features put/put/ vs pAt/
isoseme semantic features dinner (mid-day meal) vs (evening meal)
feature might be left isolated, as a result of linguistic
change affecting the areas around it: these 'islands'
of more conservative usage were called relic areas.
Dialectologists have mixed feelings about iso-
glosses. There is often too much variability in the
way a linguistic feature is used for the data to be
easily summarized in a single isogloss. Also, the
relative significance of different isoglosses remains
to be interpreted. Some isoglosses mark distinc-
tions that are considered to be more important than
others (such as the contrast between short and long
a in words like bath in British English, which has
long been the focus of special comment). Isoglosses
(a)
are an important visual guide, but they need to
be supplemented by other criteria if they are to Focal and transitional On
The expectation Isoglosses The reality Isoglosses criss- a
display, and not to obscure, the true complexity willform neat bundles, demar- cross an area, with no clear larger scale, the isoglosses
of regional variation. cating dialect A from dialect B. boundary between A and B, are seen to constitute a transit-
ional area between the focal
areas A and B.

28 II LANGUAGE AND IDENTITY

THE RHENISH FAN
One of the best examples of the way isoglosses
fail group themselves into bundles is in northern
to
Europe. A set of isoglosses runs east-west across
Germany and Holland, separating Low German,
in the north, from High German, in the south. They
reflect the different ways in which these dialects
have developed the voiceless plosive consonants of
Indo-European (p. 328). In Low German, the
sounds have remained plosives (/p, t, k/); but in
High German, these have generally become frica-
tives. For example, 'village' is [dorp] in the north,
[dorf] in the south; 'that' is [dat], as opposed to
[das]; 'make' and Tare [makan] and [ik] respec-
tively, rather than [maxan] and [icj.
The map shows the location of the isoglosses
Through most of Ger-
that distinguish these words.
many, they are close together, displaying only
minor variations; but where they meet the River
Rhine, the isoglosses move in quite different direc-
tions, in a pattern that resembles the folds in a
fan. It thus becomes impossible to make simple
generalizations about dialect differences in this
area.A speaker in a village near Cologne, for exam-
ple,would say [icj and [maxan], as in High Ger-
man, but say [dorp] and [dat], as in Low German.
What accounts for the Rhenish fan? It has been
suggested that several of the linguistic features
could be explained with reference to certain facts
of social history. For example, the area between
the [dorp/dorf] and [dat/das] isoglosses was co-
extensive with the old diocese of Trier; the area
immediately north was coextensive with the old
diocese of Cologne. The linguistic innovations
seem to have spread along the Rhine from southern
Germany to the cities, and then 'fanned out'
throughout the administrative areas these cities
controlled. Rural speakers were naturally
influenced most by the speech of their own capital
cities, and political and linguistic boundaries gra-
dually came to coincide. (After L. Bloomfield,
1933.)

The two halves of France shows six items that are used. A different legal
sys-
One of the main findings of used differently on either tem existed the early
until
the Atlas linguistique de la side of an isogloss (J. K. 1 9th century, using a writ-

France (p. 26) was the bun- Chambers & P. Trudgill, ten code inspired by Ro-
dle of isoglosses that runs 1980, p. 111). man traditions. And there is

across France from east to
west, dividing the country
into two major dialect
areas. The areas are tradi-
tionally known as langue
d'oil (inthe north) and
langue d'oc (in the south) -
names based on the words
for yes' current in these
areas during the 1 3th cen-
tury, when the division was
first recognized. The map
The distinction corres- a major difference in archi-
ponds to several important tectural style, the roofs be-
social and cultural differ- ing generally flat, and not
ences, some of which can steeply pitched (as they are
stillbe observed today. For to the north of the bundle).
example, to the south of Such clear correlations be-
the isogloss bundle tween language and cultur-
(roughly where the Proven- al identity illustrate the way
gal region begins), a bien- inwhich dialect studies
nial (as opposed to a trien- form an important part of
nial) method of crop the study of social history.
rotation is traditionally

8 GEOGRAPHICAL IDENTITY •
29
THE LINGUISTIC ATLAS OF
ENGLAND
Three of the maps from the English Dialect Survey,
carried out by Harold Orton (1898-1975) and
Eugene Dieth (1893-1956), are illustrated here.
The field survey was undertaken between 1950 and
1961 in 313 localities throughout England. The
localities were usually not more than 15 miles
apart, and generally consisted of villages with a
fairly stable population. The informants were
natives of the locality, mainly male agricultural
workers, with good mouths, teeth, and hearing,
and over 60 years of age.
The principal method was a questionnaire that
elicited information about phonological, lexical,
morphological, and syntactic features. Tape
recordings of informal conversation were also
made. Questionnaire responses were transcribed
using the International Phonetic Alphabet (p. 158).
Over 1,300 questions were used, on such themes
as farming, animals, housekeeping, weather, and
social activities; and over 404,000 items of infor-
mation were recorded.
Between 1962 and 1971 the basic material of
the survey was published in an introduction and
four separate volumes; in 1977 the Linguistic Atlas
of England was published, containing an interpre-
tation of a selection of the data. The maps below
provide an example of the Survey's basic material
for the item snack and two interpretive maps, based
on this material. The first map is a display of all
the responses obtained, which are listed in the top
right-hand corner. The other maps pick out various
trends in usage, and are a considerable simplifica-
tion. (After H. Orton, S. Sanderson 8c J. Widdow-
son, 1978.)

FORENOON -
DRINKING
FORENOONS
JOWER
LOWANCE
LUNCH
M MINNING-ON
X NAMMET(S)
X NAMMICK
. NINESES
X NUMMET
X NUMMICK
N NUNCH
NUNCHEON
PROGGER
PUTTING-ON
SANDWICH l£S)
_ SNACK
A SNACK -BIT
S SNAP
SNAPPING
SUP t A BITE
TEN-0-CLOCK(S)
> TENNER
< TENSES
W TOMMY

30 II LANGUAGE AND IDENTITY

THE LINGUISTIC GEOGRAPHY OF
WALES
One of the most recent dialect surveys was carried
out in Wales in the 1960s under the direction of
Alan R. Thomas (1935 -) and published in 1973.
It was based on 1 80 points of enquiry in the Welsh-

speaking areas, the localities being selected on the

basis of their position relative to the physical geo-
graphy of the country and to the main communi-
cation routes.
The survey was based on a postal questionnaire,
with questions using both Welsh and English.
There were over 500 questions, which dealt largely
with domestic, rural, and farming vocabulary;
about 130,000 responses were received. The ques-
tionnaire was sent to a person of educated back-
ground, who supervised its completion by local
informants, using spelling that reflected regional
pronunciation. Informants were of the older gener-
ation, with little formal education, and had spent
no prolonged periods away from their native area.
The main part of the atlas discusses the distribu-
tion of regional words for around 400 items, on
the basis of which the main Welsh speech areas
are drawn up.The illustration (right) shows the
distribution of Welsh words for pane of glass, an
item in which two distinct patterns of use can be
clearly seen: paen and its variants in the north-east
and the midlands, cwalar and its variants in most
other places. (After A. R. Thomas, 1973.)

THE LINGUISTIC ATLAS OF THE

UNITED STATES
This survey began in 1931, under the direction of
Hans Kurath (1891 -), as part of an ambitious pro-
gramme to establish a linguistic atlas of the United
States and Canada. The region was divided into
survey areas, and the first atlas to appear, dealing
with New England, was published in 1939—43. The
project is ongoing, with informant interviews com-
plete in many areas, but the amount of work
involved means that publication is a slow and
irregular process.
The illustration (right) is taken from Kurath's
Word Geography of the Eastern United States
(1949) — a survey area that included the coastal
Atlantic states from Maine to Georgia, Penn-
sylvania, West Virginia, and eastern Ohio. Dialec-
tologists went to nearly every county in these states
and interviewed two people in each — one older-
generation and unschooled, the other a member
of the middle class with some degree of education.
In the larger cities, people with a more cultured
background were also interviewed. All were natives
of their area, and had not moved much outside
DARNING NEEDLE
it. Interviewers spent from 10 to 15 hours with
MOSQUITO HAWK
each informant, dealing with over 1,000 points of
SPINDLE
usage. More than 1,200 people were interviewed,
SNAKE FEEDER
and information was obtained about the diffusion
SNAKE DOCTOR
of around 400 regional expressions for domestic
SNAKE WAITER
and agricultural items.
The map records the distribution of words for
dragonfly.

8 GEOGRAPHICAL IDENTITY- 31
Modern dialect studies
Traditional dialectology studied geographical var-
iation, generally using elderly, untravelled, and
uneducated speakers from rural areas. Modern dia-
lectology has moved in other directions.
Social factors now provide the focus of investiga-
tion. Speech variation can be partly understood
ation. A linguistic variable is a unit with at least
two variant forms, the choice of which depends
on other factors, such as sex, age, social status,
and situation. For example, in New York, speakers
sometimes pronounce /r/ in words like car and
sometimes they do not. This unit can thus be seen
as a variable, (r), with two variant forms, /r/ and
zero. usual to transcribe linguistic variables
(It is

with reference to regional location and movement, in parentheses.) It is then possible to calculate the

but social background is felt to be an equally if extent to which individual speakers, or groups of
not more important factor in explaining linguistic speakers, use /r/, and to determine whether there
diversity and change. Modern dialectologists there- is a correlation between their preferences and their

fore take account of socioeconomic status, using backgrounds. Several interesting correlations have
such indicators as occupation, income, or educa- in fact been found (see also p. 332).
tion, alongside age and sex. Ideally informants are
found in all social groups, and the traditional focus
on the language of older people of working-class
backgrounds has been replaced by the study of
speakers of all ages and from all walks of life (§10). Dropping the h
Brad- Nor-
Dialect studies have moved from the country to In British English, the accent which carries Class ford wich
the city. The description of rural dialects led to most prestige (p. 39) pronounces /h/ at

fascinating results, but only a small proportion of the beginnings of words such as head. But Middle middle (MMC) 12% 6%
a country's population was represented in such in most other accents of England and Lower middle (LMC) 28% 14%
Wales, it is common to omit /h/ in this Upper working (UMC) 67% 40%
studies. In many countries, over 80% of the popu- position. Regions do not pronounce or omit Middle working (MWC) 89% 60%
lation live in towns and cities, and their speech /h/ with total consistency, however, as Lower working (LWC) 93% 60%
patterns need to be described too — especially as can be seen from the results of two studies
linguistic change so often begins when people from of this variable carried out in Norwich and The correlation is clear. In both areas,
Bradford. there is more /h/-dropping as one moves
the country imitate those from urban areas. This
The speakers were grouped into five down the social scale. Moreover, the pro-
approach, accordingly, is known as urban dialec- social classes,based on such factors as portion is always greater in Bradford, sug-
tology. their occupation, income, and education. gesting that the phenomenon has been
Informants are now randomly selected. In the The proportion of /h/-dropping was calcu- longer established in that area. (After J. K.

older studies, small numbers of speakers were care- lated, with the following results: Chambers & P. Trudgill, 1980.)

fully chosen to represent what were thought of as

'pure' forms of dialect. Today, larger numbers of
people are chosen from the whole population of Reading aloud in Norwich
a city - perhaps using the electoral register or a People of different social levels were Class
telephone directory. Also, the earlier approach asked to read aloud a list of isolated words
generally asked for one-word responses to a range (A) and a piece of continuous text (B), and MMC 3 28
their pronunciations when reading were LMC 10 15 42
of carefully chosen questions. This produced useful
compared with their formal (C) and casual UWC 5 15 74 87
data, but these speech patterns were unlikely to
(D) speech. MWC 23 44 88 95
have been typical. When people have their atten- The shows whether the variable
table LWC 29 66 98 100
tion drawn to the way they speak, they usually such words as walking was pro-
(ng) in The consistency with which speakers
adopt a more careful and unnatural style. Attempts nounced /rj/ or /n/. (0 = no use of /n/; increase their use of n/ as their language
100= 100% use of /n/.) becomes more spontaneous and casual is
are therefore now made to elicit speech that is more
reflected at every social level. (After P.
spontaneous in character by engaging informants Trudgill, 1974.)
in topics of conversation that they find interesting
or emotionally involving (p. 332). The question-
naire has been largely replaced by the tape recorder. /l/-dropping in Montreal
The consonant I is often dropped in the il (personal) 94 84
LINGUISTIC VARIABLES pronunciation of //('he, it'), e//e('she, it), elle 67 59
Traditional dialectology studied the fact that differ- ils ('they'), la (her, it, the), and les (the, les (pronoun) 53 41
ent people do not speak in the same way. Contem- them). The prestige forms retain the /I/. la (article) 34 25
When usage is analysed by sex of la (pronoun) 31 23
porary dialectology adds to this study the fact that
speaker, a clear pattern emerges. (The les (article) 25 15
the same person does not speak in the same way numbers represent the percentage of /l/- Women are much more likely to use the
all the time. Individuals vary in their pronuncia- dropping.) higher-prestige variant than men - a
tion, grammar, and vocabulary. Is there a reason Male Female pattern of differentiation that has often
for this variation, or is it random - 'free' variation, been found in studies of urban dialecto-

as it is often called? The current belief is that most /'/(impersonal) 99 97 logy. (After G. Sankoff & H. Cedergren,
is 94 90 1971.)
of the variation is systematic, the result of the inter-
play between linguistic and social factors.
In the 1970s, the notion of the linguistic variable
was developed, as a means of describing this vari-

32 -II LANGUAGE AND IDENTITY

Linguistic areas Front-rounded vowels
These vowels, such as in
Geographical identity can sometimes be estab- German mude ('tired') or
lished within a broader context than that provided French soeur ('sister'), are
by rural or urban dialectology. Certain features of found along an axis which
runs diagonally across
speech can identify someone as coming from a par-
northern Europe. They are
ticular part of the world, but the area involved may heard in French, Dutch,
extend over several countries, languages, or even German, Danish, Norwe-
language families (§50). The study of 'areal fea- gian, Swedish, and Fin-
nish. The feature cannot be
tures' of this kind is sometimes referred to as areal
explained on historical
linguistics.
grounds: German and Eng-
Features of pronunciation are often shared by lish are closely related, but
adjacent, but historically-unrelated languages. In the latter does not have
the indigenous languages of southern Africa (p. front-rounded vowels; nor
does Spanish, which is clo-
315), the use of click sounds in speech identifies
sely related to French. The
speakers of the Khoisan languages as well as of main factor seems to be
local Bantu languages, such as Zulu and Xhosa. geographical proximity - as
In the Indian sub-continent (p. 308), languages that further illustrated by the

belong to different families (such as Indo-European way in which many south

German dialects lack these
and Dravidian) have several important phonologi-
vowels, whereas they are
cal features in common — the use of retroflex con- found in north-east Italy. (J.
sonants (p. 155) is particularly widespread, for K. Chambers & P. Trudgill,
example. In Europe the distribution of the affricate 1980, p. 185.)

[tf] is interesting: it is found in many of the lan-

guages on the periphery of the area, such as Lapp,

Romanian, Hungarian, Spanish, Basque, Italian, A genetic explanation?
The distinctive European
Gaelic, English, and the Slavic languages. The lan-
distribution of such sounds
guages within this periphery, such as Danish, Ger- as front-rounded vowels,
man, and French, do not use it. affricates, and dental frica-

Grammatical features can also cross linguistic tives has been studied from
a genetic point of view. The
and national boundaries. The use of particles to
geneticist C. D. Darlington
mark different semantic classes of nouns (§16) can (1 903- proposed in the
)

be found throughout South-east Asia. In Europe, 1 940s that the genetic

the Balkans constitutes a particularly well-defined composition of a commu-
linguistic area. For example, Albanian, Romanian, nity would determine
partly
its preferences for types of
Bulgarian, and Macedonian all place the definite
sound. The maps show the
article after the noun, as in Romanian lup ('wolf') distribution of dental frica-
and lupul ('the wolf), whereas historically-related tives in western Europe
languages outside of the Balkans area (such as (above, left), and the fre-

do not.
Italian)
quency with which the O
blood-group gene is distri-
How do areal features develop? In some areas, buted in the population
Dental fricative as Dental fricative
have probably helped to dif-
dialect chains (p. 25) I 1

(below, left). There is an in-

phoneme today in the past
throughout an area. Con-
fuse a linguistic feature triguing correlation in : popu-
centrations of bilingual speakers along lines of I Dental fricative .is i i No dental lations where fewer than
^BBI phoneme variant today
1

fricative recorded 60% have the gene, there

communication would also play a part, and politi-
is no history of these
cal factors will have exercised their influence. sounds; and in those where
Sometimes, the progress of an areal feature can more than 65% have the
be traced - an example being the uvular pronuncia- gene, the sounds are well
represented. Unfortuna-
tion of /r/. Originally, speakers of European lan-
tely, proposals of this kind
guages pronounced /r/ with the front of their have not been followed up,
tongue; but, in the 17th century, Parisians began and remain only sugges-
to use a uvular variant. The variant caught on, tive. Social explanations of

spreading first throughout most of France, then to such distributions are cur-
rently felt to be far more like-
parts of Italy, Switzerland, Luxembourg, Belgium,
ly. (After L. F. Brosnahan,
Holland, Germany, Denmark, and (by the end of 1961.)
the 19th century) to southern Norway and Sweden.
Spain, Austria, England, and other countries were
not affected. The historical reasons for this com-
plex state of affairs are little understood, and require
investigation on several fronts. In such cases, the
facts of dialectology, social history, and political
history merge.

8 GEOGRAPHICAL IDENTITY •
33
9 Ethnic and national identity

Nowhere does the issue of personal linguistic iden- reaction is a desperate attempt to keep it, and the Religious identity Perhaps
the clearest case of a lan-
tity emerge more strongly than in relation to ques- community it represents, alive; in others, a minor-
guage fulfilling the need to
tions of ethnicity and nationhood. Ethnic identity- ity group may be rapidly growing in numbers, so define a national identity in
is allegiance to a group with which one has ances- that its language begins to compete with the estab- modern times is Hebrew.
tral links. It is a general notion, which applies to lished languages of the country for educational, When the state of Israel was
everyone, and not just to those who practise a tradi- established in 1948, there
media, and other resources; in still others, the
was an urgent need to unify
tional rural culture (a current usage of the term number of speakers may be stable, but there has its heteroge-
linguistically
'ethnic'). However, questions of ethnolinguistic been an awakening (or reawakening) of cultural neous population. Classical
identity in fact arise most often in relation to the identity, with a subsequent demand for recognition Hebrew was the obvious
demands and needs of those who are in an ethnic and (usually) territorial independence. These situ- candidate, in view of its an-
cient history and continued
minority within a community-, such as the many ations are discussed further in §61.
use as the religious lan-
groups of immigrants, exiles, and foreign workers Why should language be such a significant index guage of Judaism (and even
in Europe and the USA, or the tribal divisions that of ethnic or nationalistic movements? One reason as a secular language for
characterize several African countries. is undoubtedly that such a widespread and
it is some purposes among eas-
life. To choose one
ternEuropean Jews). The
Questions of ethnicity are closely related to those evident feature of community
complex stages that led to
of national identity. Once a group becomes aware language over another provides an immediate and the successful revival of He-
of its ethnic identity, it will wish to preserve and universally recognized badge of identity. Another brew provide a particularly
strengthen its status, and this often takes the form reason is that language provides a particularly clear clear example of the nature
of a desire for political recognition, usually self- link with the past - often the only detailed link, and procedures of language
planning (§61).
government. Political commentators have stressed in the form of literature. This link exists even after
The picture shows frag-
the subjective element in the idea of a 'nation' — ability in the language has been lost; for example, ments of one Dead
of the
the difficulty of defining the psychological bond many present-day Italian-Americans and -Austra- Sea Scrolls.
that motivates a nationalistic movement, or pre- lians know very little Italian, but they still see
dicting which elements will contribute most to a Italian as a symbol of their ethnic identity. There
group's sense of identity. Religious practices, long- is also a tendency for language to act as a natural
standing institutions, and traditional customs are barrier between cultural groups, promoting con-
all important in this respect; but perhaps the most flict rather than cooperation- as has often been
widely encountered symbol of emerging nation- seen in political meetings between opposed groups,
hood is language. In the 18th and 19th centuries, when the question of which language to use in the
in particular, linguistic nationalism was a domi- discussion has become a major procedural de-
nant European movement, with language seen as cision. In bilingual communities, or areas where
the primary outward sign of a group's identity there is a recognized lingua franca, this factor is

(§§10, 61). Today, a comparable concern can be
observed in many areas of the world, as part of
separatist political demands.
important to recognize the extent to which
It is
national diversity can give rise to linguistic issues.
Political entities that comprise a homogeneous
less important; but even here, language can focus
the sense of political grievance in a clearer way
than any other factor. There is no more awesome
testimonial to the power of language than the fact
that there have been so many people ready to die,
if their demands for linguistic recognition were not

national group are quite rare. A study of the 132 met (p. 308).
states existing in 1971 found that only 12 were
true nation-states; 50 contained a major ethnic Basque
group comprising more than three-quarters of the The way language can be- and all names in the lan- accepted responsibility for
population; and in 39 states, the largest ethnic come a symbol of national guage on official docu- Basque teaching pro-
group comprised less than half the population (W. identity is very clearly ments were translated into grammes at all levels of

Connor, 1978). National and state loyalties thus seen inthe history of
Basque (Euskera), and the
rarely coincide, and when different languages are
attitude towards it of the
formally associated with these concepts, the prob- Spanish government under
ability of conflict is real. Franco, from 1 937 until the
Linguistic conflicts due to divided ethnic and mid-1 950s. The teaching of
the language in schools
national loyalties are often bitter and violent. In
was forbidden, as was
recent years, there have been major incidents in use the media, church
in
several countries, such as India (p. 308), Spain, ceremonies, and all
Canada (p. 367), Belgium, Corsica, the USA, places.Books in the lan-
South Africa, and the Celtic-speaking areas (p. guage were publicly burnt.
Basque names were no
303). The reasons for conflict vary greatly: in some longer allowed in baptism,
cases, the use of a language is declining, and the
Spanish. Inscriptions on
public buildings and tomb-
stones were removed.
By the early 1 960s, offi-
cial policyhad changed.
Basque came to be permit-
ted in church services, and
its then in church schools and
broadcasts. 1968, a gov-
In
public ernment decree authorized
the teaching of regional
languages at the primary
level in Spain. By 1979, the
Ministry of Education had
education. In March 1980,
the first Basque Parliament
was elected, with Euskera
recognized as an official
language along with
Spanish in the Basque pro-
vinces. Current discontent,
as a consequence, is
focussed more on the
region's future socio-
economic development,
associated with persistent
demands for political auto-
nomy.

34 II LANGUAGE AND IDENTITY

 fierce arguments surrounding two viewpoints. The Ann Arbor trial
Ethnic varieties BEV was nothing more than a
Some argued that In 1 977, an important case
Varieties of language can also signal ethnic identity. restricted code 40), the result of verbal depriva-
(p. was brought by the children
In fact, probably themost distinctive feature of tion. Others, that the whole thing was a myth of the Martin Luther King

immigrant groups is not their mother

ethnicity in devised by white liberals, or an attempt to further Elementary School in Ann
Arbor, Michigan, against the
tongue (which may rarely be heard outside the But
discredit blacks. in the late 1970s, these argu-
Ann Arbor School District
home), but the foreign accent and dialect that ments were largely resolved — at an academic level, Board. The racial balance of
characterizes their use of the majority language. at least. Most contemporary linguists who have children attending the school
In the course of time, many
of these features have studied this topic accept a version of the creole at the timewas 80% white,
become established, resulting in new varieties of hypothesis, because of the striking phonological
13% black, and 7% Asian
and Latino. Some of the
the majority language. Well-known cases include and grammatical similarities between BEV and black children, who came
the range of English accents and dialects associated other Creoles, such as those of the West Indies; but from a local low-income
with speakers from the Indian sub-continent, from they allow for the probability that some features housing area, were found to
the West Indies, or from Puerto Rico. A non-regio- of BEV may have arisen partly or wholly as the be doing extremely badly in
the school.
nal example would be people with a Jewish back- result of white dialects.
The mothers of these
ground, whose speech has had a distinctive There is a continuing need to disseminate the children believed that this
influence on many European languages. facts about the relationship between standard Eng- situation was due to the
lish and non-standard varieties, such as BEV, school's failure to take into
account the children's racial
BLACK ENGLISH VERNACULAR because the principle of mutual recognition and
and sociocultural back-
One of the clearest examples of ethnic linguistic respect is constantly being challenged. In particular ground. It was argued that

variety is provided by the contrast between the

one has to anticipate the severe linguistic disadvan- there was a 'linguistic bar-

tage that affects children from these dialect back- rier', in the form of BEV,
speech of black and white Americans. There is no
which impeded their aca-
simple correlation between colour and language, grounds when they go to school, where the medium
demic performance, and that
because there is considerable linguistic variation of instruction and criterion of successful perfor-
this barrier prevented the

within both racial groups, and it is perfectly pos- mance is standard English. These days there is an children from having the

sible for black speakers to 'sound' white, and vice increasing understanding of the educational issues equal educational opportu-
(§44); but an enlightened approach to the problem nity that was their rightunder
versa, depending on educational, social, and re- 20
Title of the U.S. Code. Al-
is by no means universal.
gional factors (p. 18). The term 'Black English' has ternative educational pro-
been criticized, therefore, because of its suggestion grammes should have been
that all blacks use the same variety, and has been
Some grammatical features of BEV provided to cater for their

replaced in academic study by 'Black English Ver-

• No third-person singular present
final 5 in the unique linguistic needs.
tense, e.g. he walk, she come.
The case thus depended
nacular' (BEV), referring to the speech of the on whether BEV was so dif-
group most often studied in this context - the non- • No use of forms of the verb be in the present ferent from standard English
standard English spoken by lower-class blacks in tense, when it is used as a copula, or 'linking' as to constitute a barrier.
urban communities. verb, within a sentence, e.g.They real fine, If Other considerations, of a
you interested. and economic kind,
cultural
Some features of BEV are given below. It is not were judged irrelevant. Re-
• The use of the verb be to mark habitual meaning,
clear just how widespread these features are cordings were played in
amongst the black community; nor is it obvious but without changing its grammatical form court of the children's spon-

where they come from. In one view, all BEV fea- ('invariant be''), e.g.
round here.
tures can be found in white English dialects (espe-
• Use of been to express
cially those of the southern USA), suggesting that
black English historically derived from white. The with current relevance,
Sometime they be walking taneous speech, which was
shown to be similar to the
a meaning of past activity
BEV used by black children
elsewhere; and a team of
e.g. / been know your linguistics experts testified to

association with blacks is then explained as a result name. the extent of the language

of their emigration to the northern cities, where

• Use of be done in the sense of 'will have', e.g. differences, and to the creole

these features were perceived as a distinctive We be done washed all those cars soon. history of BEV, which indi-
cated that these differences
marker of ethnic, as opposed to regional, identity. • Use of it to express 'existential' meaning (cf.
were the result of racial
With the development of urban ghettos, the con- standard English there), e.g. It's a boy in my segregation.
class name Mike. The won
trast became more marked over time. The alter- plaintiffs their
• Use of double negatives involving the auxiliary case, and the School Board
native view argues that the origins of BEV lie in
verb at the beginning of a sentence, Won't was directed to take steps to
e.g.
the use of a creole English (p. 336) by the first help the teachers identify
blacks in America. This language, originally very nobody do nothing about that. children speaking BEV, and
different from English as a result of its African to use that knowledge in

teaching the children to read

linguistic background, has been progressively
standard English. Since
influenced by white English so that it now retains then, several other school
only a few creole features. districtshave developed pro-
It is often difficult to obtain an objective discus- grammes, influenced by this
sion and evaluation of the linguistic evidence decision. The Ann Arbor
judgment can therefore be
because of the existence of strong emotions around
seen as a landmark in the
the subject, and the colour prejudice which has pro- slow process towards the
moted the view that black English is necessarily public recognition of ethnic
inferior to white — a view that has no linguistic linguistic identity. (After W.
Labov, 1982.)
validity ($2). During the early 1970s, there were

9 ETHNIC AND NATIONAL IDENTITY .

35
 .

GASTARBEITER THE ETHNICITY BOOM

There are now
over 24 million migrant workers
(often called Gastarbeiter 'guest workers') and
their dependants in north-west Europe. They come
from several countries, such as Turkey, Yugoslavia,
Greece, Italy, Japan, and the Arabic-speaking
countries. The demands of their new life require
a level of adaptation that transcends language fron-
tiers, and these workers often do not make an issue
of their linguistic identity. On the other hand, their
communication skills are usually limited, and the
social and educational problems of the receiving
country are considerable.
In the early 1980s, for example, there were over
700,000 foreign pupils in German
over 900,000 in French schools. In 1981, minority
languages being taught in French schools included
German, English, Spanish, Italian, Portuguese,
Arabic, Hebrew, Russian, Japanese, Dutch, Serbo-
Croat, Chinese, and Turkish. Even in a small
country, significant minority language problems
exist: in Denmark, for example, migrants from
Yugoslavia, Turkey, and the Nordic countries have
to be catered for; in Luxemburg, there are many
Italians and Portuguese. In Britain, there are
around 100 minority languages, about a quarter
of which are taught in schools to over 400,000
pupils.
The situation is likely to become
plex in Europe with increasing international mo-
bility within the European Economic Community,
where member-states are still investigating solu-
tions to the problem of language teaching and
learning (§62). But at least the problem is now for-
mally recognized. In 1977, the Council of the Euro-
pean Economic Community issued a directive on
the education of children of migrant workers in
Europe. The directive applied only to member-
states, but the Council resolved to extend the
measures to include all immigrant children within
the Community (over li million). The aim of the
exercise was to adapt school structures and curri-
cula to the specific educational needs of these
children without losing sight of their cultural and
linguistic identity.
Between the mid-1960s and the mid-1970s, west-
ern Europe and North America experienced an
'ethnicity boom'. Considerable progress was made
in integrating minority indigenous or immigrant
groups within their host communities, and there
was a widespread raising of consciousness about
ethnicity issues. This was especially noticeable in
the USA, where 1970 census data showed that
17% of the American population (over 33 million)
claimed a mother tongue other than English — the
largest claims relating to Spanish, German, Italian,
French, Polish, and Yiddish. This was a dramatic
increase of 71% compared with 1960 (though the
schools, and total population increased by only 13% during that
decade) and a marked reversal of the decline seen
in the period 1940-60.
However, during the 1970s a further change
took place. There was still an overall increase in
the number of people claiming a mother tongue
other than English, but this increase was largely
due to Spanish. For many other languages, es-
pecially German, Yiddish, and the Scandinavian
languages, there was a notable decline. Evidently,
large numbers of the younger generation, from
mainly North European backgrounds, were ceas-
ing to claim these languages as their mother tongue.
On the other hand, the claims increased for some
yet more com- South European languages (e.g. Greek and Portu-
guese, as well as Spanish) and for most Asian lan-
guages. (After J. A. Fishman, 1984.)
It is perhaps too early for these changes to be
given a social interpretation. One analysis has
drawn attention to the contrast between the (de-
creasing) languages of white North and Central
European Christendom, which were among the
earliest settlers, and the (increasing) non-European
languages associated with largely eastern religious
groups, whose arrival in the USA is more recent.
The former have now become a part of the Ameri-
can mainstream, it is argued, whereas the latter
have still to find their identity within that culture.
Because they are less accepted, they are more aware
of the importance of maintaining traditional
linguistic ties.

Article 2 Member States shall, in accordance with

their national circumstances and legal systems, take
Mother-tongue claim ng
appropriate measures to ensure that free tuition to
Mother tong ues claimed by ove •100,000 Czech 452,812 + 15%
facilitate initial reception is offered in their territory to
people in the U.S. in 1 970, with an esti- Hungarian 447,497 + 17%
the children .including, in particular, the teaching -
. .
mate of the percentage increase (+) or Dutch 412,627 -6%
adapted to the specific needs of such children - of the decrease (- ) in 1979 (after J. A Fishman, Japanese 408,504 +30%
official language or one of the official languages of the 1984). Portuguese 365,300 +30%
host State. Chinese 345,431 +87%
English 160.717,113 +6% Russian 334,615 + 17%
Article 3 Member States shall, in accordance with Spanish 7,823,583 +46% Lithuanian 292,820 +6%
their national circumstances and legal systems, and in German 6,093,054 -10% Ukrainian 249,351 +6%
cooperation with States of origin, take appropriate Italian 4,144,315 + 5% Serbo-Croatian 239,455 + 17%
measures to promote, in coordination with normal French 2,598,408 +7% Tagalog 217,907 + 75%
education, teaching of the mother tongue and culture Polish 2,437,938 +5% Finnish 214,168 -10%
of the country of origin for the children . .
Yiddish 1 ,593,993 -24% Danish 194,462 -10%
Swedish 626,102 -11% Arabic 193,520 + 17%
Norwegian 612,862 -2% Hebrew 101,686 +500%
Slovak 510,366 + 17% Armenian 100,495 + 17%
Greek 458,699 +25%

36 • II LANGUAGE AND IDENTITY

Minority languages
1 . Britain The dramatic in-
crease in immigrant numbers
in 960s has resulted in
the 1 status as a symbol of Maltese
over 100 languages being
used in Britain by ethnic
minority communities.
most widely spoken immi-
grant languages appear to
be Panjabi, Bengali, Urdu,
Gujarati, German, Polish,
Italian, Greek, Spanish, and
Cantonese. For the situation
of the Celtic languages, see
p. 303.
2. France There are several
minority languages indige-
nous to France- Basque,
Breton (p. 302), Catalan,
Corsican, Alsatian, Flemish,
and Occitan. In a 1 978 sur-
vey, three-quarters of the
population wished to retain
this diversity, but only
actually understood or
spoke one of these lan-
guages.
3. Luxembourg Letzebuer-
gesch, related to German, is
spoken as a mother tongue,
and it is taught in schools,
along with French and Stan-
dard German. It retains a
strong popular appeal as a
symbol of national identity.
4. Spain The history of Cata-
lan, centred on the Barce-
lona area, is similar to Bas-
que (p. 34), with an early
history of repression, and the
recent acquisition of a de-
gree of autonomy. In the
north-west corner of the
country, the Galician dialect,
closely related to Portu-
guese, provides a link
the old Kingdom of Galicia.
5. Switzerland German is
spoken by nearly 70% of the
Swiss population, French by
around 1 9%, and Italian by
10% (most of the latter living
in the canton of Ticino). This
leaves Romansch, spoken
by fewer than 50,000 in the
canton of the Grisons (Grau-
bunden). The language is
rapidly declining, under the
influence of German, though
it continues to be the early
medium of education in the
region, and there has been a
recent attempt at cultural
revival. The Romansch
League looks after all
conservation measures
relating to the language.
6. Malta Maltese is the
national language of Malta.
English also has official
status,and Arabic and Italian
are widely known. Since
receiving official status
(1934), its role in written
contexts has increased, and
in Europe
the island's television chan-
nel broadcasts in Maltese. Its
identity is widespread. (See
further, p. 316.)
The 7. Belgium A linguistic and
cultural boundary runs
across Belgium. In the north
and west are the Flemings,
descendants of the Franks,
who speak dialects of Dutch
(known as Flemish, or
Vlaams). In the south and
east are the Walloons, de-
scendants of the Romano-
Celts, who speak dialects of
French. There is also a small
German-speaking area in
the east. The capital, Brus-
sels, is officially bilingual,
though predominantly
French.
35% Early Belgian history saw
French as the dominant lan-
guage, and this situation
continued until the 1930s,
when official status was
given to Flemish. Since then,
linguistic issues have come
to dominate Belgian politics,
as efforts were made to es-
tablish language frontiers,
and to provide satisfactory
representation and
political
educational resources. In
1 968 serious rioting over
plans to expand the French-
speaking section of the Uni-
versity of Louvain (Leuven)
brought down the govern-
ment. The four linguistic
areas have now been offi-
cially recognized, with each
responsible for its own
with affairs; but the complex
social situation by no
is
means and several
resolved,
further governments have with the result that there is in their mother tongue, if their languages. However, a Lan-
fallen as a result of linguistic local German concern about parents request it; and these guage Board was set up in
policies. future separatist develop- classes are growing. By 1971.
8.Netherlands Frisian is ments. There is also a large 1 980, in comprehensive 13. Yugoslavia Serbo-
accepted as an official lan- Gastarbeiter population. schools, about 50,000 stu- Croat, Slovenian, and Mace-
guage in the Netherlands, 10. Sweden Until the 1930s, dents were being taught in donian have official status,
and has its own academy.
it Sweden was ethnically 60 different languages. but others (Albanian, Hun-
It is used in schools and homogeneous; but following 1 .Finland There were garian) have been given
courts, especially in the theSecond World War, there around 300,000 Swedish some degree of autonomy in
Friesland area; but generally was a large influx of refu- speakers in Finland in 1970- their own provinces. It is
its use is diminishing, under gees, mainly from Finland, about 7% of the population. government policy to separ-
the influence of Dutch. The but also from Italy, Hungary, Swedish is an official lan- ate linguistic from nationality
minority population consists Austria, West Germany, guage, alongside Finnish, rights, with only the former
mainly of over a million Suri- Greece, and Yugoslavia. Im- and education is available in being supported.
namese, Indonesians, Mo- migration has been con- Swedish at all levels; but 14. Romania The many min-
luccans, and Frisians- with trolled since 1 967, but there most Swedish speakers are ority languages of this
a sizeable gastarbeiter are still around a million bilingual, and the language country include Hungarian
group. people from non-Swedish generally seems to be in (15 million), German, Ukrain-
9.Germany In North Frisia, backgrounds- about 10% of decline. ian, Romany, Russian, Serbo-
Low and High German are the population. In 1975, Par- 12. Lapland Around 35,000 Croat, Yiddish, Tatar,
used, alongside Danish, Ju- liament recognized that Lapps live in Norway, Swe- Slovak, Turkish, Bulgar-
tish, and Frisian -the latter in these groups should have den, Finland, and the USSR. ian,and Czech. The larger
several divergent dialects freedom of choice to retain The Lappish language has languages have media
(p. 365). There are both Ger- their identity. There is an no official status in any of coverage; but all have some
man and Danish Frisians, active bilingualism policy. these countries, and educa- official status.
but the most active group is Foreign children may have tional practice is primarily
oriented towards Denmark, some educational instruction concerned with the majority
9 ETHNIC AND NATIONAL IDENTITY • 37
10 Social identity

In addition to the questions 'Who are you?' and

Castes
'Where are you from?', which have been addressed
Probably the clearest examples of social distinctionsbetween the phonology, vo-
from a linguistic viewpoint in §§6—9, there is also cabulary, and grammar of Brahmin and
dialects are those associated with a caste
'What are you, in the eyes of the society to which system. Castes are social divisions based non-Brahmin speech. The former also
you belong?' It is a complex and multi-faceted solely on birth, which totally restrict a per- tends to use more loan words, and to pre-
question, to which there is no easy answer. People son's way of life- for example, allowing serve non-native patterns of pronunciation.
only certain kinds of job, or certain mar-
acquire varying status as they participate in social
riage partners (p. 401 ). The best-known Non-
structure; they belong to many social groups; and
system is that of Hindu society in India, Brahmin Brahmin
they perform a large variety of social roles. As a which has four main divisions, and many Vocabulary
consequence, no single system of classification is sub-divisions -though in recent years, the tungu sheep' orangu
likely to do justice to the task of defining a person's caste barriers have been less rigidly alambu 'wash' kaluyu
enforced. The Brahmins (priests) consti- jalo 'water' tanni
social identity in linguistic terms, especially when Phonology
tute the highest class; below them, in des-
the vast range of the world's cultural patterns is
cending order, are the Kshatriyas krafu 'haircut' krappu
taken into account. This section, therefore, has to (warriors), Vaisyas (farmers and mer- jini sugar' cini

be extremely selective, in order to represent the chants), and Sudras (servants). The so- varepparo 'banana' vareppolo
range of sociolinguistic and ethnolinguistic vari- called 'untouchables', whose contact with valeppolo
the other castes is highly restricted, are the Grammar
ables involved.
lowest level of the Sudra caste. -du 'if -ecu
Linguistic correlates of caste can be vandudu came'
'it vanduccu
found at all levels of structure. For exam- panra he does' pannuha
ple, in Tamil, there are several clear-cut (After W. Bright & A. K. Ramanujan, 1 964.)
Social stratification
One of the chief forms of sociolinguistic identity
Speech and silence in Kirundi
derives from the way in which people are organized
In the Central African kingdom of Burundi, further proceedings are effectively
into hierarchically ordered groups, or
social
age and sex combine with caste to con- negated.
classes. Classes are aggregates of people with simi- strain the nature of linguistic interaction in To speak well is considered a mark of
lar social or economic characteristics. Within several ways. Seniority (ubukuru) governs good breeding in men. From their tenth
all behaviour. There are clear caste divi- year, boys in the upper castes are given
sociology, the theoretical basis of social class has
sions; older people precede younger; and formal speech training- how to use social
been a controversial subject, and it has not always
men precede women. The order in which formulae, talk to superiors and inferiors,
proved easy to work consistently with the notion, people speak in a group is strictly gov- and make speeches for special occasions.
especially when cross-cultural comparisons are erned by the seniority principle. Males of Upper-caste girls do not take part in public
involved. Factors such as family lineage, rank, highest rank must speak first, regardless speaking, but they do develop effective
of age. Females do not speak at all, in the bargaining skills, for use behind the
occupation, and material possessions often conflict
presence of outsiders, unless spoken to. scenes. They are also trained to listen with
or are defined with reference to different criteria. Upper-caste speakers seem never to great care, so that they can accurately
But for most sociolinguistic purposes to date, it raise their voices, or allow emotion to recount to the men of the family what has
has been possible to make progress by recognizing show. In group discussion, for the senior been said by visitors. (After E. M. Albert,
only the broadest distinctions (such as high vs low, person to be silent implies disapproval. As 1964.)
others must then also stay silent, any
or upper vs middle vs lower) in order to determine
the significant correlations between social class
background and language. Examples of some of The John Betjeman poem, How to get on in society',
these correlations are given (below) and also on originally setas a competition in Time and Tide, was
p. 32. included in the book Noblesse Oblige as part of the
One does not need to be a sociolinguist to sense U/non-U debate (see facing page).
way people talk has something to do with
that the
their social position or level of education. Everyone
has developed a sense of values that make some
How to get on in society
accents seem 'posh' and others 'low', some features
Phone for the fish-knives, Norman, And Howard is out riding on horseback
of vocabulary and grammar 'refined' and others
As Cook is a little unnerved; So do come and take some with me.
'uneducated'. We have a large critical vocabulary You kiddies have crumpled the serviettes
for judging other people's language in this way.
Now here is a fork for your pastries
And must have things daintily served.
I

And do use the couch for your feet;

But one does need to be a sociolinguist to define Are the requisites all in the toilet? Iknow what wanted to ask you -
I

precisely the nature of the linguistic features that The frills round the cutlets can wait Is trifle sufficient for sweet?
are the basis of these judgments of social identity. Till the girl has replenished the cruets
Milk and then just as it comes, dear?
And it is only as a result of sociolinguistic research And switched on the logs in the grate.
I'm afraid the preserve's full of stones;
that the pervasive and intricate nature of these cor- It's ever so close in the lounge, dear, Beg pardon, I'm soiling the doilies

relations has begun to be appreciated. But the vestibule's comfy for tea, With afternoon tea-cakes and scones.

38 •
II LANGUAGE AND IDENTITY
 SOME ENGLISH MARKERS OF SOCIAL today (p. 32), but a century ago this pronunciation Social identity and
CLASS was a desirable feature of speech in the upper mid- other factors
Long before the days of 20th-century linguistics dle class and above — and may still occasionally It is never possible to make

and phonetics, English novelists and dramatists,
especially in the 18th and 19th centuries, were
observing the relationship between language and
social class in Britain and using it as a basis for
characterization and social comment.
• George Gissing, about Mrs Yule,
Grub Street (1891, Chapter 7).
Mrs Yule's speech was seldom ungrammatical, and her
intonation was not flagrantly vulgar, but the accent of
the London poor, which brands as with hereditary
baseness, still clung to her words, rendering futile such
propriety of phrase as she
with educated people.
• Mrs Waddy, about Harry Richmond's father,
in George Meredith's The Adventures of Harry
Richmond (1871, Chapter 3).
'More than his eating and
be heard. The change to [rj] came about under the a simple statement about lan-
influence of the written form: there was a g in the guage variation and social
class because other influen-
spelling, and it was felt (in the late 19th century)
factors are involved, such
tial
that it was more 'correct' to pronounce it. As a
as the sex of the speaker,
result, 'dropping the g' in due course became stig- and the formality of the situa-
matized. There is also an
tion (p. 42).
in his New important interaction be-
U AND NON-U tween social and regional
factors (§8), as illustrated
In 1954, A. S. C. Ross (1907-) published an article below for British English.
entitled 'Linguistic class-indicators in present-day The two pyramids deal with
English' in a Finnish philological journal. It was differences of accent and
read by Nancy Mitford, who wrote an Encounter dialect, and represent the re-
owed to years of association
article based upon it. The result was an enormous
lationship between where' a
speaker is, both socially (the
public reaction, with immediate recognition for the vertical dimension) and geo-
terms U and non-U. Two years later, Ross's essay graphically (the horizontal
dimension). At the top are
was reprinted, with some modifications and a new
the speakers of the highest
title ('U and Non-U: an essay in sociological
social class: they speak the
linguistics'), in Noblesse Oblige, which included
standard dialect with very
his drinking, that child's contributions on the same subject by Nancy Mit- little regional variation. Also

father worrits about his learning to speak the language ford, Evelyn Waugh, and John Betjeman. at the top are those who
of a British gentleman Before that child your "h's" The aim was speak Received Pronuncia-
. . .
essay's to investigate the linguistic
must be panting of an engine - to please his
like the tion (RP), the educated ac-
demarcation of the British upper class. U stood cent which signals no regional
father and I'm to repeat what I said, to make sure
for 'upper class' usage; non- U stood for other kinds
. . .

.' information at all (within

the child haven't heard anything ungrammatical . .

of usage. It looked at distinctive pronunciation and Britain). The further we move

vocabulary, as well as written language conven- down the class scale, the
• Pip to Biddy, in Charles Dickens' Great Expec- more we encounter regional
tions, such as how to open and close letters. It was
tations (1861,Chapter 35). accent and dialect variation.
a personal account containing many subjective And when we reach the low-
'Biddy,' said I, in a virtuously self-asserting manner, 'I judgments and disregarding the subtle gradations est social class, we en-
must request to know what you mean by this?' in usage intermediate between the two extremes; counter the widest range of
'By this?' said Biddy. local accents and dialects.
but it was also highly perceptive, drawing attention
'No, don't echo,' I retorted. 'You used not to echo,
to a largenumber of distinctive features. The nature Dialects
Biddy.'
of upper-class language has changed over 30 years Highest class:
'Used not!' said Biddy. 'O Mr Pip! Used!' o
later, but the terms U and non-U are still well standard
known. English
• Elfride Swancourt to Mrs Swancourt, in Thomas
Hardy's A Pair of Blue Eyes (1873, Chapter 14).
Some of the lexical oppositions proposed by Lowest
Ross: class:
i have noticed several ladies and gentlemen looking at most
me.' u non-U Regional non-
'My you mustn't say "gentlemen" nowadays
dear, . .
have a bath take a bath variation standard
varieties
We have handed "gentlemen" to the lower classes, bike, bicycle cycle
where the word is still to be heard at tradesmen's balls luncheon dinner
Accents
and provincial tea-parties, I believe. It is done with riding horse riding
here.' sick ill Highest class:
'What must I say then?' Received
knave jack
Pronunciation
'
"Ladies and men" always.'
mad mental
looking-glass mirror Lowest
Dropping the g
writing-paper note-paper class:
'Where on earth did Aunt Em learn to drop her g's?' most
'Father told me once that she was at a school where
jam preserve
Regional localized
an undropped "g" was worse than a dropped "h". They wireless radio variation accents
were bringin' in a country fashion then, huntin' people, table-napkin serviette
Thus, for example, speakers
you know.' lavatory-paper toilet-paper
from the top social class will
rich wealthy all use the same word head-
This conversation between Clare and Dinny Cher-
vegetables greens ache, and give it the same
rel, in John Galsworthy's Maid in Waiting (1931, (RP) pronunciation, but
pudding sweet
Chapter 31), famous linguistic signal
illustrates a speakers from the lowest
telegram wire
of social class in Britain - the two pronunciations class will use skullache,
England Britain head-wark, head-warch, sore
of final g in such words as running, [n] and [rj].
Scotch Scottish head, and other forms, in a
But it also brings home way very well the arbitrary variety of pronunciations, de-
in which linguistic class markers work. The [n] pending on where they are
variant is typical of much working-class speech from. (After P. Trudgill, 1 983.)

10 SOCIAL IDENTITY -39

RESTRICTED AND ELABORATED or registers. Certainly, it is possible to show that Wolof greetings
CODES a lower-class speaker can handle abstract concepts Greeting behaviour has a
Do people from different social classes display dif- in restricted code. For example, in one of the special place among the
ferent abilities in their use of language? This was recordings made by William Labov (1927-), a Wolof of Senegal, and well
illustrates the link between
one of the questions widely discussed in the 1970s, black 15-year-old was asked why he thought a God
language and social identity.
as a result of a distinction proposed by the sociolo- would be white. He replied: 'Why? I'll tell you why! Every interaction must begin
gist Basil Bernstein (1924-).The concepts of 'ela- Cause the average whitey out here got everything, with a greeting.
borated code' and 'restricted code' attempt to you dig? And the nigger ain't got shit, y'know? In the country, a greeting

explain how a society's distribution of power and Y'understan'? So - um — for in order for that to occurs between any two per-
sons who are visible to each
its principles of control shape and enter different happen, you know it ain't no black God that's doin'
other- even if one person
modes of communication which carry the cultures that bullshit.' There is plainly abstract reasoning has to make a detour to ac-
of different social classes and that of the school, here, despite the non-standard language, and the complish it. In crowded
and so reproduce unequal educational advantages. restricted code. areas, everyone close to the

The theory proposes speaker must be greeted. In

that the sets of social relation- Studies of this kind show that the correlation
a conversational gathering,
ships in which people are embedded act selectively between the use of language and social class is evi- everyone must be greeted at
on the production of meanings, and so upon dently not simple: other factors intervene, such as the outset; and if, in the
choices within common linguistic resources. the context in which learning takes place, and the course of the conversation,
Codes are said to have their origins in different way family life is structured. These factors always someone leaves and then re-
turns, itoften necessary to
is
family structures, associated (but not inevitably) need to be borne in mind when debating levels of
pause while all are greeted
with social classes, and are relayed through crucial linguistic 'deficiency' or 'difference' between individually again.
socializing contexts, instructional and regulative, people of different social classes. Wolof society is divided
which differently orient children to the roles, mean- into several castes, and a
ings, and values of the school. Restricted codes THE LANGUAGE OF RESPECT person's social identity is in-
volved in every greeting. The
arise where meanings are particular to and embed- Many communities make use of a complex system most senior people present
ded in a local context, and the need to make mean- of linguistic levels in order to show respect to each are greeted before those of
ings specific and explicit is reduced by the other. The levels will partly reflect a system of social lower rank; and in any meet-
ing, those of lower rank must
foregrounding of shared understandings, values, classes or castes, but the choice of forms may be
speak first. When two people
and identifications. By contrast, the forms of ela- influenced by several other factors, such as age,
meet, they must reach a tacit
borated codes arise out of social relations where sex, kinship relationships, occupation, religious agreement about their rela-
less is taken for granted, where shared understand- affiliation, or number of possessions. In Javanese, one who talks
tive status: the

ings, values, and identifications are less fore- for example, choice of level can in addition be first accepts the lower role.
Variations in status also
grounded, and so where explicitness and specificity affected by the social setting of a conversation, its
occur. For example, an
are more demanded. Middle-class chil-
likely to be subject matter, or the history of contact between upper-caste person may not
dren are said to have access to both codes, whereas the participants. Other things being equal, people wish to adopt the higher-
lower-working-class children are more likely to be would use a higher level at a council meeting than ranking position, because
initially limited to a restricted code, and to experi- about religious matters than
in the street; in talking that would oblige him to sup-
port the lower-ranking per-
ence difficulty in acquiring the form of the elabor- about buying and selling; and when addressing
son with a gift at some future
ated code required by the school, and thus the someone with whom they had recently quarrelled. point. He would therefore
meanings and pedagogical practices regulated by Similar constraints have been noted for several lan- attempt to lower himself by
that code. guages, such as Japanese (p. 99), Korean, Tibetan, speaking first in a conversa-
tion.
The complexities of this theory were sometimes Samoan, and Sundanese.
A Wolof proverb sums up
reduced to the proposition that middle-class Devices for conveying relative respect and social this principle of social in-
children are able to abstract, but working-class distance can be found in all languages. What is equality: sawaa dyi, sawaa
children are not; this difference was then attributed distinctive about 'respect' languages is the way dif- dyi,gatyangga tya,
to differences in the children's linguistic resources. ferences of social level have been so extensively ndamangga ca When two
persons greet each other,
Bernstein argues strongly that there is no basis for coded in the grammar and vocabulary. In Javanese,
one has shame, the other
either of these propositions in his theory. Misread- the differences between levels are so great that has glory.' (After J. T. Irvine,
ings of the theory can also occur through a too- equivalent sentences may seem to have very little 1974.)
ready association of codes with language varieties in common.

Level are you going to eat rice and cassava now Complete Five status levels, in one
Javanese dialect (after C.
dabar Menapa pandjenengan bade dabar Geertz, 1 968), using the sen-
krama inggil pandjenengan
sekul kalijan kaspe samenika? tence Are you going to eat
menapa bade kalijan samenika
and cassava now? The
rice
Menapa sampejan bade neda sekul
krama bias J
sekul
kalijan kaspe samenika?
names krama, madya, and
ngoko refer to high', mid-
napa sampejan adjeng neda kaspe saniki Napa sampejan adjeng neda sekul dle',and 'low' respectively. In
madya Ian kaspe saniki? addition, the high and low
levels each have two divi-
Apa sampejan arep neda sega Ian sions, depending on whether
ngoko madya Ian
kaspe saiki? honorific words are used, to
apa arep sega saiki
produce krama inggil vs kra-
kowe mangan Apa kowe' arep mangan sega Ian
gnoko biasa
kaspe saiki?
ma biasa, and ngoko madya
vs ngoko biasa.

40 • II LANGUAGE AND IDENTITY

Social status and role Ceremonial language
Probably all communities there are grammatical and many stylistic devices,
'Status' is the position a person holds in the social
have developed special lexicalchanges. They use such as metaphors, pro-
structure of a community — such as a priest, an uses of language for ritual many more Spanish loan verbs, and comparisons.
official, a wife, or a husband. 'Roles' are the con- purposes. Distinctive forms words than in everyday The genre uses traditional
ventional modes of behaviour that society expects are employed by those who speech (60%, compared ways of speech, handed
when holding have official status in the with 20%), and there is a down from ancestors. To
a person to adopt a particular status.
ceremony, as well as by marked increase in the speak Malagasy well
Public roles often have formal markers associated those who participate. This number of affixes in a word means to be in command
with them, such as uniforms; but among the chief may extend to the use of (as many as 1 1 attached to of this style; and it is com-
markers of social position undoubtedly language.
is totally different languages a root, compared to the six mon to hear speakers'
People exercise several roles: they have a particular (without regard for listener or fewer heard in ordinary abilities discussed and
intelligibility), or be no more use). evaluated.
status in their family (head of family, first-born,
than selective modifica- Often, ceremonial In a marriage request
etc.), and another in their place of work (super- tions of everyday speech - genres are marked by con- ceremony, for example, the
visor, apprentice, etc.); they may have a third in such as prayers and siderable verbal ingenuity. girl's family gather in her
their church, a fourth in a local sports centre, and speeches that are dis- For example, among the village, and await the arri-
tinguished only by a more llongot of the northern val of the boy's family.
so on. Each position will carry with it certain
careful articulation, abnor- Philippines there is a Each is represented by a
linguistic conventions, such as a distinctive mode speech-maker. As the
mal prosody, and the occa- speech style known as
of address, an 'official' manner of speech, or a sional use of exceptional 'crooked language' (qam- boy's family approaches,
specialized vocabulary. During the average life- vocabulary and grammati- baqan), used in oratory, no official notice is taken of
time, people learn many such linguistic behaviours. cal forms. play, song, riddles, and them until their speech-
Among the Zuni, for ex- public situations, such as maker makes a series of
It only occasionally that the adoption of a
is
ample, 'sacred words' (t6w- debates. It is a style rich in requests to enter the vil-
social role requires the learning of a completely usup£na' we), usually witty repartee, puns, meta- lage. Unless the girl's
different language. For instance, a knowledge of prayers, are pronounced in phor, elaborate rhythms, speech-maker judges that
Latin is required in traditional Roman Catholic rhythmical units, resem- and changes in words. In these speeches are per-
bling the lines of written Malagasy, there is a con- formed adequately, ac-
practice; a restricted Latin vocabulary was once
poetry, with a reversal of trast between everyday talk cording to the traditional
prerequisite for doctors in the writing out of pre-
the expected patterns of (resaka) and oratorial per- standards of the kabary,
scriptions; students in some schools and colleges stress and intonation: formance (kabary), which they will not be allowed to
still have to speak a Latin grace at meal-times; and strongly stressed syllables is used in ceremonial situa- proceed to the formal mar-
Latin may still be heard in some degree ceremonies. become weak, and the tions such as marriages, riage request, and the
weakest syllable in the unit deaths, and bone-turnings, speech-maker must re-
More usually, a person learns a new variety of lan- and also in formal settings, double his efforts. Sub-
is pronounced most
guage when taking up a social role — for example, strongly. Ceremonial such as visits. An obliga- sequent steps in the
performing an activity of special significance in a speech among the Kamsa tory feature of kabary is ceremony are evaluated in

culture (such as at a marriage ceremony or council Indians of Colombia also 'winding' speech, in which the same way. (After
involves distinctive intona- male speakers perform a E.Keenan, 1974.)
meeting), or presenting a professional image (as
tion and timing, reminiscent dialogue a roundabout,
in
in the case of barristers, the police, and drill
of chant, but in addition allusive manner, using
sergeants). The use of new kinds of suprasegmental
feature (§29) is particularly important in this
One of the most distinctive indications of Kabary in progress An orator at a Malagasy marriage ceremony.
respect.
professional role is the intonation, loudness,
tempo, rhythm, and tone of voice in which things
are said.
In many cases, the linguistic characteristics of
social roles are fairly easy to identify; but often they
are not, especially when the roles themselves are
not clearly identifiable With un-
in social terms.
familiar culturesand languages, too, there is a pro-
blem in recognizing what is really taking place in
social interaction or realizing how one should
behave when participating in an event. How to
behave linguistically as a guest varies greatly from
culture to culture. In some countries, it is polite
to comment on the excellence of a meal, as one
eats it; in others, it is impolite to do so. In
some countries, a guest is expected to make an
impromptu speech of thanks after a formal meal;
in others there is no such expectation. Silence, at
times, may be as significant as speech (p. 38).

10 SOCIAL IDENTITY -41

 sharply maintained — for instance, several places Avoidance
Social solidarity and distance have separate names in all the languages, and the languages
One of the most important functions of language Indians themselves emphasize their mutual unintel- Among Australian abori-
variation is to enable individuals to identify with ligibility. In such circumstances, the languages act gines, it is common for a man
a socialgroup or to separate themselves from it. as badges of membership of the tribal units. An to 'avoid' certain relatives -

The markers of solidarity and distance may relate Indian will often speak initially in his own father often his wife's mother and
maternal uncles, sometimes
to family, sex 46), ethnicity, social class (p. 38),
(p. language to acknowledge publicly his tribal affilia-
her father and sisters as well.
or to any of the groups and institutions that define tion. And language acts as a criterion for all kinds Brothers and sisters, too,
the structure of society. They may involve tiny sec- of social behaviour. For example, when the investi- may not be allowed to con-
tions of the population, such as scout groups and gator asked a Bara Indian about marriage sanc- verse freely, once they grow
street gangs, or complete cross-sections, such as tions, she was told: 'My brothers are those who
up. In some tribes, avoid-
ance of taboo relatives
religious bodies and political parties. The signals share a language with me. Those who speak other means total lack of contact;
can be as small as a single word, phrase, or pronun- languages are not my brothers, and I can marry in others, a degree of normal
ciation, or as large as a whole language their sisters.' On another occasion, when she asked speech is tolerated; but the
an Indian why they spoke so many languages most interesting cases are
those where special lan-
DIFFERENT LANGUAGES instead of using the lingua franca, she received the
guages have developed to
Probably the clearest way people have of signalling reply: 'If we were allTukano speakers, where enable communication to
their desire to be close to or different from those would we get our women?' (After J. Jackson, take place. These are
1974.) usually referred to as
around them is through their choice of languages.
'mother-in-law' languages,
Few societies are wholly monolingual, and it is thus but all taboo relatives are in-
possible for different languages to act as symbols DIFFERENT VARIETIES cluded under this heading.
of the social structure to which their speakers In monolingual communities, a major way of In Dyirbal (now almost ex-

belong. The test sentence 'If they speak language marking factors such as solidarity, distance, inti- tinct), the everyday language

name, they must be —

can be completed using
'
macy, and formality is to switch from one language is known as Guwal, andthe
mother-in-law language is
geographical terms (p. 24), but social answers are variety to another. A Berlin businessman may use
called Dyalnguy. The latter
available as well: the blank can be filled by such standard German at the office and lapse into local would be used whenever a
phrases as 'my tribe', 'my religion', 'immigrants', dialect when he returns home. A conference lec- taboo relative was within ear-
turer in Paris may give a talk in formal French, shot. The two languages
'well educated', 'rich', 'servants', and 'the enemy'.
have virtually the same
The use of a different language is often a sign and then discuss the same points with colleagues
grammar, but no vocabulary
of a distinct religious or political group - as in in an informal variety. A London priest may give in common. Dyalnguy also

the cases of Basque, Latin, Welsh, the many official a sermonin an archaic, poetic style, and talk collo- has a much smaller vocabu-
languages of the Indian sub-continent, and the quially to his parishioners as they leave. During lary than Guwal.

the service, he might have used a modern English Guugu-Yimidhirr, there

In
pseudolinguistic speech known as glossolalia (p.
is no contact
at all with the
11). Switching from one language to another may translation of the Bible, or one which derives from
mother-in-law, and a strong
also be a signal of distance or solidarity in everyday the English of the 16th century. taboo also affects speech to
circumstances, as can be seen in strongly bilingual Languages have developed a wide range of var- brothers- and fathers-in-law.
ieties for handling the different kinds and levels There are important differ-
areas, such as Paraguay. Here, the choice of Span-
ences in vocabulary, style,
ish or Guarani is governed by a range of geographi- of relationship which identify the social structure
and prosody. Sexual topics
cal and social factors, among which intimacy and of a community. These varieties are discussed in are proscribed. One must
formality are particularly important. In one study other sections (§§11, 63), because they partly ref- speak to these relatives
lect such factors as occupation, subject matter, slowly, in a subdued tone,
(J. Rubin, 1968), bilingual people from Itapuami
social status, and setting; but it is important to without approaching closely
and Luque were asked which language they would
The style is
or facing them.
use in a variety of circumstances (e.g. with their note that they may also be used as symbols of social
sometimes described as
spouse, sweetheart, children, boss, doctor, priest, identity. In English, for example, forms such as dani-manaamaya, being
etc.) For most, Guarani was the language of inti- liveth and reigneth, givest, vouchsafe, and thine soft/ slow', or diili yirrgaalga,
have long been distinctive in one variety of religious speaking 'sideways'. (After J.
macy, indicating solidarity with the addressee. The
B. Haviland, 1979.)
use of Spanish would indicate that the speaker was language; but in the 1960s, as proposals for the
The avoidance languages
addressing a mere acquaintance or a stranger. modernization of Christian liturgical language of Australia illustrate yet
Spanish was also the language to use in more for- were debated, this variety came to be seen as a another means of marking
mal situations, such as patient—doctor, or student- symbol of traditional practice with which people social distance. The people
chose to identify or from which they dissociated turn away, linguistically and
teacher. Jokes would tend to be in Guarani. Court-
physically, from their taboo
ship often began in Spanish, and ended in Guarani. themselves. The case is worth citing because the
relatives. Similar taboos
The adoption of a local language as an emblem world-wide status of Christianity meant that many have also been observed in

of group identity is well illustrated by the Vaupes speech communities were involved, and over a many other parts of the
Indians of Colombia, who live in more than 20 quarter of the world's population was affected. No world, such as among the
Plains Indians of North
tribal units, each of which is identified by a separate other linguistic change can ever have raised such
America. These languages
language. Despite the existence of a lingua franca personal questions of linguistic identity on such can therefore be contrasted
(Tukano), a homogeneous culture throughout the a global scale. with those (in South-east

region, and the small numbers of speakers (around Asia, for example) where
social relations are expres-
5,000 in total, in the early 1960s), the Indians all
sed by adding complexity to
learn at least three languages - some, as many as ordinary speech (p. 40).
ten. The identity of the different languages is

42 • II LANGUAGE AND IDENTITY

Diglossia H form is often believed to be the more beautiful
Perhaps the clearest use of varieties as markers of and logical, and thus the more appropriate for reli-
in the case of diglossia — a lan-
gious expression — even if it is less intelligible. In
social structure is
Greece, there were serious riots in 1903, when the
guage situation in which two markedly divergent
varieties, each with its own set of social functions,
New Testament was translated into Dhimotiki.
coexist as standards throughout a community. One
And strong views are always expressed by Arabic
of these varieties is used (in many localized variant
speakers about Classical Arabic, which, as the lan-
forms) in ordinary conversation; the other variety guage of the Qur'an, belongs to God and heaven
is used for special purposes, primarily in formal
(p. 384).
speech and writing. It has become conventional in
Diglossic situationsbecome unstable in the face
linguistics to refer to the former variety as 'low'
of large-scale movements for a single standard —
such as might be found in programmes of political
(L), and the latter as 'high' (H).
unification, national identity, or literary reform. In
Diglossic situations are widespread, some of the
such circumstances, there are arguments in favour
better-known ones including Arabic, Modern
of either H or L varieties becoming the standard.
Greek, and Swiss German. These speech communi-
Supporters of H stress its link with the past, its
ties recognize the H/L distinction and have separ-
claimed excellence, and they contrast its unifying
ate names for the two varieties:
function with the diversity of local dialects. Sup-
High Low porters of L stress the need to have a standard
Greek Katharevousa Dhimotiki which is close to the everyday thoughts and feelings
(Demotic) of the people, and which is a more effective tool

Arabic
D
al-fusha
3
al-
c
ammiyyah of communication at 'Mixed' positions,
all levels.

(Classical) (Colloquial) setting up a modified H or L, are also supported;

Swiss German Hochdeutsch Schweitzer- and the steady emergence of L-based standards has
(High deutsch (Swiss been noted in Greece, China, Haiti, and several
German) German) other areas.

A personal column from

The functional distinction between H and L is the Basle daily newspaper
Perseenlig
generally clear-cut. H
is used in such contexts as Basler Ze/fung This item
sermons, lectures, speeches, news broadcasts, pro- shows an interesting con-
Herzlichen Dank
trast between High German
verbs, newspaper editorials, and traditional poetry. and Swiss German. The rest fur die vielen lieben
Aufmerksamkeiten zur
It is a language that has to be learned in school. of the newspaper is written in
L used in everyday conversation and discussion,
is High German, but in the Per- goldenen Hochzeit von
seenlig column (High Ger- Theres und Beat Wager-Biehler
radio 'soap operas', cartoon captions, folk litera-
man personlich), the last two Gross war die Freude uber die sehr
ture, and other informal contexts. einfallsreichen Uberraschungen!
items are entirely in Swiss
H and L varieties can display differences in pho- German (apart from the
Speziellen Dank den lieben Nachbarn
und <Ex-Nachbarn»!
nology, grammar, and vocabulary. For example, words in English). One is a Familie Wager
the sound systems of the two Swiss German var- humorous announcement of 65247 44-414946

the opening of a medical

ieties are strikingly different. Classical Arabic has
practice; the other is a birth-
three noun cases, whereas Colloquial Arabic has
day greeting. Besammlung
none. And in Greek there are many word pairs, Why are the remaining ads
Miinsterplatz
such as inos (H) and krasi (L) ('wine'): the word H not in Swiss German? This is

would be written on Greek menus, but diners probably because of their 17.00
content and level: the first
would ask for their wine using the L word. All beim Brunnen
item expresses the thanks of
three kinds of distinctiveness are illustrated in the an old married couple to their
following sentence given first in Hochdeutsch (H) neighbours for all they did at
and then in Schweitzerdeutsch (L): Nicht nur die their golden wedding celeb- Juhuill!
ration; the second an-
Sprache hat den Auslander verraten, sondern auch Kinder. d'Jugend derfe freue sich
alii
nounces the assembly point trotz Pflaschter. Impfig. Noodlestich
seine Gewohnheiten; and Niid nu s Muul had de dr Unggla Doggter
and time for a meeting of the
Usslander verraate, au syni Moodeli. 'It was not fire service association. Peter Gordon und sy Babbe
only his language that showed he was a foreigner, Even so, the second item harm ghisst d'PraxisEroffnigs-Flagge
zem grosse Anlass wmsche - mer nur's
his way of life showed it too.' (After P. Trudgill, has one distinctive feature: Bescht
Besammlung ('meeting') is - vill gsundi Kinder - und iedes Johr e
1983.) Fescht
an example of 'Swiss High
In diglossic situations, the choice of H vs L can German', midway between
Marguerite.
Primo und Jan
Shan

easilybecome an index of social solidarity. A Swiss High German (Versamm- 651908 03 383351

German speaker who used Hochdeutsch in every- lung) and Swiss German
(Besammlig).
day conversation would be considered snobbish or Happy birthday
artificial — and if the context were a political discus-
sion, could even raise questions of national
it dear Katrin!
loyalty, as Hochdeutsch is used as the everyday Alles Gueti wunscht Dir
language by people outside the country. Religious Dini liebi Familie

as well as political attitudes may be involved. The 651916 03-351574

10 SOCIAL IDENTITY 43
•
DIFFERENT WORDS AND PHRASES In addressing people whose names are known,
We recognize varieties of language as a result of kinship is a major criterion. If the speaker is related
perceiving several distinctive linguistic features to the addressee ('alter'), two factors are relevant:
being used together in a social situation. But often 'ascending generation' (e.g. aunt as opposed to cou-
a single linguistic feature is enough to indicate sin) and age. If the speaker is not related to alter,
social distance — such as the particular words or the factor of familiarity is relevant: whether or not
phrases used when people meet, address each other alter is a friend or colleague. If familiarity applies,
by name, or select pronouns for talking to or about the next factor is social rank, here defined with
each other. reference to a professional hierarchy. A senior alter
has the option of offering or accepting FN, instead
Modes of address of TLN ('dispensation' - Call me Mike), though
One of the most significant ways of signalling social this situation is often ambiguous. Age difference

intimacy and distance is through the use of a per- is not significant until there is a gap of nearly a
son's name in direct address. In English, the basic generation.
choice between first name (FN) or title with last
is

name (TLN), but several other conventions are

American address system (after S. Ervin-Tripp, 972)
possible in certain settings, such as the use of LN 1

only in business or academic settings {Now look

here, Smith .), or the use of abbreviations (Is JM
. .

in?). The range of possible forms is easy to state;

but the factors that govern the choice of forms are
often complex and difficult to summarize. When
would two people use FNs or TLNs reciprocally
to each other? When would one speaker use FN
and the other TLN?

Charting address relationships Several studies

have attempted to explicate these factors. The flow-
chart (right) was devised by Susan Ervin-Tripp
(1927-) as a means of specifying the factors that
condition a speaker's choice of address in American
English. The chart is simply a logical statement of
the various possibilities, given a context such as
'Look, — it's time to leave'; it is not an account
,

of what goes on in the speaker's mind. The know-

ledge structure represented is that of an American
academic; but dialect differences, idiosyncratic pre-
ferences, and other variants are not taken into ac-
count.
The entrance point to the diagram is at the bot-
tom left. Each path through the diagram leads to Nuer modes of address
one of the possible modes of address, listed verti- Address systems vary 'theone who goes ahead' their part. A man would
cally at the right. Alternative realizations of these greatly from culture to cul- and Duoth 'the one who normally be addressed us-
ture.Among the Nuer (Su- follows'. ing the name of his father
address modes are not given (e.g. a first name may man
dan), a system of multiple The social setting is an (his patronymic). But a
alternate with a nickname). For example, as one names and titles marks a important factor in the visiting maternal relatives
enters the diagram, the first choice which has to person's place in social selection of a mode of ad- will be greeted primarily by
be made is whether the addressee is a child structure. Every Nuer is dress. Every child inherits his mother's name (his
given a personal name, an honorific, or clan name, matronymic). The naming
(- Adult) or an adult (+ Adult). If the former,
shortly after birth, which he which tends to be used of people after their eldest
one follows the line downwards, where the only retains through life; but as only in ceremonies or on child (teknonymy) is also
distinction drawn is that between name known + ( ) an adult, it is used only by special occasions (such as heard, especially when
or not — ). If the child's name is known, one uses
( close relatives and friends. a return after a long ab- talking to in-laws. For ex-

the first name; if not, one does not use a name These names usually refer sence). When a boy is ample, a woman's status in
to the place of birth, or to initiated to manhood, he is her husband's home is
at all (0). The diagram does not give criteria for
given an ox, and from the based on her having borne
events that took place at
deciding when a child becomes an adult.
the time, such as Nhial distinctive features of this him a child, and this is the
Along the adult path, several decisions have to 'rain', Duob 'path'. Maternal animal he takes his 'ox- link that binds her to her

be made. 'Status-marked setting' refers to special grandparents often give the name', which is used only husband's social group. It
child a second personal by people of the same or is therefore natural for that
occasions (such as a courtroom) where forms of
name, which is used by similar ages. There are group to address her using
address are rigidly prescribed (e.g. your honour, kinsfolk on the maternal also dance-names' - more the child's name. (After
Mr Chairman). The 'identity set' refers to the list side. Twins are given elaborate versions of ox- E. E. Evans-Pritchard,
of occupational or courtesy titles that may be used special personal names, names that are used only 1948).
alone to mark social identity (e.g. Father, Doctor, which immediately identify at dances.
their status, such as Both Kinship roles also play
Mr, Miss).

44 II LANGUAGE AND IDENTITY

TorV? more likely to use T to fellow female students than
A well-studied example of address is the use of either French or Germans. There were psychologi-
the familiar and polite pronouns found in many cal as well as geographical differences. Radical stu-
languages, as in French tu/vous, German du/Sie, dents used more T forms than did conservatives.
Welsh ti/chwi, and so on. These forms (generally One of the conclusions of the study was that 'a
referred to as T forms and V forms, respectively, Frenchman could, with some confidence, infer that
from Latin tu and vos) follow a complex set of a male university student who regularly said T to
rules that foreigners never find easy to master. female fellow students would favour the national-
Terms such as 'familiar' and 'polite' capture aspects ization of industry, free love, trial marriage, the
of their use, but are inadequate summaries of all abolition of capital punishment, and the weakening
their social functions,and ignore important differ- of nationalistic and religious loyalties'. Inferences
ences between languages. of this kind are difficult to confirm on a larger scale,
In Latin, the T forms were used for addressing partly because of the speed of linguistic change
one person, and the V forms for more than one; (since the early 1960s, when this study was done,
but from around the 4th century bc, the convention student use of T has become much more wide-
developed of referring to the Roman Emperor using spread). But hypotheses of this kind are well worth
the plural form vos. Gradually, this 'royal you' following up, as they bear directly on the task of
extended to others who exercised power, so that establishing the basis of sociolinguistic identity.
by medieval times, the upper classes were showing (After R. Brown &
A. Gilman, 1968.)
mutual respect through the use of V forms only.
The historical picture is complicated and not en-
tirely understood, but medieval nobles would
generally address each other as V, whether talking
Flow-charts These charts provide an opportunity to make hypotheses about naming
to one person or more than one, and would address
and help to clarify interlanguage differences. For example, this kind
practice precise,
the lower classes as T. By contrast, the lower classes of diagram has been used to identify the factors governing the use of T or V forms
would use T to each other, and V to their superiors. in Yiddish (S. Ervin-Tripp, 1972).
Later the V forms began to be used in other cir-
cumstances, not simply as a mark of respect due
to those with power but as a sign of any kind of
social distance. T forms, correspondingly, began
to be used as markers of social closeness and inti-
macy. Thus, between equals, it became possible
to use either T or V, depending on the degree of
solidarity one wished to convey. Lower-class
friends would address each other as T, and use
V to strangers or acquaintances. Upper-class
people would do likewise.
In these circumstances, where there is a power
relationship motivating one usage (T = lack of res-
pect), and a solidarity relationship motivating
another (T = social closeness), situations of uncer-
tainty would often arise. For example, during a
meal, should diners address servants as T or V?
The diners are more 'powerful' (and so should use
T), but they are also socially distant from the ser-
vants (and so should use V). Similarly, should chil-
dren address their parents as T (because they are
intimates) or V (because there is a power differ-
ence)? By the 20th century, such conflicts had in
most cases been resolved by following the dictates Farr's Law of Mean Familiarity
of the solidarity dimension: these days, diners ad-
... asdiscovered by Lumer Farr, one of the senior
dress waiters as V, and children address parents
lifemen in Stephen Potter's One-upmanship
asT. naming
(1952), identifies a well-known inverse
But some fascinating differences remain. In the
relationship in the following way:
first systematic T/V study, male students from dif-

ferent linguistic backgrounds were asked about The Guv'nor addresses:

their pronoun preferences. The sample was rela-

Co-director Michael Yates as Mike
Assistant director Michael Yates as Michael
tively small, but it clearly emerged that Italians used
Sectional manager Michael Yates as Mr Yates
T more than the French, and the French more than Sectional assistant Michael Yates as Yates
the Germans. There were several interesting points
Indispensable secretary Michael Yates as Mr Yates
of detail: for example, Germans used T more to Apprentice Michael Yates as Michael
distant relations than did the French; Italians were Night-watchman Michael Yates as Mike

10 SOCIAL IDENTITY .
45
 .

Sexist language
Sexism
Maintaining sexual stereotypes in People would bring their
The relationship between language and sex has language wives, mothers, and chil-
dren.
attracted considerable attention in recent years,
This is the list of lecturers from the University of Rise Up, O Men of God . .

largely as a consequence of public concern over Reading's Department of Linguistic Science in 1983, Man, being a mammal,
male and female equality. In many countries, there as printed in the University calendar. Although gender breastfeeds his young.
is irrelevant to the job, the women in the Department
is now an awareness, which was lacking a gener- Mind that child - he may be
are clearly identified by the use of a full first name,
ation ago, of the way in which language can reflect deaf!
and/or by the use of Mrs. It is not possible to tell if Man overboard!
and help to maintain social attitudes towards men the male members of staff are married.
and women. The criticisms have been directed These randomly selected
almost exclusively at the linguistic biases that con- Lecturers: cases of sexibt language
stitute a male-orientated view of the world, foster- C. Biggs, MA, Oxford; PhD, Cambridge; may provoke ridicule, anger,
or indifference, but they
ing unfair sexual discrimination, and, it is argued, Diploma Cambridge
in Linguistics,
would be unlikely to warrant
leading to a denigration of the role of women in R. W. P. Brasington, MA, Oxford
a legal action to determine
society. English has received more discussion than A. R. Butcher, MA, Edinburgh; MPhil, their meaning. However,
any other language, largely because of the impact London; Dr phil, Kiel there are other examples
of early American feminism. F. Margaret Davison, BA, Sussex; MA, where a legal decision could
hang on the sex-specific vs
Several areas of grammar and vocabulary have Reading; Cert T Deaf, Manchester sex-neutral senses of man.
been cited. In grammar, the issue that has attracted P. J. Fletcher, BA, Oxford; MPhil, Read- In the U.S., for example,
most attention is the lack of a sex-neutral, third- ing;PhD, Alberta there has been legal contro-
person singular pronoun in English, especially in M. A. G. Garman, BA, Oxford; PhD, versy over the application of
its use after indefinite pronouns, e.g. If anyone Edinburgh; Diploma in General the generic male pronoun in
cases where it was disputed
wants a copy, he can have one. (In the plural, there Linguistics, Edinburgh
whether such phrases as 'a
is no problem, for they is available.) No natural- G. A. Hughes, BA, Montreal; Diploma in reasonable man' could legiti-
sounding option exists: one is considered very for- English as Second Language, Wales mately be applied to women.
mal, and forms such as he or she are stylistically K. Johnson, BA, Oxford; MA, Essex And in a case heard in 1977,
an appeal was made against
awkward. As a result, there have been many propo- Carolyn A. Letts (Mrs Letts), BA, Wales;
a woman's murder convic-
sals for the introduction of a new English sex-neu- MCST tion on the grounds that in-
tral pronoun — including tey, co, E, ne, thon, mon, K. M. Petyt, MA, Cambridge; MA, PhD, structions to the jury were
heesh, ho, hesh, et, hir, jhe, na, per, xe, po, and Reading; Diploma in Public and Social phrased using the generic
person. None of these proposals has attracted Administration, Oxford (Director of male form; this, it was
argued, could have biased
widespread support, but co, for example, has been Extramural and Continuing Education)
the jury's response, giving
used in some American communes, and na and per Marion E. Trim (Mrs Trim), MSc, Lon- them the impression that the
have been used by some novelists. Less radical don; LCST objective standard to be ap-
alternatives include advice to restructure sentences Irene P. Warburton (Mrs Warburton), BA, plied was that applicable to
Athens; PhD, Indiana an altercation between two
to avoid the use of he-torms.
men. Traditional safeguard
Many other examples of linguistic bias have been phrases such as 'the mascu-
given. In the lexicon, particular attention has been line pronoun shall import the

paid to the use of 'male' items in sex-neutral con- feminine' have turned out to

texts, such as man in generic phrases {the man in

be less than satisfactory in
resolving such issues.
the street, stone-age man, etc.), and the potential
for replacing it by genuinely neutral terms [chair-
man —» chairperson, salesman —* sales assistant, Sexual stereotyping has
etc.). Another lexical field that is considered proble- Sex-role stereotyping in schoolbooks
been especially noted in
matic is marital status, where bias is seen in such traditional children's read-

phrases as X's widow (but not usually Y's ing books and textbooks.
There were always more
widower), the practice of changing the woman's
male characters than fe-
surname at marriage, and the use of Mrs and Miss male, and they took part in
(hence the introduction of Ms as a neutral alterna- a greater variety of roles
tive). The extent of the bias is often remarked upon. and activities. In early read-
ing books, it was always
In one computer analysis of child school books,
the boys who were daring,
male pronouns were four times as common as the girls who were caring.
female pronouns. In another study, 220 terms were Pictures in science books
found in English for sexually promiscuous women, would show experiments
and only 22 for sexually promiscuous men. It is being conducted by boys,
while girls looked on. There
easy to see how sexual stereotypes would be rein-
is now a widespread trend
forced by differences of this kind. to avoid sex-role stereo-
types in children's books,
and to prepare children for
a more egalitarian society.

46-H LANGUAGE AND IDENTITY

THE PROPER STUDY OF MANKIND IS
MAN?
What has happened to sexist language, as a result
of feminist criticism? So far, the effect has been
far more noticeable in writing than in speech.
Several publishing companies have issued guide-
lines about ways of avoiding its use, and several
writers and editors, in many important areas, now
make a conscious effort to avoid unintentional
biases— including such well-known bodies as the
American Library Association, and writers such as
Dr Benjamin Spock and (for the record) the present
author. Legal changes, such as the Sex Discrimina-
tion Act in Britain (1975), have caused job titles
and much of the associated language to be altered.
But is there any evidence of a significant change
in practice throughout the language as a whole?
In 1984, an American study investigated the use
of man and its compounds to refer to all humans,
and the use of he and its inflected forms to refer
to females as well as males, in a selection of publica-
tions taken at intervals between 1971 and 1979.
The texts were samples of 75,000 running words
from American women's magazines, science maga-
zines, several newspapers, and both prepared and
spontaneous remarks from the Congressional
Record; a sample from The Times Literary Sup-
plement was used, as a British comparison. The
total sample was over half a million words.
The results were dramatic. In the American cor-
pus, the use of these forms fell from 12.3 per 5,000
words in 1971 to 4.3 per 5,000 in 1979. Women's
magazines showed the steepest decline, followed
A WERO 15 A MALE, ANP / UJHEN 1 CATCH THE BALL,
A HEROINE IS A FEMALE,.' iTLL BE THE HEROINE.'
by science magazines. By contrast, results for con-
gressmen showed no decline at all, and results for
congresswomen were mixed. There was no clear
decline in the British publication, but rates were
very low, and little can be deduced from such a
small sample. (After R. L. Cooper, 1984.)
What took the place of these forms? There was
no evidence that a straightforward replacement by
such forms as he or she was taking place. Rather,
it seems likely that people were using alternative
linguistic devices to get round the problem, such
as they along with a plural noun. (This is the solu-
tion I have found most congenial in the present
work, in fact.) Skar God
There is thus clear evidence that the feminist
movement had an observable impact in the 1970s
on several important genres of written language
— publications aimed at general audiences, not
/\f~G- boys
solely at women. Plainly, there has been a general
raising of consciousness about the issue of linguistic
he Her H Qt]
sexism, at least as regards the written language.
Whether this same consciousness would be found
qirh,£ ftnou,
in everyday speech is unclear, as is the question yoq are one
of how long-term these linguistic effects will be.
A
great deal of social change has taken place in
a decade, and could be enough to make the
this
i«f +17+0
associated linguistic changes permanent; but a
decade is as nothing within the large time-scale of
language change, and it remains to be seen whether
the new trends in usage will continue, or whether
there will be a reversal, with public opinion react-
ing against the extreme positions taken by some
Child's letter from Children's
militant feminists.
Letters to God.
100 DON'T KNOW
ANHTHING, PO HOO,
y
I I }< siBH^
CHARLIE BROWN?
J^V*^
10 SOCIAL IDENTITY 47 •
 . .

11 Contextual identity

The question 'Where are you from?', which signals

Setting
geographical identity (§8), can be balanced by
another locational question, 'Where are you now?' The particular time and place in which people inter-
Many features of language correlate directly with act will exercise its influence on the kind of com-
the characteristics of the context, or situation, in munication may occur - or whether
that
which a communicative event takes place. Classifi- communication permitted at all. In institution-
is

cations vary, but most approaches recognize the alized settings, such as a church or a court of law,
central role played by the following factors: the effect on language use is clear enough. But in

• Setting. The time and place in which a

many everyday situations, and especially in cul-
communicative act occurs, e.g. in church, during a
tures we find alien, the relationship between setting
meeting, at a distance, and upon leave-taking. and language can be very difficult to discover. At
• Participants. The number of people who take part dinner parties, funerals, interviews, council meet-
in an interaction, and the relationships between them, ings, weddings, and on other occasions, linguistic
e.g. addressee(s), bystander(s). norms of behaviour need to be intuitively recog-
• Activity. The type of activity in which a participant nized if people are to act appropriately, but they
is engaged, e.g. cross-examining, debating, having a are not always easy to define. For example, how
conversation. would one begin to define the optimum length of
The interaction between these factors produces a an after-dinner speech, or the proportion of
set of constraints on several features of language humour its subject matter should contain? In differ-
(discussed in Parts iii-vi, and x), notably:
ent times and places we may be obliged, permitted,
encouraged, or even forbidden to communicate;
• Channel. The medium chosen for the
and the quality or quantity of the language we use
communication (e.g. speaking, writing, drumming) and
will be subject to social evaluation and sanction.
the way it is used.
• Code. The formal systems of communication
The extent to which people recognize, submit to,
shared by the participants (e.g. spoken English, Russian, or defy these sanctions is an important factor in
etc., deaf sign languages). any study of contextual identity.
• Message form. The structural patterns that identify
the communication, both small scale (the choice of
specific sounds, words, or grammatical constructions)
and large scale (the choice of specific genres).
• Subject matter. The content of the communication, How to answer the telephone
both explicit and implicit.
Telephone conversations following practice seems to irritated when a French
provide one of the clearest be more usual (after D. checks their number,
Each of these plays a crucial part in the identifica- caller
examples of the influence Godard, 1977): when they themselves
tion of a communicative event. For example, a ser-
of setting upon language, 1 Telephone rings. have just said it. Or again,
mon is normally given in a church
(activity) because of the lack of 2. Answerer: Alio.' in trying to reach a third
(setting),by a preacher addressing a congregation visual feedback, and the 3. Caller verifies number. party, a French caller
(participants), primarily using speech (medium), in constraints of time and 4. Answerer: Oui.' would expect French
money. The opening and 5. Caller identifies self, answerers to reciprocate
a monologue in a single language (code), involving
closing phases of such apologizes, and asks for in- with a self-identification or
religious forms and genres (message form), and conversations are particu- tended addressee. some degree of small talk,
about a spiritual topic (subject matter). This kind larly distinctive, with rules The different conventions before going to get the third
of characterization needs immediate refinement, of governing sequences of can have several con- party, whereas an English
course. Some sermons permit dialogue as well as acceptable and unaccept- sequences - not least, the answerer would have no
able utterances. Certain has been
possibility (which such expectation. The
monologue; some use chant and song alongside features of the language seriously mooted) that sequence:
speech; some introduce different languages. But an are universal, but there are French people have 1 Telephone rings.
initial simplified analysis is useful, because it also interesting cultural dif- greater difficulty remem- 2. Answerer gives number.
enables a comparison to be made between different ferences, which often make bering their own telephone 3. Caller asks for third party.
themselves felt whenever number, because they do 4. Answerer: get her.'
kinds of communicative event, which points the I'll

one attempts telephone

to not have to verify it them- (Leaves phone.)
way towards a typology of communication. Several someone abroad. selves when they pick up is normal in England, but
contextually distinctive uses of language are illus- In British English, for ex- their phone! An English abnormal in France, where
trated in §63. ample, the normal se- caller in France could un- there would be a further
quence for a call to a pri- intentionally offend, by us- interaction before the
vate residence is as follows: ing the British pattern, answerer left the phone.
1. Telephone rings. which lacks the caller's Several such differences
2. Answerer gives number. self-identification and apo- exist,which, if not correctly
3. Caller asks for intended logy for troubling the understood, can easily lead
addressee. answerer. And, conversely, to unfortunate stereotypes
By contrast, in French, the English answerers can be about foreign attitudes.

48 II LANGUAGE AND IDENTITY

Maori greetings chant of welcome, using A Maori karanga exchange
In some cultures, rituals of
rhythmical actions and loud LOCAL: Haere mai ra e te mana ariki e, mauria mai o taatou tini aituaa!

greeting or leave-taking are

shouts. Welcome, prestige of chiefs, bring our many dead!
• The tangi is a high wail- Haere mai, haere mai!
marked by elaborate and
highly conventionalized
ing and sobbing, on a sin- Welcome, welcome!
forms of expression, often
gle vowel, uttered for the VISITOR: Karanga ra te tupuna whare ki te kaahuipani!
reflecting the social stand-
dead. Call, ancestral house, to those who mourn!
• The whaikoorero is the Ki ngaa iwi e, karanga ra!
ing of the speakers (§10).
oratory that is the main part Call to the tribes!
Among the Maori, for ex-
ample, distinctive beha-
of the ritual. The locals and LOCAL: Nau mai ngaa karanga maha o te motu!
viour and language identify
each group of visitors have Draw near from all corners of the island!
the ritual encounter at the
a 'team' of orators. Mauria mai ngaa mate kua ngaro ki te poo!
Speeches alternate, each Bring the dead who have gone into the night!
beginning of the ceremo-
nial gathering (or hui)
speech beginning with a VISITOR: Hoki wairua mai raa e koro e!
warning shout, and being Return in spirit, old man!
which takes place on such
followed by an archaic Ki te karanga ki te poowhiri i taa koutou kaahui pani!
occasions as weddings,
funerals, and visitations by
chant, greetings for the To the call and welcome of those who mourn you!
dignitaries.
dead and perhaps a
living, Hoki wairua mai e Paa e!
There may be as many topic for discussion,and Return in spirit, father!
concluding with a traditio-
as seven stages in the en-
nal song by the group as a
counter ritual, all but two in-
whole.
volving language. In each
• The hongi, or pressing of
case, accuracy of expres-
noses, concludes the ritual.
sion is essential, otherwise
evil will result. (After A. Sal-
mond, 1974): When high-ranking for-
• The waerea is a protec- eigners make an official
visit to New Zealand, they
tive incantation chanted
are usually greeted by the
upon entry to a gathering.
elaborate leaping and gri-
Its words are archaic, and

are often not understood.

• The wero is a ritual
lenge, involving noise and
actions, but no language.
• The karanga is an ex-
change of high, chanted
calls of greeting, and invo-
cations to the dead, be-
tween the old women of the
and visiting parties.
local
• The poowhiri is an action
macing of a Maori ceremo-
nial challenge {wero). Such
chal-
ritual displays of strength
were always customary on
thefirst encounter with
strangers- though early
settlers often took them for
displays of real belliger-
ence, with deadly results!
Maori dancers in ceremo-
nial costume, 1972.

Speech-making in Samoan
A study of formal speech-making in the village of
Falefa, in Western Samoa, provides a good illustration
of the effect of setting on language. The village council
(fono) consists of around 100 adults (matai), who are
chiefs and orators, all with special titles. Meetings of
the fono are called to discuss crises in village life; but
before the main issue is discussed, orators make one
or more formal speeches (lauga). The lauga seems to
function as an affirmation of the need for a stable
society, at a time when conflict and dissent are
present. It contains seven distinct parts (though these As an example of the speech style, part of the Mornings section of one lauga is given below
may be reduced in number, and their performance (from A. Duranti, 1983):
varies from one type of social event to another).
• Kava: an acknowledgement of the person who has . . . O iku
kaeao i . . Moving on to the mornings, . .

called out the titles of those who were served kava kaeao masagi lava
la 'o well (they) are very well-known mornings
roots in the opening ceremony. oleaukugu'u of our country
• Thanksgiving to God: for allowing the people to kaeao (o) le Loku the morning of the Church
gather in this way. ma kaeao- Kusi Pa'iale and the morning (of) the Bible
• Mornings: a metaphor for important events, which la ... o kaeao lava . . Yes . real mornings
. . . .

symbolizes the performing of good deeds, and Ua kuaga'i ia kaeao Those mornings have gone
focusses attention on the present meeting. ma kaeao- fo'i sa fa' asilisiliga and the mornings that have been indicated
• Dignity of the sacred names: an acknowledgement io(u)kou figagalo by the wish of you (chiefs)
of the dignity of the matai and their titles. ma o kakou fa'amoemoe . . and the hope of us (orators) . .

• Formal greeting: praise and greeting for all the matai la 'ae o le kaeao sili a legei well this is the most important morning
titles. ua kakou aulia maguia when we meet in good spirit
• Agenda of the fono: the official reason for the legei kaeao fou (on) this new morning
meeting, stated in very general terms. ma legei aso fou . . and this new day . .

• Clearing of the sky: the speaker wishes a good and fa'akaugu'uiga ai to accomplish
long life to all present, using this metaphor, which le- le kofa ma le fa' aukaga. the decision of the chiefs and of the orators.
represents a life with no problems.

1 1 CONTEXTUAL IDENTITY 49
 .

Participants Cuna curing ritual

The simple opposition of message 'sender' and mes- Some cultures introduce to carry out the curing, and or plants - or even western
unusual participants into someone who knows the medicines- can be ad-
sage 'receiver' needs considerable refinement if we
the speech event, especi- appropriate ikar is used to dressed in this way, as can
are to classify communicative events satisfactorily. ally where special powers instruct them. The sick are be seen in the following ex-
Normally a single person acts as sender, or addres- need to be invoked. For ex- not themselves active parti- tract (from J. Sherzer,
sor; but we have to allow for unison speech, as ample, in Cuna (Panama), cipants in the event: they 1974), in which the ikar
church or other
in the case of liturgical responses in language combines with may be asleep, or unaware calls on certain trees to use
medicine to help cure dis- of what is taking place their strength to help some-
group teaching (where the whole class may
rituals,
ease. There are many around them, or even ab- one suffering from severe
respond together), popular acclamations (such as speech forms (ikai) that are sent.Nor would they (or headaches. The italicized
during a political address, or in a sports arena), thought to effect cures. other onlookers) usually words are used only in this
and speeches by the players in a theatrical presen- These vary according to understand the special lan- particular type of ikar,
the nature of the ailment, guage of the ikar. In sev- whose fixed pattern must
tation. The linguistic characteristics of such speech
but all have the same basic eral ways, this ritual is simi- be accurately repeated in
(especially the prosody (§29)) will obviously be structure. lar to the western religious order to be effective. As
very different from those found when a person The sick person lies on a tradition of praying over or curing /Tcargenerally last for
speaks alone. hammock, under which is a for the sick. about 1| hours, those who
Similarly, a single person is the usual receiver, box of wooden dolls; is it Other objects of medici- speak them need to have
the dolls which are thought nal value,such as tree bark great powers of memory.
or addressee, of a message; but here too we must
allow for variations. We may address someone di-
rectly, or through an intermediary, such as a secre- kurkin /pekantiye olopillise pupawalakan akkuek"iciye,
tary, interpreter, or spokesman. A third party may kurkin /pekantiye olopillise pe maliwaskakan upoek"iciye,
overhear what we are saying, or see what we have kurkin /pekantinaye olopillise pe maliwaskakana pioklekeklciye,
kurkin /pekantinaye olopillipiye apikaek w iciye
written, and we may consider this desirable or
. .

'trees, your roots reach the level of gold,

undesirable. And speech addressed to a group of trees, your small roots are placed into the level of gold,
people is common enough in everyday conversa- trees, your small roots are nailed into the level of gold,
.'

tion, as well as in more formal contexts, such as trees, within the very level of gold you are resisting . .

sermons, toasts, and lectures, and the whole range

of circumstances that define the world of spoken
and written mass communication (§63).
All of these contexts can influence the language
used by the speaker. For example, to know that
one is being overheard by one's superior can lead
to marked alterations in speech, even to the extent
of adopting a completely different stylistic level (as
has been observed in Persian). One may need to
defer to the broader audience by altering pronoun
forms and using various politeness strategies, as
well as by modifying non-linguistic behaviour
(such as body movements and eye contact). In some
circumstances, the knowledge that one is being (or
even, is likely to be) overheard may lead to non-
fluency or a breakdown in communication,
patient—doctor conversation, or the well-known
effects that take place when people are asked to
speak into a microphone.
In multilingual environments, there will usually
be language switching (p. 363) when a conversa-
tion is joined by a third party who is not at ease
in the language being used. However, language
switching may not take place if the participants
wish to exclude the third party - a common reac-
tion to tourists visiting rural communities abroad.
Nevertheless, circumstances vary greatly, and reac-
tions are difficult to predict. One empirical study
encountered a group of bilinguals at an inn in Aus-
tria switched from Hungarian to German
who
when asked to do so by people at a nearby table;
but the study of a similar situation in Scotland
found that a request to switch from Gaelic to Eng-
lish was refused.
READING THE NEWS
The effects of a mass audience on speaker style
can be illustrated from studies of variation in the
speech of radio newscasters. In New Zealand, for
example, the same group of newscasters read the
news on a number of different radio stations that
share the same suite of studios. In one study, indivi-
dual newscasters were monitored when they read
the news on a higher-status station (YA) and on
a lower-status station (ZB). In every case their pro-
nunciation changed in the same direction. For
example, /t/ between vowels (as in butter) was pro-
as in duced with voicing far more on ZB than on YA,
as shown in the diagram (after A. Bell, 1984):
50% -
YA ZB YA ZB YA ZB YA ZB
Because the voiced variant is normal in New Zea-
land, the announcers' use of the alternative must
be demonstrating the external influence of a model
of acceptable public speech (in this case, Received
Pronunciation) — an instance of positive accommo-
dation to an audience (p. 51).

50 •
II LANGUAGE AND IDENTITY
NEW LANGUAGE FOR OLD
Varieties of language can alter completely if there
is a change in one's view of audience needs. In
recent years, for example, there has been a radical
shift in the way theologians have begun to talk
about God, in the light of their perception that
people have become dissatisfied with traditional
images and are searching for new ones. Such images
covered a wide area of language, including terms
that were highly abstract and mystical (supreme
being, infinite one, the unknowable, essence), meta-
phorical and personal (father, lord, judge, saviour),
psychological and ethical (forgiveness, love, com-
passion).
The dissatisfaction is well illustrated by the suc-
cess of Bishop John Robinson's Honest to God
(1963), which sold over a million copies. This book
questioned the tradition of talking about God in
crude spatial metaphors, as if he were 'up there',
or 'out there'. It argued that, to modern audiences,
such language was outmoded and acted as a barrier
to understanding, whereas images such as 'ground
of our being' could more easily be related to current
ways of thinking. Several experiments in religious
communication followed, in the spirit of this
approach, and a new academic discipline has even
been proposed to study this area - theography, a
term coined on analogy with 'geography', which
aims to 'draw the map' of language that people
use to talk about God.
LINGUISTIC ACCOMMODATION
When two people with different social back-
grounds meet, there is a tendency for their speech
to alter, so that they become more alike - a process
known as accommodation, or convergence. Modi-
fications have been observed in several areas
of language, including grammar, vocabulary, pro-
nunciation, speech rate, use of pause, and utterance
length. Everyday examples are the slower and
simpler speech used in talking to foreigners or
young children; the way technical information is
presented in a less complex manner to those who
lack the appropriate background; the rapid de-
velopment of catch phrases within a social group;
and the way many people cannot stop themselves
unconsciously picking up the accent of the person
they are talking to. The process has even been
observed with babies 'talking' to adults: at 12
months, they were babbling at a lower pitch in
the presence of their fathers, and at a higher pitch
with their mothers.
These shifts take place in order to reduce the
differences between participants, thus facilitating
interaction, and obtaining the listener's social
approval (p. 23). It should be noted that linguistic
accommodation also has its risks, such as the loss
of personal (and sometimes group) identity, or the
perceived loss of integrity, such that the listener
may react against the speaker's new style. Much
depends on how speakers view themselves and the
group to which they belong (the 'in group') in com-
God is for real, man
This is the title of a book by Carl Burke, an American
prison chaplain, who hoped to make the biblical
message meaningful to people from New York's
toughest areas, by 'translating' passages into their
everyday style of speech. The first three
commandments read:
You shall have no other gods before me... Means
1
God's the leader- nobody, but nobody, man, gets in
the way. This is the top. He is Mr. Big, real big.
2. You shall not make for yourself a graven image . .
This means no making things that look like God in the
craftshop at the settlement house. No worshipping
things like rabbits' foots and lucky dice and, damn it,
dolls.
3. You shall not take the name of the Lord your God
in vain ... It means knock off the swearing or you better
watch out.
parison with the group to which the listener
belongs (the 'out group'). But on the whole, the
benefits of convergence seem to outweigh these
risks, with several social psychological studies
showing that people react more favourably to those
who move linguistically closer to them.
Divergence
Speech divergence also takes place when people
wish to emphasize their personal, social, religious,
or other identity. There may be quite elementary
reasons for the divergence, such as a dislike of the
listener's appearance or behaviour; or there may
be more deep-rooted reasons, such as the deliberate
use of a minority language or ethnically distinctive
accent or dialect (§9). Threatening contexts readily
result in divergence, as has been demonstrated
experimentally. In one study, a group of people
in Wales were learning Welsh in a language labora-
tory. During one of the sessions, they were asked
to answer some questions about language learning.
The questions were presented to them in their indi-
vidual booths by an English speaker with an RP
accent (p. 39), who at one point arrogantly chal-
lenged their reasons for learning what he called
'a dying language with a dismal future'. The accents
used in their replies were then compared with those
used in responding to a previous question that
was emotionally neutral. The test sentence replies
showed immediate divergence (as well as an aggres-
sive tone of voice): speakers used a broader Welsh
accent, and some introduced Welsh words into
their speech. In a similar study, in Belgium (p. 37),
the divergence took the form of a complete lan-
guage shift. Here, the aggressive question was
spoken by an unsympathetic Walloon (French)
speaker to Flemish learners of English. Although
replies to other questions were in English, half the
learners switched into Flemish in their replies to
the question which threatened their ethnic identity.
11 CONTEXTUAL IDENTITY • 51
Activity
quently requiring the application of different Some English
grammatical rules from those found in good
politeness formulae
The kind of activity in which we engage will di- writing or recommended by traditional gram-
Greetings Good morning,
rectly influence the way we communicate. At one mars (§1).
Hello, Hi
level, our activities reflect the social status we have
• The vocabulary of everyday speech tends to be Farewells Good night, Bye,
and the roles we perform
(§10). But status and informal and domestic, limited and inexplicit, See you, Cheers
as speakers cope with difficulties of memory, Introductions How do you
role are very general notions, within which it is
do?, How's things? Hi
possible to recognize a much more specific notion attention, and perception. In extreme (though
Thanks Thank you, Ta,
of 'activity type'. For example, priests have a well- not uncommon) cases, empty nonsense words Thanks a lot
defined status and role within a community; but may be used, e.g. tbingummajig, whatcha- Toasts Good health,
while exercising their role as priests, they engage macallit, doo-da. Cheers, Here's to...
• There is a great deal of usage variation on the Seasonal greetings Merry
in a wide range of activities, such as leading a ser-
Christmas, Happy Birthday
vice, giving a sermon, exorcizing spirits, hearing part of individual speakers, often involving the
Apologies Sorry, beg your I

confession, baptizing, and visiting the sick. Many unconscious use of non-standard or deviant pardon, My mistake
other occupations involve a similar variety; and forms. Responses to apologies
That's OK, Don't mention it,
in all cases there are linguistic consequences of the Certain other features of this activity are Never mind
shift from one activity to another. Linguistically- included in §31. The subject of 'conversation Congratulations Well done,
distinct activities are often referred to as genres Right on, Congratulations
analysis', which deals with the rules governing
or registers, though these terms are sometimes used turn-taking between speakers, is introduced in §20.
Public noises Encore, Hear
to refer to all the contextually influenced varieties hear, Goal
Body noises Excuse me,
presented in this section. Courtesy expressions Bless you, Pardon me
Activity influence is not restricted to occu- Ritual expressions of politeness are a common fea-
pational environments. We also engage in many ture of social interaction in all forms of spoken
kinds of activity in everyday speech and writing, and written dialogue, but especially in conversa- An Arabic farewell
such as gossiping, discussing, quarrelling, petition- tion. They are of considerable importance in The normal exchange of fare-
ing, visiting, telephoning, and writing out lists. accounting for the way people judge each other, wells in Syrian Arabic is a
Here too there are linguistic norms and conven- three-part sequence. If A is
and in explaining the success or failure of an inter-
said first, the addressee
tions, although they are usually more flexible, and action. The omission of a politeness formula, when must reply with B, and the
the genres are not always as easy to define as those one is expected, or the failure to acknowledge one speaker may then use C;
first
associated with more formal activities. appropriately, can lead to a tense atmosphere, or but if B is said first, C is obli-
even social sanctions — as children who fail to say gatory.

please sometimes find to their cost. In some lan-

SPOKEN VARIETIES guages, complex formulaic politeness sequences
A. (b)xatrak by your leave'
B. mafssa/ame 'with peace'
reflect levels of social structure and long-standing C. c
alla ysallmak God keep
Conversation
social traditions, as in the case of Wolof or Maori you'
Everyday conversation is so habitual that it is easy
greetings (pp. 40, 49). English has only a small
to forget its status as a genre, with its own norms This language also illustrates
number of expressions, by comparison. the principle of replying to
and conventions, often very different from those
Languages display many differences in politeness greetings by adding' to the
used in written language (§31).
expression. For example, phrases such as good original, as in

• The language is often inexplicit, because the par- morning and good evening are by no means univer-
A. marflaba 'hello'
ticipants can rely on context to clarify their sal: salutations related to time of day are normal
B. marttabten 'two hellos'
meaning, e.g. A: That's a nice one. B: It sure in many languages, but not, say, in Bengali or or mit marfiaba "1 00 hellos'.
is. Wolof; and the distinctions found in English are
• There is no careful thematic planning governing lacking in French (which uses one expression, bon- The Qur'an in fact says at
one point (Surah IV, verse
the way a conversation proceeds; there are often jour, more widely). Foreigners do not always find
86): someone greets you,
If
changes of subject matter, and alterations in it easy to work out the pragmatic rules that govern either return the greeting or
level (even, in multicultural contexts, switching the use of these expressions, for arbitrary conven- greet him better, for God
between dialects or languages, p. 363). tions are often involved. For example, the 'morn- takes everything into ac-
• A degree of non-fluency is normal, while partici- ing' in good morning does not coincide with the count.' (After C. A. Ferguson,
1976.)
pants spontaneously construct their sentences; chronological period from midnight to noon: in
one expects to hear false starts, hesitation noises normal use, it does not extend from midnight, but
{er, um), pauses, repetitions, and other 'errors' only from waking up; and it may extend beyond
of performance. midday, until the midday meal. Outside of this
• Speech is usually quite rapid, with many of the period, its use is ironic, as when it is said late at
sounds of careful pronunciation being omitted night to someone who was expected earlier, or said
or altered in the interests of preserving natural- mid-afternoon to someone who has overslept.
ness and fluency; a wide range of prosodic effects Moreover, it may be used only once to a person
(§29) is heard, signalling the diverse emotions during the day (unlike Hello), and an echo of the
which are encountered in conversations. greeting is expected (unlike Thank you). But good
• The clear-cut sentence patterns known from the morning is simple compared with good evening,
written language are often missing; in their place where use is affected by variations in social back-
are more loosely-connected constructions, fre- ground, habits of work, and the onset of darkness.

52 •
II LANGUAGE AND IDENTITY
PROVERBIAL EXPRESSIONS to induce intimacy Slang samples
In every culture there are nuggets of popular wis- to show that one belongs Cockney rhyming slang
dom, expressed in the form of succinct sayings. to exclude others Cain and Abel table
to be secret cows and kisses the missus
These are usually referred to as proverbs, though
Gawd forbids kids
several other terms are also used (e.g. adage, Hampstead Heath teeth
maxim, precept). Proverbs are not commonly But one theme recurs among all these reasons: the lean and lurch church
encountered in everyday speech in English, but in use of slang as a means of marking social or linguis-
many cultures (e.g. in most parts of Africa), they tic identity. In Partridge's book Slang: Today and U.S. hospital slang
crispy critter severe burn
are an important and frequent element in ordinary Yesterday (1933), the group-identifying function
patient
conversation. in fact provides the basis for most of the detailed pre-stiff close to death
Several extensive collections of proverbs have illustrations, which come from a wide range of geo- prune old, dehydrated
been made, which provide evidence for consider- graphical areas and occupational activities. Slang patient

— Zorro belly someone with

able similarities across cultures similarities that is, by definition, a colloquial departure from stan-
surgical scars on abdomen
are largely due to the universality of human experi- dard usage; it is often imaginative, vivid, and (From D. P. Gordon, 1983.)
ence (though there are often signs of linguistic bor- ingenious in its construction — so much so that it
rowing). For example, many languages have has been called the 'plain man's poetry'. It thus British prison slang
parallels for such proverbs as the Somali Kaadsade especially attracts those who, for reasons of person- filth detectives
LTI long-term inmate
ma kufo 'He who takes his time does not fall.' ality or social identity, wish to be linguistically dif-
nick prison
Structurally, also, proverbs display interlanguage ferent — to be 'one of the gang', whether the 'gang' screw prison warder
similarities with their reliance on vivid images, in question be soldiers, nurses, actors, footballers, snout tobacco
domestic allusions, and word play. One of the most prisoners, warders, linguists, gays, or pop singers
But remember
interesting features is the way many can be divided (see also pp. 56, 59).
. .

The slang of one generation

into two parts that balance each other, often dis- can be the standard English
playing parallel syntax and rhythm, and links of of the next:
rhyme and alliteration. bus from omnibus
zoo from zoological garden
• English: Least said, soonest mended. piano from pianoforte
• Maori: Ka whakaiti koe i te manuhiri, ka whak-
aiti koe i a koe. 'In demeaning the visitor, you
From speech to poetry
lower yourself.'
In many speech situations. mulae is clear. per second) only occasion-
• Latin: Praemonitus, praemunitus. 'Forewarned contextual factors combine Keep your hand in God's allyexceeds that found in
forearmed.'
is with the skill of the speaker hand. a normal conversation on a
• Somali: Beeni marka bore waa malab, marka to produce genres that dis- And your eyes on the star- familiar topic. The special

dambe na waa malmal. 'Lies are honey at first, play many of the character- posts in glory. prosody has been studied,
istics of poetry. The main Lord said he would fight for example, in the mono-
later they are myrrh.' logues of New Zealand
comparison is with the your battles,
• Chinese: di wu ji wu.
'if you love a house, you techniques used in the oral If you'd only be still. livestock auctioneers. Dur-
love its 'Love me, love my dog.')
crows.' (cf. formulaic poetry of early You may not be a florist. ing the opening phase of

• Samoan: E mafuli le ului, ae tupu le suli. 'The European culture (in the Am 1 right about it? the auction, the stock is
Homeric epics, in particu- But you must tell them, that described using a loud,
parent tree has fallen over, but one of its saplings
lar), and still found earlier He's the Rose of Sharon. high-pitched drone. When
is growing.'
this century in the singing 1 know that's right. the bids begin, many of the
• Welsh: Cenedl heb iaith, cenedl heb galon. 'A of oral epics by the Serbo- You may not be a geologist. speaker's rhythm units
nation without a language is a nation without Croatian guslars. The But you must tell them, that start with a stylized shout.
a heart.' rhythm and intonation He's the Rock of Ages. The last bid generally has
changes from that of nor- 1 know that's right. its own tune- a prosodic

mal speech, so that proso- You may not be a physician. warning that the auction is
THE CHIEF USE OF SLANG . . dic lines' can be heard. But you must tell them, that about to end. Then the
Is to show one of the gang! In fact,
that you're The speech contains many He's the Great Physician. gavel falls, and the auction-
slang has so many uses that it is difficult to choose memorized formulae, Am 1 right about it? eer's speech returns to a
which can be embellished You may not be a baker. normal mode. Extracts
one as central. Eric Partridge (1894-1979) was
or modified as occasion But you must tell them, that from one auction illustrate
able to distinguish as many as 15 different reasons arises. He 's the Bread of Life. its formulaic character
for the use of slang: Am 1 right about it?... (from K. Kuiper & D. Hag-
Sermons go, 1984):
An example of this inter- Auctioneer speech What do you think, Sir?
for the fun of it mediate stage between Auctioneers all over the Sell 'em Sir?

as an exercise in wit or ingenuity speech and oral poetry is in world impress lay audi- Are they on the market, Sir? . .

the spontaneous sermons ences with their fluent ver- I'll sell 'em.
to be different
of black preachers in the bal skills; but much of their Right, I'll sell 'em.
to be picturesque southern United States. performance is based on Right, I'll sell.
to be arresting The text below has been the use of linguistic formu- We'll sell 'em.
to escape from cliches transcribed in lines (from lae, uttered in a distinctive Right, we'll sell 'em.
B.A.Rosenberg, 1970), prosodicform. Repeated I'm gonna sell 'em
to enrich the language . .

identified by the preacher's phrases and an absence of 1 got twenty dollar twenty
to add concreteness to speech own rhythms and the oral pauses contribute to an im- bid twenty bid twenty got
to reduce seriousness response of the congrega- pression of rapid speech - twenty bid forty twenty
to be colloquial tion (Amen, Hallelujah, though in fact their speed dollar forty twenty forty I'm

for ease of social interaction etc.). The heavy use of for- (as measured in syllables bid...

11 CONTEXTUAL IDENTITY • 53
VISUAL VARIETIES
Ethnographers of communication have largely of stylistically distinct varieties available in a liter-
focused on the study of speech events in a wide ate culture, there space here to select only a tiny
is

range of cultural contexts, because of a previous fraction, in order to illustrate the range of vari-
lack of research in these domains. But it must not ability and some of the linguistic features involved.
be forgotten that the various activities of the writ- Further illustration of contextually influenced
ten language also display the influence of context visual language activities will be found in §63, in
— often in a highly distinctive manner, because relation to advertising, law, science, the press, and
of the visual contrasts available in the written other specialized fields, and in §32, in relation to
medium, especially in print (§31). With thousands the field of typography.

Information materials example, constitute a clearly identifiable linguistic

This is an enormous field, including works of refer- variety but one that is made up of a large number
ence (dictionaries, catalogues, almanacs, govern- of 'sub-varieties' such as news reports, letters, edi-
ment leaflets, etc.), instructional material (phrase torials, and crosswords (p. 388). But at the opposite
books, recipe books, do-it-yourself manuals, etc.), extreme, there are many informational linguistic
newspapers, documents, reports, teletext, and all activities that are limited in scope and fairly homo-
kinds of academic publication. Some of these geneous in content and structure - some so severely
materials are so wide ranging and diverse that it constrained that they fall within the category of
is impossible to make simple generalizations about 'restricted language' (p. 56) - as with cooking
their linguistic distinctiveness. Newspapers, for recipes, phrase books, and commercial advertising.

CREOLE OYSTER Cl'StBO

i 4 pound toting chicken i small can tomatoes ;
my medi- Prosze. mi (lac mo] proshen mi dahch
: of ham i teaspoon cum powder lekarstvxu moie lehkahrstvcj ~
6 tablespoons butter Salt and ground pepper
freshly Wh»t is the matter Co panu brakoje 1 co pahnoo brahko*.- Revcf-
5 tabtopoons Sour W teaspoon, red pepper with you ? ieh 'tn iiKii ti:ni
; cups chopped onions 2 teaspoons each tin me and basil Is your digestion Czy trawUz pan chi frahvish pahn
i Urge green pepper, sliced i pound fresh shrimp all right dobrze ''

dobzheh
2 clones garlic, crushed i poond ofal Are you cold ?
t cup sliced celery and leaves
Czy panu zimno ? chi pahnoo zimno
Are you warm ? Czy panu goraco ? chi pahnoo goronco
2 quarts chicken stock i tablespoon gumbo file
Cover yourself well Prosze. si^ przvkryc proshen sien pzhi-
Be to get a young chicken. Quarter the chicken and cot the ham into dobrze krich dobzheh
Vi-inch cubes Melt 3 tablespoons of the butter in a heaw pan and heat Where do you feel Gdzie pan czujesz gdzieh pahn chuu-
the flour in it brown without burning it Put the browned flour
to a dark pain now > bol teraz ? iesh br561 tehrahz
tn a large pot.nnsing out the browning pan with a little stock. Put m the How long hare you yahk dwoogo biw
Jak dlugo byt pan
chicken arid ham. Saute the onions and pepper in the other 3 tablespoons been ill
>
chorv
pahn horif

of butter for 5 minutes, and add them to the pot with the garlic, celery You ha»e Masz pan
a, slight mahsh pahn lehkki
Ickki atak
and leaves, sliced very thin, the chicken stock, tomatoes, curry powder. attack of ferer goraczki ahtahk goronchki
salt, pepper, thyme, and basil Simmer iVi hours. Meanwhile wash the
Can you account Czy mow mi pan chi mozheh mi pahn
shnmp and cook 4 minutes in 1 cup of water. Let cool. peel, and set aside.
for it dac preyzcyne ? dahch pzhichin.n
Return the shells to the shnmp broth and simmer 1 ; minutes, then strain
the broth into the stewing kettle. When the chicken is tender, remove it,
He is not going on Jego zdrouit &ie, nie yehgo zdrovieh >ien
PUIGMAPTl 44
cut all the meat from the bones, and return the meat to the pot Clean *ery well polepsza nieh polepshah TEL 213 68 94
okra, removing the ends, add it to the pot. and simmer another half-hour. Call the nurse Prosze, zawoiac pie- proshen zahvowahch RICO— -»— = '«
When ready to serve, reheat, and when hot add the shnmp and o-. iters legniarke piehlengniahrkcn
and stir in the file powder. Do not cook further but senc immediately. Lie down Prosze sie. poiozyc proshen sien powo-
The file ts crushed dried sassafras, a thickening agent, and 1$ not to be zhirh
cooked Serve with bowls of steaming nee Serves 10. or 12, Are you comfor- Czy pana wvgodnie chi pahnoo vigtttl-
tA ESTAS HORAS
table ? nieh ^QUIEN NOS
1 do not feel well Nie czuje si< dobrze nieh chooien sien
bzheh
rt.i-
**SOWCIONA
I am hot Goraco mi goronco mi 'ken UNAAVERIA? '
I have got a hic- Ja mam czkawke. van mahm chkahv-
cough

Ceremonial materials
Some most ornate forms of visual language
of the inscriptions. Formal and often archaic language is

will be found in materials intended for use on reli- usual, reflecting the special significance society
gious and ceremonial occasions, such as in books attributes to these activities. A birth certificate, a
of religious significance, memorial plaques, certifi- tombstone inscription, and part of a religious ser-
cates (examination, birth, marriage, etc.), and vice leaflet illustrate this.
1 ms.", j »j lb* SaprrtoMaAcftl

» Vol'-'
S*»enib Sunday of Eaater (€}

CERTIFICATE OF BIRTH
BH7W A.XB DlATXS RXGUTUTIQX Ad» (NO*TDU« l*ML*XT>)-

MICHAE,
mm, Jd*M.«L JltyibU.
WHO FIRS':
\ MINOA:
i^ hf y. SCRIPT AS GREEK.
1922^,1956
Place of Birth Reflatrmr'i Dlstrtet of &<W**r*

I Bnm Conn that the foregoing parUcolan

ban been couME0ta<&!?i'^: of Birth* to my

<4yy ^-t* now.

N« 87697 S

54 II LANGUAGE AND IDENTITY

 Grune Feig«r nut worn Fioanzamt •uSiufuHcn
Dialogue materials
There are many linguistic activities where the iden-
1985
71 10 02 85
tity of the visual variety is partly dependent on the
An das Finjnzjmt
active participation of the user. Either space is left

for the users to or opportunity is given for

fill in,
Antrag
.
auf Lohnsteuer- EinkommensteuererkJarung
them to reply in their own
terms. Included in the Jahresausgleich
first category are questionnaires, official forms, i iSuA-6mca«*j m«
diaries, and various kinds of stationery; in the
MMM M3Sk,De* e'-e<rrw L^nfe* N*

second are postcards, circulars, letters, and graffiti. i StauarpnicMigar

Stauarpnichllgar (Stpfl
iStpfl ),
). b« Ehagatten Ehamann
tw« E

FamMMRMM
In the illustration (right), typographical design,
Vamame
technical vocabulary,and reduced syntax provide
an unmistakeable linguistic identity for a German Tag atonal Janr Ration AuageuMar Ban*

income tax form, which is little different from its SuaAe una Hmanjmmm
counterparts in other languages, and which would PoatleftzaN derzettioef Wohnort

doubtless provide its taxpayers with a commensur-

Anaor«l»«aU»4? VooZ«a»nSung6aB»acliandarWonnaCam31 12 19SS
ate degree of difficulty. In recent years, government
departments in several countries have tried to make
such forms easier to use, with some success (§63). Ehafrau IVomame)

It is possible to make progress in clarifying layout

and question structure, but there is a limit to the Tag Monat Janr Ration AtaafaflMai BvU
degree of simplification one can introduce when GatxtaOatt^n

Strate woo Hauanummer. Poaitartzahl. o»7»itigai Wotvxyl

one is dealing with such a complex area of human
v<yiZe«aiJab**cnana»Wohniieani3l 12 198S
activity.

"""iot sas

Identifying materials Some linguistically distinctive graffiti

Probably the most widely encountered variety of Scots rule, och aye!

visual language is that used for identifying persons, French diplomats rule, au quai.
places, and objects. This includes street names, Oedipus was a nervous rex.

public signs, name tags, compliment slips, publica- Mod au Shah - et aux souris (Paris).
tion titles, identity cards, product labels, house Town criers rule, okez, okez, okez.
numbers, registration plates, letter headings, tick- Ave Maria - 1 don't mind if I do.
ets,shop facias, and much more. Typographical Synonyms govern, all right.
clarity and distinctiveness are the main characteris-
Roget's Thesaurus dominates, regulates,
tics,along with considerable grammatical abbre- rules,OK, all right, agreed.
viation and the use of specialized vocabulary. There
are marked linguistic similarities between lan-
guages. used symbols, such as
Internationally
numerals and trade marks, are routinely involved. "/ suppose it makes a

A bilingual (Welsh/English) membership card, change to see all that

foreign graffiti.
road signs, and car registration plates illustrate >" r^ -*\ T',
several of these features.

CYNGOR BWRDEISTREF CERD YN

AELODAETH

YMYSMOIV £>
ISLEOfdNGIiSet'
Canolfannau

MEMBERSHIP
CARD
Hamdden

BOROUGH COUNCIL Leisure Centres

TEULU - FAMILIES, 1986-87

F(6n/Phone :

LLANGEFNI 722966
ZONE
K- 1743 CAERGYBI/HOLYHEAD 4111 Mon Fn
AMLWCH 830060 8 30 am 6 30 pm
Saturday
8 30 am 1 30 pm C944 RFL
1 1 CONTEXTUAL IDENTITY •
55
 .

Restricted languages Heraldry a few of these are in fre- shieldis known as a

The British linguist J. R. Firth (1890-1960) intro-

The description of the sym- quent use. charge, traditionally an ani-
bols used on shields, flags, The distinctive termin- mal or geometric shape,
duced the phrase 'restricted language' to refer to seals,and other objects in- ology can be illustrated but these days an increas-
a severely reduced linguistic system used for a spe- volves an archaic grammar from some of the main ing set of modern objects is

cial activity. The language is so tightly constrained and vocabulary, much in- features of a shield. The used, as new coats of arms
by context that only a small degree of linguistic
its fluenced by French, to pro- ground of the shield can be come to be devised. The
duce phrases such as a colour, metal, or fur, in- main shield positions have
variationis permitted. These 'languages' are both
'three bars gemel sable volving contrasts such as own terms (such as
their
spoken and written, and can be found in everyday surmounted of a lion ram- argent (silver), or (gold), chief = top third; dexter =
as well as specialized contexts. They usually consist pant gules, armed and gules (red), azure (blue), as seen by the
right side,

of routinely used formulaic constructions, with a langued azure'. Heraldic sable (black), verf (green), bearer; pale = vertical

conventionalized prosody or typographical layout, glossaries often contain purpure (purple), and vair centre),and some of the
around 800 terms, but only (squirrel). The device on a patterns are drawn below.
and a limited vocabularv.

Broadcasting scores
The reporting of sports scores and stages of play is
always a highly stereotyped activity. For example, in
American baseball, there is the 'count' routine, which
specifies first

following sequence:
One and one.
the
number of strikes (0-3)

Count of one and one to M.

number of
of
balls (0-^4)
a player at bat,
and then the
as in the
^^^ Indexing
Trucker talk
The jargon American truck
of
One and oh.
Alphabetical organization a crucial feature of index-
is band
drivers using citizen
Two and oh. (CB) radio has been widely
ing style, but there are two competing principles in
Oh and one.
regular use - letter-by-letter and word-by-word. The publicized since themedium
It's one and one.
difference can be seen by comparing two small sec- became available in 1958.
Nothing and one count . .

tions of an index: The language contains a

This can be compared with the equivalent reporting in
large number of stereotyped
Japanese sportscasts, where the English vocabulary for communicating
dialect 16,42,70-90 phrases
continues to be used, although in Japanese
accent vs 3 routine messages, using a
pronunciation (e.g. three is surii, strike is sutoraiku).
dialectic 40 special numerical code (the
The conventions are different (the order is strikes
dialect mixture 80-1 CB-1 system). More com-
before balls, and there are no plural endings or
dialectology 36 plex messages use everyday
connecting words), but the stereotyped nature of the
dialect standards 65, 84-5 English, peppered with CB
language is maintained:
research into 77 slang, which makes it attrac-
Two strike two ball. tive to initiates and largely
No strike two ball. dialect16,42,70-90 unintelligible to outsiders. In
Two strike nothing. accent vs 3
this special lexiconare such
Two nothing. dialect mixture 81
items as: affirmative (yes),
Two two. dialect standards 65,84-5 bears (police), anklebiters
(After C. A. Ferguson, 1983.) research into 77
(children), doughnuts (tyres),
In reporting final scores, the convention in America, dialectic 40 eyeballs (headlights), five-
Japan, and many other countries is to read the higher dialectology 36 finger discount (stolen
score first; whereas in such countries as Britain and
goods), grandma lane (slow
Germany, the home team is read first, with intonation These samples show some of the idiosyncratic fea-
lane), handle (CBer nick-
being the signal of which team has won. Another tures ofgrammar that characterize this style, especially name), mobile mattress (car-
common convention is for the two team names to be the inverted and telegraphic syntax.
avan), motion lotion (fuel),
read together, followed by the two scores.
rubber duck (the first vehicle
in a convoy), smokey (police-
man), and super cola (beer).
Some of the main CB-1
codes are given below.
1 0-1 Poor reception
Lagg upp 90 (98) 106 m pa st 31 Byt 0-2 Good reception
Language boundaries 1st row— (K.l, P.l) twice, * K.l, 1

1 0-3 End transmission

w.f., K.3, w.f., sl.l, K.l, p.s.s.o., K.l, till st 2 och sticka 8 cm resar 2 am,
Two knitting pattern extracts, 10-4 Message under-
one English, one Swedish, K.2 tog., w.f., K.3, w.f., (K.l, P.l) 4 2 rm. Forsta v ar avigsida. Byt till st
stood
illustrate the way in which the times, rep. from * to last St., K.l. 3r, sticka ratst (= alia v stickas rata) 10-5 Relay message
features of restricted lan- 2nd row— K.l, P.l, * (K.l, P.l) 3 och oka jamnt over forsta v till 99 10-6 Standby
guage cut across linguistic times, P. 16, rep. from * to last 3 sts., (107) 115 m. Nar arb mater 46 (47) 10-7 Leaving air
boundaries. K 1 PI K 1 48 cm avmaskas den mittersta m for 10-8 In service
3rd row— (K.l, P.l) twice, * K.l, v-ringn och var sida stickas for sig. 10-9 Repeat
w.f., sl.l, K.l, p.s.s.o., K.l, K.2 tog., Minska 1 m for v-ringn = pa hoger 1 0-1 Monitoring without
sida stickas 2 rm tills, och pa vanster transmitting
w.f., sl.l, K.2 tog., p.s.s.o., w.f., sl.l,
10-20 My position is
K.1, p.s.s.o., K.l, K.2 tog., w.f., (K.l, rm tills, bakifran.
sida stickas 2
10-100 Stop at lavatory
P.l) 4 times, rep. from * to last St., Denna hoptagn gors vartannat v 21 10-200 Police needed
K.l. ggr = 28 (32) 36 m kvar for axel.

56 II LANGUAGE AND IDENTITY

SEASPEAK A Alpha AL-FAH N November NO-VEM-BER
Some rules of
B Bravo BRAH-VOH O Oscar OSS-CAH
There have been major changes in modern sea C Charlie CHAR-LEE P Papa PAH-PAH Seaspeak
transport in recent years. Larger and faster ships D Delta DELL-TAH Q Quebec KEY-BECK
E Echo ECK-OH R Romeo ROW-ME-OH • A set of standard phrases
pose greater navigational hazards. Shipping routes F Foxtrot FOKS-TROT Sierra SEE-A/R-RAH
S is recommended, to avoid
alter and present fresh problems of traffic flow. G Golf GOLF T Tango TANG-GO the many alternative ways
VHF radio permits direct communication between H
1
Hotel
India
HOH-TELL
/N-DEE-AH
U
V
Uniform
Victor
YO U-NEE-FOR.M
V/K-TAH
there are in everyday lan-
ship, shore, and aircraft, and satellite systems J Juliet 7EW-LEE-ETT w Whiskey W7S5-KEY guage of expressing the
extend a ship's communicative range indefinitely. K Kilo KEY-LOH X Xray ECKS-RAY same meaning. For exam-
L Lima LEE-MAH Y Yankee YANG-KEY ple, Say again means What
In such circumstances, mariners need to make their
M Mike MIKE z Zulu ZOO-LOO did you say?', can't hear
'I

speech as clear and unambiguous as possible. you', 'Would you repeat

Bridge officers, however, come from a variety of that?' These phrases are ita-
language backgrounds. Similarly,some numbers change their pronunci- licized in the transcribed con-
versation below.
Although English is already recognized as the ation, so that they will bemore clearly received.
• There are fixed syntactic
international language of the sea, it is essential that Large numbers have their own grammar.
and lexical routines for giving
the language should follow clear rules, so as to information. For example,
reduce the possibilities of ambiguity and confusion zero ZERO 8 eight AIT bearings and courses using
in the sending and receiving of messages. In 1980, 1 one WVN 9 nine N/NER the 360-degree figure no-
2 two TOO IS one-five WUN-FIFE tation must give three-figure
a project wasup to produce Essential English
set 3 three TREE 215 two-one-five TOO-WUN-F1FE
values: 009 degrees, not 9
for InternationalMaritime Use (referred to as Sea- 4 four FOWER 1,000 thousand TO US AND degrees, etc. Dates are sig-
s FIFE 24,000 two-four-
speak) in Britain. The recommendations relate five
nalled using prefixes, e.g.
6 six SIX thousand TOO-FOWER-
mainly to communication by VHF radio, and 7 seven SEVEN TOUSAND 'day one-three, month zero-
five, year one-nine-eight-
include procedures for initiating, maintaining, and
five'. Days of the week are
terminating conversations, as well as a recom- never used. When giving
mended grammar, vocabulary, and structure for A conversation in Seaspeak reasons, sentence construc-
messages on a wide range of maritime subjects. Western Sky (WS) is approaching Singapore tion is simplified. Everyday
The language thus has considerable expressive (SPO). English has such connec-
tives as since, because, so
power, though it is far more restricted than every-
ws: Singapore Port Operations. This Western Sky.
is that, in order to, and as, but
day language. Information: My ETA* position: East Johore pilot in Seaspeak, only reason is
station is Over.
time: one-three-four-five UTC.f used, e.g. Iintend to enter
Call-signs stern reason: my port
spo: Western Sky. This is Singapore Port Operations. first,

When sending call-signs in Seaspeak, as in air-traf- Mistake. Time is: one-four-three-zero UTC now. Stay thruster is damaged'.
• Everyone knows Mayday
fic control, police communication, and other radio on. Over.
is the marker word for Dis-
contexts, the nato phonetic alphabet is used to ws: Singapore Port Operations. This is Western Sky.
tress; but there are also
spell a word or speak out individual letters. Each Correction. My ETA is one-five-four-five UTC. Over. marker words for Urgency
letter has its own name and pronunciation (italics spo: Western Sky. This is Singapore Port Operations. (pan-pan) and Safety (Se-
Information-received: Your ETA position: East curity, say-cure-e-tay), the
mark the stress), which is given as follows in the
Johore pilot station is time: one-five-four-five UTC. latter being used when send-
Seaspeak manual. ing a message containing an
Instruction: anchor in the General Purpose
In addition to the equipment Anchorage, reason: your berth is occupied. Over. important navigational or
meteorological warning. In-
required for routine radio WS: Singapore Port Operations. This is Western Sky.
itial distress messages are
communications, several Instruction-received: anchor in the General Purpose
special-function aerials can repeated three times, and
Anchorage. Nothing more. Over.
be seen on this Batch 2 type take priority over all other
spo: Western Sky. This is Singapore Port Operations. communications.
42 destroyer. They are used
Out. • Special markers indicate
for short- and long-range
surveillance, target identifi- message type. The opening
cation, helicopter navigation,
* ETA = estimated time of arrival, word is spoken aloud, e.g.

and missile control. t UTC = coordinated universal time. Question, Instruction, Advice,
Warning, Intention, and each
has its own reply marker,
e.g. Answer, Instruction-
received, Advice-requested.
Each form has its own rules.
For example, only certain
question-forms are allowed:
rising intonations and
tag questions (e.g. isn't
it?) are not permitted. Use is

also made of turn-taking de-

check or correct
vices, to
messages, mark speaker-
change, and so on, e.g.
Understood, Mistake, Over,
Out.

1 1 CONTEXTUAL IDENTITY 57
 To associate (in argot): Marti jabrl ho. Khalag thilav. A ciphering machine (the
Hidden and secret language ('The client is rich. Make him sit separate.') Typex Mark III) used in Bri-

To client (in Hindi): Han, hath joriye. ('Yes, fold your tainaround 1 945. The lid of
Why should people deliberately use language that
hands.')
the machine is raised, show-
is unintelligible to all but a few initiates? There (Sanskrit verses are then said, while the client takes
ing the rotors that control the
permutations.
are three general reasons: to mark a person's mem- his bath.)
letter

bership of a group, to provide a pastime, and to

The argot words are sometimes of unknown origin
ensure secrecy when performing a particular ac-
(e.g.ragul 'thief, khotar 'policeman'), but often
tivity. When viewed as linguistic games, they are
they are distortions of everyday words (e.g. mandir
often seen as a creative form of play (e.g. by the
'temple' —» jhandir, ghar 'house' —» ragha) or com-
Cuna of Panama), or even as a means of improving
mon words that have been given a special meaning
competence in speaking and language learning (as
(e.g. baja 'musical instrument' is used for 'gun').
in Thai). Genres of secret language can thus be
Secret names for numbers are especially common
found in many cultures and in a wide range of
because of their role in financial transactions.
&&£\?\Am>
human contexts, especially those where there is a /
Within a radius of one kilometre around the Vish-
concern to avoid detection (as in criminal argot,
wanath temple in Vanarasi, one investigator found *""muii«fl"tftfBr
or cant), or to keep something hidden from lay
several distinct sets of secret number-names, used
people (as in magical formulae). Apart from the
by such people as diamond dealers (A), silk mer-
cases presented below, therefore, reference should
chants (B), fruit and vegetable merchants (C), and
also be made to several other instances of hidden
Pandas (D). Numbers 1-5 are given here, with
language described in this book: glossolalia (p. 11),
Hindi numbers for comparison (after R. R. Mehro-
in-law taboos (p. 42), trucker talk (p. 56), whistle
tra, 1977):
speech (p. 400), and various forms of slang (p. 53).
No. A B C D Hindi
CRIMINAL CODES
There have been few studies of the secret languages 1 airan pa sang nima sang ek
used by underworld groups - for obvious reasons: 2 thaipa swan jor javar do
it is not difficult to imagine the problems faced by 3 babarpa ikwai rag singhara tin
academic researchers who have entered dens, par- 4 airvan fok fok fok car
lours, and red-light areas, armed only with a tape 5 sutpa bud bud panro pane
recorder and an innocent smile! And even if they
can extricate themselves safely, the risk continues. The most noticeable kinds of criminal argot, or
One scholar, who studied an underworld language 'speech disguise' as it is sometimes called, are those
in a city in India, was severely beaten at a later where utterances are totally or partially unintel-
date for publishing something its speakers did not ligible to the outsider because of the distinctive
like. sounds, grammar, or vocabulary. But a great deal
In a study of the Vanarasi Pandas (those who of argot occurs that appears to be in ordinary lan-
look after Hindu pilgrims to the city), it was found guage, though in fact the utterances have a special
that they use an argot alongside Sanskrit and Hindi meaning. An example of the latter was recorded
during transactions with the pilgrims. A great deal in another study in India when the sentence Jao
of code-switching (p. 363) takes place, as can be katori manj lo 'Go, clean the bowl' was used by
seen in this sequence where a Panda talks to both a murderer to an associate in front of his victim.
an associate and a pilgrim about to bathe: The intended sense was: 'Prepare a grave'!

Cryptology
It is a short step from the have been intercepted. various ways, by substitu- genious methods that have transformations. It is also

secret languages of children
and the underworld to the
world of secret intelligence,
with its dual concern to pre-
serve the military, commer-
cial,and scientific security
of one side, and to pene-
trate the corresponding se-
curity systems of the other.
Cryptology is essentially a
two-part science. One
branch of the subject, cryp-
tography, deals with the
task of making messages
secure, so that they cannot
be understood by an
enemy. The other branch,
cryptanalysis, is concerned
with extracting the meaning
from enemy messages that
The two branches are
often referred to as code-
making' and code-break-
ing'; but these popular
names are inadequate, be-
cause they fail to distinguish
the special sense of 'code'
in this field. In cryptology,
code has to be dis-
tinguished from cipher.
code a system of
is
phrases, words, syllables,
or letters, each of which has
an associated code word'
or 'code number'; it may be
'decoded' using a 'code
book'. A cipher, by contrast,
is a system in which a mes-
sage is 'enciphered' by
transforming its letters in
tion or transposition; it may
be 'deciphered' using a
'key'. For example, the mes-
sage Crystal escape
planned Friday could be en-
coded as follows:
182 636 24 812
where each code number
would correspond to the
A words as listed in a code
book. It could also be enci-
phered as follows:
NLGHCZM YHNZPY
PMZEEYV SLAVZG
using a simple 'cipher
phabet' in which each
has been substituted (a
o = d, etc.).
The history of cryptology
illustrates the many in-
been devised to maintain
secrecy. Cipher alphabets,
for example, can be made
more complex by using
several equivalents for a let-
ter ('homophones'), as
when c is replaced by dx, re,
or pj. Several such alpha-
bets can be used at once: a
message can be enci-
phered using alphabet A,
and the result further
changed using alphabet B,
and so on; or certain letters
al- of the message can be enci-
letter phered using A, others
= z, using B. Modem cipher
machines can produce
these 'polyalphabetic' ci-
phers using millions of such
possible to use codes and
ciphers simultaneously. The
result of all this ingenuity is
the secret message, or cryp-
togram.
These days, there are
several other aspects to sig-
nal security and intelli-
gence, such as altering ra-
dio and radar frequencies,
radio silence, and stegano-
graphy- the use of techni-
ques that conceal the exis-
tence of a message, such
as invisible inks, microdots,
or the use of electronic de-
vices which hide a message
in a signal. But these are of
more interest to spies than
linguists!

58 •
II LANGUAGE AND IDENTITY
PLAYING WITH SOUNDS AND LETTERS • is a switch of initial and
In another form, there

There are several 'languages' in which words are
systematically altered, through the addition, sub-
traction, substitution, or transposition of sounds.
Some are purely phonetic changes; others require
a knowledge of spelling and alphabet order. Even
in the more complex cases, practitioners can teach
themselves to talk at great speed.
Records of some of the languages go back over
100 years in many parts of Europe. They have
finalconsonants, and of initial consonants and con-
sonant clusters of successive syllables, e.g. in a
French version, parler 'speak' —* larper, boire
'drink'—* roib. Not all classes of words are affected,
however, e.g. Je bouffe pas 'I'm not eating' —» Je
foub pas. A similar game has been found in Java-
nese, e.g. rupiab 'rupees' —» puriah, nduwe 'have'
— » wunde.
• Some secret languages involve sound substitu-

mainly been found among children, but there are tions that resemble written language codes. One
also reports of their use among adults, especially Javanese game is based on the order of the 20 con-
in contexts where secrecy is required (e.g. in front sonant letters of the alphabet. The first ten letters
of customers or small children). (h, d, p, m, n, t, d, g, c, s) are matched against the
• In back slang, words are spelled backwards, second ten (/', b, r, w, y, t, k, I, n, rj), in reverse

and then the new arrangement of letters is given order; and the members of each pair are made to

a plausible pronunciation. has been observed in

It
substitute for each other (h for T)and vice versa,
the UK among soldiers, barrow-boys, shopkeepers, d for n and vice versa, and so on). Thus the sentence
thieves, and public school pupils. First World War {h)aku gawe layag 'I'm writing a letter' emerges

examples include kew 'week', neetrith 'thirteen', as rjamu rade patarj. It is reported that some
tekram 'market', and tenip 'pint'. In French, parler speakers develop great skill in producing such
a I'envers ('speaking backwards') is found in forms at speed.
several variants, e.g. copains 'friends' —* painsco, • 'T-ing in i' (talking in initials) has been
mari 'husband' —» rima, I'envers 'backwards' —* reported, in which certain words are replaced by
verlen. The same game also appears in Javanese, their first letters. A case from a school in Texas
e.g. Bocah iku dolanan asu 'The boy is playing showed examples such as Some p l-ed the m 'Some
with a dog' —> hacob uki nanalod usa. In Thai, people liked the movie', She's a v p g 'She's a very
there are several variants involving consonants, pretty girl'. Parents also sometimes use this form
vowels, and even tones, e.g. krab bdan 'return of abbreviation in front of their young children,
home' —» kdan bdb, ydak kin khaw 'I'd like to eat' along with the other spelling conventions, such as
—* yaw kin khaak. It's time for b, e, d.

• In what
is sometimes called centre slang, the

centralvowel of a word, along with its following Mystical letters

consonant, is placed at the beginning, and a non-
Inthe Middle Ages, there
sense syllable added, e.g. eekcher 'cheek', hoolerfer arose a Jewish (later a
'fool', ightri 'right'. Christian) system of mysti-
• In eggy-peggy oraygo-paygo speech, an extra cal practices based on an
syllable added,Pugut thagat begook dow- esoteric interpretation of
is e.g.
Old Testament texts,
gun. Similar are cases where an extra vowel or con- known as the Kabbala
sonant is inserted between each syllable: using f, (from Hebrew qabbalah
for example, Where are you going becomes Wheref something received'). It
aref youf gofing. In Cuna (Panama), there is a form was thought that language
in general, and biblical lan-
in which pp or r is inserted, along with the vowel
guage in particular,
of the preceding syllable, e.g. ua 'fish' —» uppuappa, tained coded secrets about
tanikki 'he's coming' —» taranirikkiri. In another God and the world, based
Cuna game, ci is prefixed to every syllable, e.g. on the way the letters of the
maceret 'man' —* cimaciceciret. In Javanese, games text were arranged, and the
numerical values which
using an inserted for p plus vowel repetition have
could be assigned to them.
been recorded, e.g. Aku arep tuku klambi 'I want Some books, such as the
back to the early Christian
era, was known as gema-
tria. Here, numbers were
substituted for letters,
values compared in order
to provide fresh insights
into themeaning of texts.
the most commonly used
system, the first
Hebrew alphabet are
of the
numbered from 1 to 1 0; the
con- next eight are given the
values 20, 30, etc.; and the
last four letters
values 100, 200, 300, and
400. In English, the 26 let-
ters are valued 1 to 26, in
order. On this basis, all
kinds of curious and (some
Jesus, Messiah, son, God,
cross, and gospel all = 74.
Gematria is also occasion-
and ally practised outside of the
religious context. For ex-
ample, in deciding whether
In one should carry out a cer-
tain activity at a certain
ten letters time, believers may look to
see whether the numerical
value of their name and
that of the day or date cor-
respond in any way.
have the
Mystical sums
Part of the arithmetic used by
those who argue for the vali-
dity of gematria.

to buy a dress' —> afakufu afarefep tufukufu 1 3th-century Sefer ha-zo- still believe) significant cor- Bad Hide
klafambifi (after J. Sherzer, 1974). har (Book of Splendour), relations can be obtained. + Language + Listen

• In Pig Latin, the first consonants are put at viewed by many as a Linguists take note: Profane Eavesdrop
sacred book, went into the tongue and lexicon = 82.
the end of the word, and ay or e added, e.g. Utpay Torah texts in minute de- sibilantand hissing are ad- Arm All

atthay ookbay ownday 'Put that book down'. In tail, in a search for mystical jacent numbers. + Bend + Vote
a variant of this, last consonants are put at the values. Every word, letter, etymology, Indo-European Elbow Democracy
beginning of the word, with extra sounds to aid vowel point, and accent and West Germanic all = 1 37.
mark was evaluated, to Those interested in deeper
the pronunciation, e.g. Tepu tatha keboo nadaw. Not King
determine its hidden mean- matters will note that:
+ Same
A similar phenomenon has been studied in Cuna, + Chair
ing. The method lost its reli- man and Eden = 28.
where it is known as arepecunmakke (from Spanish Different Throne
gious popularity by the 18th Bible and Holy Writ are
al reves 'backwards' and Cuna sunmakke 'to century. separated by 100. Good Keep
speak'). Here, the first syllable placed at the end,
is
One exegetical techni- Mount Sinai and the laws of + Deeds + Off

—» que, which can be traced God = 135. Grass

e.g. takke 'to come' ketak, ipya 'eye' —» yaip. Scout

11 CONTEXTUAL IDENTITY •
59
 .

Verbal art Flyting

laic phrases, which may be in sequence or separated Among the Germanic peo-
SPEAKING IN PAIRS by several other 'lines'. The genre includes pro- ples, ritual cursing and
The use of parallelism to mark certain kinds of verbs, songs, and chants and is used in relation boasting, known as flyting,

speech activity is widely known. Semantic couplets to many formal activities, such as greetings, fare- often took place between
are found extensively in the Indian languages of poets or chiefs. One of the
wells, petitions, courtship, funerals, negotiations,
earliest exchanges of this
Middle America, such as Nahuatl and Yucatec, as and ceremonies of all kinds. Bini varies in length kind is recorded between the
part of formal speech genres. In Rotinese (eastern from two to several hundred 'lines'. The example English and Viking leaders in
Indonesia), for example, parallel speech (or bint) below is of a 'succession' bini, in which imagery the Anglo-Saxon poem The
is used as a form of ritual language, in which past
Battle ofMaldon (ad 991 ).
of renewal is used to express the continuation of
The form is also found in
events are recounted following a fixed ancestral lineage. The numbers refer to the lines that are in
Gaelic tradition, being best
pattern. Bini involves speaking in pairs of formu- parallel. (After J. J. Fox, 1974.) developed among the Scot-
tishpoets (makaris) of the
15th and 16th centuries.
Their ferocious exchanges of
extravagant invective are
well illustrated in The Flyting
Oe No Dain biin The goat of Oe No from Dai 1 of Dunbar and Kennedie by
Na biima-pau henuk The goat has a yellow-necklaced beard 2 William Dunbar (1460-
Ma Kedi Poi Selan manun And the cock of Kedi Poi from Sela 1
1 521 ?). The exact meaning
of some of the words un-
Na manun ma-kao lilok. The cock has gold-stranded tailfeathers 2 certain, but there is
is

De ke heni pau biin
Te hu ela lesu biin
De se lesun na pau seluk
Fo na pau henu seluk;
Ma fed heni koa manun
Te sadi ela nggoti manun
no doubt-
Cut away the goat's beard 3 ing their malicious intent!
Leaving but the goat's throat 4 Mauch muttoun, byt buttoun,
That throat will beard again 5 peilit gluttoun, air to Hil-
And the beard will be a yellow necklace again; 6 hous;
And pluck out the cock's tailfeathers 3 Rank beggar, ostir, dregar,
Leaving only the cock's rear 4 foule fleggarin the flet;

Fo nggotin na koa seluk That rear will feather again 5 Chittirlilling, ruch lilting, lik
Fo na koa lilo seluk. And the tailfeathers will be gold strands again. 6 schilling in the milhous;
Fo bei teman leo makahulun Still perfect as before 7 Baird rehator, theif of natur,
Ma tetu leo sososan. And ordered as at first. 7 fals tratour, feyindis gett . .

(II. 145-8)

VERBAL DUELLING recorded in early Chinese and Germanic languages.

Informal linguistic contests, in which people attack Among the Eskimo there are song duels, in which
each other through their forceful or ingenious use all forms of insults are exchanged. The West Indian

of language, can be found in all pans of the world, calypso was originally a type of verbal insult di-
and in all kinds of social settings. In everyday con- rected at political figures. Among black American
versation there are numerous occasions where peo- youths in ghetto areas, various kinds of exchange
ple have to fight to speak first, avoid interruption, are known as 'sounding', 'signifying', 'woofing',
and have the last word. The subject matter ranges or 'playing the dozens' — a sequence of ritual insults
from subtle forms of intellectual sarcasm and (or 'raps') followed by replies ('caps'). Among Tur-
humour to the crudest possible attacks on a per- kish boys, from around ages 8 to 14, the exchanges
son's courage, sexual prowess, or relatives. At one are phonologically linked: the retort must rhyme
level, attacks may be subtle and indirect, involving with the insult, and each new insult must be linked
allusion and figurative speech; at another, there in some way with a previous part of the sequence.
may be explicit taunts, boasts, name calling, and The exchanges are all to do with virility and homo-
jokes at the other's expense. sexuality. They are delivered with great fluency and
Often these duels take the form of set sequences speed, and may continue for some time. A fragment
of challenges and replies according to certain rules. from one exchange illustrates the rhyming pattern:
They involve a great deal of skill, as participants
A: Ustiine binek Let me
ride you
have to master special techniques of sentence con-
B: Halebe gidek go to Aleppo
Let's
struction, remember a large number of fixed
A-.Halep yikildi Aleppo was flattened
phrases, and be able to modify them in ingenious
B: Iqine tikildi It was crammed inside (you)
ways as they come under verbal attack. These duels
have been studied in places as far afield as Africa, In such sequences, A can win only if B fails to reply
the Near East, Greenland, and the Americas. They with an appropriate retort. If B succeeds, A must
Exchanging angry words in
seem to function as a means of discovering the rules continue with more taunts. The more rhymes B
the playground.
governing the social structure of the peer group. has memorized, the more he is safe from sudden
One can discover and test the dominance of others, verbal attacks. He loses the contest if he answers
without recourse to fighting and bloodshed. without rhyme, or fails to answer at all. (See
Politeness duels and boasting contests have been further, riddles, p. 63.)

60-11 LANGUAGE AND IDENTITY

CURSING AND SWEARING Four letters and
A remarkable variety of linguistic forms can be con- the law
sidered as cursing and swearing. At one extreme In 1 936, Eric Partridge
there are the complex and sophisticated expres- (1894-1979) included fuckin
sions that may be found in religious, legal, and his Dictionary of Slang and
Unconventional English.
other formal contexts. At the other, there are the
Despite his use of an aster-
many daily examples of taboo speech, usually pro- isk for the vowel, the result
fanities or obscenities, that express such emotions was a storm of complaints to
as hatred, antagonism, frustration, and surprise. schools, libraries, and the
police. Even today, the book
The most common utterances consist of single
is not always available on the
words or short phrases (though lengthy sequences
open shelves of public librar-
may occur in 'accomplished' swearers), conveying ies.
different levels of intensity and attracting different An even greater furore
degrees of social sanction. English examples range took place in 1959, when
from 'mild' expletives, such as heck and dash, to Grove Press of New York
published the unexpurgated
the two maximally taboo words, fuck and cunt. edition of D. H. Lawrence's
The functions of swearing are complex. Most Lady Chatterley's Lover,
obviously, it is an outlet for frustration or pent-up which contained several in-
emotion and a means of releasing nervous energy stances of the word. The edi-
tion was banned on grounds
after a sudden shock (§4). It has also been credited
(IKfC* of obscenity, and court cases
with various social functions as a marker of group followed, first in the U.S.,
identity and solidarity (§10), and as a way of Punch cartoon of 2 April 1 91 3 Old Lady: I shouldn't cry if then in Britain. The trial of the

expressing aggression without resort to violence. Iwere you, little man. British publishers, Penguin
Little Boy: Must do sumping; bean't I old enough to swear. Books, at the Old Bailey,
In these social contexts swearing can become a
took place in October 1 960,
dominant linguistic trait, with sentences often con- other class deals with the names of gods, devils, and a verdict of not guilty
taining many taboo words. sacred places, the future life, and anyone or any- was returned. As a result, the
Sex, excretion, and the supernatural are the main thing that holds a sacred place in the belief systems word quickly appeared in the
sources of swear-words. One important class of daily press, and it has since
of the community: God, Dear Lord, By the beard
become widespread in liter-
items deals with words to do with body parts and
of the prophet, By the holy sacrament, Heavens, ary work. In the context of
functions that society considers taboo, such as Hell Sometimes expressions from other belief
. . . public speech, however, a
tnerde, balls, and other 'four-letter' words. The systems are used (e.g. by Jove). In the course of strong prohibition remains.
Despite the development
time, euphemistic forms of words can obscure their
of liberal attitudes, there is
original meaning [hell —* heck, bloody —> bloom- still a strong antagonism to
ing, and such ingenious distortions as Geraniums the use of four-letter words in
and Gee Whiskers from Jesus). In fact, it can be public speech; and they are
still not always to be found in
argued that the real meaning of the expressions
dictionaries.There was
used swearing is rarely a factor governing their
in
nothing between fuchsite
use (thus allowing a contrast to be drawn with blas- and fucoid in Webster's Third
phemy, where the speaker has a definite intention New International Dictionary
to vilify religious matters). (1 96 1 for example but the
) , ;

never possible to predict the range of experi-

It is
gap was filled in the 1 983
addendum.
ence a culture will use to curse or swear by. It may
Roman tablet of 50 bc Cursing tablets were be the name of a dead relative, a ruler or famous Rabelaisian curses
commonplace among the ancient Greeks and
person, symbols of power, natural forces (Donner- Gargantua and Pantagruel
Romans. A curse would be inscribed on a tablet, which
would then be buried or thrown into deep water. The wetter), a part of the body (Stap me vitals), an (both 1532), by Francois Ra-
belais (c. 1495-1553), con-
lengthy inscription on one such tablet begins (after W. animal (Rats), or even a plant. One of the most
tainswearing performances
Sherwood Fox, 1919): famous oaths of ancient Ionia was ma tin krambin that have never been sur-
Good and beautiful Proserpina (or Salvia, shouldst 'By the cabbage!' - an expression that seems to passed. In the 1694 English
thou prefer), mayest thou wrest away the health, body, have originated in the special status of this vege- translation by Peter Motteux
complexion, strength and faculties of Plotius and table as an antidote to hangovers! Baudelaire swore (1660-1718), Book IV be-
consign him to thy husband, Pluto. Grant that by his gins:
own devices he may not escape this penalty. Mayest by Sacre-Saint-Ciboire 'Sacred Saint Onion'; Soc-
Ods-bodikins. What a devil.
thou consign him to the quartan, tertian and daily rates swore ni ton kuna 'By the dog'; and Pytha- Codzooks. By the mass.
fevers to war and wrestle with him until they snatch goras is said to have sworn ma tin tetrakton 'By With a pox to them. vow I

away his soul ... give thee his ears, nose, nostrils,
I
the number four'. Even nonsense words can be and swear by the handle of
tongue, lips, and teeth, so he may not speak his pain;
invoked: Robert Southey (1774—1843) swore by my paper lantern. Ad-
his neck, shoulders, arms, and fingers, so that he may zookers. Zwoons. A pox on
not aid himself. .
the great decasyllabon Aballiboozobanganovribo. it. A murrain seize thee for a

Some languages, such as Arabic and Turkish, are blockheaded booby ... By
and the curse continues through the whole anatomy
famous for the range and imagination of their the worthy vow of Charroux.
of poor Plotius, in a most comprehensive way.
swearing expressions ('You father of 60 dogs', By St Winifred's pocket. By
A similar curse was levelled against a Parisian
St Anthony's hog. By St Fer-
woman, and published in a Nancy newspaper, as 'You ride a female camel', etc.). By contrast, several
reol of Abbeville. By St
recently as 1910! peoples, such as the Amerindians, Polynesians, and Patrick's slipper. By our Lady
Japanese, swear very little, or not at all. of Riviere . .

11 CONTEXTUAL IDENTITY •
61
 .

with their fixed sentence sequence and

Language and humour word play:
final-line
Universal jokes?
Many cultural differences ex-
The story is told of a man who was
carrying out joke and subject
A: Knock, knock! ist in telling
research into the language of jokes in the Reading matter. is a fairly common
B: Who's there? It

Room of the British Library but who had to be A: name. (Fred)

experience not to see why a
expelled for laughing too loud. The story is improb- foreign language joke is
B: name who? (Fred who?)
funny. On the other hand,
able. Nothing is more likely to kill a good joke A: name + extension. (Fred (= afraid) I can't tell
certain themes are found in
than a linguistic analysis. The examples in this sec- you!) many languages. An exam-
tion, therefore, are not offered by way of enter- ple is the way certain social
tainment but solely to illustrate some of the
A classification of the types of linguistic devia-
or regional groups are
tion and incongruity would be hard to achieve, for stereotyped as stupid, so
conventions that make the expression of humour
probably all aspects of language structure have that merely by saying There
one of the most distinctive of all linguistic contexts. was this man from X', the lis-
been used as the basis of an effect at one time or
The choice of funny or silly words, grammatical tener knows that a foolish ac-
other. Further examples would include effects
patterns, pronunciations, and tones of voice is a tion is to follow. 'Irish jokes'
based on word-structure and word-class (e.g. 'Can the tradition Eng-
normal part of informal conversation. In one study illustrate in
the match box? No, but the tin can.'), idiomatic land, but the Irish should not
of an evening's conversation, the participants 'tam-
shifts (e.g. 'A: Who's that at the door? B: The in- take this personally, for they
pered with' several linguistic features, for humor- too have Irish jokes. In Dub-
visible man. A: Tell him I can't see him.'), incon-
ous effect: one person talked of climbing 'an Ande'; lin, such jokes are often
gruous themes (such as the 'elephant' jokes, which made about people from
another coined a false gender, saying 'a customs
must now number thousands), as well as the many Cork; and in Cork, the jokes
officer-ess'; and a third speaker, talking about foot-
puns and riddles, which are discussed on the facing are often made about people
ball, adopted a mock-American accent, comment-
page. For the cognoscenti, there are even jokes that from Galway. (I have no data
ing 'We wuz robbed.' In a recent conversation on who Galway people joke
cross the boundary between languages, such as 'Pas
between several teenage boys, this kind of word- about!)
de deux. Father of twins', and 'Coq au vin. Chicken Similarly, in Tonga, such
and pronunciation-play proved to be a dominant
motif, acting more as a marker of solidarity than
on lorry.' A good
joke classification would also jokes are made about people
have to deal with the contexts in which jokes are from Ena, an island off the
of humour — for the linguistic changes per se pro- coast of Tongatapu. In Jor-
used (or not used), the attitudes and expectations
voked little laughter. dan, there are jokes about
of the people who use them, and the conventions
people from the village of Al-
that listeners have to follow while a joke is being Sareeh. Several Central Afri-
JOKES told — such as not interrupting, not anticipating can tribes refer to pygmy
Modifications of these kinds happen so often that groups in this way. What is
the punch-line, and (if the joke is truly 'awful')
we hardly notice them; but they use the same prin- interesting is the way in
making a disparaging remark when it is all over.
ciple, of deviating from language norms, as is found which the same joke turns up
in more structured forms of humour, such as jokes invery different cultures. For
example, there's the one
and riddles, where the 'punch-line' frequently relies Comic alphabets
about the Sareehi boy who
on breaking the linguistic expectations of the lis- There are hundreds of poems and puns based on chased a bus all the way
tener. This can be observed even in jokes that are reciting the letters of the alphabet. Widely known home, then boasted to his
(fortunately) quite short, especially those with in the 19th century, they seem to have originated mother that he had saved 20
pence. But his mother called
highly stereotyped openings: as an adult reflex of the rhyming alphabets that
him foolish, saying that if he
came to be used in schools ('A for an Apple, an had followed a taxi, he would
What do you get if you cross . .
Archer, and Arrow; B for a Bull, a Bear, and a have saved over a pound! Of
an elephant with a mouse? Barrow', etc.). One of the alphabets reproduced course, the joke was origi-
nally told in Arabic, and the
Large holes in the skirting board. in Eric Partridge's Comic alphabets (1961) runs
monetary units were in local
as follows:
Where is Felixstowe? currency, but the same joke
is familiar in English, and is
On the end of Felix' foot.
for 'orses. N for mation. heard, with minor variants, in

for mutton. O for the rainbow. several other languages.

important that jokes have some degree of
It is
for yourself. P for soup.
stereotyping, in everyday contexts, for other-
initial
for dumb. (i.e. 'deaf Q for the bus.
wise it would not be clear what the speaker's or dumb') R for mo', (i.e. 'half
intentions were. Common markers in English are for brick, (i.e. 'heave a ['arf] a mo[ment]')
.'
such phrases as 'Did you hear the one about . .
brick') for you. (i.e. 'as for
and 'There was this man .'
Often, a sub-genre . . for vescence. you')
of joke is established through the use of a specific for police, (i.e. 'chief T for two.

opening, such as 'There was an Englishman, an of police') U for me.

Irishman, and a Scotsman .', or 'Waiter, there's for beauty, (i.e. 'age V for la compagnie.
w
. .

before .') for a quid. (i.e.

a fly in my soup.' Children's jokes in particular
. .

for Novello. 'double you' in

rely on a small number of set openings, or fixed
for oranges, (i.e. 'Jaffa betting)
internal structures, which permit a large number oranges') for breakfast.
of follow-up sequences — witness the traditional for teria. (i.e. for mistress, (i.e. 'wife
success of such patterns as 'What's the difference 'cafeteria') or mistress')
between a noun and a noun?', 'What did the L for leather. for the doctor, (i.e.

noun say to the noun?', 'Why did the noun M for sis. (i.e. 'send for.. .')

verb?', and, above all, the 'Knock knock' jokes 'emphasis')

62 II LANGUAGE AND IDENTITY

RIDDLES
Riddling is a kind of intellectual linguistic game
or contest, which in some ways is similar to verbal
duelling (p. 60). It is found in many cultures, in
all continents, and throughout history, but it is not
universal (observers have reported no riddles in
Manus, Miao, and Pukapuka, for example). A
satisfactory definition encompassing the whole of
the genre is difficult to achieve because riddles come
in several linguistic forms and are used for a variety
of purposes. It is also not easy to draw a clear
distinction between riddles and other kinds of
linguistic game, such as puns, and 'catch' questions.
But essentially, riddles are traditional utterances
intended to mystify or mislead: objects, animals,
people, and events are deliberately described in
such a way that their description suggests some-
thing quite different. The task of the listener is to
resolve the ambiguity and arrive at an appropriate
interpretation.
In Europe, riddles usually take the form of short
questions, generally with humorous Eng-
intent. In
lish, the genrefound largely in children's games
is
and conversation, from around 7 to 10 years of
age, and there are few things that make more
demands on parental patience than learning to cope
with the persistent riddle. In Africa, by contrast,
riddles are widely used by adults: they are often
cryptic statements, of a poetic or philosophical
pierre is used for both), or the puns used by the
oracle at Delphi (such as the ambiguous reply to
the general who wished to know whether he should
go on a journey: Domine, stes vs Domi ne stes
'Master, stay' vs 'At home do not stay'). Shake-
speare was one of the greatest users of puns. In
France, one of the most famous punsters was the
Marquis de Bievre, in whose never-acted play Ver-
cingetorix (1770) there is an italicized pun in every
line.
Puns are a feature of many linguistic contexts,
such as black comedy, sick humour, T-shirts, lapel
badges, car stickers, trade names, book titles, and
graffiti (p. 55). The world of advertising (p. 390)
makes great use of the economical impact and
freshness of a pun (e.g. the slogan for a new kind
of adhesive, 'Our word is your bond'). But the best
and worst of them are found in everyday conversa-
tion. Puns that have been justly lauded include the
response of the disappointed recipient of a gift of
poor quality flowers ('With fronds like these, who
needs anemones?'), the comment made by the cir-
cus manager to the human cannonball who wanted The oldest
to leave ('Where will I find another man of your English riddles
calibre?'), and the comment about the Spanish girls In the oldest collection of
in a certain town, that they are 'senoreaters'. Anglo-Saxon poetry, the
Exeter Book, there are 95
Puns have been called verbal practical jokes, and riddles, which probably date
are either loved or hated according to tempera- from the 8th century. The rid-
ment. Their popularity varies greatly between lan- dles are generally in the first

character, which do not contain any question ele- guages and cultures, though the reasons for this person, as illustrated by the
opening lines of the Book'
ment. In the ancient world, riddles had a serious are unclear; it has been said, for example, that they
riddle (translation R. K. Gor-
purpose, being used by kings, judges, oracles, and are far more popular in Britain than in the USA, don):
others to test a person's wisdom or worthiness. and in France than in Germany. But punning is
A foe deprived me of life,
Riddles vary greatly in grammatical and phono- not without its dangers. The Gnat, in Lewis Car- took away my bodily
logical form. They may be single phrases, or have roll's The Hunting of the Snark, dies of a pun. strength; afterwards wet me,
several short lines. They may be introduced by spe- And punsters should beware the phenomenon of dipped me in water, took me
cial formulae, such as What is it?, A noun, A four- compulsive punning, first recorded by a German out again, set me in the sun
where quickly lost the hairs
letter word. They may display rhymes, parallel surgeon in 1939, and now known as 'Forster's syn-
I

had. Afterwards the hard

I

rhythms, and other special effects, often (as during drome'. edge of the knife cut me . .

the Renaissance) involving intricate and sophisti-

The riddle ends:
cated forms of expression. This three-line Persian
Ask what is my name, useful
riddle from Teheran has an equivalent number of
to men; my name is famous,
syllables in each 'line' (after C. T. Scott, 1965): of service to men, sacred in
A page from the Exeter Book myself.
jdota baeradaeraend/haerce bedaevaend/ behaem
nemiraesaend/ 'They are two brothers. However much
they run, they do not reach each other.' WcZ~£T
''

/caerxdye docaerxej 'Wheels of a bicycle.'

Examples such as this also illustrate the way in

which riddles can cut across linguistic boundaries,
for the same subject matter will be found in the
riddle collections of many languages.

IN THE BEGINNING WAS THE PUN

— „ „,
|v , ^ ....
^ £k *

.
T"*™ T*»l* *,„,„„„ _
This heading is a quotation from Murphy (Samuel I *""»!*.
-

Beckett) and represents one view about the impor- *pir pi- ,

tance of puns; John Dryden's comment, that they I i*'H>i»l,

are 'the lowest and most grovelling kind of wit'
represents the other. There is truth in both. Puns
have always been known, and some have achieved
great fame — notably the Peter/ rock play on words
in the New Testament (clearer in French, where

1 1 CONTEXTUAL IDENTITY • 63
 . -

Word games Grid games Word-squares

A square of letters is con-
Playing with words is a universal human activity, There are now innumerable games that all operate on
structed, using words of
the principle of building up words on a predetermined
but it is particularly noticeable in the way literate equal length, which read in
grid. Some are intended for individual use, such as
societies have devised word games, based largely Word Search (a large letter-grid in which words have
horizontal, vertical, and oc-
casionally diagonal direc-
on the written language. People delight in pulling to be found by moving from one square to the next,
tions. Usually the words are
words apart and reconstituting them in a novel in any direction). Others are for several players, such
the same in each direction,
guise, arranging them into clever patterns, finding as Lexicon, Kan-U-Go, and Boggle. In Scrabble -the
but double word-squares',
in
most famous game of this type- points are assigned
hidden meanings inside them, and trying to use they read differently, as in:
based upon how many letters are used, with higher
them according to specially invented rules. Word points for the rarer letters, and with certain squares ORAL
puzzles and competitions are to be found in news- in the grid more valuable than others. This game now MARE
papers, at parties, in schools, on radio and tele- has its own national championships, in which expert EVEN
vision, and in all kinds of individual contexts -
players display rare feats of lexical awareness to NEAT
achieve high scores. Clement Wood's Deaf/7 of a
aswhen an adult completes a crossword, or a child A famous Roman word-
Scrabble Master cleverly portrays some of the special
game of Hangman. Something of the enor- square was part acrostic,
plays a knowledge required to keep on winning:
part palindrome. It may be
mous diversity of the 'ludic' function of language Thiswas the greatest of the game 's great players: read in four directions.
is illustrated in this section (see also pp. 59, 62-3). you played BRAS, he d make it HUDIBRASTIC.
If

He ruled a world 15 by 15 squares, S A T O R

Peopled by 1 00 letters, wood or plastic. A R E PO
TENET
He unearthed XEBEC, HAJI, useful OAID, OPERA
Acrostics Found OUOS (see pi.
QUOTHA,
of QUID PRO QUO) and ROTAS
These are compositions, usually in verse, in which Its literal translation is 'The
certain letters within the text form
Discovered AU, DE, DA all unitalicized
a word, phrase, or sower, Arepo, guides wheels
special pattern. Some are written as puzzles; in others,
(AU JUS, DA CAPO, ALMANACH DE GOTHA).
with care', but it may well
there is no attempt to conceal the answer'. Generally Two-letter words went marching through his brain, have had special signifi-
the initial letter of each line provides the clue, but Spondaic-footed, singing their slow litany: cance to Christians of the
sometimes it is based on the last letter of the line (a AL (Indian Mulberry), Al (a sloth), EM, EN, time: the middle lines form a
telestich'), combinations of and last letters (a
first BY, MY, AX, EX, OX, LO, IT, AN, HE... cross, and the letters can be
double acrostic), or more complex sequences. rearranged to form several
The day his adversary put down GNASHED,
He laid - avirtuoso feat- beneath it GOUTIER,
significant messages, like
A 'triple acrostic' The solution is based on initial,

medial, and final letters in each clue word. So placed, that six more tiny words were hatched: A
Left, middle, and right GO, NU, AT, SI, then (as you've seen, no doubt) HE, P
Give us a choice of a light.
ER. A
1 The kind of glance which he who's lost his heart
T
E
Bestows on her who wears the latter part.
2. Here is one
R
A PATERNOSTER O
With a gun. Decoding crossword clues O
3. This is bound
The crossword is undoubtedly the most popular of all S
To go round.
word games. Its origins are unclear, but became it T
4. Simplify taste
widely known in 1 91 3, when a U.S. journalist, Arthur E
And eliminate waste.
5. My meaning is made plain
Wynne, devised a newspaper puzzle, called a 'word R
By my saying again. it
cross', which quickly became a craze. But for anyone O
who has tried writing a good puzzle turns out to be
it,

far more difficult than solving The construction of the

(where A and stand for AL-
Ado Rin G it.
PHA and OMEGA).
Musk Etee R interlockingwords within the puzzle is not the issue:
Other intriguing word-
Ban Dag E the main problem is devising clues which are
shapes have been invented,
Econ Omiz E ingeniously ambiguous, but do not unintentionally
such as diamonds, pyra-
Reite Ratio N mislead.
mids, and half-squares. Also
The more difficult puzzles make use of cryptic clues,
which require the solver to understand several special
of interest is the maximum
size of such shapes. In
conventions. An anagram might be signalled by a
English, nine-word squares
figure of speech expressing disorder, such as A youth
have been completed, con-
is all mixed up 'Used in' may mean that the required
.'

Chronograms . .

word is hidden within a phrase forming part of the clue.

taining several rarewords
A date is hidden in a series of words, by using the and places, but so far no ten-
If the clue contains a parenthetic phrase such as we

letters for Roman numerals, C, D, L, M, V (used for

word squares using ten
I, hear', two similar-sounding words are involved.
different words have been
U), and X. The significant letters are usually written Punning clues often end with an exclamation or
producing an odd graphic appearance to completed - even with the
in capitals, question mark. And a large number of conventional
help of a computer.
the Chronograms were often used on medals,
line. expressions are used to symbolize certain letters, such
tombstones, foundation stones, bells, and title pages as 'left' (=/), 'north' (=n), a sailor' (=ab), or a
A nine-word square:
of books, to mark the date of an event. thousand' (=m). Q U A R
In the chronogram used in the tower vaulting of In the specialized world of the serious' crossword U P P E
Winchester Cathedral, the verse of scripture reads: compilers, the rules governing the construction of clues PO
slnt DoMVs hVIVs pll reges nVtrltll, reglnae nVtrlCes are strictly adhered to, and much pleasure is obtained OM
plae' (Isaiah 49:23, 'Kings shall be the nursing fathers by making them really difficult and ingenious. In Britain, I E
and queens the nursing mothers of thy house'). the symbol of this state of mind has been the choice N T
MDCWVWIIIIIIIIII = 1 635, the date of completion of ofpseudonyms of some of the great compilers: T E
the roof. Torquemada, Ximenes, and Azed (Deza in reverse) E R
all names of leaders of the Spanish Inquisition! R S

64 II LANGUAGE AND IDENTITY

 Anagrams
The letters of words and
phrases are rearranged to
make new words - a proce-
dure which at one time was
thought to disclose signifi-
cant information about a
person's character or
future,and even to carry
mystical meaning or magi-
cal power (p. 59). People
would sometimes live ac-
cording to the 'real' mean-
ing of their names, and in
post-Renaissance Europe,
it was commonplace to
work out laudatory ana-
grams from the names
the famous. Louis
France even had an official
anagrammatist within his
court. Jonathan Swift, on
Lipograms
These are compositions
which contain no instances
of a particular letter of the al-
phabet. An early master of
the genre was the 5th-cen-
tury bc Greek poet Tryphio-
dorus, who wrote an epic of
24 books, each omitting a
different letter of the
alphabet. One of the most
famous lipograms of recent
times is Gadsby (1 939), a
50,000-word novel by Ernest
Wright, that makes no use of
the most frequent letter of
the English alphabet,
from this remark-
tiny extract
able work illustrates how it
can be done:
Upon this basis am going to I
show you how a bunch of
bright young folks did find a
champion; a man with boys
and own; a man
girls of his
of so dominating and happy
individuality that Youth is
drawn to him as is a fly to a
sugar bowl. It is a story about
a small town . .
Sign outside The Plough at
East Hendred, Oxfordshire.
Hiding words within other
words, or spreading them
across the words in a sen-
tence (e.g. maniac within
the man I accuse) is a well-
known feature of cross-
word puzzle clues, and
can become a game in its
own right. Altering word
boundaries can also lead to
initially confusing results,
as is shown by this sign.
Rebuses
the other hand, was one of be analysed into the follow-
A rebus mixes letters, pictures and logograms (p. 200)
many who ridiculed the ing words, 'Resist- a plot
make words and sentences. Often, the sentences
to
pomposity and superstition is brought home - the tour'.
make sense only when read aloud in a certain way,
of those who dealt in ana-
As a game, however, ana- as in this famous rebus:
grams. In Gulliver's grams can provide a great
Travels, natives of Tribnia
deal of fun, especially
YY U R Too wise you are.
(= Britain) discover plots when an anagram relates
YY U B Too wise you be.
using the 'anagrammatic to the meaning of the origi-
C U R
I see you are,
/
method': nal in some way: YY 4 ME Too wise forme.
By transposing the letters astronomers —> moon-starers Other conventions are shown in H& (=hand), XQQ
of the alphabet in any sus- conversation—> voices rant on (= excuse), and in such ingenious forms as timing
pected paper, they can lay Margaret Thatcher-* Meg, timing {= split-second timing) and FECpoxTION
open the deepest designs the arch-tartar
(=smallpox infection).
of a discontented party. So, mother-in-law -* woman
for example, if should say,
I
Hitler
in a letter to a friend, 'Our parishioners -* I hire par- A typical rebus game from a children's annual
of brother Tom has just got sons
XIII of the piles', a skilful deci- —
pherer would discover that
the same letters that com-
revolution > to love ruin
sweetheart —> there we sat
total abstainers -» sit not at
FIND THE ANIMAIS
can vou orseo rne am/ma ls vefz
pose that sentence, may
FR.O/V\ THE PICTUR.6 CCU6S P
ale bars
Palindromes Tongue twisters
There are words or One of the few word games
phrases - and sometimes that relate purely to the spo-
much larger units of ken medium. Words that
language - that read the contain the same or similar
same in both directions. Sim- sounds are juxtaposed, and
pleexamples are found in the exercise is to say them
such everyday examples as as rapidly as possible, as in:
madam and Eve; but the real The Leith police dismisseth
Greek challenge is to construct long
us.
sequences that make sense,
The sixth sheikh's sixth
such as:
sheep's sick.
She sells sea-shells on the
Draw, o coward!
sea-shore.
Sex noon taxes.
at
Eh, cava, lavache? This fine Italian specimen is
e. A worth recording:
or the palindrome attributed
Se I'Arcivescovo di
to Napoleon:
Costantinopoli si voiesse
disarcivescoviscostantino
Univocalics sible. The game was in-
Able was ere I I saw Elba.
politannizzare, vi disarcives- vented by Lewis Carroll, who
By contrast with lipograms,
coviscostantino politanniz- gave as one of his first exam-
Longer sequences tend to univocalics are compositions
zereste voipernon fare dis- ples, Drive pig into sry.' His
deteriorate into nonsense, thatuse only one vowel. The
arcivescoviscostantino answer involved five steps:
though there are exceptions:
politannizzare lui?
possibilities for expression pig-wig-wag-way-say-sty.
are much more limited, but
Doc, note, dissent. A fast
I (If the Archbishop of Con- several clever poems have
never prevents a fatness. I stantinople wished to give up been constructed in this way,
diet on cod. his archbishopric, would you as is illustrated by this coup-
do the same order that he
in let from a 1 6-line work by
Pangrams
The longest palindrome is re- may not give up his arch- C.C. Bombaugh(1890): These are sentences that
putedly over 65,000 words. bishopric?) No cool monsoons blow soft contain every letter of the al-
on Oxford dons, phabet - ideally, a single in-
Orthodox, jog-trot, book- stance of each. The typist's
worm Solomons! sentence The quick brown
fox jumps over the lazy dog
HERESTO PANlts PEN D ASOCl satisfies the first criterion, but
has several duplications. A
Doublets 26-letter pangram devised in
AL HOU R 1NHAR M(LES SMIRT) One word is changed into 1984 is Veldt jynx grimps
another in a series of steps, Waqfzho buck (all words to
it
HA ND FUNLET FRIENDS <*> each intervening word
ing from its neighbours
differ-
in
be found
ary). In
in a large diction-
French, the shortest
only one letter. The chal- pangram is listed as Whisky
HI PRE IGN BE cJUSTAN DK lenge is both to form the vert.jugez cinq fox d'aplomb,
chain of linked words, and to but there are three duplica-
IN DAN DEVIL SPEAKOF NO NE do so in as few steps as pos- tions.
1 1 CONTEXTUAL IDENTITY 65
12 Stylistic identity and literature

The way people use language gives us information periods, or places. In this sense, we talk of 'Shakes- Who's who?
about their physical type, their geographical, eth- pearean style', the 'house style' of an institution,
The importance of personal
nic, and social background, and the type of context and all the variations in expression that relate to linguistic identity is often re-
in which they are communicating (§§6—11). In each psychological or social states ('informal style', cognized in the study of liter-

case, the distinctive features mark someone as 'legal style', etc.). ature, where an author's ex-
belonging to a group, or performing a particular Both these general senses are widely used in lan- pression may be analysed in
detail by literary critics to de-
type of activity along with others — 'female', 'upper guage study. Evaluative notions are an essential termine its specific effect,
class', 'from Glasgow', 'black', 'praying', and so part of aesthetic approaches to language, and are meaning, and significance.
on. But in addition, a person's language use con- implicit in such areas as elocution, oratory, and But critics are not the only
veys information of a purely idiosyncratic kind. literary criticism. Descriptive approaches are found professionals involved in the
study of language individua-
We observe the language and conclude that it is more in scientific studies, such as the various
lity.
'William Brown' communicating — or, of course, branches of linguistics, where there is a concern • Lawyers have a particular
William Wordsworth, or William Shakespeare. for objective identification without evaluation. But language of
interest in the
In everyday life, unless cognitive faculties are there is a common strand running through these their clients, especially when
impaired, people have an ability to recognize indi- various traditions: style always involves an appre- questions of libel, slander,
perjury, and mistaken iden-
vidual voices and handwriting style, and this faci- ciation of contrast between alternative locations,
are raised. In recent
tity
lity has prompted a great deal of research into periods, appearances, or behaviour. As language years there have been
voiceprints (p. 20) and graphology {$33). But adult observers, we distinguish 'Shakespearean' from several stylistic analyses of
voice quality and handwriting are relatively perma- 'not Shakespearean', 'formal' from 'informal', speech samples, where the
nent, background features of communication — 'scientific' from 'religious', and so on. And as pro-
aim has been to establish the
similarities between an ac-
general physiological and psychological reflexes ducers of language ourselves, we can to a large cused person's language
over which there is little conscious control. They extent choose the linguistic 'guise' in which we wish and that heard in a tape re-
therefore contrast with those personal linguistic to appear. cording (p. 69). However,
characteristics that are relatively temporary, in the such forensic linguistic' en-
This concept of choice is central to stylistic study,
quiries are persuasive only
foreground, changing and changeable as people whatever our approach. Style is seen as the (con- when the stylistic features
make conscious choices about what they want to scious or unconscious) selection of a set of linguistic are particularly clear-cut.
express and the way they want to express it. It features from all the possibilities in a language. The • Psychiatrists, especially
is these characteristics that provide the subject effects these features convey can be understood those practising psycho-
therapy, spend a great deal
matter of this section, often referred to cumulat- only by intuitively sensing the choices that have
of their professional lives at-
ively under the heading of style. been made (as when we react to the linguistic tempting to understand the
impact of a religious archaism, a poetic rhyme idiosyncratic linguistic beha-
scheme, or a joke), and it is usually enough simply viour of their patients - an
approach that stems from
to respond to the effect in this way. But there are
Style the detailed analyses of Sig-
often occasions when we have to develop a more mund Freud (1856-1939).
Style one of the thorniest concepts to be dealt
is analytical approach, as when we are asked our opi- By studying patients' favour-
with in this encyclopedia. To Samuel Wesley, it nion about a particular use of language. Here, itewords and sense associa-
was 'the dress of thought'; to Jonathan Swift, it when we need to explain our responses to others, tions, their errors, and the

was 'proper words in proper places'; to W. B. words they avoid, analysts

or even advise others how to respond (as in the
may draw up a linguistic pic-
Yeats, it was 'high breeding in words and in argu- teaching of literature), our intuition needs to be ture of the disorder and use
ment'. And so we could continue, through several supplemented by a more objective account of style. as a basis for treatment.
it

hundred definitions and characterizations. It is a It is this approach which is known as stylistics. Psychotherapy, indeed, has
remarkable career for a word that originally meant The notion of stylistic choice could be used to been called the 'talking cure'.

no more than a 'writing-implement' — a pointed
object, or stilus, for inscribing wax.
The many senses of style can be classified into
two broad types: the evaluative, and the descrip-
tive. Under the first heading, style is thought of
in a critical way: the features that make someone
or something stand out from an 'undistinguished'
background. In this sense, it implies a degree of
excellence in performance or a desired standard
of production, as when someone is complimented
for 'having style', or condemned for writing 'with-
out style'. The second sense lacks these value judg-
ments and simply describes the set of distinctive
characteristics that identify objects, persons,
explain many of the effects used in the expression
of social and contextual identity (§§10, 11); and
indeed, several stylisticians do adopt this wider
approach. For them, 'style' is any situationally dis-
tinctive use oflanguage — a characteristic of groups
as well as individuals. In the present volume, how-
ever, a narrower definition is used: 'style' is viewed
as the set of language features that make people
distinctive - the basis of their personal linguistic
identity.

66 •
II LANGUAGE AND IDENTITY
STATISTICS AND STYLE Yule's
The recognition and analysis of all forms of linguis- Characteristic
tic on the making of compari-
variation depends George Udny Yule (1871-
sons. We intuitively sense that individuals and 1 951 was a Cambridge
)

statistician who pioneered

groups differ and develop, and we seek to explain
several important stylostatis-
our intuition by systematically comparing the way tical measures. His main
in which they make use of specific linguistic fea- concern was to devise a
we wish to make our account objective,
tures. If criterion which would apply
largely independently of
sooner or later we need to count the frequency of .'
nlniiir sample size. 'Yule's
these features, plot their distribution in controlled Characteristic' (K) is a mea-
samples, and quantify the extent of the difference Hfralb sure of the chance that any
e rib ii nc
— at which point, we would be engaging in stylo- two nouns selected at ran-
statistics, or stylometrics. Such studies comprise a
dom from a text will be identi-
cal. It is thus a means of
major part of the field of statistical linguistics (§15) measuring the repetitiveness
— a field which investigates not only the differences ofa work's vocabulary, ex-
between samples or texts, but also the properties pressed as a single value.
that samples (and, ultimately, whole languages,
and all languages) have in common, as part of the
search for linguistic universals (§14).
~^r?tfTnnrr

>runf, llrl(
— George Udny Yule

Nobody can count everything; and even if .

rA|(!lnn(i

modern computers printed out comprehensive ac-

counts of the linguistic structure of texts, there
would not be enough time available to analyse
them. On the other hand, the larger the sample
of data analysed, the more confident our conclu- Institutions, as well as people, need to be considered
sions will be. Stylostatistical studies thus tend to in relation to the definition of style as individual
identity'. There are certain distinctive linguistic
use a small number of carefully chosen textual fea-
characteristics of newspaper language,for example,
tures and to search for these in as large a body
which be found in all instances of the genre
will
of text as is practicable. Where possible, compari- (p. 388); but each paper has its own linguistic identity
sons are made with statistical data available for too, which makes different from the others. The same
it

the whole language (such as large-scale counts of
word frequency). In this way the language acts as
a 'norm' against which the idiosyncratic features
are made to stand out.
Typically, stylostatistics investigates matters of
frequency and distribution in three main areas:
principle applies to the study of banks, commercial
products, broadcasting channels, and any organization
which requires an identity and public image. House
styles, letter-heads, newspaper titles, advertising
slogans, and many kinds of trade mark illustrate some
of the ways in which institutions rely on stylistic
features as a means of promoting corporate identity.

• formal characteristics that do not relate directly

to the meaning of a text, such as parts of speech,
source of vocabulary (e.g. Romance vs Germanic),
and the length of words, sentences, or lines;
• characteristics that relate directly to meaning,
such as the size and diversity of an author's voca- Groucho Marx The success of a public entertainer
bulary; and
may depend on linguistic idiosyncrasy. Public
recognition can come from the clever use of a single
• the detailed study of single words, or small sets catch phrase. In many cases, the image involves an
of words, such as and, or the use of on vs upon; entire way of speaking, a well-known example being
particular attention is paid to words that occur only the professional tone of voice of Groucho Marx.
once in a text, in the works of an individual author,
or in the language as a whole {hapax legomena).

Quantitative studies using these variables date

from the 19th century. Much effort was devoted
to devising measures that were statistically satisfac-
tory as well as stylistically interesting.
Stylostatistics would not normally analyse those
features over which individuals have little or no
control because they are part of the obligatory
structure of a language - such as the letter sequence
q + u in English, or the use of the article before
the noun. Where there is no choice, there is no
basis for making a stylistic contrast. Style is thus
seen as an author's regular selections from the op-
tional features of language structure.

12 STYLISTIC IDENTITY AND LITERATURE •

67
 puter programs are becoming available for the
Authorship identification study of stylistic differentiation.
One of the most important applications of stylo-
has been in relation to cases of
statistical studies
WHO WAS JUNIUS?
disputed or unknown authorship. The frequency
If we see them obedient to the laws,
[the people]
prosperous in their industry, united at home, and
and distribution of a small number of linguistic
respected abroad, we may reasonably presume that their
features in aproblem text is compared with the affairs are conducted by men of experience, abilities and
corresponding features in texts where the author- virtue. If, on the contrary, we see an universal spirit of
ship is known. Given a judicious selection of fea- distrust and dissatisfaction, a rapid decay of trade,
tures for comparison, it is often possible to make dissensions in all parts of the empire, and a total loss
an identification, though with varying levels of con- of respect in the eyes of foreign powers, we may Augustus de Morgan
fidence. In this way, several important authorship pronounce, without hesitation, that the government of
questions have been illuminated or solved. that country is weak, distracted and corrupt.
Mind, I told you so!
Sometimes a conclusion can be reached using This is an extract from a series of political letters Augustus de Morgan (1 806-
a very small number of variables. For example, in 71), British mathematician
written under the pseudonym Junius, which
one study (E. L. Moerk, 1970), as few as 20 gram- appeared in the London
daily newspaper, The Pub-
and logician, who first saw
matical features proved enough to distinguish the possibility of stylostatistic
lic 1769 and 1772. The letters
Advertiser, between authorship identification,
between 1,000-word samples of six Greek and were much appreciated and were reprinted in wrote in a letter in 1 851 : 'I

Roman writers (Herodotus, Thucydides, Xeno- pamphlet form several times. During the years that should expect to find that
phon, Tacitus, Caesar, Livy). The measures followed, and throughout the 19th century, hunt- one man writing on two dif-
included counts of main and subordinate clauses, ferent subjects agrees more
ing for the identity of Junius became a popular
nearly than two different men
certain types of connecting word, nouns in various
sport, with several well-known names being pro- writing on the same subject.
cases, and several other parts of speech. However, posed. Some of these days spurious
it is often necessary to use much larger textual sam-
An investigation in 1962 by the Swedish linguist, writings will be detected by
ples (samples of over 10,000 words are common), Alvar Ellegard (1919—) counted the words in the
this test. Mind, Itold you so.'

where special attention has to be paid to their The letter was not pub-
letters (over 80,000) and compared them with a lished until 1 882, when it was
homogeneity; sophisticated statistical measures million-word norm of political literature from the read by the American geo-
may have to be used; and a wider range of linguistic same period. Some words were found to be more physicist, T. C. Mendenhall
criteria may need to be involved, and given a precise (1841-1924). He likened the
common in the letters than in the norms, and some
definition. frequency distribution of
were found to be less common. Altogether, 458 words of different lengths to
There are several technical problems. If we are lexical features were used, along with 51 synonym the spectrum of light, and be-
counting parts of speech, then it is important to choices (such as whether Junius used on or upon, gan to search for word-
use clear criteria of identification (are London, boy, commonly or usually, till or until, know not how length profiles in several
and the rich called nouns?' (p. 91). If a word English authors - 'word
all or do not know how). For example, Junius pre-
spectra', which he thought
count is being made, a precise definition of 'word' ferred until to till in 78% of possible instances — could be as specific as me-
isof paramount importance (how are hyphenated a feature shared by only one in seven contemporary tallurgical spectrograms. He
forms to be handled? are bear 'carry' and bear 'ani- writers in Ellegard's sample. These features were made several pioneering
mal' counted as one word or two? is The Hague then compared with a sample of over 230,000 contributions to authorship
one word or two?) (p. 104). Certain kinds of data studies, including the Bacon/
words taken from the known works of the most Shakespeare controversy.
may need to be excluded from the sample (e.g. quo- likely contender for authorship, Sir Philip Francis.
tations, translated forms, proper names). And a The similarities were so significant that Ellegard
decision must be made about the range of stylistic was able to conclude with confidence (p. 15), 'We
variation to be permitted within the supposedly have identified Junius with Francis.'
'homogeneous' sample (e.g. how to take account
Junius
of different levels of intimacy and formality, or dia-
lect mixing?). Above all, the basic question must
be addressed: is the text too small to warrant any
kind of stylostatistical study — as is often the case
with poems, letters, or police statements?
It is difficult to take account of all these technical
factors, and even more difficult to anticipate the
range of external factors that interfere with linguis-
tic judgments. A text might have been written by
more than one person. It might be a deliberate imi-
tation or forgery. And there may be a large number
of potential candidates for authorship, all of whom
need to be systematically compared. But with meti-
culous care and a great deal of motivation, it is
possible to reach reliable conclusions, and several
fascinating and successful studies have been carried
out. Moreover, there is well-founded optimism for
the future, now that several specially designed com-

68 •
II LANGUAGE AND IDENTITY
FORENSIC LINGUISTICS features were examined, all to do with the way
Most stylostatistical studies are of literaryworks; Evans connected his clauses: [a) clauses lacking any
but the same techniques can be applied to any formal linkage; and clauses linked by (b) words
spoken or written sample, regardless of the 'stand- like and, (c) words like then, (d) ellipsis, (e) words
ing' of the user. In everyday life, of course, there like if, and (/) words like which. The results were
is usually no reason to carry out a stylistic analysis as follows (after J. Svartvik, 1968):
of someone's usage. But when someone is alleged
to have broken the law, stylisticians might well be Type 1 Type 2
involved, in an application of their subject some-
times referred to as 'forensic' linguistics.
Typical situations involve the prosecution argu-
ing that incriminating utterances heard on a tape
recording have the same stylistic features as those
used by the defendant — or, conversely, the defence
arguing that the differences are too great to support
this contention. A common defence strategy is to
maintain that the official statement to the police,
'written down and used in evidence', is a misrepre-
sentation, containing language that would not be
Histographic presentation of
part of the defendant's normal usage. Type 1 and Type 2 sen-

Arguments based on stylistic evidence are usually tences in the Evans state-
very weak, because the sample size is small, and A E D C B F ments.

the linguistic features examined are often not very

discriminating. But in several cases they have cer-
Typel Type 2
tainly influenced the verdict; and in one well- No. % No. %
known case subsequent analysis definitely sup- (a) 92 (37.1) 10 (20.0)
ported the contention that there had been a mis- (b) 17 (6.9) 15 (30.0)
carriage of justice. (c) 30 (12.1) 1 (2.0)

10 Rillington Place
(d) 50 (20.2) 17 (34.0)
Shakespeare -
(e) 45 (18.1) 5 (10.0)
In 1950, Timothy Evans was hanged for the murder 14
or Bacon?
(£) (5.6) 2 (4.0)
of his wife and child at this address in London. The question of whether the
Three years later, following the discovery of several works of Shakespeare could
Total 248 50
be attributed to Bacon at-
bodies at the house, John Christie was also hanged.
tracted particular interest
After considerable discussion of the case, a public The differences turn out to be highly significant. towards the end of the 1 9th
enquiry was held, which led to Evans being granted With so few criteria, and a small sample, conclu- century. T. C. Mendenhall
a posthumous pardon in 1966. sions must be tentative; but the analysis undoub- (p. 68) counted the lengths of

A central piece of evidence against Evans was about 400,000 words from
tedly corroborates Evans's denial: from a linguistic
Shakespeare's plays, and an
the statement he made to the police in London on point of view, the paragraphs that he later claimed unspecified but large sample
2 December 1949, in which he confessed to the were untrue are very different indeed from the rest from the writings of Bacon.
murders. Evans was largely illiterate, so the state- of his statement, which to the end he continued These large totals were
ment was made orally, and written down by the made up of a number of sep-
to assert was the truth.
arate counts, based on
police. At the trial, he denied having anything to
single works. He found that
do with the murders, claiming that he was so upset in each single count from
that he did not know what he was saying, and Shakespeare there were
that he feared the police would beat him up if he more four-letter than three-
letter words, whereas the
did not confess.
O O Shakespeare reverse was the case with
In an analysis of the Evans statements, amount- Bacon Bacon also had a
ing to nearly 5,000 words, it proved possible to •— — —• Bacon
much higher proportion of
show that the language contained many conflicting longer words than Shake-
stylistic features, such as those italicized below: speare. The graph (left)
shows the frequency distri-
Type 1: 1 done my day's work and then had an butions, using data derived
argument with the Guvnor then I left the job. He give from Mendenhall's original
me my wages before I went home. graphs (after C. B. Williams,
Type 2: She was incurring one debt after another and 1970).
However, statistical evi-
I could not stand it any longer so I strangled her with
dence convinces only those
a piece of rope and took her down to the flat below
who wish to be convinced.
the same night whilst the old man was in hospital.
As one sceptic remarked, in
1 901 following the publica-
The incriminating statement was analysed in five
,

tion of Mendenhall's findings:

sections, which contained background
three of if Bacon could not have writ-

information (Type 1), two of which contained the 4 6 8

ten the plays, the question
Letters per word
details of the murders (Type 2). Six grammatical still remains, who did?'!

12 STYLISTIC IDENTITY AND LITERATURE • 69

 .

categories are illustrated here, along with a selec-

Stylistic distinctiveness Classical rhetoric
tion of other kinds of figurative expression.
Rhetorical ability was prized
How do we set about the task of isolating and
• metaphor Two unlike notions are implicitly
in classical times, when
identifying the linguistic features that constitute a several major works were
related, to suggest an identity between them: written on the art of public
person's style? Traditionally, this activity was car-
speaking, including Aristo-
ried out as pan of the field of rhetoric, the study When have seen the hungry ocean gain
I
tle's Rhetorica, Quintilian's
of persuasive speech or writing (especially as prac- Advantage on the kingdom of the shore Institutio Oratoria, and
(William Shakespeare, Sonnet 64)
tised in public oratory). Several hundred 'rhetorical Cicero's De Oratore. Five
steps were thought to be
figures' were introduced by classical rhetoricians, • simile Two unlike things are explicitly com-
involved in successful rhetor-
classifying the way words could be arranged in pared, to point a similarity, using a marker such ical composition, identified
order to achieve special stylistic effects. Many were as like or as: here with their Latin and (in
restricted to the patterns found in Latin or Greek, parenthesis) Greek labels.
I wandered lonely as a cloud
but some achieved a broader currency, especially • inventio (heuresis)
That floats on high o'er vales and hills
Relevant subject matter is
after the Renaissance, in studies of poetry. (William Wordsworth, Daffodils)
brought together (a process
The traditional classification of rhetorical fig- of 'invention').
• personification A type of metaphor in which
ures distinguished between schemes and tropes. • dispositio (taxis) The
an object or idea is represented in human terms: material is organized into a
Schemes (such as alliteration) were considered to
structural form appropriate
alter theformal structure of language to create sty- And all the little roofs of the village bow low, pitiful,
for oratory ('disposition').
without altering the meaning. Tropes
listic effects, beseeching, resigned . .

• elocutio {lexis) Lan-

(such as metaphor) were thought to alter the mean-
(D. H. Lawrence, End of Another Home Holiday)
guage is chosen to suit the
ing of the language in some way. However, the • paradox A statement that is contradictory or subject matter, speaker, and
theoretical principleon which this distinction relies occasion ('style' or 'elocu-
absurd on the surface, which forces the search for
tion').
(the relationship between form and meaning, §13) a deeper level of meaning: • memoria (mneme) The
is not straightforward, and its application to the various elements of the dis-
War is peace. Freedom is slavery. Ignorance strength.
is
vast range of literary effects led to controversy, course have to be retained in
(George Orwell, 1 984)
especially over the extent to which changes of form memory.
• metonymy The use of an attribute in place of • actio {hypocrisis) The
inevitably result in changes of meaning.
speech is delivered using the
In present-day stylistic analysis (e.g. in schools), the whole, e.g. the stage (the theatrical profession),
most effective techniques
the distinction is usually not made; inventories or
the bench (the judiciary). ('delivery' or 'pronunciation').
In the middle ages, the
simple classifications of 'figures of speech' are used • oxymoron Two semantically incompatible
study of rhetoric became part
instead. Also, only a tiny number of the traditional expressions are brought together, thus forcing a
of the scholastic trivium,
labels (alliteration, simile, hendiadys, homo- non-literal interpretation, e.g. Emerson's delicious along with grammar and log-
ioteleuton, epanalepsis, etc.) continue to be taught. torment, Milton's living death. ic (§65). Post-Renaissance

It is recognized that the task of mastering long lists theorists reduced the five
• apostrophe Objects, ideas, places, dead or ab- parts to two, 'style' and deli-
of labels can alienate readers and encourage them
sent people are directly addressed: very', and the subject, as a
to go in for mechanical 'figure spotting' ('Ah, I can
result, became particularly
see three similes in that poem'), without pausing Milton! Thou should'st be living at this hour associated with techniques
to reflect on the role the figures play in the meaning (William Wordsworth, London, 1802) of verbal expression, espe-

of a text. But a selective and sensitive introduction • chiasmus A balanced structure, in which the
cially in relation to reading
aloud (the concern of 'elocu-
to figures of speech retains its value, in helping stu- main elements are reversed: tion'). Because of this in-
dents of style to see the many ways in which a
fluential tradition, many peo-
text is linguistically distinctive. Love's fire heats water, water cools not love
ple think of rhetoric as
(William Shakespeare, Sonnet 154)
essentially a matter of 'verbal
ornament'.
FIGURES OF SPEECH • Combinations of figurative effects are also com-
The modern academic
Metaphor and simile are the most widely recog- mon, especially in literary writing.
view of the subject, however,
nized figures of speech, being commonly used in The silted flow involves far more than the
many everyday varieties of language as well as in Of years on years special effects of language
production. deals with the
rhetorical and literary contexts. Some analysts con- Is marked by dawns
It

whole study of creative dis-

sider metaphor, in particular, to be the core of As faint as cracks on mud-flats of despair
course, in both speech and
linguistic (and especially poetic) creativity. Both (Stephen Spender, The Prisoners)
writing, including the use of
language in the mass media,
• anacoenosis asking the denouncing future events; speech or gestures; and the way in which audi-
Emotional ences react to and interpret
opinion of listener or reader; • diasyrmus disparaging • mempsis complaining
appeals • asphalia offering one- the arguments of one's op- against one's injuries; communications directed at
The complex universe of selfas surety for a bond; ponent; • ominatio prophesying them (p. 393). In effect, it is
traditional rhetoric is clearly • bdelygma expressing • ecphonesis using an evil; the analysis of the theory
illustrated from this small abhorrence; emotional exclamation; • paramythia consoling and practice of techniques of

selection of classical terms
which described the types
of emotional appeal.
• amphidiorthosis
modifying a charge made in
• cataplexis threatening
punishment or disaster;
• comprobatio com-
plimenting one's listeners or
judges;
• eucharistia giving those who grieve; argumentation, involving list-
thanks; • peroration summarizing eners as well as speakers,
• eulogia commending or in an impassioned manner; readers as well as writers. In
its broadest sense, there-
blessing a person or thing; • thaumasmus exclaim-
• hypocrisis mocking an ing in wonder. fore, modern rhetoric studies

anger; • diabole predicting or opponent by exaggerating the basis of forms of effec-

all

tive communication.

70 II LANGUAGE AND IDENTITY

LITERARY VS NON-LITERARY
LANGUAGE
In principle, a detailed stylistic study could be made
of a press report, a television commercial, or any
other 'everyday' use of language, and examples of
this 'general' approach to stylistics can be found
in §§11 and 63. In practice, most stylistic analyses
have attempted to deal with the more complex and
'valued' forms of language found in works of litera-
ture ('literary stylistics'). Moreover, it is possible
to see in several of these studies a further narrowing
of scope, with analysts concentrating on the more
striking areas of literary language. Poetic language
has attracted most attention, and within this there
has been a marked predilection to investigate auth-
ors who make use of highly abnormal or 'deviant'
features of language (such as Dylan Thomas or
cummings). The bias is less obvious in contem-
e. e.
porary stylistic work, but it is still present.
This concentration on the more distinctive forms
of literary expression is not difficult to explain. It
reflects the fact that linguistic analytic techniques,
as developed during this century, are more geared
to the analysis of the detailed features of sentence
structure than of the broader structures found in
whole texts or discourses (Part in). The more com-
pact and constrained language of poetry is far more
likely to disclose the secrets of its construction to
the stylistician than is the language of plays and
novels,where the structuring process is less evident,
and where dialogue and narrative is often indis-
tinguishable from the norms of everyday speech.
Most work, accordingly, has been in the area of
poetic language.
Bottom up vs top down
A more balanced account of the language of litera-
ture gradually emerging in contemporary stylis-
is
tics, with two main approaches to the subject
plainly in evidence. The first approach begins by
identifying the smallest features felt to be distinc-
tively used in a work - minimal contrasts of sound,
grammar, or vocabulary - and proceeds to build
up more complex patterns of use. The second
moves in the opposite direction, beginning with
the broadest possible statements about an author's
style, then studying particular aspects of the lan-
guage in detail.
In the first approach,we might start by consider-
ing the distinctive way in which a novelist, for
example, favours certain adjectives, varies tenses,
or coins idiosyncratic words. We might count the
frequency with which these features are used in
a particular novel, and contrast them with the
frequencies found in other works, by the same or
different authors. In the second approach, we might
start by discussing the structure of the novel as
a whole, with reference to plots and sub-plots,
favourite themes, and the way characters inter-
relate. In due course, we might proceed to look
more closely at how particular linguistic features
signal the author's intentions, and again make com-
parisons with other works.
The two approaches - sometimes referred to as
the 'bottom up'and 'top down' (or 'micro' and
'macro') approaches - can both be illuminating,
and neither excludes the other. Quite often, micro-
and macro-stylistic procedures are simultaneously
used in investigating the same work. To some
extent, the approaches are complementary, and it
might be thought that they would meet 'in the mid-
dle'. In practice, because of the multifaceted nature
of stylistic analyses, and the different theories used
by stylisticians, this hardly ever happens.
T.S.Eliot (1888-1965)
A linguistic perspective
Linguistic approaches to the study of literary style
importance of seeing an individual auth-
stress the
or's use oflanguage in the context of the language
as a whole. It is pointed out that this language
cannot be studied in isolation from other varieties.
Literature reflects the whole of human experience,
and authors thus drawing on all
find themselves
varieties of language (or even on different lan-
guages) as part of their expression. In a single work,
they might make considerable use of a non-literary
variety, or allude to several such varieties. This hap-
pens most markedly in drama and the novel, as
can be readily observed from the vast linguistic
range of Shakespeare or Moliere, or the regional
and class varieties found throughout the writing
of Dickens or Hardy. But the tapping of language
varieties can be found in poetry too, as in The
Waste Land, where T. S. Eliot draws on linguistic
features belonging to conversational, religious,
medieval, and musical varieties, as well as a wide
range of literary forms. Indeed, this work clearly
illustrates the way literature knows no linguistic
bounds, for it also includes lines in French, Ger-
man, Italian, and Sanskrit (p. 73).
Nor can language be isolated from the
literary
least situationally specific language variety of all
— conversation (p. 52). When an author uses lan-
guage ingeniously, we instinctively relate the spe-
cial features to our own spoken norms, and any
explanation of the effect ultimately depends on our
awareness of these norms, and of how the features
relate to them. To use modern stylistic terms, we
see how the features have been 'foregrounded' —
made to stand out from the background of normal,
unremarkable usage. Robert Graves recognized the
importance of this principle when he said that a
poet should 'master the rules of grammar before
he attempts to bend or break them'. So too critics
— and indeed all who enjoy literature - need to
be aware of the normal constraints on language
use before they can explain the effects authors
achieve when rules are bent or broken.
12 STYLISTIC IDENTITY AND LITERATURE •
71
THE EDGES OF LANGUAGE We might dispute the particular ordering of items
Authors take when
they push language to its
risks on this scale,but the general move away from
limits. If they break too many rules, they can fall literal meaning is clear enough, as is the growing

over the edge of language into unintelligibility.
Even well-known authors, such as James Joyce and
Dylan Thomas, have been criticized for verbal
excesses - for sacrificing meaning to the seductive
patterns of sound or graphic form. Is it possible
to arrive at a satisfactory explanation of
distinctive graphic features in e.e.
difficulty we encounter as we attempt to provide
a plausible context for each use. It would also be
possible to construct even more bizarre examples
{an incompleteness ago), and thus to suggest how,
with the more deviant kinds of poetic language,
all the a reader might simply give up the struggle to decode
cummings's Four its meaning. There is nothing more likely to crush

III, for instance? the desire to read poetry than having to resort to
The move from centre to edge of language is cryptanalysis.
a gradual one. We
can take an everyday construc- Poets are not the only ones who push language
tion, and manipulate its use to show increasing beyond its normal limits. All who engage in literary
levels of inventiveness — and thus increasing diffi- or quasi-literary activity, from novelists and
culties of interpretation. One construction that has dramatists to journalists and commentators, face
been well studied from this point of view is the similar problems. Nor is the wrestle with words
use of ago with a noun phrase to express various restricted to literature. Humourists, both amateur
temporal meanings. It is possible to construct a and professional, are another group who con-
continuum that has mundane uses at one end and stantly tease new effects out of old words, in their
bizarre uses at the other, as in this example (from search for good punch-lines. And a further example
G.N. Leech, 1969): is provided by the heading of the present section,

which is the title of a book by the German theo-

several hours ago MUNDANE logian, Paul van Buren, about how people use the
many moons ago word 'God' as part of religious discourse (p. 384).
ten games ago In his view, theistic language is 'a case of walking

several performances ago language's borders' - an attempt to express insight

a few cigarettes ago at the very edge of the 'platform of language',

three overcoats ago where, if we try to go further, 'we fall off into a
two wives ago misuse of words, into nonsensical jabbering, into
a grief ago (Dylan Thomas) the void where the rules give out'. Theologians,
a humanity ago ABNORMAL like poets, it seems, are continually striving to say
what cannot be said.

Bending the rules e. e. cummings

Dylan Thomas's poetry re- Four III
peatedly illustrates the way
in which some poets bend
grammatical rules as they
strive to express their in-
here's a little mouse) and
sights. A feature of his sty- what does he think about, i

technique is the use of

listic
wonder as over this
unexpected associations
floor (quietly with
between words (p. 105) -
as well as a grief ago, ex- bright eyes) drifts (nobody
amples (from Fern Hilt) in- can tell because
clude happy as the heart Nobody knows, or why
was long, all the sun long, jerks Here &, here,
and once below a time. gr(oo)ving the room's Silence) this like
Such effects can be for- a littlest

mally identified only by poem a

drawing attention to the (with wee ears and see?
everyday meaning of the
tail frisks)
underlying phrase. In such
(gonE)
cases, usage norms pro-
'mouse',
vide a relevant perspective
for the discussion of stylis-
We are not the same you and
tic effects, and often act as i,since here's a little he
a stimulus to critical think- oris
ing. it It

? (or was something we saw in the mirror)?

Dylan Thomas
therefore we'll kiss; for maybe
what was Disappeared
into ourselves
who (look), .startled.

72 II LANGUAGE AND IDENTITY

Literary genres GRAMMAR AND VOCABULARY Multilingual poetry
• archaisms The use of grammar and vocabulary may even
Lexical effects
Genres of literature are established categories of no longer current is a well-established feature of cross language boundaries.
In Verlaine's Sonnet boiteux,
composition, characterized by distinctive language poetry (though not so common today). Examples
English words heard in a
or subject matter. The most widely recognized are include the use of grammatical forms such as 'twas
London fog are interspersed
poetry, drama, and the novel, but several other and quoth, words such as e'en, fain, and wight, within the French text:
categories exist, such as the short story, auto- and spellings such as daunsynge and olde. Tout I'affreux pass6 saute,
biography, and essays. Each major category can • neologisms The invention of new words is per- piaule, miaule etglapit
be further classified — for example, epic, lyrical, haps the most obvious way to go beyond the nor- Dans le brouillard rose et
jaune et sale des sohos
and narrative genres within poetry; comedy, mal resources of a language: completely fresh
Avec des indeed et des all
tragedy, and farce within drama; and romance, creations, such as Shakespeare's incarnadine: new rights et des haos.
crime, and science fiction within the novel. constructions, such as Hopkins's 'widow-making (The whole hideous past
unchilding unfathering deeps' {The Wreck of the jumps, whines, mews and
Deutschland) and new parts of speech, as in Othel-
;
yelps in the pink and yellow
and
Poetry lo's verb lip (= kiss), To
lip a wanton in a secure
with indeeds
dirty fog of the sohos,
and all rights
couch.' and hey-o's.)
There has always been controversy over the nature • poetic diction In a narrow sense, this term But one of the best-known
of poetic language. To some, poetic language refers to vocabulary that is typically poetic, and examples is the cluster of
foreign language elements in
should be special, removed from the language of that would rarely be used in other contexts; more
the closing lines of T. S.
everyday (thus, Thomas Gray's dictum, 'The lan- broadly, it can mean any use of words thought
Eliot's The Waste Land:
guage of the age is never the language of poetry'). to be effective by the poet, whether or not it occurs
I sat upon the shore
To others, it should be closely in touch with every- elsewhere. The traditional sense can be illustrated Fishing, with the arid plain
day, or, perhaps, be 'current language heightened' by nymph, slumber, woe, and billows, or many behind me
(Gerard Manley Hopkins). To Ralph Waldo Emer- lines from 18th-century poetry, such as the opening Shall I at least set my lands
order?
in
son, the whole of language is in any case 'fossil of Thomas Gray's Elegy in a Country Churchyard:
London Bridge is falling
poetry'.
The curfew tolls the knell of parting day, down falling down falling
Statements of this kind to some extent miss the down
The lowing herd winds slowly o'er the lea
point, which is to stress the enormous range of Poi s'ascose nel foco che gli

linguistic expression that is found under the head- For the broader sense, there is the beginning of affina
Quando fiam chelidon -
ing of poetry. At one extreme, there are poems that Stephen Spender's The Exiles: uti

O swallow swallow
are as far removed from everyday speech as it is History has tongues Le Prince d'Aquitaine a la
possible to imagine; at the other, there are poems Has angels has guns - has saved has praised - tour abolie
that, if it were not for the division into lines, would Today proclaims These fragments I have
closely resemble prose. Poetic movements often Achievements of her exiles long returned. shored against my ruins
Why then lie fit you.
swing between these poles, as people respond to Hieronymo's mad againe.
• word order Abnormal word order is common,
the competing linguistic influences of old traditions Datta. Dayadhvam. Dam-
as when nouns (e.g. Mil-
adjectives are placed after
and contemporary realities. It is not possible to yata.
ton's 'Anon out of the earth a fabric huge /Rose Shantih shantih shantih
make simple general statements about the form of
like an exhalation' {Paradise Lost)), or the normal
poetic language, therefore; all one can do is identify
order of elements in a clause is reversed (e.g. Ham-
a number of recurrent notions that are part of the
let's 'I might not this believe . .
.').
traditional image of poetic language, and that enter
into what is often called 'poetic licence'.
The creativity poets seek takes many forms. It
The very language
may involve the invention of totally new linguistic
of men
William Wordsworth, who
features, as in the neologistic vocabulary of James
made a strong statement
Joyce, or the typographical design of a poem by about the relationship of
e.e. cummings. But it more often takes the form poetry to prose. In the
of a fresh use of familiar language, as when John Preface to the Lyrical Ballads
Donne compares himself and his mistress to the (1800), he wrote:

legson a pair of compasses, or T. S. Eliot's Prufrock

My purpose was to imitate,
compares the evening laid out against the sky to and, as far as is possible, to
a 'patient etherised upon a table'. Above all else, adopt the very language of
poets fear banality. Whatever the literary era or men There will be found
. . .

in these volumes little of

tradition in which they find themselves, they are
what is usually called poetic
concerned to avoid what is linguistically boring or diction;as much pains has
predictable, and to discover ways in which words been taken to avoid it as is
can come alive, to convey fresh worlds of meaning. ordinarily taken to produce it

T. phrase vividly captures the essence of

S. Eliot's
... It may be safely affirmed
that there neither is, nor can
their predicament: 'the intolerable wrestle with
be, any essential difference
words and meanings'. between the language of
prose and metrical
composition.

12 STYLISTIC IDENTITY AND LITERATURE .

73
 .

SOUNDS AND RHYTHMS is classified on the basis of the number of stressed The sound of
For many people, it is the sound or 'music' of poetry syllables it contains, such as the monometer (1), silence
that chiefly identifies the genre - the distinctive use dimeter (2), trimeter (3), tetrameter (4), pentameter Lateral consonants (p. 157)
of vowels, consonants, cadences, and rhythms. (5), and hexameter (6). Combinations of foot-type are used by poets in several
Several different phonetic and phonological stylis- and line-length produce such designations as 'iam- languages to suggest soft-

bic pentameter' - the so-called 'backbone' of Eng- ness and silence:

tic which is
features contribute to the total effect,
often studied under the separate heading of phono- lish metre.
Wi/d thyme and va//ey- /i/ies
stylistics. I come not, friends, to steal away your hearts whiter sti//

(William Shakespeare, Julius Caesar) Than Leda's love, and

• Individual sounds can be used in an onomato-
cresses from the rill
poeic or symbolic way (§30) for expressive pur- The rising world of waters dark and deep (Keats, Endymion)
poses. Poetry sometimes uses vowels and (Milton, Paradise Lost)
Les souff/es de /a nuit f/ot-
consonants to reflect the noises of real life, or to
Many famous studies (such as G. Saintsbury's taient sur Ga/ga/a.
symbolize other sensory or abstract notions, such (Victor Hugo, Booz Endormi)
three-volume A History of English Prosody,
as colour, texture, character, or mood (e.g. The breezes of the night
1906-10) have been devoted to plotting the metri- floated over Galgala.'
Milton's 'The serpent subtlest beast of all the
cal norms in a language's poetry, and evaluating
field'). Poetic rhythms, too, can directly evoke real Dir in Liedern, /eichten,
the kinds of deviations from these norms that poets
world sounds and events, as in schne//en,
use. As systems of description, they work quite well Wa//et kiih/e F/uth.
I sprang to the stirrup, and Joris, and he; in analysing the regular lines of traditional poetry. (Goethe, West-ostlicher
I galloped, Dirck galloped, we galloped all three.
But they have been criticized on several points. Divan)
(Robert Browning, How They Brought the Good News from
Ghent to Aix) They tend to be applied in too mechanical a way; 'For you the cool waves lap
it is often difficult to decide which analysis to assign in songs light and nimble.'
• A network of associations can be built up
to a line containing an abnormal rhythm; and they Ah! Lagyan ke/az eji sze/
between sounds. By repeating vowels or conso- break down completely when they encounter the Mi/ford bbo/fe/e.'
nants at different points, words and phrases can markedly irregular lines of modern 'free verse': (Janos Arany, A Walesi ba
be formally linked, sometimes to achieve a purely rdok)
What is the use of talking, and there is no end of talking,
aesthetic effect, sometimes to force the listener to Oh! The night breeze rises
There is no end of things in the heart.
consider their possible relationships of meaning. towards Milford
softly
I call in the boy,
Three cases are usually recognized, involving the Haven.'
Have him sit on his knees here
(After S. Ullmann, 1964.)
repetition of initial consonants {alliteration, as in To seal this,
fine friend), vowels (assonance, as in roll/moan), And send it a thousand miles, thinking.
and final syllables {rhyme, e.g. gladness /madness). (Ezra Pound, Exile's Letter)
The poet's struggle
But other effects are possible, such as the repetition In such cases, necessary to devise alternative
it is T. S. Eliot's lines, in Easf
of initial syllables, as in state/stayed {reverse metrical models, using different analyses of stress, Coker, sum up what for
rhyme), or the simultaneous repetition of initial and introducing other prosodic notions, such as many is the essence of the
and final consonants, as in bend/bound {para-
tempo, pause, and intonation (§29), to identify the linguistic task facing the
poet.
rhyme). The opening lines of Coleridge's Kubla patterns that emerge when readers utter the lines.
Khan illustrate the overlapping use of several of So here am,
I in the middle
these effects. way, having had twenty
• The language is organized into rhythmical years . .

units, which appear in print as lines. In European In Xanadu did Kubla Khan Trying to learn to use words,

poetry, the traditional study of versification, or J and every attempt

Is a wholly new start, and a
prosody, was based on the rules of Latin scansion, different kind of failure
and many generations of schoolchildren have had Because one has only learnt
to learn to scan verse on the assumption that poetry A stately p leasure -dome decree to get the better of words
For the thing one no longer
in their language used similar rhythms to those of
has to say, or the way in
Latin. But the quantitative metrical system of Latin,
which
based on a classification of syllables into long and One is no longer disposed to
short durations, is by no means universal. English Where Alph, the sacred river, ran say it. And so each venture
Is a new beginning, a raid on
and German use an accentual system, in which
the inarticulate
heavy and light syllables alternate. Classical Chin-
With shabby equipment al-
ese used a tonal system, which alternated classes ways deteriorating
of even and changing tones. Sometimes only the Through caverns measureless to man In the general mess of impre-
number of syllables in a unit is critical, regardless cision of feeling,
Undisciplined squads of
of their pitch, loudness, or duration, as in Mord-
emotion.
vinian. And several 'mixed' metrical systems have
been found, such as French, where syllable number
and accentuation combine.
Down
T ^
to a sunless sea,
T "c -
Traditional analyses of English metre divide poe-
tic lines into combinations of stressed (') and
Sound-patterns in the opening lines of Coleridge's
unstressed (w) syllables known as 'feet'. Four types Kubla Khan
are prominent in English verse: the iamb {»'), tro-
chee ('u ), anapaest w
'), and dactyl
(
(' ). A line w
74 II LANGUAGE AND IDENTITY
Concrete poetry
This 1950s movement produced a form of poetry,
variously known as 'shaped', pattern', concrete', or
'Cubist' poetry, which blurred the boundary between
literature visual arts. In concrete poems, the
and the
primary consideration is the way in which letters and
words are arranged on the page, so that they visually
reinforce, or act as a counterpoint to, the verbal
meaning.
Shaped poems can in fact be traced to classical
Greek times, and they emerge over the centuries in
the work of several western writers, such as Apollinaire
(p. 1 1 ), Mallarme, Mayakovsky, Dylan Thomas, and
cummings (p. 72), as well as being very popular in
eastern literature, where they may have originated.
One of the best-known examples is The Altar, by the
1 7th-century poet George Herbert. Representative of
the more recent movement is Au Pair Girl by Ian
Hamilton Finlay.
f-»wn Qi
Ian Hamilton Finlay's nfllT
Au Pair Girl y»
rlau pair
Drama
There has been remarkably little study of the genre
of drama from a linguistic point of view. This is
partly because so much of the language it contains
is traditionally analysed under other headings —
reference is often made to the 'poetry' of Shake-
speare's plays, for instance, which is then investi-
gated using metrical and rhetorical techniques. And
if the language of drama is not poetry (it might
be argued), then it is prose, and thus analysable
using the techniques of other prose genres, such
as the novel or short story.
But drama is neither poetry nor novel. It is first
and foremost dialogue in action. With few excep-
tions, there is no narrative framework other than
that provided by the language of the characters
and by the visual setting in which they act. The
author cannot step back and provide an opinion
or manipulate our point of view, as happens rou-
tinely in novels. The dialogue must do everything.

;airgirl au Dramatic dialogue also has to be convincing, as

a representation of conversation. But to be convinc-
au pair girl ing is not to be real. No dramatist presents us with
the equivalent of a tape recording of everyday
ju pair girl a speech, with all its hesitations, broken syntax, and
inexplicit vocabulary (p. 52). Even the most collo-
au pairgirlo
,rl quial of dramatic conversations, whether it is
written by Harold Pinter, Arthur Miller, or Shakes-
jairgirl au pairgin peare, presents us with an exercise in linguistic arti-

jirlau pair girl au pair fice, the extent of which is only beginning to be

appreciated, as techniques become available that

jair girl au pair girl au pa allow us to make comparisons with real conversa-
tion.
air girl au pairgirl au pair
pair girl au pair girl au pa
>u pairgirl au pair girl at
r^au pairgirl au pai r
i I

- Significant silence
~*irlau pairgirl Even pauses can be mani- be well out of the draught ASTON: Are you?
pulated for special drama- there. DAVIES: Always have
tic effect, as is well illus- ASTON: You don't get been.
trated by this extract from much wind. (Pause.)
Harold Pinter's The Care- DAVIES: You'd be well out You got more rooms then,
George Herbert's The Altar taker (1959), where three of it. It's different when have you?
A broken Altar, Lord, thy servant rears, degrees of pause are writ- you're kipping out. ASTON: Where?
Made of a heart, and cemented with tears; ten into the dialogue. Apart ASTON: Would be. DAVIES: mean, along the
I

Whose parts are as thy hand did frame; from controlling the pace of DAVIES: Nothing but wind landing here ...up the
No workman's tool hath touch'd the same. the drama, the pauses also then. landing there.
A Heart alone underline the uncertain re- (Pause.) ASTON: They're out of
Is such a stone, lationship between the ASTON: Yes, when the commission.
As nothing but characters, keep the at- wind gets up it . . DAVIES: Get away.
Thy pow'r doth cut. mosphere tense, and help (Pause.) ASTON: They need a lot of
Wherefore each part to promote our sense of ap- DAVIES: Yes... doing to.
Of my hard heart prehension. ASTON: Mmnn... (Slight pause.)
Meets in this frame, DAVIES: You sleep here, (Pause.) DAVIES: What about
To praise thy name. do you? DAVIES: Gets very downstairs?
That if I chance to hold my peace, ASTON: Yes. draughty. ASTON: That's closed up.
These stones to praise thee may not cease. DAVIES: What, in that? ASTON: Ah. Needs seeing to . . . The
O let thy blessed Sacrifice be mine, ASTON: Yes. DAVIES: I'm very sensitive floors . .

And sanctifie this Altar to be thine. DAVIES: Yes, well, you'd to it. (Pause.)

12 STYLISTIC IDENTITY AND LITERATURE 75

 .

CONTROLLING THE DIALOGUE ROS: (irritated by the repetition): What? A novel for voices
The stylistic distinctiveness of a dramatic text lies guil: Don't you discriminate at all?
It would take only the ad-
primarily in the conventions of layout, abbrevia- ROS: (turning dumbly): Whs.'} dition of character names
(Pause.) and a change
tion, and direction that the dramatist employs to to present
tense in the last line to turn
and the movement
indicate the nature of the action,
and interpretation of the dialogue. But there are FROM DRAMA TO NOVEL the following extract into a
play. It is the whole of a
great variations in approach. Some authors restrict chapter from Hank Stine's
themselves to the essential directions about the ac-
A play for voices The Prisoner (1 970, p. 63),
tions of the characters, leaving it to the actors and In Dylan Thomas's Under Milk Wood (1954), the a novel based on the cult

producer to infer interpretations about character narrator's voice interweaves with the voices of the television series starring

characters to produce a work that, though cast in Patrick McGoohan. The

or tone of voice from the text. In Shakespeare, for Prisoner is under arrest.
example, we find the bare minimum of comment, dramatic form, permits several of the effects of the
novel. On what charge?'
as in this scene from Hamlet:
'A complaint has been
SCENE I. Elsinore. The guard-platform of the Castle. ACT I brought against you.'
'By whom?'
[francisco at his post. Enter to him Bernardo.] first voice: In the blind-drawn dark dining-room of By a party who considers
ber: Who's there? School House, dusty and echoing as a dining-room himself aggrieved.'

fran: Nay, answer me. Stand and unfold yourself.

in a vault,Mr and Mrs Pugh are silent over cold grey 'In what way?'
cottage pie. Mr Pugh reads, as he forks the shroud 'It is not my place to
ber: Long live the King!
meat in, from Lives of the Great Poisoners. He has know.'
fran: Bernardo? 'Whose place is it?'
ber: He. bound a plain brown-paper cover round the book.
Those whose place it is
Slyly, between slow mouthfuls, he sidespies up at Mrs
fran: You come most carefully upon your hour. to know such things.'
ber: Tis now struck twelve; get thee to bed, Francisco. Pugh, poisons her with his eye, then goes on reading. 'And when will be ad- I

He underlines certain passages and smiles in secret. vised of the nature of the
By contrast, some authors make great use of mrs pugh: Persons with manners do not read at table, charge?'
stage directions — in effect, giving the reader a par- first voice: says Mrs Pugh. She swallows a digestive 'At the proper time.'

tial interpretation of events. Along with the formal tablet as big as a horse-pill, washing it down with 'And when will that be?
clouded peasoup water. When it is deemed
indications of dialogue, they provide the chief
[Pause] necessary.'
marker of linguistic identity of the genre in its writ- 'And who will decide it is
mrs pugh: Some persons were brought up in pigsties.
ten form. Extracts from Act of Tom Stoppard's
I necessary?'
MR pugh: Pigs don't read at table, dear. The proper authorities.'
Rosencrantz and Guildenstern Are Dead (1967) first voice: Bitterly she flicks dust from the broken 'Just who are these
way stage directions can give basic
illustrate the cruet. It settles on the pie in a thin gnat-rain. "proper authorities"?'
information about character, setting, and plot. MR pugh: Pigs can't read, my dear. Those who have been
MRS pugh: I know one who can. duly constituted.'
ACT I And who constituted
them?'
Two Elizabethans passing the time in a place The people.'
without any visible character. Charles Dickens 'Which people?'
They are well dressed - hats, cloaks, sticks and all. Ina speech made in 1 858 boundary between plays and 'The people of this village.
:

Each of them has a large leather money bag. (For the Royal Theatrical novels Every writer of f ic-
'
Now, you will come with me.
Fund), Dickens made a re- tion, though he may not There will be no further ar-
guildenstern's bag is nearly empty.
mark which underscores the adopt the dramatic form, guments.'
rosencrantz's bag is nearly full.
problem facing anyone who writes in effect for the stage.' They went out into the
The reason being: they are betting on the toss of a
wishes to argue for a definite rain.
coin, in the following manner: guildenstern
(hereafter 'guil'J takes a coin out of his bag, spins it,

letting it fall, rosencrantz (hereafter 'ros'j studies

it, announces it it happens) and puts it
as 'heads' (as
in his own Then they repeat the process. They have
bag.
apparently been doing this for some time.
The run of 'heads' is impossible, yet ros betrays no
surprise at all - he feels none. However, he is nice
enough to feel a little embarrassed at taking so much
money off his friend. Let that be his character note.

ros: Now why exactly are you behaving in this

extraordinary manner?
guil: I can't imagine! {Pause.) But all that is well
known, common property. Yet he sent for us. And
we did come.
ros (alert, ear cocked): I say! I heard music -
guil: We're here.
ros: - Like a band - 1 thought I heard a band.
guil: Rosencrantz...
ros: (absently, still listening): What?
(Pause, short.)
guil: (gently wry): Guildenstern . . Charles Dickens (1812-1870)

76 • II LANGUAGE AND IDENTITY

The novel
Who spoke?
Since the 18th century, the novel has become the The characters a novel
in
major genre of literature in most literate societies. can be presented in sever-
It has attracted a vast range of literary criticism, al ways: the author can
but few large-scale linguistic investigations. Enor- describe them directly,
other characters can talk
mous variations in the size and scope of different about them - and they can
novels make it difficult to arrive at satisfactory talk for themselves. One of
generalizations about linguistic form and content, the most important linguis-
other than to identify its essentially narrative pur- tic techniques of character-
ization is through the use of
pose. The problem has long been recognized:
a distinctive style of
indeed, it was present from the earliest years of
speech, which emphasizes
the genre, when authors searched for a label to features of regional or
identify their new product. Henry Fielding, for class background, or per-
example, called his Joseph Andrews (1742) 'a sonal idiosyncrasies. The
effect may be conveyed by
comic Epic-Poem in Prose'.
the habitual use of a single
Part of the analytical problem lies in the way word (as in the first extract
novels contain so much variety mixing. They tap below), by more funda-
the resources of a language's stylistic range more mental changes in gram-
matical construction (as in
than does any other genre. In principle, no charac-
the second and third ex-
ter, situation, theme, plot, or point of view is
tracts), or by a completely
excluded. All language varieties might expect to different orthographic sys-
be represented in a novel sooner or later, from the tem, specially devised to
most colloquial to the most formal, from the most capture features of pronun-
ciation (as in the fourth ex-
mundane to the most arcane. Even other major tract).
genres come to be swallowed up by the novel, as
may be seen in an early work, Samuel Richardson's
Clarissa (1747-8), where one of the letters from
Lovelace to Belford opens with the heading 'Act
II, Scene: Hampstead Heath, continued. Enter my
Rascal', and continues in dialogue form with stage
Speech presentation
directions interspersed. James Joyce's Ulysses has grammars, a
In traditional
basic distinctionis drawn
over 100 pages in dramatic form, representing the
between 'direct' and 'indir-
fantasy world of one of the characters. And, as speech.
ect' (or 'reported')
a recent example, D. M. Thomas's The White Hotel In directspeech, some-
(1981) contains long extracts of poetry, alongside one's words are quoted ex-
actly as they were said
documentary prose, imaginative prose, postcards,
(e.g. 'Did the man see you
letter-writing, scholarly writing, and even foot-
yesterday?' Mary asked
notes! John.); in indirect speech,
In recent years, linguistic (or textual) analysis we express what was said
has examined several aspects of the way language in our own words (e.g.
Mary asked John if the man
is used to identify the various themes, characters,
had seen him the day be-
settings, plots, and viewpoints which are intro-
fore.). Several important
duced into the narrative. There have been several linguistic changes take
studies of the experiments in linguistic technique place in moving from direct
that have been a major feature of novel-writing to indirect speech, such as
the removal of inverted
during the past century, especially in relation to
commas, the change of
the ways in which a character's consciousness pronouns, and the 'back-
might be portrayed (p. 78). Authors such as Henry shift' in tense forms and
James, Virginia Woolf, and James Joyce have associated adverbs. Both
styles of speech presen-
attracted particular attention from this point of
tation are widely used in
view. Another well-investigated area is the author's
the novel.
use of linguistic devices to maintain realistic dia- But there are several
logue and to identify character. And a third area other, more subtle modes
is the study of the movement and direction of plot,
of presenting speech,
which can be illuminated by the detailed study of where the distinction
patterns of sentence and paragraph connectivity -
a major feature of the emerging field of text-
linguistics (§20), and a preoccupation of structur-
alist approaches to literature (p. 79).
'1 think you are wrong,
Uriah,' Isaid, dare say
'1
there are several things
could teach you, if you
would like to learn them.'
'Oh, 1 don't doubt that,
Master Copperfield,' he
answered; 'not in the least.
But not being umble your-
self, you don't judge well,
perhaps, for them that are.
1won't provoke my betters
with knowledge, thank you.
I'm much too umble. Here
is my umble dwelling, Mas-
ter Copperfield.'
(Charles Dickens, David
Copperfield, 1 849-50,
Ch.17)
'No,' El Sordo said and
ted his shoulder. 'Joke.
Comes from La Granja.
Heard last night comes
English dynamiter. Good.
Very happy. Get whisky.
For you. You like?'
(Ernest Hemingway, For
between direct and indirect
does not easily apply. 'Free
indirect speech' (le style in-
direct libre, or erlebte
Rede) is one such mode,
which, typically, the report-
ing clause of indirect
speech is dropped, but the
other conventions are re-
tained (e.g. Had the man
seen John the day be-
fore?). But there may be
further variants, in which
only some forms are al-
tered (e.g. Had the man
seen John yesterday?).
And thereis also a cate-
gory of 'free direct speech',
in which the writer moves
from narrative to direct
speech without the use of
the usual markers (e.g.
Mary approached John.
Did the man see you
yesterday? John looked
away.).
Charles Dickens was
one who experimented
successfully with modes of
speech presentation -
12 STYLISTIC IDENTITY
Whom the Bell Tolls,
1 940,
Ch.11)
1
While narrating these
things, every timeQuee-
queg received the toma-
hawk from me, he flour-
ished the hatchet-side of it
over the sleeper's head.
What's that for, Quee-
queg?' 'Perry easy, kill-e;
oh! perry easy!'
(Herman Melville, Moby
Dick, 1851, Ch. 21)
'Noa!' said Joseph . .
Noa!' that manes nowt -
Hathecliff makes noa
'cahnt uh t'mother, nur yah
norther - bud he'll hev his
pat- lad; und Aw mun tak him -
soa now ye knaw!'
(Emily Bronte, Wuthering
Heights, 1847, Ch. 19)
especially the use of free
indirect speech. The Coro-
ner's interrogation of the
crossing sweeper, Jo, in
in Bleak House (Ch. 1 1 ) is a
good illustration of the way
this style can add speed
and economy to a narra-
tive, as characters interact
without the cumbersome
use of such conventions as
'said X ... said Y'. It also
conveys something of the
character's typical speech
style, and his unspoken
process of reflection:
Name, Jo. Nothing else
thathe knows on. Don't
know that everybody has
two names. Never heard of
sichathink.
Don't know that Jo is
short for a longer name.
Thinks it long enough for
him. He don't find no fault
with Spell it? He can't
it.
spell No father, no
it.
mother, no friends. Never
been to school . .
AND LITERATURE. 77
 .

Taking texts to pieces

Some ways of telling
• The author, in the first
person, takes on the per-
sona of someone in the
story. This convention
allows a great sense of in-
volvement and immediacy,
and a personal relationship
to develop with the reader;
but inevitably there is a
limited perspective:
1801 - have just returned
. 1
from a visit to my landlord - the
solitary neighbour that shall
a story
The present century has witnessed a series of aca-
vention, adopting the view- to think that his wife was com-
point of the people
demic approaches to the study of style, deriving
in the vi- ing to see him. There was no-
cinity, first through the use thing concrete she could do to partly from linguistics and partly from literary
of question-forms, then help him, of course, but it
theory. It is a complex period, raising questions
through speculative com- would mean a lot to be able to
unburden himself, to tell her
about literary interpretation and evaluation that
mentary:
how awful he felt, how the in- go well beyond the scope of this encyclopedia. In
A fine
night, and a bright large jection hadn't done him any this section, therefore, it is possible only to hint
moon, and multitudes of stars. good, and how horrible the
MrTulkinghorn, in repairing to people in the ward were. She at the overlap that exists between the study of lan-
his cellar, and
opening and
in would sympathise with him, guage and this broader domain.
shutting those resounding and he would feel better. He
The end of the 19th century saw a reaction
doors, has to cross a little might ask her to bring him a
prison-like yard. He looks up book, some cheerful modern against the traditional view of a literary work as
I

be troubled with. This is cer- casually, thinking what a fine book, and his fountain pen, so the product of an author's way of thinking about
tainly a beautiful country! In all what a bright large
night, there'd be no recurrence of that
the world, in which the writer's personal history
England, do not believe that
I
moon, what multitudes of stars! ridiculous situation yesterday

Icould have fixed on a situation A quiet night, too. when he'd had to borrow the and cultural milieu were crucial factors in arriving
so completely removed from young boy's pencil to write at an interpretation. To some critics, this view of
the of society. A perfect down the prescription. Yes,
stir
What's that? Who fired a gun and most important of all, he literature was too subjective and 'mystical', and
misanthropist's Heaven - and
Mr Heathcliff and are such a
I
or pistol? Where was it? could get her to find out about drew attention away from what they saw to be
The few foot-passengers that fungus for him, the birch
suitable pair to divide the deso- the one definite fact about a work: the language
lation between us. A capital and stare about
start, stop, fungus.
fellow!
them. Some windows and (Alexander Solzhenitsyn, in which it was written. Only a close analysis of
(Emily Bronte, Wuthering doors are opened, and people Cancer Ward, 1 968, Ch. 1 3) textual language, it was felt, would place the study
come out to look. It was a loud
Heights, 1847. Ch. 1) of literary texts on a firm, objective footing.
report, and echoed and rattled • The story may be told by
• A third-person narrative, heavily. It shook one house, or
the characters themselves, The first method to attempt this task developed
with an omniscient narra- so a man says who was pass- in the form of a represen- in France, to be known as explication
and came
tor,can provide a compre- ing .. Has Mr Tulkinghorn
.

stream of
tation of their de This used analytical techniques deriving
texte.
hensive account of all as- been disturbed? His windows
consciousness. In its most
pects of the story, including
are dark and quiet, and his from the study of classical languages. Texts were
door is shut. must be some- extreme form, there are no
the characters' motiva-
It

quotation marks or report- seen as unique, autonomous units of meaning, with

thing unusual indeed, to bring
tions, without personal in- him out of his shell . .
ing verbs; sentences are a complex internal structure that could be disco-
volvement: (Charles Dickens, Bleak short and elliptical; topics vered only through a meticulous language analysis.
House, 1853, Ch. 48) change suddenly:
Emma Woodhouse, hand- The linguistic features identified (such as figurative
some, clever, and rich, with a • The author may switch O, look we are so! Chamber language or metrical structure) were then related
comfortable home and happy from third-person to first- music. Could make a kind of to the text's historical background and to the
disposition, seemed to unite
person and back again, pun on that. It is a kind of music
some of the best blessings of reader's aesthetic response.
existence: and had lived nearly
often a highly indirect
in I often thought when she.

twenty-one years in the world and subtle manner. In this Acoustics that is. Tinkling. The 'formalist' school of Russian (later Czech)
Empty vessels make most
with very little to distress or vex extract, the two viewpoints critics in the early 20th century focussed on the ana-
noise. Because the acoustics,
her. continue almost simulta- lysis of the literary text as an end in itself, without
the resonance changes ac-
(Jane Austen, Emma. 1816, neously, as first-person
cording as the weight of the reference to social history, the writer's intention,
Ch. 1) words like yesterday and water is equal to the law of fall-
the reflective yes combine or the reader's reaction. In this approach, literary
• The author may cease to ing water. Like those rhapso-

be omniscient, adopting with third-person words dies of Liszt's, Hungarian, language was seen as a special variety, whose aes-
the point of view of the such as he, and words that gipsyeyed. Pearls. Drops. thetic effects could be explained by a systematic,
could represent either Rain. Diddle iddle addle addle
reader, or of another char- technical analysis.
viewpoint, such as might oodleoodle. Hiss. Now. Maybe
acter the novel. In this
in
and could:
now. Before. These principles were introduced into modern
extract, the author breaks (James Joyce, Ulysses, 1 922, English criticism in the 1920s, through the work
the usual third-person con- It cheered Pavel Nikolayevich p. 281 (Penguin edn.))
of I. A. Richards (1893-1979), William Empson
(1906-1984), and Cleanth Brooks (1906-), and
indue course became associated with the American
school of critical theory known as 'New Criticism'.
In linguistics, the ideas of the Russian emigre
linguist Roman Jakobson (1896-1983) were
influential in the development of a specifically
Leo Spitzer (1887-1960)
INTUITION linguistic approach to the study of style, which later
This Austrian-born jjterary theorist provided an early ac-
Identifies
came to be called linguistic stylistics (sometimes
count of the link between stylisticJeatures and aesthetic Significance
response. His approach to stylistic study involved the use of patterns features stylolinguistics) . In its early period, this was much
of what he called a 'philological circle' (right) of reasoning. confirmed taken up with microanalytical studies of word use,
The first step is to use our intuition to identify stylistic fea- metrics, and sentence structure. In recent years,
tures. The second step is to analyse these features and
efforts have been made to broaden the scope of
find a pattern. As a result, we confirm the validity of our
the subject, through the study of patterns above
original intuition. Patterns Features
discovered described
the sentence (§20).
and analvsed

78 . II LANGUAGE AND IDENTITY

STRUCTURALISM AND AFTER • 'Helper' vs 'Opponent', as characters assist or hinder
Formalist approaches displayed several limitations. the course of events.

They were unable to handle types of literature that These notions, it was argued, identified a common
did not use specifically 'literary' language (p. 71), structure of themes, actions, and character types
and their microanalytic techniques were not suit- underlying all kinds of narrative.
able for larger texts, such as the novel. As a result, Structuralism paid little attention in its analyses
an alternative approach developed during the to the role of the human mind or social reality.
1950s based on the principles of structural linguis- A poem, for example, was to be understood not
tics (§§16, 65). This provided a fresh focus for tex-
by studying the experience of the poet, the reader,
tual analysis concentrating on the function of the
or the world, but by studying the text. The author
various elements in a text. The insights of
the
was no longer the authority for interpretation; the
founder of modern linguistics, Ferdinand de Saus- meaning of a text was to be found in its individual
sure (p. 407), were used to hypothesize rules gov- use of language. This meaning was accessible to
erning the underlying system of meaning that a the critic because author and critic both belonged
literary text expressed. The aim was not to interpret to the same community language system (or langue, Roland Barthes French
texts, in the traditional way, but to define universal p. 407). Language had been handed down to an theorist Barthes (1 91 5-80)
principles of literary structure using linguistic tech- was a major influence on
author, who used it to construct a text. In this view,
early structuralist thought,
niques. As one critic put it, the aim was to 'trans- language did not reflect reality, but created it. and he continued to play an
form literary studies into a scientific discipline' (T. Structuralism brought a valuable objectivity into important role in the post-
Todorov). literary analysis, but at the expense of the total structuralist period. In his
The approach used by the French anthropologist neglect of an author's individuality, the social con- later thinking, the focus on a
Claude Levi-Strauss 1908 ( —
and others was to take
)
text, and the varying historical situation. In the
text's formal structure is
replaced by an emphasis on
the basic notion of a contrastive unit (or '-erne', late 1960s, accordingly, there developed a reaction the active, creative process-
as in 'phoneme', p. 160), and apply it to the analysis to this 'logocentric' view, which came to be known ing carried out by the reader
of behaviour (kinship, eating, etc.). In literary stu- as 'post-structuralism' -a set of ideas whose impli- (cf . Chomsky's emphasis, in

dies, research focused on finding a common struc- cations have still to be fully explored. Here, lan-
the same decade, on the
creative abilities of the
ture underlying the many kinds of narrative text guage seen not as a static structure, existing
is speaker (p. 409)).
(e.g. folk tales, myths, detective stories). For exam- regardless of social, historical, or personal Several reader-oriented
ple, significant basic units of myth ('mythemes') considerations, but as a system whose values shift approaches to literature
were recognized, and organized as a set of binary
in response to these factors, and whose meaning have now been proposed,
oppositions, in thesame way as phonemes. In one and a number of controver-
is too complex to be demonstrable by structuralist
sial issues have emerged or
study (A. Greimas, 1966), it was suggested that
J. techniques. A range of post-structuralist view- re-emerged as to whether
three basic thematic contrasts occur in all narra- points has developed which emphasize the limita- readers can be credited with
tive:
tions of binary analyses, draw attention to the a 'literary competence' cap-
able of handling the special
multiple and overlapping meanings of words, and
• 'Subject' vs 'Object', which relate to the desire or properties of literary lan-
stress the role of mental processes in interpreting guage. But there a com-
search that motivates a character at the beginning of is

the story (e.g. a detective searching for a murderer).

linguistic relationships. The approach is highly mon emphasis on the opi-
• 'Sender' vs 'Receiver', as people communicate with critical of the scientific aims of structuralism, deny- nion that meaning is not to

be found the language of

in
each other about relevant events (e.g. establishing ing the possibility of objectivity in textual interpre-
the text. Rather, it is the
various facts about the murder). tation.
reader who constructs the
text's meaning, always read-
ing in meanings which can-
not be found within the text
itself. Texts, in this view,

Deconstruction have no separate identity:

they exist only when they are
The methods and principles tion begins by isolating a construct the opposition, the presence, both can be seen
read.
come under
of structuralism specific structural relation- critic reverses the expected as derived). In this way,
most severe attack in the ship (e.g. speech' vs 'writ- priorities (showing that, in readers are forced to rethink
approach known as 'decon- ing'), and identifying the certain respects, writing the validity of the sets of op-
struction', associated pri- priorities that give the struc- might be closer to self-con- positions they use to think
marily with the writing of ture its centre (in structura- sciousness than speech, about the world.
Jacques Derrida (1930-). list thinking, speech is held and speech less so). The The approach has at-
This approach aims to show to be more fundamental, result, however, is not to tracted a great deal of inter-
inherent contradictions and closer to thought, express- see the alternative term as est among literary theorists
paradoxes in the way that ing the presence' of the in some way superior (to in recent years. Whether a
structuralism demonstrates author more directly; writing see writing as fundamental, coherent critical position
the rules governing the is a derived medium, with and speech as derived). can ultimately emerge from
structure of texts, especially an independent existence Rather, the whole basis of such radical questioning is a
its reliance on binary oppo- on paper that makes it less the opposition is called into major theme of contempor-
sitions. able to maintain the author's question {both speech and ary critical debate.
The task of deconstruc- presence). In order to de- writing can be shown to lack

12 STYLISTIC IDENTITY AND LITERATURE •

79
PART HI
The structure of language
The structure of language is something most of us
take completely for granted. We are so used to
speaking and understanding our mother tongue
with unselfconscious ease that we do not notice
the complex linguistic architecture that underlies
almost every sentence. "We forget the years we
expended in mastering this skill, so that when we
encounter the structural complexity of a foreign
language as an adult, we are often amazed at the
level of difficulty involved. Similarly, when we hear
of people whose ability to control the structures
of their language has broken down, as in the case
of aphasia (p. 270), we can be surprised at the
amount of structural planning involved in the
linguistic analysis and treatment of their handicap.
Such instances suggest the central importance of
the field of linguistic structure, not only to such
specialists as teachers or therapists, but to all
wish to further their understanding of the pheno-
menon of language.
A simple but effective way of sensing the variety
and complexity of language structure is to turn a
radio dial slowly from one end of a waveband to
the other. The first reaction to the auditory tangle
of sounds and words must be one of confusion;
but if we stop and listen for a while to one of the
foreign-language stations, a pattern will gradually
emerge. Some words will stand out, and some (such
as international products or political names) may
be recognizable. The pronunciation will become
less alien, as we detect the melodies and rhythmical
Fragments of the grammatical and semantic structure of
language.
patterns that convey such information as 'stating'
and 'questioning'. Wemay even find ourselves dis-
tinguishing familiar from unfamiliar sounds.
The same kind of reaction takes place when we
scan an array of foreign-language publications.
Instead of sounds and rhythms, we are now dealing
with shapes and spaces; but the principle is the
same. The multifarious variety of visual forms,
many of which are expressing similar meanings,
is a striking manifestation of the diversity and
depth of language structure.
In this part of the encyclopedia, we therefore
examine the factors involved in carrying out a
structural analysis of language, whether spoken,
written, or signed, and illustrate the main compo-
nents, or levels, that linguists have proposed in
order to elucidate the way languages operate. The
who largest section will be devoted to the field of gram-
mar, which is at the centre of most linguistic investi-
gation, but several pages will also be given to
semantics, the study of meaning in language, and
to the associated themes of dictionaries, place
names, and personal names. We begin with a
review of some general issues that form part of
any structural study of language, and address the
question of whether all languages have properties
in common. Part in concludes with a discussion
of some of the more recent movements in linguistic
study which analyse conversations, narratives, and
other kinds of spoken or written discourse.
13 Linguistic levels
There is too much going on in a piece of speech,
writing, or signing to permit us to describe its
characteristics in a single, simple statement.
in a short spoken sentence such as Hello there!,
several things are taking place at once. Each word
conveys a particular meaning. There is a likely
order in which the words may appear — we would
not say There hello! Each word is composed of
a specific sequence of sounds. The sentence as a
whole is uttered in a particular tone of voice (poorly
signalled in writing through the exclamation mark
(§29)). And the choice of this sentence immediately
constrains the occasions when it might be used —
on a first meeting (and not, for example, upon
leave-taking). While we say or hear the sentence,
we are not consciously aware of all these facets
of its structure, but once our attention is drawn
to them, we easily recognize their existence. We
could even concentrate on the study of one of these
facets largely to the exclusion of the others — some-
thing that takes place routinely in language teach-
ing, for instance, where someone may learn about
aspects of 'pronunciation' one day, and of 'voca-
bulary' or 'grammar' the next.
Selective focusing of this kind in fact takes place
in all linguistic studies, as part of the business of
discovering how language works, and of simplify-
ing the task of description. The different facets are
usually referred to as levels of linguistic organiza-
tion. Each level is studied usingterms and
its own
techniques, enabling us to obtain information
about one aspect of language structure, while tem-
porarily disregarding the involvement of others.
The field of pronunciation, for example, is basically
analysed at the level of phonetics, using procedures
that are quite distinct from anything encountered
at other linguistic levels. When we do phonetic
research, we try to disassociate ourselves from the
problems and practices we would encounter if we
were carrying out a study at the level of, say, gram-
mar. Similarly, grammatical study takes place using
approaches that are in principle independent of
what goes on in phonetics. And other levels, like-
wise, provide us with their own independent 'slant'
on the workings of language structure.
The notion of levels widely applicable, espe-
is
cially when we engage of a range
in the analysis
of languages, as it enables us to see and state pat-
terns of organization more clearly and succinctly
than any other way that has so far been devised.
Levels appear to have a certain empirical validity
in psychological and neurological contexts also
(§45). At the same time, we must never forget that,
when we isolate a level for independent study, we
are introducing an artificial element into our
enquiry, whose consequences must be anticipated.
82 III THE STRUCTURE OF LANGUAGE
The sounds of speech that we study via phonetics Space-station
substance through which the pat-
are, after all, the linguistics?
Even terns of grammar are conveyed. There will there- In this diagram, each circle
fore be interrelationships between levels that need represents a linguistic level.
to be taken into account if we wish to understand If the whole were drawn as
the way language as a whole is organized. As with a complex of modules and
corridors, the result would be
any structure, the whole cannot be broken down
not unlike a space station
into its constituent parts without loss; and we must matter, Charles
(or, for that
therefore always recollect the need to place our De Gaulle Airport in Paris).
work on individual levels within a more general Linguists may enter the sys-
structural perspective.
tem atany level, with imme-
diate access to all other le-
vels. In this respect, the
HOW MANY LEVELS? linguist bears a striking re-
It isnot difficult to sense the complexity of language semblance to the star-child
structure, but it is not so easy to say how many of Arthur C. Clarke's 2010\
levels should be set up in order to explain the way
this structure is organized. Some simple models of
language recognize only two basic levels: the set
of physical forms (sounds, letters, signs, words)
contained in a language, and the range of abstract
meanings conveyed by these forms. More com-
monly, the notion of forms is sub-divided, to dis-
tinguish different kinds of abstractness. In speech,
for example, the physical facts of pronunciation,
as definedby the processes of articulation, acoustic
transmission and audition, are considered to be the
subject matter of phonetics (§27). The way differ-
ent languages organize sounds to convey differ-
ences of meaning is the province of phonology Key
(§28). And the study of the way meaningful units L Linguist
are brought into sequence to convey wider and M Morphology
more varied patterns of meaning is the province P Phonetics
of grammar. The term semantics is then used for
1
P Phonology
the study of the patterns of meaning themselves. S Syntax
1
Four-level models of language (phonetics/ S Semantics
phonology/grammar/semantics) are among the O Other levels
most widely used, but further divisions within and
between these levels are often made. For example,
within the level of grammar, it is common to recog-
nize a distinction between the study of word struc-
ture {morphology) and the study of word sequence
within sentences (syntax) (§16). Within phono-
logy, the study of vowels, consonants, and syllables
(segmental phonology) is usually distinguished
from the study of prosody and other tones of voice
(suprasegmental phonology) (§29). Within seman-
tics, the study of vocabulary (or lexicon) is some-
times taken separately from the study of larger
patterns of meaning (under such headings as text
or discourse) (§20). All of these are regularly
referred to as 'levels' of structure.
We could continue, making divisions within divi-
sions,and recognizing more subtle kinds of struc-
tural organization within language. We could
extend the notion to include other aspects of lan-
guage functioning apart from structure (as when
some scholars talk of a pragmatic 'level' (§21)).
But there comes a point when the notion ceases Models of spoken language structure
to be helpful. When a theory sets up a large number
of levels, it becomes difficult to plot the relation- 2-level language 3-level language
ships between them, and to retain a sense of how
they integrate into a single system. At that point,
alternative models need to be devised.
form meaning pronunciation
WHICH LEVEL FIRST? (phonology)

Is there a 'best' direction for the study of a lan-

guage, using the framework of levels? The Ameri- 4-level language

can linguist Leonard Bloomfield (1887-1949)
recommended an approach in which one worked
through the various levels in a particular order,
pronunciation grammar meaning
beginning with a phonetic description, proceeding (semantics)
through phonology, morphology, and syntax, and
concluding with semantics. In this view, the analy-
phonetics phonology
sis at each level apart from the first is dependent
on what has gone before. Workers in the Bloomfiel-
dian tradition would talk about starting at the 'bot- 5-level This model (after M. A. K. Halliday, 1 961 ) recognizes three primary levels
tom', with the phonetics, and working 'up' to the (substance, form, context); substance and form are related by the 'interlevel' of
semantics - though in view of the complexity of phonology; form is divided into grammar and lexis.

the task facing phoneticians, phonologists, and

grammarians, it ismoot point whether anyone
a language
who strictly followedthis approach would ever
arrive there. Apocryphal stories abound of infor-
substance extra-linguistic
mants in field studies who have died before the
situation
investigating linguist got around to studying the
meaning of the speech patterns that had been so phonetics phonology grammar lexis context
painstakingly transcribed! (semantics)
In any case, it is now recognized that it is possible
to carry out an analysis at one level only if we
6-level approach (after S. M. Lamb, 1 966), the various levels are
In this referred to
make certain assumptions about other levels. Our
as strata, and the model as a whole is known as stratificational grammar.
choice of sounds to describe phonetically depends
to some extent on our awareness of which sounds
play an important role in a language (phonology), language

which in turn depends on our awareness of the

way sounds distinguish words (grammar) enabling phonology grammar semology
them to convey differences in meaning (semantics).
Similarly, when we study grammatical patterns,
such as sentence structure, we need to be aware hypophonemic phonemic morphemic lexemic sememic hypersememic
of both semantic factors (such as the relationships (= phonetic) (= semantic)

of meaning that bring the patterns together) and

phonological factors (such as the features of into-
The present work In this encyclopedia, we shall be making most use of a 6-level
nation that help to identify sentence units in
model which uses three basic notions (transmitting medium, grammar,
of structure
speech). In a sense, when we work with levels, we semantics), each containing a twofold division. The model also incorporates the
need to be able to move in all directions at once. dimension of language in use, which is related to the concerns of language structure
The British linguist J. R. Firth (1890-1960) once through the notion of pragmatics. The diagram gives only the distinctions required for
the spoken medium of transmission: these are reviewed in detail in Part iv. The
likened the business to a lift that moves freely from
properties of the written medium are reviewed in Part v, and of the signing medium
one level to another, in either direction, without in Part vi. The various facets of language in use are discussed in Part n and §63. The
giving priority to any one level. The simile makes remaining levels, including pragmatics, are dealt with in later sections of Part in.
its point, but the two-dimensional analogy is still
language
misleading. To
capture the notion of levels, multi-
dimensional geometries are required.

phonetics phonology morphology syntax lexicon discourse

13 LINGUISTIC LEVELS •
83
14 Typology and universals
Expressing
comparison
The languages of the world present us with a vast In Chomsky's view, therefore, the aim of linguis- The English comparative
construction, 'X is bigger
array of structural similarities and differences. Why tics is to go beyond the study of individual lan-
than V" involves three parts:
should this be so? One way of answering this ques- guages, to determine what the universal properties the adjective (big), the
tion is to adopt a historical perspective, investigat- of language are, and to establish a 'universal gram- markers of comparison (-er
ing the origins of language, and pointing to the mar' that would account for the range of linguistic and than), and the standard
importance of linguistic change — a perspective that variation that is humanly possible. The question of comparison ( Y). This way
of putting it is shared by
is discussed in Part IX. An alternative approach is simply: What are the limits on human language
many languages, including
is to make a detailed description of the similarities variability? Languages do not make use of all pos- Berber, Greek, Hebrew,
or differences, regardless of their historical antece- sible sounds, sound sequences, or word orders. Can Malay, Maori, Songhai, Swa-
dents, and proceed from there to generalize about we work It might be possible to
out the reasons? hili, Thai, Welsh, and Zapo-
tec.
the structure and function of human language. draw between the patterns that are essential
a line
However, the opposite
There are two main ways of approaching this features of language, and those that no language
order, in which the standard
latter task. We might look for the structural fea- ever makes use of. Or perhaps there is a continuum of comparison is expressed
tures that all or most languages have in common; between these extremes, with some features being first, is also common. In Jap-
or we might focus our attention on the features found in most (but not all) languages, and some anese, for example, it is 'Y
yon okii (literally 'Y than
that differentiate them. In the former case, we are being found in very few. Questions of this kind
big'), and this way of putting
searching for language universals; in the latter case, constitute the current focus of many linguists'
it shared by Basque, Bur-
is
we are involving ourselves in language typology. attention. mese, Chibcha, Guaranf,
In principle, the two approaches are complemen- Hindi, Kannada, and Turkish,

tary, but sometimes they are associated with differ- among others. Finnish is a
language which uses both
ent theoretical conceptions of the nature of
constructions.
linguistic enquiry.

SIMILARITY OR DIFFERENCE? The Port-Royal

Since the end of the 18th century, the chief concern
has been to explain the nature of linguistic diver-
grammar
Contemporary ideas about GRAMMAIRE
This was the focus of comparative philology
sity.
and dialectology, and it led to early attempts to
the nature of linguistic uni-
versals have several ante-
cedents in the work of 1 7th-
GENERALE
set up genetic and structural typologies of lan-
guages (§50). The emphasis carried through into
century thinkers. The
Grammaire ge'ne'rale et rai-
,E T RAISONNE'E,
sonn6e (1 660) is widely re- n »
the 20th century when the new science of linguistics
continually stressed the variety of languages in the
cognized as the most in-
CONTENANT v..

*2_
fluential treatise of this Zes fondemens de fartde farler ? expli-
world, partly in reaction against the traditions of period. It is often referred to
19th-century prescriptivism, where one language, as the Port-Royal gram-
quex^ £vne manure da ire & naturellei
Latin, had been commonly regarded as a standard mar', because was writ- it
Zes raifons de ce qui communa. toutes
of excellence (§1). ten by scholars who be- eft

Since the 1950s, the focus on diversity has been

longed community of
to the les Ungues, & des principalis differences
intellectuals and religious qui s'y reruontrent
replaced by a research paradigm, stemming from established between 1 637
j

of the American linguist Noam Chomsky

work
the
(1928- ), in which the nature of linguistic univer-
and 1660 in Port-Royal,
Versailles.
Itplufteurs remarques nouueUes fur U
Zangue Francoife.
Although published ano-
sals holds a central place. Chomsky's generative
nymously, the authorship
theory of language proposes a single set of rules of the grammar has been
from which all the grammatical sentences in a lan- ascribed to Claude Lance-
*3?*
guage can be derived (p. 97). In order to define lot (161 5-95) and Antoine

these rules in an accurate and economical way, a Amauld (1612-94). Its sub-
title, referring to 'that which
*
grammar has to rely on certain general principles is common to all lan-
- abstract constraints that govern the form it takes guages, and their principal
and the nature of the categories with which it oper- differences provides a
. .
.'
A PARIS,
ates. In this approach, these principles are con- neat summary of the cur-
rent preoccupation with
Chez Pierre le Petit, Imprimcur 8c
ceived as universal properties of language — Libraire ordinaire du Roy, rue S. Jacques,
universals and typology.
properties that are biologically necessary and thus However, the approach of a la Croix d'Or.
innate (p. 234). The notion of universals is impor- modern linguistics is less

tant, it is argued, not only because it deepens our concerned with how lan- M. DC. LX.
understanding of language in its own right, but guage relates to logic and A* tc Pri»ilt£e def* Majtjic.
reality, and more with its
because it provides an essential first step in the task arbitrary properties.
of understanding human intellectual capacity.

84 •
III THE STRUCTURE OF LANGUAGE
BREADTH OR DEPTH? — which can be given statistical expression. For
The distinction between typological and universa- example, in over 99% of languages whose word
lis! approaches to language study is doubtless order has been studied, grammatical subjects pre-
ultimately an arbitrary one; and both have con- cede objects. And in a phonological study of over
siderable insights to offer. But the two approaches, 300 languages (p. 165), less than 3% have no nasal
as currently practised, differ greatly in their proce- consonant. Linguistic features that are statistically
dures. Typologists typically study a wide range of dominant in this way are often referred to as
languages as part of their enquiry, and tend to make 'unmarked'; and a grammar that incorporates How many
generalizations that deal with the more observable norms of this kind is known as a 'core grammar'.
languages?
aspects of structure, such as word order, parts of
it is impossible in principle to

speech, and types of sound. In contrast with the study all human languages,
empirical breadth of such studies, universalists rely in order to find out about

on in-depth studies of single languages, especially universals, for the simple

in the field of grammar — English, in particular, Three types of universals reason that many languages
are extinct, and there is no
is a commonlanguage of exemplification — and Substantive
Substantive universals comprise the set of categories way of predicting what lan-
tend to make generalizations about the more guages will emerge in the
thatis needed in order to analyse a language, such
abstract, underlying properties of language. as 'noun', 'question', 'first-person', 'antonym', and future. To be practical, typo-

This focus on single languages might at first seem logical or universal studies
Do all languages have nouns and vowels? The
'vowel'.
answer seems to be yes. But certain categories often therefore need to be based
strange. If we are searching for universals, then
thought of as universal turn out not to be so: not all on a sample of the 4,000 or
surely we need to study many languages? Chomsky so current languages of the
languages have case endings, prepositions, or future
argues, however, that there is no paradox. Because world (§47). But how should
tenses, for example, and there are several surprising
English is a human language, it must therefore limitations on the range of vowels and consonants that a representative sample be
incorporate all universal properties of language, as typically occur (§28). Analytical considerations must
achieved?
also be borne in mind. Do all languages have words? Several projects on lan-
well as those individual features that make it speci-
The answer depends on how the concept of 'word' is guage universals have had
fically 'English'. One way of finding out about these defined (p. 91). to address this basic ques-
properties, therefore, is the detailed study of single tion. The aim is to include as
languages. The more languages we introduce into Formal many different kinds of lan-

Formal universals are a set of abstract conditions that guage as possible. Lan-
our enquiry, the more difficult it can become to
govern the way in which a language analysis can be guages are selected from the
see the central features behind the welter of indivi- main branches every lan-
of
made- the factors that have to be written into a
dual differences. grammar, if it is to account successfully for the way guage as
family, insofar
On the other hand, it can be argued that the sentences work in a language. For example, because these are known. They are
detailed study of single languages is inevitably all languages make statements and ask related
not selected from the same
questions (such as The car is ready vs Is the car local geographical area, in
going to produce a distorted picture. There are fea-
ready?), some means has to be found to show the case they display a high de-
tures of English, for example, that are not com- gree of mutual influence.
relationship between such pairs. Most grammars
monly met with in other languages, such as the derive question structures from statement structures And the number of lan-
use of only one inflectional ending in the present by some kind of transformation (in the above example, guages within each family
'Move the verb to the beginning of the sentence'). If has to be carefully consi-
tense (third-person, as in she runs), or the absence
it is claimed that such transformations are necessary
dered. It would not be right
of a second-person singular/plural distinction (cf. to select an arbitrary five lan-
in order to carry out the analysis of these (and other
French tu/vous). Without a typological perspec- kinds of) structures, as Chomskyan theory does, then guages from each family -
tive, some say, it is not possible to anticipate the they would be proposed as formal universals. Other bearing in mind that Indo-
Pacific, for example, has
extent to which our sense of priorities will be upset. cases include the kinds of rules used in a grammar,
or the different levels recognized by a theory (§13). over 700 languages, where-
If languages were relatively homogeneous entities,
as Dravidian has only about
like samples of iron ore, this would not be a prob- 20 (§52). The languages of
Implicational
lem. But, typologists argue, languages are unpre- Implicational universals always take the form 'If X, then New Guinea ought, statisti-

dictably irregular and idiosyncratic. Under these being to find constant relationships cally speaking, to constitute
Y', their intention

circumstances, a focus on breadth, rather than between two or more properties of language. For about 20% of any sample.
example, three of the universals proposed in a list of surveys have
In practice,
depth, is desirable. tobe satisfied with what they
45 by the American linguist, Joseph Greenberg
(1915- ) are as follows: can get. As few of the New
Guinea languages have
Universal 1 7. With overwhelmingly more-than-chance been studied in depth, for in-
Relative or absolute? frequency, languages with dominant order VSO stance, currently imprac-
it is
[= Verb— Subject-Object] have the adjective after the ticable to achieve the target
The is to be able to make succinct
universalist ideal noun. of 20%. For such reasons,
and statements that hold, without
interesting Universal 31. If either the subject or object noun agrees even the largest surveys
exception, for all languages. In practice, very few with the verb in gender, then the adjective always
work under considerable
limitations. For example, in
such statements can be made: the succinct ones agrees with the noun in gender.
an American study of phono-
often seem to state the obvious (e.g. all languages Universal 43. If a language has gender categories in
logical universals (§28), the
have vowels); and the interesting ones often seem the noun, it has gender categories in the pronoun. database was provided by a
to require considerable technical qualification. total of only 31 7 languages -
As is suggested by the phrasing, implicational
less than 1 0% of the whole.
Most of the time, in fact, it is clear that 'absolute' statements have a statistical basis, and for this reason
are sometimes referred to as 'statistical' universals But the study nonetheless
(or exceptionless) universals do not exist. As a provided an enormous
(though this is a somewhat different sense from that
result, many linguists look instead for trends or amount of valuable informa-
used in §15).
tendencies across languages - 'relative' universals tion (I. Maddieson, 1984).

14 TYPOLOGY AND UNIVERSALS •

85
 .

15 The statistical structure of language

Within any level of linguistic structure (§13), it is occurrence of the letters of the alphabet. Here is
possible to count the different units that occur, and a selection of frequency orders found in one com-
interrelate the frequencies we obtain, to see if there parative study of different styles of American Eng-
are statistical governing their use.
regularities lish (after A. Zettersten, 1969, p. 21): (a) press
Many aspects of grammar, vocabulary, sound sys- reporting, (b) religious writing, (c) scientific writ-
tem, and writing system have been studied in this ing, (d) general fiction. The average rank order,
way, and several interesting patterns have emerged. based on a description of 15 categories of text
It has even been possible to propose statistical pro- words, is given as (e). Col-
totalling over a million
perties that are in common to all languages; these umn (f) gives the order used by Samuel Morse
are sometimes referred to as statistical laws or (1791-1872) in compiling the Morse Code. His fre-
'
universals' quency ordering was based on the quantities of
Statistical regularities are independent of speaker type found in a printer's office (see column (g)).
or writer, or subject matter. While in a sense we
are free to say whatever we want, in practice our (b) (d) (e) (f) Monosyllabic
(g)
linguistic behaviour conforms closely to statistical or polysyllabic?
expectations. We can say with confidence that if e e e e e e 12,000
The most frequent words are
we write a q in English, it is almost always going t t t t t t 9,000
monosyllabic. This effect is

to be followed by u (though not always, because a i a a a a 8,000 clearly seen

a study of tele-
in

of Iraq, and other exceptions). Less obviously, but o a i o o i 8,000 phone conversation. There
n o o h n 8,000 were few words of 3 or more
equally confidently, it emerges that just over 60% i

syllables in the 800 most fre-

of everything we say will be made up of consonants, i n n n n o 8,000
quently occurring words.
and just under 40% of vowels. About a third of s s s i s s 8,000 (After N. R. French, et al.,
all the syllables we use in everyday speech will have r r r s r h 6,400 1930.)

the structure of consonant 4- vowel + consonant, h h h r h r 6,200

as in cat. The 50 most commonly used words in 1 1 1 d 1 d 4,400
the language will make up about 45 of everything % d d c 1 d 1 4,000
-T 600
we write. c c d u c u 3,400
The remarkable thing about such facts is that, m u u w u c 3,000
while we are engaged in communication, we do u m m m m m 3,000
not consciously monitor our language to ensure f f f c f f 2,500 a 400
that these statistical properties obtain. It would be P P P g P w 2,000
impossible to do so. Yet, without any deliberate g y g f
g y 2,000
effort on our part, we will find the same underlying w w y y w g 1,700
Z 200 i
regularities in any large sample of our speech or y g b P y p 1,700 Number of
writing. The study of these regularities, and of the b b w b b b 1,600 syllables = 1

factors that constrain them, is the province of sta- V V V k V V 1,200

tistical linguistics. k k k V k k 800
400
X X X q 500
i i
Number of
LETTER FREQUENCY X j q X j i
400 different words
One of the simplest demonstrations of statistical q q j
z q X 400
regularity within a language is the frequency of
z z z q z z 200 Lexical top twenties
The 20 most-frequently
occurring words in stu-
Written Spoken Written Spoken
Rank German dies of newspaper writing
Rank French German English English French English English
in English, French, and
1 de der the the 12 dans mit at yes German are shown (after P.

2 le (a.) die of and 13 il sich he was M. Alexejew et al., 1968). For

3 la (a.) und to I 14 a daR with this comparison, the last column
to 15 en dem by but lists the most-frequent words
4 et in in

des and of 16 ne sie be on in the London-Lund corpus

5 les
des den a a 17 on ist well of spoken conversation (p.
6 it

7 est zu for you 18 qui im an he 411). The importance of the

8 un (a.) das was that 19 au eine as have speech writing distinction is

9 une (a.) von is in 20 se DDR his for evident: note the frequency

10 du fur that it of /, yes, and well in spoken

11 que (p.) auf on IS English, and the occurrence
of DDR (German Democra-
a. article p. pronoun tic Republic) in the German
list.

86 •
III THE STRUCTURE OF LANGUAGE
Zipf s laws example, the most frequent word, I, occurs 5,920
times (r.f = 5,920), and the 100th word, he's,
One of the first demonstrations of the existence occurs 363 times (r.f = 36,300). The size of the
of major statistical regularities in language was car- sample is also a critical factor.
ried out by the American philologist George Nonetheless, the 'standard curve' of word fre-
Kingsley Zipf (1902—50). His best known 'law' quency, summarized as f.r=C is an interesting
proposes a constant relationship between the rank observation about language patterns. Moreover,
of a word in a frequency list, and the frequency the same kind of curve has been found in many
with which it is used in a text. If you want to test languages. For example, in a French word-fre-
the validity of the law, you have to carry out the quency book, the 100th word was used 314 times
following operations: (= 31,400), the 200th, 158 times (31,600), and the
1. Count all the instances of different words in a 1,000th, 31 times (31,000).
text - the 364, 251, table 4, etc.
is

2. Put them in descending rank order of frequency, OTHER RELATIONSHIPS G.K. Zipf (1902-50)
and give each rank a number — (1) the 364, Zipf also showed that there is an inverse relation-
ship between the length of a word and its fre-
(2) K 251, (3) of 166, etc.
3. Multiply the rank number (r) by the frequency quency. In English, for example, the majority of
(f), and the result is approximately constant (C).
the commonly used words are monosyllables. The Syllables
same relationship obtains even in a language like Take a tape recording of
For example, the list below gives the 35th, 45th, German, which has a marked 'polysyllabic' voca- some spoken English, and
55th, 65th, and 75th most-frequently occurring bulary. This effect seems to account for our ten- transcribe Mark the boun-
it.

words in one category of the London— Lund corpus daries between the syllables.
dency to abbreviate words when their frequency You should find that 1 2 syl-
of spoken conversation. The values come out at of use rises, e.g. the routine reduction of micro- lables make up 25% of the
around 30,000 each time. phone to mike by radio broadcasters. It would also speech /5a/. /sv /. /in/,
:

seem to be an efficient communicative principle to /aend/./i/./a/./tu/./in/,/*/.

= C /n/,/it/,/Saet/ (see Appen-

f have the popular words short and the rare words
dix II for transcription). Half
long.
35 very 836 = 29,260 the speech will use only 70
Factors such as efficiency and ease of communi- different syllables. But to ac-
45 see 674 = 30,330
cation appealed strongly to Zipf, who argued for count for 90% of the speech,
55 which 563 = 30,965 you will need to recognize
a principle of 'least effort' to explain the apparent
65 get 469 = 30,485 over 1 ,300 syllable types.
equilibrium between diversity and uniformity in
75 out 422 = 31,650 /da/ alone makes up 7% of
our use of sounds and words. The simpler the allspoken syllables; it turns
sound and the shorter the word, the more often up on average every 14 syll-
In other words, the relationship is inversely propor- will human beings want to use it. There are, how- ables. (After G. Dewey,

it was thought to obtain regardless of

and 1923.)
tional, ever, several difficulties facing this explanation (e.g.
subject matter, author, or any other linguistic vari- how to quantify the 'effort' involved in articulating
able. However, it was subsequently shown that the sounds, and the exceptions to the law referred to
relationship does not obtain for words of highest above), and today a more conventional explana-
and lowest frequencies. In the same corpus, for tion in terms of probability theory is accepted.

Length/frequency relationship
The relationship of syllable length and frequency of
occurrence was charted in a study of nearly 1 1 million
German words (after F. W. Kaeding, 1898). Take a text, any text . .

Take a text, in any language, count for 60%; and the first
Number of and count the words. Order 1 ,000 for 85%. The first

syllables in Number of word Percentage the words in terms of de- 4,000 will account for 97.5%.
word occurrences ofwhole creasing frequency. Accord- In short samples, however,
ing to statistical prediction, considerable variation from
1 5,426,326 49.76
the first 1 5 words will ac- these proportions will be
2 3,156,448 28.94
count for 25% of the text. found.
3 1,410,494 12.93
The first 1 00 words will ac-
4 646,971 5.93
5 187,738 1.72
6 54,436 0.50
7 16,993
8 5,038
9 1,225
10 461 Dictionaries
11 59 0.22
Take a dictionary and count the meanings of each
12 35
word, as indicated by the sub-entries. The number of
13 8
words (n) that have a particular number of meanings
14 2
(m) is inversely proportional to the square of the
15 1
number of meanings (nm 2 = C).

15 THE STATISTICAL STRUCTURE OF LANGUAGE • 87

 16 Grammar

It is difficult to capture the central role played by

grammar in the structure of language, other than Six types of grammar Two of the most important
by using a metaphor such as 'framework' or 'skel- Descriptive grammar An approach that describes symbols in modem gramma-
eton'. But no physical metaphor can express the grammatical constructions that are used in a tical analysis. An asterisk is
language, without making any evaluative judgments placed before a construction
satisfactorily the multifarious kinds of formal
about their standing in society. These grammars are to show that it is ungram-
patterning and abstract relationship that are commonplace in linguistics, where it is standard matical. A question-mark
brought to light in a grammatical analysis. practice to investigate a corpus' of spoken or written shows that the construction
Two steps can usually be distinguished in the material, and to describe in detail the patterns it is of doubtful grammaticality.
contains (p. 410).
study of grammar. The first step is to identify units For example, there is no
in the stream of speech (or writing, or signing) - doubt about the ungrammati-
Pedagogical grammar A book specifically designed cality of
units such as 'word' and 'sentence'. The second forteaching a foreign language, or for developing an
step is to analyse the patterns into which these units awareness of the mother tongue. Such teaching *Who and why came in?
grammars' are widely used in schools, so much so that *That book looks alike.
fall,and the relationships of meaning that these
patterns convey. Depending upon which units we many people have only one meaning for the term But the status of the follow-
'grammar': a grammar book. ingsentences is less certain.
recognize at the beginning of the study, so the defi-
Both are in use, yet there is
nition of grammar alters. Most approaches begin Prescriptive grammar A manual that focusses on something odd about them.
by recognizing the 'sentence', and grammar is thus constructions where usage is divided, and lays down
rules governing the socially correct use of language ?Don't forget yours and my
most widely defined as 'the study of sentence struc- books.
(§1 ). These grammars were a formative influence on
ture'. A grammar of a language, from this point ?This is the car of the family.
language attitudes in Europe and America during the
of view, is an account of the language's possible 1 8th and 1 9th centuries. Their influence lives on in the
One of the main aims of
sentence structures, organized according to certain handbooks of usage widely found today, such as A linguistic analysis is to dis-
general principles. For example, in the opening Dictionary of Modern English Usage (1926) by Henry cover the principles enabling
Watson Fowler (1 858-1 933). us to decide the grammati-
pages of the most influential grammatical treatise
cality of a sentence.
of recent times, the American linguist Noam Reference grammar A grammatical description that
Chomsky (1928— ), writes that a grammar is a be as comprehensive as possible, so that it
tries to
'device of some sort for producing the sentences can act as a reference book for those interested in So much grammar
of the language under analysis' (1957, p. 11), to establishing grammatical facts (in much the same way in a language
as a dictionary is used as a reference lexicon' (§18)).
which is added the rider that the sentences pro- Probably the largest gram-
Several north European grammarians compiled
duced must be grammatical ones, acceptable to the handbooks of this type in the early 20th century, the mar produced for any lan-
native speaker. best known being the seven-volume Modern English guage: A Comprehensive
Grammar (1909-49) by the Danish grammarian Otto Grammar of the English Lan-
Within this general perspective there is room for guage (1 985), by Randolph
Jespersen (1860-1943), and A Comprehensive
many different positions. In particular, there are
Grammar of the English Language 1 985) by Randolph
(
Quirk, Sidney Greenbaum,
two quite distinct applications of the term 'gram- Quirk (1920- )etal.
Geoffrey Leech, and Jan
mar', yielding a specific sense and a general one. Svartvik. The amount of de-
tail in its 1 ,779 pages comes
The specific sense is the more traditional: here, Theoretical grammar An approach that goes
beyond the study of individual languages, to determine as a surprise to many people
grammar is presented as just one branch of lan- who, because of the traditio-
what constructs are needed in order to do any kind
guage structure, distinct from phonology and of grammatical analysis, and how these can be applied
nal focus on grammar as a

semantics. This is the approach used in this encyc- matter of word-endings,

consistently in the investigation of a human language.
It is thus a central notion in any investigation of
have been brought up to
lopedia (§13):
think of English as a lan-
linguistic universal (§14).
language structure guage lacking in grammar.
grammar A term But this book stands on the
Traditional often used to
shoulders of even more
summarize the range of attitudes and methods found
detailed treatments of areas
phonology grammar in the period of grammatical study before the advent
of the language; for example,
of linguistic science (§65). The tradition' in question
a and the alone have war-
is over 2,000 years and includes the work of
old,
The general sense of the term, popularized by ranted a 200-page study (P.
classical Greek and Roman grammarians,
Chomsky, subsumes all aspects of sentence pattern- Renaissance writers, and 1 8th-century prescriptive Christopherson, 1939).
ing, including phonology and semantics, and intro- grammarians. It is difficult to generalize about such a
duces the term 'syntax' as the more specific notion: wide variety of approaches, but linguists generally use
the term pejoratively, identifying an unscientific
approach to grammatical study, in which languages
grammar
were analysed in terms of Latin, with scant regard for
empirical facts. However, many basic notions used by
modem approaches can be found in these earlier
phonoiogy- writings, and there is now fresh interest in the study
of traditional grammar, as part of the history of
linguistic ideas.

88 • III THE STRUCTURE OF LANGUAGE

PARSING vs CREATING language learner, or the translator. Grammar need Poodles wearing
Traditional grammars taught people to 'parse', or not be dry, unreal, arcane; it can be alive, relevant, jeans?
analyse, a sentence, by making a series of divisions entertaining. As with so many subjects, it depends It is not difficult to think up
within it. The man saw the cow, for example, only on how it is put across. dramatic or entertaining sen-
would be divided into a 'subject' {the man), and tences that would motivate a
child to carry out a grammati-
a 'predicate' {saw the cow). The predicate would
cal analysis, because of their
then be divided into its verb {saw) and the 'object'
Grammatical nature rambles? ambiguity or stylistic effect.
{the cow). Other divisions would be made until Here are some nice cases of
Imagine teaching a child about the structure of a flower
all had been identified.
the features of the sentence ambiguity, taken from W. H.
in the following way. A hypothetical plant is drawn on
Mittins, A Grammar of Mod-
It is an approach to language that many people the board, and its parts labelled: stamen, pistil, stalk,
ern English (1 962), all of
recall with distaste. Grammar, for them, was a dry, etc. Each term is defined, and the children write them
which can be explained
boring, and frustrating subject. Why should this in their books. They have to learn them off by heart,
through a single principle:
have been so? and until they do they will not be allowed to see or
work with any real plant! The girl was followed by a
There were several reasons. All too often, in the It is unlikely that anyone ina modern biology class small poodle wearing jeans.
traditional grammars, insufficient reasons were would be taught this topic through such an approach. Next came a mother with a
given for making a particular sentence analysis. As The teacher would arrive armed with real plants, and very small baby who was
a consequence, it was common to find children give them out; then the children would search for the pushing a pram.
parts, all the while meeting problems, and asking for
learning analyses and definitions off by heart, with- I always buy my newspapers
help with the labels as they went along. Later, the
out any real understanding of what was going on. teacher would get them to write up their project in a at the shop next to the police
In particular, they had to master the cumbersome, book, and then might ask for some terms to be learned. station in which cards, maga-
That is the modern way: discovery first, definitions zines, and fancy goods are
Latin-based grammatical terminology as an end in
of terms last. But grammar continues to suffer, in many displayed.
itself (terms such as 'accusative', 'complement',
schools, by being taught the other way round (when Asailor was dancing with a
'apposition'), and apply it to examples of language it is taught at all!). A hypothetical sentence is put on wooden leg.
that were either artificially constructed, or taken the board, and the required grammatical terminology
In each case, the construc-
from abstruse literature. It was all at a considerable has to be learned, before any attempt is made to
tion at the end of the sen-
remove from the child's real language world, as grapple with real sentences in a real world. Often,
tence has been separated
even, no attempt at all is made to go searching for
found in conversation or the media. Little attempt from the noun to which be- it
interesting, real, sentence specimens. If is as if the
was made to demonstrate the practical usefulness longs. If one wished to avoid
children's knowledge of plants were to remain forever
the unintentionally humorous
of grammatical analysis in the child's daily life, solely on the blackboard. No one would tolerate such
effects, the sentences would
whether in school or outside. And there was no a silly pedagogical approach for biology. But for many
need to be reformulated with
decades, just such an approach was actively practised
interest shown in relating this analysis to the this construction immediately
for grammar- and it is by no means extinct.
broader principles of grammatical patterning in the following, or postmodifying',

language as a whole. It is not surprising, then, that the noun ('The girl wearing
jeans .').
most people who were taught parsing in school
. .

ended up unable to see the point of the exercise,

and left remembering grammar only as a dead, irre-
levant subject. A page from Maureen Vldler's Find a Story (1974) The
page is cut horizontally, so that as each strip is turned
The reality is quite the opposite. The techniques over, a new sentence and picture results. On the next
of grammatical analysis can be used to demonstrate page, for instance, the top strip reads 'My silly cousin Nell',
the enormous creative power of language — how, and the bottom one along the sea shore'. There are only
1 2 pages in the book, but there are over 20,000 possible
from a finite set of grammatical patterns, even a
grammatical combinations. In a similar approach ('Roll a
young child can express an infinite set of sentences.
Story'), the child makes a sentence by rolling a series of
They can help us all to identify the fascinating blocks on which words have been printed. Such
'edges' of language, where grammaticality shades approaches are an entertaining means of drawing young
into ungrammaticality, and where we find the children's attention to sentence structure. Grammar can,
many kinds of humorous and dramatic effects, both at times, be fun.

in literature and in everyday language (p. 72). As

we discover more about the way we each use gram-
mar as part of our daily linguistic survival, we inevi-
tably sharpen our individual sense of style, and thus
promote our abilities to handle more complex con-
structions, both in speaking/listening and in read-
ing/writing. We become more spot likely to
ambiguities and loose constructions, and to do
something about it. Moreover, the principles of
grammatical analysis are general ones, applicable
to the study of any language, so that we find our-
selves developing a keener sense of the similarities
and differences between languages. And many
kinds of specialized problems can be illuminated
through the study of grammar — such as the difficul-
ties facing the language-handicapped, the foreign-

16 GRAMMAR -89
Basic grammatical notions know how to analyse irregular nouns and verbs: New words out
feet the plural of foot, but it is not obvious how
is of old
The range of constructions that is studied by gram- to identify a plural morpheme in the word, analo-
There are four normal pro-
mar is very large, and grammarians have often gous to the -5 ending of horses. In the Turkish word cesses of word formation in
divided it into sub-fields. The oldest and most evinden 'from his/her house', there is the opposite English:
widely-used division is that between morphology problem, as can be seen from the related forms: • prefixation a prefix is
and syntax. placed before the base of the
ev house
word, e.g. disobey,
evi his/her/its house • suffixation a prefix is
evden from the house placed after the base of the
Morphology It -i ending marks 'his/her/its', and
seems that the
word, e.g. kindness;
• conversion a word
This branch of grammar studies the structure of the -den ending marks 'from' — in which case the changes its class without any
words. In the following list, all the words except combination of the two ought to produce eviden. change of form, e.g. (the)
the last can be divided into parts, each of which But the form found in Turkish has an extra n, which carpet (noun) becomes (to)
carpet (verb); and
has some kind of independent meaning. does not seem to belong anywhere. Its use is auto-
• compounding two base
matic in this word (in much the same way as an forms are added together,
unhappiness un- -happi- -ness
extra r turns up in the plural of child in English e.g. blackbird.
horses horse- -s
— child-r-en). Effects of this kind complicate mor- There are also some less
talking talk- -ing usual ways of making new
phological analysis — and add to its fascination.
yes yes words:
To those with a linguistic bent, there is nothing • reduplication a type of
Yes has no internal grammatical structure. We more intriguing than the search for regularities in compound in which both ele-
could analyse its constituent sounds, /')/, /e/, /s/, a mass of apparently irregular morphological data. ments are the same, or only
but none of these has a meaning in isolation. By Another complication is that morphemes some- slightly different, e.g. goody-

contrast, horse, talk, and happy plainly have a goody, wishy-washy, teeny-
times have several phonetic forms, depending on
weeny,
meaning, as do the elements attached to them (the the context in which they occur. In English, for • clippings an informal
'affixes'): un- carries a negative meaning; -ness example, the past-tense morpheme (written as -ed), shortening of a word, often to
expresses a state or quality; -s expresses plural; is pronounced in three different ways, depending a single syllable, e.g. ad,

and -ing helps to convey a sense of duration. The gents, flu, telly,
on the nature of the sounds that precede it. If the
• acronyms words formed
smallest meaningful elements into which words can preceding sound is /t/ or /d/, the ending is pro- from the initial letters of the
be analysed are known as morphemes; and the way nounced /id/, as in spotted; if the preceding sound words that make up a name,
morphemes operate in language provides the sub- is a voiceless consonant (p. 128), the ending is pro- e.g. nato, unesco, radar
ject matter of morphology. nounced /t/, as in walked; and if the preceding (= radio detection and rang-
ing); a sub-type is an alpha-
It is an easy matter to analyse the above words sound is a voiced consonant or a vowel, the ending
betism, in which the different
into morphemes, because a clear sequence of ele- is pronounced /d/, as in rolled. Variant forms of
letters are pronounced, e.g.
ments is involved. Even an unlikely word such as a morpheme are known as allomorphs. VIP, DJ-, and
antidisestablishmentarianism would also be easy • blends two words merge
to analyse, for the same reason. In many languages into each other, e.g. brunch
(from 'breakfast' + 'lunch'),
(the so-called 'agglutinating' languages (p. 293)), INFLECTIONAL AND DERIVATIONAL telex ('teleprinter' + 'ex-
it is quite normal to have long sequences of mor-
Two main fields are traditionally recognized within change').
phemes occur within a word, and these would be morphology. Inflectional morphology studies the
analysed in the same way. For example, in Eskimo way in which words vary (or 'inflect') in order to Abso-blooming-
the word angyaghllangyugtuq has the meaning 'he express grammatical contrasts in sentences, such lutely
wants to acquire a big boat'. Speakers of English as singular/plural or past/present tense. In older
Morphemes can be classi-
find such words very complex at first sight; but grammar books, this branch of the subject was fied into 'free' and 'bound'
things become much clearer when we analyse them referred to as 'accidence'. Boy and boys, for exam- forms. Free morphemes can
into their constituent morphemes: ple, aretwo forms of the 'same' word; the choice occur as separate words,
e.g. car, yes. Bound mor-
angya- 'boat' between them, singular vs plural, is a matter of
phemes cannot occur on
-ghlla- an affix expressing augmentative meaning grammar, and thus the business of inflectional mor- their own, e.g. anti-, -tion.

-ng- 'acquire' phology. Derivational morphology, however, stu- The main classes of bound
-yug- an affix expressing desire dies the principles governing the construction of morphemes are the prefixes
new words, without reference to the specific gram- and suffixes; but infixes are
-tuq- an affix expressing third person singular. - an
also possible affix which
matical role a word might play in a sentence. In
English has relatively few word structures of this is inserted within a stem. The
the formation of drinkable from drink, or disinfect nearest we get to this
Eng- in
type, but agglutinating and inflecting languages,
from infect, for example, we see the formation lish is emphatic forms such
such as Turkish and Latin, make widespread use of different words, with their own grammatical as abso-blooming-lutely aw-
of morphological variation. A verb in the African fut, but in many languages,
properties.
language, Bilin, could appear in over 10,000 vari- infixation is a normal mor-
phological process. In Taga-
ant forms.
log, for example, the form
/urn/ 'one who does' is in-

MORPHEME PROBLEMS fixed within the form/pi:lit/

'effort' to produce/pumhlit/,
Not all words can be analysed into morphemes which means one who com-
so easily. In English, for example, it is difficult to pelled'.

90 •
III THE STRUCTURE OF LANGUAGE
 Even in very hesitant speech, pauses come at There are no word spaces in
Words talk.
intervals -
usually between major grammatical
the 4th century ad Greek
Codex Sinaiticus. These were
Words sit uneasily at the boundary between mor- units, such as phrases or clauses (p. 95). So if there introduced by the Romans.
phology and syntax. In some languages - 'isolating' are no audible 'spaces', how do we know what
K*feTicecTlHonx
languages, such as Vietnamese (p. 293) — they are the words are? Linguists have spent a great deal fAAiKOYcceTOTT'
plainly low-level units, with or no internal
little of time trying to devise satisfactory criteria - none oYNi^u>Norieif-
structure. In others - 'polysynthetic' languages, of which is entirely successful. eiTieNTUJiYOTT-
such as Eskimo — word-like units are highly com-
plex forms, equivalent to whole sentences. The con-
Five tests of word identification
cept of 'word' thus ranges from such single sounds
Potential pause this criterion is not perfect ple, long words generally
as English a to palyamunurringkutjamunurtu
Say a sentence out loud, such
either, in the light of have their stress on the
('he/she definitely did not become bad') in the and ask someone to re- penultimate syllable, e.g.
forms as absoblooming-
Western Desert language of Australia. peat it very slowly, with lutely. 'cartrefhome', car'trefi
Words are usually the easiest units to identify, pauses'. The pauses will 'homes'. In Turkish, the

language. In most languages, they tend to fall between words, Minimal free forms vowels within a word har-
in the written
and not within words. For The American linguist Leo- monize in quality (p. 161),
are the entities that have spaces on either side. (A
so that if there is a marked
example, the/ three/ little nard Bloomfield (1887-
few languages use word dividers (e.g. Amharic), /pigs/ went/ to/ market. 1 949) thought of words as change in vowel quality in
and some do not separate words at all (e.g. Sans- But the criterion is not fool- 'minimal free forms' - that the stream of speech, a
krit).) Because a literate society exposes its proof, for some people will is, the smallest units of new word must have be-
break up words containing speech can meaning- gun. But there are many
members to these units from early childhood, we that
exceptions to such rules.
more than one syllable, fully stand on their own.
all know where to put the spaces — apart from
e.g. mar/ket. This definition does handle
a small number of problems, mainly to do with the majority of words, but it Semantic units
hyphenation. Should we write washing machine or Indivisibility cannot cope with several In the sentence Dog bites
Say a sentence out loud, items which are treated as vicar, there are plainly
should it be washing-machine} Well informed or
and ask someone to add words in writing, but which three units of meaning, and
well-informed} No one or no-one} each unit corresponds to a
extra words' to it. The extra never stand on their own in
It is more difficult to decide what words are in items will be added be- word. But language is often
natural speech, such as
the stream of speech, especially in a language that tween the words and not English the and of, or not as neat as this. In /
has never been written down. But there are prob- within them. For example, French je (T) and de ('of'). switched on the light, the
the pig went to market has little clear meaning',
lems, even in languages like English or French. Cer-
might become the big pig Phonetic boundaries and the single action of
tainly, it is possible to read a sentence aloud slowly, 'switching on' involves two
once went straight to the It is sometimes possible to

so that we can 'hear' the spaces between the words; market, but we would not tell from the sound of a words.
but this is an artificial exercise. In natural speech, have such forms as pi-big- word where it begins or
pauses do not occur between each word, as can g or mar-the-ket. However, ends. In Welsh, for exam-

be seen from any acoustic record of the way people

WORD CLASSES 'name' of a colour? In place of definitions based Classifying nouns

Since the early days of grammatical study, words on meaning, there is now a focus on the structural Distinctionssuch as mascu-
have been grouped into word classes, traditionally features that signal the way in which groups of line/feminine and human/
labelled the 'parts of speech'. In most grammars, words behave in a language. In English, for exam- non-human are well known
up sub-classes of
in setting
eight such classes were recognized, illustrated here ple, the definite or indefinite article is one criterion
nouns, because of their
from English: that can be used to signal the presence of a follow- widespread use in European
ing noun (the car); similarly, in Romanian, the arti- languages. But many Indo-
nouns boy, machine, beauty cle (ul) signals the presence of a preceding noun and African lan-
Pacific
guages far exceed these in
pronouns she, it, who (avionul 'the plane').
the number of noun classes
adjectives happy, three, both Above all, the modern aim is to establish word they recognize. In Bantu lan-
verbs go, frighten, be classes that are coherent: all the words within a guages, for example, we find
prepositions in, under, with class should behave in the same way. For instance, such noun classes as human
jump, walk, and cook form a coherent class, beings, growing things, body
conjunctions and, because, if
parts, liquids, inanimate ob-
adverbs happily, soon, often because all the grammatical operations that apply
jects, animals, kinship
interjections gosh, alas, coo to one of these words apply to the others also: names, abstract ideas, arte-
they all take a third person singular form in the facts, and narrow objects.
In some classifications, participles {looking, taken) present tense [he jumps / walks / cooks), they all However, these labels
have a past tense ending in -ed {jumped/ walked/ should be viewed with cau-
and were separately listed.
articles (a, the)
tion, as they are no more ex-
Modern approaches classify words too, but the cooked), and so on. Many other words display the
act semantically than are the
use of the label 'word class' rather than 'part of same (or closely similar) behaviour, and this would gender classes of European
speech' represents a change in emphasis. Modern lead us to establish the important class of 'verbs' languages. In Swahili, for ex-
linguists are reluctant to use the notional defini- in English. Similar reasoning would lead to an ana- ample, there are sub-classes
for human beings and in-
tionsfound in traditional grammar — such as a logous class being set up in other languages, and
sect/animal names, but the
noun being the 'name of something'. The vagueness ultimately to the hypothesis that this class is generic words insect' and
of these definitions has often been criticized: is required for the analysis of alllanguages (as a 'sub- 'animal' in fact formally be-
beauty a 'thing'? is not the adjective red also a stantive universal', §14). long to the 'human' class!

16 GRAMMAR -91
 -

Gradience What part of speech Noun tenses?

Word classes should be coherent. But we do not
if is round? Some languages formally mark the expression of time
want to up hundreds of classes,
set we have to You cannot tell what on word classes other than the verb. In
relations
class a
Japanese, adjectives can be marked in this way, e.g.
let some irregular forms into each one. For exam- word belongs to simply by
looking at it. Everything de- was white', shirokute being
shiroi white', shirokatta
ple, for many speakers house is the only English
pends on how the word Potowatomi, the same ending that
white', etc. In
'be-
noun ending in/ s/, where the/s/ becomes/ z/ when expresses past time on verbs can be used on nouns
haves' in a sentence. Round
the plural ending is added {houses). Although in is a good illustration of this
also:

theory it is 'inown', in practice it

a class of its principle in action, forcan it /nka§atas/ I am happy
is grouped with other nouns, with which it has belong to any of five word /nkasatsapan/ I was once happy
a great deal in common. classes, depending on the /nos-/ my father
grammatical context. /nosp-an/ my dead father
Because of the irregularities in a language, word /nciman/ my canoe
classes are thus not as neatly homogeneous as the Adjective /ncimanpan/ my former canoe (lost, stolen)
theory implies. Each class has a core of words that Mary bought a round table. (After C. F. Hockett, 1958, p. 238.)

behave identically, from a grammatical point of Preposition

view. But at the 'edges' of a class are the more The car went round the Five moods
irregular words, some of which may behave like corner. A range of attitudes can be expressed by the mood
words from other classes. Some adjectives have a Verb system of the verb. In Fox, one mood expresses the
The yacht round the buoy
meaning 'God forbid that this should happen!; another,
function similar to nouns (e.g. the rich); some will
'What if it did happen! What do care!' In Menomini, I

nouns behave similarly soon.

to adjectives (e.g. railway there is a five-term mood system:
is used adjectivally before station). Adverb
/pi-w/ he comes/is coming/ came
The movement from a central core of stable We walked round to the
/pi -wen/ he is said to be coming/it is said that he
shop.
grammatical behaviour to a more irregular peri- came
phery has been called gradience. Adjectives display Noun /pi'V is he coming?/did he come?
your round.
It's have a /piasah/ so he is coming after all!
this phenomenon very clearly. Five main criteria
I'll

whiskey. /piapah/ but he was going to come! (and now it turns

are usually used to identify the central class of Eng-
out he is not)
lish adjectives: (After C. F. Hockett, 1 958, p. 237.)
(A) they occur after forms of to be, e.g. he's sad;
A dustbin class? Dual and number
(B) they occur after articles and before nouns, e.g. trial
Several of the traditional
the big car; Four numbers are found in the language spoken on
parts of speech lacked the
(C) they occur after very, e.g. very nice; Annatom Island (Melanesia): singular, dual, trial, plural
coherence required of a well-
(excl./incl. = exclusive/ inclusive of speaker):
(D) they occur in the comparative or superlative defined word class - notably,
form, e.g. sadder/ saddest, more/ most impres- the adverb. Some have lik- /ainjak/ I

ened this class to a dustbin, /aijumrau/ we two (excl.)

sive; and /aijumtai/ we three (excl.)
into which grammarians
(E) they occur before -ly to form adverbs, e.g.
would place any word whose /aijama/ we (excl.)
quickly. grammatical status was un- /akaijau/ we two (incl.)
/akataij/ we three (incl.)
We can now use these criteria to test how much clear. Certainly, the following
words have very /akaija/ we (incl.)
little struc-
like an adjective a word is. In the matrix below, /aiek/ you (sing.)
turally in common, yet all
candidate words are listed on the left, and the five have been labelled adverb' /aijaurau/ you two
criteria are along the top. If a word meets a criter- in traditional grammars: /aijautaij/ you three
ion, it is given a + sad, for example, is clearly
/aijaua/ you (pi.)
;
tomorrow very no (After L. Bloomfield, 1933, p. 257.)
an adjective {he's sad, the sad girl, very sad, sadder/
however quickly when
saddest, sadly). If a word fails the criterion,
—
it is not just the A fourth person
given a (as in the case of want, which is nothing
A fourth-person contrast is made in the Algonquian
likean adjective: "'he's want, '"the want girl, *very
The, an adverb? In such con-
languages, referring to non-identical animate third
want, * wanter / wantest, *wantly). persons in a particular context. In Cree, if we speak
texts as The more the mer-
of a man, and then (secondarily) of another man, the
rier.
A B C D E forms are different: /'na:pe:w/vs/'na:pewa/. This
fourth person form is usually referred to as the
happy + + + + + 'obviative'.

old + + + + - (After L. Bloomfield, 1 933, p. 257.)

top + + + - -
two + + — — — Fifteen cases
asleep + - - - - Nominative (subject), genitive (of), accusative (object),

want — — — — — inessive (in), elative (out of), illative (into), adessive

(on), ablative (from), allative (to), essive (as), partitive
(part of), translative (change to), abessive (without),
The pattern in the diagram is of course wholly arti- and comitative (with).
instructive (by),
ficial because it depends on the way in which the The
Finnish case system seems fearful to those
criteria are placed in sequence; but it does help brought up on the six-term system of Latin. But the
less familiar cases are really quite like prepositions
to show the gradual nature of the changes as one
except that the forms are attached to the end of the
moves away from the central class, represented by noun as suffixes, instead of being separate words
happy. Some adjectives, it seems, are more adjec- placed before, as in English.
tive-like than others.

92 •
III THE STRUCTURE OF LANGUAGE
GRAMMATICAL CATEGORIES
In many languages, the forms of a word vary, in
order to express such contrasts as number, gender,
and tense.These categories are among the most
familiar of grammatical concepts, but their ana-
all
lysis can lead to surprises. In particular, it emerges
that there is no neat one-to-one correspondence
between the grammatical alterations in a word's
form and the meanings thereby conveyed. Plural
nouns do not always refer to 'more than one'; a
first-person pronoun does not always refer to the
person who is talking; and masculine nouns are
not always male.

Category Typical formal Typical mean- Examples But note

contrasts ings conveyed
aspect.
(verbs) perfect(ive), completeness, Russian ya vs on c7fa/(impf.),
pro6ital(p\.) Adverbs can change the meaning, as
imperfect(ive) habituality, con- roughly 'I used to read/was
read' vs 'I when always changes the in progress'
tinuousness, reading'; English she sings (as a job) vs meaning of John is driving from London to
duration, progres- she's singing (now). a habitual (and often irritated) meaning:
siveness John's always driving from London.

(nouns, pronouns, nominative, voca- actor, possession, English gen. boy's, girls'; Latin nom. puella Cases may have several functions. The
adjectives) tive,accusative, naming, location, gen. puellae 'of the girl'; Serbo-Croat
'girl', English genitive is sometimes called the
genitive, partitive motion towards grad 'town', loc. gradu at a town'. possessive', but it can express other
meanings than possession, e.g. the man's
release, a week's leave, a summer's day.
gender.
(nouns, verbs, ad- masculine, femi- male, female, Spanish masc. elmuchacho 'boy', fern, la There is no necessary correlation between
jectives) nine, neuter, ani- sexless, living muchacha 'girl'; German masc. derMann grammatical gender and sex. In German,
mate, inanimate 'the man', fern, die Dame 'the lady', neut. spoon' is masculine (derLoffel); 'fork' is
das Ende 'the end'; Russian past tense feminine (die Gabef); 'knife' is neuter {das
singular masc. dital, fern, iitala, neut. titalo Messer). French 'love' amour is masculine
'read'. in the singular, but often feminine in the

plural.

mood.
(verbs) indicative, sub- factuality, possibi- Latin regu/esc/f 'he/she/it rests' vs Although a major section in traditional
junctive, optative lity, uncertainty, requiescat 'may he/she rest'; English God grammars, many European languages no
likelihood save the Queen, if I were you. longer make much use of the subjunctive.
It is often restricted to formulaic phrases
or very formal situations.
number.
(nouns, verbs, singular, dual, one, two, more Swedish oil"car', hilar 'cars'; Dutch ikroep Nouns plural in form may refer to singular
pronouns) trial, plural than one, more 'I call', wij roepen 'we call'; Samoan/'oe/ entities (e.g. binoculars, pants), and some
than two, more 'you' (sing.),/'oulua/ 'you two'./'outou/ nouns functioning as singulars refer to
than three 'you' (pi.). several events (e.g. athletics, news). The
two crops known as wheat and oats look
very similar; but in English one is singular
and the other is plural.
person
(pronouns, verbs) first person, speaker, addres- Welsh mi'Y, n/'we'; Menomini/nenah/T, First person can refer to addressee (Doc-
second person, see, third party, /kenah/'thou',/wenah/'he'; Latin amo'l tor (to patient) : How are we today?) or to
third person, fourth party love', amas 'you love' (sing.), amat 'he/ a third party (Secretary (to friend, about the
fourth person she/ it loves'. boss) We 're not in a good mood today)
:

Third person can refer to self (Wife: How's

my husband? Husband: He's hungry).
tense
(verbs) present, past, present time, past Italian io parlo 'I speak', io ho parlato 'I Tense and time do not always correspond.
future time, future time have spoken', ioparlavo 'I was speaking'; Present tense-past time: Minister dies
Gaelic chuala mi'\ heard', cluinneadh mi (headline). Present tense-future time: I'm
'I'll hear'. leaving tomorrow. Future tense-past time:
John will keep phoning us every day.

(verbs) active, passive, who did action, Classical Greek active didasko 'I teach', There are several active verbs in English
middle, causative what was acted middle didaskomai'l get myself taught'; which have no passive (e.g. She has a car
upon, what Portuguese active cortou 'cut', passive foi will not transform into 'A car is had by her),

caused action cortada 'was cut'; Tigrinya active qatale he and several passives which have no active
killed', causative yaqtale'he caused to kill'. (e.g. He was said to be angry will not trans-
form into 'Someone said him to be angry).

16 GRAMMAR -93
 .

ends at the points marked by pauses (-), or whether Minor sentence

Syntax this is all one, loosely constructed sentence: types
Syntax is the way
which words are arranged
in
when they fed the pigs/ they all had to stand well back/ A language contains many
to show meaning within (and
relationships of - and they were allowed to take the buckets/ - but they sentence-like units which do
sometimes between) sentences. The term comes weren't allowed to get near the pigs/ you see/ - so they not conform to the regular
from syntaxis, the Greek word for 'arrangement'. patterns of formation. Here is
weren't happy . .

a selection from English:

Most syntactic studies have focused on sentence
structure, for this is where the most important Linguistic approaches Yes.
Gosh!
grammatical relationships are expressed. Despite all we continue to employ
the difficulties, Least said, soonest mended.
the notion of 'sentence', and modern syntacticians How come you're early?
THE SENTENCE try to make sense of it. But they do not search Oh to be free!
Traditionally, grammars define a sentence in such for a satisfactory definition of 'sentence' at the out- All aboard!
— an Down with racism!
terms as 'the complete expression of a single set enterprise that is unlikely to succeed, with
No entry.
thought'. Modern studies avoid this emphasis, over 200 such definitions on record to date. Rather, Taxi!
because of the difficulties involved in saying what they aim to analyse the linguistic constructions that Good evening.
'thoughts' are. An egg can express a thought, but occur, recognizing the most independent of them Happy birthday!
Checkmate.
it would not be considered a complete sentence. as sentences. Thus, because the following construc-
/ shut the door, as it was cold is one sentence, tions can stand on their own as utterances, and
but it could easily be analysed as two thoughts. be assigned a syntactic structure, they would be
Some traditional grammars give a logical defini- recognized as sentences:
tion to the sentence. The most common approach she asked for a book/
proposes that a sentence has a 'subject' (=the come in/
topic) and a 'predicate' (= what is being said about the horse ran away because the train was noisy/
the topic). This approach works quite well for some
The following combination of units, however,
sentences, such as The book is on the table, where
could not be called a sentence:
we can argue that the book is what the sentence
is 'about'. But in many sentences it is not so easy
will the car be here at 3 o'clock/ it's raining/.

to make this distinction. It's raining is a sentence, The syntax of the first unit and that of the second
but what is the topic? And in Michael asked Mary do not combine to produce a regular pattern. It
for a pen, it is difficult to decide which of Michael, would be just as possible to have:
Mary, or the pen is the topic - or whether we have
it's raining/ will the car be here at 3 o'clock/
three topics!
In some written languages, it is possible to arrive or either unit without the other. Within each unit,
at a working definition of 'sentence' by referring however, several kinds of rules of syntactic order
to the punctuation one is taught to use in school. and selection are apparent. We may not say:
Thus, an English sentence for many people 'begins
*will be here at 3 o'clock the car/
with a capital letter and ends with a full stop' (or Exceptions These signs -
*will be here the car/
the first from a school play-
some other mark of 'final' punctuation). The prob- *car at 3 o'clock/.
ground, the second from an
lem is that many languages (e.g. in Asia) do not
office building - are presu-
make use of such features; and even in those that Each unit in the sequence, then, is a sentence; but
mably sentences, but they
do, punctuation is not always a clear guide. It may the combination does not produce a 'larger' syntac- do not follow the expected
tic unit. rules of punctuation.
be omitted (in notices and legal documents, for
example); and it proves difficult to prescribe rules A
sentence is thus the largest unit to which syn-
governing its use other than 'good practice'. People apply - 'an independent linguistic form,
tactic rules

therefore often disagree about the best way to not included by virtue of any grammatical con-
struction in any larger linguistic form' (L.
Bloom-
punctuate a text. In some manuals of style, it is
recommended that one should not end a sentence field, 1933, p. 170). But
approach has its
this
exceptions, too. In particular, we have to allow
before a coordinating conjunction [and, or, but).
for cases where sentences are permitted to omit
But there are often cases where an author might
feel it necessary - for reasons of emphasis, perhaps
part of their structure and thus be dependent on
— to do a previous sentence {elliptical sentences), as in:
the opposite.
It is even more difficult to identify sentences in A: Where are you going?
speech, where the units of rhythm and intonation B: To town.
often do not coincide with the places where full
Several other types of exception would be recog-
stops would occur in writing. In informal speech,
nized in a complete grammatical description.
in particular, constructions can lack the careful
organization we associate with the written lan-
guage is not that conversation lacks
(p. 52). It

grammar: simply that the grammar is of a

it is

rather different kind, with sentences being particu-

larly difficult to demarcate. In the following
extract, it is not easy to decide whether a sentence

94 • III THE STRUCTURE OF LANGUAGE

ASPECTS OF SENTENCE SYNTAX Concord
Clauses Grammatical links between
Hierarchy words are often signalled by
The various units that make up the structure of a
Hilary couldn't open the windows. concord, or agreement'. A
clause are usually given functional labels, such as
form of one word requires a
Subject (S), Verb(\l), Complement (C), Object (O),
One of the first things to do in analysing a sentence and Adverbial (A). A number of clause types can be
corresponding form of
it - sets of words
another, as when in English
is to look for groupings within identified in this way, such as:
a singular noun agrees with'
(or morphemes, p. 90) that hang together. In this
S + V The dog + is running. a singular verb in the present
example, we might make an initial division as fol- S + V+ The man + saw + a cow. tense: the man walks vs the
lows: S + V+C The car + is + ready. men walk.
S + V+A A picture + lay + on the ground. The purpose of concord
Hilary / couldn't open / the windows. S + V+ O+ O 1+ gave + John + a book. varies greatly between lan-
S + V+O+C He + called + John + a fool. guages. In Latin, it is an es-
Units such as couldn't open and the windows are S + V+ O+A Mary + saw + John + yesterday. sential means of signalling
called phrases. The would be called
first of these which words go together. In
Several approaches to grammatical analysis make use
a verb phrase, because its central word (or 'head') the absence of fixed word-
of elements of this kind, though there is considerable
is a verb, open; the second would be called a noun order patterns, sentences
variation in definition and terminology. Languages also
would otherwise be un-
phrase, because its head is a noun, windows. Other vary greatly in the way in which these elements are
interpretable. For example,
types of phrase also exist — adjective phrases, for identified. In English, for example, word order is the
in parvum puerum magna
main factor, with only occasional use being made of
example, such as very nice. puella vidit the tall girl saw
morphology (e.g. he (subject) saw (verb) him (object)).
Phrases may in turn be divided into their consti- the small boy', we know that
In Latin, word-endings provide the main clues to
the boy is small and the girl
tuent words (p. 91): element function, word order being irrelevant (e.g. puer
is tall only through the agree-
puellam vidit 'the boy saw the girl'). In Japanese, basic
couldn't + open the + windows grammatical relations are marked by special particles:
ment of the endings, -urn vs
-a.
ga (subject), o (direct object), ni (indirect object), and
And words may be divided into their constituent On the other hand, con-
no (genitive). For example,
morphemes, if there are any: cord plays much less of a
kodomoga tomodachino inuni mizuo yaru role in modern French, in
could + n't window + s the child friend's to dog water gives cases such as le petit garcon
et la grande title 'the little boy
This conception of sentence structure as a The child gives water to his/her friend's dog.'
and the big girl'. Because the
hierarchy of levels, or ranks, may be extended position of adjectives is fixed
'upwards'. The sentence can be made larger by link- (before the noun, in these
ing several units of the same type: Phrases cases), it would not pose any

Most phrases can be seen as expansions of a central problems of intelligibility if

Hilary opened the windows, but David couldn't there were no difference be-
element (the head), and these are often referred to
open the doors. as 'endocentric' phrases: tween the masculine and
feminine forms:
cars
Here, too, we have a sentence, but now we have le petit garcon
the cars
to recognize two major units within it - each of 'la petit title
the big cars
which has a structure closely resembling that of all the big cars 'le petite garcon
la petite title
an independent sentence. These units are traditio- all the big cars in the garage
nally referred to as clauses. In the above example, Phrases which cannot be analysed in this way are then If French allowed free word
order, as in Latin, so that one
the clauses have been 'coordinated' through the called exocentric': inside/the cars.
could say 'le garcon et la fille
use of the conjunction but. An indefinite number The internal structure of an endocentric phrase is
petit grande, then concord
of clauses can be linked within the same sentence. commonly described in a three-part manner:
would be needed to show
A five rank hierarchy is a widely used model of all the big cars in the garage which adjective should go
syntactic investigation: PREMODIFICATION HEAD POSTMODIFICATION with which noun - but this
does not happen. The
gender system is thus of
sentences sentences
limited usefulness, though it
are analysed into are used to build Coordination vs subordination still has a role to play in cer-

clauses clauses tain syntactic contexts, such

Coordination is one of two main ways of making
are analysed into are used to build as cross-reference (J'ai vu
sentences more complex; the other is known as
un livre et une plume. II 6tait
phrases phrases subordination, or embedding'. The essential difference
nouveau. saw a book and
I

are analysed into is that in the former the clauses that are linked are
are used to build a pen. It [i.e. the book] was
of equal grammatical status, whereas in the latter, one
words words new.').
clause functions as part of another (the main' clause).
are analysed into are used to build Compare:
morphemes morphemes
Coordinate clause:
The boy left on Monday and the girl left on Tuesday.
Morphemes are the 'lower' limit of grammatical
Subordinate clause:
enquiry, for they have no grammatical structure. The boy left on Monday when John rang.
Similarly, sentences form the 'upper' limit of gram-
The phrase on Monday is part of the clause, giving
matical study, because they do not usually form
the time when the action took place. Similarly, the unit
a part of any larger grammatical unit. when John rang is also part of the clause, for the same
reason. But when John rang is additionally a clause
in its own right.

16 GRAMMAR -95
IMMEDIATE CONSTITUENT
DIAGRAMS
One of the most widely used techniques for display-
ing sentence structure is the use of immediate
constituent (IC) analysis. This approach works
through the different levels of structure within a
sentence in a series of steps. At each level, a con-
struction is divided into its major constituents, and
the process continues until no further divisions can
be made. For example, to make an IC analysis of
the sentence The girl chased the dog, we carry out
the following steps:
1. Identify the two major constituents, the girl and
chased the dog.
2. Divide the next-biggest constituent into two,
viz. chased the dog into chased and the dog.
3. Continue dividing constituents into two until
we can go no further, viz. the girl and the dog
into the + girl, the + dog, and chased into
chase + the -e d ending.
PHRASE STRUCTURE
A good way of putting more information
into an
analysis would be to name, or
elements label, the
that emerge each time a sentence is segmented. It
would be possible to use functional labels such as
'subject' and 'predicate', but the approach that is
most widely practised has developed its own termi-
nology and abbreviations, so these will be used
here. Taking the above sentence (S), the first divi-
sion produces a 'noun phrase' (NP) the girl and
a 'verb phrase' (VP) chased the dog. (This is a
broader sense of 'verb phrase' than that used on
p. 95, as it includes both the verb and the noun
phrase that follows.) The second division recog-
nizes a 'verb' (V) chased and another noun phrase
the dog. The next divisions would produce combi-
nations of 'determiner' (DET) and 'noun' (N) the
+ girl, the + dog. This is the 'phrase structure' of
the sentence, and it can be displayed as a tree dia-
gram:

The order of segmentation can be summarized

using lines or brackets. If the first cut is symbolized
by a single vertical line, the second cut by two lines,
and so on, the sentence would look like this:

the HI girl / chase /// -ed // the /// dog

However,
stituent structure
grams':
a much clearer
is
way of representing con-
through the use of 'tree dia- A
DET N
A
DET N
The girl chased the dog
The girl chased the dog
This kind of representation of the phrase structure
of a sentence is known as a 'phrase marker' (or
'P-marker'). Phrase structures are also sometimes
represented as labelled sets of brackets, but these
are more difficult to read:

[stNpfDETthe] [ N girl]] [yptvchased] [Np[ DET the]

[N
dog]]]]

The girl chased the dog

The second kind of tree diagram
normal convention in modern linguistics.
Such representations of structure are very help-
ful, as far as they go. But not all sentences are as
easy to analyse in IC terms as this one. It is some-
times not clear where the cuts should be made (e.g.
whether to divide the three old men into the + three
Diagramming
is in fact the A frequent practice in American schools is the use of
a system of vertical and slanting lines to represent the
various relationships in a sentence. The
representations are often called Reed & Kellogg'
diagrams, after the authors of a 1 9th-century English
textbook. A long vertical line marks the boundary
between subject and predicate; a short vertical line
divides verb and direct object; and a short slanting line

old men or the three old + men, or the three + old marks off a complement. Other items are drawn in
beneath the main parts of the sentence.
men). More important, the process of segmenting
individual sentences does not take us very far in The old man called me a crazy inventor.
understanding the grammar of a language. IC man | called |
me / inventor
analyses do not inform us about the identity of
the sentence elements they disclose, nor do they
provide a means of showing how sentences relate
The approach shows the relationships between words
to each other grammatically (as with statements
clearly, but it cannot handle variations word order:
in
and questions, actives and passives). To develop both / turned off the light and / turned the light off would A little-used Chinese box'
a deeper understanding of grammatical structure, be diagrammed in the same way. representation of sentence
structure
alternative approaches must be used.

96 •
111 THE STRUCTURE OF LANGUAGE
RULES Transformations Generative notation
Analyses of single sentences are illuminating, as This tiny fragment of a generative grammar suffices A major feature of generative
grammar is the way special
far as they go, but grammarians are concerned to only to illustrate the general conception underlying
notations have been devised
move beyond this, to see whether their analyses the approach. 'Real' grammars of this kind contain toenable rules to be ex-
work for other sentences in the language. To what many rules of considerable complexity and of dif- pressed in an economical
other sentences might the above sequence of steps, ferent types. One special type of rule that was pro- way. In particular, different
and the resulting P-marker, also apply? In Noam types of brackets, such as (
posed in the first formulations became known as ),

[], and { } are given different

Chomsky's approach, first outlined in Syntactic a transformational rule. These rules enabled the
meanings. Round brackets,
Structures (1957), the jump from single-sentence grammar show the relationship between sen-
to for example, enclose a gram-
analysis is made by devising a set of rules that tences that had the same meaning but were of dif- matical element that is optio-
would 'generate' tree structures such as the above. ferent grammatical form. The link between active nal in a sentence; that is, the
sentence would be gramma-
The procedure can be illustrated using the follow- and passive sentences,for example, could be shown
tical even if the element were
ing rules (but several details from the original — such as the horse chased the man (active) and left out. The rule
approach are omitted for clarity): the man was chased by the horse (passive). The
NP— DET(ADJ)N
S^NP + VP kind of formulation needed to show this is:
means that a noun phrase
VP-*V + NP NP, + V + NP 2 -* NP 2 + Aux + Wen + by + NP, can consist of either a deter-
miner, adjective, and noun or
NP-»DET + N which an economical way of summarizing all
is simply a determiner and
V— chased the changes you would have to introduce, in order noun (the old man or the
DET^ the to turn the first sentence into the second. If this man). A grammar could, of
course, list the two possibili-
N-> girl, dog formula were to be translated into English, four
ties separately, as
separate operations would be recognized:
The first rule states that a sentence can consist of NP— DET+N
a noun phrase and a following verb phrase; the (i) The first noun phrase in the active sentence
NP— DET + ADJ + N
second, that a verb phrase can consist of a verb (NP,) is placed at the end of the passive sen-
tence. but collapsing them
a into
plus a following noun phrase; the third, that a noun single rule, through theuse
phrase can consist of a determiner plus a noun. (ii) The second noun phrase in the active sentence of the convention, saves a
( )

Each abstract category is then related to the appro- (NP2 is placed at the beginning of the passive
) great deal of space, and re-
priate words, thus enabling the sentence to be sentence. presents something we all
The verb (V) is changed from past tense to 'know' about the structure of
generated. Grammars that generate phrase struc- (iii)
the noun phrase.
tures in this way have come to be called 'phrase past participle (Ven), and an auxiliary verb
structure grammars' (PSGs). (Aux) is inserted before it.
If we follow these rules through, it can be seen (iv) A particle by is inserted between the verb and
that there is already a significant increase in the the final noun phrase.
'power' of this grammar over the single-sentence This rule will generate all regular active-passive
analysis used previously. If we choose the sentences.
NP, and the dog for the second,
girl for the first In subsequent development of generative gram-
we generate the girl chased the dog; but if the mar, many kinds of transformational rules came
choices are made the other way round, we generate to be used, and the status of such rules in a grammar
the sentence the dog chased the girl. By the simple has proved to be controversial ($65). Recent gener-
device of adding a few more words to the rules, ative grammars look very different from the model
suddenly a vast number of sentences can be gener- proposed in Syntactic Structures. But the funda-
ated: mental conception of sentence organization as a
V— chased, saw, liked .
single process of syntactic derivation remains
DET-> the, a influential, and it distinguishes this approach from
N—*girl, man, horse . .
those accounts of syntax that represent grammati-
cal relations using a hierarchy of separate ranks
the girl chased the horse
man saw the girl
the (p. 95).

the horse saw the man etc.

However, went were introduced into the rules,

if

as a possible V, ungrammatical sentences would

come to be generated, such as *the girl went the Rules and 'rules'
man. In working out a generative grammar, there- The 'rules' of a generative grammar are not to be
fore, a means has to be found to block the gener- identified with the prescriptive 'rules' that formed part
ation of this type of sentence, at the same time of traditional grammar (p. 3). A prescriptive
permitting such sentences as the man went to be grammatical rule is a statement - such as 'You should
never end a sentence with a preposition'- that tells
generated. The history of generative syntax since
us whether we are right or wrong to use a particular
1957 is the study of the most efficient ways of writ- construction. Generative rules have no such
ing rules, so as to ensure that a grammar will gener- implication of social correctness. They are objective
ate all the grammatical sentences of a language and descriptions of the grammatical patterns that occur.

none of the ungrammatical ones.

16 GRAMMAR -97
WORD ORDER perhaps these did not exist. But a group of OVS Lisu
The term 'word order' is somewhat ambiguous, languages have now been found, all in the Amazon This Lolo-Burmese language
for it can refer both to the order of words in a basin, mainly belonging to the Carib family, e.g. seems to have free word
phrase, and to the order of multi-word units within Hixkaryana, Apalai, Bacairi, Makusi. A few other order, yetit has no morpho-

logicalcases to mark Subject

a sentence. Given the sentence languages (e.g. Jamamadi, Apurina) seem to be
and Object. A sentence
OSV. But there is some variability in the data that Noun-Verb-Noun might
The cat sat on the mat
have been collected so far, with both OVS and OSV mean either N1 did
therefore
both the following involve word-order problems being used by some languages. VtoN2'or'N2didVtoN1\
— but they are of very different kinds: In theory, such a language
Word-order generalizations often need careful
ought be unintelligible! But
to
qualification. Latin, for example, is said to have
* cat the sat mat the on in fact the speakers survive,
a free word order, but in fact SOV is a very common by relying on context, the use
*sat the cat on the mat
pattern in that language. Modern Hebrew is SVO, of alternative grammatical
In linguistic description, word-order studies but Classical Hebrew seemed to favour VSO. Ger- constructions, and a modi-

usually refer to the second type of problem — SOV cum of common sense.
that man prefers SVO in main clauses, but in
is, the sequence in which grammatical elements subordinate clauses. In Tagalog, the V usually
such as Subject, Verb, and Object occur in sen- comes first, but there is great variation in what
tences. A great deal of attention has been paid to follows, with both OS and SO being widely used.
the way in which languages vary the order of these In Japanese, SOV is favoured, but OSV is also very
elements, as part of typological studies (§14). Word common.
order, it is hoped, will be a more satisfactory way
of classifying languages than the older morphologi-
cal method (which recognized such types as isolat- OSV in space
ing and inflecting, p. 293), into which many Sick I've become.
languages do not fit neatly. Strong with the Force you
In comparing word orders across languages, it are.
Your father he is.
is important to appreciate that what is being com-
When nine hundred
pared is the 'basic' or 'favourite' pattern found in years you reach, look as
each language. For example, in English, we will good you will not.
encounter such sequences as: The rarity of OSV con-
structions and languages
SVO the boy saw the man
perhaps explains the im-
OVS Jones I invited — not Smith pact of this strange
VSO govern thou my song (Milton) speech style, used by the
osv strange fits of passion have I known Jedi Master, Yoda, in the
film Return of the Jedi
(Wordsworth)
(1983).
sov pensive poets painful vigils keep (Pope)

However, only the first of these is the natural,

usual, 'unmarked' order in English; the others all

convey special an emphatic or poetic kind.

effects of
The same principle must apply in studying word
order in all languages, but it is often not so easy
to establish which is the normal word-order pattern
and which is the pattern that conveys the special
effect. The mere fact of talking to a foreigner, for
instance, might motivate a native speaker to change
from one order to another, and it often requires Deep and surface structure
on the part of the linguist to deter-
great ingenuity In the standard approach to we would hear if the sen- have different surface forms
mine whether such stylistic changes are taking generative grammar, sen- tence were spoken. but the same underlying
place. tences are analysed in terms This distinction was used meaning, as in the case of
oftwo levels of organization, to explain sentence ambigui- activeand passive sen-
known as deep structure and ties, by arguing that in such tences. Cats chase mice and
Typology surface structure. At the cases a single surface struc- Mice are chased by cats
Apart from cases of free word order (e.g. Latin, 'deep' (or 'underlying') level, ture correlates with more were said to have different
a sentence structure is re- than one deep structure.An surface structures, but the
Quechua, Navajo, Fore), there are six logical possi- presented in an abstract early Chomskyan example same deep structure.
bilities: SVO, SOV, VSO, VOS, OSV, OVS. Of way, displaying all the fac- was Flying planes can be The interpretation and
these, over of the world's languages use SVO
75% tors that govern how its dangerous, which can be re- status of the two notions has
French, Hausa, Vietnamese) or SOV
(as in English, meaning should be inter- lated to two underlying sen- altered greatly in generative
tences: Planes which fly can theory over the years (§65),
(as in Japanese, Amharic, Tibetan, Korean). A
preted. At the 'surface' level,
there is a more concrete be dangerous and To fly but the basic insight is one
further 10-15% use VSO (e.g. Welsh, Tongan, representation, giving the planes can be dangerous. that has achieved wide-
Squamish). Examples of VOS are Malagasy, Tzot- string of morphemes that The distinction was also spread recognition in linguis-

zil, and Houailou. closely corresponds to what used to relate sentences that tics.

Until recently, Object-initial languages were con-

spicuous by their absence, and it was thought that

98 •
III THE STRUCTURE OF LANGUAGE
Honorific grammar
Several languages make use of a special set of gram- Miscellany
matical contrasts, in which different levels of polite- Even a brief survey of four-hundreds and four for- 'that in there', 'that high up

ness or respect are expressed, according to the grammatical issues leaves and eight'. Some
ties there', that unseen'.
one somewhat in awe at number systems involve • English has a single pair
mutual status of the participants (§10). An 'honori- the extraordinary variety of counting backwards: Eng- of response words that can
fic' system, as it is is well developed
often called, patterns that exist in the lishsees 1 99 as '1 00 plus be used to reply to all ques-
in several oriental languages, such as Korean, Java- languages of the world. Re- 99';Yoruba sees it as '200 tions (other than those be-

nese, Tibetan, and Japanese; and although its use peatedly the lesson is
brought home that there is
is changing, especially among younger generations
nothing sacrosanct or su-
of speakers, it still plays an important role in the perior about thegrammar
marking of social relationships. of any one language - a
Japanese honorific expression shares with many lesson that is particularly
other languages certain characteristics of formal apposite for English users,
whose language holds a
speech. Local dialect forms are avoided; loan
special position in
words are often used (Chinese loans, in the case society (§59). The following
of Japanese); sentences are longer and involve more structural differences illus-
circumlocution and negative expression (cf. Eng- trate this important princi-
lish 'I wonder whether you mightn't .'). What
ple still further.
. .
less 1 Several languages
'.
have no number system:
Andamanese makes do
with two number-words,
one and one-plus. Khoisan
languages express one,
two, and occasionally
three, but rarely more.
modern • English has a single pair
of demonstratives, this and
that,which basically refer
to near' vs further away'.
To make other semantic
ginning with a question-
word, such as why): yes
and no. In Welsh, there is
an indefinite number of res-
ponse forms, the choice
depending on the gramma-
ticalform of the question.
For example, a question
beginning A oes ...? (Is
there . . .?) is replied to by
oes (yes) or nag oes (no);
Ydy Gwen yn mynd? (Is
Gwen going?) -> Ydy/Nag

• English counts in tens distinctions, we have to use ydy, Ydych chi'n mynd?
differentiates Japanese from European languages and units, as reflected in a circumlocution, e.g. that (Are you going?) - Ydw/
is the way in which pronouns, verbs, adjectives, our number-names: 41 = one over there'. Japanese Nag ydw, Allwch ...? (Can
and many types of grammatical construction tens one. Welsh
'four has a three-way system: you . . .?) - Gal la N a alia.
change their form depending on their honorific sta- counts in a mixture of tens kono = near the speaker, The principle underlying
and twenties: 20 = again, sono = near the hearer, this proliferation of forms is
tus. A large number of special forms are permitted,
30 = deg ar hugain ten on ano = distant from both (in straightforward, however.
which are classified into 'respect words' {sonkei- twenty', 40 = deugain two time or place). The Austra- In most cases, the reply

go), 'condescending words' (kenzyoo-go), and twenties', 50 = deg a deu- lianlanguage Alyawarra simply repeats the verb
'polite words' (teinei-go). gain 'ten and forty'. French has a four-term system: form, allowing for changes
also makes some use of nhinha = this', yanha = in pronouns. as if in It is
Honorific markers in the morphological system
twenty: 91 = quatre-vingt- nhaka = 'that
'that (near)', English there was a sys-
include: (a) a specific honorific prefix, o- or go-; onze 'four-twenty-eleven'. and awutha = 'the
(far)', tem:
(b) the complete replacement of a word, e.g. iw-u Old Hawaiian made use of one mentioned before'. Es-
Are you going? Yes-l-go.
'say' becomes ossyar-u; and (c) a complex system forty as a counting unit: 50 kimo has around 30 separ-
Is he there? No-he-isn't.
of titular forms (where English would say 'Mr, was forty and ten'; 968 ate demonstrative forms,
Did see? Yes-you-saw.
I

was expressed as two expressing such notions as

Mrs, Miss'), all suffixes attached to the name:
-sama very polite
-san neutral
-chian diminutive
"I miss the good old days
-kun for men only
sensei traditionally used to a person who was when all we had to worry about
'born earlier', but now used to someone was nouns and verbs.
whose capabilities are respected, especially
a teacher or politician
A wide range of pronoun forms is used. Among
the first-person forms, we find:

watakushi very formal male; less formal female

watasbi formal male; neutral female
atakushi rare male; snobbish female
atashi chiefly female, colloquial
washi dialectal, chiefly male, older gener-
ation
boku exclusively male, proscribed in talking
to superiors (but cf. p. 21)
ore colloquial male
Among the second-person forms, we find:

anata standard, polite, not used to superiors

anta informal
sochira polite, very formal
kimi chiefly men to men of equal or lower sta-
tus
omae informal, colloquial, somewhat pejora-
tive
kisama and feme derogatory, very impolite
(After S. I. Harada, 1976.)

16 GRAMMAR -99
 '

17 Semantics —T '&-. L-L

=53
Sw
Semantics is the study of meaning in language. The
term did not come to be widely used until the 20th
century, but the subject it represents is very old, r-i-~*

reaching back to the writings of Plato and Aristotle,

and attracting the special interest of philosophers,
logicians, and (these days) linguists {$65). The
linguistic approach aims to study the properties
of meaning in a systematic and objective way, with
reference to as wide a range of utterances and lan-
guages as possible. It is thus broader than the Semantics and Alice
approach taken by many logicians and philoso- One of the favourite The question is,' said
phers, who have tended to concentrate on a res- quotations of semanticists is whether you can
Alice,

tricted range of sentences (typically, statements, or

from Lewis Carroll's Through make words mean so man
the Looking Glass (1 872, different things.'
'propositions') within a single language. But logical Chapter 6), in which Humpty The question is,' said
analysis nonetheless exercises a major influence on Dumpty turns our Humpty Dumpty, 'which s to
i
yt
contemporary linguistic semantics (p. 107). conventional understanding be master -that's all.'

of meaning on its head, and Alice was too much

Any approach to semantics has to be
scientific
thus makes us see more puzzled to say anything: so
clearly distinguished from a pejorative sense of the clearly what it has to be after a minute Humpty
term that has developed in popular use, when peo- about. everyone were to
If Dumpty began again.
ple talk about the way language can be manipu- use words in an idiosyncratic 'They've a temper, some of
lated in order to mislead the public. A newspaper way, as Humpty suggests, them - particularly verbs,
the result would be they're the proudest -
headline might read 'Unemployment reduced to
communication anarchy. adjectives you can do
semantics' - referring to a new way of counting Only in certain fields - such anything with, but not verbs
the unemployed which makes it appear that there as literature (§1 2) - do we - however, /can manage the
are fewer of them. Or someone might say in an tolerate personal deviations whole lot of them!
from the semantic norms of Impenetrability! That's what /
argument, 'That's implying that the
just semantics',
the language. say!'
point is purely a verbal quibble, bearing no rela-
'Would you tell me,
tionship to anything in the real world. This kind There's glory for you!'
please,' said Alice, 'what that
'I don't know what you
of nuance is absent when we talk about semantics means?' said in a thoughtful tone.
mean by "glory," Alice said.
'

from the objective viewpoint of linguistic research. 'Now you talk like a 'When make a word do a
Humpty Dumpty smiled I

reasonable child,' said lot of work like that,' said

contemptuously. Of course
Humpty Dumpty, looking Humpty Dumpty, always 'I
you don't -till I tell you. I

very much pleased. 'I meant pay extra.'

it
meant "there's a nice knock-
THE MEANINGS OF MEANING down argument for you!"
by "impenetrability" that 'Oh!' said Alice. She was
we've had enough of that too much puzzled to make
In an important early book on the subject, C. K. 'But "glory" doesn't mean
subject, and it would be just any other remark.
Ogden & I. A. Richards's The Meaning of Meaning "a nice knock-down
as well if you'd mention what 'Ah, you should see 'em
argument,'" Alice objected.
(1923), 16 different meanings of the words 'mean/ you mean to do next, as I come round me of a
'When /use a word,'
meaning' were distinguished. Here are some of suppose you don't mean to Saturday night,' Humpty
Humpty Dumpty said, in
them: stop here all the rest of your Dumpty went on, wagging
rather a scornful tone, 'it
life.' hishead gravely from side to
means just what choose I it
That's a great deal to side, for to get their wages,
to mean - neither more nor
make one word mean,' Alice you know.'
John means to write, 'intends' less.'

A green means go. 'indicates'

light
Health means everything, 'has importance' only a part. Of greater importance
it is is the study
His look was full of meaning, 'special import' of the way in which words and sentences convey
What is the meaning of life? 'point, purpose' meaning in the everyday situations of speech and
What does mean
'capitalist' to you? 'convey' writing.
What does 'cornea' mean? 'refer to in the world'
THREE CONCEPTIONS OF MEANING
It is the last kind of use that comes closest to the
focus of linguistic semantics; but even this is a spe- Words —» things

cial kind of enquiry. The question asks for a defini- A popular view is that words 'name' or 'refer to'
tion,which is a somewhat unusual form of reply, things - a view that can be found in the pages
found more in dictionaries than in everyday speech, of Plato's Cratylus. Proper names like London, Bill
that involves the 'translation' of the difficult word Brown, and Daddy illustrate this conception, as
into 'easier' words. The study of the properties of do several other words and phrases - the labels
definitions is an important part of semantics, but attached to objects for sale in a shop, or those found

100 •
III THE STRUCTURE OF LANGUAGE
on a paint colour chart. But there are large numbers
of words where it is not possible to see what 'thing' n-LMH»l**r tffu
the word refers to: verbs such as ask or find; adjec-
tives such as difficult or popular; nouns such as
consistency or tradition. In fact, the majority of
words seem unable to be related to things, in any
clear way.

Words-* concepts-* things

This view denies a direct link between words and
things, arguing that the relationship can be made
only through the use of our minds. For every word,
there is an associated concept. One of the best-
known formulations of this position is the 'semiotic
triangle' of Ogden and Richards (1923, p. 99):

Thought

Symbol Referent

The main criticism of this approach is the insuper-

able difficulty of identifying 'concepts'. The 'con-
cept' underlying a word such as tradition is no
easier to define than the 'thing' referred to by tradi-
tion. Some words do have meanings that are rela-
tively easy to conceptualize,but we certainly do
not have neat visual images corresponding to every
word we say. Nor is there any guarantee that a
concept which might come to mind when I use the A design by Isidore of Seville (c. ad 555-636) The
word table is going to be the same as the one you, design attempts to show a link between a word's shape
the reader, might bring to mind. and its meaning. Isidore believed that the basic
meaning of a word could be found if it could be traced
back to its primitive shape. The discussion is found in
the ninth book of his Originum sive etymologiarum libri
Stimuli—» words —> responses
XX, which is largely about questions of semantic
Leonard Bloomfield (1887-1949) expounded a history and the origins of language.
behaviourist view of meaning in his book Language
(1933): meaning is something that can be deduced
solely from a study of the situation in which speech
is - the stimulus (S) that led someone to speak
used
Natural or conventional?
(r),and the response (R) that resulted from this
The Greek philosophers and even they change and things - a principle ac-
speech (s). He draws this as follows: were the first debate the
to greatly from language to cepted by modern seman-
nature of meaning, from language. But naturalistic ticists. There is nothing in

R which two main views
emerged. The naturalist
view, deriving largely from
In Bloomfield's example, Jill hungry, sees an
is
Plato (427-347 bc), main-
apple (S) and asks Jack to get for her (r); this
it tained that there was an in-
linguistic stimulus (s) leads to Jack getting the apple trinsic connection between
(R). Bloomfield argues that you can tell what the sound and sense. The con-
ventionalist view, largely
meaning of r...s must be just by observing the
Aristotelian, held that this
events that accompanied it. However, in very many connection was purely ar-
situations it is difficult to demonstrate what the bitrary (§65).
relevant features of the stimulus/response are — a In their extreme forms,
both views are untenable. If
real problem when events are not clearly visible
the naturalist view were
in physical terms (as in the expression of feelings).
valid, we would be able to
And it proves even more difficult to handle cases tell the meaning of words
where people do not act in the 'predicted' way (if just by hearing them. Only
Jack did not fetch the apple, perhaps because of onomatopoeic words (§30),
such as bow wow and
a quarrel with Jill at Monte Carlo two years
splash, come close to this,
before).
thinking is still widely en-
countered, especially in the
concern many people have
over the use of certain
words (to do with death or
sex, for example, p. 61 ), or
in the readiness with which
they make judgments
about the appropriateness
of words. 'Look at them,
sir,' says Aldous Huxley's
character Old Rowley,
pointing to swine wallowing
inthe mud, Rightly is they
called "pigs".' (Crome Yel-
tow, 1921).
The conventionalist posi-
tion is nearer the truth,
itemphasizes the arbitrary
relationship between words
the form of the word pig
that bears any direct rela-
tionship to the thing'. But it
is equally untenable to
think of language, as the
conventionalists did, solely
as the result of an agree-
ment between people to
use words in a certain way.
Such a procedure would
presuppose the prior exis-
tence of language, to for-
mulate the agreement in
the first place. Diodorus of
Megara (4th century bc)
nonetheless supported the
conventionalist position to
as the extent of calling his
slaves by the names of
Greek particles!

17 SEMANTICS- 101
Modern semantics
In the past, semantic debate has been largely con-
cerned with discovering what 'meaning' is, as a
concept in its own right. The enquiries have
undoubtedly increased our understanding of the
nature of the problem, but an accepted definition
of 'meaning' is as far away today as it was in Plato's
time. Why should this be so?
It is now widely held that 'meaning' is not some
kind of 'entity' separate from language — any more
than measures such as 'height' or 'length' have
What is plain, though, is that the same biological
relationships are given quite different linguistic
treatment between the two languages. Family
photographs would look the same, but the words
would have different senses (see below).
But even within a single language, we need to
distinguish sense from reference, to explain the way
language makes divisions where there are none in
reality. The neat scientific classifications of fauna
and flora, where each name has its place in a system
of terms, are not typical of language. In everyday
life, we use such words as hill and mountain, cup

some kind of independent existence. To say that
objects 'have height' means only that they are so
many units high; it does not mean that there is
an abstract property of 'height' that exists indepen-
dently of objects. In the same way, it is argued,
to say that words 'have meaning' means only that
they are used in a certain way in a sentence. We
can examine the meaning of individual words and
sentences — but there is no 'meaning' beyond that.
In modern linguistics, then, meaning
by making detailed analyses of the way words and
sentences are used in specific contexts. It is an
approach shared by several philosophers and
psychologists (p. 412). Ludwig
(1889-1951), in particular, stressed its
in his dictum: 'the meaning of a word
in the language'.
SENSE vs REFERENCE
Semantics is not directly concerned with the study
of the external world, or its conceptualization. The
world of non-linguistic experience is the province
of physicists, geographers, psychologists, and
others. Nor, as we have seen (p. 101), is semantics
easily able to cope with the study of how language
refers to this external world - the notion of 'refer-
ence'. Rather, the primary focus of the modern sub-
ject is on the way people relate words to each other
within the framework of their language - on their
'sense', rather than their reference.
and glass, or stream and river, where the real-world
notions are quite indeterminate. When does a
stream become a river, or a hill a mountain? And
would all agree about which of the pictures (right)
count as a chair}
There is also the problem of how we explain
what a word's meaning is. Let us imagine someone
who had encountered the word chair and did not
know what it meant. One procedure would be to
is studied explain its reference: we could take the person to
a chair and point to it. But this would be of limited
help, for how would the person know from that
experience which other objects in the world should
Wittgenstein also be called chairs? The wrong deduction might
importance also be made, that what we were pointing at was
is its use the quality 'wooden', or the concept of 'furniture'
— the kind of error children make when they learn
vocabulary (§42). A better procedure would be to
explain the sense of the word, using a rough defini-
tion such as 'a seat with four legs and a back'.
Such a definition would enable the person to look
out for other objects with similar properties, and
thus use the word appropriately. The definition
could then be sharpened, as related words were
met (e.g. armchair, stool). But this whole process
of vocabulary learning continues without any
direct reference to the objects in the real world:
there is total reliance on the use of words to explain
the sense of other words - a process that reaches
its logical conclusion in a dictionary (§18).

The distinction between sense and reference is

a critical one, because itallows us to study the

many cases where we happily use words, even
though they do not naturally correspond to the
way may be difficult
things are in the world. This
to see we restrict our study to a single language,
if brother mother brother
but when we look at how different languages 'par- uncle uncle

cel out' theworld, the distinction is forced upon

us. For example, in the 'real' world, mothers
and fathers have brothers and sisters. In English,
there are no single words expressing the notions
'mother's brother', 'father's brother', 'mother's sis-
ter', or 'father's sister', and we have to use a circum-
locution to make the distinction. In the Australian Biological Biological Female Biological Biological

language Pitjanjatjara, however, we have a differ- brother of of

sister parent brother of sister of
female parent female parent male parent male parent
ent situation: ngunytju = 'mother's sister', kamuru
= 'mother's brother', kurntili = 'father's sister',

and mama = 'father's brother'. There is also a com-

kamuru ngunytju
I

null
I

c
plication (to English ways of thinking): mama also
means 'father', and ngunytju also means 'mother'.

102 -III THE STRUCTURE OF LANGUAGE

SEMANTIC SPACE (N) Indians, the first two groups being fairly close
Psychologists also share the concern to establish together (after H. Maclay &
E. E. Ware, 1961).
the semantic properties of individual words, and
several approaches have been proposed to plot dif-
ferences and quantify the psychological 'distance'
between words.
A pioneering work in this field was C. E. Osgood,
G. Suci, &c P. Tannenbaum, The Measurement of
Meaning (1957), which was a study of 'affective'
meaning - the emotional reactions attached to a
word. Each word was subjected to a test that they
J
called a 'semantic differential' - the name reflecting
the view that it was possible to analyse meaning
into a range of different dimensions. Osgood
likened his procedure to a game of Twenty Ques- Charles E. Osgood (1916- )

tions, in which each question (e.g. 'Is it good or The semantic procedure is a limited
differential
bad? fast or slow? small or large?') would aim to one. It does not provide information about the
locate a concept in semantic space. The questions basic meaning of a word but only about the emo-
were presented as seven-point scales, with the tions the word generates. It tells us, for example,
opposed adjectives at each end, such as that mother might be 'very good', 'slightly strong',
etc., but it does not tell us that the word means
good bad
'adult female parent'. To display this kind of infor-
and subjects were asked to rate words in terms mation, other ways of working with semantic space
of where they would fall on these scales. If they are required. We can illustrate this using the results
felt that car was 'good', for example, they would of a technique in which people judge the similarities
place a mark towards the 'good' end of the first between words. In the diagram, mammal names
scale; if 'bad', towards the other end. The seven are located in a space where the horizontal dimen-
positions allowed for variations in degree of feel- sion represents size and the vertical dimension
ing. Ten of the scales are illustrated below, giving represents ferocity (after L. J. Rips et al., 1973).
the average responses from two groups of 20 sub- Larger animals are on the left; more ferocious
jects to the word polite (after C. E. Osgood, 1952): animals are towards the bottom. The more similar
any two animals are, the closer they are placed
in the space.
"Polite"

Rounded

goat • P'g
sheep*
cow

horse*

dog
animal
•
•
mammal * rabbit

• deer
mouse
•
bear*
•cat

lion

1 Stale

Size

The method was also used to make comparisons

between cultural groups. For example, noise is a
highly affective concept for the Japanese, who
tended to react to it using the extremes of the polar
scales; it is not so for Americans or Kannada-speak-
ing Indians. The word male varies in its conno-
tations between Hopi (H), Zuni (Z), and Navaho
This is a very simple analysis, which it would be
more difficult to make for words where the relevant
dimensions of meaning are less clear-cut (items of
furniture, for example). But the general approach
is illuminating, with considerable research poten-
tial.

17 SEMANTICS •
103
SEMANTIC STRUCTURE 'colour'. It has been argued that the whole of a
One most productive approaches to the
of the language's vocabulary is structured into fields; but
semantic analysis of vocabulary has come from the there is in fact a great deal of variation as we move
application of structuralist ideas (§65). From this from one part of the language to another. There
viewpoint, language is a network of systematic would be little difficulty gathering together all the
relationships between phonology, for
units. In English lexemes for 'body parts', for example; but
example, the relationships exist between sounds — it would be very difficult to do the same job for

or phonemes (§28). What are the equivalent 'noise' or 'ornaments'.

semantic units, and how are they related? There have been many philosophical and linguis-
tic attempts to classify the concepts or words in
Lexemes a language — notably, those associated with the
So far in this section, we have used the term 'word' 17th-century quest for a universal language (§58).
to discuss semantic units, and this is the traditional In recent times, the most influential and popular
use. People readily talk about the 'meaning of work has been the Thesaurus of Peter Mark Roget
words'. However, if we wish to enquire precisely (1779-1869), first published in 1852. Roget Roget (1779-1869)
into semantic matters, this term will not do, and divided the vocabulary into six main areas: abstract
an alternative must be found. There are three main relations, space, matter, intellect, volition, and
reasons. affections. Each area was given a detailed and
exhaustive sub-classification, producing 1,000 Semantic change
1. The term word used in ways that obscure
is
semantic categories in all. One path through the The linguistic approach to
the study of meaning. The forms walk, walks, semantic was first pro-
fields
thesaurus is illustrated below:
walking, and walked could all be called 'different pounded by German scho-
words'; yet from a semantic point of view, they lars in the 1 930s. In one of
the earliest studies (J. Trier,
are all variants of the same underlying unit, 'walk'.
affections 1 934), the approach showed
If the variants are referred to as 'words', though,
how the structure of a se-
what should the underlying unit be called? It would mantic field can change over
not be particularly clear to say that 'these four general terms religious time. Middle High German

words are different forms of the same word'. terms for 'knowledge'
changed greatly between
2. The term word is useless for the study of
1200 and 1300. In 1200, a
idioms, which are also units of meaning. A much- German had no separate
used example is kick the bucket (= 'die'). Here we lexeme for the quality of
have a single unit of meaning, which happens to obligation institutions cleverness. The language

consist of three words. Again, it would hardly be contained kunst ('courtly

and list ('non-courtly
skills')
clear to talk of this unit as a 'word', if we then
and there was also
skills'),
go on to say that this word consists of three words. wisheitlor any form of know-
3. The term word has in any case been appro- ledge, whether courtly or not,
priated for use elsewhere in linguistic study — in temperance intemperance sensualism asceticism etc. mundane or divine.
A hundred years later,
the field of grammar, where it does sterling service
everything was different.
at the junction between syntax and morphology Wisheit had developed the
(p. 90).
Groups of words are then listed under each of these restricted meaning of reli-
headings and classified into the main parts of gious experience'; kunst was
For such reasons, most linguists prefer to talk speech. For example, in the 1962 edition of the beginning to take on the
about the basic units of semantic analysis with fresh work (p. 625), we find the following items listed meaning of 'art skill', and
terminology, and both lexeme and lexical item are wizzen (modern Wissen) had
as a section within temperance (numbers refer to
in common use. We may now avoid the lack of
more the meaning of know-
other thesaurus sections; keywords are in italics): ledge'. List had left the field
clarity referred to above, and say that the 'lexeme'
entirely, as it had begun to
walk occurs in several variant forms — the 'words' abstainer, total a., teetotaller 948n. sober person;
develop pejorative conno-
walk, walks, etc. Similarly, we can say that the prohibitionist, pussyfoot; vegetarian, fruitarian, tations (cf. its sense of cun-

'lexeme' kick the bucket contains three 'words'; Pythagorean; Encratite; dieter, banter, faster; ning' or trick' in Modern Ger-
enemy of excess, Spartan 945n. ascetic. man). The whole of this
and so on. It is lexemes that are usually listed as
change can be summarized
headwords in a dictionary. Accordingly, we shall Thesauri of this kind have now been produced in the form of two diagrams:
put this term to use in the remaining parts of this for several languages, and prove to be a useful
section. adjunct to many practical linguistic activities, such
as professional writing, translating, and setting or
solving crosswords. For the semanticist, however,
SEMANTIC FIELDS their value is limited, as they contain no informa-
One way of imposing some order on vocabulary tion about the sense relationships between indivi-
is to organize it into 'fields' of meaning. Within dual lexemes, and items that come from different
each field, the lexemes interrelate, and define each regional, social, or professional varieties (§§8-11)
other in specific ways. For example, the various are juxtaposed without comment. To study the
For a similar use of dia-
lexemes for 'parts of the body' (head, neck, structure of a semantic field, more precise means grams in the comparison of
shoulders, etc.) form a semantic field, as do the of plotting the sense relations between lexemes modern languages, see
different lexemes for 'vehicles', 'fruit', 'tools', or need to be used. p. 106.

104 III THE STRUCTURE OF LANGUAGE

SENSE RELATIONSHIPS The company
How are the lexemes of a language organized? To The 'animal' kingdom lexemes keep
think of them as a list, such as we might find in Animalis a strange lexeme in English, because it can
'You shall know a word by
a dictionary, is highly misleading. There is no be used at three levels in a hierarchy of inclusion:
the company it keeps', said
1 in a classification of living things, it contrasts with
semantic reality in alphabetical order; on the con- the British linguist J. R. Firth
vegetable, to include birds, fishes, and insects;
trary, alphabetical order destroys semantic struc- (1890-1960) in 1957, refer-
2. it contrasts with bird, fish, and insect to include
ring to the syntagmatic ten-
ture, keeping apart lexemes that should belong humans and beasts; dency of lexemes to work to-
together (such as aunt and uncle, or big and little). 3. it contrasts with human.
gether ('collocate') in
Rather, we need to develop an alternative concep- predictable ways. Blond col-

tion, based on our intuitions that groups of lexemes LIVING locates with hair, floc/cwith
sheep, neigh with horse.
are related in sense.
Some collocations are totally
Accounts of semantic structure recognize several predictable, such as spick
kinds of sense relations between lexemes. Some vegetable animal with span, or addled with
result from the way lexemes occur in sequences brains or eggs. Others are
much less so: letter collo-
(syntagmatic relations); others from the way in
cates with a wide range of
which lexemes can substitute for each other (para- lexemes, such as alphabet
digmatic relations) (§65). For example, in the sen- and spelling, and (in another
tence It was a very auspicious — , English speakers sense) box, post, and write.
'know' that the omitted word will be one of a very human animal Yet other lexemes are so wi-
dely used that they have no
small set occasion, event) — unless, of course,
(e.g.
predictable collocates at all,
a literary or humorous point is being made (It was such as have and get.
a very auspicious kilt). This would be a syntagmatic — '. Likewise, potato is a hyponym of vegetable
Collocation should not be
semantic relationship. By contrast, the relationship confused with association of
in English, but Kartoffel isnot included among ideas'. The way lexemes
between the following two sentences is a paradig- Gemiise in German (after J. Lyons, 1963). work together may have no-
matic one: Is that a new
radio? No, it's an old • Antonymy This is the relationship of 'opposite- thing to do with 'ideas'. We
radio. The substitution of old for new results in say in English green with jea-
ness of meaning'. Antonyms are often thought of
a change of meaning that we recognize as an 'oppo- lousy (not blue, red, etc.),
in the same breath as synonyms, but they are in
site'.
though there is nothing liter-
fact very different. There may be no true synonyms, ally 'green' about 'jealousy'.
Several types of paradigmatic relationship have
but there are several kinds of antonyms. Some of Coffee can be white, though
been recognized, some of which form a familiar the most important types are: the colour is brown. Both
part of language syllabuses in school. These lads and lasses may be well
include: rounded enough to be called
—gradable antonyms, such as big/ small, good/ bad,
buxom, but this lexeme is
which permit the expression of degrees (very big, used only with the latter.
• Synonymy This is the relationship of 'same- quite small, etc.); Collocations differ greatly

ness' of meaning, e.g. kingly/ royal/ regal, — nongradable antonyms (also called complemen- between languages, and
provide a major difficulty
pavement/ sidewalk, youth /youngster. The search tary terms), which do not permit degrees of con- in
mastering foreign lan-
for synonyms is a longstanding pedagogical exer- trast, such as single /married, male/ female; it is not guages. In English, we 'face'
cise, but it is as well to remember that lexemes possible to talk of very male, quite married, etc., problems and interpret'
rarely (if ever) have exactly the same meaning. except in jest; and dreams; but in modern He-
There are usually emotional, or
stylistic, regional, - converse terms: two-way contrasts that are inter- brew, we have to 'stand in
frontof problems and 'solve'
other differences to consider. And context must be dependent, such as buy / sell or parent/ child; one
dreams. In Japanese the
taken into account. Two lexemes might be synony- member presupposes the other. verb for 'drink' collocates
mous in one sentence but different in another: with water and soup, but also
range and selection are synonyms in What a • Incompatibility Under this heading are with tablets and cigarettes.

nice —
of furnishings, but not in There's the grouped sets of lexemes that are mutually exclusive The more fixed a colloca-
mountain — members of the same superordinate category. For tion is, the more we think of
itas an idiom' - a pattern to
• Hyponymy This less familiar relationship example, red, green, etc. are incompatible lexemes be learned as a whole, and
refers to the notion of 'inclusion', whereby we can within the category colour: it would not be possible not as the 'sum of its parts'.
say that 'an X is a kind of Y'. For example, rose to say 'I am thinking of a single colour, and it is Thus we find French broyer
green and red.' On the other hand, red is not incom- du noir (lit. 'grind' + 'black'),
is a hyponym of flower, car of vehicle. Several lex-
meaning to 'have the blues'
emes will be 'co-hyponyms' of the same superordi- patible with such lexemes as round or dirty (some-
or 'be browned off' - a nice
nate term: rose, pansy, tulip Once again, it must
. . . thing can be at once 'red and round'). Terms for instance of the arbitrary use
be stressed that this is a linguistic, and not a real- fruit, flowers, weekdays, and musical instruments of colour terms.

illustrate other incompatible sets. Once again, we Collocations are quite dif-
world, classification. Languages differ in their
ferent from the idiosyncratic
superordinate terms, and in the hyponyms they ac- must be prepared for some unexpected usages —
linksbetween ideas that can
cept under one such term. For instance, in Classical as in English, where black, white, and grey are not be verbally expressed. On a
Greek the lexemes for 'carpenter', 'doctor', 'flau- always included within the category of colour (as psychiatrist's couch, we may
tist', and other occupations are all hyponyms of with black-and-white films and TV sets), and where 'free associate', responding
to farm with Easter, or jam
demiourgos; but there is no equivalent superordi- red can be excluded from this category (as with
with mother. This is not collo-
nate term in English. We simply do not have a snooker, where one may proceed to play the 'col-
which is a link be-
cation,
single 'occupational' term that would allow us to oured' balls only after all the red balls have been tween lexemes made by all
say 'A carpenter/doctor/flautist, etc. is a kind of potted). who speak a language.

17 SEMANTICS •
105
COLOUR LEXEMES a single morpheme (excluding light brown, etc.), Kinship contrasts
The colour spectrum is a continuous band, lacking were in common use (excluding indigo), applied Another semantic field which
any clear physical boundaries. The semantic field to many objects (excluding blond), and were not has been much studied is
of colour has therefore attracted particular atten- contained within another colour (excluding scar- that of kinship. Here too

let). They also claimed (p. 25) that these basic terms
there are interesting differ-
tion because it demonstrates very clearly the differ-
ences between languages:
ent patterns of lexical use in a language. English were ordered, as follows: • Hungarian had no terms
has 11 basic colour lexemes: white, black, red, for 'brother' or sister' until

green, yellow, blue, brown, purple, pink, orange, the 19th century, though it

white green purple did have separate terms for

and grey. In contrast: pink
<[red]< <[blue]<[brown]< 'elder' and younger' brothers

• There were no generic lexemes for 'brown' or black yellow orange and sisters.
'grey' in Latin; modern Romance forms (such
L -1 L J Lgre y . • Malay has a generic term
for both sibling' and cousin'.
as French brun, gris) have been borrowed from • There is no single term for
Germanic. Navaho has a single lexeme for both. grandfather' or grand-
If a language has a term to the right of the sign
• Navaho also makes no lexical distinction mother' Swedish: farfar =
in
<, it will also have all the terms to the left. mortar =
between 'blue' and 'green'. On the other hand, 'father's father',
These claims are not without controversy. mother's father', farmor =
it has two terms for 'black', distinguishing the
Obtaining reliable data from native speakers about father's mother', and mor-
black of darkness from the black of such objects mor= mother's mother'.
such matters is a problem, especially as their judg-
as coal. • Njamal (Australia),
In
ments might have been coloured by their exposure
• Russian makes a distinction between two kinds some terms express gener-
to other languages. Some languages, also, seem to ation distance, e.g. a man
of 'blue', sinij vs goluboj, where English has to
have 12 basic terms (e.g. Russian). But the research can use maili both for
use circumlocutions: 'dark blue' vs 'sky blue'. father's father' and
has demonstrated some impressive similarities
Hungarian has two terms for 'red'. 'daughter's son's wife's sis-
across a wide range of languages.
• Japanese awo can mean 'green', 'blue', or 'pale', ter' - both are two gener-
depending on context (e.g. vegetables, sea, ations away.

clouds).
POLYSEMY OR HOMONYMY? • Latin distinguished
• Polysemy refers to cases where a lexeme has father's brother' (patruus),
• In Hanunoo, there are just four basic colour
more than one meaning: for example, chip can father's sister' (matertera),
terms, 'black', 'white', 'red', and 'green'. 'mother's brother' (avun-
mean a piece of wood, food, or electronic circuit.
• Some New Guinea Highland languages have People see no problem in saying that 'the word
culus),and 'mother's sister'
terms only for 'black' and 'white' — perhaps bet- modern Ro-
(amita), but
chip has several different meanings in English'. mance languages have re-
ter translated as 'dark' vs 'light'.
• Homonymy refers to cases where two (or more) duced these to two (e.g.
• In some languages the situation is more difficult French oncle and tante, de-
different lexemes have the same shape: for exam-
to express in words, and a field diagram is rived from the maternal
ple, bank is both a building and an area of ground.
clearer. LiteraryWelsh, for example, divides the terms).
Again, people see no problem in saying that 'these
green— brown part of the spectrum quite differ-
are two different words in English'. Deixis
ently from English:
Every language has a set of
This second reaction would also be given to those
lexemes which can be inter-
cases where lexemes were only 'half identical in preted only with reference to
guryrdd
green shape: the speaker's position in
space or time. These are
-homophones, which have the same pronunci- known as forms (from
deictic
ation, but different spelling (e.g. threw vs the Greek word for point-
blue glas
through); ing'), and the conditions gov-

— homographs, which have erning their use have at-

the same spelling, but
tracted especial attention in
grey different pronunciation (e.g. wind— air move- recent semantics. They fall
ment vs bend). into three main types.
llusyd
brown • Personal deixis The
The distinction seems clear enough, and diction- use of pronouns, such as /
aries treat cases of multiple meaning either as poly- and you, which identify who
Modern Welsh is similar to English, but even so, semy or as homonymy. But in fact it is not always is taking part in the dis-
glas used for the colour of growing things
is easy to decide which we are dealing with, and dic- course.
• Spatial deixis Forms
(though it otherwise is equivalent to blue). tionaries sometimes differ in their decisions. Are
that distinguish the position
table (furniture) and table (arrangement of data) of the speaker in relation to
Colour universals? two different words, or the same word with two other people or objects, such
The differences between the colour terms of various meanings? Dictionaries usually go for the latter as this that, here there
languages are striking, and might lead us to con- solution, on grounds of a shared etymology. On (p. 99), bring/take, come/go.
Come, example, implies
for
clude that each language has worked out a unique the other hand, pupil (in school) and pupil (of the
direction towards the
system in a totally arbitrary way. A 1969 study eye) are usually listed as different words - though speaker - Come here! (but
by B. Berlin &: P. Kay, however, argued the oppo- in fact they have the same historical origin. French not 'Go here!).
site. After studying the colour systems of 98 lan- voler 'fly' and voler 'steal' are similar: they are now • Temporal deixis Forms
that distinguish time with
guages, they concluded that there is a universal thought of as different words, but both derive from
reference to the speaker,
inventory of only 11 basic colour categories, and Latin volare. There is often a conflict between his- such as now, yesterday,
all languages use either these 11 or fewer. 'Basic' torical criteria and present-day intuition, in sorting then, and the various kinds
was interpreted to mean that the terms used only out cases of polysemy and homonymy. of tense marker.

106 •
III THE STRUCTURE OF LANGUAGE
SEMANTIC COMPONENTS • Prosodic meaning The way a sentence is said, Grammar or
A further way to study lexical meaning is by analys- using the prosody of the language (§29), can radi- semantics
ing lexemes into a series of semantic features, or cally alter the meaning. Any marked change in
The uncertain boundary be-
components. Man, for example, could be analysed emphasis, for example, can lead to a sentence being tween semantics and gram-
as adult, human, and male. The approach was interpreted in a fresh light. Each of the following mar is a classic problem in
originally devised by anthropologists as a means sentences carries a different implication, as the linguistic theory. It can be il-

lustrated by the many sen-

of comparing vocabulary from different cultures, stress (indicated by capitals) moves:
tences that are used in a ha-
and it has been developed by semanticists as a John's bought a red CAR (not a red bicycle). bitual manner, and are thus
general framework for the analysis of meaning. John's bought a REDcar (not a green one). semi-idiomatic in type, falling
Whole systems of relationships can be estab- JOHN's bought a red car (not Michael). midway between the
'straightforward' idioms such
lished, using a small set of components. For exam- The prosody informs us of what information in
as raining cats and dogs and
ple, the components adult/non-adult and the sentence can be taken for granted (is 'given') clear cases of sentences
male/female can be used for the following: and what is of special significance (is 'new'). which follow the normal rules
of grammar, such as The
man (adult, male), woman (adult, female) • Grammatical meaning The categories that
man kicked the ball.
boy (non-adult, male), girl (non-adult, female). are established by grammatical analysis can also
In one study, a large
be analysed from a semantic point of view. A sen- number of habitually used
Many animals display a similar pattern (though
tence such as John read a book yesterday consists expressions were collected,
lacking a male/female non-adult distinction):
of Subject + Verb +Object + Adverbial (p. 95); based on the lexeme think.
MALE FEMALE NON-ADULT They included:
but it can also be analysed as an 'actor' performing
bull cow calf
an 'action' on a 'goal' at a certain 'time'. There Come to think of
ram ewe lamb it . .

is a great deal to be said about the 'semantic roles' What do you think?
boar sow piglet
played by syntactic elements — an area of study I thought better of it.
In componential analysis, contrasts are usually that falls uneasily between semantics and grammar. Think nothing of it.
presented in terms of + or — and often drawn
,
• Pragmatic meaning The function performed
Think it over.
It doesn't bear thinking
in a matrix. Thus, we could use -I- male and by the sentence in a discourse needs to be consi- about.
— male of course, + female and
(or, —female) dered. The meaning of the sentence There's some I thought you knew.
to summarize the above possibilities: chalk on the floor seems plain enough; but in some I think so.
What think is
situations it would be interpreted as a statement
I . .

bull ram boar cow ewe sow calf lamb piglet

was just thinking aloud.
MALE + + + - - ~ +- +~ +- of fact ('Have you seen any chalk?') and in others
I

FEMALE - - - + + + +-+-+- as a veiled command (as when a teacher might

Who'd have thought it?
Who do you think you are?
The analyses become more interesting, as the lex- point out the chalk to a child in class). The pragma- (After A. Pawley&F. H.

study of sentence function is reviewed in §21, Syder, 1983, pp. 213-14.)

emes become more complex. Here, for instance, tic

is a possible matrix for some human motion verbs. but it overlaps greatly with the field of semantics. It is argued that people

• Social meaning The choice of a sentence may have memorized expres-

NATURAL HURRIED FORWARD ONE FOOT ALWAYS
ON GROUND directly affect the social relationships between the sions of this kind, as part of
walk + - + + participants. We may convey such impressions as the process of building up

march — + + + politeness, rudeness, competence, or distance, and

fluent connected speech (the

run - + + — phenomenon is less obvious

- — this will affect our status and role within a commu- in the written language). On
limp + +
nity. 'What do you mean by talking to me like the other hand, these lexi-
It is easy, using a system of this kind, to see what that?' is a question that raises larger issues than calized sentence stems', as

gaps there are in a language. For example, they were called, are plainly
lexical the meaning of the individual lexemes and sen-
not as 'fixed' in their structure
this matrix suggests there is no single English lex- tences that have been used. as conventional idioms, and
eme expressing the notion of 'human using legs • meaning Perhaps the most
Propositional their meaning can be pre-
to move backwards'. On the other hand, it is not important trend modern semantics is the investi-
in dicted quite accurately from
always so easy to decide which are the relevant gation of sentence meaning using ideas derived their constituent lexemes
(unlike, say, raining cats and
components of a lexeme and whether they can be from philosophy and logic. In this kind of
dogs). The result is an area
applied in a binary (+/— ) way. Would swim be approach, a careful distinction is drawn between of usage that lies midway be-
-I- hurried or -hurried in this matrix? Or, in sentences (grammatical units, p. 94) and proposi- tween the domain of gram-
other fields, would soup be +eat or —eat, and tions. A proposition is the unit of meaning that mar, which focusses on pro-
ductive sentence types, and
porridge -(-liquid or -liquid? identifies the subject matter of a statement; it des-
that of the lexicon, which
cribes some state of affairs, and takes the form of focusses on the properties
SENTENCE MEANING a declarative sentence, e.g. Mary loves Michael. In of particular lexical items.
The study meaning takes us by degrees through
of such theories as 'truth-conditional semantics', sen-
the whole of a language, and it proves difficult to tences are analysed in terms of the underlying pro-
draw a neat line around the semantic component positions they express, and these propositions are
of any linguistic framework (§13). Much of the then tested to see whether they would be true or
focus of traditional semantics has been on vocabu- false, in relation to the real world. The theories
lary, but contemporary semantics is increasingly are controversial, and require not a little expertise
concerned with the analysis of sentence meaning in formal logic to be understood. But they may
— or, at least, of those aspects of sentence meaning in due course provide a level of general explanation
that cannot be predicted from the 'sum' of the indi- for semantic observations that the subject has
vidual lexemes. hitherto lacked.

17 SEMANTICS • 107
 UB

18 Dictionaries

A dictionary is a reference book that lists the words How big is a

of one or more languages, usually in alphabetical dictionary?
order, along with information about their spelling, Dictionaries usually claim to
df
pronunciation, grammatical status, meaning, his- !.
contain X,000' words. But
tory, and use. The process of compiling dictionaries this grand total can mean sev-
eral different things. It might
is known as lexicography, and the people who
refer just to the number of
carry out this task are lexicographers — 'harmless
headwords in the dictionary
drudges', as Dr Johnson defined them. - that is, the bold-face items
most homes have a diction-
In literate societies, that occur at the beginning of

ary, but there enormous variation in the way

is each entry. Or it might in-
clude in the sub-
addition all
this is used. Some people constantly use them as
sidiary bold-face items that
a serious educational tool, aiming to improve their occur within an entry: under
own or their children's 'word power'. Others use quick, for example, there will
them only for fun — as the arbiter in a game of be ~/yand ~ness. Different
Scrabble, for instance (p. 64). Others do not use word classes might be
counted separately (e.g. play
them at all and do not replace them when they A page from the noun vs play verb), as might
fall badly out of date. The continued use of 10- thumb' dictionary (real size)
idioms, and irregular gram-
or 20-year-old dictionaries is by no means uncom- The Random House Dictionary of the English Lan- matical forms (e.g. go, went).
mon. guage (c. 260,000 headwords, 1987) Depending on what you de-
cide to count, you can end up
For a book that is viewed with a level of respect (in'sa lar, ins'ya-), adj. 1. of or pertaining to an island
with two very different totals
normally accorded only to the Bible, it is remark- or islands: insular possessions. 2. dwelling or situated
for the same dictionary.
able how casually dictionary-users treat their dic- on an island. 3. forming an island: insular rocks. 4.
Claims about size should
detached; standing alone; isolated. 5. of, pertaining to,
tionaries. When people are asked what factors or characteristic of islanders. 6. narrow-minded or
therefore be viewed with
caution.
govern their choice of dictionary, most cite linguis-
illiberal; provincial: insular attitudes towards The best way to evaluate
tically irrelevant matters, such as price, pictorial
foreigners. 7. Pathol, occurring in or characterized by the coverage of a dictionary
content, and size — not in terms of number of one or more isolated spots, patches, or the like. 8. Anat. is to compare the words and
entries, but whether it would fit on a shelf, or in pertaining to an island of cells or tissue, as the islets senses it includes with
a pocket. Many people expect a dictionary to con- of Langerhans. —
n. 9. an inhabitant of an island; another dictionary of about
the same size. It is notable
tain encyclopedic information about historical islander. [1605-15; <LL insuldr{is). See insula, -ar']
- in'su lar ism, n. - in'su - in 'su how even the largest diction-
events, people, and places. Most admit they have . . . lar'i . ty, n. . lar ly,
.

aries present great differ-

never bothered to read the Preface to their diction- adv.
ences in their coverage - the
ary — the place where the layout and conventions Longman Dictionary of the English Language variation being particularly

of the book are systematically explained. As a con- (c. 90,000 headwords, 1984) noticeable in the way they
/'insyoola/ ad] la of or being an island b living or treat world regional vocabu-
sequence they are unable to say what the various lary(how many Australian,
abbreviations and symbols mean, or why they are situated on an island (—residents) 2 of a plant or animal
South African, or West In-
having a restricted or isolated natural range or habitat
there. The general conclusion is inescapable: most dian forms does an English
3a of island people (surviving customs) b that results
people who would check out every tiny feature of dictionary include, for exam-
(as if) from lack of contact with other peoples or ple?), local dialect words, ab-
their new car before buying it are unaware of the
cultures; narrow-minded, illiberal 4 anatomy of an breviations, slang and sub-
power that lies under the bonnet of their dictionary. island of cells or tissue [LL msularis, fr L insula island] standard forms, new coin-
- insularism n, insularly adv insularity n , ages and borrowings. The
use of illustrations and the in-
Penguin English Dictionary (c. 40,000 headwords,
clusion of encyclopedic infor-
THE RANGE OF DICTIONARIES 2nd edn, 1969) mation (names of people,
Dictionaries come shapes and sizes, from the
in all [msewler] adj of or like an island; of or like inhabitants places, historical events,
massive unabridged works, such as the 2,662-page of an island; cut off from general currents of thought; etc.) is also a major source
Merriam-Webster Third New International Dic- narrow-minded, smugly intolerant. of difference, especially be-
tween British and American
tionary of the English Language (1961), to the For further comparison, Nimmo's tiny book says
dictionaries. Ithas been esti-
tiniest of pocket-size works, such as the 386-page simply: 'a. surrounded by water'. mated that the lack of corres-
Nimmo's Thumb English Dictionary, each page of Apart from variations of format — the use of bold pondence in large English
which is less than 9 cm high. Coverage (the number face, numbered senses, etc. — there are major differ- dictionaries is often as great

of headwords the work contains) and treatment ences in the range and depth of information pro- as 30% - indicating that a
truly comprehensive diction-
(the kind of information provided under each head- vided. It is worth spending five minutes making ary of the language has yet
word) thus vary enormously. The simplest way of a point-by-point comparison, to see exactly how to be compiled.
showing this is to compare the amount of detail much information is lost as the dictionaries become
given for the same word in dictionaries of different smaller. And the moral is plain: for serious study
sizes. Here are the entries for insular in a large, of a language's word-hoard, only the largest diction-
medium-sized, and small dictionary: aries will suffice.

108 III THE STRUCTURE OF LANGUAGE

 «v,"'n s

DICTIONARY
OF THE

ENGLISH LANGUAGE *•**/

4
Co

IN WHICI i>/«
The WORDS are deduced from iheir ORIGINALS, «,>*/-/;
- "> tr^'"' e
'
«W """Wj •]?«**.
's

ILLUSTRATED thcil DIFFERENT S I r. N I F ICAT ONS

I

EXAMPLES from the bca WRITERS.

A HISTORY of fc LANGUAGt,
An ENGLISH GRAMMAR.
SAMUEL JOHNSON, AM.

TWO VOLUMES

«^^S§?
*I'4
Oh!*«.*
r
«Oi-
g. *«3r
rf*<£*

L N D M
'°'

"Me,
n ' J" "*">
p '
"W
r

Pnnted bj W. S

T. uxi T. Lo» CM M C. H
!.« inJ R. Utl I Da a.
MDCCl

The title page and facing picture, together with an extract from
Dr Samuel Johnson's Dictionary of the English Language

This great dictionary was con, Pope, and the Bible.

published in two volumes in The dictionary is very much
1 755, and its influence on the language of the best'
subsequent lexicography authors.
was unequalled. Noah • Dictionaries became more
Webster went so far as to authoritative - and authori-
compare Johnson's contribu- - as a consequence.
tarian
tion to Newton's in mathema- They became increasingly
tics. The dictionary had four used in a normative way, as
main features that set it off guides to good usage - a
from previous English works: bias which 20th-century dic-
• It aimed to be a scholarly tionaries are only just begin-
record of the whole of a lan- ning to correct.
guage - a marked contrast However, Johnson's was
with the haphazard dictionar- not the first monolingual
ies of hard words' previously English dictionary. The credit
compiled. for this must go
Robert to
Noah Webster (1758-1843)
• It was based on a corpus Cawdrey'sA Table Alphabe- Webster's works on spelling,
of examples of usage, large- tical! a 2,500-entry work
.,
. .

grammar, and lexicon consti-

ly from the period 1 560 to
1 660; certain types of words
were excluded (e.g. proper
names - a decision which re-
sulted in a major difference
between British and Ameri-
printed in 1 604.
James Murray (1 837-191 5) The first editor of the tuted the first major account
Oxford English Dictionary, is seen at work in his of American English, and
From Johnson's Preface Scriptorium. Murray, the son of a village tailor from gave the variety a clear iden-
My purpose was to admit no Hawick, Scotland, was largely self-educated. He left tity and status. His An Ameri-
testimony of living authors,
can Dictionary of the English
school at 1 4, and was a teacher and bank clerk before
that might not be misled by
becoming a lexicographer. His Scriptorium, where Language, published in two
I

can dictionaries, for while the partiality, and that none of

former followed Johnson, the
latter did not).
my contemporaries
have reason to complain
most of the editing took place, was built in the back volumes in 1828, consisted
might
garden of his house in Oxford. Murray planned the of around 70,000 entries.
whole of the New English Dictionary, and he edited
. .

Particular attention was paid

• It introduced a literary di-
mension, departing from the
previous concentration on
technical language. Half of
Johnson's quotations
all
come from Shakespeare,
Dryden, Milton, Addison, Ba-
have studiously endea-
I
voured to collect examples
and authorities from the
writers before the restor-
ation, whose works regard
as the wells of English unde-
filed, as the pure sources of
genuine diction.
more than half of its first edition himself. The first to the inclusion of scientific
instalment was published in 1884: A-ANT, 352 pages, terms, and to etymological
price 1 2s. 6d. (or 62fp today). It took 44 years to background. The latest revi-
complete the dictionary, in 1 25 instalments - four times sion the Third New Inter-
is
I
longer than had been expected. The complete work, national Dictionary (Merriam,
totalling 424,825 entries, was then published as the 1 961 ), containing over
Oxford English Dictionary (OED) from 1 933 onwards, 450,000 entries, which took
with four supplements issued from 1 972. The task of 757 editor-years to com-
transferring the OED database to computer storage plete. Supplements
began in 1984. appeared in 1976, 1983, and
1986.

18 DICTIONARIES • 109
 45 R0SC0E — ROSE
at fit 4* La'M da *t*™ (1T*S< ; te fU « fa Paait -aw, laaa a* raat. % »U toat ca baaa. liw aaa d» 0a> daafai M«
laadW. | C* «« eat lait tr«c a
Ml a« Uaa J (IMS) p<*a?w naataa. Pacaawrv k aM au raaat. V pot. euCc. etc, dc coaiear raw raaatUer de aoaa
—
:
|

R0a001
L««n
(atr
em IBM. a>ert
acad (BarreyJ. 1»IS «
Hearr Karma).
•
Pn*ww
ckwWt aarlau. ad k
Weat-tiocaln. area Lataer-
de ckisw as 0««w'
*
I CiK
fraad. k
la ^h aelfe fwat aV aaa dkaaaaa.
plat preoeax a*aata(e «a'U poaarde.
Can k piv.
| H
—
An**, aier La pU d* roae as iyaar-.U. V n
Bat. Mia-)*
A.i. •
mu
t-rrraav :
Ul li raaaaf V Ocaraoaa. THE HISTORY OF LEXICOGRAPHY
C*i«BC * Maata»atw da l»ifl a ISM. mtmbn * la
The earliest dictionaries had a practical aim. They
aaJaaf layafc ate Laadjaai (MM .-.:*- a -. -: • ah Mai U bk
T«n de lode faanaat de aUlbrrt* a Dwserter ear
• alia
Tr*a* decana* «e rrwa *«Jaaaai (ISTT-lSel)- aaa ba>
grapa* or Jaa* Dalit* (l**Si. Uli aaa* C^pcnaacaa — Arefcit fnn
kUlaerta. drt-oa, a»eit d'abord ecrtt : were often bilingual or polyglot word-lists aimed
l^"*!l
Idaavaat. '
T*!!^i^^ i^-^
V«Wa de M«a, kaT^L^n*^i^
awanf J h^TaaaWai" a%7 caa«aa«aa cartattam. I Bam Wa%jami at the traveller and the missionary, or glossaries
la I I

fwawa Crawk fc-etre clr»U»re. lerawe de .iuiax.

10 V ton ftaat a* ea-aart..
i
|
written to help people understand words which
were dialectal, technical, or rare. From as early as
the 5 th century BC, the Greeks were compiling glos-
sal, explaining difficult words in such authors as
Homer. The first vocabulary lists in English were
similar: these were 8th-century Anglo-Saxon
glosses, in which English words were written
between the Latin lines. Later, these glosses were
collected together as lists. But random collections
of words or glosses are not dictionaries: to count
as a dictionary, the words need to be organized
in a systematic way — such as through the use of
the alphabetic principle.
The history of lexicography goes back over
2,000 years, to ancient China, Greece, and Rome.
But there has never been a time when some kind
of lexicographical work was not in progress. Some
periods were particularly fruitful.

• Arabic dictionaries flourished from around the

8th century onwards.

Left: a page from the Larousse duXXsiecle (1 966)

It shows the mixture of lexical, literary, and encyclopedic
information which is the hallmark of this tradition.

Below: a page from a German-Latin school book, Teutsch

lateinisches Worter-Buchlein
Printed around 1 722, it contains about 6,000 words, each of
which is illustrated.
afartaaj
33<ftun$ Diverfon-
Arx.f. }.

3}arfu<$e
Canpona.
f.l.

flanflabta
•tram T*»
beraaf, I,

>

5[<lf<$6anr
Macellum.
". ».

$<finqnufc
Circer.
m.J.

2>rurf«
Pont, m, ).

ffraurmarf
Forum oli-
onum.n.l.

-rum e.
3'Jirium.
o. I.
Oc&ftn'
mardf Po.
runt Boiri-
M,D, 2,

Imanft
i Forum c*»

um.Q, i.

110 •
III THE STRUCTURE OF LANGUAGE
• There was a flurry of activity in several lan- THE PRESENT . . . THE FUTURE . . Words that have
guages following the invention of printing. Since the 1970s, the flow of dictionaries has been never been used-
• The Accademia della Crusca produced its dic- unabated, as publishers try to meet the needs of outside the
tionary in 1612 (the first to be compiled by a an increasingly language-conscious age. In English, dictionary!
team of people), and prompted several other for example, new and supplements to the
editions
Early dictionaries were often
national dictionary projects. well-known dictionaries have appeared, and
little more than lists of 'hard'
• Polyglot dictionaries were particularly numer- several publishers have launched new general series words. After all, it was rea-
ous in the 17th century, with the development (notably, Longman). Reader's Digest produced its soned, why should a diction-
of trade and missionary activities around the Great Illustrated Dictionary in 1984, the first full- ary include the words every-

world. colour English dictionary, in the encyclopedic tradi- one already knew? The re-
was the inclusion of
sult
• The 18th century saw a fresh direction in lexico- tion of Larousse. Prominent also have been the dic-
many very rare words - and
graphy, following the discoveries of the compar- tionaries for special purposes (foreign language some whose usage has
ative philologists (§50), and the first major teaching, linguistics, medicine, chemistry, etc.). For never been officially re-

historical dictionaries began to be compiled. the first time, spoken vocabulary has begun to find corded, outside of the diction-
aries in which they appear!
• The 19th century saw many large-scale diction- its way into dictionaries (though by no means all
Examples are commemor-
ary projects, produced by teams of compilers, are yet willing to include the more colloquial words able and liquescency.
and several specialized dictionaries (such as of and uses). But for non-words, the
dialect or technical words). Different kinds of But this outpouring is as nothing compared with prize must go to Dord, first

dictionaries began to be produced, notably the the flood yet to come. The 1980s will one day be used in a 1 930s dictionary,
and found subsequently in
Larousse series (from 1856), with its distinctive seen as a watershed in lexicography — the decade
several others. It seems
pictorial and encyclopedic character. in which computer applications began to alter radi- there was a file in the com-
• The 20th century has seen the development of cally the methods and the potential of lexicogra- piler's officewhich contained
lexicography as a scholarly subject, largely phy. Gone are the days of painstaking manual the entry 'D or d' as abbre-
viations for the word den-
under the influence of linguistics, and promoted transcription and sorting on paper slips: the future
sity'. The information was ac-
especially by the growth of academic societies, is on disk, in the form of vast lexical databases,
cidentally put into the
such as the Dictionary Society of North America continuously updated, that can generate a diction- meaning
dictionary as Dord,
(1975), and the European Association for Lexi- ary of a given size and scope in a fraction of the and a new word
'density',

cography (euralex, 1983) time it used to take. Special programs will become was bom.
available enabling people to ask the dictionary spe-
cial questions (such as: 'find all words that entered
the language in 1964' or 'find all words ending
in -esse'). Access to large machine-dictionaries will
Twenty questions to ask when you buy
become routine in offices and homes. One day, we
a dictionary
shallnot look up a word in a dictionary on a shelf
1 Is the paper of good, hard-wearing quality?
but ask our home computer for the information
2 Will the binding allow to be opened flat?
it

3 Are (especially long) entries clearly laid out? we need. That day is not far off.
4 Does it have the words you most want to look up?
(Keep a note of some words which have caused you
problems, and use them as a quick check.)
5 Does it have good international coverage? Some important events bj (Before Johnson)
6 Does it contain encyclopedic information? 5th bc Protagoras of Ab-
c. Hindu grammarian, Amara- tedesco (Venice).
7 Does it have illustrations of difficult concepts? dera compiled a glossary of simha. 1499 Probably the first dic-
8 Are the definitions clearly distinguished, and unfamiliar words in Homer. 8th c. The first general tionary to be printed in Eng-
organized on a sensible principle? 3rd c. bc The poet Philetos Arabic dictionary, Kitab al- land: the Latin-English
9 Are the definitions easy to understand, and helpful of Cos compiled a glossary 'ayn, compiled by AI-KhalTI Promptuorium parvulorum
(e.g. avoiding vicious circularity, as when X is defined of unusual poetic, techni- Ibn Ahmad. (London, Richard Pynson).
as Y, and Y is then defined as X)? cal, and dialect words. 10th-11thc. Compilation 151 1 The first printed
10 Does give citations (examples of usage), and are
it
2nd c. bc Aristophanes of ofa Byzantine encyclo- Dutch dictionary: Noel de
they real or artificial? Byzantium compiled a dic- pedic dictionary, the Berlaimont's Vocabulaire.
1 Does it give guidance about usage? and obso-
tionary of current Suidas. 1539 Compilation of
12 Does it use a good set of stylistic labels (e.g. lete words. 11 the. First Chinese- Robert Estienne's Diction-
formal, slang, medical, archaic)? 1st c. bc Marcus Verrius Japanese encyclopedic naire francois-latin.
1 Does it give etymological information? Flaccus compiled the first dictionary, by Minamoto no 1596 The first published
14 Does give guidance about capitalization, spelling
it
Latin lexicon, Libri de signi- Shitag6. Russian dictionary: Laur-
variation, and where syllable boundaries go (i.e. where ficatu verborum. 12th c. Compilation of the Leksys
entii Zizanii's . .

to hyphenate)? 1st c. Valerius Harpocra- Greek Etymologicon mag- synonima sloveno-ross-

1 Does it give pronunciation variants, and is the tion compiled a lexicon of num, author unknown. kaia.
phonetic transcription easy to follow? the Attic orators. 13th c. Joannes Balbus 1606 Publication of Jean
16 Does it contain idioms, phrases, proverbs, etc.? 2ndc. First systematic Januensis compiled the Nicot's Thresor de la Ian-
1 Does it contain lists of synonyms and antonyms? Chinese dictionary, Hsuo encyclopedic dictionary, gue francoyse, the first sys-
1 Does it give useful cross-references to other words Win, compiled by Hsu the Catholicon, one of the tematic French dictionary.
of related meaning? Shen. most influential dictionaries 1611 Publication of the
1 Does it give information about word class, 5th c. Hesychius of Alex- of the middle ages, and the major Spanish diction-
first
inflectional endings, and other relevant features of andria compiled a large to be printed ary,Tesoro de la lengua
first (in460).1

grammar? lexicon of Classical Greek. 1477 The earliest printed castellana o espanola of
20 Are there useful appendices (e.g. abbreviations, 6th c. Compilation of a bilingual dictionary: the Covarrubias y Horozco.
measures)? Sanskrit dictionary by the Vocabolista italiano- (After R. L Collison, 1982.)

18 DICTIONARIES- 111
19 Names

A name is a word or phrase that identifies a specific variation in levels of importance. In Britain, for
person, place, or thing. We see the entity as an example, the first name is the important one —
individual, and not as a member of a class: Everest, David Michael Smith would usually be referred to
for example, is a unique name (a 'proper noun'), as 'David'; in Germany, the name nearest to the
whereas mountain applies to a whole class of surname is more important - Johann Wolfgang
objects (a 'common noun'). In the written lan- Schmidt would usually be referred to as Wolfgang.
guage, European languages generally recognize the Some languages make use of patronymics — a
distinction by writing names with an initial capital name derived from the father's given name: in Rus-
letter. But most other writing systems do not dis- sian, Ivan's son would be known as Ivanovich, and
tinguish upper- and lower-case letters (§33), and his daughter as Ivanovna. The opposite practice,
even in Europe there are several arbitrary conven- of naming a parent after a child (teknonymy) is
tions and points of uncertainty. English, for exam- less common, but is widespread in the Arab world,
ple, is idiosyncratic in its use of capitals for days for example, where a parent is often called 'father
of the week and proper adjectives (as in the Chinese of or 'mother of the eldest son. In Russian, the
language) (p. 194). And decisions have to be made patronymic is placed between the child's given and
over whether one writes catholic or Catholic, the family names. In Icelandic, the patronymic serves
church or the Church, bible or Bible. as the surname, which then changes with each gener-
The science that studies names is known as ono- ation. Amharic names consist simply of the child's
mastics, usually divided into the study of personal given name plus the father's given name. In English,
names (anthroponomastics) and place names patronymic prefixes and suffixes are used only in
(toponomastics). In more popular usage, however, familynames (e.g. Robertson), and this is common
the term onomastics is used for the former, and throughout Europe (e.g. 'son of appears in Scots
toponymy for the latter. The division is ultimately Mac/Mc-, Irish O', Welsh Ap, Polish -ski, Greek
an arbitrary one, as places are often named after -poulos).
people (e.g. Washington) and vice versa (e.g. Israel
is sometimes used as a first name). Other categories

of name (e.g. ships, trains, yachts, domestic pets,

SIMILARITIES AND DIFFERENCES
There are some impressive similarities in naming
race horses, commercial products) also need to be
practice across different languages, such as the use
taken into account. But most name studies fall
of names based on professions. Smith, and its for-
under one of the two major headings.
eign-language equivalents, is the best-known case,
being the most common surname in many parts
of Europe: Arabic Haddad, Hungarian Kovdcs,
Russian Kuznetsov, Portuguese Ferreiro, German
Personal names Schmidt, Spanish Hernandez/ Fernandez, French
Most people are familiar with only one personal Le Fevre/La Forge, and so on. But the differences
naming system, and are surprised to learn that in naming practices are far more striking.
practices differ greatly from language to language. The possibilities of variation seem endless. We
Even such a basic distinction as 'given name' (or find the tripartite personal names of the Romans
'Christian name') and 'family name' (or 'surname') (e.g. Gaius Julius Caesar), the compound names

is not universal (in Europe, it began to be used of early Germans and Celts (e.g. Orgetorix 'king
in the late middle ages, reaching some areas only of killers'), and the use of 'by-names' to distinguish

as recently as the 19th century). These names are people who have the same name (e.g. Welsh Dai
also often referred to as 'first' and 'last' names; Jones-the-milk vs Dai Jones-the-post). In Europe,
but this nomenclature is ambiguous when compar- there is a great diversity of given and family names;
ing languages, as there is considerable variation by contrast, in several oriental societies the possi-
in the order in which such names occur. In most bilities are highly restricted — for example, just
European languages, the family name follows the three family names, Kim, Pak, and Yi, are used
given name; but the reverse is the case in, say, Hun- by most of the people of Korea.
garian and Chinese (e.g. Mao Ze-Dong). We find children named after saints, events,
In some societies, a middle name is also regularly places,omens, personal traits — even animals (as
used. This is the case in America, for instance, with north American Indians, e.g. Little Bear). In
where an initial is especially favoured (e.g. John some societies, divine names can be used ('theo-
H. Smith). In Europe, middle names are less com- phoric' names, such as Greek Herodotus 'given by
mon, unless acquired at a special occasion (such Hera' or Arabic Abd Allah 'slave of Allah'). At
as the Catholic ceremony of Confirmation). Where the opposite extreme, children might be named
there is a sequence of names, there may also be after unpleasant notions to make them undesirable

112 III THE STRUCTURE OF LANGUAGE

 to evil spirits ('apotropaic' names, such as 'cripple' itwas usually as a female name. Its peak British
or 'ugly', p. 9). Where personal names are con- usage was during the 1950s and 1960s.
cerned, there seems to be no limit to parental idio- Interesting though the origin of a name may be,
syncrasy and invention, two fine examples being it exercises very little influence on most parental
the Puritan given name Kill Sin, and the Russian choice and is of little value in the study of naming
concoction Mels — an acronym for 'Marx-Engels- trends. Far more important is the recent history
Lenin-Stalin'. of the name in a society — whether ithas been used
by the parents or near relatives, or by famous indi-
FIRST NAMES viduals such as film stars, pop stars, or members
In 1623, the historian and antiquary William Cam- of a royal family. Nations have different traditions
den (1551—1623) published an appendix to his and America permit all pos-
in this respect. Britain
guide book to Britain, in which he included a long siblenames, whereas in France and Germany there
list of the most popular given names and surnames are approved lists of names that must be used if
of his time. Since then, there have been several aca- a child is to be legally recognized. The influence
demic studies of given names in a wide range of of a religious tradition (as in the Catholic use of
European languages, and many popular accounts, saints' names) is often apparent. Statistical studies
aimed especially at providing information for par- of name use are thus of particular significance.
ents who do not know what to call their baby. Information about name use over the years
The studies are both etymological and statistical. comes from a variety of sources such as parish
The former have universal appeal. People are fasci- records, national censuses, newspaper birth
nated by the history of names — in particular, where announcements, and special surveys by name scho-
their own name comes from and how its usage has lars and enthusiasts. There has, for example, been
changed over the centuries. For example, Hilary a survey of every first name used by the Smiths
is from Latin hilarius, meaning 'cheerful'. It has in England and Wales since 1837. In the case of
been used by three male saints (including one English, it is also important to consider all parts
Pope), and it has continued as a male name in of the English-speaking world to discover whether
Europe. However, in Britain, it fell out of use in there have been directions of influence.
the 17th century, and when it revived in the 1890s,

General findings about first

names are tentative but Top ten first names

intriguing. Names are much
influenced by fashion, and BOYS (England and Wales) BOYS (USA) GIRLS (England and Wales) GIRLS (USA)
very few names retain great 1925 1950 1925 1950 1925 1950 1925 1950
popularity from one John David Robert Robert Joan Susan Mary Linda
generation to the next. There William John John Michael Mary Linda Barbara Mary
is much greater variation in
George Peter William James Joyce Christine Dorothy Patricia

girls' names: in the lists

James Michael James John Margaret Margaret Betty Susan
(right), hardly any girls' Ronald Alan Charles David Dorothy Carol Ruth Deborah
names carry over from one Robert Robert Richard William Doris Jennifer Margaret Kathleen
generation to the next, Kenneth Stephen George Thomas Kathleen Janet Helen Barbara
whereas several boys' 8 Frederick Paul 8 Donald Richard 8 Irene Patricia 8 Elizabeth Nancy
names do. The white/non- 9 Thomas Brian 9 Joseph Gary 9 Betty Barbara 9 Jean Sharon
white difference in the USA 10 Albert Graham 10 Edward Charles 10 Eileen Ann 10 Ann(e) Karen
is noticeable for girls, also.

And in English, it seems that 1965 1975 1970 : 962 (white) 1965 1975 1970 1982(wh/fe)
traditionally-male names are Paul Stephen 1 Michael Michael 1 Trac(e)y Claire Michelle Jennifer

likely to be used for girls at

David Mark 2 Robert Christopher 2 Deborah Sarah Jennifer Sarah
any time, whereas female Andrew Paul 3 David Matthew Julie Nicola Kimberly Nicole

names are rarely used for Stephen Andrew 4 James David Karen Emma Lisa Jessica
boys (after L. Dunkling & W. Mark David 5 John Jason Susan Joanne Tracy Katherine
Gosling, 1983). Michael Richard 6 Jeffrey Daniel Alison Helen Kelly Stephanie
Ian Matthew 7 Steven Robert Jacqueline Rachel Nicole Elizabeth
8 Gary Daniel 8 Christopher Eric 8 Helen Lisa 8 Angela Amanda
9 Robert Christopher 9 Brian Brian 9 Amanda Rebecca 9 Pamela Melissa
10 Richard Darren 10 Mark Joseph 10 Sharon Karen 10 Christine Lindsay

1981 1 982 (non-white) 1981 1982 (non-wn7fe)

Andrew Michael 1 Sarah Tiffany
David Christopher 2 Emma Crystal
Daniel James Claire Ebony
Christopher Jason Kelly Erica
Stephen Robert Rebecca Lakisha
Matthew Anthony Gemma Latoya
Paul Brandon Rachel Nicole
8 James 8 Kevin 8 Lisa 8 Candice
9 Mark 9 David 9 Victoria 9 Danielle
10 Michael 10 Charles 10 Laura 10 Brandi

19 NAMES 113
 names Washington, Everest, Reading, London (town The old and the
Place - 'the bold one '); new
of Londinos
The names people give to their surroundings pro- • memorable incidents or famous events, such as The two faces of the moon
vide aunique source of information about a a battle, e.g. Waterloo, Crimea, Blenheim, Cape present a marked contrast in
society's history, beliefs, and values. There are so Catastrophe, Anxious Bay, Manhattan ('the place names. The near side
reflects2,000 years of study,
many aspects of a country's development that place of great drunkenness'); and
with most
of the prominent
achieve linguistic recognition in its place names. • other place names, such as a famous city, or features being of Latin or
The various steps in the exploration of America, a town from an immigrant's home-land, e.g. Greek origin. The far side

for example, can be seen reflected in the 'layers'
of Spanish, French, Dutch, Indian, and English
names introduced by different groups of explorers;
and Celtic, Roman, Anglo-Saxon, Scandinavian,
and Norman names provide a similar insight into
British history. Often, a place name is the only
record of a historical event or of a person's exis-
tence. The name of Reada, 'the red', lives on in
the town of Reading, in Berkshire ('the people of
Reada'); but of his life and deeds, nothing else is
known.
The study of place names includes the 'small'
places and institutions (such as the names of streets,
houses, inns, and fields) as well as the main geo-
graphical features of the world (such as seas, rivers,
mountains, cities, and towns); but most academic
study has been in relation to the latter. Place names
are sometimes fanciful and idiosyncratic
Rabbit Hash, Britain Thertheoxlaydede 'there the
ox lay dead', or jocular house names such as Webil-
tit and Noname), but the vast majority can be
explained with reference to a small set of creative
processes. With geographical names, some of the
most widespread types of derivation include the
following:
• natural features, such as hills, rivers,
lines, e.g. Dover (water), Staines (stones),
lulu (safe harbour), Rotorua (two lakes),
Kalgoorlie (a native shrub), Twin Forks, South
Bend;
• special
caster
mound);
such as camps and forts, e.g. Don-
sites,
(camp on the Don), Barrow (burial
US A Paris, Memphis, Troy, Hertford, London. reflects the results of 20th-
century space exploration.
Many other factors have been recognized. There Though some traditional la-
names introduced by explorers bels are used (such as Mare
are the appealing
'sea'), the majority of new
as they encountered good and bad fortune on their
names are those of Soviet
travels (e.g. Cape Tribulation and Weary Bay in
and American astronauts,
Australia). Animal names are sometimes used (e.g. astronomers, and scientists.
Beaver City, Buffalo). And there are many names
of a purely descriptive type, such as North Sea,
1 Mare Tranquillitatis
South Island, and - perhaps the most common 2 Montes Apenninus
place name of all - the 'new town' {Newtown, Neu- 3 Clavius
ville, Naples, Villanueva, Novgorod, Neustadt, 4 Lacus Somniorum
and - less obviously, because of its Phoenician ori- Cleomedes
5
gin - Carthage). By contrast, there seem to be few 6 Archimedes
names derived from famous writers and artists:
there is a distinct absence of towns called Shake-
speare, Voltaire, or Tolstoy.
Place names have an intrinsic fascination, and
(e.g. USA many specialized studies have been undertaken.
But it must not be forgotten that many thousands
of names have an unclear or unknown etymology,
and it is this which provides a continuing motiva-
tion for place-name study, such as is carried on
by the English Place-Name Society, the American
Name Society, and similar bodies. These studies
also relate to matters of practical import. To facili-
and coast- tate international communication by post, telex,
rar s iae ^-^
Hono- and telephone, the various problems posed by
linguistic place-name variation need to be antici- /J*
pated. Place names can vary greatly between lan-
guages (e.g. Munich vs Munchen) and be f w ^J <*-\«
unrecognizable in different scripts. Names can
change along with governments (p. 9). New sys-
tems of naming may need to be introduced for spec-
I6 k
-\~~o\(a«
•» 7

• religious significance, such as gods, saints, and ial purposes, such as the means of designating

churches, e.g. Providence, Godshill, Axminster, stellar objects (e.g. NGC4565, M101). There is
St Neots, Sacramento, Santa Cruz, Thorsley thus a pressing need for international cooperation Mare Moscoviense
(fromThor); in the coining and use of place names - a need Gagarin
• royalty, e.g. Queensland, Victoria Falls, New that can only become more urgent as the explor- Apollo

ation of space proceeds. Korolev

York, Carolina, Kingston, Louisiana, Maryland,
Campbell
Fredericksburg;
Tsiolkovskiy
• explorers, e.g. America, Cookstown, Columbus,
Flinders;
• famous local people, such as presidents, politi-
cians, tribesmen, e.g. Delaware, Baltimore,

^X ^± -£X
LLANFAIRPWLLGWYNGYLLGOGERYCHWYRNDROBWLLLLANTYSILIOGOGOGOCH

J**~* **~ \v^ ,A AA.«j«L

114 •
HI THE STRUCTl'RF. OF LANGUAGE
Some common pi, ace-name elements in English
Source Language Meaning Common modern forms International brand names
International companies are finding it increasingly
ac Anglo-Saxon oak Ac-, Aik-, Oak-, Oke-, -ock
important to develop brand names that can be used
baile Gaelic farm, village Bally-, Balla-, Bal-
in a wide range of countries. To have a product with
bearu Anglo-Saxon grove, wood Barrow-, -bere, -beare, -ber
a single, universally recognized name can lead to
beorg Anglo-Saxon hill, burial-mound Bar-, Berg, -borough, -burgh
major savings in design, production, and promotion
brycg Anglo-Saxon bridge Brig-, -bridge, -brig
costs- especially now that satellite television has
burh Anglo-Saxon fortified place Bur-, -borough, -burgh, -bury
made world advertising a reality in such contexts as
burna Anglo-Saxon stream, spring Bourne-, Burn-, -burn, -borne
major sporting events. Also, from a legal point of view,
by Old Norse farm, village -by
an international trade mark has a clearer status than
caer Welsh fortified place Car-
a national one, in cases of dispute.
ceaster Latin Roman town, fort Chester-, -caster, -c(h)ester
is said that more time is spent deciding the name
It
cot Anglo-Saxon shelter, cottage Coat-, Cote, -cot, -cote
of a new product than on any other aspect of its
cwm Welsh deep valley Coombe-, Combe-, Comp-, -combe development. The problem is not easily solved. In the
daire Gaelic oak wood Derry-, -derry, -dare
European Economic Community alone, there are over
dalr Old Norse dale, valley Dal-, -dale
5 million registered trade marks, and inventing a new
denn Anglo-Saxon swine pasture -dean, -den
name that does not conflict with existing practice is a
dun Anglo-Saxon hill, down Dun-, Down-, -down, -don, -ton
highly complex and time-consuming process. Several
ea Anglo-Saxon water, river Ea-, -ey, Ya-, -eau
hundred names may need to be proposed, each of
eg Anglo-Saxon island Ea-, Ey-, -ey, -y
which has to be checked from a linguistic, marketing,
ey Old Norse island -ey, -ay
and legal viewpoint. In such fields as cars, perfumes,
gleann Gaelic narrow valley Glen- and soft drinks, thousands of possible names may
graf Anglo-Saxon grove Graf, -grave, -greave, -grove
need to be investigated to find one that is
ham Anglo-Saxon homestead, vil- Ham-, -ham internationally acceptable and registrable.
lage
An indication of the scope of the problem can be
hyrst Anglo-Saxon wooded hill Hurst-, -hurst, -hirst
seen from the experience of Dunlop, who spent over
-ing Anglo-Saxon place of -ing
two years researching a name for a new tyre, to no
lean Anglo-Saxon glade, clearing Leigh, Lee, -leigh, -ley, -le
avail. They then launched an international competition
loch Gaelic lake, inlet Loch-, -loch
amongst their employees, receiving over 10,000
mere Anglo-Saxon lake, pool Mer-, Mar-, -mer, -mere, -more
entries. Around 30 names were selected from the
nes Old Norse cape, headland -ness, Nas-, Nes-
enormous number submitted - but not one was found
pwll Welsh pool,anchorage Pol-, -pool
to be legally available in more than a small number
rhos Welsh moorland Ros-, Ross, -rose
of countries. After further work, a viable name was
stan Anglo-Saxon stone Stan-, Ston-, -stone, -ston
found (the tyre was called Denovo); but often
stede Anglo-Saxon place, site -sted, -stead
companies are not so successful, ending up with a
stoc Anglo-Saxon meeting place Stoke-, Stock-, -stock, -stoke
name that is unusable for legal or linguistic reasons.
stow Anglo-Saxon meeting place Stow(e)-, -stow(e)
A word that is pronounceable in one language may be
street Latin Roman road Strat-, Stret-, Streat-, -street
quite impossible to say in another; and there is always
tun Anglo-Saxon enclosure, village -ton, -town, Ton-
the danger of unfortunate connotations creeping in,
porp Old Norse farm, village Thorp(e), -thorpe, -throp, -trop
because of the name overlapping with words of an
pveit Old Norse glade, clearing Thwaite, -thwaite
irrelevant or taboo meaning.
WIC Anglo-Saxon dwelling, farm Wick-, Wig-, -wick, -wich

Country
The longest place name etymologies
Argentina (Spanish) 'the
A long place name would Fairbourne narrow track The new name has 66 let- (in the 1 9th century). But the silver republic'
appear tobe an unnecess- railway in North Wales was ters, thus beating the Angle- unofficial name of a 300 m Canaries (Latin) 'dogs' (not
ary complication, and is it deliberately renamed, so as sey village whose unofficial hill in Southern Hawke's Bay, canaries)
usually shortened to a more to be longer than the pre- name of 58 letters had also New Zealand, has 85 letters. Cn/7e (Araucanian) end of
speakable and writable vious British record-holder. been artificially constructed the land'
size. In this way, 'El Pueblo
Cyprus (Greek) copper'
de Nuestra Sehora la Reina Gorsafawddacha'idraigodanheddogleddolonpenrhynareurdraethceredigion
Ethiopia (Greek) 'burning
de los Angeles de Porciun- 'The Mawddach station and its dragon teeth at the Northern Penrhyn Road on the golden
face'
cula' has been reduced to beach of Cardigan Bay'
Jamaica (Carib) 'well
the more manageable 'Los
Llanfairpwllgwyngyllgogerychwyrndrobwllllantysiliogogogoch watered'
Angeles'.
St Mary's Church in a hollow by the white hazel, close to the rapid whirlpool, by the Japan (Chinese) 'sun-root'
However, an exception-
red cave of St Tysilio' (the sun rose over Japan)
allylong place name proves
Mexico (Aztec) war god,
to be a tourist attraction, Taumatawhakatangihangakoauauotamatea(turipukakapikimaungahor-
Mextli'
and in 1 984 a new candi- onuku)pokaiwhenuakitanatahu
Pakistan (an acronym)
date for longest place name 'The place where Tamatea, the man with the big knee who slid, climbed and swallowed
P(unjab) + A(fghan tribes) +
emerged. A station on the mountains, known as Land-eater, played on his flute to his loved one'
K(ashmir) + S(ind) +*
(Baluchis)tan

-£X J^± £^
GORSAFAWDDACHA'IDRAIGODANHEDDOGLEDDOLONPENRHYNAREURDRAETHCEREDIGION

ii K , n J* M.H

19
kM-

NAMES- 115
20 Discourse and text

The traditional concern of linguistic analysis has range of social situations, on such a variety of Conversation
been the construction of sentences (§16); but in topics, and with such an unpredictable set of parti-
analysis
recent years there has been an increasing interest cipants, it has proved very difficult to determine
In recent years, the phrase
in analysing the way sentences work in sequence the extent to which conversational behaviour is sys-
'conversation analysis' has
to produce coherent stretches of language. tematic, and to generalize about it. come to be used as the
Two main approaches have developed. Dis- There is now no doubt that such a system exists. name of a particular method
course analysis focusses on the structure of natur- Conversation turns out, upon analysis, to be a of studying conversational
structure, based on the tech-
ally occurring spoken language, as found in such highly structured activity, in which people tacitly
niques of the American
'discourses' as conversations, interviews, commen- operate with a set of basic conventions. A compari- sociological movement of
taries, and speeches. Text analysis focusses on the son has even been drawn with games such as chess: the 1 970s known as ethno-
structure of written language, as found in such conversations, it seems, can be thought of as having methodology.
'texts' as essays, notices, road signs,and chapters. an opening, a middle, and an end game. The partici- The emphasis in previous
sociological research had
But this distinction is not clear-cut, and there have pants make their moves and often seem to follow
been deductive and quantita-
been many other uses of these labels. In particular, certain rules as the dialogue proceeds. But the anal- tive, focussing on general
both 'discourse' and 'text' can be used in a much ogy ends there. A successful conversation is not questions of social structure.
broader sense to include all language units with a game: it is no more than a mutually satisfying The new name was chosen
to reflect a fresh direction of
a definable communicative function, whether linguistic exchange. Few rules are ever stated expli-
study, which would focus on
spoken or written. Some scholars talk about citly (some exceptions are 'Don't interrupt', and
the techniques (or methods')
'spoken and written discourse'; others about 'Look at me when I talk to you'). Furthermore, used by people themselves
'spoken and written text'. In Europe, the term text apart from in certain types of argument and debate, (oddly referred to as 'ethnic'),
linguistics is often used for the study of the linguis- there are no winners. when they are actually en-
gaged in - and thus
social
tic principles governing the structure of all forms
linguistic- interaction. The
of text. Conversational success
central concern was to deter-
The search and struc-
for larger linguistic units For a conversation to be successful, in most social mine how individuals exper-
tures has been pursued by scholars from many dis- contexts, the participants need to feel they are con- ience, make sense of, and
report their interactions.
ciplines. Linguists investigate the features of tributing something to it and are getting something
In conversation analysis,
language that bind sentences when they are used out of it. For this to happen, certain conditions the data thus consist of tape
in sequence. Ethnographers and sociologists study must apply. Everyone must have an opportunity recordings of natural conver-
the structure of social interaction, especially as to speak: no one should be monopolizing or con- sation, and their associated

manifested in the way people enter into dialogue. stantly interrupting. The participants need to make transcriptions. These are
then systematically analysed
Anthropologists analyse the structure of myths and their roles clear, especially if there are several possi-
to determine what properties
folk-tales. Psychologists carry out experiments on 'Speaking as a mother/linguist/Catho-
bilities (e.g. govern the way in which a
the mental processes underlying comprehension. lic. .'). They need to have a sense of when to speak
. conversation proceeds. The
And further contributions have come from those or stay silent; when to proffer information or hold approach emphasizes the
need for empirical, inductive
concerned with artificial intelligence, rhetoric, phi- it back; when to stay aloof or become involved.
work, and in this it is some-
losophy, and style (§12). They need to develop a mutual tolerance, to allow times contrasted with 'dis-
These approaches have a common concern: they for speaker unclarity and listener inattention: per- course analysis', which has
stress the need to see language as a dynamic, social, fect expression and comprehension are rare, and often been more concerned
interactive phenomenon - whether between the success of a dialogue largely depends on people with formal methods of ana-
lysis (such as the nature of
speaker and listener, or writer and reader. It is recognizing their communicative weaknesses,
the rules governing the struc-
argued that meaning is conveyed not by single sen- through the use of rephrasing (e.g. 'Let me put that ture of texts).
tences but by more complex exchanges, in which another way') and clarification (e.g. 'Are you with
the participants' beliefs and expectations, the me?').
knowledge they share about each other and about There is a great deal of ritual in conversation,
the world, and the situation in which they interact, especially at the beginning and end, and when
play a crucial part. topics change. For example, people cannot simply
leave a conversation at any random point, unless
they wish to be considered socially inept or ill-man-
nered. They have to choose their point of departure
Conversation (such as the moment when a topic changes) or con-
Of the many types of communicative act, most struct a special reason for leaving. Routines for
study has been devoted to conversation, seen as concluding a conversation are particularly com-
the most fundamental and pervasive means of con- plex, and cooperation is crucial if it is not to end
ducting human affairs (p. 52). These very charac- abruptly, or in an embarrassed silence. The parties
however, complicate any investigation.
teristics, may prepare for their departure a long way in
Because people interact linguistically in such a wide advance, such as by looking at their watches or

116 •
III THE STRUCTURE OF LANGUAGE
giving a verbal early warning. A widespread con-
vention is for visitors to say they must leave some CONVERSATIONAL MAXIMS
The success of a conversation depends not only on
time before they actually intend to depart, and for
what speakers say but on their whole approach to the
the hosts to ignore the remark. The second mention interaction. People adopt a cooperative principle'
then permits both parties to act. when they communicate: they try to get along with
The topic of the conversation is also an impor-
each other by following certain conversational
'maxims' that underlie the efficient use of language.
tant variable. In general it should be one with which Four basic maxims have been proposed (after H. P.
everyone feels at ease:
'safe' topics between strangers Grice, 1975):
in English situations usually include the weather,
• The maxim of quality states that speakers'
pets, children, and the local context (e.g. while
contributions to a conversation ought to be true.
waiting in a room or queue); 'unsafe' topics include They should not say what they believe to be false,
religious and political beliefs and problems of nor should they say anything for which they lack
health. There are some arbitrary divisions: asking adequate evidence.
• The maxim of quantity states that the contribution
what someone does for a living is generally safe:
should be as informative as is required for the
asking how much they earn is not. Cultural vari- purposes of the conversation. One should say
ations can cause problems: commenting about
the neither too little nor too much.
cost of the furniture or the taste of a meal may • The maxim of relevance states that contributions
should clearly relate to the purpose of the exchange.
be acceptable in one society but not in another.
• The maxim of manner states that the contribution
It is difficult to generalize about what is normal, should be perspicuous - in particular, that it should
polite, or antisocial in conversational practice, as be orderly and brief, avoiding obscurity and
there is so much cultural variation. Silence, for ambiguity.
example, varies in status. It is an embarrassment Other maxims have also been proposed, such as 'Be
in English conversations, unless there are special polite', Behave consistently.' The principle of
reasons (such as in moments of grief). However, relevance is most attention, as
currently attracting it

some has been proposed as a fundamental explanatory

in cultures (e.g. Lapps, Danes, the Western
principle for a theory of human communication (D.
Apache) it is quite normal for participants to Sperber& D. Wilson, 1986).
become silent. Often, who speaks, and how much Listeners will normally assume that speakers are
is spoken, depends on the social status of the parti- following these criteria. Speakers may of course break

cipants - for example, those of lower rank may (or 'flout') these maxims - for example, they may lie,

be sarcastic, be different, or clever - but

try to
be expected to stay silent if their seniors wish to
conversation proceeds on the assumption that they are
speak (p. 38). Even the basic convention of 'one not doing so. Listeners may then draw inferences from
person speaks at a time' may be broken. In Antigua, what speakers have said (the literal meaning of the
for example, the phenomenon of several people utterance) concerning what they have nor said (the
implications, or 'implicatures' of the utterance). For
speaking at once during a whole conversation is
example,
a perfectly normal occurrence.
A: I need a drink.
B: Try The Bell.

If B is adhering to the cooperative principle, several

implicatures arise out of this dialogue: for example,
The Bell must be a place that sells drinks; it must be
open (as as B knows); it must be nearby. If B is
far
not being cooperative (e.g. if he knows that The Bell
is closed, or is the name of a greengrocer's), he is

flouting the maxims of quality and relevance.

Deliberate flouting of this kind is uncommon, of
course, and only occurs in such special cases as
sarcasm, joking, or deliberate unpleasantness. More
likely is the inadvertent flouting of conversational
maxims - as would happen if B genuinely did not know
that The Bell was closed, and accidentally sent A on
a wild goose chase. In everyday conversation,
misunderstandings often take place as speakers make
assumptions about what their listeners know, or need
to know, that turn out to be wrong. At such points, the
conversation can break down and may need to be
repaired', with the participants questioning, clarifying,
and cross-checking. The repairs are quickly made in
the following extract, through the use of such pointers
as told you' and sorry'.
'I

Bob Newhart A: Have you got the time?

Newhart's comedy routines often rely on the B: No, I told you, I lost my watch.
audience's awareness of discourse conventions. His A: Oh, sorry, I forgot.
'driving instructor' sketch, for example, gives us only
half of the conversation, from the instructor's viewpoint,
But it is quite common for participants not to realize

leaving the responses of the learner driver to our

that there has been a breakdown, and to continue
conversing at cross purposes.
imagination. Joyce Grenfell's teaching young children'
sketches were based on the same principle.

20 DISCOURSE AND TEXT •

117
 . . . . . . . . . . . . .

CONVERSATIONAL TURNS are used in sequences. The units have been called Misunderstandings
Probably the most widely recognized conversa- 'exchanges' or 'interchanges', and in their minimal An important aim of dis-
tional convention is that people take turns to speak. form consist simply of an initiating utterance (I) course analysis is to find out
But how do people know when it is their turn? followed by a response utterance (R), as in: why conversations are not
Some rules must be present, otherwise conversa- always successful. Mis-
I: What's the time? understanding and mutual
tions would be continually breaking down into a
disorganized jumble of interruptions and simulta-
R: Two o'clock. recrimination is unfortunately
fairly common. Participants
neous talk. In many formal situations, such as com- Two-part exchanges (sometimes called 'adjacency often operate with different
mittee meetings and debates, there are often commonplace, being used in such con-
pairs') are rules and expectations about

explicit markers showing that a speaker is about texts as questioning/answering, informing/ack- the way in which the conver-
sation should proceed -
to yield the floor, and indicating who should speak nowledging, and complaining/excusing. Three-
something that is particularly
next ('I think Mr Smith will know the answer to part exchanges are also important, where the res- evident when people of dif-
that question'). This can happen in informal situa- ponse is followed by an element of feedback (F). ferent cultural backgrounds

tions too ('What do you think, John?'), but there Such reactions are especially found in teaching sit- interact. But even within a
culture, different 'rules of in-
the turn-taking cues are usuallymore subtle. uations:
terpretation' may exist.
People do not simply stop talking when they are teacher: Where were the arrows kept? has been suggested, for
It
(I)
ready to yield the floor. They usually signal some pupil: In a special kind of box. (R) example, that there are dif-
way in advance that they are about to conclude. ferent rules governing the
teacher: Yes, that's right, in a box. (F) way in which men and
The clues may be semantic ('So anyway, .', 'Last . .
women participate in a con-
but not least .'); but more commonly the speech
What
of particular interest is to work out the
is
. .
versation (p. 21 ). A common
itself can be modified to show that a turn is about
constraints that apply to sequences of this kind.
source of misunderstanding
to end - typically, by lowering its pitch, loudness,
The teacher— feedback sequence would be inappro- is the way both parties use

or speed. Body movements and patterns of eye con- priate in many everyday situations: head nods and mhm noises
while the other is speaking -
tact are especially important. While speaking, we A: Did you have a good journey? something that women do
look at and away from our listener in about equal B: Apart from a jam at Northampton. much more frequently than
proportions; but as we approach the end of a turn, A: *Yes, that's right, a jam at Northampton. men. Some analysts have
we look at the listener more steadily. Similarly, suggested that the two sexes
Unacceptable sequences are easy to invent: mean different things by this
when talking to a group of people, we often look
behaviour. When a woman
more steadily at a particular person, to indicate A: Where do you keep the jam ? does it, she is simply indicat-
that in our view this should be the next speaker. B: *It's raining again. ing that she is listening, and
Listeners are not passive in all of this. Here too encouraging the speaker to
On the other hand, with ingenuity it is often pos- continue, but the male inter-
there are several ways of signalling that someone
sible to imagine situations where such a sequence prets it to mean that she is
wants to talk next. Most obviously, the first person agreeing with everything he
could occur (e.g. if B were staring out of the win-
in a group actually to start speaking, after the com- is saying. By contrast, when
dow at the time). And discourse analysts are always
pletion of a turn, will usually be allowed to hold a man does he is signal-
it,
on the lookout for unexpected, but perfectly accep- ling that he does not neces-
the floor. More subtly, we can signal that we want
table, sequences in context, such as: sarily agree, whereas the
to speak next by an observable increase in body
woman interprets it to mean
tension - by leaning forward, or producing an aud- A: Goodbye. that he is not always listen-
ible intake of breath. Less subtly, we can simply B: Hello. ing. Such interpretations are
interrupt — a strategy which may be tolerated, if plausible, it is argued, be-
(used, for example, as A is leaving an office, passing cause they explain two of the
the purpose is to clarify what the speaker is saying,
B on his way in). Many jokes, too, break discourse most widely reported reac-
but which more usually leads to social sanctions. tions from participants in
rules as the source of their effect:
cross-sex conversations -
EXCHANGES A: Yes, I can. the male reaction of It's im-
Because conversational discourse varies so much B: Can you see into the future? possible to say what a

in length and complexity, analysis generally begins woman really thinks', and the
female reaction of 'You
by breaking an interaction down into the smallest never listen to a word say.' I

possible units, then examining the way these units (AfterD.N.Maltz&R.A.

Borker, 1982.)

Conversation manoeuvres
Conversational turn-taking Did you hear the one Are you with me? Changing topic Let's wait and see.
is often marked by clear sig- about . . Do make myself clear?
I Introducing a new topic:
nals of direction. Can you spare a minute? Don't you think? That reminds me . . Ending
Halt! Who goes there? Let me put it another way . Incidentally . . Sorry, but I have to go now.
Openings But not: *How much do you Don't get me wrong . . That's a good question. Nice talking to you.
Guess what . . earn? What I'm trying to say is . . By the way . . Well,must get back to work.
Sorry to trouble you . . By the listener: Speaking of John . . Gosh, is that the time?
Lovely day! Ongoing checks You mean . . Where was I? Imustn't keep you.
Got a match? By the speaker: Have got you right?
I Concluding a topic:
Can help you?
I Do you see? Mhm. So it goes.
Good morning. Can you guess what he I don't get you. That's life.
Excuse me . . said? Let's get that straight . . Makes you think, doesn't it.

118 • HI THE STRUCTURE OF LANGUAGE

 A text plainly has to be coherent as well as cohesive, Macrostructures
Textual structure in that the concepts and relationships expressed Not all textual analysis starts
To call a sequence of sentences a imply 'text' is to should be relevant to each other, thus enabling us with small units and works
that the sentences display some kind of mutual to make plausible inferences about the underlying from the 'bottom up' (p. 71 );

some approaches aim to

dependence; they are not occurring at random. meaning.
make very general state-
Sometimes the internal structure of a text is imme- ments about the macrostruc-
diately apparent, as in the headings of a restaurant ture of a text. In psychology,
menu; sometimes it has to be carefully demon- example, attempts have
Two ways of demonstrating for
been made to analyse narra-
strated, as in thenetwork of relationships that enter
cohesion tives into schematic outlines
into a literary work. In all cases, the task of textual that represent the
Paragraphs are often highly cohesive entities. The elements
analysis is to identify the linguistic features that cohesive ties can stand out very clearly if the in a story that readers re-
cause the sentence sequence to 'cohere' — something sentences are shuffled into a random order. It may member. These schemata
that happens whenever the interpretation of one even be possible to reconstitute the original sequence have been called story-
solely by considering the nature of these ties, as in grammars' (though this is an
feature is dependent upon another elsewhere in the
the following case: unusually broad sense of the
sequence. The ties that bind a text together are term 'grammar', cf. §16).
1 However, nobody had seen one for months.
often referred to under the heading of cohesion 2. He thought he saw a shape in the bushes. In one such approach

(after M. A. K. Halliday R. Hasan, 1976). & 3. Mark had told him about the foxes. (after P. W. Thorndyke,

4. John looked out of the window. 1977), simple narratives are

Several types of cohesive factor have been recog-
5. Could it be a fox? analysed into four com-
nized: ponents: setting, theme, plot,
(The original sequence was 4,2,5,3,1 .)
• Conjunctive relations What is about to be said We can use graphological devices to indicate the and resolution. The setting
has three components: the
is explicitly related to what has been said before, patterns of cohesion within a text. Here is the closing
paragraph of James Joyce's short story A Painful characters, a location, and a
through such notions as contrast, result, and time: time. The theme consists of
Case'. The sequence of pronouns, the anaphoric
I left early. However, Mark stayed till the end. definite articles, and the repeated phrases are the main an event and a goal. The plot
cohesive features between the clauses and sentences. consists of various episodes,
Lastly, there's the question of cost.
Several of course refer back to previous parts of the each with its own goal and
• Coreference Features that cannot be semanti- story, thus making this paragraph, out of context, outcome. Using distinctions
of this kind, simple stories
cally interpreted without referring to some other impossible to understand.
are analysed into these com-
feature in the text. Two types of relationship are
ponents, to see whether the
recognized: anaphoric relations look backwards He turned back the way he had come, the rhythm same kinds of structure can
for their interpretation, and cataphoric relations of the engine pounding in his ears. He began to be found in each (p. 79). Cer-
look forwards: doubt the reality of what memory told him. He halted tain similarities do quickly
emerge; but when complex
under a tree and allowed the rhythm away. He
Several people approached. They seemed angry. to die
narratives are studied, it

Listen to this: John's getting married. could not feel her near him in the darkness nor her proves devise
difficult to

voice touch his ear. He waited for some minutes more detailed categories that
• Substitution One feature replaces a previous are capable of generaliza-
listening . He could hear NOTHING: the night was
expression: tion, and analysis becomes
perfectly silent. He listened again perfectly silent.
: increasingly arbitrary.
I've got a pencil. Do you have one} He felt that he was ALONE.
*~

Will we get there on time? I think so.

• Ellipsis A piece of structure is omitted, and can

be recovered only from the preceding discourse:
Where did you see the car? a In the street. U> |i»| -y

• Repeated forms An expression is repeated in

whole or in part:

Canon Brown arrived. Canon Brown was cross.

• Lexical relationships One lexical item enters
into a structural relationship with another (p. 105):
The flowers were lovely. He liked the tulips best.

• Comparison A compared expression is pre-

supposed in the previous discourse:
That house was bad. This one's far worse.

Cohesive links go a long way towards explaining Conceptual structure One a New Mexico desert. ae affected entity
way of representing the con- Empty, it weighed five tons. at attribute of
how the sentences of a text hang together, but they
ceptual structure of a text For fuel it carried eight tons co containment of
do not tell the whole story. It is possible to invent (after R. de Beaugrande & of alcohol and liquid oxygen. to location of
a sentence sequence that highly cohesive but
is W. Dressier, 1981, p. 100). pu purpose of
The abbreviations identify
nonetheless incoherent (after N. E. Enkvist, 1978, This 'transition network' qu quantity of
the types of semantic links
summarizes the following sp specification of
p. 110): which relate the concepts
paragraph: st state of
(following the direction of the
A week has seven days. Every day I feed my cat. su substance of
A great black and yellow V-2 arrows):
Cats have four legs. The cat is on the mat. Mat rocket 46 feet long stood in
has three letters.

20 DISCOURSE AND TEXT 119

21 Pragmatics
Pragmatics studies the factors that govern our
choice of language in social interaction and the
effects of our choice on others. In theory, we can
say anything we like. In practice, we follow a large
number of (most of them uncon-
social rules
sciously) that constrain the way we speak. There
is no law that says we must not tell jokes during
a funeral, but it is generally 'not done'. Less
obviously, there are norms of formality and polite-
ness that we have intuitively assimilated, and that
we follow when talking to people who are older,
of the opposite sex, and so on. Writing and signing
behaviour are constrained in similar ways.
Pragmatic factors always influence our selection
of sounds, grammatical constructions, and vocabu-
lary from the resources of the language. Some of
the constraints are taught to us at a very early age
— in British English, for example, the importance
of saying please and thank you, or (in some fami-
lies)of not referring to an adult female in her
presence as she. In many languages, pragmatic
distinctions of formality, politeness, and intimacy
are spread throughout the grammatical, lexical,
and phonological systems, ultimately reflecting
matters of social class, status, and role (§10, p. 99).
A well-studied example is the pronoun system,
which frequently presents distinctions that convey
pragmatic force - such as the choice between tu
and vous in French.
Languages differ greatly in these respects. Polite-
ness expressions, for instance, may vary in fre-
quency and meaning. Many European languages
do not use their word for please as frequently as
English does; and the function and force of thank
you may also alter (e.g. following the question
'Would you like some more cake?', English thank
you means 'yes', whereas French merci would mean
'no'). Conventions of greeting, leave-taking, and
dining also differ greatly from language to lan-
guage. In some countries it is polite to remark to
a host that we are enjoying the food; in others
it is polite to stay silent. On
one occasion, at a
dinner in an Arabic community, the present author
made the mistake of remarking on the excellence
of the food before him. The host immediately apo-
logized, and arranged for what was there to be
replaced!
Pragmatic errors break no rules of phonology,
syntax, or semantics. The elements of How's tricks,
your majesty? will all be found in English language
textbooks and dictionaries, but for most of us the
sequence is not permissible from a pragmatic view-
point. Pragmatics has therefore to be seen as separ-
ate from the 'levels' of language represented in
linguistic models of analysis (§13). It is not a 'part'
of language structure, but its domain is so closely
120 •
III THE STRUCTURE OF LANGUAGE
bound up with structural matters that it cannot
be ignored in this section of the encyclopedia.
THE IDENTITY OF PRAGMATICS
Pragmatics is not at present a coherent field of
study. A large number of factors govern our choice
of language in social interaction, and it is not yet
clear what they all are, how they are best inter-
related, and how best to distinguish them from
other recognized areas of linguistic enquiry'. There
are several main areas of overlap.
Semantics (§17) Pragmatics and semantics both
take into account such notions as the intentions
of the speaker, the effects of an utterance on lis-
teners, the implications that follow from expressing
something in a certain way, and the knowledge,
beliefs, and presuppositions about the world upon
which speakers and listeners rely when they inter-
act.
Stylistics (§12) and sociolinguistics (§§10, 63)
These fieldsoverlap with pragmatics in their study
of the social relationships which exist between par-
ticipants, and of the way extralinguistic setting, ac-
tivity,and subject-matter can constrain the choice
of linguistic features and varieties.
Psycholinguistics (§§7, 38) Pragmatics and psy-
cholinguistics both investigate the psychological
states and abilities of the participants that will have
a major effect upon their performance — such fac-
tors as attention, memory, and personality.
Discourse analysis (§20) Both discourse analysis
and pragmatics are centrally concerned with the
analysis of conversation, and share several of the
philosophical and linguistic notions that have been
developed to handle this topic (such as the way
information is distributed within a sentence, deictic
forms (p. 106), or the notion of conversational
'maxims' (p. 117)).
As a result of these overlapping areas of interest,
several conflicting definitions of the scope of prag-
matics have arisen. One approach focusses on the
factors formally encoded in the structure of a lan-
guage (honorific forms, tu/vous choice, and so on).
Another relates it to a particular view of semantics:
here, pragmatics is seen as the study of all aspects
of meaning other than those involved in the analy-
sis of sentences in terms of truth conditions
(p. 107). Other approaches adopt a much broader
perspective. The broadest sees pragmatics as the
study of the principles and practice underlying all
interactive linguistic performance — this including
all aspects of language usage, understanding, and
appropriateness. Textbooks on pragmatics to date,
accordingly, present a diversity of subject matter,
and a range of partially conflicting orientations and • Expressives The speaker expresses an attitude Indirect speech
methodologies, which proponents of the subject about a state of affairs, e.g. apologize, deplore, acts
have yet to resolve. However, if we take diversity congratulate, thank, welcome. Some speech acts directly
of opinion to be a sign of healthy growth in a sub- • Declarations The speaker alters the external address a listener, but the

status or condition of an object or situation majority of acts in everyday

ject, it must be said that few other areas of language
conversation are indirect.
study have such a promising future. solely by making the utterance, e.g. / resign, I
For example, there are a
baptize, You're fired, War is hereby declared. very large number of ways of
asking someone to perform
Speech acts FELICITY CONDITIONS anaction. The most direct

The British philosopher J. L. Austin (1911-1960) Speech acts are successful only if they satisfy several
way is to use the imperative
construction (Shut the door),
was the first to draw attention to the many func- criteria, known as 'felicity conditions'. For exam- but is easy to sense that
it

tions performed by utterances as part of interper- ple, the 'preparatory' conditions have to be right: this would be inappropriate
sonal communication. In particular, he pointed out the person performing the speech act has to have in many everyday situations
that many utterances do not communicate infor- the authority to do so. This is hardly an issue with - too abrupt or rude, per-
haps. Alternatives stress
mation, but are equivalent to actions. When such verbs as apologize, promise, or thank, but
such factors as the hearer's
someone says 'I apologize .', 'I promise
. . .', 'I . . it is an important constraint on the use of such
ability or desire to perform
will' (at a wedding), or 'I name this ship .', the . . verbs as fine, baptize, arrest, and declare war, the action, or the speaker's
utterance immediately conveys a new psychologi- where only certain people are qualified to use these reasons for having the action
done. These include the fol-
cal or social reality. An apology takes place when utterances. Then, the speech act has to be executed
lowing:
someone apologizes, and not before. A ship is in the correct manner: in certain cases there is a
I'd be grateful if you'd shut
named only when the act of naming is complete. procedure to be followed exactly and completely the door.
In such cases, to say is to perform. Austin thus (e.g. baptizing); in others, certain expectations Could you shut the door?
called these utterances performatives, seeing them have to be met (e.g. one can only welcome with Would you mind shutting
the door?
as very different from statements that convey a pleasant demeanour). And, as a third example,
It'd help to have the door
information {constatives). In particular, performa- 'sincerity' conditions have to be present: the speech shut.
tives are not true or false. If A says 'I name this act must be performed in a sincere manner. Verbs It's getting cold in here.
ship .', B cannot then say 'That's not true'!
. . such as apologize, guarantee, and vow are effective Shall we keep out the
In speech act analysis, we study the effect of only if speakers mean what they say; believe and draught?
Now, Jane, what have you
utterances on the behaviour of speaker and hearer, affirm are valid only if the speakers are not lying. forgotten to do?
using a threefold distinction. First, we recognize Ordinary people automatically accept these con- Brrr!
the bare fact that a communicative act takes place: ditions when they communicate, and they depart Any of these could, in the
the locutionary act. Secondly, we look at the act from them only for very special reasons. For exam- right situation, functionas a
request for action, despite
that is performed as a result of the speaker making ple, the request Will you shut the door? is appro-
the fact that none has the
an utterance - the cases where 'saying = doing', priate only if (a) the door is open, (b) the speaker clear form of an imperative.
such as betting, promising, welcoming, and warn- has a reason for asking, and (c) the hearer is in But of course, it is always
ing: these, known as illocutionary acts, are the core a position to perform the action. If any of these open to the hearer to mis-
understand an indirect re-
of any theory of speech acts. Thirdly, we look at conditions does not obtain, then a special interpre-
quest - either accidentally or
the particular effect the speaker's utterance has on tation of the speech act has to apply. It may be
deliberately.
the listener, who may feel amused, persuaded, intended as a joke, or as a piece of sarcasm. Alterna- Teacher: Johnny, there's
warned, etc., as a consequence: the bringing about tively, of course, there may be doubt about the some chalk on the floor.
of such effects is known as a perlocutiottary act. speaker's visual acuity, or even sanity! Johnny: Yes, there is, sir.
Teacher: Well, pick it up,
It is important to appreciate that the illocutionary
then!
force of an utterance and its perlocutionary effect
may not coincide. If I warn you against a particular The first part of this notice is

cast in the form of a statement,

course of action, you may or may not heed my
but there is no doubt as to the
warning. directive illocutionary force
There are thousands of possible illocutionary intended by the writer. The
acts, and several attempts have been made to clas- perlocutionary effect, however,
is not as anticipated!
sify them into a small number of types. Such classi-
fications are difficult, because verb meanings are
often not easy to distinguish, and speakers' inten-
tions are not always clear. One influential approach
sets up five basic types (after J. R. Searle, 1976):

• Representatives The speaker is committed, in

varying degrees, to the truth of a proposition,
e.g. affirm, believe, conclude, deny, report.
• Directives The speaker tries to get the hearer
to do something, e.g. ask, challenge, command,
insist, request.
• Commissives The speaker is committed, in
varying degrees, to a certain course of action,
e.g. guarantee, pledge, promise, swear, vow.

21 PRAGMATICS • 121
 PART IV
The medium of language: speaking
and listening
The core of language study is the analysis of
ing and grammatical expression - matters that
its
are treated under the heading of 'The Structure of
Language' in Part in. These properties of language
underlie all forms of linguistic communication, and
must be recognized whether we choose to commu-
nicate through the medium of speech, writing, sign-
ing, or any other. At the same time, each medium
has its own distinctive properties that require inde-
pendent study. These are the subject matter of Parts
IV-VI.
Webegin with the study of sound, which is
the most universal and natural medium for the
transmission and reception of language. Writing
holds a less central place, in the history of both
the individual and the human species. No commu-
nity has ever been found to lack spoken language,
but only a minority of languages have ever been
written down. Likewise, the vast majority of
human beings learn to speak (for those who do
not, see Part vm); but it is only in recent years
that some of these people have learned to write.
Part iv provides an account of the chief factors
involved in the production, transmission, and
reception of speech. It begins with a description
of the aspects of human anatomy and physiology
involved in speech production — the vocal organs
(§22). This is followed by an account of the acous-
An X-ray radiographic slide showing a speaker pronouncing
an [i] vowel.
mean- tic properties of speech — the nature of the sound
waves that are set in motion as a result of vocal
organ activity (§23). Some of the main instruments
used in the study of speech physiology and acous-
tics are reviewed in §24. The process of speech
reception then introduced (§25), with reference
is
to both the mechanism of hearing and the little-
understood task of speech perception. And all three
stages - production, transmission, and reception
- are brought together, in order to attack the pro-
blem of how to promote speech interaction with
machines (§26).
Phonetics, the science of speech sounds, provides
the framework for §27, which reviews the whole
range of sounds that the vocal tract can produce,
and presents the symbols used in making a phonetic
transcription. This leads naturally to a discussion
of the kinds of sound most commonly used in the
world's languages, and of the way sounds interre-
late as part of a language system (§28). The main
notions in the subject of phonology, which studies
the properties of sound systems, are introduced
during this section. Part iv then concludes with
an account of such effects as intonation, rhythm,
and tones of voice - the 'suprasegmentaP aspects
of spoken language (§29) - and a discussion of
sound symbolism (§30).
22 The anatomy and physiology of speech

The vocal organs General arrangement

organs
of the vocal
The vocal organs body that
are those parts of the
are involved in the production of speech. The name
'vocal organs' is not entirely appropriate, as their palate
main function is in fulfilling the basic biological
needs of breathing and eating. But there seems to
have been considerable evolutionary development
.tongue
in their form, which enables them to function effi-
ciently for the act of speech (§49). .pharynx
Perhaps the most striking feature of any diagram .epiglottis
of the vocal organs is the amount of the body
involved in speech. It is not simply the mouth and .larynx
throat; we have to show the involvement of the
lungs, the trachea (or windpipe) and the nose. trachea (windpipe)
Inside the mouth, we have to distinguish the ton- clavicle
gue, and the various parts of the palate. Inside the
throat, we need to distinguish the upper part, or sternum (ribs and
breastbone)
pharynx, from the lower part, or larynx, which
contains the vocal folds - commonly called the .lungs
vocal 'cords'. The pharynx, mouth, and nose form
a system of hollow areas, or cavities, known as
the vocal tract (though this term sometimes
includes larynx and lungs as well). When we move
the organs in the vocal tract, we alter its shape,
thoracic (chest)
and it is this which enables the many different .

cavity
sounds of spoken language to be produced.

THE LUNGS
Before any sound can be produced at all, there has
to be a source of energy (p. 132). In speech, the
energy takes the form of a stream of air, which
has in normal circumstances been set in motion
by the lungs.
The lungs are found in a cavity in the chest (or
thorax) known as the thoracic cavity. This cavity diaphragm

is bounded back by the spinal column, at

at the
the front by the ribs and breastbone (or sternum),
and at the bottom by the dome-shaped muscle
known as the diaphragm, which separates the lungs
from the lower cavities of the abdomen. The struc-
ture surrounding the thoracic cavity is referred to
as the thoracic cage. The act of respiration takes
place through the action of the thoracic cage, which
enables the lungs to act as a kind of bellows, allow-
ing air to flow inwards and outwards. equalize the pressure with that of the atmosphere
In order to speak, we must first inhale. Signals outside the body.
from the nerve centre stem (where res-
in the brain We then exhale. We contract the chest, and thus
piration is controlled) cause the muscles of the thor- the lungs, by lowering the ribs and raising the
acic cage to contract: in particular, the muscles diaphragm, forcing the air out. But we never exhale
between the ribs (the intercostal muscles) cause the all the air. Only about a quarter of the air in the

ribs to move upwards and outwards, and the dia- lungs is used while we are engaged in normal con-
phragm to move downwards. The result is to versation — though the amount increases to some
expand the chest, and thus the lungs, temporarily extent if our speech becomes loud or effortful, as
causing the air pressure in the lungs to be reduced. in shouting, acting, public speaking, or producing
Air immediately flows into the lungs, in order to a 'stage whisper'.

124 IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

USING THE LUNGS FOR SPEECH The respiratory cycle
Lung often referred to as pulmonic air. When
air is The sequence of events involved when we breathe
pulmonic air flows outwards, it is said to be egres- in [inspiration) and out [expiration) is known as
sive. The vast majority of speech sounds are made the respiratory cycle. Normally, the two halves of
using pulmonic egressive air. this cycle are nearly equal in duration; but when
It is also possible - though not usual - to speak we speak, the pattern changes to one of very rapid
while the air stream is flowing inwards to the lungs inhalation and very slow exhalation. The rate at
(pulmonic ingressive air). We occasionally hear this which we breathe also changes. When we are silent
air stream used in English when someone is trying and at rest, our average rate is 12 breaths a minute,
to talk while laughing or crying, or when out of so the time we take to inhale and exhale is about
breath. Words such as yes and no are sometimes 2\ seconds each. During speech, we cut down the
said with an ingressive air stream, when we use time for inhaling to as little as a quarter of a second,
a 'routine' tone of voice to acknowledge what and we regularly extend the time for exhaling to Right and wrong
someone is saying. An alternate use of egressive 5 or 10 seconds - even 20 seconds is possible, ways to breathe
and ingressive streams is sometimes heard when
air depending on the speaker's voice control, emotion- Techniques of breath control
people are counting rapidly, 'under their breath'. al state, and other such factors. This altered pattern form an essential part of any
training programme to im-
But ingressive speech is of poor quality, muffled, of breathing enables our exhalations to 'carry'
prove the use of the voice -
and croaky, and many people find it unpleasant much larger amounts of speech than would other- as in singing, drama, or
to listen to. It is never put to routine use in everyday wise be the case. In everyday conversation, it is speech therapy (§46). The
English speech. quite normal to produce from 250 to 300 syllables most efficient breathing re-
in a minute. quires a rapid inspiration and
a measured expiration that
exactly meets the needs of
the voice. The careful control
and diaphragm move-
of rib
Pneumotachograph (Right) Inside the face mask of this ment is the main feature of
instrument there are separate meters for monitoring the the intercostal diaphragmatic
volume of air flow from the mouth and the nose. method of breathing, which
specialists consider to be the
most efficient for most pur-
poses. It contrasts with sev-
Spirometer (Below) The spirometer measures the volume of methods,
eral less efficient
air produced by the lungs. The information is plotted on a such as the tense, nervous
revolving drum. form of breathing in the up-
per part of the chest, known
as clavicular breathing. If it is
habituallyused during
\A speech, the excessive mus-
cular tension required to
maintain this breathing pat-
tern can strain the vocal
folds, causing hoarseness
and other abnormal voice
qualities.The diagrams
> > above show correct (a) and
***«, incorrect (b) body positions
for singing.
t

An 18th-century plethys-
mograph Plethysmography
is a technique for determin-
ing the size of a part of the
body, and it can be used to
investigate the lung volume.
This plethysmograph was
simply a barrel with an air-
tight collar at the neck. The
act of breathing alters the
subject's body volume,
which in turn alters the
amount of air in the barrel. A
small pipe at the top of the
barrel permits the pasage of
air. Monitoring the changes
in airflow through the pipe
thus provides a measure of
the changes in lung volume.
A development of the same
technique is still used today.

22 THE ANATOMY AND PHYSIOLOGY OF SPEECH •

125
Speaking without the lungs
The vowels and consonants of English, as of most
languages, are all made using pulmonic egressive
air. But there are several other types of speech
sound which do not use an air-stream from the
lungs, and these are encountered in many languages
of the world.

CLICKS
One of the most distinctive types of non-pulmonic
sound is the click. Click sounds are sharp, suction
noises, made by the tongue or lips. For example,
the noise we write as tut tut (or tsk tsk) is a pair
of click sounds, made by the tongue against the
top teeth. While making a click sound, it is possible
to breathe in and out, quite independently, showing
that the lungs are not involved in their production.
In European languages, isolated click sounds are
often heard as meaningful noises, but they are not
part of their system of vowels and consonants
(§28). The tut tut click, for example, expresses dis-
approval in English, but the sound is not used as Miriam Makeba Her record-
part of a word, in the way that /t/ and /p/ are. were
ings of click songs'

However, in many other languages, clicks are used popular in the 1960s. A
native speaker of Xhosa, she
as consonants. Most well known are some of the
used words containing click
languages of southern Africa, often referred to as consonants in her songs,
'click languages'. !Xu is one such language, with achieving notable effects by
many articulating them with great
as as 48 clicks (p. 168). The Khoisan
resonance.
languages, which include the languages of the
Khoikhoin (Hottentot) and San (Bushmen) tribes,
have the most complex click systems, using many How click sounds are made
different places of articulation in the mouth, and Near East. Lateral clicks made
A click sound is produced at the front of the
involving the simultaneous use of other sounds solely in the mouth. The air (phonetic symbol [a]) are mouth, using the tongue or
made in the throat or nose. flow is controlled by move- made with the sides of the lips. This forms a cavity in

ments which take place tongue, and are heard in the mouth, cut off from the
against the back part of the the noises of encourage- air outside.
GLOTTALIC SOUNDS roof of the mouth, known ment made to horses or
The space behind the Adam's apple, between the as the velum (p. 1 30). Be- other animals (including 2. The body of the tongue
vocal folds, is known can use
as the glottis. We cause of this, click sounds the human). A click sound is moved slightly down-
are described as using a made at the lips would be wards and backwards, so
the glottis to start an air-stream moving, and -
ve/ar/c air-stream mechan- known as a bilabial click as to form a partial vacuum
several languages make use of sounds based on this ism. Tut tut is an example made with the lips puck- inside the cavity.
principle, referred to as the glottalic air-stream of a double dental click ered, it is often used as a
mechanism. When the glottis makes the air move (phonetic symbol [j]), be- kiss at a distance'. 3. When the tongue is sud-
cause the teeth have been denly lowered, or the lips
inwards, the sounds are called implosives. An
involved in its production. A 1. The back of the tongue opened, air rushes in from
implosive consonant is a glottalic ingressive sound. raised so that it presses outside, to produce the
single dental click is widely is
When the air is made to move outwards, the sounds used as a noise expressing against the velum. At the sound we hear as a click.
are called ejectives. An ejective consonant is a glot- negation throughout the same time, a closure is

talic egressivesound.
Implosive consonants occur in many languages,
but are particularly common in American Indian
and African languages (such as Shona and Ijo).
Ejective consonants are widely used in the lan-
guages of the Caucasian family, and also in many
American Indian and African languages (such as
Hausa and Amharic). They may even be heard in
certain accents and styles of English. Speakers from
the north of England quite often use them at the
ends of words, in place of the usual pulmonically tongue
produced [p], [t] or [k]. And regardless of the accent body
we use, if we speak in a tense, clipped manner,
these sounds will often be 'spat' out at the end
of a word.

126 IV THE MEDIUM OF LANGUAGE:. SPEAKING AND LISTENING

 OTHER TYPES OF SOUND
How ejective sounds are made The vocal tract can produce many other kinds of
The essential feature of an and the closure
glottis and the sound pops' out. sound, but they do not seem to be used with any
ejective sound is that the higher up the vocal tract. regularity in spoken language. Scraping the teeth
closed, so
glottis is tightly 4. The opens, and
glottis
together, flapping the tongue against the floor of
that no air can get to or from 2. We contract some of the lung air rushes up the vocal
the lungs. We
are, in effect muscles of the larynx, so tract, to act as a source of the mouth, or making a sucking noise with the ton-
holding our breath' for a brief as to make the glottis move power for the next speech gue against the inside of the cheek - sounds like
moment. an upwards direction - sound.
in these are perceived as idiosyncrasies of the speaker.
a movement which com- The listener does not usually interpret them as
1. At the same time as the presses the air in the The whole process, from
attempts at communication.
glottis closes, we make cavity. closure to final
initial glottal

ready to articulate a conso- glottalrelease takes on aver- On the other hand, other air-stream mechanisms
nant sound - for [p'] we 3. The increased pressure age only a twentieth of a se- are occasionally used when people communicate.
suddenly released by re- cond, though there is a great
close the lips, for [t'j or [k'] is
A velaric egressive sound (the same mechanism as
we raise the tongue. A moving the closure in the deal of timing variation
for a click, but with the sound sent outwards rather
body of air is thus trapped mouth - opening the lips, among languages.
in the cavity between the or lowering the tongue - than sucked inwards), made with the lips, is fairly
cavity
common in French, where, along with a distinctive
hand gesture and shrug of the shoulders, it means
roughly 'I couldn't care less' or 'It's not my fault.'
The similar sound, but with the tongue protruding
slightly, is a signal of contempt in many languages
— what in Britain is called a 'raspberry'.

closed
glottis
How implosive sounds are made
Implosive sounds use a pro-
cess that is to a large extent
the reverse of that used for
ejectives.
1. We
make a closure in
the vocal tract - for [6] at
the lips, for [cf] or [gf] with
the tongue. Note the spe-
cialphonetic symbols, to
distinguish these sounds
from pulmonic [b], [d] and
[g].
closed
glottis
2. The muscles of the lar-
ynx are used to partially
close the glottis, and move
it in a downwards direction,
so that air pressure in the
cavity above the glottis is
somewhat reduced. The
glottis is not closed com-
pletely (unlike ejective
sounds), so that a certain
amount of lung air is still
able to move between the
\—m\
J y^
vocal folds, causing them
to vibrate.
3. When we open the lips,
or lower the tongue,we
release the vocal tract
closure, and outside air is
sucked into the mouth. This
mixes with the lung air in
the glottis, to produce a
sound which has a muffled,
hollow resonance.
Abnormal air-stream mechanisms are also used
possible to compress
in special circumstances. It is
airwithin the cheek-space and use it to carry speech
— so-called buccal voice, most well-known through
the voice of Walt Disney's Donald Duck. It is also
possible to make sounds using air rising from the
stomach or oesophagus (the pipe leading from the
pharynx to the stomach), as in a belch. Oesopha-
geal voice is used in a sophisticated way by many
people whose diseased larynx has been surgically
removed (p. 276).
It is usual for a language to use only one or two
air-stream mechanisms for the production of vow-
els and consonants. All languages make use of pul-
monic egressive air. Glottalic egressive air (for
ejectives) is also widely used (though not in Euro-
pean languages). Glottalic ingressive air (for implo-
sives) is much rarer; and velaric air (for clicks) is
used only in a small number of African languages.
It is uncommon to find a language using more than

one or two of these mechanisms regularly. A few

languages use three. Damin, a ritual language of
a north Australian aboriginal tribe, the Lardil, is
unique in that it is reported to use no fewer than
five air-stream mechanisms. Pulmonic egressive,
glottalic egressive and velaric ingressive sounds are
used, but this language also has a pulmonic ingres-
sive [1] sound, and a velaric egressive [p] sound.

No other languages have been discovered with con-

sonants involving these latter types of sound, which
has led some scholars to speculate that perhaps
the sound system of this language was specially
invented to perform some ritual function.

22 THE ANATOMY AND PHYSIOLOGY OF SPEHCH • 127

 f

THE LARYNX of the many sets of muscles controlling laryngeal epiglottic

cartilage
Before we can speak, lung air has to be converted movement. These movements take place very
into audible vibrations, using the various organs rapidly during speech, and account for several
within the vocal tract. The most important source kinds of auditory effect.
of vibration for the production of speech sounds
is in the lower region of the tract, at the larynx. Voicing The most important effect is the produc-
The larynx is located in the upper part of the tion of audible vibration - a buzzing sound, known
trachea. It is a tube consisting of cartilages as voice or phonation. All vowels, and most of the
with connecting ligaments and membranes, within consonants (e.g. [m], [b], [z]) make use of this
which are housed the two bands of muscular tissue effect. It is in fact possible to feel the vibration
cricoid
known as the vocal folds. The location of the larynx — for example, by placing the and thumb
forefinger cartilagt

can be easily felt because its front part, the thyroid on either side of the Adam's apple, and comparing
(or 'shield-like') cartilage, forms a prominent angle the effect of saying [zzz] and [sss] loudly. Alterna-
in the neck, known as the 'Adam's apple' (it stands tively, theresounding effect of vocal fold vibration
out more sharply in men). Two other cartilages can be sensed by making these sounds while closing
work along with the thyroid to define the area of one's ears with the fingers.
the larynx — the cricoid (or 'ring-like') cartilage, Each pulse of vibration represents a single open-
and the two arytenoid (or 'ladle-shaped') carti- ing and closing movement of the vocal folds. In thyroid
cartilage
lages. The movements of all three help to control adult male voices, this action is repeated on average
the way the vocal folds vibrate. Their anatomical about 120 times (or 'cycles') a second - corres-
arrangement is shown (right). ponding to a note on the piano about an octave
arytenoid
The opening between the vocal folds, known as below middle C. In women, the average is just less
cartilage Y cricoid
cartilaj
the glottis, is men, the inner
quite a small area. In than an octave higher, about 220 cycles a second. —-—
-3
edge of the folds is usually between 17 and 24 mm; The higher the pitch of the voice, the more vib- trachea

in women it is even smaller, from about 13 to rations there will be (p. 133). A new-born baby's (b)

17 mm. The folds at the glottis are often referred cry averages 400 vibrations a second. epiglottic
to as the 'true' vocal folds because slightly above cartilage
them in the larynx is a second constriction, called Pitch We are able to alter the frequency of vocal
the 'false' vocal folds, or 'ventricular' folds. It is
fold vibration at will, within certain limits, to pro-
uncommon to hear these used in speech sound pro- duce variations in pitch and loudness. The linguis-
duction, though they are often involved in certain tic use of these connection with
features (in
types of voice quality (a notable example being the 'intonation', 'stress', and the 'tones' of tone lan-
'gravelly' voice of the jazz musician, Louis Arm- guages) are described separately in §29.
thyroid
cartilage
strong), and the effect of 'two-toned' voice is heard

in certain types of voice disorders (p. 276). Glottal stop The vocal folds may also be held cricoid
cartilage
tightly closed (when holding one's breath, for trachea
The vocal folds example). When they are opened, the released lung
The vocal folds are remarkably versatile. Their ten- air causes the production of a glottal stop [?], heard
sion, elasticity, height, width, length, and thickness most clearly in the sharp onset to a cough, but The structure of the larynx
can all be varied, owing to the complex interaction also commonly used as a sound unit in many lan- (a) frontview
(b) back view
guages and dialects (§27). In British English, for side view
(c)
example, the glottal stop is most commonly heard
An evolutionary perspective in those dialects that have been influenced by

From a biological point of view, the larynx acts as a

London speech (in such words as bottle, where it
valve, controlling the flow of air to and from the lungs, replaces the sound [t]).

and preventing food, foreign bodies, or other

substances from entering the lungs. Also, by closing Glottal friction If the vocal cords are kept wide
the vocal folds, it is possible to build up pressure within
apart, air expelled with energy will produce audible
the lungs, such as would be required for all forms of
muscular effort (e.g. lifting, defecation, coughing). glottal friction - an effect that is often used as an
In the course of evolution, the larynx has come to [h] sound in languages.
be adapted to provide the main source of sound for
speech. However, its position in front of the lower Voice qualities Other vocal fold movements can
pharynx (which leads to the stomach) presents a
complication, because food and liquids must therefore
be initiated to produce such sound effects as
pass the entrance to the trachea on their way to the whisper, breathy voice (heard in the 'bedroom'
stomach. (This complication does not exist in other voice of many female film stars and singers), and
animals, where the larynx is positioned higher up creaky voice (heard, for example, in the menacing
(§49).) To solve the problem, a leaf-shaped cartilage
low tones of the horror-film actor, Vincent Price),
known as the epiglottis is pulled across the entrance
to the larynx as part of the mechanism of swallowing,
where the vibrations are extremely slow (about 30
thus preventing these substances going in the wrong times a second). These and other similar effects
direction. involve complex patterns of vibration, and their
physiological mechanism is not entirely under-
stood.

128 •
IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING
 How is vocal fold
vibration
produced?
It is possible to see vocal fold
vibration using a laryngeal
mirror or a fibre-optic
laryngoscope. Rapid vib-
rations appear as a trembling
movement, somewhat like
that of a vibrating guitar
string. To study the effects in
detail, the vibrations need to
be filmed, and observed in
slow motion, and this is now
routinely done in instrumen-
tal phonetics (§24).

Several theories of vocal

cord vibration have been
proposed. At first, it was
thought that the vocal folds
vibrated in the manner of a
stringed instrument - the dir-
ect result of nerve impulses
moving the muscles of the
larynx (a theory that in more
recent times has been called
the 'neurochronaxiac'
theory). The contemporary
view, however, explains vo-
cal fold activity in terms of
the way the mus-
laryngeal
cles change the folds' ten-
sion and elasticity in res-
ponse to the air stream from
the lungs: the myoelastic
aerodynamic' theory of
phonation.
The movement of the vocal In this view, air pressure
folds, filmedby a high-speed causes the vocal folds to
camera (1 0,000 frames per open for each vibration; they
second) before the start of then close, partly because
phonation, as the vocal folds the natural elasticity of the
are in the process of closing folds makes them 'bounce
(above), and during phona- back', and partly because
tion, withthe folds open and the folds are sucked' to-
then closed (below). The gether, due to a sudden drop
posterior part of the folds is pressure in the glottis,
in air
towards the bottom of each as the air rushes through the
frame. narrowed larynx (the 'Ber-
noulli effect', named after the
Swiss mathematician Daniel
Bernoulli (1700-82)).

A laryngeal mirror (Above, left) The mirror is carefully inserted through the mouth until it
is larynx. The angle of the mirror then permits a clear view. The investigator needs
above the
to hold down the tongue, and must be careful not to let the mirror touch the walls of the
throat, which would produce a 'gagging' reflex in the subject.

A fibre-optic laryngoscope (Above, right) The flexible tube is inserted up through the nose
and nasal cavity until hangs down behind the soft palate. Some of the glass filaments in
it

the tube provide a strong light source; others bring an image back to the eye-piece (which
can be linked to a camera, p. 141). The main advantage of this approach, compared with
the laryngeal mirror, is that the subject is able to speak with the device in place.

22 THE ANATOMY AND PHYSIOLOGY OF SPEECH 129

 ARTICULATION
Once the air stream passes through the larynx, it
enters the long tubular structure known as the
vocal tract. Here it is affected by the action of
several mobile vocal organs - in particular, by the
tongue, soft palate, and lips — which work together
to make a wide range of speech sounds. The pro-
duction of different speech sounds through the use
of these organs is known as articulation.
In addition, sounds produced within the larynx
or vocal tract are influenced by the inherent proper-
of the cavities through which the air stream
ties
passes - the pharyngeal, oral, and (from time to
time) nasal cavities. These cavities give sounds their
resonance. Several kinds of resonance can be pro-
duced because the vocal tract is able to adopt many
different shapes.
In describing articulation, it is usual to dis-
tinguish between those parts of the vocal tract that
are immobile ('passive articulators') and those that
can move under the control of the speaker ('active
articulators'). Within the first category, we need
to recognize:
• the upper teeth, especially the incisors, which
are used to form a constriction for several
sounds, such as the first sound of thin [0];
• the ridge behind the upper teeth, known as
the alveolar ridge, against which many speech
sounds are made, such as [t], [s]; and
• the bony arch behind the alveolar ridge, known
as the hard palate, which is used in the articula-
tion of a few sounds, such as the first sound
ofyo«[j].
The soft palate may be raised against the naso-
pharyngeal wall to make a 'velopharyngeal
closure', so that air escapes only through the
mouth. This produces a range of oral sounds
— such as all the vowels and most of the conso-
nants of English.
The soft palate may be lowered to allow air
to escape through mouth and nose. This is the
position required to produce nasalized vowels,
as in French (e.g. bon 'good'), Portuguese, and
many other languages.
(iii) The soft palate may be lowered, but the mouth
remains closed. In this case, all the air is
released through the nose, as in such nasal con-
sonants as [m] and [n].
Lips The orbicularis oris ('muscle that encircles
the mouth') is the main muscle controlling lip
movement, though several other facial muscles are
also involved. The lips may be completely closed
(as for [p] or [m]), or held apart in varying degrees
to produce the different kinds of rounding or
spreading used on vowels (e.g. [u] vs [i], p. 153)
or the friction of certain kinds of consonant (as
in the b of Spanish saber 'know'):
Jaw The mandible bone permits a large degree
of movement. It controls the size of the gap between
the teeth and strongly influences the position of
the lips. Speakers sometimes adopt open or closed
jaw positions - as when someone speaks 'through
gritted teeth'.

All other organs are mobile, to a greater or lesser

An X-ray radiograph giving
extent. a side view of the organs of
speech during the articula-
Active articulators an [i] vowel. The
tion of
'humped' outline of the
Pharynx This is a long muscular tube leading
tongue can be clearly seen.
from the laryngeal cavity to the back part of the For a traditional anatomical
oral and nasal cavities. The areas adjacent to these diagram of this area, see
cavities provide a means of dividing the pharynx p. 383.
into sections: the laryngopharynx, oropharynx,
and nasopharynx. The pharynx can be narrowed
or widened. Certain types of consonant can be pro-
duced by making a constriction here (p. 155), and
movements of the larynx, soft palate, and tongue
may also involve pharyngeal modifications that
affect the quality of a sound. 'Pharyngealized' con-
sonants and vowels can be heard in several lan-
guages (e.g. Arabic).

Soft palate, or velum This is a broad band of mus-

cular tissue in the rear upper region of the mouth,
whose most noticeable feature is the uvula — an
appendage that hangs down at the back of the
mouth, easily visible with the aid of a mirror.
In normal breathing, the soft palate is lowered,
to permit air to pass easily through the nose -
though of course the mouth may be open as well.
In speech, there are three main positions that affect
the quality of sounds:

130 • IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

 soft palate
The tongue
Of all the mobile organs, the tongue is the most
capable of adopting more shapes and
versatile. It is
positions than any other vocal organ, and thus
styloglossus
enters into the definition of a very large number muscle (E)

of speech sounds: all vowels, and the majority of

consonants. The tongue is a three-dimensional
muscle, the whole of which can move in any of
three main directions through the action of the var-
ious 'extrinsic' muscles: upwards/forwards (e.g.
for [i]); upwards/ backwards (e.g. for [u]); and
downwards/backwards (e.g. for [a]). In addition, genioglossus
muscle (E)
several 'intrinsic' muscles determine the shape of The movements of the
the tongue, in any position. For example, some tongue
muscles raise or lower the tongue tip, or move it In relation to the position of
the main extrinsic (E) and
to the left or the right. Others move the tongue
intrinsic (I) muscles.
sideways, or form a groove along the middle (as
is needed for the articulation of [s]).

There are no obvious anatomical sections to the The nerves that control the vocal organs
tongue, so to classify sounds, arbitrary divisions
There are twelve cranial
have to be made using the position of the tongue nerves, whose role
in relation to the upper part of the mouth. The the brain with the head and
main areas are best located when the tongue is at neck. Some perform a mo-
tor' function, controlling the
rest, with its tip behind the lower teeth.
action of muscles; others
• front the part opposite the hard palate perform a 'sensory' func-
tion,sending signals to the
• back the part opposite the soft palate
brain. (The main areas of
• centre the part opposite where the hard and the brain involved in the
soft palate meet production and perception
of speech are introduced in
(Front, back, and centre are often jointly called §45.) Seven of the cranial
the dorsum of the tongue.) nerves are brought into
service as part of the pro-
• blade the tapering part opposite the teeth ridge cess of speech and hear-
• tip or apex the front extremity ing, and the relevant func-
tions of these nerves are
• rims the edges of the tongue
listed below. (It is
listing cranial nerves.)
is to link
V The trigeminal nerve
acts as a motor nerve to
the muscles of the jaw and
to one of the muscles con-
trolling the soft palate. It
also acts as a sensory
nerve from the back two-
thirds of the tongue.
VII The fac/a/nerveisa
motor nerve supplying the
muscles of the lips.
VIM The auditory or acous-
tic nerve acts as a sensory
nerve from the ear (p. 143).
IX The glossopharyngeal
nerve acts both as a motor
usual to nerve to the pharynx, and
the back of the tongue.
X The vagus nerve is a
motor nerve that supplies
the muscles of the pharynx
and larynx.
XI The accessory nerve
acts as a motor nerve to
the muscle that controls the
raising of the soft palate.
XII The hypoglossal nerve
is a motor nerve that sup-
plies the muscles of the
tongue.
In addition, the relevance
of several spinalnerves
should be noted, some of
which control the chest
muscles involved in

(Variations in tongue size are discussed in §6.) use roman numerals when as a sensory nerve from respiration (p. 124).

The structures of the oral

cavity
hard palate

teeth
(alveolar)

alveolar
ndge
ridge soft palate

uvula 1 J

\~5 "P

epiglottis p

\| N.XX blade

J ^\ \. \ front
The main parts of the tongue ^\^ \ centre
V ^^ back

11 THE ANATOMY AND PHYSIOLOGY OF SPEECH •

131
 23 The acoustics of speech

Sound energy is a pressure wave consisting of vib- be seen we

follow the progress of a series of
if

rations of molecules in an elastic medium — such particles, once they have been
set in motion by a
as a gas, a liquid, or certain types of solid. For source of vibration. In the following diagram, the
the study of speech production, we usually deal movement of each particle is plotted at successive
with the propagation of sound through the air: air moments of time (X+ 1, X
+ 2, etc.), imagining
particles are disturbed through the movements and the source of vibration to be on the left.
vibrations of the vocal organs, especially the vocal
folds (§22). But when we study speech reception
(§25), air is not the only medium involved. The Air particles

process of hearing requires the sound vibrations Source of vibration PI P2 P3 P4 P5 P6 ... (etc.)

in air to be transformed into mechanical vibrations Air particles (P) at rest O

(through the bony mechanism of the middle ear), I Vibration starts at time X O O I PI hits P2.
hydraulic changes (through the liquid within the II Time X + 1 O IIP2hitsP3;
inner ear), and electrical nerve impulses (along the PI returns towards its rest point.
auditory nerve to the brain).
When an object vibrates, it causes to-and-fro
III Time X+ 2 O Ill P3 hits P4;
P2 returns towards its rest point;
movements in the air particles that surround it. PI has passed its rest point.
These particles affect adjacent particles, and the IV Time X+ 3 •O IVP4hitsP5;
process continues as a chain reaction for as long P3 returns towards its rest point;
P2 has passed its rest point;
as the energy lasts. If there is a great deal of energy
PI returns towards its rest point.
in the original vibration, the wave of sound that
is produced may be transmitted a great distance,
before it dies away. But the air particles themselves
do not travel throughout this distance. The move- If we draw a line between the position of each Pure tones
ment of each particle is purely local, each one particle, the wave-like motion will be apparent, The simplest waveforms,
affecting the next, in much the same way as a long here presented vertically. such as those illustrated on
series of closely positioned dominoes can be this page, are sinusoidal in
knocked over, once the first domino is moved. shape, and known as sine
Air particles waves. They consist of a sin-
However, unlike dominoes, air particles move back PI P2 P4 P5
P3 P6...(etc. gle pulse of vibration that re-
towards their original position once they have At rest peats itself at a constant rate
transmitted theirmovement to their neighbours. Time X
and produces a pure tone.
The movement of sound waves in air is some- Pure tones are rarely
Time X + 1
heard in everyday life. Most
times explained by analogy with a stone dropped Time X + 2 sounds are complex, con-
into a pool of water, causing ripples, or waves; Time X + 3 sisting of several simulta-
but this ignores the essentially three-dimensional Time X + 4 neous patterns of vibration.
nature of the activity. The domino simile, likewise, Time X + 5 To produce a pure tone, you
gives only a limited impression of the movement TimeX + 6 need a special electronic
Time X+ machine, or a device such as
involved. A would be with an
better parallel 7
a tuning fork.
(etc.)
expanding balloon, which grows in all directions When a tuning fork (below)
at once. Sound waves, too, move simultaneously is struck, it vibrates with a

in all directions from their source. single tone. The prongs of

A
graph can be drawn of the pressure wave that the fork move to and fro at
is up when particles move in this way, and
built a fixed rate. When the fork is
WAVEFORMS this is known as a waveform. It is usual to draw held to the ear, a pure tone
The way air particles move can be compared to waveforms as patterns from left to right, on either can be heard.
a pendulum or a swing. At rest, a swing hangs side of a horizontal line representing the passage
down vertically. When
put in motion, a back-
it is of time. The simple movement of a single particle
wards movement is followed by a forwards move- would look like this:
ment, on either side of the rest point, as long as
there is energy available to keep the swing moving.
This to-and-fro movement is known as oscillation. X+4
Similarly, air particles oscillate around their rest
point. As a particle moves forward, it compresses
the adjacent particles and causes a tiny increase Rest

in the air pressure at that point. As it moves back,

it decompresses these particles and causes a de-

crease in pressure. The motion is wave-like, as can X+ 2 X+ 6

132 • IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

FREQUENCY COMPLEX TONES Familiar
A single to-and-fro movement of an air particle Most sources of sound produce complex sets of frequencies
is called a cycle, and the number of cycles that vibrations, and this is always the case with speech. One way of relating the phy-
occur in a second is known as the frequency of Speech involves the use of complex waveforms sical notion offrequency to
a sound. Frequency used to be measured in 'cycles because it results from the simultaneous use of our sense of pitch is to relate
familiar musical notes to fun-
per second' (cps), but this unit has been renamed many sources of vibration in the vocal tract (§22).
damental frequency. Middle
hertz (after the German physicist, Heinrich Rudolf When two or more pure tones of different frequen-
C has a frequency of 264 Hz,
Hertz (1857—94), who first broadcast and received cies combine, the result is a complex tone. and the notes above it in the
radio waves), and abbreviated as Hz. The basic There are two kinds of complex tone. In one diatonic scale of C have fre-

frequency at which a sound vibrates is known as type, the waveform repeats itself: a periodic pattern quencies as follows:

the fundamental frequency, generally abbreviated
as F and pronounced 'F nought'.
The range of frequencies that
adult can hear extremely wide — from about 20
is
to 20,000 Hz. It is not possible to hear vibrations
lower ('infrasonic') or higher than this ('ultraso-
nic'). However, the frequencies at both ends of this
range are of little significance for speech: the most
important speech frequencies lie between 100 and
4,000 Hz. The fundamental frequency of the adult
male voice, for example, is around 120 Hz; the
female voice, around 220 Hz (p. 128).
The frequency of a pure tone correlates with the
sensation of pitch — our sense that a sound is
'higher' or 'lower'. On the whole, the higher the
frequency of a sound, the higher we perceive its
pitch to be. But our perception of pitch is also
affected by the duration and intensity of the sound
stimulus. The notions of 'frequency' and 'pitch' are
of vibration. In the other, there is no such repeti-
tion: the vibrations are random, or aperiodic.
a young normal Speech makes use of both kinds. The vowel sounds,
for example, display a periodic pattern; sounds
such as [s] are aperiodic.
Harmonics
The sound produced by an object vibrating in a
periodic way involves more than the simple sine
wave (p. 132). Other amounts of energy are also
generated by the same vibration, all of which are
correlated with the basic sine wave in a simple
mathematical relationship: they are all multiples
of the fundamental frequency. Thus an F of A is the note sounded by the
oboe when an orchestra is
200 Hz will set up a 'sympathetic' set of frequencies
tuning up. By comparison,
at 400 Hz, 600 Hz, and so on. These multiples are the top note of a seven-
known as overtones, or harmonics, and numbered octave piano is 3,520 Hz,
in sequence. In physics (but not in music), F counts and the bottom note is 27.5
as the first harmonic. So, in this example, 400 Hz Hz.

not identical: frequency is an objective, physical would be the 'second harmonic', 600 Hz the 'third
fact, whereas pitch is a subjective, psychological harmonic', and so on. This kind of framework is
sensation (p. 144). especially useful in analysing vowels, certain conso-
nants, and intonation patterns (pp. 135-7).
Depending on the nature of the vibrating object
(forexample, the material it is made of, or its thick-
ness), different sets ofharmonics are established, Vocal waveforms
Wavelength and these are heard as differences in sound quality, A typical waveform of the

The rate at which sound energy travels through air is or timbre. The difference we hear between two vowel [a:] and of the conso-
nant [sj. The time segments
known as its velocity, and this is a constant - usually voices, or two musical instruments, when they pro-
(depending on temperature conditions) about 343 displayed are the same for
duce a sound of the same pitch and loudness, is each sound. The periodic
metres per second. All sounds that have the same
a contrast of timbre caused by the different har- pattern can clearly be seen
energy will get from A to B in the same time.
During the time it takes for there to be a single cycle monics. inthe case of the vowel
of vibration, a sound wave travels a certain distance.
sound, but no pattern is vi-

This is known as the wavelength of the sound. sible in the case of [sj.

Because of the constant rate of sound travel, it thus

follows that the higher the frequency of a tone, the
shorter its wavelength will be. A simple formula
expresses this relationship: A = (C/F), where Cis
sound velocity, Fis frequency, and A is the wavelength.
Thus, a tone of 500 Hz has a wavelength of
(343 m/500 m) = 69 cm; a tone of 1 ,000 Hz would be
34 cm.
The importance of wavelength can be seen in
relation to the way we receive sound. When a sound
wave approaches an object, if its wavelength is greater
than the size of the object, it will tend to 'bend' around
it; if the wavelength is smaller, it will tend to be

reflected. Thus, for example, as sound waves

approach the head, the lower frequencies, having a
longer wavelength, will be more likely to be retained,
whereas the higher frequencies will not- a factor that
may be of considerable significance when considering
how to assist people whose hearing is impaired
(p. 266).

23 THE ACOUSTICS OF SPEECH 133

AMPLITUDE AND INTENSITY that it has an intensity which is 90 dB greater than Speech and
The extent to which an air particle moves to and the reference level. everyday sounds
fro around its rest point is known as the amplitude We are able to hear a vast range of sound The relative intensity of dif-
of the vibration. The greater the amplitude, the intensities. A loud shout is a million times more ferent kinds ofspeech can
greater the intensity of the sound, and along with powerful than a whisper. It has been estimated that be seen by relating them to
other factors (such as frequency and duration) the the human ear is sensitive to about 10 million mil- the average intensities of
some everyday sounds (after
greater our sensation of loudness. In the following lion (10 13 ) units of intensity. To enable analysts
D.B. Fry, 1979).
diagram, we see three sine waves of equal frequency to cope with such large amounts, sound intensities
threshold of audibility
but of different amplitude. In each case, one com- are related to each other as ratios, using a logarith-
plete vibration lasts 10 msec (the frequency is thus mic scale. An increase of 10 dB is equivalent to 10 rustle of leaves

100 per second, or 100 Hz). But (a) has twice the a doubling of loudness. 30 dB is twice as loud as 20 ticking of watch (at ear);

amplitude of (b), and (b) twice that of (c). 20 dB, 40 dB is twice as loud as 30 dB, and so on. radio studio

In this way, 10 13 units can be 'reduced' to a scale 30 quiet garden; whispered

of 130 decibels - a scale that more accurately ref- conversation
lects the way in which we sense differences of loud- 40 residential area, no traf-

ness between sounds. fic

It is possible to work out average intensity values 50 quiet office; typewriter

for individual speech sounds. In the following table
60 conversation at 1 m; car
(after D. B. Fry, 1979, p. 127), the values for Eng- at 10m
lish sounds, expressed in decibels, have been related
70 very busy city traffic at
to the sound with the lowest intensity, [9] (as in 30 m
thin), which is given the value 0. Open vowels are
75 telephone bell at 3 m;
the most intense sounds, followed by close vowels shouting
and continuants; the weak fricatives and plosives
80 noisy tube train; loud
occur at the opposite end of the scale (§27; for radio music
10 15 20 25 30 35 40 45 50
transcription conventions, see Appendix 2). In a
Time (ms) 90 pneumatic drill at 1 m
word like thorn, accordingly, the increase in inten-
To measure the loudness of a sound, we need sity from the first sound to the second is nearly 100 car horn at 5 m; orches-
tra fortissimo
to take into account the contribution of both am- 30 dB.
plitude and frequency — factors that relate to the 110 boilermakers' shop
energy with which the sound is produced. The term o: 29 e 23 1 20 3 13 6 10 120 pneumatic hammer,
intensity is used to refer to the overall power of 28 i: 22 / 19 12 1 m; amplified rock

a sound — a useful notion for the study of speech, a: 26 u: 22 18 s 12 band

where sound waves are complex, and the loudness a 26 1 22 m 17 t 11 1 30 four-engined jet aircraft,

of a sound does not relate clearly to any one of 3: 25 w 21 tf 16 g 11 30 m

its acoustic components. a 24 r 20 n 15 k 11 At around 1 20 dB, the sensa-
tion of hearing is replaced by
u 24 j 20 03 13 v 10
that of pain.
Decibels
To measure sound intensity, we need a basic, inter-
nationally accepted reference level for sound pres-
sure in air. This reference sound pressure level (or
Sound pressure level (dB) The threshold of audibility
for normal young adults has
SPL) identifies the threshold at which a sound can a characteristic 'U' shape,
be heard (it is traditionally defined as 0.0002 dynes when it is presented in the
per square centimetre, dyne being the unit of mea- form of a graph. The vertical
surement for pressure). Departures from this refer- scale represents increments
in sound pressure level,
ence level are then measured in units known as
measured in decibels. The
decibels (dB) (named after Alexander Graham Bell horizontal scale represents
(1847—1922), the American inventor of the tele- increments in the frequency
phone). Thus, to say that a sound is 90 dB means of a sound, measured in
hertz. The diagram shows,
for example, that the ear will
pick up a sound of 250 Hz at
The sound around us around 1 5 dB, and will hear
it throughout the whole deci-
The time takes for sound waves to die away to an
it
bel range, until the loudness
inaudible levelis known as the 'reverberation time'. In
discomfort level (LDL) is
a room, the walls and furnishings cause the energy
reached. The area of great-
to be absorbed. Fibrous materials, such as curtains
est sensitivity to sound is be-
and carpets, absorb sound well, whereas hard, dense
tween 500 and 5000 Hz. The
surfaces cause sound to be reflected. Modern
audiogram, used in the study
classrooms tend to use the latter material, thus
of hearing impairment, is
producing high levels of noise ('ambient noise') that
based on this kind of repre-
can often make it difficult for children to hear what is
sentation (p. 266).
being said. Frequency (Hz)

134 • IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

SPECTRA
It is make an acoustic analysis of a com-
possible to Relating acoustics to articulation
plex wave and present its various components in By studying several remarkably similar
tern is to because the articulatory
the form of a sound spectrum. A spectral analysis speakers, it is possible to that displayed when vowels chartis based only on the
is a graph in which the horizontal axis represents work out the mean frequen- are described according to point of greatest tongue con-
frequency and the vertical axis represents ampli- cies of the first and second their place of articulation in striction, whereas the acous-
formants of the vowels in a the vocal tract (the vowel chart derives from the re-
tude. For example, the many acoustic components
tic
language. This has been 'quadrilateral', p. 154). How- sonances of the whole vocal
of the vowel [iz], representing the way the vocal done for British English, ever, the match is not exact tract.
tract resonates during its articulation,can be with the following results
shown as a spectrogram in the following manner: (after J. C.Wells, 1962):
Frequency of F2 (Hz)

heed 300 Hz 2300 Hz

hid 360 Hz 2100 Hz
Fi
head 570 Hz 1970 Hz
[':] had 750 Hz 1750 Hz
hard 680 Hz 1100 Hz
Fi
hod 600 Hz 900 Hz
= hoard 450 Hz 740 Hz
5
n d:

u hood 380 Hz 950 Hz

u: who 300 Hz 940 Hz
3 hub 720 Hz 1240 Hz
3: herb 580 Hz 1380 Hz
600 200 1 800 2400 3000
Frequency (Hz)
We may now transfer such fi-
gures onto a graph in which
the frequency of F, is dis-
played vertically and F hori-
2
zontally. The resulting pat-
The vowel has a quite distinct spectral
[a:]

character, reflecting the very different configu-

ration of the vocal tract.

Fi
Fi
[»:]

F3
e
<

600 1 200 1 800 2400 3000

Frequency (Hz)

Note that, in these spectra, the amplitude of

some frequencies is much greater than others. In
fact, possible to see various 'peaks' of acoustic
it is

energy in each case, reflecting the main points of

resonance in the vocal tract. These peaks are
known as formants, and they are numbered from
lowest to highest: the 'first formant' (Fj), the
'second formant' (F2 ), and so on. In the spectrum
of [i] above (which was spoken by a man at a funda-
mental frequency of 120 Hz), Fj peaks at 360 Hz,
F2 at 2,280 Hz, and F 3 at 3,000 Hz.
Formant structure is a major feature of speech
sounds. All vowels and some consonants have for-
mants. It is the formant pattern (especially the dis-
position of the first two formants) that enables us
to differentiate vowels, or to recognize repetitions
of a vowel as being the 'same', even when produced A real-time spectral analyser This instrument uses a cathode ray tube to display the
continuously changing spectra of complex waveforms. Because the changes take place so
by different speakers. And vowel formants can also
quickly, further analysis requires that the investigator 'stop the action' - for example, by holding
help in identifying the character of adjacent conso- the waveform in a storage oscilloscope, photographing the screen, or providing a computer
nant sounds. display.

23 THE ACOUSTICS OF SPEECH- 135

 .

THE SOUND SPECTROGRAPH made by the stylus. Frequencies of little or no inten-

During the 1940s, the sound spectrograph was sity thus appear as areas of clear paper.
designed to analyse and display speech spectra.
Although the limitations of the instrument do
This machine records speech, analyses the sound
not allow perfect accuracy of measurements from
waves into their different frequencies using an array
the paper strip, the visual display allows an imme-
of electronic filters, measures the intensity of each
diacy of recognition, to anyone trained in acoustic
frequency, and then presents the result as a visual
analysis, which has made it a popular research tool.
display, using a stylus to make marks on a strip
Contrasts between individual sounds can be clearly
of specially coated paper. Spectrograms illustrating
shown, as can the way sounds influence each other
a variety of sounds are shown on this and the oppo-
as they combine in connected speech. These days,
site page.
too, there have been developments in spectro-
Three dimensions of sound are represented on
graphic design that permit more accurate and
this kind of spectrogram.
extensive analyses, using computational techni-
1. Time is displayed horizontally: in the most ques. In particular, digital spectrographs are now
widely used spectrograph (the Kay Sona-Graph), available that can store two spectrograms on a
it is possible to record 2.4 seconds of speech on screen, thus enabling a more systematic and precise
the paper strip, which is 'read' from left to right. comparison of points of detail.
Each half-inch of paper thus displays a tenth of
a second of speech.
2. The vertical dimension displays information
about frequency — from Hz (the bottom line) to
Types of spectrogram
8,000 Hz. The scale is linear: each vertical inch of
It is possible to choose one much broader bands of fre- ponent analysed and
is
of two settings when mak- quency (usually 300 Hz), printed out asan amplitude
paper represents 2,000 Hz. ing a spectrogram. If the and this will make the for- display along the top of the
3. The third dimension is intensity, represented machine is set to narrow', mants stand out clearly. spectrogram.
by the degree of darkness of the marks on the paper. it will analyse the range of For most purposes, wide- Narrow-band and wide-
The more intense the signal, the blacker the mark speech frequencies into band analysis is more use- band versions of the sen-
small bands (usually ful in speech science. tence This is a spectro-
45 Hz), and this will make It also possible to ob-
is gram' are shown below, the
individual harmonics show tain information about the latter accompanied by an
up very clearly. If the ma- amplitude of a sound using amplitude display.
switched on, the drum re- chine is set to 'wide', an a spectrograph. The inten-
How a spectro- volves. The stylus then analysis is made using sity of each frequency com-
gram is made moves up the drum, ana-
lysing the different frequen-
1 A stretch of speech is cies in the speech. If there
recorded on the spectro- is energy present at a given

graph, either through a frequency, the stylus marks

microphone or from a tape the paper.
recorder. 4. When the stylus has
2. A specially coated reached the top of the
paper is placed around the drum, the machine is
drum at the top of the ma- switched off, and the paper
chine, and a stylus is is removed. The spectro-

placed against the paper. gram is now available to be

3. When the machine is interpreted.
2

This is a spectrogram'

i
^^/ * w

dis iz ei S pe krr 3 u gram

136 -IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING
The acoustic features of vowels and consonants The vowels

Vowels All vowel sounds in identifiedby a short period of voicing may actually
begin ."H.llil!;!.

normal speech display two,
and usually three, formants.
These appear on a wide-
band spectrogram as thick
dark bars. They can be
clearly seen in the spectro-
grams of the long vowels [h],
[a:], and [u:], said in isola-
tion. The vertical striations
represent vocal fold vib-
ration.
Semi-vowels The sounds
and [w] (as in you and we)
function as consonants
many languages, but they
have the acoustic features
vowels - [i] and [u]
respectively - and are thus
often called semi-vowels'
(p. 1 52). Their vowel-like
character can be clearly
seen on a spectrogram
where they are articulated
between [a] vowels. In both
cases, the formants are bent
as the vowel changes its
quality. For example, during
silence, while a closure is before the noise burst.
made in the mouth, followed The different places of
by a short burst of noise, articulation of plosive
when the closure is released consonants can also be seen
(p. 157). Both of these from spectrograms. The
features can be clearly seen bilabial sounds [p, b] have a
on a spectrogram, especially burst of noise at low
when these consonants are frequencies; the alveolar
articulated between vowels. sounds have the burst
[t, d]
The silence is shown by the at high frequencies; and the
vertical strip of clear paper; velar sounds [k, g] have the
and the release by a thin burst within the middle
[j] spike' of marks spread quite range. There is also a clear
widely across the spectrum. difference at the point of
in The onset of the following transition between
vowel is shown by the consonant and vowel: a
of appearance of the black rapid articulatory movement
formant bands. is involved, and this is
The differences between reflected in the sharp bend in
voiceless [p, t, k] and voiced the formants at the onset of
[b, d, g] plosives (p. 128) can the vowel - a bend that
be clearly seen. varies in direction between
• In the voiced sounds, the each pair of consonants.
intensity of the noise burst is
much less than in the Fricative consonants
voiceless sounds. These consonants make use
• The duration of the silence of random acoustic energy,
isshorter for voiced sounds. or noise, which is

[aja], the first formant bends • An important variable is represented on the

downwards, and the second the time between the release spectrogram as a broad area
formant bends upwards as of the plosive and the onset of disturbance at certain
the tongue moves from [a] to of vocal cord vibration (which frequencies. This is most
[i]; they then bend back as is represented by the clearly seen in the case of
the tongue resumes its formants). This known as
is such as [s]
sibilant fricatives,
original position. The bend 'voice onset time', or VOT. and which are high-
[J],

also affects the third formant; There a noticeable VOT

is energy sounds. The energy
but a fourth formant, higher gap the case of each
in for [s] is largely above
up, is much less affected. voiceless plosive (marked as 4000 Hz; that for [/] begins
Xon the spectrograms); but lower, at around 2500 Hz.
Plosive consonants A the voiced plosives have little The contrast is presented
plosive consonant is typically or no gap. In such cases, the here between [a] vowels.

Connected speech
The spectrograms at right between them become more a process of
articulation is
are of carefully articulated difficult to distinguish. continuous change. It is a
sound units, said in isolation. Spectrograms of connected fact we should always bear
In connected speech, the speech (see facing page) in mind as we read the neat,

sounds are produced more emphasize the essential separate symbols of a

rapidly, they influence each continuity of spoken phonetic transcription.
other, and the boundaries language - the fact that

avak u: stikfan
"The study of acoustic phonetics"
a s a i

23 THE ACOUSTICS OF SPEECH 137

24 The instrumental analysis of speech

A vast array of instruments is today available for DISPLAYING SPEECH SIGNALS

the study of speech production. A well-equipped The most commonly used instrument for observing
phonetics laboratory includes equipment for sound waves is the oscilloscope, which displays the
recording speech, analysing the acoustic properties frequency and amplitude of a waveform. Some
of the sound signal (§23), and investigating the scopes provide only a temporary image; others
physiology of the vocal organs (§22). Related tech- (storage oscilloscopes) are able to hold a waveform
niques are available for the study of hearing and on the screen for more detailed study. In these
speech reception (§25). The field as a whole is cases, the images can be photographed directly, or
known as instrumental (or experimental) pho- some kind of chart recorder can be used to obtain
netics. a visual trace on paper.
There are many kinds of chart recorder, reflect-
RECORDING SPEECH ing various technologies. All involve the use of a
Any scientific investigation into the nature of mechanically unrolling sheet of paper, on which
speech requires the keeping of permanent records. a trace can be made with one or more ink jets or
The speech signal itself can be recorded on disk pens (depending on how many channels are being
or (more usually) audiotape, and displayed in recorded simultaneously), or using a system based
visual form on a paper chart, computer printout, on heat recording, fibre-optics, or ultraviolet light.
or screen (which can then be photographed). Simi- Various forms of computer processing of speech
larly, photographic and other techniques are avail- signals are now routinely available, in which infor-
able for recording and displaying the speaker's mation can be rapidly displayed in graphic or
physiological activity. numerical form.
The careful choice of methods and instruments The sound spectrograph, the most valuable of
is of the highest priority in acoustic research. Unless all instruments for the visual display of speech, is
special precautions are taken, recordings may not illustrated on p. 136.
be sufficiently clear to enable accurate acoustic
analysis to be carried out. If a recording contains
a lot of background noise, or if the signal is weak
or distorted, the speech sound waves will be ob-
scure. Particular attention must therefore be paid
to the limitations of the recording instrument
(usually a tape recorder), the microphone, the play-
back system, and the location in which the record- A Visicorder (below).
ing is made. In thistechnique, a light
For best results, recordings should be made in
beam responds to the electri-
cal characteristics of the
a special studio,which has been shielded from
speech signal and makes
external sounds and which has sound-absorbent traces on light-sensitive
walls. If this is not available, recordings should be paper.
made in a quiet room containing sound-absorbent
material (such as soft furnishings). In this way, it

should be possible to minimize problems of echo

and other interference.

An anechoic chamber (above).

The ceiling, walls, and floor have been covered
with a material designed to absorb sound and
cut out reverberation.
\

A storage oscilloscope (left)

138 • IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

Physiological investigations
Speech physiology studies all aspects of the disposi-
tion and use of the vocal organs during the act
of speaking. Precise information can be obtained
about the functioning of the larynx, soft palate,
tongue, and lips, and the role of other organs and
cavities in the production of individual sounds and
in connected speech. Particular attention is paid
to patterns of air pressure, volume, and flow, and
to the activity of the underlying muscles and nerves
— including the way in which speech movements
are controlled by the brain (§45). The work
involves both the analysis of natural speech sam-
ples and the experimental investigation of sound
production under carefully controlled conditions.
AIR FLOW
How isbreathing modified in order to facilitate
speech? What happens to the air stream as it passes
through the vocal tract? The field of aerometry
investigates these questions,and some of its most
important instruments are illustrated on p. 125.
Several instruments provide air-flow data, such
as the electroaerometer and the pneumotacho-
graph. In pneumotachography, for example, a
specially designed face-mask provides separate
measures of air flow from the mouth and nose.
For example, the following traces were obtained
from a speaker uttering the word smile. Trace (a)
shows the amount of nasal air flow, measured in
litresper minute (1/m) - absent for [s], strong for
[m], then gradually reducing during the rest of the
word. Trace (b) shows the corresponding amount
of oral air flow - strong at the beginning and end,
but completely absent for [m], where the lips are
closed. What is interesting, in this speaker, is the
continued nasality throughout the vowel, which
shows the influence of the preceding [m].

Muscle movement A hooked-wire electrode, as fine as a human hair, is inserted into the neck of a subject, to monitor
the activity of a laryngeal muscle. This is one of the techniques available for use in electromyography (EMG).

Muscles produce tiny amounts of electrical activity when they contract. This can be monitored by placing an electrode
on the skin, or (to produce a clearer and more specific signal) directly into a muscle. If the muscle is involved in
the production of a speech sound, it sends signals to a monitoring device; the signals are amplified and then displayed
on an oscilloscope or on paper.

24 THE INSTRUMENTAL ANALYSIS OF SPEECH •

139
THE TONGUE AND PALATE
It is easy enough to see the tongue in a mirror,
but it is one of the most difficult organs to monitor
during speech. Many techniques have been tried,
with varying levels of success.
Direct photography This has been possible in a
few cases where had an operation
a subject has
involving the partial removal of the cheek. The
movements of the tongue can be clearly seen, but
the data are of limited value, because a subject's
speech is never normal in such circumstances.
Intraoral devices A mechanical device can be
placed inside the mouth to monitor aspects of ton-
gue movement, such as a miniature camera, a pneu-
matic bulb, a wire electrode, or an instrument for
recording pressure. The main problem with such
procedures is that the presence of a foreign body
inside the mouth is likely to interfere with normal
articulation. Microelectronic techniques are likely
to make this problem less serious in future.
Plaster casts A
three-dimensional impression of
the front part of the tongue during the articulation
of a sound can be obtained in the form of a plaster
cast. However, the method has many limitations.
Only single sounds can be studied, and even in these
cases, thetongue position is likely to be abnormal,
because of the weight of the material pressing down
on it.
Plastographic techniques A material is placed in
the mouth that deforms during tongue contact. An
early user of this technique was Erasmus Darwin
(1731—1802), who used tinfoil to study vowel
sounds in this way. An obvious limitation is that
only one extreme movement of the tongue can be
recorded at any time.
X-rays Several X-ray studies have been made
using static, cine, and video photography, and there
are many specialized techniques. For example, in
cinefluorography, X-ray images are photographed
after being projected onto a fluorescent screen.
However, the radiation hazard severely limits the
amount of data that can be obtained, and it is not
always easy to identify soft-tissue areas clearly,
even when the surface of the tongue is outlined
with a radio-opaque material. This difficulty can
be overcome to some extent through the use of
tomographic (or laminagraphic) techniques, which
enable an X-ray to be taken of a predetermined
layer of body structure. Recently, too, computer-
controlled X-ray microbeams have been devised,
which track the movements of the tongue through
the use of small lead pellets fixed to its surface.
This technique involves very low doses of radia-
tion. But even the best pictures cannot provide a
sense of the three dimensions involved in tongue
movement (p. 131).
140 •
IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING
Palatography This is the main technique for
obtaining information about the exact location of
tongue contact with the palate. In 'direct' palato-
graphy, the palate is painted with a special mixture;
if the tongue makes contact with the palate during
the articulation of a sound, some of the mixture
will be wiped off, and the pattern can then be
photographed. palatography uses an arti-
'Indirect'
ficial palate, which over the subject's palate,
fits
and is painted with material that adheres to the
tongue after a contact is made. The palate can then
be removed from the mouth for a detailed examina-
tion. A great deal of data about individual articula-
tions can be obtained using these approaches, but
the technique can provide no information about
the movement of the articulators in connected
speech.
Electropalatography By contrast, this technique
gives detailed information in real time about the
location and sequence of tongue-palate contacts.
An artificial palate is used, which incorporates an
array of metal contacts. When the tongue touches
a contact, a signal is sent to a recording device.
Different patterns of contact can then be displayed
on a screen or printed out on paper. Computer
processing enables large quantities of data to be
quickly analysed (see illustration [facing page]).
Ultrasonics In recent years, have
researchers
begun to use ultrasound techniques, in which a
beam of very high frequency sound waves is used
to monitor the position and movement of internal
structures. Although it is still in the experimental
stage, the safety of this approach makes it one of
the most promising avenues for future speech
research.
Monitoring an ultrasonic
echo of the vocal tract
(p. 277).
THE LARYNX
The larynx can be observed through a laryngeal
mirror or a fibre-optic laryngoscope (both illus-
trated on p. 129). High-speed films can then be
made of the vocal fold vibrations and replayed at
normal speed to give a slow-motion effect, or ana-
lysed frame by frame. A similar effect can be
obtained by using a stroboscope to illuminate the
folds: the frequency of its flashing light is adjusted
until it corresponds to that of the fold vibrations,
at which point the folds appear to vibrate slowly.
A glottograph can also be used - a device that
monitors the amount of light passing through the
glottis, and thus indicates the extent of glottal
opening during speech.
The main disadvantage of laryngoscopy techni-
ques arises from their invasive character: they are
introduced into the vocal tract and thus to a greater
or lesser extent interfere with the comfort of the
speaker and the naturalness of speech. A technique
of larynx observation that avoids this problem is
provided by electrolaryngography. Two small elec-
trodes are applied to the skin of the neck on either
side of the thyroid cartilage (p. 128). During
Electropalatography The
speech, information about the way the vocal folds shows a sub-
top illustration
come together is displayed on a screen as a wave- ject making an electropalato- OCT.
KC.
443
Jot OCT. 447
form {Lx). Here is the Lx waveform of an adult graph recording. This system DCC. 311
00. . . .00
uses 64 electrodes arranged OOO. .Of) 00. . .00
female speaker producing a steady [a] vowel; the 00. . .00 00.. .00
to cover the surface of the 00. . .00 00.. -on
full speech waveform is also given for comparison. artificial palate (above). Pat-
00.. .00 00.. 00
00.. .00
terns of contact are then dis-
played as a changing pattern
of lights on a display screen.
Speech a computer
Alternatively,
printout can be made, in
Vocal folds closed which tongue contacts are

AAAAAAAAAJl
Vocal folds open
Lx
represented by zeros (see
illustration above right). The
velar area of the palate is at
the top of each small dia-
gram, and the alveolar area
When the electrolaryngograph is incorporated
the bottom. The printout
is at
into a device known Voiscope®, it is possible
as a isread from left to right, with
to display fundamental frequency (Fx) on an oscil- samples occurring at
loscope screen in a manner that corresponds to our 1 msec intervals. The illus-
tration shows the articulation
perception of pitch. High fundamentals appear of the word pigsty, from the
towards the top of the screen; low ones towards release of the [p] to the [s].

the bottom. The Fx contour of an utterance is given The sequence of tongue-

below, showing two falling-rising pitch patterns on palate contacts made during
the word can be clearly seen.
the first four words, and a falling pitch at the end.
Note, in particular, the clear
The breaks in the trace are due to the occurrence velar closure made for [g].
of the sounds [st] and [h], where there is no vocal W.J. Hardcastle &
(After
fold vibration. A.Morgan, 1982, p. 51.)
A subject making an
electrolaryngograph
recording
R. The screen shows
two intonation pat-
terns (§29), one
above the other, for
the sentence 'It's
easy, isn't it?' The
contrast between ris-
ing and falling tones
can be clearly seen.
Right: Lying in front of
the equipment are the
guard-ring electrodes
used, which are held
in place by an elastic
band.

<- t n qvr th 3 hew

24 THE INSTRUMENTAL ANALYSIS OF SPEECH •
141
 25 Speech reception

The ear 3,000 and 4,000 Hz, p. 135), thus making weak Do we need two
sounds at these frequencies more perceptible. It ears?
The first step in the reception of speech takes place eardrum to some extent
also helps to protect the Two ears - bilateral hearing
when sound waves arrive at the ear. From there, from changes in temperature and humidity as well - are a great asset. They en-
sound is transmitted along the auditory nerve to as from physical damage (though no canal has yet able us to be more precise in
our judgment of the position
the brain (§45). The process is a complex one, proved capable of withstanding the ingenious
of a sound source - an im-
involving several distinct stages which reflect the attempts of young children to insert all kinds of portant factor in listening to
main anatomical division of the ear into outer ear, implements inside their ears!). people a group, or heeding
in

middle ear, and inner ear. the direction of a vocal warn-

THE MIDDLE EAR ing. This happens because a

sound source is usually
THE OUTER EAR The eardrum, or tympanic membrane, separates
nearer one ear than the
The outer ear consists of two parts. The visible the outer ear from the middle ear. It isroughly other; as a result, the signals
part is known pinna — a structure
as the auricle, or circular in shape, lying at an angle of about 55° to each ear will be slightly out

consisting of several rounded prominences formed across the whole of the external auditory canal. of phase, and one will be
more intense. The brain re-
mainly from cartilage. The pinna has a minor role It consists of a fibrous tissue with important elastic
solves these differences and
to play in the reception of sound: it helps to focus properties that enable it to vibrate when sound makes a judgment about
sound waves into the ear, and assists our ability waves reach it. The shape and tension of the ear- localization. Sometimes
to detect the source of a sound. It also protects drum cause the vibrations to be focussed at a pro- there ambiguity (when a
is

the entrance to the auditory canal, both from physi- minence near its centre, from where they are sound reflected by a
is

nearby object, for example),

cal attacks and from excessive amounts of sound. transferred to the first of the bones of the middle
in which case we have to
By pressing the central part of the pinna with the ear, which is firmly attached to the membrane. search' for the sound source
finger, it is possible to cover the entrance to the The chamber of the middle ear, known as the by moving the head.
canal, thus considerably reducing the amount of tympanic cavity, lies within the bones of the skull, The value of two ears is
sound entering the ear. about 15 mm high. It is filled with air, because there most evident in cases of
hearing loss in one ear (uni-
From here, the external auditory canal leads to is a direct connection to the nose and throat via
laterahoss, p. 266). The
the eardrum. The canal is about 2.5 cm long and the Eustachian tube (named after the Italian ana- good' ear copes well with a
contains hairs and glands that secrete wax {ceru- tomist, G. E. Eustachio (1520—74)). This tube is single speaker in a quiet
men), a substance that acts as a filter for dust, normally closed, but such activities as yawning or room; but in contexts where
sound is coming in from sev-
insects, and other tiny substances that might swallowing open it. In this way, the air pressure
eral directions (such as in a
approach the eardrum. The canal acts as a small level on either side of the eardrum is maintained. meeting), the listener finds
amplifier for certain sound frequencies (between The primary function of the middle ear is to turn localizing the source of
sound very difficult, and may
look for the speaker in the
wrong direction.
Weshould also note that
the brain uses our two ears
ways. One ear
in different

may have an advantage over

the other for certain types of
sound. This can be shown in
tests of dichotic listening,
where different signals are
presented simultaneously to
each ear, and listener res-
ponses show that one ear
transmits a sound to the
brain more readily than the
other (p. 259).

142 • IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

 '

the sound vibrations at the eardrum into mechani- thin at the base of the cochlea (about 0.04 mm)
cal movement — which be transmitted
will in turn
to the fluid-filled inner ear. It does this using a sys-
and
0.5
gets thicker as
mm). It is
approaches the apex (about
it

thus able to respond differentially

fg%*
tem of three tiny bones, known as the auditory incoming vibratory pressures: high frequencies 1 ! 1
'

to 1 1 1

1
1 1 1! 1 ' I

'
1

ossicles.These bones are the smallest in the body primarily affect the narrow end; certain low fre-
2 3
and are the only bones to be fully formed at birth. quencies the thicker end; and other low frequencies
They are suspended from the walls of the tympanic activate the entire membrane (p. 144).
cavity by ligaments, and are delicately hinged Resting on this membrane is the highly sensitive
together so that vibrations can pass smoothly organ of hearing, called the organ of Corti (dis-
The tiny size of the ossicles
between them into the inner ear. The three bones covered by the Italian anatomist, Alfonso Corti can be seen in this drawing.
have been named according to their shape: the mal- (1822-76)), which translates the mechanical move-
leus ('hammer'), which is attached to the eardrum, ments of the membrane into nerve impulses. It con-
the incus ('anvil'), and the stapes ('stirrup'). The tains a systematic arrangement of cells covered
stapes fits into the oval window — an opening in with very fine hairs, distributed in rows and layers
the bony wall separating the middle ear from the along the membrane. These hair cells act as sensory
inner ear. receptors, picking up the pressure movements in
This may seem an unnecessarily complicated sys- the endolymph. Electrochemical changes take
tem of getting vibrations from point A to point place, which activate the fibres of the auditory
B, but it is known to have several advantages. In nerve (the vm
cranial nerve, p. 131). The signals
particular, the process acts as a kind of leverage are then sent the short distance along this nerve
system, enabling the vibrations to be greatly ampli- to the temporal lobe, via the brain-stem and mid-
fied (by a factor of over 30 dB) by the time they brain (p. 258).
reach the inner ear. As the inner ear is filled with
fluid, vibrations would very readily get lost without
.oval window
cochlear partition
this amplification. Also, the bony network of the
(containing basilar membrane)
middle ear helps to protect the inner ear from sud-
den, very loud sounds. The muscles that control
the movement of the eardrum and the stapes func-
tion in such a way that they lessen the chances
of massive vibrations damaging the inner ear (the
'acoustic reflex'). However, the time it takes for
apex
these muscles to reactnot so rapid that the inner
is
perilymph
ear can be protected from all such sounds; and
cases of damage to the eardrum or inner ear do A schematic view of an uncoiled cochlea The movement of the stapes at the oval window
is transmitted to the cochlear duct via the fluid (perilymph) in the two chambers.
occur.

THE INNER EAR scala vestibuli perilymph Reissner's membrane

This is system of small interconnecting cavities
a
and passageways within the skull. It contains the cochlear
semi-circular canals, which control our sense of duct
balance, and the cochlea, a coiled cavity about
35 mmlong, resembling a snail's shell. The main
function of the cochlea is to turn the mechanical
vibrations produced by the middle ear into electri-
cal nerve impulses capable of being transmitted to
the brain.
endolymph
The cochlea is divided along most of its length
into an upper chamber (the scala vestibuli) and a
lower chamber (the scala tympani), separated by organ of Corti

the cochlear duct. Both chambers are filled with basilar

membrane
a clear, viscous fluid known as perilymph. Vib-
rations enter this fluid via the oval window and hair cells
the scala vestibuli, and are transmitted all the way
around the cochlea. They pass from upper to lower
chamber through an opening in the cochlear duct
at its apex, and finish at a sealed opening in the
wall of the middle ear, called the round window.
This structure can be clearly seen in a diagram of
an uncoiled cochlea (above right).
The cochlear duct is separated from the scala
perilymph scala tympani
tympani by the basilar membrane, and is filled with
fluid known as endolymph. This membrane is very A cross-section through one cochlear duct, showing the location of the organ of Corti.

25 SPEECH RECEPTION 143

THEORIES OF HEARING impulses.More complex theories of pattern-recog-
nitionseem to be required, involving both place
Pitch perception and temporal elements, and perhaps involving cen-
How does the frequency information in a sound tral as well as peripheral neural processes. A great
wave as it enters the cochlea (p. 143) come to be deal of research therefore continues to be devoted
transformed into a pattern of nerve impulses that to this topic.
will enable the sound to be perceived? Several
theories have been proposed since the matter was Loudness perception
first systematically investigated in the mid- 19th It is unclear how the cochlea detects the loudness Width at apex 0.50mm
Average width 0.2 mm basal turn
of a sound. One factor is the rate at which indivi-
1

century.
Resonance or place theory This is the classical
theory of pitch perception, deriving from the work
of the German scientist, Hermann von Helmholtz
(1821—94). In this approach, individual fibres in
the cochlea were thought to resonate to a particular
frequency. As the frequency changed, so the place
of vibration along the basilar membrane would
change. However, it has since been demonstrated
that the basilar membrane vibrates, not at single
points, but along most of its length.
Temporal or frequency theory In this view, first
proposed by William Rutherford in 1886, the fre-
quency of a wave was thought to be transmitted
by the number of pulses per second in a nerve fibre.
Every hair cell was thought to respond to every
tone, the cochlea acting as a kind of telephone
on frequency informa-
transmitter, directly passing
tion to the auditory nerve. This view had to be
modified when it was discovered that no nerve fibre
is capable of firing at more than 1,000 Hz, and
dual nerve fibres fire: the louder a sound stimulus,
the faster the rate of firing. But this cannot be the
whole explanation, as individual fibres seem able
to cope with changes of only around 40 dB. Other
The basilar membrane A
mechanisms have to be involved. One proposal
schematic diagram showing
arises out of the finding that some of the hair cells the points at which tones of
are tuned to respond to certain frequencies. A very different frequencies cause
loud sound at a given frequency would quickly 'use maximum amplitude of vib-
ration (from O.Stuhlman,
up' the coding potential of that fibre; and other
1952).
fibres might then be brought in as 'reinforcement',
thereby transmitting a proportionately greater sti-
mulus to the brain. The study of such matters pro-
vides another important focus of contemporary
research into auditory perception.

that most firings take place at much lower rates.

As humans can respond to speech frequencies rang-
ing up to 20,000 Hz, a purely temporal theory is
inadequate.

Volley theory This view, proposed by E. G.

Wever in 1949, represents a compromise between
place and temporal theories. It proposes that below
5,000 Hz temporal patterning is important, with The case of the missing fundamental The place
pitch perception being dependent on the synchro- theory of pitch perception encounters a problem with
nized action of several nerve fibres, firing in volleys. the case of the missing fundamental. If a complex tone
is generated, consisting (for example) of three pure
Above 5,000 Hz, place analysis is well preserved.
tones at 300, 600, and 900 Hz, a listener will perceive
the tone as having a pitch which corresponds to the
Travelling wave theory This account was pro- fundamental frequency of these harmonics - 1 50 Hz
- despite the fact that no energy has been generated
posed by the Hungarian physiologist Georg von
at that frequency.
Bekesy (1899-1972). By illuminating the vibrating
basilar membrane with a stroboscopic light, he was
able to show that sound vibrations move through
the cochlea in the form of a wave. The point on
the basilar membrane at which the wave amplitude
is greatest corresponds to the frequency of the sig-

nal.

After a century of debate and experimentation, it

is still not possible to be sure about the relative

roles of spatial and temporal factors in the hearing

process. The acoustic and physiological aspects of
the speech signal do not seem to interact in a simple
or direct way; the patterns of fluid movement in Hermann von Helmholtz Georg von Bekesy (1899-
the cochlea do not relate clearly to patterns of nerve (1821-1894) 1972)

144 IV THE MEDIUM OE LANGUAGE: SPEAKING AND LISTENING


Speech perception
Just as the vocal organs have evolved to facilitate
the production of speech, so the auditory system
seems 'tuned' to receive speech patterns. When we
hear sounds, we hear them as either speech or non-
speech; there seems to be no middle ground. No
matter how hard we try, we cannot hear speech
as a series of acoustic hisses and buzzes, but only
as a sequence of speech sounds. This is the kind
of observation that has motivated the field of
speech perception — the study of the way speech
sounds are analysed and identified by ears and
brain.
Even after several years of research, the process
of speech perception is little understood. Several
questions illustrate the problems faced by
researchers in this area.
• We hear a sequence of sounds, words, and
other units in speech; but when we look for these
units in the speech waveform, it is difficult to see
them. The linguistic units are not neatly demar-
cated by pauses or other boundary markers: the
speech signal varies continuously (p. 137). How,
then, is the brain able to analyse this signal so that
the language units can be identified?
• When several people are talking at once in a
crowded room, we are able to 'tune in' to one
speaker and to ignore the others. How does the
brain select auditory information so impressively?
• When we hear different instances of a sound,
sequences, and cope with these modifications. For
example, in the word handbag, the nd is pro-
nounced as [m], because of the influence of the
following [b] (p. 164); but the word is still inter-
preted as hand and not ham. How does the brain
carry out such partial identifications?
Indirect procedures
A further difficulty is that the link between speech
and listener perception cannot be studied in a direct
manner. The movements within the ear and audi-
tory nerve cannot easily be observed, nor can the
associated activity taking place in the brain. Speech
perception studies therefore have to rely on a range
of indirect methods.
• The acoustic properties of the speech signal are
analysed and related to the way in which people
judge sounds to be the same or different.
• Experiments are devised in which sounds are
obscured or distorted ways, to see how
in various
far identification continues tobe possible.
• Speech sounds are created artificially using spe-
cial instruments (p. 146), and systematically varied
to see what effect this has on our perception - a
procedure that also helps to determine whether our
assumptions about speech analysis have been cor-
rect.
In all cases, problems of experimental design and
data interpretation have caused progress to be
painstakingly slow. But a substantial core of theor-
etical and empirical research is now available.

we have no difficulty recognizing them as 'the

same'; but when we examine the waveforms, we
find that they are not physically identical. A [b]
before an vowel does not have exactly the same
[i]

waveform as [b] before [a], or [b] at the end of

a word. Moreover, the articulation of [b] by differ- 'I heard you'
ent people will result in different waveforms Auditory perception studies deal with the way in which
an organism detects, discriminates, interprets, and
because their regional accents and individual voice
reacts to a sound stimulus. Several factors are
qualities will not be the same (p. 20). It will vary, involved, and thus to say that we have 'heard' a sound
further, when people adopt different tones of voice can mean several different things.
(such as a whisper), or when it is said in a noisy • The body may react to the presence of a sound
situation. How does the brain recognize sounds stimulus, but we are not consciously aware of it. Such
involuntary reflexes (of respiration and heartbeat, for
when there is so much variation?
example) provide an important source of evidence for
• Many pairs of words differ by only one sound the reception of sound by infants.
— cap and cab, for instance; but when we examine • A sound is consciously detected, or 'heard'. For this
the waveforms of such words, we find that the dif- to happen, there has to be a certain minimum of
stimulation - the absolute threshold of the stimulus.
ferences between them are often simultaneously
• Sounds may be recognized as being the same
located in several parts of the speech signal. The (recognition) or different (discrimination). An important
contrast between cap and cab is partly due to question is how different two sounds have to be in
acoustic differences found at the ends of these order for the brain to perceive that they are different.
words, but it is also due to differences between This minimum difference in magnitude (the 'just
noticeable difference') is known as the difference
the vowels - the [a] of cab being much longer than
threshold. Our ability to detect and discriminate sound
that of cap. Yet in listening to speech we ignore isknown as auditory acuity.
the vowel difference and 'hear' only the consonant • The brain is able to focus on certain aspects of a
difference. How does the brain bring this informa- complex auditory stimulus and to ignore others: the

tion together? phenomenon of auditory attention. When we begin to

'hear attentively', we are said to be listening. The
• normal speech, people produce sounds very
In
concepts of 'hearing' and listening' are therefore not
quickly (12 or more segments per sound), run the same and should always be carefully distinguished.
sounds together, and leave sounds out. Nonethe-
less, the brain is able to process such rapid

25 SPEECH RECEPTION • 145

ACOUSTIC CUES
One reason why we are able to recognize speech,
despite all the acoustic variation in the signal,
even in very difficult listening conditions, is that
the speech situation contains a great deal of redun-
dancy — more information than is strictly necessary
to decode the message. There is, firstly, our general
ability to make predictions about the nature of
speech, based on our previous linguistic experience
- our knowledge of the speaker, subject matter,
language, and so on. But in addition, the wide range
of frequencies found in every speech signal presents
us with far more information than we need in order
to recognize what is being said. As a result, we
are able to focus our auditory attention on just
the relevant distinguishing features of the signal
- features that have come to be known as acoustic
cues.
What are these cues, and how can we prove their
7 as [d]; and patterns 13 and 14 as [g] (after A.
M. Liberman, et al., 1957).
and Such findings have laid the foundation for speech
perception studies; but a great deal still remains
to be explained. For example, it is not obvious how
listeners handle the difference between stressed and
unstressed sounds, or other modifications that
result from the speed of connected speech. More-
over, the acoustic values cited for the various
sounds are averages, and do not take into account
the many differences between speakers. Males,
females, and children will produce the same vowel
with very different formants, and it is not yet clear
how listeners make allowances for these differences
- for example, enabling them to judge that a male
[a] and female [a] are somehow the 'same'. Presu-
mably, they work out some way of relating vowel
values to the dimensions of the speaker's vocal
tract; but experiments have not yet been able to

role in the perception of speech? It is not possible establish exactly what this might be.
to obtain this information simply by carrying out
an acoustic analysis of natural speech (as in §23):
this would tell us what acoustic information is pre-
sent but not which features of the signal are actually
used by listeners in order to identify speech sounds.
The best an acoustic description can do is give us
a rough idea as to what a cue might be - for exam-
ple, a formant (p. 135) at a certain frequency, or
the duration of a burst of noise. But to learn about
listeners' perception, we need a different approach.
The main technique has been to create artificial
Time in msec
sounds using a speech synthesizer — an electronic
device that generates sound waves with any A series of control patterns for a speech synthesizer, in which F2 transitions are
systematically varied.
required combination of frequency, intensity, and
time (§§23, 26). In the classic experiments using
this device, the synthesizer was fed simplified pat-
terns of the kindproduced by a sound spectrograph
(p. 136). For example, it could be programmed
Early speech synthesis
to produce a sound with two formants at certain
A member of the Haskins
frequencies, and one could then see whether the Laboratories research
sound that emerged was recognizable as a certain team is shown using one
vowel. Or, a sequence of formants, formant transi- of the early speech syn-
thesizers, the Pattern
tions, and bursts of noise could be synthesized, to
Playback. A syllable was
see if listeners would perceive a particular sequence
painted onto an acetate
of consonant and vowel. film loop. The pattern was
Using this technique, in the 1960s researchers then read photoelectri-
at Haskins Laboratories in the USA found it cally, and converted into
was
an acoustic signal.
possible to establish the crucial role of the first two
formants for the recognition of vowels. Similarly,
the technique confirmed the importance of voice
onset time (p. 137) for discriminating voiceless and
voiced consonants. And, in an important series of
experiments, it was shown how the transitions of
the second formant are especially important as a
cue for place of articulation. The 14 patterns
(above right) represented 300 msec syllables differ-
ing from each other only in their second formant
transition. Fj was held steady, and F2 was made
to vary, as shown. These sounds were then gener-
ated in a speech synthesizer and presented to
listeners in random order. It was found that pat-
terns 1 and 2 were heard as [b]; patterns 6 and

146 -IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

 Categorical perception
When [pa] and
nounced in isolation,
consonants clearly differ in
voice onset time (VOT) -
the time between the re-
lease of the lips
set of vocal cord vibration
(p. 1 The average VOT
37).
value for [p] is +0.06 sec,
and for [b] is 0.0. What
would happen to a conso-
nant that was synthesized
midway between these va-
lues? Would be identifiedit
as or as [b]?
[p]
one study (L. Lisker &
In
A. Abramson, 1970), artifi-
cial syllables were synthe-
sized with VOT values
ranging from -0.1 5 sec to
+0.1 5 sec in steps of
0.01 sees. They were
played randomly to
[ba] are pro- teners, who had to
the whether they were [pa] or
[ba]. The graph below
gives the results.
When VOT was
and the on- than +0.03, the syllable
was usually identified as
when was more than
[ba]; it
+0.03, was heard as [pa].
it
For most of the study, lis-
teners were in total agree-
ment; only in a narrow
boundary region was there
any confusion. What is sig-
nificant is the very sharp
nature of this boundary (a
differential of only
0.05 sec). It seems that lis-
teners hear two possible
types of sound, regardless
of VOT variation. There is
very little room for uncer-
lis- tainty.
Vcce Onset Time
Selective listening
say Several experiments on
When listening to someone in a crowded room full of
other sound contrasts have
other conversations, we are able to attend to the
shown the generality of this
speaker and ignore the others. However, we hear if
finding, especially for con-
our name spoken nearby, we readily tune in to that
less sonants. It is even present
conversation, at the risk of ignoring the person we are
in infants (p. 238). Lis-
supposed to be listening to. Such 'cocktail-party
teners seem able to discri-
phenomena' illustrate the human ability to pay
minate items on an acous-
attention to some incoming sound stimuli and to ignore
tic continuum in a
others - what is known as selective listening.
categorical manner.
In early studies of selective listening, listeners were
Sounds near a boundary
presented with two spoken messages at once. They
area are interpreted as be-
were instructed to ignore one and to repeat the other
longing to one or other of
while listening to it at the same time ('shadowing'). It
the categories, and are not
turned out that people were able to do this very well.
felt to be uncertain cases.
Moreover, afterwards they had hardly any memory for
Categorical perception is
the ignored message - not even noticing when it was
plainly an important char-
changed from a male to a female speaker, or from one
acteristic speech percep-
language to another.
- and seems to
tual ability
This kind of study can be used to help demonstrate
be important for other as-
the fundamental role that meaning plays in the process
pects of sound discrimina-
of speech perception. For example, if one message
tion too.
is semantically similar to the other, or consists of
strings of cliches, listeners switch from one message
to the other without realizing it; whereas if the two
messages are semantically quite distinct, there is no
interference.
This point also emerges from studies of people
shadowing a single speech message. They can do this
very quickly - sometimes being only a quarter of a
second behind the speaker (no more than a syllable
or two). If errors are then deliberately put into the
speech stimulus (e.g. tomorrance for tomorrow, or pen-
knife for petrol), they are often spontaneously
corrected by the shadower. This too suggests that
linguistic factors have an important part to play in the
perception of continuous speech.
m Seconds

PERCEIVING CONTINUOUS SPEECH

A great deal of research has been carried out on when asked if there were any sounds missing, said
the perception of isolated sounds, syllables, or no; and even told that a substitution had been
if

words. In connected speech, however, very differ- made, most were unable to locate it. In another
ent processes seem to opeate. We do not perceive study, people were presented with one of four sen-
whole sentences as a sequence of isolated sounds. tences, in which a sound (marked =) had been rep-
Grammar and meaning (§§16-17) strongly laced by a cough, and were asked to identify the
influence our ability to identify linguistic units. word =eel.
Several experiments have pointed to these differ-
It was found that the =eel was on the axle.
ences. In one study, acoustically distorted words
It was found that the =eel was on the shoe.
were presented to listeners in isolation and in con- =eel was on the orange.
It was found that the A cocktail-party distraction.
text: the isolated words were identified far less ac-
It was found that the =eel was on the table.
curately. In another study, single words were cut
out of a tape recording of clear, intelligible, con- People responded with wheel, heel, peel, and meal
tinuous speech: when these were played to list- respectively, demonstrating the influence of gram-
eners, there was great difficulty in making a correct matical and semantic context in perceptual deci-
identification. Normal speech proves to be so sion-making (after R. M. Warren R. P. Warren, &
rapidly and informally articulated that in fact over 1970).
half the words cannot be recognized in isolation Results of this kind suggest that speech percep-
- and yet listeners have little trouble following it, tion is a highly active process, with people making
and can repeat whole sentences accurately. good the inadequacies of what they hear, arising
Another feature of continuous speech perception out of external noise, omitted sounds, and so on.
is that people 'hear'sounds to be present, even if A further implication is that models of speech per-
they are not. In one experiment, sentences were ception based on the study of isolated sounds and
recorded with a sound electronically removed, and words will be of little value in explaining the pro-
replaced with a cough or buzz. Most listeners, cesses that operate in relation to connected speech.

25 SPEECH RECEPTION 147 •

THEORIES OF SPEECH PERCEPTION Compromise
Theories of speech perception are usually classified Both approaches have their strengths and weak-
into two general types, in which listeners adopt nesses. Active approaches plausibly explain how
very different roles. listeners are able to adjust for such differences as
speaker accent, voice quality, and speed of speech.
Listeners are active And several kinds of experiment can be interpreted
In this view, listeners are thought to play an active to support this view. In 'shadowing' studies, for
role in speech perception, in the sense that when example, people are asked to repeat what someone
they hear a message, the sounds are decoded with says as quickly as possible, without waiting for the
reference to how they would be produced in speech. speaker to finish (p. 147). Listeners are evidently
The listener's knowledge of articulation (§22) acts able to carry out this task at great speed, copying
as a bridge between the acoustic signal and the sounds even before they have heard all the acoustic
identification of linguistic units. cues. To do this, they must be making active use
One major view, proposed in the 1960s, is called of their knowledge of linguistic structure.
the 'motor theory' of speech perception. This However, there are arguments against a wholly
theory argues that people internally model the arti- active view of speech perception. There are many
culatory movements of a speaker. They identify cases of people who cannot speak, for pathological
sounds by sensing the articulatory gestures that reasons (§46), but who can understand well. And
must have produced them — as if they were 'saying' it is possible to understand the speech of stutterers,

words to themselves to match the incoming speech. foreigners, young children, and others where it is

Another approach is known as 'analysis by syn- not possible to make a simple articulatory match.
thesis'. Here, listeners use a set of rules to analyse The passive approach does not encounter the pro-
an incoming acoustic signal into an abstract set blems introduced by postulating an intermediate,
of features. The same rules are used to synthesize articulatory step in speech perception. On the other
a matching version in production. The listener's hand, it has in turn been criticized for underesti-
perceptual system then compares the acoustic fea- mating the variability of the link between acoustic A model of the hypotheti-
cal brain mechanisms
tures of the incoming signal with the ones it has signals and linguistic units (p. 145) and for present-
used in speech perception
generated itself, and makes an identification. ing an account in which the processes of speech and production (after G.
production and those of speech perception are seen Fant, 1967). This is a pas-
Listeners are passive as entirely separate. It therefore seems likely that sive model, which speech
in
perception proceeds along
In this view, listeners play a passive role. They hear some combination of active and passive theories
the route ABCDE. In an ac-
a message, recognize the regular distinctive features will be required, in order to provide a satisfactory tive model, by contrast, the
of the waveform, and decode it. Listening is there- explanation of the process of speech perception. route would be ABCKFE.
fore essentially a sensory process, with the pattern
of information in the acoustic stimulus directly trig-
gering the neural response. No reference is made
to a mediating process of speech production I
1

(except in difficult conditions, such as noisy speech A B Primary C Auditory D

situations). The ears auditorv
analysis (subphonemic)
Several mechanisms have been proposed. One
Sensorv side
approach proposes a system of 'template matching' K
— listeners match incoming auditory patterns to a
.

Phonemes E Received
set of abstract speech patterns (such as phonemes Distinctive features Syllables message

and syllables, §28) that have already been stored Words Transmitted
Prosodemes message
in the brain. Another postulates the use of 'feature
K
detectors' — special neural receptors (analogous to Motor side
those known to exist in visual processing) that are Innervation Motor t i

I Speech H G patterns
organs of speech
capable of responding to specific features of the (subphonemic)
organs
sound stimulus, such as a particular formant, noise
burst, or other universal feature.

148 • IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

26 Speech interaction with machines

Is it possible to construct machines that will talk speakers, as long as the speech is not too rapid,

and understand speech? As early as the 18th cen-
tury, attempts were being made to devise ways of
mechanically reproducing the human voice. The
Austrian inventor, Wolfgang von Kempelen
(1734—1804) built one such machine, consisting
of a bellows to produce air flow, and other mech-
anisms to simulate parts of the vocal tract. Alex-
ander Bell (1847-1922) also constructed a 'talking
head', made out of various synthetic materials, that
was able to produce a few distinct sounds.
Modern techniques have led to massive progress
in this field. It is no longer necessary to build physi-
cal models of the vocal tract; sound waves can be
generated electronically by synthesizing the differ-
ent components of the sound wave. Early results
sounded very much like machines; more recently,
the quality of synthesized speech has greatly
improved — so much so, that with some devices
it is impossible to tell whether a machine or a
human being is talking. In most cases, however,
there are still problems of intelligibility and
naturalness to be overcome, especially in producing
speech with an acceptable intonation and rhythm.
In general, automatic talkers are programmed
with what to say. But current work in artificial
intelligence has led to speech that has been synthe-
sized on the basis of concepts derived from a
machine's own internal knowledge structure. This
is a major area of contemporary research, which
informal modifications are avoided, and the vo-
cabulary is limited. More advanced techniques of
pattern matching are now available as a result of
research into artificial intelligence, and projects are
under way to use the new generation of computers
to tackle these problems - but progress also
depends on filling the gaps in our understanding
of the acoustics of speech (§23).
Speech interaction with machines is unlikely to
become routine (in homes and offices) until the turn
of the century, though some devices will be avail-
able sooner. For example, currently being devel-
oped is a 10, 000- word, voice-activated typewriter,
for which the manufacturers claim 97% accuracy
in word recognition. The core vocabulary in this
machine can be used by any speaker; but the
machine has to be trained by individual speakers
before they can get access to the whole vocabulary
store. However, it remains to be seen whether such
devices can cope with the biggest problem facing
routine office use - the level of background noise.

draws greatly on the fields of psychology, computer

science, and linguistics. It is also hampered by the
limitations of these fields — for example, the primi-
tive accounts of human discourse structure avail-
able in linguistics (§20) make it difficult to devise
satisfactory question-answering systems for use
with machines.

SPEECH RECOGNITION
The problem of automatic speech recognition is
a more complex one. It requires the automation
of the processes of auditory perception and com-
prehension - neither of which is well understood
in its own has to cope with
right (§25). Also, it

the large amount of variation in speech (such as

regional accent and voice quality) and the noise
found in everyday speech situations. However,
there are now several devices that can recognize
a small vocabulary of words spoken clearly and
in isolation, or separated by pauses (as in a list
of telephone numbers), in a quiet environment. The A surviving fragment of
Alexander Graham Bell's
machines are particularly successful if they have
'talking head', made around
been given some information about the speaker's 1 863, and discovered many
voice characteristics in advance (a 'template' of the years later in the attic of

speaker). It is becoming increasingly possible

also Bell's house in Washington
to recognize chunks of continuous speech by single DC.

26 SPEECH INTERACTION WITH MACHINES •

149

WAYS OF MAKING THEM TALK
Different techniques are available for providing a
machine with a voice. The voices (human or synthe-
sized) can be pre-recorded, so that they can be pro-
duced at a predetermined point (when an engine
requires servicing, for example), or when a button
is pressed (as in many children's toys, such as
'Speak and Spell'). They are also widely used in
aids for the handicapped, and in computer assisted
terminals. But the approach has several serious
limitations. The linguistic needs have to be pre-
cisely anticipated, and changes are difficult to intro-
duce. Also, accessing the data takes a lot of time,
especially if a large vocabulary has been stored in
the machine.
The alternative is a system of speech 'synthesis
by rule'. Here, there are three main methods of
language analysis.
• The machine can translate whole words, by
looking them up in a table in which each word
has been given a synthesizer code. The words
are then pronounced as wholes, in a kind of
'look and say' approach (§34). This is easily
done when the vocabulary is small; but when
many words are involved, the approach becomes
150 IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING
The Voder Speech Synthe-
demonstrated at
sizer being
the 939 World's Fair by Bell
1
Telephones
cumbersome and restricting. Retrieving the
words is very time-consuming, and the system
cannot cope with new or unusual words (such
as proper names).
The machine analyses words into their consti-
tuent parts (morphemes, p. 90), and links these
with appropriate speech sounds. It has been esti-
mated that a dictionary of only 8,000 mor-
phemes is sufficient to handle 95% of texts.
However, a large computer system is needed in
order to achieve real-time performance.
The machine analyses words letter-by-letter, and
matches the patterns with a stored list of sound
correspondences, using a predetermined set of
rules (about 400, in one approach). The usual
method synthesizes speech on the basis of the
acoustic features of sounds (§23), which are then
combined into the vowel and consonant seg-
ments, words, and sentences, according to the
language's phonological rules (§28). Some
approaches use a database of already synthe-
sized segments. This is a kind of 'phonic' decod-
ing (§34) — essential if the system is to handle
all the words in a language. But it is difficult
to devise satisfactory rules to handle stress and
intonation.
Stages in human-machine speech • Speech interaction may be more expensive in
interaction running costs than using written or coded text.
Recognition Synthesis It is unclear whether these costs are outweighed

by the advantages.
Acoustic air pressure Goal/intention
waveform into microphone • The commercial demand for quick results has
led to some poor-quality research and develop-
ment.
• The ability to produce hardware is now far
Semantic construction
ahead of the relevant theoretical research fields,
Electrical waveform of the message
especially in speech perception and acoustics.
• Speech input to machines will be unreliable for
the foreseeable future because of interference
from other speech in the environment.
Analysis of waveform Syntactic construction • Routine speech input to machines leaves no per-
into frequencies of the message
manent record, whereas written or coded input
does.
• The concern to make machines sound human
is unnecessary. Sometimes mechanical voices are

Relevant features of
patterns (phonemes, etc.)
Decision about which
recognized unit occurred,
allowing for noise, etc.
Recognition of sequences
(words, phrases)
Analysis of syntactic
structure, allowing for
omitted words, errors
in unit recognition, etc.
It*
Analysis of semantic
structure
Orthographic form of more easily understood, especially in noisy work
the words conditions. And 'friendly' voice systems may
persuade non-expert users that the machines are
capable of more than they really are.
• Many people do not like talking machines
Phonological
because this suggests they have intelligence (a
specification currently controversial issue, in its own right).
In favour
At the same time, the field of human-machine
speech interaction has had many defenders, whose
Sounds analysed into arguments have been of the following kind.
acoustic parameters
• Speech interaction is the easiest and most natural
approach, and is less prone to errors than typing.
• It permits other activities to proceed at the same
time.
Electrical waveform
using speech synthesizer • It is more convenient inmany respects. Interac-
tion can take place even when the user is out
of sight or reach of the machine. An interaction
can be easily monitored by third parties and is
compatible with other everyday methods of
Audible pressure waveform
using loudspeaker
communication (e.g. by telephone).
• There are a large number of social and business
applications that promote accuracy, efficiency,
and cost-effectiveness.

The future
* The relativesequence of These arguments will doubtless continue. In the
events at the syntactic and
meantime, machine speech, of varying quality, is
Action semantic stages is theoreti-
cally controversial.
rapidly becoming a routine part of everyday life,
as the various technical problems are overcome.
We can hear automatic talkers in elevators,
announcing systems, assembly lines, calculators,
WHY BOTHER? children's toys, cars, speaking clocks, telephone
We are so used to hearing human voices emerging
answering services, and business offices (in provid-
from robots and computers in science fiction films
ing stock quotation reports). The potential of
that the desirability of human-machine speech
human-machine speech interaction is also being
interaction is often taken for granted. However,
actively researched in such areas as air-traffic con-
has not been without its critics.
this field of research
connection information, shop checkout
trol, airline
The arguments have been based on several theoreti-
monitoring, medical screening, robotics, and com-
cal and commercial considerations.
munication aids for the handicapped. There are
• The research (into speech recognition, in parti- still many problems to be solved, but a world in

cular) is unnecessary to social needs, and the which speech interaction with machines is routine
money would be better spent elsewhere. may now be only a generation away.

26 SPEECH INTERACTION WITH MACHINES 151

27 The sounds of speech

The description and classification of speech sounds In very precise descriptions of speech sounds,
is the main aim of phonetic science, or phonetics. other factors may also be noted, such as the relative
Sounds may be identified with reference to their position of the jaw or the overall shape of the
production (or 'articulation') in the vocal tract, tongue.
their acoustic transmission, or their auditory recep-
tion. The most widely used descriptions are arti-
culatory, because the vocal tract provides a
Vowels and consonants
convenient and well-understood reference point These two labels are probably the most familiar
(§22); but auditory- judgments play an important of all the terms used in the description of speech,
part in the identification of some sounds (vowels, but they nonetheless need to be used with great
in particular). The more precise and comprehensive care, to avoid mixing up two different kinds of
possibilities of acoustic description are outlined in definition.
§23. In a phonetic definition, vowels are distinguished
An articulatory phonetic description generally from consonants in terms of how they are articu-
makes reference to six main factors. lated in the vocal tract, and the associated patterns
of acoustic energy. In this approach, consonants
Air stream The source and direction of air flow
are defined as sounds made by a closure in the
identifies the basic class of sound. The vast majority
vocal tract, or by a narrowing which is so marked
of speech sounds are produced using pulmonic
that air cannot escape without producing audible
egressive air (p. 125). Non-pulmonic sounds
friction. Vowels are sounds that have no such stric-
include the clicks, implosives, and ejectives des-
ture: air escapes in a relatively unimpeded way
cribed on pp. 126—7.
through the mouth or nose. It is therefore relatively
easy to 'feel' the articulation of consonants; where-
Vocal folds The variable action of the vocal folds
as vowels, involving only slight movements of the
must be considered — in particular, the presence
tongue and lips, are difficult to locate in this way,
or absent of vibration (p. 128). Voiced sounds
and are easier to distinguish on auditory grounds.
are produced when the vocal folds vibrate; voice-
In a linguistic — strictly, a phonological — defini-
less sounds are produced when there is no vib-
tion (§28), vowels are distinguished from conson-
ration, the folds remaining open. Other vocal fold
ants in terms of how these units are used in the
actions are sometimes referred to (e.g. the closed
structure of spoken language. In this approach,
glottis used to produce the glottal stop).
consonants (C) are defined as the units that typic-
ally occur at the margins of syllables (p. 164);
Soft palate The position of the soft palate
(p. 130)
must be noted. When
lowered, air passes
it is
vowels (V) are the units that typically occur at the
through the nose, and the sound is described as centre of syllables. For example, in the syllables
nasal or nasalized; when it is raised, air passes pet /pet/, cat /kat/, and bus /bAs/, the syllable
through the mouth, and the sound is oral.
structure in each case consists of a central unit and
two marginal units, to produce the pattern CVC.
Place of articulation This parameter can be used In the case of most sounds, the phonetic and
to make several precise phonetic distinctions. It the phonological approaches coincide. For exam-
refers to the point in the vocal tract at which the ple [p], [f], and [m] are 'consonants' from both
main closure or narrowing is made, such as at the points of view. Phonetically, they involve closure
lips, teeth, or hard palate. Accompanying 'second- or audible friction. Phonologically, they function
ary' constrictions or movements may need to be at syllable margins, e.g. map /map/, puff /pAf/;
taken into account as well. there are no such syllables as */mpf/ or */ fp/- m
Similarly, [a], [i], and [o] are 'vowels' from both
Manner of articulation This is also a major des- points of view: phonetically, they are produced
criptive parameter, referring to the type of constric- without audible friction; and phonologically, they
tion or movement that takes place at any place occur at the centres of syllables, in such words as
of articulation, such as a marked degree of narrow- cap /kap/, hit /hit/, and hot /hot/.
ing, a closure with sudden release, or a closure with
slow release. The problem cases
In such cases as English [1], [j], [w], and [j], the
Lips The position of an important fea-
the lips is two sets of criteria conflict. From a phonological
ture of the description of certain sounds (especially point of view, these units typically occur at the
vowels), such as whether they are rounded or margins of syllables, as in let /let/, rat /rat/, wet
spread, closed or open. /wet/, and you /ju:/ and they must therefore be

152 • IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

considered as consonants. But from a phonetic
point of view, they are articulated without audible Ways of transcribing English vowels
friction, and acoustically they display a similar Several phonetic transcriptions for English vowels
energy pattern to that displayed by [a], [i], etc. (p. have been devised. The changes in symbol reflect
137). They must therefore be considered as vowels. different interpretations by the authors of the
relationships between the sounds. For example, Daniel
There are only two ways out of this problem.
Jones (p. 1 54) represents the difference between sear
One is to say that these four units are neither conso- and sit as essentially a contrast of vowel length, using
nants nor vowels but midway between these cate- the symbols [i:] and [i] respectively; whereas A. C.
gories. The terms semi-consonant or semi-vowel Gimson (191 7-85) represents it as a contrast of both
have often been introduced for this purpose, and length and quality, using a different symbol as well as
a length mark, [i:] and [i] respectively. It is important
this is the usual solution. The other, more radical,
to realize that all the authors are transcribing the same
solution is to introduce two completely different set of contrasts (apart from the case of the back open
sets of labels for talking about these units. The vowels, where American and British English accents
American phonetician Kenneth Pike (1912- differ). Their symbols simply draw our attention to
)
different aspects of the way the vowels are produced.
suggested the terms vocoid and contoid for the pho-
netic distinction, reserving vowel and consonant
Trager Kenyon
for the phonological one. Thus, all vowels are
Jones Gimson & Smith & Knott
vocoids; but consonants may be either contoids
(1956) (1962) (1951) (1935)
([p], [f], [h], etc.) or vocoids ([1], [i], [w], [j]). (The
further possibility of a contoid acting as a vowel seat i: i: iy i

is of marginal linguistic significance, e.g. the [s]

sit i i i i

of psst.)
set e e e e
It is certainly helpful to have two sets of terms
sat ae ae ae ae
to make it clear whether we are talking about sound
cut 3 A 3 A
units from a phonetic or a phonological point of
cart a a: a a
view. However, as only a small number of units
raise the problem in a severe form, the distinction cot D a a

has not been universally adopted. Moreover, it caught d: d: oh 3

needs further interpretation in the light of the way curt a: 3: ar 3-

a language's phonological system is organized full u u u U

(§28). fool u: u: uw u

This encyclopedia uses Gimson's system, with the

substitution of [a] for [x].

Vowels
Vowels are normally described with reference to
four criteria: Things that can happen to vowels
Vowel qualities can be much influenced by articulatory
• The part of the tongue that is raised - front,
movements elsewhere in the vocal tract, especially at
centre, or back (p. 131). the back of the mouth and in the throat. The
• The extent to which the tongue rises in the direc- possibilities include the following.

tion of the palate. Normally, three or four Nasalization

degrees are recognized: high, mid (often divided The soft palate is lowered, allowing some of the air
into mid-high and mid-low), and low. Alterna- stream escape through the nose- an important
to
feature of many vowels in French and Portuguese, for
tively, tongue height can be described as close,
example. The diacritic for nasalization is ~.
mid-close, mid-open, and open.
• The position of the soft palate - raised for oral Widening
In some languages, variations in the size of the
vowels, and lowered for vowels which have been pharynx (§22) affect vowel quality. 'Wide' vowels are
nasalized. produced when the tongue root is pulled forward and
• The kind of opening made at the lips — various the larynx is lowered, thus enlarging the pharynx.
degrees of lip rounding or spreading. Vowels where this does not take place are then termed
'narrow'. The West African language, Twi, illustrates
It is difficult to be precise about the exact articu-
this kind of contrast.
latory positions of the tongue and palate because
Rhoticization
very slight movements are involved, which give us
In some languages, vowels are articulated with an
very little internal sensation. Absolute values are additional feature: /--colouring, or rhoticization. These
not possible (such as saying that the tongue has vowels are usually articulated with the tip or front of
moved n millimetres in a certain direction), because the tongue raised, as would be the case for a
retroflex [r] (p. 1 55). They are widely used in American
the mouth dimensions are not the same between
and British English dialects in words where r follows
speakers. Vowel judgments therefore tend to be the vowel, as in sir, word. The phonetic symbol for a
made on the basis of auditory criteria, in associ- rhoticized vowel is
J
as in [a J ], though sometimes
,

ation with a limited amount of visual and tactile special symbols are used, such as [&].

information.

27 THE SOUNDS OF SPEECH •

153
 \ ,.

The Cardinal Vowel system The main vowel

The first widely used system for classifying vowels Using the Cardinal Vowel diagram diacritics
was devised by the British phonetician, Daniel Once the cardinal vowel values have been learned (a
Symbol T or
Jones (1881-1967). The Cardinal Vowel diagram matter of auditory practice), it is possible to place the
(or quadrilateral) is a set of standard reference vowels of a speaker of any language onto the chart articulation lower than cardi-
T
in a fairly precise way- if necessary, confirming the nal value, e.g. ?, e
points based on a combination of articulatory and
auditory judgments by acoustic measurements. In this
auditory judgments. The front, centre, and back way, typical articulations Symbol
in different languages can be - or
of the tongue are distinguished, as are four levels compared - the five-vowel system (p. 1 67) of Spanish,
articulation higher than cardi-
of tongue height: alongside that of Japanese, for example. The two
nal value, e.g. e, e.
systems are very similar, but most of the Japanese
• the highest position the tongue can achieve with- vowels are articulated in slightly more open positions +
Symbol or +
out producing audible friction; than the Spanish; the close back vowel also shows a
difference in lip rounding. (It is standard practice to use articulation further forward
• the lowest position the tongue can achieve; and
the nearest (rounded or unrounded) cardinal vowel than cardinal value, e.g. o +
• two intermediate levels, dividing the intervening symbols, when locating real' vowels on the chart.)
,

9
space into auditorily equidistant areas. English has a more complex vowel system, as can
Symbol ' or
_
be seen from the following chart (the accent
The vowel classification,
grid provides a basis for represented is received pronunciation', p. 39). articulation further back than
along with information about the accompanying cardinal value, e.g. i-, i

Symbol"
l.i 9.y 8.u J6.UJ
'\°
\ nasalized vowel, e.g. a

e\o \ Symbol ."

15. Y centralized vowel, e.g. e

2.e 10.0

Japanese \ Symbol:

<>6.d 14. long vowel, e.g. i:

3.t 11

V ^
shoe
Symbol

htd\_ half-long vowel, e.g.

4.a 12.CE 5.a U.o o\ put i-

Symbol >

head\p bird saw

position of the lips. Jones gave the main vowel sofa vowel more rounded than
points numbers, distinguishing a primary series normal, e.g. 0=
had*
(1—8) from a secondary series (9-16), and adding
two further points (17-18). Each of these vowel-
V\ ^ cut Symbol «

.h <RP\
0° hod
points was also given a phonetic symbol. The dis- vowel less rounded than nor-
hard
mal, e.g. 3<
tinction between primary and secondary cardinal
vowels is based on lip position. The first five pri-
mary vowels are all unrounded: front [i], [e], [e],
and [a], and back [a]. The remaining three back Vowel glides
vowels are rounded: [o], [o], and [u]. In the second- Languages frequently make use of a distinction
between vowels where the quality remains constant
ary series, the lip position is reversed: the first
throughout the articulation (pure or monophthong
five are rounded: front [y], and [ce], and
[0], [ce], vowels) and those where there is an audible change
back [d]. The remaining three back vowels are of quality.The latter are known as vowel glides. a If

unrounded: [a], [y], and [ui]. The two other vowels
represent the high points achieved by the centre
of the tongue: they are unrounded [i] and rounded
single movement of the tongue is involved, the glides
are called diphthongs; a double movement produces
triphthongs. Diphthongal glides in English can be
heard such words as say, fine, cow, boy, and so.
in

Triphthongal glides are found in certain pronunciations

Several other proposals have been made about of such words as fire, power, and sure.
ways of dividing up the vowel area that reflect arti- The tongue movements of several diphthongs,
represented by arrows, are drawn on the Cardinal
culatory movements (as established from a study
Vowel diagram. The accent is Midwestern American.
of X-ray photographs) more accurately; but Jones's
diagram continues to be widely used, especially in
Europe. It should be emphasized that the cardinal
vowels are not real vowels: they are invariable
reference points (available on record) that have to
be learned by rote. Once phoneticians have learned
them, they can be used to locate the position of
the vowels in any speaker. For more precise des-
criptions of vowels occurring within a broad area
of the diagram, it is possible to use diacritic marks
along with the vowel symbol, as shown on p. 153,
and below.

154 • IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

 . 1

Consonants • The position of the soft palate — whether raised

{oral) or lowered {nasal) (p. 130).
Consonants are normally described with reference • The place of articulation in the vocal tract.
to six criteria. • The manner of the articulation.

• The source of the air stream — whether from
the lungs {pulmonic) or from some other source
(non-pulmonic) (pp. 124—7).
• The direction of the air stream - whether moving
outwards Regressive) or inwards {ingressive) (pp.
126-7).
• The state of vibration of the vocal folds -
whether vibrating {voiced) or not {voiceless) (p.
128).
Sounds using non-pulmonic and ingressive air
streams (clicks, ejectives, and implosives) are des-
cribed on pp. 126-7. The present section therefore
deals largely with pulmonic egressive sounds,
which in fact constitute the vast majority of the
sounds of speech. Within the remaining criteria,
place and manner of articulation provide the main
possibilities for consonant variation.

PLACE OF ARTICULATION
Two reference points are involved in defining con- culator) (p. 130). Eleven possible places are used
sonantal places of articulation: the part of the vocal in speech, as indicated in the figure. (A full list
tract that moves (the 'active' articulator) and the of phonetic symbols is given on p. 159 and in

part with which it makes contact (the 'passive' arti- Appendix II.)

1 Bilabial. Both lips are in- as in some pronunciations of front of the tongue towards 10. Pharyngeal. The front
volved in the articulation, e.g. red) are sometimes clas- the hard palate, e.g. [TJ, [3], wall of the pharynx (in the
[Pi. [b], [mj. sified separately as post- as in shoe and French ye. region of the epiglottis) arti-
alveolar. culates with the back wall,
2. Labio-dental. The lower e.g. [h], both common in
7. Palatal. The front of the [?],
lip articulates with the upper 5. Retrof lex. The tip of the tongue articulates with the Arabic.
teeth, e.g. [f [v]. tongue is curled back to
j,
hard palate, e.g. [5], [j], as in
articulate with the area be- 1 . Glottal. The vocal folds
German ich and ja respec-
3. Dental. The tongue tip tween the rear of the alveolar come together to cause a
tively.
and rims articulate with the ridge and the front of the closure or friction, e.g. [h], [']

upper teeth, e.g. [8], [5], as hard palate, e.g. [tj, [dj, as (the glottal stop, p. 1 28) -a
in thin and this respectively. heard in many Indian English
8. Velar. The back of the rather different method of
tongue articulates with the articulation from any of the
accents.
soft palate, e.g. [k], [g]. other consonants.
4. Alveolar. The blade (and
sometimes the tip) of the 6. Palato-alveolar. The
tongue articulates with the blade (and sometimes the 9. Uvular. The back of the Other ways of describing
alveolar ridge (p. 1 30), e.g. tip) of the tongue articulates tongue articulates with the articulation, in the context of
[t], [s]. Sounds articulated at with the alveolar ridge, with uvula, e.g. [r], as in French phonology, are discussed in

the rear of this ridge (e.g. [j], a simultaneous raising of the rue (certain accents). §28.

Some consonant places of articulation

Bilabial [pi and [b] Alveolar [t] and [d] Velar [k] and [g], when fol-

lowed by an [i] vowel

Labio-dental [f] and [v] Dental [9] and 6] Alveolar [s] and [z]

27 THE SOUNDS OF SPEECH 155

 . . 1 - — i —

Coarticulation
Measuring coarticulation
The vocal organs do not move from sound to sound
is possible to measure
It articulations was plotted normal speech. In the case
in a of separate steps. Speech is a con-
series the degree of overlap be- using an electropalato- of [tk], one subject made an
tinuously varying process (p. 137), and sounds con- tween consonant articula- graph (p. 1 41 It might be
. almost simultaneous
tinually show the influence of their neighbours. For tions. Inone study, sub- expected that the [k] ob- closure of the two sounds,
jects pronounced words sure at the end of cock as can be seen from this
example, if a nasal consonant (such as [m]) pre-
containing the adjacent would be completely re- palatograph print-out of the
cedes an oral vowel (such as [a]), some of the nasal- consonants [kt] (as in cock- leased before the [tj clo- middle portion of the word
carry forward, so that the onset of the vowel
ity will fa/7), and [tk] (as in Wat- sure for tail was begun but ; Watkins.
will have a somewhat nasal quality. The reason kins), and the sequence of this does not happen in

is simply that it takes time for the soft palate to (Each circle represents a 401 402 403 404
move from its lowered position (required for [m]) coo. .0000 conn. ooooo 0000000
tongue and palate. The 00. . 000 000 . . ooo. 0000
to the raised position (required for [a]). It is still velar area is at the top of 00. .000 . 000. .000
each diagram; the alveo-
in the process of moving after the articulation of 00. . .CO . 00.. ..00
lar area is at the bottom.
OCO. ..0 000. .0
[a] has begun. Similarly, if [a] were followed by
. . .
Frames are at intervals
of msec (From W. J.
000. .00 . 000. ..00
[m], the soft palate would begin to lower during
1

Hardcastle & P. Roach,

OOOCOCCO 0000
the articulation of the vowel, to be ready for the 1977, p. 39).)
00.. .. oooouo
following nasal consonant. |t| closure made |k| closure made
When sounds involve overlapping or simulta- 405 406 407 408 409
neous articulations in this way, the process is ooooooocoo oocooooooo 0000000000 0000000000 cooooocooo
cooooooo ooooocoo 00000000 00000000 OODC 0000
known as coarticulation. If the sound becomes
COO.. 000 000.. 000 OCO.,.000 000. .OCO OCO. 0000
more following sound (its 'target'), we are
like a 00.... 00 000. ..00 CJOO...00 OCO. .000 000. .000
dealing with anticipatory coarticulation; if the 000... .0 000. ..0 . 00C. . .0 . 000... .0 000. ...0
000.. .00 000... 00 ooc. ..00 OOC. . .00 COO. .00 .

sound displays the influence of the preceding oocooooo 00000000 00000000 00300000 OGOCOOOO
sound, we are dealing with perseverative coarticu- 000000 000000 000000 000000 000000
lation. Anticipatory effects are far more common:
a typical example in English is the way vowel lip 410 411 412 413 414
0000000000 0000000000 cooooooooo OQOOCCCCOO OCCOG oooco
position affects a preceding [s], in such words as 00000000 cooooooo 00000000 00000000 OOCOOOOO
see (where the [s] is pronounced with spread lips) 00000000 ooooocoo cooooooo OCOOOOOO OOOOCOOC
000.. 000 000.0000 oocooooo 00000000 00000000
and sue (where [s] is pronounced with rounded coo. ..00 000. .00 000. ..00
000... 00 000... 00 .

lips). 000... 00 0000.000 ocoo .000 OCO. .000 000. ..00

00000000 00000000 ooco .000 0000. .00 000. .. .0
000,.. . 00 • • .
• 00. o
Secondary articulations [t) closure released
Often a consonant is produced using two points of arti- 415 416 417 419
418
culation, one closure (the 'primary' articulation) being 0000000000 ooooooooco cooooocooo COOO. 00000 0000. .000 .

more marked than the other (the secondary' articula- 00000000 00000000 00000000 000. .000 000. .000
tion). There are four main kinds of secondary articula-
00000000 OOOOODOO oocooooo 000. .000 000. 0000
000.. 000 000.. 000 coo. ..00 00.... 00 00.. ..00
tion.
000. .00 . OCO. .. .0 00. .

OCO.. . .0
Labialization The lips are rounded at the same time
as the primary articulation is made, as in the initial con-
sonants of sue and shoe. A labialized consonant is
[1 c] closure released [i] vowel begins
indicated by [J placed beneath the main symbol, or
H placed after it, as in [sj, [s*].

Palatalization The tongue is raised to a high front increased level

Parametric
position at the same time as the
primary articulation
is made. An [i] vowel resonance is added to the conson- phonetics Respirator)' ^—
activity
ant, and is symbolized by a small raised [ ]. Palatal- J
The notion of coarticula- open vibrating
ized consonants are found, for example, in Slavonic Vocal
tion applies to every arti-
folds
languages - as in the contrast between Russian palat- culatory component in- raised
alized [t
J
]
(e.g. [brat j ] 'to take')and non-palatalized [t] volved in the production Soft lowered y^~
(e.g. [brat] 'brother'). of a stretch of speech.
palate -^ raised
This can be seen if the Back of rest ^*-
Velarization The tongue is raised to a high back
functioning of each com- tongue raised
position at the same time as the
primary articulation ponent is plotted separa-
is made. An [ui] vowel resonance is added to the con- Front of
tely, toshow the relative tongue
sonant, and is symbolized by the sign [~] through the
movements of the articu-
consonant symbol. A velarized [1] can be heard in Eng- lators over time. The dia- Jaw rest position
lish, as in pool. A series of velarized sounds occurs
gram shows seven para-
in Arabic.
meters representing the lowered
Lips closed
action of the lungs, vocal
Pharyngealization The pharynx is narrowed at the
folds, soft palate, front
same time as the primary articulation is made. An [a] opened and rounded
and back of the tongue,
vowel resonance is added to the consonant, and is
jaw, and lips during the
symbolized by the [~] diacritic - the same as for velar-
articulation of the word
— 1
1

ization, but as no language is known to contrast these 0-2 0-4 0-6 0-8 10 1-2 1-4 1-6
horse {after L. F. Brosna-
articulations, there is no ambiguity. Pharyngealized Seconds
han & B. Malmberg,
consonants are found, for example, in Arabic.
1970, p. 70).

156 • IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

MANNER OF ARTICULATION lar ridge or the tongue back against the uvula, in
There are four main kinds of constriction made the different kinds of trilled r, heard for example
by the articulators in producing consonants, and in many English, French, and German accents.
these are used to provide a further dimension of
classification. (The full range of phonetic symbols Flap A single tap is made by one articulator
is given on p. 159.) against another, as in some pronunciations of the
r in v ery, or the d in ladder, where the tongue tip
Total closure taps once against the alveolar ridge. In Spanish,
Plosive A complete closure is made at some point a contrast is made between a trilled and a flapped
in thevocal tract; the soft palate is raised. Air pres- r, as in perro [pero] 'dog' and pero [pero] 'but'.

sure thus builds up behind the closure, which is Partial closure

then released explosively, as in [p] and [b]. The
Lateral A is made at some point
partial closure
broader category of stop includes closures pro-
in the mouth, such a way that the air stream
in
duced by other air streams (pp. 126—7), as well
is allowed to escape around the sides of the closure.
as plosives.
Various kinds of / sound are the result.

Nasal A complete closure made

at some point
is
Narrowing
in the mouth; the lowered, so that
soft palate is Fricative Two vocal organs come so close
air escapes through the nose, as in [m] and [n]. together that the movement of air between them Vowel-like
Voiceless nasals are shown with the diacritic [ ], causes audible friction, as in [f ], [z], [h]. Some frica- consonants
as in [ml. tives have a sharper sound than others, because
L J
o Certain consonants have
of the greater intensity of their high frequencies some of the phonetic proper-
Affricate A complete closure is made at some (p. 137): [s], [z], [/] (as in shoe), and [3] (as in ties of vowels: two alveolar

French These are known as sibilants. sounds, [1] and [j], the bila-
point in the mouth; the soft palate is raised. Air je).
bial [w], and the palatal [j].
pressure builds up behind the closure, and is then Linguistic range They are usually referred to
released relatively slowly (compared to a plosive The above descriptions give only a limited impres- as approximants (or friction-
release). The first element of the sound has a sharp less continuants), though the
sion of the range of sounds found in the languages
latter two are commonly
plosive character, but this is followed by an element of the world. A more accurate impression emerges called semi-vowels, as they
of audible friction, as in [tf] and [dj] (heard in Eng- when we examine some of the possibilities of arti- have exactly the same
lish church and judge respectively). culation within a single category of consonant. articulation as vowel glides.

Several kinds of stop consonant, for example, are Although phonetically vowel-
like, these sounds are
Intermittent closure illustrated in the following list (which does not
usually classified along with
Roll or Trill One articulator taps rapidly against show details of vowel quality). (After P. Ladefoged, consonants on functional
another - typically the tongue tip against the alveo- 1982.) grounds (p. 152).

Description Symbol Example Language

Manner
voiced b banu Sindhi 'forest'
voiceless unaspirated P psnu Sindhi 'leaf
h h
aspirated P p snu Sindhi 'snake hood'
murmured (breathy) b fi
b 3n3.nu
fi
Sindhi 'lamentation'
implosive 6 6sni Sindhi 'curse'
laryngealized (creaky) h baabe Hausa 'quarrel' (verb)
ejective k' k'aaka Hausa 'how'
nasal release dn dno Russian 'bottom'
prenasalized nd ndizi Swahili 'banana'
lateral release ti tiah Navaho 'oil'
ejective lateral release ti' ti'ee? Navaho 'night'
affricate ts tsait German 'time'
ejective affricate ts' ts'aal Navaho 'cradle'
Place
bilabial pb pig English
dental td muttu Malayalam 'pearl'
alveolar td muttu Malayalam 'density'
retroflex td. muttu Malayalam 'knee'
palatal cj ciri Quechua 'cold'
velar kg kara Quechua 'expensive'
uvular qG qara Quechua 'skin'
glottal ? ? alla Arabic 'God'
labio-velar kpgb akpa Yoruba 'arm'

27 THE SOUNDS OF SPEECH •

157
 PHONETIQUE
L' ASSOCIATION main principles were that there should be a separate
INTERNATIONALE (INTERNATIONAL letter for each distinctive sound, and that the same
PHONETIC ASSOCIATION) symbol should be used for that sound in any lan-
This Association was inaugurated in 1886 by a guage in which it appears. The alphabet was to
small group of language teachers in France who consist of asmany roman alphabet letters as pos-
had found the practice of phonetics useful in their sible,using new letters and diacritics only when
work and wished to popularize the methods. It was absolutely necessary. These principles continue to
first known as The Phonetic Teachers' Association, be followed today.
changing to its present title in 1897. The IPA has been modified and extended several
One of the first activities of the Association was times, and is now widely used in dictionaries and
to produce a journal in which the contents were textbooks throughout the world. Some of its spe-
printed entirely in phonetic transcription. The idea cial letters have even been accepted as part of the

of establishing a phonetic alphabet was first pro- new orthographies devised for previously unwrit-
posed by Otto Jespersen (1869-1943) in 1886, and ten languages, such as in certain parts of Africa.
Paul Passy, founder of the
the first version of the International Phonetic The Association is now based at University College International Phonetic
Alphabet (IPA) was published in August 1888. Its London. Association

The index page to the last number of Le Maitre

Phonetique, which appeared in 1970. The headings
da laist m.f.
are in French, the official language of the Association.
az membaz wil nau, ois iz 5a la:st nAmbar av Si m.f. in its preznt
Each article has been written in a transcription that
partly reflects the pronunciation of the author. For fo:m. a: d33:nl waz pAbliJt fa 09 f3:st taim in 1889, 5au priivjash, fram
example, Soravia uses [ou] to represent the diphthong 1886, it ad apiad az " 5a fanetik ti:t|a ". in 1889, a:r asausieijn haed
found in such words as know ([founetiks] = phonetics),
whereas Lewis uses [au] (as in [taunl] = tonal) and 321 membaz in 18 kAntnz, 5a mad3orati kAmirj fram *swi:dn,
Fox uses [aQ] ([taQn] = tone). The asterisk is used *d33:mani an *fra:ns. tadei, wi: haev mo: 5n 800 membaz in auva 40
before a word that is a proper name.
kAntnz, 5a greit mad3Drati kAmirj fram 5a *junaitid steits an *greit
bntn.
nau 5at wi: av disaidid ta print a: nju: Journal in o:6Dgrafi, fa 5a
13
f3:st taim in d3u:n 1971, it iz haupt Sat 5a ri:dajip wil bi inland an
Sat kontribju:Jnz wil bi nsi:vd fram a waida S3:kl av faunitijnz an
fonetik ti:tjaz. maust av a: membaz hu: av nplaid ta 5a S3:kjalar in 5a la:st

me ;tr8 ,

ionetik
etetnasional,
1970
m.f. hav signifaid Sat 5ei wif ta kantinju: ta sabskraib ta 5a nju:
Journal. Sauz hu: av nt?t jet mfo:md as av Sear mtenfnz ar 3:d3d ta du:

asosiasp sau wiSaut dilei, sins a: famsensiz wil not alau as ta send 5a Journal ta
l
organ d9 fo:ma membaz hu:z sabsknpjnz a not nnju:d.
wi ikspekt 3a nju: Journal ta kantem about 50 peid3iz at 5l

de matH-t
pa--3 outset, fa Sis ri:zn, in 5a f3:st n.\mbaz at h:st, wi: wil limit 5a lerjkO av
tabla
kDntnbju:Jnz tu a maeksimam av about 3,000 W3:dz. Sa folauirj nauts fa
, •
i„tt (G.
Soravia) .
kantnbjutaz giv an indikeijn av 5a nkwaiamants av preznteijn fa 5a
Journal Sei wil in fju:t/a bi printid Dn 5a Journalz kAva.
;

The Association Secre- though previously, from States and Great Britain.
K1
*rDd53 V" tary's statement explain-
Dietnch)
•
•
1886, it had appeared as Now that we have de-
(G. ing the demise of Le Maitre The Phonetic Teacher'. In cided to print our new Jour-
Phonetique. 1889, our association had nals orthography, for the
kowksjo •

Canepati) '
-
- in
d The notice, which appeared 321 members in 18 coun- firsttime in June 1 971 it is
,

in the 1970 issue, was tries, the majority coming hoped that the readership
on headed The last m.f.': from Sweden, Germany will be enlarged and that
9 naut Leung) .
'As members will know, this and France. Today, we
ksentani-z ^- • ~ Hoc\ ek)
contributions will be re-
is the last number of the have more than 800 ceived from a wider circle
mo* about 3*9
^^ • •
m.f. in its present form. Our
journal was published for
members in over 40 coun-
tries, the great majority
of phoneticians and
.'

o^nnaunW^- .
. •
the first time in 1889, coming from the United
teachers . .

36

09 t3 onl
-tani a
^^ •

^
37

38

WW'.****-*-

158 •
IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING
 : : : : :

THE INTERNATIONAL PHONETIC ALPHABET (Revised to 1979)

Dental, Palate-
Bilabial Labiodental Alveolar, or Retro/lex alveolar Palatal Velor Uvular Labial- Labial- Pharyngeal GloOai
Poet-alveolar Palatal Velar

o,
^aeal m "0 n 1 Ji 3 N

Plonve P b t d k kp gb ?
t 4 J 9 q
" I (Median)
3 Fnvatxve * P f V SA ii 1 3 * Y X « M 6 ? h 6

E (Median)
j Appraximanl u i i M W
•* *
«(

AT
<w
El * *

pulmonic

ll i 1 I X

frill r R
CO
Top or Flap ' t

* Ejective P' k*

/mpioflt* 6
1 | g
cf

o 1 C

w Lateral Click 6

DIACRITIC8 OTHER SYMBOLS Front Back VOWELS Front Back STRESS, TONE (PITCH)

. Voiceless a • or . Raised e-, e, e w p, 1 Alveolo-palatal fricatives

placed at begin-
<J
i i in Close y u '
stress,
. Voiced 5 \
• or ,
Lowered e-, e, e |r L 3 Palatalized /, 3 1 Y a ning of stressed syllable
b Aspirated t" , Advanced u+, u r Alveolar fricative trill
secondary stress " high :
,

Breathy-voiced b s> - or - Retracted i, i-, t J Alveolar lateral flap

level pitch, high tone
..

"
c Y Half-close e
. Dental ( Centralized e fj Simultaneous
J and x a e low level ' high rising:

Labialized - Nasalized a Variety of J resembling s,

. t J" low rising ' high falling
: :

Palatalized J *, " r-coloured oJ etc.

j ; ,
e A Half-open oe low falling * rise-fall :

- Valorized or Phoryn- : Long a: 1 = 1

'fall-rise.
gealized t, 1
'
Half-long a- = e »
, Syllabic 9 |
" Non-syllabic 3 = Variety of a AFFRICATES can be
"
or . Simultaneous sf (but see > More rounded »• » = r-coloured 3
a a Open (X written as digraphs, as
also under the heading < Less rounded y< ligatures, or with slur
Affricates) Rounded marks thus to, t( dj
Unrounded ; ,

»tjd5: tstjfs.
c, ] may occasionally be
The International Phonetic Alphabet (1979 revision) used for t/, dj

Category Examples Left: Some of the symbols

partially voiced (for used by a British research
Place
bilabial
segments normally team for transcribing dis-
trills
PPP b ordered speech (after P.
voiceless) .s s.
lingualabials (tongue tip to h h(
preaspirated p Grunwell, era/., 1980).
upper lip)
Note the final section re-
plosives P B Air-stream mechanism
commending symbols for
nasal M pulmonic ingressive s m use when the phonetician Below: Some of the symbols
*-~ *~~
lateral L zero air stream with used by an American re-
isunsure which sound has
reverse labio-dentals articulation present search team for transcribing
been used - a common
(lower teeth to upper lip) p m f ('mouthing') (0 (m) problem when transcribing the speech of children (after
u u u
bidentals (lower teeth to C. Bush, et a/., 1973)
Duration and pause the unstable pronuncia-
upper teeth) h *
n excessively short m i tions of handicapped
Manner prolonged (using existing speakers.
nasal fricatives m F
n
F
I PA symbol) m: p:
strong/tense articulation f m silence: short, long, extra-
weak/lax/tentative long IV
articulation f m
Data inadequate
reiterated articulation p-p-p
'not sure' Ring doubtful \
plosive with non-audible
symbol \
release P'b'
e.g. unspecified \
Vocal fold activity segment
unaspirated (marked unspecified ©
explicitly) P
= r consonant
prevoiced (voicing starts unspecified vowel ®
earlier than normal); unspecified fricative ©
postvoiced (voicing probably [t], but not ©
starts later) -V z- sure

27 THE SOUNDS OF SPEECH 159

28 The linguistic use of sound

Phonetics is the study of how

speech sounds are kinds of abstract representation it is necessary to Some minimal
made, transmitted, and received (§27). It is a sub- set up in order to explain the range and distribution pairs for English
ject that requires as its source of data a human of phonetic segments found in languages. And in phonemes
being with an intact auditory mechanism and a the context of generative linguistics (p. 409), there
(southern British)
functioning set of vocal organs. The person's parti- is an even more ambitious aim: to arrive at phono-
Vowels
cular language background is not strictly relevant: logical analyses that have a demonstrable mental
N-N seat - sit
phoneticians would draw the same conclusions reality for the language users (p. 163).
N-N sit- set
about the production and reception of speech N-N set - sat
whether they were dealing with speakers of Eng- PHONEMES N-N cat - cut

lish,Hindi, or Chinese. Although the categories Phonological analysis relies on the principle that N-N/ cut -cart

outlined in §27 can be used for the analysis of any certain sounds cause changes in the meaning of
Nl-N cart - cot

language, that section provides no information a word or phrase, whereas other sounds do not.
N-N/ cot -caught
Nl-N cord - could
about the way these categories are actually used, An early approach to the subject used a simple N-N/ pull -pool
in the languages of the world. methodology to demonstrate this. It would take NI-NI pool - pearl
By contrast, the primary aim of phonology is a word, replace one sound by another, and see NI-NI pearl - pale

to discover the principles that govern the way whether a different meaning resulted. For example,
Nl-N/ day -die
NI-NI buy - boy
sounds are organized in languages, and to explain
the variations that occur. A common methodology
we hear pig in English as consisting of three separ-
ate sounds, each of which can be given a symbol
N-H
M-/au/
toy - toe
hoe - how
is to begin by analysing an individual language, in a phonetic transcription, [pig]. If we replace [p] /au/-/iV now -near
/I3/-/C3/ tear (noun) -
to determine which sound units are used and how by, say, [b], a different word results: big. [p] and tear (verb)
they pattern —
the language's 'phonological struc- [b] are thus important sounds in English, because /ea/-/oV tear - tour
ture'. The properties of different sound systems are they enable us to distinguish between pig and big, Nl-N sure -she
then compared, and hypotheses developed about pan and ban, and many more word pairs. /a/- zero waiter - wait

the rules underlying the use of sounds in particular In a similar way, [i] and [e] can be shown to
Consonants
groups of languages, and ultimately in all languages be important units, because they distinguish N-N pig - big
('phonological universals', §14). between pig and peg, pin and pen, and many other N-N bee - tea
The distinction between phonetics and phono- pairs. And so we could continue, using this techni- N-N tin -din

logy can be seen from a second point of view. The que — the 'minimal pairs' test - to find out which N-N din -kin

human vocal apparatus can produce a very wide N-N cap -gap

range of sounds; but only a small number of these

sound substitutions cause differences of meaning.
The technique has its limitations (it is not always
N-N gag - hag

are used in a language to construct all of its words

N-N/ hen - men
possible to find pairs of words illustrating a particu- Nl-N map - nap
and sentences. Phonetics is the study of all possible lar distinction in a language), but it works quite N-N sin -sing

speech sounds; phonology studies the way in which well for English, where it leads to the identification N-N sink -silk

a language's speakers systematically use a selection of over 40 important approach N-N lid -rid

of these sounds in order to express meaning.

units. In the earliest
to phonological analysis, these 'important units'
N-H red- wed
- yell
Nl-lil well
There is a further way of drawing the distinction. are called phonemes. nl-Ni you - chew
No two speakers have anatomically identical vocal Phonemes are transcribed using the normal set N-N/ chin -gin
- laugh
tracts, and thus no-one pronounces sounds in of phonetic symbols 159), but within slant lines, NI-IU large

exactly the same way as anyone else (a motivation

(p.
— /p/, /b/, /i/, etc. This shows N-N - vat
fat

for the study of voiceprints, §6). There is even a

not square brackets
that the units are being seen as part of a language,
N-N heave -
heath
considerable amount of variation in the sounds of and not just as physical sounds. N-N wreath -
wreathe
a single speaker. Yet when using our language we
are able to discount much of this variation, and Allophones N-N though -so

focus on only those sounds, or properties of sound, In working out the inventory of phonemes in a
N-N bus - buzz
zoo - shoe
N-lll
that are important to the communication of mean- language, using this approach, we soon come lll-N Confucian -
ing. We think of our fellow-speakers as using the across sounds that do not change the meaning confusion

'same' sounds, even though acoustically they are when we make For example, the
a substitution. N-N beige - bait

not. Phonology is the study of how we find order
within the apparent chaos of speech sounds.
In its search for significant generalizations about
sound systems, phonology is continually looking
beneath the 'surface' of speech, to determine its
underlying regularities, and to establish how these
relate to other areas of language, notably syntax
and morphology (§16). Much of present-day pho-
nological theory is thus concerned with the various
consonants at the beginning of shoe and she have
very different sound qualities. For shoe, the lips
are rounded, because of the influence of the follow-
ing [u] vowel; for she, the lips are spread. If we
now substitute one of these sounds for the other,
we do not get a change of meaning — only a rather
strange-sounding pronunciation. There is only one
phoneme here — the voiceless palato-alveolar pho-
neme l\l (p. 155) — but it turns up in two different

160 • IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

phonetic 'shapes', or variant forms, in these two
words. These phonetic variants of a phoneme are
Grouping sounds
known as allophones. In the phonemic approach
tophonology, linguists
When we study a new language, it is important faced with an array of
to pay careful attention to the phonetic variations sounds usually use three
which occur, to ensure that we make the right deci- deciding whether
criteria in
sions about which sounds count as phonemes and these sounds belong to the
same phoneme.
which count as allophones. We do not know this Complementary distribu-
information in advance; we have to work it out.
into phonemes
voiceless plosive sounds at ples, satisfy this criterion,
the end of words are some- as the variants in each
times articulated in a re- case have a great deal in
laxed way, and sometimes common - the/J/s are both
are pronounced quite, voiceless palato-alveolar
strongly. The/p/ of cup and the/p/s are
fricatives,
might be heard with a tiny
both voiceless bilabial plo-
amount of audible breath sives. However, is some-
it

tionThe sounds must ('aspiration') following its times possible to find

And in doing so we have to be ready to cope with complement each other, in
differencesbetween the way sounds work in differ- terms of where they occur
in words. For example,
ent languages. For example, English does not dis-
the case of the two/ J/
tinguish the meanings of words using a contrast sounds in shoe and she,
between [[*] and [J"], but some other languages do the rounded variety occurs
(e.g. Lak). Sound
differences that separate allo- only before rounded
phones in English may separate phonemes in vowels, and the spread
variety only before non-
another language, and vice versa - a principle that
rounded vowels. Where we
is clearly illustrated by the / sounds of such words find the one, we do nor find
as leaf and pool. The first / ('clear' /) is articulated the other: they are mutually
much further forward in the mouth than the second exclusive, never occurring
in the same phonetic
('dark' /) - as can be felt, if the sounds are said
environment. Such sounds
slowly to oneself. In English, these are allophones are said to be in 'comple-
of a single /l/ phoneme. In Russian, however, they mentary distribution'.
are different phonemes. Free variation If the
sounds do occur in the
same place in a word, then
they can belong to the
same phoneme only if
do not change the meaning
of the word. For example,
release, or a relatively
large amount. But the
in ferent amounts of aspira-
tion do not affect the mean-
ing of the word: replacing
weakly aspirated [p
h
l by
strongly aspirated [p"]
does not thereby change
cup into some other word.
Such sounds are said to be
in 'free variation' - though
whether the variation is in
fact genuinely free, and not
conditioned by such factors
as social class or regional
background, is an interest-
ing question (p. 332).
Phonetic similarity To be-
long to the same phoneme,
sounds ought to display a
reasonable amount of phy-
they sical similarity. The two
kinds of /// or the two kinds
of /p/, in the above exam-
sounds in complementary
dif- distribution that are nof
phonetically similar, and
in
these cases analysts would
be reluctant to treat them
as members of the same
phoneme. A case in point
is English [h] and [rj]: the
former occurs at the begin-
ning or in the middle of
words; the latter only in the
middle or at the end. They
therefore rarely contrast.
Could they, then, be taken
as allophones of a single
phoneme? No, because
they have nothing phoneti-
cally in common, apart
from both being conson-
ants - [h] is a voiceless
glottal fricative; [rj] is a
voiced nasal continuant.

How many minimal pairs are there?

A convenient way of dis- vertically on the left, and late the amount
work a of

Beyond the segment playing a language's pho- the vowel sounds are listed phoneme has to do in a
nemic substitutions is to horizontally across the top. language. For example,
Several approaches to phonology have assumed that
construct a chart of pos- Charts of this kind have English does not use final
a language's sound system can best be analysed in
sible words or syllables. all kinds of practical appli- consonants with equal fre-
terms of a series of individual segments ([b], [a], [s],
Below is part of a chart cations. They can help lan- quency, as can be seen
etc.). But there are a number of phonological
adapted from Denyse guage teachers and from the following list,
characteristics that affect units that are much larger
Rockey's Phonetic Lexicon speech therapists in pro- which is derived from
than the individual segment, such as syllables, words,
(1973, pp. 56-7). It shows nunciation work. They can Rockey's data. Each figure
phrases, and sentences.
some of the 1 1 7 monosyll- be a source of information refers to the number of
Several segments in a word or phrase may display
ables in English that end to budding poets and monosyllabic words ending
the same phonetic feature - for example, they may all
with/b/ (though this figure Scrabble-masters (p. 64). with the consonant listed. It
be lip-rounded or nasalized (p. 153). In particular, includes several obsolete, Linguists can compare the shows, for instance, that
languages often display cases of harmony between dialect,and technical use a language makes of over twice as many mono-
consonants or vowels. In certain kinds of 'vowel words). The initial sounds individual combinations of syllables end in/k/as end in
harmony', for example, all the vowels within a word of these words are listed phonemes and thus calcu- hi-
have to be of the same general type. Turkish is such
a case, where words contain (with certain exceptions) D(+r) d/q o(+r) 429
Ad/
only front vowels or back vowels. Thus we find [verdim]
'Igave' with front vowels, and [tfocfeuk] 'child' with back
ebb abb ob orb
H
H
383
376
vowels. But no words are formed with front + back
combinations, such as [e] + [o]- a situation quite unlike
bib . barb bob
NV 330
313
English, where the sequence of vowels in a
polysyllabic word is not predictable in this way.
Tib tab . AV 304
dieb dib deb dab Dob daub /-nV 240
The analysis of phonological features in terms of keb cab cob corbe 223
/-P/
units larger than the segment is a preoccupation of 212
gib gab garb gob gaub As/
several current theories, such as 'prosodic' phonology 153
fib fob l-V
and 'autosegmental' phonology. Patterns of pitch, 138
Itl
loudness, tempo, rhythm, and tone of voice provide 132
another set of data which cannot be analysed with
reference to single segments. These aspects of
sib sab sob sorb
H
Atf/

AV
122
117
phonology are usually studied under a heading that 105
HI
well reflects this different emphasis: 'suprasegmental' /-e/ 104
phonology (§29). A specific approach that emphasizes
the relationship between segments and syllabic
hi 87
hob AdV 84
sequences of rhythm and stress is known as 'metrical' A6/ 32
phonology. jab job

28 THE LINGUISTIC USE OF SOUND •

161
DISTINCTIVE FEATURES theory, these features are given two values, symbo- Distinctive-feature
In a phonemic necessary to recognize
analysis, it is lizedby the signs -I- and -, as in [±voice], [±nasal]. matrices
smaller units than the segment, in order to explain For example, [n] is both [+nasal] and [+voice];
The features are listed on the
how sets of sounds are related. This can be seen [p] is[-nasal] and [-voice]. A small set of these left each matrix, and the
of
by comparing any two contrasting segments, using contrasts is worked out and applied to all the segments are listed along
the articulatory criteria introduced in §27. sounds that turn up in a language. Results may the top. Each segment is
analysed in terms of all fea-
be presented in the form of a matrix, in which the
• English /p/ and /b/ differ in one respect only: tures. The terminology used
presence or absence of each feature is noted (see
in these particular matrices
/p/is voiceless, and /b/ is voiced. In other res-
below). relates to the traditional arti-
pects, they are the same: they are both bilabial,
Distinctive feature theory has been primarily culatory terms used in §27 in
and pulmonic egressive.
plosive, oral, the following way (V = vowel,
used by generative approaches to linguistics (§65),
• /p/ and/g/ differ in two respects: there is a con- C = consonant):
where the aim is to provide an account of phono-
trast of voicing, and there is also a contrast in + compact lowV
logy that can be integrated within a theory of gram-
the place of articulation - bilabial vs velar. - compact high and mid
mar (§16). It is argued that distinctive features are
• /p/ and /z/ differ in three respects: this time, V
the important facts to take into account when
there is a contrast in the manner of articulation + consonantal obstruction in
carrying out a phonological analysis, as they reveal
(plosive vs fricative), alongside the contrasts in vocal tract
more about the way in which the sounds of a lan-
voicing and place. - consonantal no vocal tract
guage are organized, and more readily permit
obstruction
All segments in a language can be analysed in generalized statements within and between lan-
+ continuant fricative/
this way, either from an articulatory or an acoustic guages, than do descriptions based on phonemes approximant
(p. 146) point of view, and the result is a set of and allophones. A particular advantage is that the C
contrasting components known as distinctive fea- same set of terms can be used for describing both - continuant stop/ affricate
tures. The English segment /p/, for example, is a vowels and consonants - something traditional C
combination of the features of 'voicelessness', 'plo- articulatory descriptions were unable to do (as can + diffuse high V;
siveness', and 'bilabiality'. In distinctive feature be seen from the diverse terminology of §27). labial/dental/
alveolar C
- diffuse lowV;
palatal/velar/
backC
-t-flat rounded V
English consonant matrix -flat unrounded V
pbfvmtd08szntfc%J'3kgl rwj hrj + grave back V;

consonantal
__________________
+ + + + + --
____ + + + + + + + + + + + + + + + — +
labial/velar/
backC
vocalic
diffuse + + + ----
________________________
+ + + + + + + + + — —
+
+
+
+ — — — -
- grave front V;

compact
grave
flat
+ + + ----
____________________
+ -+--___
+ — — — - — — — + + —
+
+ +
dental/
alveolar/
palatal C
voice - + --+ +
-- + - + - + + - + - + - + + + + + — + + nasal nasal C
continuant
strident
nasal
-
__ _____
— + + ++
+_ ________
—
____ ______ ___________
+ +
+
— - + +
+
—
+
+ +
+
+
+
+
— — + + + + + -

+
- nasal
+ strident
oralC
fricative/
affricate C
with high-
frequency
noise
- strident C with low-
Matrix for a seven-vowel system frequency
noise
i e a u a
+ vocalic glottal
consonantal vibration with
vocalic + + + + + + + free passage
diffuse + - - + - - - of air through
compact - - + - - + + vocal tract
grave - - - + + + + - vocalic no glottal
flat — — — + + + — vibration or
voice + + + + + + + free passage
continuant + + + + + + + of air
strident
nasal
+ voice voiced C
- voice voiceless C
(After R. Jakobson & M.
Halle, 1956.)

162 •
IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING
PHONOLOGICAL RULES domain of connected speech, would be the change
In traditional accounts of phonology, a sound is of [n] to [m] in the phrase ten boys, because of
described as occurring in a particular position the influence of the following [b]. Here, the rule
within a syllable or word, and that is all. No refer- would summarize the fact that 'an alveolar nasal
ence is made to our knowledge of the relationships becomes bilabial before a following bilabial conso-
that exist between the various types of sound in nant'.
different contexts. Yet this information is essential Other rules add or delete segments. An addition
if we are to understand the way sounds systemati- rule accounts for the way
which some English in
cally relate to each other and to the grammar and accents add vowels between certain consonant seg-
lexicon of a language. ments, as in the pronunciation of film as [fitam].
To illustrate this point, we may consider such A deletion rule occurs when vowel segments are
pairs of words as telegraph and telegraphy. A pho- regularly omitted from such phrases as J am (—*
nological analysis of these words is not complete I'm) in certain grammatical contexts. There are
simply by giving each a phonemic transcription: also rules that combine two segments
as one, as
/tebgraif/ vs /talegrsfi/. We also need to show when would + you become [wudju:] (p. 164).
that, despite the different patterns of strong and Phonological rules are not restricted to making
weak vowels within them, the pronunciations are statements about the sound patterns of a particular
systematically related, with other pairs of words language. They are also used to demonstrate the
in the language displaying the same kind of rela- similarities and differences between the sound sys-
tionship (such as microscope /microscopy). In tems of different languages. Is the rule about conso-
recent years, relationships of this kind have become nant devoicing at the ends of words found only
a major focus of phonological investigation. And in English, or does it apply to a larger group of
one of the main techniques for demonstrating such languages, or possibly to all languages? The formu-
regularities in the sound patterns of language has lation of phonological rules is thus seen as an
been through the use of phonological rules. important step towards the phonologist's goal of
Phonological rules are general statements about discovering the universal principles governing the
the relationships between sounds, or classes of use of sound in language.
sound. They summarize what happens when
sounds occur in particular grammatical or phonetic
contexts. In English, for example, [b] is used at Abstract or concrete?
the beginning and at the end of words, but espe- In order to arrive at satis- ciations, each of which is tion, which lacks clear pho-
cially in the latter position it loses some of its voic- factory generalizations, nasal. But what happens if netic motivation. Moreover,

ing: we say [ctjab] (jab), with a 'devoiced' sound. phonologists have often in- we extend the example to it is not immediately ob-

troduced abstract underly-
This observation can be summarized in the form
ing forms into their rules
of a rule: '[b] becomes [b] at the end of a word.' from which several pronun-
The validity of the rule can then be tested against ciations can be derived.
other examples, to see if there are exceptions. For example, the words im-
Phonological rules are expressed in a special possible, indecisive, and
inconclusive all begin with
notation to make the description as clear and suc-
the same prefix, meaning
cinct as possible and (according to some analysts) 'not', but the pronuncia-
to identify the essential theoretical properties of tions differ. In the case, first
sound systems. The above rule could be written it is [im] (because of the fol-
lowing bilabial); in the sec-
as follows (the symbol '—*' means 'becomes'; '/'
ond case it is [in] (before
means 'in the context of; and '#' means 'word the alveolar consonant);
boundary'): and in the third case, for
many speakers, it is [irj]
[b] -[&]/-# (before the velar conso-
nant). How can this varia-
In generative phonology, such rules would be writ- tion be explained?
ten using a distinctive feature notation: is not very convincing
It
to suggest that one form is
+ consonantal + consonantal more important than the
others, and set up a rule
- nasal / - #
which two of these forms
are derived from the third.
It is more plausible to say
that all three are 'equal',
(or, 'voiced oral consonants become voiceless oral and to derive them from a
consonants before a word boundary'). Several such single 'underlying form'.
notational conventions have been devised in order One such representation
would be [iN], where
to cope with all the types of phonetic relationship
stands for a nasal feature.
that have been observed.
This solution seems rea-
There are many kinds of phonological rule. Some sonable, as 'N' is clearly re-
include such forms as
irregular and illiberal?
Again, the prefix means
'not'; and the differences
seem to result from the fol-
lowing sounds. Should we
therefore group [1] and [r]
along with [m], [n], and [n],
and have a single rule for
all five possibilities?
If we do, we must set up
an underlying form from
which all can plausibly be
derived. [iN] no longer
seems appropriate, as two
of the sounds are not
nasal. [iC] (where 'C
stands for consonant')
would be too general, as
not all consonants are used
as part of the set of nega-
tive prefixes. Some inter-
in mediate category needs to
be devised, which is suffic-
iently abstract to enable all
the sounds to be grouped
together, yet sufficiently
concrete (that is, phoneti-
cally real) to provide a
meaningful explanation
'N' about what is taking place.
It would be possible to in-
vent a category [X] (where
= [m, n, rj, r]), but this
'X' 1,
vious how this category
would be useful in describ-
ing other areas of the lan-
guage.
Problems of this kind
have attracted a great deal
of discussion in phono-
logical theory in recent
years. There is much dis-
agreement about the ex-
tent to which phonological
analyses of this kind do or
should express psycho-
logical reality - that is, re-
present the native
speaker's intuitions about
the way the sound system
works (p. 409). And the de-
gree of abstractness that
should be allowed into an
analysis is especially con-
troversial. Some ap-
proaches permit the use of
symbols in the underlying
representations that have
no phonetic reality at all.
Other (so-called 'natural')
approaches require that all
symbols introduced into an
analysis bear a clear rela-
tionship to the physically
real processes of articula-
tion.

rules, such as the above, change the distinctive fea- lated to the three pronun- seems an arbitrary solu-
tures of segments. A further example, from the

28 THE LINGUISTIC USE OF SOUND •

163
Syllables of a sound), making the notion difficult to define Possible syllables
objectively. Also, prominence theory does not The number of possible syl-
The syllable is of considerable relevance to the task always give a clear indication of where the bound- lables combinations of
(i.e.
of phonetic and phonological description. It is a ary between syllables falls. In such words as master, different consonants and
notion that people intuitively recognize ('Shall I should the syllable division be master, mas-ter, vowels) varies greatly from
put it in words of one syllable?') and there are language to language. Totals
or mast-er} We are left with this problem, even
from the upsid survey
several writing systems in which each syllable is though in each case the relative sonority of the (p. 165) include:
represented by a symbol (p. 201). But it is by no sounds is the same.
means easy to define what syllables are or to iden- Hawaiian 162
tify them consistently. Do such words as fire, meal,
A phonological approach Rotokas 350
and schism have one syllable or two? Do meteor Phonological views of the syllable focus on the way Yoruba 582
Tsou 968
and neonate have two syllables or three? sounds combine in a language to produce typical
Ga 2,331
A syllable is a unit that is larger than a single sequences. Two classes of sound are established: Cantonese 3,456
segment and smaller than a word. However, this sounds that can occur on their own, or are at the Quechua 4,068
characterization can be seen from both a phonetic centre of a sequence of sounds {vowels (V)); and Vietnamese 14,430
Thai 23,638
and a phonological point of view. In phonetics, those that cannot occur on their own, or are at
some have attempted to identify syllables on the the edge of a sequence {consonants (C)) (p. 152).
basis of theamount of articulatory effort needed Typical sequences include CV see, CVC hat, CCVC Juncture
to produce them. The psychologist R. H. Stetson stop, etc. In this way the range of syllable types
Phonetic boundaries used to
(1892-1950) was one who argued that each syll- used in a language can be identified and different
demarcate words or other
able corresponds to an increase in air pressure, air languages compared. For example, some languages grammatical units are known
from the lungs being released as a series of chest use only V or CV syllables (e.g. Hawaiian); others as junctures. There are sev-
pulses — the pulse or motor theory of 7
syllable pro- use several consonants before and after the vowel eral phrases in English that
are distinguishable in this
duction. These pulses can often be readily felt and (e.g. English can have as many as three before and
way:
measured, especially when people speak emphati- four after — CCCVCCCC, as in some pronuncia-
cally. The main objection to the theory is that the tions of strengths). that stuff vs that's tough
pulses are sometimes very difficult to detect — for The syllable, in this view, takes its place as an an aim a name
example, in adjacent syllables when two vowels important abstract unit in explaining the way vow- I scream ice cream
nitrate night rate
co-occur (as in the word doing, which is two syll- els and consonants are organized within a sound
ables, but usually spoken with a single muscular system. There is, moreover, empirical evidence for In the first case, for example,
effort). the psychological reality of syllables, from the the [s] of stuff is stronger; and
The study of speech errors and related phenomena. In the of tough is aspirated.
Otto Jespersen (1860-1943) pre-
linguist ft]

It is not always easy to hear

sented an alternative phonetic approach, known 'slips of the tongue', for example, the kinds of sub-
the differences when the
as the prominence theory. This defines the syllable stitutions generally display the influence of syllabic phrases are said side-by-
in auditory terms, arguing that some sounds (vow- structure: initial consonants tend to replace each side; but the acoustic
els) are intrinsically more sonorous than others (p. other, as do final consonants. Thus one study changes can be readily ob-
served in a spectrogram
134), and that each peak of sonority corresponds reports many reversals of the types 'feak and wee-
(P- 136).
to the centre of a syllable. The problem with this ble' (for weak and feeble) or 'tof shelp' (for top
view is that other factors than sonority enter into shelf), but there are few reversals that mix up places
the definition of prominence (such as the pitch level in syllable structure (p. 262).

Connected speech
When words combine into lection from English, the sonants. In English, alveolar pronunciation (p. 39). In this fluenced by a following
connected speech, several pronunciations on the left consonants are commonly accent, the final ris not sound, e.g. ten bikes being
things can happen to the are heard when the words lost, especially at the ends sounded in such words as pronounced as tern baiks/.
pronunciation of their indivi- are said in isolation, or with of words, e.g. the final al- four and father, when they
dual segments. The speed emphasis; on the veolar plosive would nor-
progressive, which a in
right, are pronounced in isolation,
and rhythm can cause some when they are said nor-
sound is influenced by a
in mally be dropped in such or at the end of a sentence;
segments to adopt a weaker mal conversation. preceding sound, e.g. lunch
phrases as next day, but when followed by words
some to drop
articulation, score becomes, UnfiTo:/.
mashed potatoes, stopped that begin with a vowel, a
some to be put in, and
out, a /ei/ hi speaking, or got to go linking /r" is regularly used, coalescent (or reciprocal),
some to change character and /and/ /and, an, (='gottago'). The initial as in four o'clock or father inwhich there is mutual in-
altogether. n/ weak vowel may elide in and mother. fluence, or fusion', e.g.
could /kud/ /kad, kd/ such phrases as go away don you becomes
't

Strong and weak had /had/ /ad, d/ and try again. dauntfu/.
Words sometimes have him /him/ Assimilation
both strong ('accented') and
weak ('unaccented') forms,
IS

not
M
/not/
/im/
/s,z/
/nt, n/
Liaison
A sound may be introduced
In connected speech, adjac-
ent sounds frequently in-
These effects partly illus-
trate the role of phonetic co-

depending on whether they
are pronounced with force.
Words that express gram-
matical relationships
in
language are
particularly af-
fected. In the following se-
fluence each other so that articulation (p. 156), but
between words. Liaison is a
they become more alike, or they are also partly phono-
Elision notable feature of French,
rapid speech,
assimilate. There are three logical in character, as the
In sounds e.g. the final t of c'est is pro-
a may be main kinds of assimilation: rules differ from language to
left out, or elided, nounced when followed by
language.
especially when
they occur a vowel. It can also be illus- regressive (or anticipatory),
as part of a cluster of con- trated from English received in which a sound is in-

164 • IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

Comparative phonology Dependencies Favourite
Several important dependencies can be observed consonants
Given that the human vocal tract is capable of arti- What would a language look
between the sounds that are used in languages.
culating such a wide range of sounds (§27), several like, if it included only the
These take the form of 'implicationaP statements,
questions naturally arise. Which sounds turn up most common consonant
of the type: 'If X occurs, then Y will occur.' For segments? The 20 most fre-
most frequently in the languages of the world? Are
example, there are only four exceptions in the quent consonants were ex-
there any sounds that occur in all languages? What
upsid sample to the statement that if a language tracted from the upsid file, to
patterns of sound can be found in different lan- display the following system
contains /p/, it will also contain /k/. There is only
guages, and are there any similarities between the (alveolar and dental phones
one exception (Hawaiian) to the statement that if are grouped together):
patterns that occur?
/k/ occurs, then /t/ will occur. Similarly, if there
Questions about language universals and ten- P,b t. d k, g ?
is /g/, there will be /d/; if /d/, then /b/; and if
dencies (§14) cannot be answered in an impression- f s
/m/, then/n/. m n
istic way, nor even by comparing the language
More generally, nasals do not occur unless stops w 1
studies of several authors, whose methodology is
occur at the same place of articulation (five excep-
likely to differ. Answers require a systematic survey Most languages have be-
tions); voiceless nasals and approximants (p. 157) tween 1 4 and 1 6 of these
of a representative sample of languages, in which
do not occur unless the language has their voiced segments. No language has
the same analytic methods are used in each case,
counterparts; and mid-vowels do not occur unless exactly this system, but
and which is sufficiently large to enable some statis- some are very close to e.g.
there are high and low vowels (two exceptions). it,

tical conclusions to be drawn. The findings pre- Bambara (Niger-Congo),

sented in the following pages are based on an which lacks p], and includes
Areal statements
[z] and [<fe].
American survey known as upsid (The University
The upsid survey selects single languages from the The upsid survey shows
of California, Los Angeles Phonological Segment
main language families. There is also a need for the typical range of conso-
inventory Database). The inventories of 317 lan- nant segments to be be-
detailed phonological studies of all the languages
guages were included, with one language being tween five and 1 1 stops, one
spoken within a geographical area, to determine and four fricatives, two and
selected from each family grouping recognized (e.g.
the nature of any preferences for certain types of four nasals, and four others.
one from West Germanic, one from East Germanic,
sound. Such areal studies (p. 33) would draw atten- No one segment is found in
and so on (§50)). The segments were analysed as all languages. (After Mad- I.
tion to such features as the prevalence of click con-
phonemes (p. 160), each unit being represented by dieson, 1984.)
sonants in South Africa (and also in certain East
its most characteristic variant. (After I. Maddieson,
African languages), pharyngeals and glottals in
1984.) Why did click sounds spread
Afro-Asiatic languages, retroflex consonants in from the Khoisan languages
South Asia, or implosives and labio-velar coarticu- into other parts of South and
lation (p. 156) in African languages. Historical East Africa? One theory is
evidence is sometimes available to explain the that Zulu and Xhosa women
borrowed the clicks so as to
Number of segments development of an areal phonological feature, but
disguise words that would be
It is not yet known whether
there is an upper limit all too often the reasons are lost. taboo in their own lan-
on the number of segments that can be efficiently guages.
distinguished in speech, or a lower limit set by the
smallest number of segments needed to build up
a vocabulary. The smallest inventories in the upsid
sample contained only 11 segments: Rotokas
(Indo-Pacific) and Mura (Chibchan). Several Poly-
nesian languages are known to have very small
inventories. By contrast, the largest inventory
belonged to !Xu (Khoisan), with 141 segments,
with several other languages of this family display-
ing comparably large totals. Between these
extremes, 70% of the languages in the sample had
between 20 and 37 segments.
When the inventories are analysed into types of
sound, consonants emerge as being far more com-
mon than vowels. The number of consonants (C)
in an inventory varies between 6 and 95 (a mean
of 22.8); the number of vowels (V) varies between
3 and 46 (a mean of 8.7). If we divide V by C,
the resulting ratio varies between 0.065 and 1.308.
It is possible to say that the 'typical' language has

over twice as many Cs as Vs. Larger inventories

tend to have a higher proportion of Cs. However,
several languages do not conform to these trends,
such as Haida (Amerindian), with 46C but only
3V, and Pawaian (Indo-Pacific), which actually has
more V (12) than C (10).

28 THE LINGUISTIC USE OF SOUND •

165
STOPS a fricative (as opposed to a kind of breathy vowel), Segment frequency
All languages in the upsid survey have
stop conso- is found in 63% of the languages.
within a language
nants (p. 157), with voiceless segments occurring The found in Kabardian
largest set of fricatives is
Insouthern British English,
much more commonly than voiced (92% vs 67%). (Caucasian), where there are 22 in all, grouped into
an analysis of the frequency
Other types of stop are much less common, such eight types: of vowels and consonants in

as aspirated (29%), voiceless ejectives (16%), and conversation produced the

voiced implosives (11%). Most languages have two 7 voiceless non-sibiliant 7 voiced non-sibilant following totals (after D. B.
Fry, 1947).
types of stop, but the number varies between one 2 voiceless sibilant 2 voiced sibilant
and six. Languages with very complex sets of stops 1 voiceless non-sibilant 1 voiceless sibilant
Consonants Vowels
include Igbo (Niger-Congo) and !Xu (Khoisan), ejective ejective
each with six types. The Igbo inventory, for exam- 1 voiceless lateral 1 voiced lateral % %
ple, is as follows:
7.58 9 10.74
5 voiceless unaspirated 5 voiceless aspirated NASALS 6.42 I 8.33
plosives plosives Almost all upsid languages (97%) have at least 5.14 e 2.97
5 voiced plosives 5 breathy voiced one phoneme whose main allophone is a voiced
4.81 ai 1.83
2 voiceless implosives plosives nasal, and this is usually /n/ (in 96% of cases).
3.66 A 1.75
1 voiced implosive If there is a second nasal, it will usually be /m/.
3.56 ei 1.71
Languages with two, three, or four nasals are com-
Similarly, most languages have stops at three or
mon; the maximum seems to be six. Only four lan- 3.51 i: 1.65
four places of articulation (excluding glottal stops). guages in the whole sample have no nasal segments 3.22 30 1.51
Over 99% have bilabial, dental/alveolar, and velar at all (such as Rotokas (Indo-Pacific)). 3.09 a 1.45
stops. A
few have only two places of articulation The majority of nasal consonants are voiced: 2.81 D 1.37
(e.g. Hawaiian). Some (mainly Australian lan-
93%. Fewer than 4% are voiceless. The most com- 2.46 X 1.24
guages) have as many as six, with stops in bilabial, mon nasal segments are dental/ alveolar, followed 2.00 u: 1.13
dental, alveolar, retroflex, palatal, and velar posi-
by bilabial, velar, and palatal. 1.97 u 0.86
tions.
1.79 a: 0.79

FRICATIVES
LIQUIDS AND APPROXIMANTS 1.78 ao 0.61
The upsid analysis distinguishes between 'liquid'
At least one fricative (excluding /h/) is found in 1.46 3: 0.52
sounds (/1/and/r/) and 'approximant' sounds {/'}/
93% of the upsid languages; most of the cases
and/w/) (p. 156). Most languages (96%) have at 1.15 C3 0.34
where fricatives are absent are Australian. As can 1.05 0.21
least one liquid; 72% have more than one. /I/ seg-
13

be seen from the following graph, the majority of ments are somewhat more common than /r/ seg- 0.96 31 0.14

ments. Irish Gaelic has the largest number of 0.88 U3 0.06

liquids: 10 (2 voiced flaps, 2 voiceless flaps, 4 0.60

voiced laterals, and 2 voiceless laterals). At the 0.41
other extreme, several languages have none, such 0.37
as Nootka (Amerindian). The majority of liquids 0.10
are voiced (83%); 87% of them are dental/alveo-
lar. The most common /r/ segments are also voiced

(97%), and involve rapid tongue tip movements

taps, and flaps - 86%). Uvular [r], found
(trills,

inFrench and German, is not a common segment.

The approximants are also widely used. A /j/
segment is found in 86% of the languages; a /w/
segment in 76%.

GLOTTALICS
Ejectives are the most common consonant to use
a glottalic air stream (pp. 126—7). They are typi-
cally voiceless (99%) and are commonly stops
(60%). Two-thirds of all ejectives are found in
4 5 6 7 8 9 12* Amerindian languages, especially from North
Number of fricatives America. In 100% of cases, if a language has a
single ejective, it is /k'/. Some languages have as
languages have up to four fricatives, but some have many as five ejective consonants, e.g. bilabial, den-
12 or more. and uvular.
tal/alveolar, palatal, velar,
The most frequent fricative is a dental/alveolar The majority
of implosives are found in African
sibilant: 83% of the languages have some form languages. These are typically voiced (97%). If a
of /s/. Next comes /// and /f/, then /z/, /x/, /v/, language has a single implosive, it is usually /6/.
and/3/, m
that order. The asymmetry between /s/ Some languages have as many as four such- seg-
and/z/ is worth noting: the latter is found in only ments: bilabial, dental/alveolar, palatal, and velar
a third as many languages, /h/, when analysed as or uvular.

166 • IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

VOWELS 3-vowel systems
The 2,549 vowel segments in the upsid data can 1 u e o
a a
be classified on the basis of place and manner of
articulation as follows: Aranda (Australian) Amuesha (Andean-
front
Greenlandic Equatorial)
central (Eskimo-Aleut)
back
55 4- vowel systems
10
417 i » u
a e o
a
1 00 19
8 448 Rukai (Austronesian) Klamath (Penutian)

5-vowel systems
81 392 IJ i u i ui
lovv^
fi 1 36 e o e a
a a
Italic figures indicate unrounded vowels.
Bold figures indicate rounded vowels. Spanish (Indo-European) Japanese (Isolate)

It can be seen that front vowels are usually o i u

o o
unrounded (94%), and back vowels are usually
ae
rounded (93.5%). Low vowels are usually central a
(75%), and central vowels are usually low (69%).
Hebrew (Afro-Asiatic) Taishan (Sino-Tibetan)
High front vowels are much more common than
high back vowels. 6-vowel systems
The smallest vowel systems turn out to have i i u i u
e o o
three members (fewer than 6% of upsid lan-
e a
a a
guages). Some languages have been analysed as
having fewer than this (such as Kabardian (Cauca- Lappish (Uralic) Kanakuru (Afro-Asiatic)

sian)), but the analysis depends on how much of i u

the phonetic contrasts observed can be attributed e o e o
to the consonant system. There seem to be no clear 3 as a
a
cases of 1-vowel languages. By contrast, the largest
number of vowel segments is 24 (!Xu (Khoisan)). Romanian (Indo-European) Chamorro (Austronesian)
Most languages have between 5 and 7 vowels — 7-vowel systems
a point that can cause some surprise to speakers i u i u
of Indo-European languages, which have many e o i O
more. German and Norwegian both have 15 e 3 e o
a a
vowel-quality contrasts (disregarding length) - the
largest totals in the survey. Katcha (Niger-Congo) Kunama (Nilo-Saharan)
The more vowel qualities there are in a language,
8-vowel systems
the more likely that language is to show length iy uui 1 u
contrasts - though in fact only 20% of the lan- o e o
guages have both long and short vowel segments. e ce 3

Similarly, only 22% of the languages contrast oral a e 3

a
and nasal vowels.
There are only 83 clear cases of diphthongal pho- Osmanli (Ural-Altaic) Javanese (Austronesian)
nemes in the whole upsid sample, found in only 9-vowei system 12-vowel system
23 languages. Over a quarter of these occur in just i u i: u:
one language, !Xu, which has four series of diph- i u i u
thongs: oral, nasalized, pharyngealized oral, and e o 3:

pharyngealized nasal. e d e 3 3:
a ae ado:
Masai (Nilo-Saharan) English (Indo-European)
VOWEL SYSTEMS
Phonologists usually de- talk about vowel arrange- 14- vowel systems
scribe vowel systems with ments using an analogy with ii uu i: u: i u
reference to the articulatory basic geometrical shapes. ee oo e: o: ee
space they occupy, as repre- Most vowel systems are ee o5 e: 3: e 3

sented by such models as shape, es-

'triangular' in aa a: a
the Cardinal Vowel diagram pecially based on a 3- or 5-
Bambara (Niger-Congo) Wolof (Niger-Congo)
(p. 1 About 86% of the
54). vowel pattern. Fewer than
languages in the upsid sur- 1 0% of the languages have 20-vowel system
vey have their vowels evenly 'square' or 'rectangular' sys- i u 1 u
and widely distributed within tems. (Diphthongs are not i u I
this space (the principle of taken into account in the sys- e a o e 3 6
'vowel dispersion'), and it tems illustrated right.) ae a d ae a 6
thus becomes possible to Panjabi (Indo-European)

28 THE LINGUISTIC USE OF SOUND • 167

 From one extreme to the other
The remarkable differences between the phonological systems of the world's languages is nowhere better illustrated
than by a comparison of the smallest and largest consonant inventories in the upsid survey.

ROTOKAS
Consonants Bilabial Alveolar Velar

Voiceless plosive P t k
Voiced plosive 9
Voiced non-sibilant fricative P
Voiced tap D

!XU

x to <b
o o .N
TO
I
TO
It, 9
2 o ^ 2
Consonants
§ 5 .2 to to -I TO t> .TO
•O iS
v.

TO iS
Non-click 21 > a
TO
—I
<D
i

Voiceless plosive t k
Voiceless aspirated plosive t
h
kh
Voiced plosive d 9
Breathy voiced plosive g
Voiceless ejective stop t' k
g'
Voiced ejective stop d'

Voiceless sibilant affricate ts tf

h h
Voiceless aspirated sibilant affricate ts tf

Voiced sibilant affricate 4i 4%

Voiceless sibilant ejective affricate ts'

Breathy voiced sibilant affricate dz .03

Voiced sibilant ejective affricate dz'
Voiceless non-sibilant fricative
Voiced non-sibilant fricative
Voiceless sibilant fricative
Voiced sibilant fricative
Voiced nasal
Long voiced nasal
Breathy voiced nasal
Laryngealized voiced nasal
Voiced flap
Voiced central approximant

T3
9
? ?
TO "O
S.§ C N
TO
TO C C T3 2 c
C is 9 TO s
9 -5 -2 -~ <&
2 TO^ 2 § TO .N
S c
9
C TO 11 g ^* C
S? TO

Clicks Q TO
c Q TO SI Is I"
^ TO I
TO
Q. TO 21
Voiceless
Voiceless aspirated nc
Glottalized voiceless
Voiced
ge- w
gc ne g*
Breathy voiced nc
gc s
Glottalized voiced 96
Voiceless affricated V
Voiceless aspirated affricated oV
Glottalized voiceless affricated or n^ h
Voiced affricated 9? 0?
Breathy voiced affricated sf*' P?
Glottalized voiced affricated »"
gs
Voiceless lateral affricated C
A

Voiceless aspirated lateral affricated c

<h
oc"
Glottalized voiceless lateral affricated OC 06"
A
Voiced lateral affricated 9& QC 9C
Glottalized voiced lateral affricated 9i
Breathy voiced lateral affricated QC DC*

168 IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING

 29 Suprasegmentals

it ain't what you say, but the way that you say deliberation or emphasis. A rapid, clipped single Road hogs and
it.' This familiar comment, immortalized in song, syllable may convey irritation; a slowly drawled hot potatoes
isthe time-honoured way of briefly indicating what syllable, greater personal involvement. Compare: Why do people sometimes
'suprasegmentaF analysis is all about. The 'seg- introduce bursts of speed
'Shall Ileave now?' asked Janet. 'Yes,' snapped
ments' of spoken language are the vowels and into their speech? This ques-

consonants, which combine to produce syllables, John rudely. tion was addressed as part
a a
words, and sentences — the 'verbal' aspect of speech 'Shall I leave now?' asked Janet. 'Ye-e-s,' replied of linguistic analysis of
psychiatric interview. (The
(§§27-8). But at the same time as we articulate John, thoughtfully stroking his beard.
resultswere published in a
these segments, our pronunciation varies in other book called The First Five
and tempo together enter into
Pitch, loudness,
respects. We make use of a wide range of tones Minutes- so called, because
a language's expression of rhythm. Languages vary
of voice, which change the meaning of what we the microscopic nature of the
greatly in the way in which rhythmical contrasts analysis made it impractic-
say in a variety of different ways. It is these effects
are made. English makes use of stressed syllables able to publish more than
that provide the data of suprasegmental analysis. five minutes' worth of mater-
produced at roughly regular intervals of time (in
ial!)
fluent speech) and separated by unstressed syllables
The authors concluded
PROSODIC FEATURES - a 'stress-timed' (or isochronous) rhythm. In that there were six main con-
The basic psycho-acoustic properties of sound are French, the syllables are produced in a steady flow, ditions that led people to
the source of the main linguistic effects: pitch and resulting in a 'machine-gun' effect — a 'syllable- speak more quickly than
loudness (§§23, 25). These effects, along with those usual.
timed' rhythm. Loudness is the basis of rhythmical
arising out of the distinctive use of speed and
• Road hogs Speakers
effects in English (as shown by the way it is possible
think they are about to be in-
rhythm, are collectively known as the prosodic fea- to tap out a sentence in a 'te-tum, te-tum' way). terrupted, so they speed up
tures of language — a broader sense of prosody than By contrast, the length of a syllable (whether long in order to forestall it.

that found in the study of literature, where it refers or short) was the crucial feature of rhythm in Latin; • Hot potatoes Speakers
to the metrical patterns found in lines of poetry what they are
realize that
and pitch height (high vs low) is a central feature
saying unpleasant, so they
is
(§12). of the rhythm of many oriental languages. speed up to get it over with
The most important suprasegmental effects in a as quickly as possible.
language are provided by the linguistic use of pitch, PARALINGUISTIC FEATURES • Getaways Speakers rea-
or melody - the intonation system. Different levels Apart from the contrasts signalled by pitch, loud- lize that they have said
something unpleasant or
of pitch {tones) are used in particular sequences ness, tempo, and rhythm, languages make use of threatening, so they speed
{contours, or tunes) to express a wide range of several other distinctive vocal effects, using the up to put as much 'ground'
meanings. For example, all languages seem to make range of articulatory possibilities available in the as possible between them-
use of the difference between a falling and a rising vocal tract (§22). The laryngeal, pharyngeal, oral, selves and the distasteful
topic.
pitch pattern, and this is widely interpreted as and nasal cavities can all be used to produce 'tones • Smokescreens Speakers
expressing a contrast between 'stating' and 'ques- of voice' which alter the meaning of what is said. realize that what they have
tioning'. In English orthography, the contrast is These effects are sometimes referred to as effects said might be taken in the
signalled by the use of punctuation (p. 205), as of 'timbre' or 'voice quality', and studied under wrong way, so they speed
up, presenting new material
in They're waiting, vs They're waiting} In speech, the heading of vocal paralanguage — a term
that will capture the attention
a much wider range of tones is available to express intended to convey the less central role played by of their listeners.
various nuances and degrees of emphasis (cf. the these features in the communication of meaning, • Greener pastures
extra emotion suggested by They're waiting?}*). compared with that of prosodic features. Speakers, while speaking,
think of something more in-
Another important prosodic feature is loudness, One of the clearest examples of a paralinguistic
teresting to say, so they
which is used to convey gross differences of mean- feature is whispered speech, used in many lan-
speed up to get to the new
ing, such as the increased volume usually associated guages to add 'conspiratorial' meaning to what is topic as quickly as possible.
with anger, as well as the fine contrasts heard on said. Another is the marked lip-rounding which • Rebounds Speakers per-
the different syllables in a word (p. 164). Syllabic is widely used as a tone of voice when adults talk ceive that their speed of
speaking is inappropriately
loudness is usually referred to as stress, the syllables to babies or animals. But few of these effects are
slow, and in correcting this
being referred to as 'stressed' or 'unstressed'; but truly universal. For example, a 'breathy' or 'husky' they end up speaking more
the term accent is also often used ('accented' vs tone of voice conveys deep emotion or sexual desire quickly than they had in-
'unaccented'), referring to the way the prominence in many languages; but in Japanese, it is routinely tended.
(After R. E. Pittenger, C. F.
of a syllable is frequently due to the use of pitch used as a way of conveying respect or submission.
Hockett & J. J. Danehy,
as well as loudness (p. 171). A 'creaky' or 'gravelly' tone of voice is often used 1960.)
Variations in tempo provide a third supraseg- in English to convey unimportance or disparage-
mental parameter. It is possible to speed up or slow ment; but in Finnish, it is a normal feature of many
down the rate at which syllables, words, and sen- voice qualities, and would not have this conno-
tences are produced, to convey several kinds of tation. And there is no equivalent in English to
meaning. In many languages, a sentence spoken the use of strongly nasalized speech to convey a
with extra speed conveys urgency; slower speed, range of emotional nuances in Portuguese.

29 SUPRASEGMENTALS 169

Transcribing intonation
The first scientific attempt to transcribe the patterns of
English intonation was made by Joshua Steele in An
essay towards establishing the melody and measure
of speech in 1 775. It had been prompted by remarks
made in an essay by James Burnet the previous year.
Burnet had claimed that there was no such thing as
intonation in English. The music of our language, he
argued, is 'nothing better than the music of a drum,
in which we perceive no difference except that of
louder or softer'.
Steele rebutted this point of view in an original
manner. He pasted a piece of paper to the finger-board
of a bass viol next to the fourth string, and marked
the notes on it that corresponded to the various frets.
He then imitated the inflection of the voice by sliding
his finger up and down the string, and found that the
beginnings and ends of these inflections could usually
be located in the intervals between the frets. He then
devised a transcription which represented these
observations - the first systematic transcription of
English intonation.
In Steele's transcription of a line of poetry, as spoken
by himself, the accent marks are to be interpreted as
musical notes. The vertical symbols mark length, on
a scale from longest to shortest (
= ;
" ). Three
degrees of emphasis are indicated by the symbols a
.-. and , (from heavy to light). Pauses are also
. .
marked, using a system of rests, as in music. The
symbols underneath the text indicate an increased
level of loudness.
Largo.
MODERN TRANSCRIPTIONS
Contemporary transcriptions of intonation vary
greatly, as they reflect different theoretical views
of the nature of the subject. Some approaches
attempt to provide a faithful phonetic record of
melodic movement; others are more phonological
in character (§28), including only those aspects of
melody which seem to be crucial for expressing
contrasts in meaning. Some phonetic studies rely
on auditory judgments alone; others use a combi-
nation of auditory and acoustic analysis. Within
the phonological studies, there is difference of opi-
nion over the extent to which pitch contrasts are
capable of being analysed using the procedures
of phonemic analysis, and over the extent to
which grammatical and semantic considerations
(§§16—17) should be allowed to influence the
nature of a transcription. As a consequence, several
competing descriptive frameworks are in present-
day use.
, ,
,
Phonetic approaches
, Right: The widely used interlinear system The lines
represent the upper and lower limits of pitch range. Size ' '
of dot indicates relative loudness of each syllable, and the
' '«i
dot with a tail marks the most prominent syllable. The
method has been nicknamed the 'tadpole' transcription. Quando ci rivedremo?
The language in this example is Italian.

fe^^^
"

r^n i
t.« ?• ? r T

Let The typography in these

AA AA A A A .. me h a «ook
att cases provides an accurate
To or not to be that the queftion. Steth
be! ? is
ha, o, reflection of the phonetic

.L->in.i]. i >rW ^ o pe. changes in the melody.

(From D. L Bolinger, 1964,
pp. 282ff, and, for the ver-

whether 'tis nobler in the mind to fuffer the

35. (i)'AM,.
1
Hilda SAID
IDD NA
PREGN „ N T?' sion with prominence re-
presented by capitals, R. P.
Stockwell, 1972, p. 107.)
*? Y Y- T- - T _Y 9 Y Y- Y- i ?• .T

y JiL> ><i>\i> ^l-fe cca 86-90

m >
i

J
\m
^^
A .' A .'. A.. .'• A .*. A A.°. A..
&=£»
ftings and arrows of outrageous
t
fortune, or
A/WW\
to
F ieJ
»p p p
FE£±
hU\ ih mh]U Veg-re meg - 6r - kez-teklMar husz-6
—^-^^^^
- ve

-
nem I3t-tuk

1
egy-mastl

, 1

la djien dy J»n Jan

Qu' est on maman?
take arms againft aflail of troubles, and by op- - ce qu' a pc mr le diner.
4 3 2 2 2
Three ways of using musical notation. The first description (top) follows the musical
model very closely. The language is Hungarian. (From F6nagy & K. Magdics, 1 963.) I.

an The second (bottom left) shows four pitch levels more impressionistically drawn against
^ Ml an unspecified musical stave, the duration of each tone being represented by the length
of the marks. The numerals refer to toneme levels (p. 1 72). The language is Chinese.
(From N-C Chang, 1958.) The third (bottom right) gives a very approximate pitch
pofing, end them? — movement against an unspecified stave. The language is French. (From P. Delattre,
WVWVWWWWWA
1966.)

Above: The first few

,a'-i 400
lines of Steele's trans-
cription
$v-z Z Z Z Z II
300

Jf^^k^l^ZV JL_8 =::^:^? :

*-*3i. *
^ y
5J.200
150
£*: ::::*::::::: \f 1

g£
U£_ . .... .
»» .1 100
Right: Steele's dia-
Ejr- e nd a g smaeg a r s,s p a s erarpok arl'iohan sof
gram showing how 20 1 40 |
60 I 80 10Q 120 | 140 160| 180 200
the bass viol should
be used The information derived from a spectrographic analysis (p. 136). The language is

Norwegian. (From E. Haugen & M. Joos, 1952, p. 47.)

170 •
IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING
 Grammatical Intonation plays an important role Some ways of
Phonological systems in the marking of grammatical contrasts. The iden-
saying no in
'Come ,on. 'Let's get ,going. tification of such major units as clause and sentence
||
southern British
One of the widely used tonetic' systems of (§16) often depends on the way pitch contours
English
transcription. The accents represent pitch directions, break up an utterance; and several specific con-
placed before the word to which they apply. The double trasts, such as question and statement, or positive
line indicates the boundary between the two intonation
and negative, may rely on intonation. Many lan-
units; the raised vertical linemarks a stressed syllable.
(After J. D. O'Connor & G. F. Arnold, 1973.)
guages make the important conversational distinc- The most neutral tone; a sim-
tion between 'asking' and 'telling' in this way, e.g. ple statement of fact; de-
the |
man is 'eating a f bowl of PORhdge |
She's here, isn't she} (where a rising pitch is the tached.

In thissystem, a range of phonological symbols is used spoken equivalent of the question mark) vs She's
within the line of print. The most prominent syllable is here, isn't she\ (where a falling pitch expresses the •\
printed in small capitals, with an accent representing exclamation mark).
the direction of pitch movement above the vowel. The Emotionally involved; the
remaining symbols represent other features of pitch higher the falling tone, the
and stress, placed before the syllable to which they Information structure Intonation conveys a great
more involved the speaker.
apply. Note that sentence-initial capitals and deal about what is new and what is already known The choice of emotion (e.g.
punctuation marks are not used. (After D. Crystal, in the meaning of an utterance — what is referred surprise, excitement, irri-
1969.) to as the 'information structure' of the utterance. tation) depends largely on
context and facial expres-
If someone says 7 saw a BLUE car, with maximum
tul p iimaVih ] /he n sion.
intonational prominence on blue, this pronuncia-
tion presupposes that someone has previously •\
Pitch The movement of pitch is shown here by the queried the colour; whereas if the emphasis is on
direction of the solid and broken
Three main pitch
lines.
J, it presupposes a previous question about which A routine, uncommitted com-
The language
levels (high, mid, low) are represented.
person is involved. It would be very odd for ment; often used as a con-
is Kunimaipa (New Guinea). (From A. Pence, 1 964.)
versational 'noise' while
someone to ask Who saw a blue car}, and for the
someone else is talking.
//4 well they/ may do at/ A level// reply to be I saw a BLUE carl
Five main tones are recognized in this system, tone 4
(which is falling-rising in pitch) being shown here. The Textual Intonation not only used to mark the
is
main melodic contours are identified by//, and smaller structure of sentences; also an important ele-
it is
rhythm units by/. The main emphasis within a unit is Context and facial expres-
underlined. (After M. A. K. Halliday, 1 967.)
ment in the construction of larger stretches of dis- sion are important factors
course (§20). Prosodic coherence is well illustrated here. With a 'pleasant' face,
the tone is sympathetic and
The 'man in the 'street is in the way paragraphs of information are given
3- °2-3 3- °2-3 3- friendly, asking the speaker
a distinctive melodic shape in radio news-reading.
selling 'apples 'quickly.
to carry on; with an 'unplea-
As the news-reader moves from one item of news sant' face, it is guarded or
°2-3 °2-3 °2- -A
to the next, the pitch level jumps up, then gradually grim.
There are four pitch levels, numbered from 1 (high)
descends, until by the end of the item the voice
to 4 (low). The small circle marks the beginning of each
pitch movement. (After K. L Pike, 1945.)
reaches a relatively low level.

Psychological Intonation can help to organize Disbelief or shock - the ex-

language into units that are more easily perceived tent of the emotion depend-
ing on the width of the tone.
and memorized. Learning a long sequence of
THE FUNCTIONS OF INTONATION numbers, for example, proves easier if the sequence
Intonation, and the other suprasegmental features is divided into rhythmical 'chunks'. The ability to

of language, perform a variety of different func- organize speech into intonational units is also an Mild query or puzzlement; a
tions. important feature of normal language acquisition tone often used in echoing
— a feature that is often absent in cases of language what has just been said.

Emotional The most obvious function is to disorder (§§40, 46).

express a wide range of attitudinal meanings — ex-
citement, boredom, surprise, friendliness, reserve, Indexical Suprasegmental features also have a Bored, sarcastic, routine.
and many hundreds more. Here, intonation works significant function as markers of personal identity
along with other prosodic and paralinguistic fea-
tures to provide the basis of all kinds of vocal emo-
- an 'indexical' function. In particular, they help
to identify people as belonging to different social
V
Accompanied by a 'negative'
tional expression. groups and occupations (such as preachers, street
face, a tone of uncertainty,
vendors, army sergeants) (§§6—12).
doubt, or tentativeness; if a
'positive' face, a tone of en-
couragement or urgency.

^
A tone of emotional involve-
ment, expressing great em-
phasis. Depending on the
face and the context, so the
attitude might be impressed,
challenging, or complacent.

29 SUPRASEGMENTALS •
171
 Tone languages in an utterance is known as downdrift. It may even English stress
result in a high tone at the end of a sentence having In English words, each syl-
In well over half the languages of the world, it
the same absolute pitch level as a low tone at the lable is pronounced with a
is meaning of a word simply
possible to change the certain level of loudness, or
beginning!
by changing the pitch level at which it is spoken. stress. Usually, three levels

Languages that allow this are known as tone lan- Accent

of stress are recognized:
main stress;
primary, or
guages, and the distinctive pitch levels are known
Tone languages have to be distinguished from secondary stress; and un-
as tones or tonemes. stressed.
pitch-accent languages (e.g. Swedish, Japanese,
The number of distinctive tones in a language The main stress patterns
Serbo-Croat), in which a particular syllable in a
varies. The simplest systems have only two tones, can be seen in the following
word is pronounced with a certain tone, or 'accent'. list (' primary; secondary;
high vs low (e.g. Zulu); Yoruba has three (high, ,

For example, Japanese /sora/ 'sky' has a falling unstressed syllables un-
mid, low); Lushai has four (extra-high, high, mid,
accent on the first syllable, whereas /kawa/ 'river' marked):
low); Thai has five (low, mid, and high-falling,
has a rising accent on the second. A language may rwo syllables
high-rising, low-falling-rising); Cantonese Chinese
also contain minimal pairs that contrast only in 'finish 'female 'under
has six (mid- and low-level, high- and low-falling, be'hind Chi'nese ma'chine
word accent. In Swedish, the sentence Den bar tom-
and high- and low-rising).
ten dr bra means either 'This site is fine' or 'This three syllables
Tonal differences may affect vocabu- either the
understand .after'noon
goblin is fine', depending on the accentual pattern ,

lary or the grammar of a language. Probably the 'yesterday 'consequence

of totnten.
most widely known case of lexical contrast is Man- 'photo,graph im'portant
These languages must be distinguished from
darin Chinese, which has four tones, each of which four syllables
those where each word has a fixed place for the
has been given a 'tone letter', and also a number, re'markable .unim'portant
point of maximum prominence, though there is no 'caterpillar 'heli.copter
in systems of transcription.
restriction over which tone is used. 'Accent', in such
five syllables
cases, is synonymous with 'stress'. In Czech or Fin-
Tone Letter Example Meaning con.sider'ation
nish, for example, the main accent generally falls .satisfactory
high level |
ma' mother on the first syllable of a word; in Persian or Turkish, ad'mini.strative
on the last; in Polish or Welsh, on the penultimate six syllables
high-rising ma" hemp syllable. English and Russian are different again: de.sira'bility
low-falling-rising ^ ma J
horse in these languages, the accentual pattern of any .meteorological
given word is fixed, but there is no single pattern seven syllables
high-falling
\J ma" scold
used throughout the language. .unilateralism unreliability

eight syllables
Many tongue-twisters have been devised based on
.inter nationaliz'ation
this feature of the language, such as: Silence
An important feature of speech transcriptions is the
Compound words also need
Mama 1
qi ma 3 Ma 3 man. Mama ma 4 ma 3
.
1
. to be identified interms of
marking of pauses. Pauses are used to demarcate
'Mother rides horse. Horse slow. Mother scolds stress, for example:
linguistic units, to signal the cognitive activity of the
horse.' speaker, and to help structure speech interactions. 'wind, screen 'green, fly

Grammatical uses of tone are also common. They may be silent, or filled with a vocalization (such .broken-'hearted
In
as English ei\m)).
several languages of West Africa (e.g. Twi, Bini), Some interesting contrasts
Silence can also communicate a meaning in its own
a change of tone signals the difference between cer- occur when the stress pat-
right, as shown by the many descriptive phrases for
tain tense forms. In Bini, for instance, a low tone kinds of silence- 'threatening', 'thoughtful', and so on.
tern of compounds differs
from that of phrases (units
isused for present tense, and a high or high-low Here, silence is far more than the absence of speech.
consisting of separate
tone for past tense. Sometimes, pauses can be quite specific in intention
- such as the words). Compare:
didactic function used in teaching ('It's
The tones themselves are of two kinds: some stay called a — ?') or the social function of avoiding a taboo 'white 'house (any building)
at a single pitch level; others involve a change of (such as a host's 'Do you need to —
?'). Equally, 'White, House (the
pitch level ('gliding' tones). Tone languages are expectations of silence may be imposed by a social President's house)
usually classified into those that use gliding tones group (as in meetings), or have an institutionalized 'light 'housework (at home)
value, as in churches, libraries, and theatres. 'lighthouse ,work (at sea)
(contour tone languages) and those that do not
Cross-cultural differences are common over when to
(register tone languages). Thai and Mandarin Chi- Advertising slogan writers
talk and when to remain silent, or what a particular
nese illustrate the first type; Zulu and Hausa illus- (and linguists) enjoy playing
instance of silence means (p. 38). In response to the
question 'Will you marry me?', silence in English would
with these forms. One sign
trate the second.
be interpreted as uncertainty; in Japanese would be
outside a New York kiosk
When a sequence of tones is uttered, adjacent
it
read: Even hot dogs enjoy
acceptance. In Igbo, would be considered a denial
it
tones tend to influence each other in much the same our hot dogs' - a sentence
if the woman were to continue to stand there, and an

way as segments do (p. 164). Such assimilations any linguist would be proud
acceptance if she ran away.
of.
are known as tone sandhi. For example, a low tone
preceded by a high tone will usually begin with
a downward pitch movement. In particular, the
intonation system of the language can cause
changes in the pitch level of tones. In a sentence
where there is a gradually falling intonation con-
tour, the tones towards the beginning of the sen-
tence will be spoken at a higher level than the tones
towards the end. This gradual lowering of tones

172 •
IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING
 |

Speech and music tion are relative, not absolute. People are not
instruments. They do not speak out of tune.
Intonation has often been called the 'melody' or On the other hand, the evident similarities
'music' of speech, and musical notation has some-
between speech and music have led to several fruit-
times been used in the transcription of intonation ful developments in both subjects. Some linguists
(p. 170). But the analogy is not really a good one. have borrowed terminology from music in their
There are two main differences. Music is composed search for clear ways of describing suprasegmental
to be repeated; speech is not. And, if we examine effects; and certain composers and music analysts
modern western music, we find tones that have have, in turn, looked to suprasegmental studies for
been given absolute values, whereas those of speech ideas about the attributes and range of the voice.
are relative.
The consequences of this second point are far
reaching. Notes have fixed frequencies (e.g. middle
C now has a frequency of 264 Hz), and instruments Musics speech range are also noted, but (after I. Anhalt, 1972):
can be tuned to ensure that their notes are compat- several details of the • The virtuoso performance
There is no traditional termi-
ible. But speech is not like this. Men, women, and transcription have been of a live monody, part
nology in phonetics for des-
children use tones with the same linguistic function omitted.) speech, part song, with em-
cribing the many variations phasis on the prosody and
(for stating, asking, etc.), yet produce them at and tem-
of pitch, loudness, high paralanguage.
widely differing frequencies. Moreover, two people po which can be found in the 'book of the 'prophet • The use of a tape record-
connected speech. One se- isAiah| - 'thirtieth CHAPter| -
of the same sex may both use the 'same' rising ing of a person's voice, care-
quence of studies therefore
tone to ask a question, but one may produce it looked to music for its des-
diminuendo fully chosen in terms of lan-
guage, dialect, age, voice
with a higher frequency range than the other. And criptive terminology, and pro-
the 'fifteenth 'verse of the
CHAPter] - quality, and so on.
even within a single speaker, the pitch at which posed analogous categories
• Autopolyphony- several
crescendo
a tone is produced may vary from one moment for the analysis of the supra-
same person's
layers of the
segmental effects that can isAiah chapter 'thirty 'verse
to the next, without this affecting the meaning of
|
utterances superimposed
be heard on stretches of ut- f ifTEEN

what is said. Language is not affected by these bio- through multiple recordings.
terance. These included resonant • Polyphonies of small
logical or random variations. The tones of intona- such notions as: and rest -ye
piano/pianissimo
in re'turning
'shall be saved | — |
groups of voices, with struc-
tures saturated by certain
decreased levels of piano phonemic, prosodic, or para-
loudness in 'quietness - and in cONfi- linguistic features. Special
20 forte/fortissimo increased dence| - 'shall 'be your attention paid to the use of
is

©
levels of loudness STRENGTH - |
consonants - a feature with-
diminuendo gradually out precedent in western
allegro diminuendo
decreasing loudness music, where the emphasis
in re'turning and 'rest ye
crescendo gradually has traditionally been on
urqint ttnf miliar >n$ shall be saved in 'quietness
increasing loudness
|
vowels - and to non-periodic
and in cONfidencel - 'shall
allegro/ allegrissimo in- rhythms.
r> be 'your strength
Ml
|

creased tempo speech • Polyphonies of large

»» »*» +
«'" » » '«»
4^U of
lento/lentissimo decreased
(After D. Crystal & D. Davy,
groups, representing such
»
1969.)
tempo of speech situations as the cocktail
(torn* ) accelerando gradually party, mass responses, the
b« increasing tempo Speech —> music sound of a mob, and so on.
rallentando gradually de- • The complementation of
recent years several com-
In
creasing tempo vocal sound by percussive,
lit"//
posers, such as Luciano
electronic, and other effects.
The following extract from Berio(1925- ) and Karl-
For example, in Cento, by
the beginning of a sermon il- heinz Stockhausen (1 928-
r lustrates the way some of have been experimenting
),
Istvan Anhalt (1919- ),alive
choir of 1 2 voices is accom-
these effects combine to pro- with ways of displaying the
panied by vocal and instru-
duce the overall tone of voice voice in all its possible
mental sounds pre-recorded
Berio's Sequenza
III (1958) In this piece, lasting
associated with this variety. modes of expression, in as-
8'40", a woman
presents a sequence of disturbing
on tape. The text is a collage
Parts of the text correspond sociation with new techni- of fragments of a poem,
vocal behaviours, reflecting a range of bizarre moods. to certain prosodic effects, ques of composition involv-
The extract illustrates just a few of the composer's use some words being split and
as shown. (Certain other ing electronic devices.
other elements recombined.
of vocal labels, and clearly displays the combination prosodic effects of pitch Various procedures are used
of linguistic and musical notation that is the hallmark
of the work.
Throughout the piece, alongside periods of
sustained singing, we find such effects as: muttering,
breathy whispering, whimpering, humming, laughter,
and the use of narrow and wide pitch range; variations
Stylized tunes
Some kinds of intonation & n/Jjj levels are about a minor
in pause; a range of 'vocal tics' such as clicks, coughs, pattern have a very close third apart. This tune is also
and gasps; various trills and tremolos (e.g. striking the relationship to music - in common with warnings or
hand against the mouth, or using the hand over the particular the stylized are near absolute. reminders of a routine or
mouth as a mute); and there are several semantic tunes, or chants, used by Stereotyped tunes can cajoling kind, as in 'It's on
labels, such as tense, witty, joyful, excited, and frantic. street vendors, auctioneers also be heard in everyday the | ta J, ble' (i.e. 'I've told
Some of the notational conventions used in the work (p. 53),blues performers, speech, such as the sing- you a million times!'). Chil-
include; Sd. subdivision: =r tongue click;* brief sung trainconductors, and so song calling' intonation of dren, too, have their sty-
note; ^ cough;' o
sigh; { whine; .
*% gasp; \ finger on. These are often pro- 'Come and t get I it!' or lized tunes - notably, in
snap. duced with values which '

t Mi J,
chael!', where the cat-calling (above).

29 SUPRASEGMENTALS 173
30 Sound symbolism

It isan accepted principle of phonetic study that ner imitation (gitai-go) refers to feelings and fig-
individual sounds do not have meanings: it does urative expressions about objects and natural
not make sense to ask what [p] or [a] 'mean'. The surroundings, in which sounds play no part.
smallest units of language that are meaningful are The most popular forms are reduplications - pat-
the elements of grammar known as morphemes (p. terns of consonants and vowels that occur twice
90), such as un- and -ness. However, this ignores in immediate succession. These are used far more
the existence of an interesting (though limited) ten- than in English, which prefers simple forms to
dency in language to develop forms that speakers reduplications: compare such uncommon forms as
feel do have a close relationship to objects or states pitter-patter and ding-dong alongside the common
in the outside world. In such cases, individual bang, splash, and plop. In Japanese, reduplicated
sounds are thought to reflect, or symbolize, proper- forms occur normally in everyday conversation.
tiesof the world, and thus to 'have meaning'. Their range can be illustrated from the following
This phenomenon is known as sound symbolism, examples (after H. Kakehi, etal., 1981).
though in literary contexts it is more commonly
referred to as onomatopoeia (p. 74).
gion-go: gacha-gacha rattle
Sound-symbolic forms are usually considered to
tinkle
chirin-chirin
be features only of literary expression (of poetry,
kasa-kasa rustle
in particular) or of such restricted areas as child
language (bow-wow) or the language of comic
strips (Zap\, Pow\). In fact, several everyday lexical
gitai-go: tobo-tobo plod
items are onomatopoeic, even though the sound fura-fura roam
symbolism may not be immediately obvious. In kira-kira twinkle
English, for example, there are a number of items
beta-beta stick to

that can be given a phonetic classification in this

dabu-dabu baggy, loose
way. They can be grouped on the basis of their
initial consonants, vowels, or final consonants, as Not all onomatopoeias are reduplicative:
when words beginning with /si-/ are said to convey
unpleasant associations [slime, slither, slug, sloppy,
gisshiri packed full, crowded
etc.), or words containing high front vowels asso-
shikkari firmly, strongly
ciations of smallness (teeny weeny, wee, etc.). Final
consonants also present an interesting basis of clas-
sification. Examples include: meaning
Very often, more than one level of is

/-P/ lap, clip, rip
crack, creak, click, cluck, flick,
blob, glob, jab, rub
bubble, trickle, rustle
ooze, wheeze
1-51 smash, crash, crush, splash, slash, lash
l-V cough, woof
puff, gruff, biff,
But English, and Indo-European languages gener-
ally, are not good exemplars of the use of sound
symbolism. To see the importance of this feature,
we must look elsewhere. Korean, for instance, has
over a thousand words that are sound-symbolic
in character; and a correspondingly large onomato-
poeic vocabulary also occurs in Japanese.
JAPANESE ONOMATOPOEIA
Japanese has over three times as many onomato-
poeic expressions as English, and uses them to
express a more wide-ranging set of linguistic dis-
tinctions. Sound imitation (gion-go) is used to re-
flect physical, audible noises relating to the actions
or movements of people, animals, and things. Man-
expressed. For example, bara-bara refers to very
whack strong rain ('pelting down'); but it may also refer
to things that have been broken up, scattered, or
disorganized. It could be used in such contexts as:
'The family is split up', 'The queue is not straight',
or 'We left as a group, but came home separately.'
Goro-goro is used for sounds, such as the purring
of a cat, or for rumbling noises (such as thunder,
or heavy objects); but it is also used to express
manner, such as a state of discomfort caused by
a lump, the way in which things are strewn around
in abundance, or the state of being idle.
There are also several grammatical factors that
must be taken into account. For instance, the parti-
cle to isused to indicate that the preceding expres-
sion is ga wan-wan to naku 'The
a quotation: Inu
dog goes bow-wow' - something which in English
would be more naturally expressed by a single lexi-
cal item, 'The dog is barking.' Also, certain forms
are typically used in the expression of grammatical
meanings, such as the use of gun-gun (steadily,
rapidly) and don-don (at a great rate) in progressive
contexts (as in gun-gun ookiku-natte 'went on
growing and growing').

174 •
IV THE MEDIUM OF LANGUAGE: SPEAKING AND LISTENING
TYPES OF SYMBOLIC MEANING Nonsense verse
Several attempts have been made to find specific 'mother' 'father'
The semantic value of
correspondences between sounds and meanings. Dakota sounds is nowhere better
ena ate
For example, in several languages an association Nahuatl illustrated than in successful
naan ta'
nonsense verse, the most
has been suggested between close vowels (espe- Tiv ng ter famous example of which is
cially [i]) and smallness, and open vowels (espe-
Luo mama baba the first verse of Lewis
cially [a]) and largeness, as in English teeny, little, Hebrew ima, em aba, av Carroll's Jabberwocky.
bit, slim, -ling vs large, vast, grand, French petit
French mere, mama pere, papa Twas brillig, and the slithy
vs grand. On
the other hand, there are several Tamil ammaa appa toves
counter-examples to this tendency - most Yucatec Did gyre and gimble in the
nan tat
obviously, English big vs small. wabe:
Greenlandic anaana(q) ataataq
To be convincing, evidence of fundamental links Allmimsy were the
borogroves,
between sound and meaning needs to be provided Again, the pattern And the mome raths
is not universal. In Georgian,
from a large number of languages. The same pat- mama means 'father'; and in a number of South outgrabe.
tern must be found in languages of very different Asian languages (e.g. Tamil, Telugu), mama means Carroll also provided inter-
types and be confirmed by experiments into 'mother's brother'. pretations of some of the
speakers' intuitions. A certain amount of descrip- nonsense words, such as
work has been carried out; but corresponding slithy = 'lithe and slimy',
tive UNIVERSAL MEANINGS? mimsy = 'flimsy and miser-
psycholinguistic studies 412) have not. The
(p. There certainly limited evidence of a
is few broad able', mome = 'from home',
American linguist, Morris Swadesh (1909—67), sound/meaning correspondences and outgrabe = 'something
in language. But
brought together several such descriptive observa- there are many exceptions to the correlations that between bellowing and whist-
tions. For instance, he drew attention to the use have been proposed, and when individual features
ling, with a kind of sneeze in
the middle'.
of [i]-type vowels to express nearness (this) and are studied across a wide range of languages, a The poem has also been
[a]- or [u]-type vowels to express distance ('that/ variety of divergent meanings emerge. The use of translated. The extent to
you') in many languages: phonological reduplication within a word is wide- which the effects carry over
into foreign languages can
spread, but meanings vary greatly, as some of these
be judged from these ex-
'this'([i)) 'that/ you' ([a/ u]) examples show (M. Swadesh, 1972).
tracts:

Chinook -i- -u-

Klamath ke- ho-, ha
CONCLUSIONS Le Jaseroque
II brilgue: les toves
The examples of sound-symbolism are fascinating,
Tsimshian gwii- gwa lubricilleux

tuvicha but in the absence of frequency information about Se gyrent en dans

Guarani tye vrillant le
the phonological and lexical patterns in the various guave,
Maya li' la', lo'
languages, it is not possible to arrive at a definitive Enmimes sont les
Binga ti ta
gougebosqueux,
Fur in
interpretation. The cross-linguistic similarities may
ilia Et le momerade horsgrave.
indeed have evolved separately, indicating a basic
Didinga ici ica (F. LWarrin, 1931)
Tamil id* adi
human propensity to use certain sounds in certain
ways, or they may simply be the result of language Der Jammerwoch
Thai nii nan Es brillig war. Die schlichte
contact over a long period of time. In the absence
Burmese dii thoo Toven
of historical data, drawing conclusions from sound Wirrten und wimmelten in
symbolism about the origins of language would Waben;
Sometimes some quite specific correspondences be premature (§49). Far more descriptive data are Und aller-mumsige
have been noted, such as the tendency for languages needed, accompanied by experimental investiga- Burggoven
Die mohmen Rath'
to express 'mother' with a nasal, and 'father' with tion of the speakers' intuitions about the relation-
ausgraben.
an oral front consonant. ship between sounds and meanings. (R.Scott, 1872)

Meaning of reduplication Language Examples

Plural Bella Coola s-tn 'tree', s-tntn 'trees'

Hausa suna 'name', sunana-ki 'names'
Tsimshian am 'good', am'am 'several are good'
Repetition Karok pdchup 'kiss', pachupchup 'kiss a lot'
Intensity Karok go 'see', go-go 'look at carefully'
Scattered distribution Nootka mah'tii 'house', maamah'ti 'dispersed houses'
Space Somali fen 'gnaw at', fen-fen 'gnaw at on all sides'
Continuation Nahuatl kweyooni 'flashes once', kwe'kweyooka 'is flashing'
Smallness Nez Perce q'eyex 'club', q'eyexq'eyex 'small club'
Diminutiveness Sahaptin pshwa' 'rock', pswa'pswa 'pebble'
Past tense Greek leipo 'I leave', leloipa 'I have left'

Adjective marker Nez Perce sik'em 'horse', sik'eemsik'em 'mean'

30 SOUND SYMBOLISM •
175
 j»$
7)
/

k \

#8^6 1
ABCDE

PAI^R

ibtme

ABTAE ABBTfl ABCDE i

c
 BC
3100
3000
PARTY
2900
2800
2700
The medium of language: writing
2600
2500 and reading
2400
2300
2200
2100
2000
There is a curious ambiguity about the study of ferences between written and spoken language. A
1900
written language: certain aspects, such as the his- contrast drawn between the study of the physical
is
1800 tory of the alphabet, have been meticulously inves- properties of the graphic medium (graphetics) and
1700 tigated by generations of scholars; others, such as of the linguistic system which makes use of that
1600 the psychological processes underlying the tasks of medium (graphology). We explore the many facets
1500
reading, writing, and have attracted
spelling, of graphic expression, as found in handwriting,
serious study only in the last decade or so. From printing, and the electronic media. Several tradi-
1400
a scientific viewpoint, we know far less about the tions of study are described, such as palaeography,
1300
written language than we do about the spoken, epigraphy, and calligraphy. The writing systems of
1200 largely because of the 20th-century bias in linguis- the world are introduced through a historical per-
1100 tic studies towards the analysis of speech - a bias spective: we look at the precursors of writing, and
1000 which is only nowadays beginning to be corrected. then at pictographic, ideographic, logographic, syl-
Part v therefore has to be shorter than Part iv — labic, and alphabetic systems. Separate accounts
rhmi v
900
800
at least, for the present. are given of several individual systems, from
There are two underlying themes to this Part. ancient hieroglyphic to modern shorthand.
700
Firstit is emphasized that writing and speech are The variety of writing systems, and the historical
600 different and equal manifestations of language. emergence of the alphabet, raises many theoretical
500 Writing should not be seen as merely 'transcribed questions - not least the question of how the pro-
400 speech', because its formal characteristics, and its cess of reading and writing actually takes place
strategies of production and comprehension, are in the brain. We review the main psychological
300
quite unlike those encountered in speech. Secondly, models of reading that have been proposed, and
200
the notion of 'written language' is shown to be look briefly at the more poorly researched fields
100 extremely broad and multifaceted, subsuming any of writing and spelling. Part v then concludes with
kind of visual realization of language (such as a review of the issues involved in the many propo-
100 manuscript, typescript, and print). sals for spelling reform.

200 We begin by examining the similarities and dif-

Jf\A^cro\jj 300
400 A tiny sample of the world's alphabets over the centuries. In
the background is an engraving of a printing shop in the late
500 1 5th century by Philippe Galle, after a drawing by Jan
van
600 der Straet.

700
800
900

L 1000
1100
1200
1300
1400
1500
1600
1700
1800

5 28 1900
AD
3 1 Written and spoken language

WRITING SUPREME special skill, used only for sophisticated purposes Write - or speak?
The history of language study illustrates widely (as in scientific and by a minor-
literary expression) The functions of speech and
divergent attitudes concerning the relationship ity of communities. It was needed in order to have writing are usually said to
between writing and speech. For several centuries, access to the early history of a language (philology, complement each other. We
do not write to each other
the written language held a preeminent place. It §50), but this was felt to be a woefully inadequate
when we have the opportu-
was the medium of literature, and, thus, a source substitute for the study of the 'real' thing, speech.
nity to speak - apart from
of standards of linguistic excellence. It was felt to Writing, seen as a mere 'reflection' of spoken lan- such exceptional cases as
provide language with permanence and authority. guage, thus came to be excluded from the primary secretive children in class
The rules of grammar were, accordingly, illustrated subject matter of linguistic science. The pendulum and spouses who are 'not
talking'. Nor can we speak to
exclusively from written texts. swung to the opposite extreme in the new gener-
each other at a distance -
The everyday spoken language, by contrast, was ation of grammars, many of which presented an except in special cases in-
ignored or condemned as an object unworthy of account of speech alone. volving technical equipment.
study, demonstrating only lack of care and organi- On the other hand, there

zation. It was said to have no rules, and speakers COMPROMISE are many functional parallels
which ought not to be
were left under no illusion that, in order to 'speak It is understandable but regrettable that writing ignored, especially as these
properly', it was necessary to follow the 'correct' and speech should have been allowed to confront are on the increase in
norms, as laid down in the recognized grammar each other in this way. There is no sense in the modern society.
books and manuals of written style. Even pronun- view that one medium of communication is intrin- • The relative permanence
of written language makes
ciation could be made to follow the standard writ-
it
sically 'better' than the other. Whatever their his-
such func-
ideally suited for
ten form, as in recommendations to 'say your h's' torical relationship, the fact remains that modern tions as recording facts and
and not to 'drop your g's' (p. 32). The written lan- society makes available to its members two very communicating ideas (§4).
guage, in short, was the main plank on which the different systems of communication, each of which But these days, talking
prescriptive tradition rested (§1). has developed to fulfil a particular set of communi- books for the blind, libraries
ofrecorded sound, and other
cative needs, and now offers capabilities of expres-
ideas are providing alterna-
SPEECH SUPREME sion denied to the other. Writing cannot substitute tives.

There was sporadic criticism of this viewpoint for speech, nor speech for writing, without serious • Letters and messages for

disservice being done. The

study of speech distant contactsused only to
throughout the 19th century, but it was not until scientific
be written. Nowadays, they
the present century that an alternative approach in its own right is now a well-developed subject
can also be spoken - thanks
became widespread. This approach pointed out (Part iv). The analogous study of the written lan- to tape cassettes,telephone
that speech is many thousands of years older than guage is less advanced, but has just as promising answering machines, radio
writing (§49); that it develops naturally in children a future. phone-ins, and other such
developments.
(whereas writing has to be artificially taught); and
• The immediacy of speech
that writing systems are derivative - mostly based makes ideal for social or
it

on the sounds of speech. 'Writing is not language', 'phatic' functions (p. 10). But
insisted the American linguist, Leonard Bloomfield writing alsohas its phatic
The wheel turns full circle. Nowadays greetings cards are functions, and these seem to
(1887-1949), 'but merely a way of recording lan-
available that speak when you open them. be increasing, as suggested
guage by means of visible marks.' by the expanding business of
It was also argued that, as speech is the primary producing cards to mark spe-
medium of communication among all peoples, it cial occasions - birthdays,
should therefore be the primary object of linguistic Christmas, anniversaries,
examination results, passing
study. In the majority of the world's cultures, in
(or failing) a driving test, and
fact, there would be no choice in the matter, as many more.
the languages have never been written down. Early However, when comesit

linguistics and anthropology therefore stressed the to tasks of memory and

urgency of providing techniques for the analysis learning, speech is no substi-
tute for writing. Written re-
of spoken language - especially in cases where the
cords are easier to keep and
cultures were fast disappearing and languages were scan. Written tables and
dying out. 'When we think of writing as more figures readily demonstrate

important than speech,' wrote Robert Hall (191 1-) relationships between
things. Written notes and
in a popular paperback, Leave Your Language
lists provide an immediate
Alone (1950), 'we are putting the cart before the mnemonic. Written explana-
horse in every respect.' tions can be read often, at in-
Because of this emphasis on the spoken language, dividual speeds, until they
are understood.
was now the turn of writing to fall into disrepute.
it

Many linguists came to think of written language

as a tool of secondary importance - an optional,

178 V THE MEDIUM OF LANGUAGE: WRITING AND READING

 THE DIFFERENCES BETWEEN as they would when speaking. As a consequence,
Newspeak
WRITING AND SPEECH writing avoids words where the meaning relies on
The written language is
Writing and speech are now seen as alternative, the situation (deictic expressions, such as this one, amusingly used to point up
'equal' systems of linguistic expression, and over there, p. 106). Writers also have to anticipate the idiosyncrasies of a
research has begun to investigate the nature and the effects of the time-lag between production and speech style, in this 1963
extent of the differences between them. Most Guardian editorial:
reception, and the problems posed by having their
obviously, they contrast in physical form: speech language read and interpreted by many recipients The BBC has introduced a.
uses 'phonic substance', typically in the form of in a diversity of settings. In the absence of imme- New method of disseminat-
air-pressure movements (§23); writing uses diate feedback, available in most speech interac- ing the spoken word at any
'graphic substance', typically in the form of marks tion, care needs to be taken to minimize the effects ratewe think is new be- it

cause we don't. Remember

on a surface. But of far greater interest are the dif- of vagueness and ambiguity.
hearing it until a week or two
ferences in structure and function that follow from ago it consists of. Putting the
• Written language displays several unique fea-
this basic observation. fullstops in the middle of sen-
tures, such as punctuation, capitalization, spatial
These differences are much greater than people tences instead of at the end
organization, colour, and other graphic effects as we were. Taught at school
The contrast is greatest when written
usually think.
(§32). There is little in speech that corresponds, as a corollary to this new
texts are compared with informal conversation; sentences are run on without
apart from the occasional prosodic feature (§29):
but even in fairly formal and prepared speech set- a break readers will say we
for example, question marks may be expressed by
tings, such as a teacher addressing a class, the struc- are in. No position to talk but
rising intonation; exclamation marks or underlin- thisappears to be a deliber-
ture of the language that spoken bears very little
is
ing may increase loudness; and parentheses may ate policy on the part of the
similarity to that found in writing. It is something
lower tempo, loudness, and pitch. But the majority BBC whereas our. Misprints
that is immediately apparent if a stretch of speech are accidental.
of graphic features present a system of contrasts
is tape recorded and transcribed. Even a fluent The practice seems to
that has no spoken-language equivalent. As a
speaker produces utterances that do not read well have started as a. Means of
result, there are many genres of written language enlivening the reports of
when written down (p. 94).
whose structure cannot in any way be conveyed otherwise tedious football
The differences of structure and use between matches on a. Saturday
by reading aloud, such as timetables, graphs, and
spoken and language
written are inevitable, afternoon now it has spread
complex formulae.
because they are the product of radically different to the. News columns as it

kinds of communicative situation. Speech is time- • Grammatical and lexical differences are also were and the effect is to
make the subject matter.
bound, dynamic, transient - part of an interaction important. Some constructions may
be found only
Confusing the interest of the
in which, typically, both participants are present, in writing, as in the case of theFrench simple past listener is directed to the.
and the speaker has a specific addressee (or group tense (the passe simple). Certain items of vocabu- Manner of delivery rather
of addressees) in mind. Writing is space-bound, sta- lary are rarely or never spoken, such as many poly- than the. Events recounted
tic, permanent - the result of a situation in which, syllabic chemical terms, or the more arcane legal we tried to discover whether
the ellipses or hiatuses fol-
typically, the producer is distant from the recipient terms. Conversely, certain items of spoken voca-
lowed a. Definite pattern or
- and, often, may not even know who the recipient bulary are not normally written, such as whatcha- whether the breaks were
is (as with most literature). Writing can only occa- macallit (with no standard spelling), and certain made. Arbitrarily a pattern
sionally be thought of as an 'interaction', in the slang or obscene expressions. didemerge seems that
it

most of the breaks come

same way as speech (exceptions include personal
correspondence and, more important, the growing
field of computer-based interaction). It is therefore
not surprising to find differences emerging very
quickly when languages first come to be written
down, as has been observed in such cases as Basque
and Tok Pisin.
Points of contrast
• The permanence of writing allows repeated
reading and close analysis. It promotes the develop-
ment of careful organization and more compact,
intricately structured expression. Units of dis-
course, such as sentences and paragraphs, are
clearly identified through layout and punctuation.
By contrast, the spontaneity and rapidity of speech
minimizes the chance of complex preplanning, and
promotes features that assist speakers to 'think
standing up' - looser construction, repetition,
rephrasing, filler phrases (such as you know, you
see),and the use of intonation and pause to divide
utterances into manageable chunks (p. 52, §29).
• The participants in written interaction cannot
usually see each other, and they thus cannot rely
on the context to help make clear what they mean,
• Written language tends to be more formal than after the. Definite or indefi-
Prepo-
spoken language and is more likely to provide the nite article or after a.
sition sometimes they follow.
standard that society values. It also has a special
Verbs but they always come
status, mainly deriving from its permanence. Writ- when you. Least expect
ten formulations, such as contracts, are usually them and they constitute an
required to make agreements legally binding (p. outrage on what. We in the
trade call the. Genius of the
386). Sacred writings are used as part of the identity
language.
and authority of a religious tradition (p. 384).
Mutual influence
Despite these differences, there are many respects
in which the written language can influence the
spoken. Soon after learning to read, children use
the written medium as a means of extending their
spoken vocabulary - as indeed do many adults.
Some words may be known only in written form.
Loan words may come into a language through
the written medium. Sometimes the whole of a lan-
guage may be known only from writing (as with
Latin, or certain cases of foreign language learning,
§62). And an old written language can be the
source of a modern spoken one (as in Hebrew).
Writing systems may derive from speech, in a his-
torical sense, but in modern society the dependence
is mutual.

31 WRITTEN AND SPOKEN LANGUAGE- 179

 . .

Portraying the sound of speech PUNCTUATION Portraying the

The punctuation system can be altered or sound of silence
The differences between speech and writing are expanded, or certain features used with unexpected Graphic devices are often
most clearly displayed when people attempt to por- frequency. used to express reactions
tray the sound of the former using the graphic when no words are spoken
These two - they're twins, Sam 'n Eric. Which is Eric - ? at all.
properties of the latter. The most complex and You? No - you're Sam — 'We might go in your um-
ingenious ways of doing this are to be found in (William Golding, Lord of the Flies) brella,' said Pooh.
written literature, where authors are continually '?'

Tiens, Zazie . . regarde!! le Metro!!! (Here, Zazie . .

battling to put sounds into words.

.

'We might go in your um-

look!! the Metro!!!) brella,' said Pooh.
The graphic conventions authors use have (Raymond Queneau, Zazie dans le Metro) '? ?'
received little study, especially from a cross-linguis- 'We might go in your um-
tic viewpoint. Different languages do not display
Sometimes the author may go so far as to explain:
brella,' said Pooh.
the same range of written language conventions -Ohjj (ji c'est le point d'indignation) (Oh;; [\\ is the '!!!!!!'

for the portrayal of speech. In English, for example, indignation mark)) For suddenly Christopher
(Raymond Queneau, Le Chiendent) Robin saw that they might.
emphatic speech is not usually printed in a heavy
(A. A. Milne, Winnie the
typeface, but this is common in Chinese fiction. Pooh)
And the use of repeated letters (as in ye-e-es) can
have a range of interpretations, such as emphasis
SPELLING
The spelling can be altered, to convey the impres-
and hesitation, in different languages. This can lead
sion of regional accent, personality, or other effects
to ambiguity, especially when texts are translated.
(p. 77).
For example, a character in the English translation
of Alexander Solzhenitsyn's Cancer Ward is The family name depends wery much upon you,
recorded as saying 'No-o'. The use of this conven- Samivel, and I hope you'll do wot's right by it.
(Charles Dickens, Pickwick Papers)
tion in the original Russian would convey an
emphatic negative; but the English version is far Aw knaow you. Youre the one that took away maw
more likely to signal a hesitant one. Italics, likewise, girl. Youre the one that set er agen me. Well, I'm gowin
can be ambiguous, both within and between lan- to ev er aht.
(G. B. Shaw, Major Barbara)
guages, being used variously as a marker of foreign
words, technical terms, book titles, emphasis, and Oh, there you are steward. Ole man dlunk, bline dlunk.
several other effects. Purrimabed.
The way the written language conveys the effects (G. B. Shaw, Major Barbara)

of sound is now beginning to be studied more sys- An' they're always speshully savidge when they haven't
tematically. Some of the graphic effects in wide- any tusks.
spread use are illustrated below (after R. Chapman, (Richmal Crompton, William the Bad)

1984). A Germanic pronunciation of French is portrayed

in this extract:
VERBAL DESCRIPTION
Eh! pien, si ces tames feulent fus dennit gombagnie, dit
Probably the most common technique is to make
Nucingen, che fus laiserai sele, gar chai drop manche.
use of descriptive words and phrases. Some authors
(= Eh! bien, ci ses dames veulent vous tenir compagnie
take great pains to make their descriptions vivid,
. je vous laisserai seul, car j'ai trop mange. 'Well, if
. .

precise, and meaningful. these ladies want to keep you company, I will leave you
... a note of menace pierced through his voice. alone, because I have eaten too much.') (Honore de
(James Joyce, Dublmers) Balzac, Splendeurs et miseres des courtisanes)
e. e. cummings
'Undoubtedly, undoubtedly,' broke in Mr Verloc in a (1894-1962), whose in-
deep deferential bass of an oratorical quality genious typographic con-
. .

(Joseph Conrad, The Secret Agent)

CAPITALIZATION figurations frequently reflect
Varying the use of capital letters can convey loud- the distinctive rhythms and
His voice lifted into the whine of virtuous recrimination. ness, special significance, and several other effects. tones of voice that can be
(William Goldmg, Lord of the Flies)
heard in speech (p. 72).
'At such times as when your sister is on the Ram-page,
'Butmy darling,' he protested in the cajoling tone of Pip,'Joe sank his voice to a whisper and glanced at the
one who implores a child to behave reasonably.
door, 'candour compels fur to admit that she is a
(Aldous Huxley, Point Counter Point)
Buster.'
... a soft, greasy voice, made up of pretence, politeness Joe pronounced this word, as if it began with at least
and saliva. twelve capital B's.
(Anthony Trollope, Ralph the Heir) (Charles Dickens, Great Expectations)

Extended capitalization usually expresses loudness.

'MISS JEMIMA!' exclaimed Miss Pinkerton, in the

largest capitals.
(William Makepeace Thackeray, Vanity Fair)

Heedless of grammar, they all cried, 'THAT'S HIM!'

(R. H. Barham, The Jackdaw of Rheims)

180 •
V THE MEDIUM OF LANGUAGE: WRITING AND READING
 . .

TYPE SPACING AND SIZE

Variations in type size and spacing provide a wide Science fiction Frankly Canapes? Why
Ellery, Thanks delicious, yes,
range of possible effects (§32). The genre of science fiction I Mary, they're Tate,
often experiments with the don't I'm
Alice couldn't see who was
beyond the Beetle,
sitting written language, in its at- think treating

but a hoarse voice spoke next. 'Change engines - it ' tempts to characterize the We you'll Canapes? D'Courtney.
communicative habits of fu- brought be I
said, and there it choked and was obliged to leave off.
Galen working expect
it sounds like a horse,' Alice thought to herself. And ture times and alien beings.
along (or him
an extremely small voice, close to her ear, said,
In Alfred Bester's The De- to Monarch in
molished Man (1 953), a help him celebrate. much town
'You might make a joke on that - something about "horse" group of gifted telepathic He's longer, very
and "hoarse", you know.' people meet at a cocktail just The shortly.
(Lewis Carroll, Through the Looking Glass) party, and all begin to com- taken his Guild Exam.
municate at once. The re- If is and
Once on the bridge, every other feeling would have gone you're just been
sult is a pattern of inter-
down before the necessity - the necessity — for linked thoughts, so
interested about classed
Powell, we're ready 2nd.
making my way to your side and getting what you complex that it evokes a to
wanted. comment from one of the run rule
(G. B. Shaw, The Man of Destiny) participants. The comment you Monarch's
itself illustrates two other for espionage
graphic features, neither of Guild Canapes?
LETTER REPETITION which has a spoken lan- President.
unethical.

The repetition of letters and hyphens generally guage counterpart: the ex- Canapes?
shows extra spoken emphasis, but other effects are Why yes.
perimental use of a logo-
Thank
also sometimes conveyed by this technique. gram in the proper name,
Canapes? You,
and the use of an ironic Mary.
And I've lost you, lost myself, question-mark.
Lost all-1-1-1- '@kins! Chervil Tate! Have a heart! Will you people take a look
(Robert Browning, Men and Women) at the pattern (?) we've been weaving ..."

What a beautiful, byoo-ootiful song that was you sang

last night.
(William Makepeace Thackeray, Vanity Fair)

'Shhhhhhhhhhh! Shhhhhhhhhhhhhh!' they said. Ugh!Aargh!Haha!

(William Faulkner, Dry September)
Non-verbal vocalizations cate different types of laugh
Fuego . . . fueeego! (Fire . . . Fire!) (V. B. Ibahez, Sangre are among the most difficult - normal (a), giggle (e),

y arena) of sounds to represent in and hearty (o). They may

written form. Even a also express different char-
Tout a coup, tres bas, mais la bouche grand ouverte, 'straightforward' event, acters:
ilpsalmodie 'vaaaaaaache' et sa tete retombe. such as a laugh, can be 'Ho-ho-ho!' laughed dark
(Suddenly, very quietly, but with his mouth wide open, shown in several ways. A Cat.
he chants 'cooooooow' and his head falls back.) selection from various 'Hee-hee-hee!' laughed the
(Raymond Queneau, Le Chiendent) authors brings to light: tippling bride . .

Ha! ha! ha! 'Heu-heu-heu!' laughed

Ha, ha! dark Cat's mother . .

ITALICS Ha, ha, ha! (T. Hardy, Tessofthe

The use of italics is found in a variety of contexts, Ha-ha-ha! D'Urbervilles)
conveying loudness and other tones of special signi- Ha-ha, ha-ha! In languages that lack in-
ficance. Ha ha! itial aspiration, the conven-
He! he! he! tion may look quite differ-
'Hel-/o!' said my aunt as I appeared. He, he! ent.
(H. G. Wells, Tono Bungay) He, hee, hee, hee! - Mi fate proprio ridere,
Ho, ho, ho! scusate, ah, ah, ah. (You
'What can you mean by talking in this way to meV Haw-haw! really make me laugh, ex-
thundered Heathcliffe with savage vehemence. HA! HA! HA! HO! HO! HO! cuse me, ha, ha, ha.) (I. Si-
(Emily Bronte, Wuthering Heights) The vowel variations indi- lone, // seme sotto la neve)
Chicago will be ours! Chicago will be ours!
(Upton Sinclair, The Jungle)

i'm desperately fond of Shirley.'

'Desperately fond - you small simpleton! You don't
know what you say.'
i am desperately fond of her: she is the light of my
eyes.'
(Charlotte Bronte, Shirley)

The non-verbal vocalizations in this extract from aWelsh

edition of Asterix are the only items that do not need transla-
tion.

31 WRITTEN AND SPOKEN LANGUAGE 181

32 Graphic expression

It is traditional in language study to distinguish tists,and many others interested in both the aesthe- Writing the date
'spoken' from 'written' language; but the latter tic properties of graphic expression and its util- The graphic expression of
term does not capture the range of expression that itarian functions (in publishing, advertising, carto- dates is one of the conven-
the visual medium makes available. 'Written' graphy, etc.). However, these diverse approaches tions universally taught in

implies, firstand foremost, 'handwritten' - but have not resulted an agreed descriptive appara-
in
schools. But it is difficult for
schools to keep up with con-
plainly there are many other ways of presenting tus or terminology, and several important aspects
temporary typographic prac-
written language, using such technologies as the of the subject continue to be neglected - not least, tice, and to distinguish be-

and the video display

printing press, the typewriter, the analysis of the effects conveyed by graphic com- tween the many social
unit. The term subsumes all these modes,
'graphic' munication. usages. These are some of
the graphic conventions
and we shall therefore use it throughout this sec-
which have been used to ex-
tion, to emphasize the importance of adopting a press the date in English.
broad framework for the study of the way language 31 January 1987
is visually presented. We shall not, however, use 31st January 1987
the phrase 'graphic language', as found in such 31st January, 1987
fields as typography, because it applies to a much 31st January. 1987
January 31 1987
wider class of phenomena than that dealt with in
January 31, 1987
this encyclopedia - including the use of pictures, Modes of graphic expression January 31 1987
st
graphs, musical notation, and so on. The several modes of graphic expression are identified January 31 st, 1987
The different varieties of graphic expression in this diagram, in relation to the study of other kinds January 31 st. 1987
of human visual communication and to the linguists' 31.1.87 1.31.87
seem to have no parallel in spoken language.
use of the terms 'spoken' and 'written' language. The 31/1/87 1/31/87
Speech belongs to individuals, and is not split classificationis based on an analysis by the British 31-1-87 1-31-87
between two people (apart from such special cases typographer, Michael Twyman (1934-), who 31187
as foreign language interpreting). There is nothing subsumes all graphic effects under the heading of 31 Jan 1987
'graphic language'.
in speech that corresponds to what happens when 31 Jan 87
31 Jan. '87
someone dictates a letter, and then allows a typist Language
1987-01-31
to present the message in graphic form. There are 1987 January 31
no spoken language equivalents to the specialist
The main distinction is
scribe, editor, draughtsman, cartographer, or
whether the day precedes
graphic designer - though certain parallels can be the month, as in British and
found in the work of the barrister and actor. Continental practice, or fol-
The
properties of graphic expression are not Aural/oral lows it, as in American prac-
Thus, 6/7/41 would
widely appreciated. A few graphic conventions are ('spoken') tice.
refer to 6 July in the former
introduced in schools as prescriptions (such as how case, and to 7 June in the lat-
to lay out a letter, address an envelope, and set ter. The mixing of Arabic and

out a science experiment or a mathematics prob- Roman numerals is also a

lem); but attention is not directed to the learning British convention.
The use of these conven-
of general principles of visual language organiza- Non-graphic
tions varies over time. The
tion that would apply in different times, situations, e.g. facial expression,
current fashion is not to use
body gesture, p. 000)
and technologies. Most conventions are ignored, suffixes or punctuation (e.g.
and as a result people have only fragmentary skills 31 January 1987), although
older styles are still often
in producing and interpreting the range of forms
taught in school (e.g. Janu-
which are available for linguistic expression - ary 31st, 1987). The styles
something they discover to their cost, when they Verbal Pictorial Schematic also vary in terms of social
are faced with such tasks as preparing posters or ('written') (e.g.drawings, (e.g. charts, situation. Abbreviations are

handouts for general use, and find that they are photographs, diagrams, more likely to be used in in-

mosaics) musical notation formal letters than formal in

unable to convey the effect they require.
correspondence; and the
By contrast, there are several professions that use of the suffix commonly
have studied aspects of graphic expression in great appears in literature for spe-

detail. In particular, there has been minute analysis cial occasions, such as wed-
Hand-made Machine-made ding invitations. In works on
of the letters of the alphabet, and the thousands
astronomy or geophysics,
of forms these letters take (in the different styles the order year-month-day is
of handwriting, print, etc.). The specialists involved usual ( 1987 January 31),
include typographers, type designers and manufac- and this is also the order re-
turers, historians of printing, historians of inscrip- commended by the Interna-
tional Organization for Stan-
tions (epigraphers) and handwriting (palaeo- dardization when numbers
Print Typescript Videotex
graphers (p. 187)), art historians, forensic scien- only are used (1 987-1 -31 ).

182 •
V THE MEDIUM OF LANGUAGE: WRITING AND READING
Types of graphic expression
Lists
Verbal graphic expression is so enormously varied that it defies any simple A list is an ordered series of lines, each acting as a

system of classification. One approach, which analyses texts in terms of the semantic unit. They range from short sequences of
reading strategies they imply, is presented below (after M. Twyman, 1982, single words (as in shopping lists) to lengthy arrays
of technical description, as can be found in collectors'
which deals, in addition, with pictorial and schematic features of graphic
catalogues and restaurant menus.
communication).
^ innct

Pure linear
Nothing in graphic expres-
sion really corresponds to
the uninterrupted linearity
which is so typical of
speech. Word spaces,
endings, and pages are
^Jcets 2 Oeuotes
Pate d'Homard et Ris de Veau £6.25
(Sweetbread and lobster pate with a pink champagne )e0yi
Jambon San Dandle et Grisson aux Figues £6 75
(Cured ham and Swiss dried beef served with Bgs)
line
L Assiette de Retan et Boeuf Marin* au Tartare du March* £6.75
(Sacas of hatbut and beet marinated, dressed around minced vegetables)

normal graphic conven- Bavaroise des Legumes, Sauce Homardine £5.50

(Vegetable mousse Savoured with basil dressed on a hbstsr vinaigrette and warm trnfood)
tions, which it is usually im-
Symphonie des Poissons en Gelee des Legumes £5.50
practical to disregard. (Terrtne of seafoods in a saved vegetable JeSy)
The spiral of characters
on the two sides of the ^Jcots 3 Ueuotes CJiau2s
Phaistos disc illustrates a Tasse d'Homard Parfume au Fenouil £3.75
(Lobster bisque flavoured with fennel)
continuously linear form.
Ravioli de Fruits de Mer au Safran £5.75
This terracotta tablet, about
{Ravioli 8Sed with Mid mushrooms and seafoods coated with a saffron butter sauce)
1 6 cm in diameter, was
L' Elixir de Poireaux en CroQte £3.75
found in Crete in 1908. Se- (Chicken consomme with madeira and ieeks
cooked under a pasoy bd/
veral characters are recog-
Tarte de Poireaux Andalouse (15 minutes) £5.50
nizable (such as parts of (Stnced leeks seasoned with spices In puff pasty with a sweet pepper sauce)
the body, animals, and
tools), and interpretations P>oissons
of some sequences have Emlnce de Sole Belle Reunite £14.25
(FlBetsol sole with baby vegetables)
been proposed, but the
Cassoulet d'Homard Forest ere £22.75
disc as a whole has not
(Steamed lobster with wild mushrooms and baby vegetables In a white wine sauce)
been deciphered. Blanquette de Turbo t au Stilton. Con fit des Orgnons £14.75
(RMet of turbo) braised In red wine and Stilton with glazed onions)

Coqullles des Fruits de Mer en

Feuillete £14.50
(Seafoods encased In puff pastry and a white butter sauce)

In ttces
Interrupted linear
Emlnce de Boeuf Poche aux Capres £14.75
This is the normal conven- (Poached Met ot'beef dressed on a caper sauce with broccca and carrots)

tion used in continuous text.

The spaces occur between ^ r" ft
linguistic units (words), and
line breaks usually occur be- Matrices
tween words or syllables (us-
A matrix is a table of rows and columns containing
ing the hyphen).
linguistic,numerica ,or other information, which can
The present page illus-
be scanned vertically or horizontally. Matrices are
trates two line-break princi-
\ (?) widely used in technical publications, but several
ples in operation. The larger V JtLX
everyday topics are also conventionally treated in
size of text has a justified
matrix forrr such as this football league table.
i,
right-hand margin, with irre-

gular spaces between

Electric circuits
HRST DIVISION
words, and occasional hy- The complete path along which an Home Away
phens at graphic syllable from its start to its finish, is called
electric current flows,
an electric circuit.
P W D L F A W 1) 1- FA Pts
boundaries. The present par-
Leeds 13 5 1 12 4 3 3 1 3 20
8

agraph has an unjustified

The diagram shows a simple electric circuit.
The electric current comes out of a battery,
Arsenal 13 6 1 21 2 13 2 6 11 18
Man. C 12 3 3 12 4 3 2 1 5 4 17
('ragged-edge') right-hand goes through a switch and two light bulbs, Spurs 13 4 1 1 9 4 2 4 1 10 6 17
margin, with regular spaces and then back to the battery again. Cryst P 13 5 2 10 5 2 3 1 5 4 17
between words, and many When the switch is on, the circuit is complete Chelsea 13 3 3 11 8 2 3 2 5 5 16
and the light bulbs work. Wolves 13 3 1 2 12 13 4 1 2 14 14 16
fewer word divisions. In both
cases, line endings do not
When the switch is off, the circuit is said to be broken, L'pool 12 4 2 12 2 13 2 3 4 15
and the light bulbs do not work. Stoke 13 4 3 13 1 2 4 5 15 13
correspond to units of mean- Cov C 13 3 1 2 6 3 2 2 3 6 8 13
ing. In contrast, the following Different types of electric current Newc U 13 1 4 1 6 6 3 13 9 10 13
The current produced by a battery flows in the same S'hmptn 13 3 2 1 8 3 12 4 7 10 12
2 14
extract taken from a series
is
direction all the time. Electricity that goes Everton 13 2 3 1 9 6 9 15 12
of books designed for slow
only in one direction is known as direr! current, or D.C. Derby 13
n
.> 1 3.11 9 12 3 7 11 11
readers, in which most of the The current made in power stations flows very rapidly YVBA 13 • > I 13 9 2 4 9 21 11
line endings have been in one direction round a circuit and then in the other
Man. U 13 9 2 6 4 12 3 7 14 11
made to coincide with the direction. An electric current that changes direction
Notts F 13 3 2 1 12 fi 3 4 1 12 11
H' field 13 3 3 1 9 5 2 4 3 12 11
boundary of a grammatical in this way is known as an alternating current, or A.C.
Ipswich 13 • > 2 2 13 7 15 1 8 9
unit, such as a sentence, An alternating current, passing through the filament W Ham 13 1 4 2 9 10 3 3 6 11 9
clause, or clause element of a light bulb, flows
50 times a second.
backwards and forwards B'pool
Burnley
13
13
1 >
2
2 6
5
9
4 12
115 2
4 14
4 2 10 4
8
(§1 6). (From D. Crystals J.
Foster, 1983, p. 7.)

32 GRAPHIC EXPRESSION- 183

 —

Linear branching Graphic symbolism

START Like sound symbolism (§30), graphic properties can be
used to represent the extralinguistic world in a direct
manner. The illustrations show this principle at work
Is official transport available? in the worlds of business, education, and humour.

YES NO The distinctive typographic design used by the

I
fashion firm Streets of London How are the effects
Prior authority from Can you travel as a to be interpreted? The account of the firm's policy,
Finance Division passenger in another printed on the inside cover of one of their catalogues,
is needed to use private vehicle? provides the clue.
own vehicle.
YES NO Out on its own. A style all its own. Streets veers off

the beaten track. Exploring new trends. Going beyond

Prior authority from the norm. Bringing the remote within reach with a look
Finance Division is needed of solitary refinement. Distant. Different. And desirable.
to use own vehicle. Break away and break the monotony.

The vehicle is The vehicle is

Is your
vour own.

visit to one place only?

hired, or
otherwise ?

Prior approval from

st-e OF LONDON
YES NO Establishment Division
Is the place visited must be obtained.
more than 75 miles from Isthe outward journey
Learning to read This is
the starting point? more than 75 miles?
one of a series of animated
YES NO YES NO letters devised to help
teach sound-letter relation-
Prior approval from Prior approval Prior approval from ships to children (From L.
Finance Division is not needed. Finance Division Wendon, 1985, p. 28).
Is the return journey-
is needed for is needed for
more than 75 miles?
all journeys. all journeys. ,
SAMMY SNAKE
YES NO This snake slithers and
I
slides along making a

Prior approval from soft hissing sound in

Tree diagrams are popular Finance Division words, like this 'sssss'.
Is the distance between any
representations when there is needed for
two consecutive places
are two or more alterna- all lournevs. visited more than 75 miles? Graphic humour Cartoon-
tives which need to be ist Edward McLachlan

NO
clearly related, as in this
set of instructions, cast
binary form. Multiple
in
YES

Prior approval from

I I

Prior approval
has ingeniously exploited
the graphic nature of
these words.
DANG R
branching is found in family Finance Division is not needed.
tree diagrams. is needed for

all journeys.

Non-linear viewing
In thisform of presen- TYPEWRITER, Underwood, manual,
tation, the linesare not
read in sequence. The
typography directs the
modern.
Cambridge 4960.
CYCLE TYRES.
Price
C—29.9
26 by 1|, 26 by U, 27 by
£12 10s. - Phone
FR&GU
reader's attention to
ferent parts of the text,
dif-
H, from 11s. Tubes 5s. 6d. K. —
which may or may not be Lovegrove, 181, Oxford Road, Cambridge.
read in detail. Initial read- Tel. 51436. 19.12 P—
ing may proceed in any WEE BELLING Electric Cooker, £5;
direction (even vertically Tricycle, chain driven, age 4 to 6, £2 10s.;
upwards). These conven- Cot and Mattress, £5; Boy's Sports Cycle.
tions are standard in ad- 3-speed, for age 12 onwards. £9 10s.;
vertising (p. 390), popular Pram Seat £1 Cambridge 64297.
.
—
journalism (p.
some forms of poetry
388), and
26.9 P—
(P- 75).
MARPARTS open all day Sundays and late
night Fridays. 5B— 26.9
SOLID FUEL BOILER, Ideal Marquis,
— Cambridge
45,000 b.t.u., with
21330.
MARPARTS open
night Fridays.
flue.

all
£15.

day Sundays and

P— 26.9
5B— 26.9
late
NO ENTRY
Ivl^Uri^n w
11

184 •
V THE MEDIUM OF LANGUAGE: WRITING AND READING

Graphetics and graphology
The writing system of a language can be studied
from two points of view, which relate to each other
in the same way that phonetics and phonology do
for the study of speech (§§27-28). Graphetics, z.
term coined on analogy with phonetics, is the study
of the physical properties of the symbols that con-
stitute writing systems. Graphology, coined on
analogy with phonology, is the study of the linguis-
tic contrasts that writing systems convey. Very little
research into these domains has been carried out
within the field of linguistics (§65); in particular,
the notion of graphetics is not widely employed.
In such fields as typography, chirography (hand-
writing), and psychology, however, aspects of these
topics have now received considerable study, espe-
cially in relation to the teaching of reading and
writing, and the task of visual perception (§§34,
44).
film, and electronic display screens. Often, tech-
niques have to be devised for special functions, such
as architectural drawing, record keeping, laundry
marking, security coding, writing on glass, wood,
or film, and writing that can be read electronically,
as in department store check-outs and libraries.
The three main eras of graphic expression —
handwriting, printing, and electronic — share many
graphetic properties; but they have developed
separate traditions and disciplines of study, and
they will therefore be reviewed separately in the
following pages. Graphology, in the linguistic
sense, will be discussed separately in §33.

GRAPHETIC ISSUES Three kinds of ox-writing

A properly developed theory of graphetic science There are several examples of boustrophedon in the
history of writing. In particular, it was used in a
would deal with the range of implements and asso-
transitional period of early Greek writing. The Greek
ciated human skills required for the production and way the ox
name means 'ox turning', referring to the
reception (reading) of linguistic marks on surfaces, would pull a plough, moving first in one direction, then
screens, and other backgrounds, in any language. the other.
This would primarily involve the study of motor In theory, there are three possible ways of writing
boustrophedon. In (a), the lines reverse but the words
control and coordination of hands and eyes (cf.
do not. In (b), the words reverse as well as the lines.
the use of the vocal organs for speech, §22), and In (c)the letters reverse as well as words and lines.
of the psychological processes involved when these (a) This is an illustration of
marks are perceived and remembered by the reader writing of way possible one

(cf. the field of speech reception, §25).

in a boustrophedon style. The
but ^direction reverse lines
There is great scope for the development of gra- the words do not.
phetic studies, when we consider the range of vari- (b) This is another illustration
ation in graphic practice displayed by modern nl .gnitirw nodehportsuob fo

languages, and throughout the history of writing. thiscase, both the lines and
.desrever era sdrow eht
Most noticeably, languages vary in the direction shows
(c) The third illustration
in which they are written — left-to-right, right-to-
2i9tt9ln9(1w 3n9qqsr1 terlw
left (e.g. Arabic), top-to-bottom (e.g. traditional are reversed as well as words An unlikely - but effective -
Japanese), and the uncommon bottom-to-top (e.g. .aenilbne mode of graphic expression:
some forms of Ancient Greek). More than one di- smoke in air, lasting for only
a matter of minutes.
rection may be involved, as in the boustrophedon
method of writing lines in alternate directions, used
in several early systems (see right). A language may
worn
use several different conventions simultaneously -
such as the common use of vertical arrangement
in neon signs and on book spines in English.
The nature of the writing implement and surface Efc^QM
will have some influence on the kind of system that
develops. The history of graphic expression shows
a variety of implements, including the use of reeds,
This drawing of an early Greek treaty (6th— 5th century
quills, brushes, steel points, fountain pens, pencils,
bc) is of the third kind, as can be seen most clearly
ball-point pens, fibre-tipped pens, chalks, crayons,
from the reversed Es.
typewriters, laser printers, photocomposing sys- Boustrophedon writing has also been found in many
tems, and word processors. The implements rely other parts of the world. Several inscriptions have been
on a range of natural and synthetic products, from found in the countries in and around the Mediterranean

the early use of blood and plant juices to the

modern range of coloured inks, photochemicals,
lights, and electrical charges. Many surfaces have
been involved, such as animal bone, rock, clay,
wax, pottery, cloth, papyrus, parchment, paper,
Sea- in Crete, Cyprus, Italy (both in Etruscan and in
the Italic languages), Asia Minor (Hittite), and ancient
South Arabia. But there are also inscriptions much
further afield, in India, Central America, Easter Island,
and northern Europe (in the early runic alphabet).

32 GRAPHIC EXPRESSION 185


Handwriting
The many forms and styles of handwriting (or chir-
ograpby) have attracted a wide range of aesthetic,
psychological, and scientific studies, each with its
own aims and procedures. Moreover, each of the
main families of writing systems (European, Semi-
tic, East Asian) has its own complex history of
handwriting styles. No universally agreed system
of classification exists, and there is considerable
controversy over approach and nomenclature. All
that can be done in this section, therefore, is to
present a few of the categories and descriptions
that are widely recognized, using the European
writing tradition as the domain of illustration. (For
other traditions, see p. 188.)

Book hand This is a professional form of writ-

ing, found in many different styles, used mainly
for copying literature. It was formalized, clear,
and regular, displaying little scribal idiosyn-
crasy.

Documentary hand This form consists of a

sorudo puopftxre ©DOubiinc^
range hands used by officials and private indi-
of
viduals as part of daily routine. It is generally
found as a rapidly produced cursive, often very
irregular and difficult to read.
SKvuinarknu (XpwJOffjjGCjru'.i; cir^o
Majuscule
letters
Several forms of writing consist of
broadly contained within a single pair of ran TTrouuiraiu »i pem j. »o<
;
' rftjoai*
horizontal lines; they are usually referred to as
capital letters. The Greek and Latin alphabets Insular half-uncial writing, as found in the Irish Book of Kells, c. ad 800.
were way. The chiselled
originally written in this
inscriptions of ancient Greek served as the Cursive form of writing, the characters
In this Black letter This was a development of Caro-
model of writing on papyrus rolls (which survive are joined a series of rounded, flowing
in lingian minuscule, widely used in many varia-
from around 300 bc). The Latin form used strokes, which promotes ease and speed. It is tions between the 1 1 th and 1 5th centuries. The
throughout the Roman Empire from the 1 st cen- found in general use from around the 4th cen- rounded strokes became straighter, bolder, and
tury ad is known as rustic capitals (in contrast tury bc and in time replaced uncial and half- more pointed. Often referred to as Gothic script,
with the great square capitals chiselled on stone uncial writing as a handwriting norm. it became the earliest model for printer's type

in Roman inscriptions). in Germany.

Insular This form of writing was developed in
Ireland from around the 5th century ad. It was [kqC^UtJptft^tCi 1 dome pnim
rtlL\lNOr<USI\JXY brought to England by Irish monks, where it was
used alongside uncial writing, which it ultimately
supplanted.
I TOcq>itvafhyf0 oyib? uiea R
Early Roman rustic capitals
I nuutr^CRCtt)vmologiUuo
Carolingian minuscule This form of writing
Minuscule Several forms of writing consist of was named Emperor Charlemagne
after

letters whose parts extend above and below a (742-81 4), who promoted it throughout Europe.
pair of horizontal lines. They are usually known It was widely acclaimed for its clarity and attrac-
uim trvTaabrancfcrice butt* \tV
as small letters. Minuscule writing was a gradual tiveness, and exercised great influence on sub-
A Gothic book hand of the 1 4th century.
development, in regular use for Greek by the sequent handwriting styles. It is from this period
7th-8th century ad. The original form ('pure that we find the development of the 'dual alpha-
bet' - the combination of capital letters and Humanistic (Roman) This form of writing,
minuscule') was later modified, as uncials and
small letters in a single system. also based on the Carolingian, was devised in
other features (such as Greek accents) were
Italyby Poggio (1380-1459) as an alternative
added.
It was originally known as
to black letter writing.
C^cubcrtb:cAprxrxim xutou Antiqua, reflecting the concern of the humanist
movement of the period to return to ancient
&Ot*iurn mxnuprxcCftf toriT Latin sources. It subsequently became the basis
for roman letters in printing.

CopioftLxcnfcffluerer'lpL uiumT xxmin ctcultx> Ltmre .

r^\*incoLt*miC'j<lo(bbfhj f110^ ^*«r minufprotutnmtur

The prototype roman script of Poggio.
The Gospels Stoudion - the earliest dated true
minuscule (AD 835)
of
-zArizxcrctrnirxciilo- idquo
Italic This form of sloped cursive lettering was
developed by the Italian scribe Niccold Niccoli
Corebxr uefnzxffioxbxmix\ (1364-1437). In due course, it led to the deve-
lopment of italic letters in printing.
Uncial This form of writing was
used in Greek and
Latin manuscripts from the
especially
CcfubcxeLo- quinurnnumi
4th to the 8th century ad. It consists of large uA tUrtton cum vn oendto lat&ret
•"

(the etymological meaning of the term is 'inch-

Xmi-ocrx
high') rounded letters. A later development, half- )
uncial, prepared the way for modern small let- An example of late 8th-century Caroline
ters.
minuscule. The prototype italic script of Niccoli

186 • V THE MEDIUM OF LANGUAGE: WRITING AND READING

 T

PALAEOGRAPHY frequent on such objects as seals, rings, medals, Minim confusion

Palaeographers study ancient and medieval hand- and coins (the separate study of the latter being In some medieval styles of
writing in order to establish the provenance, date, known as numismatics). Large numbers of clay and handwriting, several letters
and correct form of a text. The subject principally papyrus inscriptions have survived throughout the were formed by a series of
joined vertical strokes ('mi-
involves the study of writing on papyrus, parch- Near East containing information about historical
nims'), without further dis-
ment (vellum), or paper, though it does not exclude events and daily business activities. And there are tinction. Sequences of m, n,
other forms (such as graffiti). Most palaeographic thousands of inscriptions of a more casual nature, v (written as u) and /(which
research has been into manuscripts within the such as graffiti. Illustrations of ancient inscriptions lacked a distinguishing dot)

Greek/Latin tradition. are shown on pp. 305, 317, and 327 would thus appear identical.
A six-stroke sequence could
There are innumerable problems facing the
be interpreted as ium, miu,
palaeographer. In olden times, few books were DIPLOMATICS iniu, niui, and many other
dated and title pages are a late medieval develop- Diplomatics, from the Greek diploma (folded), is possibilities, giving rise to

ment. The absence of spaces between words in the study of legal and administrative documents major problems of interpre-
tation.
preclassical and classical texts can lead to ambi- of all kinds. Most attention has been paid to the
guity. Also, variant texts have to be brought
together, to determine which is the original read-
ing.
The problem of
serious.
textual error is

It is not surprising, given the 'routine'

particularly
public documents of monarchs, emperors, and
popes, which are usually classified separately from
the many varieties of private document that exist.
One of the main aims of the subject is the identifi-
cation of genuine documents as distinct from
MR
nature of the task, that copyists would introduce drafts, copies, or forgeries. Particular attention is The Magna Carta
errors as they worked — errors which would be paid to the materials and inks used, as well as to This important document
compounded as further copies of a manuscript the handwriting style, the forms of seal or signa- was validated by a seal
were made. Indeed, in many cases, the scribes did ture, and the linguistic features of the text (such (below), not a signature.
King John, like many medi-
not know the language or dialect of the manuscript as the choice of language, the way information is
eval monarchs, could not
they were copying. There are even cases on record structured, and the kind of dating system write.
of scribes copying right across a two-column text, employed).
producing a totally unintelligible version.
Another big problem arises out of the use of
abbreviations, especially common in Roman times.
Letters at the end of a word would be replaced
by a point or other sign {suspension, e.g. imp. = Anglo-Saxon originals
imperator 'emperor'); and letters would be omitted In an early Anglo-Saxon technique, it was possible to
within words {contraction, as in the shortened make more than one 'equally original' version of a
document. The text would be written out two or more
forms of Jewish or Christian holy names). Whole
times on a single sheet of parchment. The space
words were sometimes replaced by shorthand between the texts would then be filled in by various
signs, e.g. Latin et = 7,est=+. By the end of the words or symbols - often the word chyrographum
middle ages, over 13,000 abbreviations and signs ('handwriting') - and the sheet would be cut irregularly

were in use. through this writing. To prove that the texts were
genuine was then a simple matter: only the original
Palaeographic detective work is assisted by a
texts would exactly match when reassembled.
detailed knowledge of the language, the historical
events of the period, the contemporary use of writ- -*=
ing materials, the mannerisms of the scribes, and irceop ^uriJi spvff$ife miDcr jwpunsc -\lcccpt m
especially the history of handwriting styles. In :ogxn jfdj-Hicrtajcj- onyiosfynr oxxfcpr Unrcfj
modern times, such techniques as the use of ultra-
i
tepjuxmcuftim rrurnnum rtornnjcrun-jcrlj
violet light (to bring out faded handwriting) have
proved invaluable.
yj-ulp ncrmcc pjicxvo^- ftrpccn -jpfiuncfir phi
xV hif vd^ ppsccrf: ")<% cfy kip v<%<
ff
crm *ilF
y<x\ia\>eaiinccv&p jpaxvlwfv
EPIGRAPHY :Si(f:e|rbiV<fi:dJt

Epigraphy is the study of ancient inscriptions -

texts that have been written on hard, durable •TFfuofi htnari}:i^v^<siie-jfoclscfto%-\hpjfs^y
material, such as stone, marble, metal, clay, pot-
taxush^f \t "Ojuhz ncp sr by n» -croc tf{- onb jr fit
tery, wood, and wax, using such techniques as
$C&rnccju<i-]$cy micbcrbet cumfcu pfrjw y
engraving, carving, embossing, and painting. Its
aim is to ascertain the nature of the original records p hva tutor) scf yfilnruTTi lire ay&roxn pi Lie- y>
of ancient civilizations, thereby providing the prim- kfiir ft&kmc on|vcm^TXrncfr*xr)T<fir \f\\^
ary data for historical and philological enquiry. In ftp. if yeo hotvo ffccfii op pob>tj- bi|Ta>j*f-.-jur
this process, it provides considerable insight into
Yyxufv.f edfjiev ct^ru:
f cynpsc c
the early development of writing systems.
Several kinds of ancient inscriptions exist. Many
rpjicofc \ pulpbun ctrjic J-tflrfrxm <

are found on or within large monuments, such as f\uffe. f bruhfrcm dqu ftrorcpnrc

the Egyptian pyramids, or the Persian rock carving pp^P-J- fulfsart ctepe J.cdfsan> <
at Blsitun (p. 301). Memorial inscriptions are also

32 GRAPHIC EXPRESSION 187

 ,

CALLIGRAPHY
The penmanship, or handwriting at its most
art of
THE
formal, is known today as calligraphy. It is a major
art form in eastern Asia, China, Korea, Japan, and
in Arabic-speaking countries. In Europe and Amer-
ica, it has been less widely practised, though there
has been a strong revival of interest since the end
of the 19th century. The artistic effect depends on
a combination of factors - good-quality materials,
the selection of an appropriate and effective writing
instrument, the correct formation of the symbols
INVITATION, ^4^^D INSTRUCTION.
according to an accepted style of writing, the plac-
ing of these symbols in an elegant sequence, and
the harmonious layout of the text on the page.
There have been several famous schools of calligra-
phy, and the subject has attracted a great deal of
historical study in which specialists identify the dif- (j/>r a. i year csnajfe/ j/aad Wakt^, _—
ferent kinds of style and how they have evolved.

(
Right: A fine example of 1 3th-century calligraphy. ^/ffa/? /r////s , JO/7y./ SfOr/t/yrttrtfxm . ".
Csf/i/f/jf.i/'/t fr/a/wi f/a/{rxl?//e/if'$fl.

& «* ft i 4
1L
^Aabcdefgh ijk Imnopqrfs tuvwwxyz. tz&.^ a
ABCDEF GHIJKLMNOPGLR £ V-K.%4 fy $1

Aabcdefghij klmnop qrfs t uvwxyz.

» * *ffi
ABCDEFGHIJKLMNOPQ
Above: A handwritten English
alphabet (top: round hand) com-
pared with Italic print (middle) and
Roman print (bottom).

Right: An elegant Arabic script,

taken from a late 1 8th-century
M J?
y
~

^*
mMf%m
*^' *s >'
f^i»
-*J
grammar. i Above: Examples of Chinese hsing shu
(running style) and ts'aoshu (grass
style) from the influential work of Wang
Hsien-chi (ad 344-388). The latter
forms are so simplified that they are

*^<?M<h^ „ .,,
capable of being read only by
phers. The primary
calligra-
interest lies in the

$$& way shapes and spaces can be created

•
to produce a harmonious and mean-
\xt ingful whole.

188 • V THE MEDIUM OF LANGUAGE: WRITING AND READING

GRAPHOLOGY (PSYCHOLOGICAL) as to determine someone's suitability for employ- Signature
A person's handwritten mark or signature holds ment). interpretations
a special place in society. It is required for legal Some examples of graphological interpretation
Rosa Baughan's Character
agreements, and its forgery can be illegal. Likewise, are given below. The analyses sometimes convey Indicated by Handwriting: a
a person's general handwriting conveys identity: an but there are many
initial intuitive plausibility, practical treatise in support
no two people's writing is the same in every respect individual differences, and there still need to be of the assertion that the
(cf. voiceprints, §6). controlled empirical investigations into the perso- handwriting of a person is an
infallible guide to his char-
from here to the view that
It is a short step nality generalizations that have been proposed.
acter is one of many books
handwriting conveys information about a person's There are also many differences of opinion among on graphology that appeared
character and personality — the subject of grapho- practitioners about the analyses (such as whether at the end of the 1 9th cen-

logy. Although speculation about this relationship meaning is best located within the different ele- tury. includes copies of
It

many famous signatures,

has existed since Roman times, popular interest ments of a letter, or within the letters and words
with personality interpre-
is quite recent, with the term itself being introduced as wholes). At present, we lack the scientific evi- tations.
towards the end of the 19th century by the French dence required to demonstrate the accuracy or
abbot, Jean Hippolyte Michon (1806-81). This reliability of graphological procedures and conclu-
sense of graphology, it should be emphasized, must sions.
be clearly distinguished from the linguistic use of
the term found on p. 185. Letter interpretations
Graphologists study handwriting variation using A Manual of Graphology
Extracts from Eric Singer's Napoleon Bonaparte
(1953), showing the kinds of relationship postulated There is the dominant will in
several parameters, notably:
between forms and personality traits. The drawings are the strongly marked "t", and
in the hard, thick line which
Size whether large or small, wide or broad, con- by Gertrude Elias.
terminates the flourish; his
stant or varying, and including the relative size of egotism and self-assertion
Different kinds of t cross
individual letters or letter elements (e.g. the length are evidenced in this flourish,
of the cross stroke of a f). his originality in the peculiar
form of the capital letter "B";
Layout the arrangement of writing on the page, but ambition is here "still the
including the size of margins and the distance lord of all".'

between lines (narrow or wide, constant or vary- Desire (o protect, authority, a fatherly spint.

ing).

Line direction whether straight, sloping upwards dr/o/vt,

or downwards, or curved.

Connection whether a sequence of

joined or separate, and how
letters
the upstrokes and
is
I
Quick thinking, enterprise, thoughts running ahead of action, haste
Alexander Pope
An elegant signature, per-
fectly freefrom pretentious-
downstrokes interconnect (curved, angular, with Angles of writing ness. The simple form of the
various flourishes). two capital letters shows cul-
ture and refinement of a high
Temporal features whether the speed of writing order, and the liaison of the
is rapid or slow — in rapid writing, for example, capital letter "A" to the name
t cross strokes and / dots may be misplaced, and Angle to the left Upright ingje shows deductiveness and
strokes may appear between adjacent words. Dislike for socio} relationships; retreat
into a private world
Independence, reasonableness logical power, of which no
poet ever had more.'
Regularity whether the size, angle of writing, and
The meanings of D/d
distance between strokes is constant or varying,
Open bottom
and whether there is an even or disjointed appear-
it

ance to stretches of writing. V Wants to knots himself

Left parts tastelessly exaggerated

Vulgarity

Mozart
Written in two pans
Letterforms whether simplified or elaborated, Second arc broad, with extended stroke
'Sensuous tenderness in the
involving different degrees of legibility. I) Individualism, lack of adjustment Underlinmg of own importance sloping movement and the
downstrokes of the letters
Angle whether letters are upright, or slanting to A quiz question
the right or left to different degrees, and whether There are two writings [below], the upper one by a
gentleman, the lower one by a lady. If you were asked
this is constant or varying.
whether you think this couple would make a good (Answer: A marriage be-
match in marriage, what would you say?' tween these two people
Shading the thickness or thinness of different
-Ua^ -aell<j>
would be a failure from the
strokes. *V start.They are too different.
Look and spacing, at
at size
Handwriting characteristics have been studied
the angle of writing and at
with reference to all kinds of normal and pathologi- the form of connection. Their
cal psychological and physiological states. Most personal experiences of
of the early publications dealt with the writing space, their social inclina-
tions, their personal ways of
of monarchs, criminals, authors, politicians, and
adjustment to life and society
other professionals, but more recent works have
are completely incompatible,
examined the writing of the general population — and will never agree to-
sometimes, from quite specific points of view (such gether.')

32 GRAPHIC EXPRESSION •
189
Print
The selection and organization of letterforms and Kind. Characters.

other graphic features of the printed page is the

concern of typography. It deals with all matters Roman lower-case . a to z and 3e oe fT fi fl &i i

that affect the appearance of the page, and that Roman small capitals a to z and £ s ^

contribute to the effectiveness of a printed message: Roman capitals A to Z and .€ffitt

the shapes and sizes of letters, diacritics, punctu- Roman figures 1234567890
ation marks, and special symbols; the distances Fractions lliiillli
between letters and words; the length of lines; the Roman points
space between lines; the size of margins; the extent iiaaa^eeij ii
Roman accents
66douiiuu»vwi;fl(^
and location of illustrations; the use of colour; the
selection of headings and sub-headings; and all Peculiars •ttnt'—*—»-—
other matters of spatial organization, or 'configu- Commercial signs @ 9> "> £ » % - - x
/

ration'. In addition, typographers need to be Italic lower-case a to x and a a ff fl fl ffl fi 33

involved in such matters as the kind of ink used, Italic capitals A to Z and MCE & £ 3°

the choice of paper, and the method of printing. Italic figures . 10

Each of these components has to be evaluated in Italic points . : :1?C1 6

its own right, as part of an overall judgment about iilaidiiUlli

Italic accents 26 Left: The ordinary fount of
6 b 6 u Q
the 'weight', 'colour', or 'atmosphere' of the page
it it it -to tt' f ft

275 characters, taken from

as a whole. '75 L A. Legros& J. C.Grant,
The design of individual letters is the main Typographical Printing-
Surfaces (1916)
concern of specialists in typography called 'type
designers'. Originally, in the western tradition, let- SCRIBAL ABBREVIATIONS. Left: Some of the special
terforms were devised to reflect the properties of
the main 15th-century manuscript hands - the
a a % fc b c c c i xSadeeSfffg 2 * h
characters illustrated by
Legros& Grant (1916).
roman, italic, and black-letter styles (p. 186). Since 1 i
J ft i 1 ft H m m (nj m n n n! 5 6

gq r I § t t T. ? Below: Part of a printer's

then, there has been a remarkable proliferation of P P p p p q q. q, list

has been estimated that over 10,000 type-

styles. It u u u v vwxxyyza of special symbols. The
numbers ensure that the
faces have been designed since the invention of
printer correctly interprets an
printing — a variety that has so far prevented the a r> E I O M P P P> I? V author's symbol require-
development of any single system of classification. ments.

Typographic measurement
A wide range of typographic
terminology has developed
to handle the many kinds of
tal letters, but no
proved possible.
A more recent approach
agreement
i Ty p oH r a S
pnyjg i
001
002
003
004
+
-
X
=
035
036
037
cf
9
{

typeface and setting, but this argues that any new system
038 }

iscurrently subject to profes- should be based on the 005 * 039 +

sional scrutiny. It is argued height of lower-case letters, 006 + 040 -
that several terms and con- which predominate in most
cepts originally devised for printed text. It proposes a
007 041 X
Greek
use with metal type are no four-level system, using the 008
•

042 =
"
longer clearly applicable in notions of 'x height' (the 009 043
£ TuimrpaffiHH
-i-
the context of photocomposi- height of the small letter x),
tion or laser printing (e.g. the 'ascenders' (a part of a letter
010 ~~
044 ±
point size of pieces of type, 011 == 045
that extends above the
Cyrillic (Russian)
or the notion of leading -{he height of the letter x, as in d) 012 = 046
t

spacing between lines of type), and 'descenders' (a part

wmt > tt

Several proposals have which extends below the x,

013 047
been made for alternative as in /). The approach is in : 1: 014 < 048
systems of measurement principle applicable to other 015 ^ 049
based on the characteristics (non-roman) writing sys-
of the printed image (as pro- tems. Greek and Cyrillic re-
Hebrew Arabic 016 ^ 050
•
duced by any method) rather
than on the characteristics of
the 'body size' of the traditio-
nal piece of type, but these
have proved be controver-
to
quire no modification.
mean
brew, and Indian scripts can
be aligned with the roman x
height. And Chinese, Kor-
The
height of Arabic, He-
A>-

C :m Chinese
017
018
019
020
<-
-
T
1
051
052
053
054
•
•
Deuandgarl (Hindi)
sial. Draft proposals for an
international standard were
ean, and other oriental
scripts can be aligned with
021 V 055
drawn up in the late 1 970s, A Capital line 022 056
capital height. (From S. C Baseline
based on the height of capi- Brogain, 1983.) B Mean line D Descender line 023 057

190 •
V THE MEDIUM OF LANGUAGE: WRITING AND READING
 A selection of modern typefaces
Complex explanations
Eurcstyle

abcdefghijklmnopqrstuvwxyz
It is not difficult to find at- such thing as a 'best' way tohave the information set
tempts at written explana- of displaying information out as a logical tree (b), be-
ABCDEFGHIJKLMNOPQRSTUVWXYZ tion that fail to communi- typographically. The back- cause the format helps the
1234567890 .,;:"»«&!? cate because of ground and experience of reader to distinguish be-
unnecessarily complex lan- the reader, as well as the tween relevant and irrele-
Kismet guage (§63). An important subject matter and the cir- vant factors. On the other
abGdefghijWmnopqpstcjvwxgz application of typographical cumstances of use, have hand, an experienced per-
research involves present- also to be taken into ac- son, familiar with the var-
/98GOeFGI?le}K2fr?n?OpQ^§5aVWXYZ
ing readers with alternative count. This was shown by ious alternatives, could find
1234*567890 .«fr!?
typographic versions of the a study that looked at var- this format too unwieldy,
same message, see
to iousways of presenting the and might prefer the more
Newtext Book italic ITC
whether their performance complex information that compact, tabular presen-
abcdefghijklmnopqrstuvwxyz is helped or hindered. might be put out by a tation (c). A further alterna-
ABCDEFGHIJKLMNOPQRSTUVWXYZ Some versions, it seems, space-age agency
travel given (d). (After P.
tive is

1 234567890 .,;:"»«&!? are much easier to read (a). Someone who has to Wright, 1977.)
and understand. sort out this situation for the

OCRB However, there is no first time will find it helpful

abcdefghij klmnopqrstuvwxyz

ABCDEFGHIJKLMNOPQRSTUVWXYZ (a) VI' here onht time

travel h\ rocket.
is limited (d)

1234567890 .,: :&!?

Where only cosl is limited
When time is limited, travel bv Rocket, unless cosl travel b\ satellite if iournc\ more than Id orbs.
is also limned, in which case go b\ Space Ship. travel h\ astrobus it |ournev less than 10 orbs.
When only cost is limited an Astrobus should be
used for journeys of less than 10 orbs, and a
Where hoth time and cost are limited
travel bv space ship.
Satellite for longer journeys. Cosmocars are rec-
ommended, when there are no constraints on time Where time und cost are m>t limited
or cost, unless the distance to be travelled exceeds travel bv super star it journey more than 10
10 orbs. For journeys longer than 10 orbs, when orbs.
time and cost are not important, journeys should bv cosmocar than 10 orbs
The Importance of typographical design be made by Super Star.
travel it journey less

Lack of thought in this area can often lead to unsatisfactory

results. The width of the table on the right-hand page of this Is time limited?
early typing manual (A. E. Morton, Modem Typewriting and I

(b)
Manual of Office Procedure, 1929) has led to its being printed I

No
sideways ('landscape'). The result is that, when the book is I

held in the normal way, some of the headings appear upside Is cost limited? Is cost limited?
down. I
I
1
I

Yes No Yes No
Modern Typewriting and Manual of Office Procedu I I

Is travelling distance Is travelling distance

more than 10 orbs? more than 10 orbs?
I
~1 r
Yes No Yes No

travel travel travel travel travel travel

5 S s » 3 by by by bv by bv
«,or>iq auorXyiax S..L 1* .2 Space Ship Rocket Satellite Astrobus Super Star Cosmocar
Xq uaaya caJou o^v
i
If journey less Itjourney more
(c)
than 10 orbs than 10 orbs
<.at09
-utciqo Xia*TiaaJ=oa
UOTiaTJlBUOa-OBUA ssssssSSssfiS travel bv travel by
Where only time is limited Rocket Rocket
k?i h oi XiaAMriaj
i:dry tjai jo aits
iliii" |nl
travel bv travel by
*uou»ura»*i t«utj puw
Where only cost is limited Astrobus Satellite
uoneiado uao*i*q sail
*«» rR^oo.T^,

•Uonrj.-iJo travel by travel by

jo eX«p o»t ui-Jita Where time and cost are not limited Cosmocar Super Star
ty»Tii oi xt»Ait«t»j
Udnd M»l jo Mis | | ||

1 Tiojc; [BUOtia - 8?1 travel by- travel by

oajrunjq
Where both time and cost are limited
Space Ship Space Ship
1
;5iJ?sS
•uot iSucS ToaiAJ^a
lo-jodna »j*t - :j.:i

'H0IlVii3d3
S5S3S35$ 5.!5S3 (

•siaoaiwadra
Jo 'on
"- >;-- -s
S

32 GRAPHIC EXPRESSION •
191
 finmfyifip nnfit3. i nftt nrfnfp man &•
Printing
There are two main dimensions to printing: the orinq; npnliqfi mDutiftnta.fr nra-
am e tjuuuur a D rra a mm rt
use of a device to make copies of an image, and
the availability of movable type. The first of these
MM
inr
darn : an noagnran m ma*
ua nlirmafnilu rt ftonna cnairirroa.
fi ttn

isknown to have been used in China from at least £<n£OmnnUuB&ruB-rtflu.£rrfmi

rt nnrinnlrra mitu t rtplrtt steam n
around the 7th century ad: the earliest known
•

uimntr ram :i fiora marairu mfabra

book, The Diamond Sutra, was printed using inked marie-i pplanlimie ah : * umngfta
amrnanbuBqur mmmmir fim mra.
wooden blocks, in 868. The second dimension fcrattp awo fat Btai antne anrat
.

emerged during the 11th century, when movable tart) am aduiutui Centra fun tntara-

blocks, carved with individual characters, came to trmunfa litma qur mMt i fnraruiia
frmttf grama fin rut mutmiua f ffta •

be used. By the beginning of the 15th century, in ? dims atarmtniB tftrt-otniqi uolum
Korea, the process had developed to the extent that nu i nmunfia q moufrur ra mra-ft f
qutbus t amraa mufa : ut hatetr as
printers were manufacturing bronze sets of type
nrfcrttml ft rami rft ua.^iomp cm a
containing 1 00,000 pieces. But none of these disco- nmda qur ftotar : n tat nalDf bona.
veries became known in the West, and they had
Gutenberg's printing press
no influence on the subsequent history of printing.
In fact, we do not know what
In Europe, the main step forward came in the Gutenberg's original printing
mid- 15th century, with the invention in Germany press looked like. This imagi-
of movable metal type in association with the hand- native reconstruction is lo-

operated printing press - developments that are cated in the Gutenberg Mu-
seum, Mainz.
generally credited to Johannes Gutenberg (1390s-
Gutenberg's 42-line Bible,
1438). Metal type was set by hand until the intro- set in Gothic type, was com-
duction of various systems of mechanized typeset- pleted in 1 456. The text is St

ting in the 19th century. The linotype machine was Jerome's Latin version. It is
introduced towards the end of the century, and as close as possible to manu-
script style, with illuminations
became standard newspaper offices. Techniques
in
added by hand. There are no
of photocomposition became a commercial reality page numbers, title page, or
in the 1950s. The latest development, computer- other identifying marks of its
ized typesetting, has been in use since the late printing provenance. Title
pages were not common
1960s'.
until the end of the 1 5th cen-
tury.

Typing
The idea for a typewriter has been known terity is not linked to letter frequency - for

since the 1 8th century, but the prototype
ofmodern machines was not constructed
until1867 (by the American inventor,
Christopher Latham Sholes (1819-90)).
Typewriters have the advantage over
everyday handwriting, in terms of clarity
and speed, but they lack the range of typo-
example, the two strongest fingers of the
right hand are used for two of the least fre-
quent letters,/ and k.
Why this arrangement was chosen is not
clear. Some letter separations are moti-
vated by the need to avoid key jamming
(e.g. q and u) But there is no simple princi-
.

graphic contrasts available in printing (e.g. pie. The second line has a largely alphabe-
variations of type size and shape, justified tical arrangement. The top line, according

setting (p. 1 83), or the provision of many
special symbols). However, modern word
Drocessors and electronic typewriters have
overcome some of these disadvantages,
offering a range of typographical options.
The standard QWERTYUIOP keyboard
arrangement dates from the first
machines. Its design has attracted gene-
to one story, contains the letters of the
word typewriter, so that salesmen could
find them easily when demonstrating early
machines!
As a result, many alternative keyboards
have been designed, to try to improve
speed and efficiency. Most are based on
statistical counts of letter frequency. But all

rations of criticism, on ergonomic grounds, attempts to reform the keyboard have

Although most typists are right-handed, failed because of the vast cost of machine
this keyboard makes the left hand do 56% replacement and typist retraining. The old
of the work. Of all movements for succes- layout is now standard in computer and
sive letters, 48% use only one hand word processor keyboards, and seems
nstead of two - most noticeable when typ- likely to remain so.

ng such words as addressed. Finger dex-

riE B BB B E BE S E ED B (Ml
nEBBBEEEEHEnBtgl
William Caxton (c. 1422-1491) Caxton, the first printer in b0 E H E E E H E m a gn L»]EEEBEEEBBESnr
England, is seen presenting the Duchess of Burgundy with
a copy of his first book, the Recuyell of the Historyes of Troye
H EBEEEEHEEGDmi^J
s- ||S~
LEBHEEEEEEBBBn^
||Z||X||C||V||B|| N ||M||.|H| ?
|
|
la. || HqIIjIIkIIxIIbIImIIwIIvIIzII^ |

(1 475). By the time of his death, he had published around f-ftfmmff^ Scans*- SOK.B.- LOO
||
100 items, including Chaucer's Canterbury Tales and M'aLIJi |

Malory's Morte Darthur.

192 •
V THE MEDIUM OF LANGUAGE: WRITING AND READING
Electronic media range of resources into another in which the range Graphic translatability
is different. A simple example of the pro-
The advent of advanced technology in the 1970s Manyof the graphic displays and character sets, blem of graphic translatabi-
has led to dramatic developments in the field of especially on the cheaper word processors and lity, using a standard format

information services and design. It is now possible computer printers, provide a very limited range of for identifying the various le-

to transmit graphic data from a central source to characters. However, technological developments
vels of biological classifica-
How can the typeface
a TV screen - a system known as teletext (or tele-
tion.
have been so rapid that little study has been devoted contrasts that identify these
tex). This approach is known primarily through to the way these constraints affect our ability to levels be presented on a vi-
its use by the broadcasting services, where it deals communicate There are several features
effectively. deo display unit, given the
with such topics as news, weather, and sports of written language data that it is difficult or impos- graphic restrictions of that
results. In Britain, the BBC service
(begun in 1974) medium?
sible to display on a screen at present. The size
is known as Ceefax, and that of the Independent of the screen, for example, limits the amount
Broadcasting Authority as Oracle. that can be displayed in a single line, usually to Subkingdom VERTEBRATA, or Vertebrates.
Class Mammalia, or Mammals.
Teletext systems are currently limited to a few 40 or 80 characters. There are therefore immediate Order Rodentia, or Rodents.
hundred pages, and the viewer cannot as yet 'talk design problems for anyone wishing to transmit Family MURID&, or Rats and Voles.
Genus Mus, Rats and Mice.
back'. By contrast, the service known as viewdata tables of data where the rows are of greater width. Genus Microtus, Voles.
is interactive. In this approach, computer-stored There are many other problems of 'translation'.
information is sent through the telephone system, How can one compensate for the absence of bold- Prestel
but viewers are allowed to return data to the com- face or italic characters (§31), which are such a
The ordinary set of 96 char-
puter, thus permitting a wide range of services, such routine feature of modern printed language (in dis- acters available to Prestel
as shopping and travel booking from home. In Bri- tinguishing the different kinds of information in users. It can be compared
tain, the trade name of this service, started by Bri- a printed bibliography, for example)? How should with the basic set of 275
tish Telecom in 1979, is Prestel. characters used in printing
such notions as emphasis be conveyed — through
(p. 190).
underlining, capitalization, spacing, flashing lights,
GRAPHIC TRANSLAT ABILITY colour? And if colour is available, how might it
The typographic implications of these develop- be best used in graphic communication (for exam- PRESTEL 'CO' CHARACTER SET

ments are far reaching. Modern printing systems ple, in highlighting the rows or columns of tabular Roman small letters a-* 26

are extremely flexible in their ability to handle and matter)? Above all, what is the communicative Roman capitals A-Z fc 2?

To what
extent can this flexibility be Numerals 1234567890 10
display text. potential of moving graphic symbols on the screen
Fractions V,*. 3
maintained in the graphic displays produced by the — a dimension not available to other kinds of writ- Punctuation .. "<>—-X7I 13
electronic media, given the current limitations of ten language? There are many such questions, each Commercial signs CS'iHI? 5

the technology? This problem has been identified Others »+-*<>•»«

of which requires careful research if satisfactory
11

as one of 'graphic translatability' - the conversion

Space 1

answers are to be found. TOTAL 96

of graphic expression from a medium with one

Dot-matrix characters
TV receivers generally dis- proving this situation. It is the Methods of filtering have
play graphic characters us- result of a 1 983 study under- alsobeen devised which eli-
ing an electronically pro- taken at the Institute for Per- minate the problem of verti-
duced array of dots within a ception Research in Eind- cal 'stepping' (the effect
matrix of tiny squares. hoven (Netherlands) using which produces jagged
These 'dot-matrix' char- matrices of 1 2 x 10 ele- edges to curved char-
acters can be seen in video- ments. Several versions of acters): different tones of
tex, electronic games, and each character were de- grey are used to achieve the
personal computing. How- signed and their legibility impression of curves ('grey
ever, the limitations of the tested in recognition experi- scaling'). This method (be-
equipment have meant that ments. The designers claim low) was first seen in public
the graphic characters are that the larger matrix pro- during the BBC's General
often represented in a fairly duces characters that are Election programme, in
coarse manner - a matrix of more readily identifiable, dis- June 1983. (From N.E.
6 (horizontal) x 10 (vertical) criminable, and acceptable. Tanton, 1984.)
elements can produce only (From F. L. van Nes, 1983.)
a schematic approximation
of printed type. Legibility dif-
ficulties are thus common,
especially when there is

poor contrast with a picture

background (as in many film

sub-titles), or when char-

acters are read from a dis-
tance (as with teletext,
which tends to be watched
at the same distance as nor- Above: Four dot-matrix confi-
mal TV programmes). gurations for the letter a that
The array of symbols were used in the discrimina-
(right)shows one way of im- tion experiments.

32 GRAPHIC EXPRESSION- 193

 . . . . .

33 Graphology

Graphology, in its linguistic sense, is the study of graphs seems to be dictated by factors that are little Language differences
the systems of symbols that have been devised to understood, such as fashion, prominence, elegance, Languages sometimes differ

communicate language in written form. It must be or personality (§32). in their choice of allographs
Graphology also makes use of the notion of dis- to mark linguistic units.
clearly distinguished from the psychological sense
of the term, which refers to the study of handwrit- tinctive features (p. 162).A grapheme is perceived Word classes
ing as a guide to character and personality (p. 189). as a single configuration, or gestalt, and not as a There are several differ-
It also needs to be seen in contrast with graphetics, set of lines and dots; but it is nonetheless possible ences between the personal
the study of the physical properties of manuscript, to analyse the shapes into their components, to pronouns:
English you.
print,and other forms of graphic expression (§32). determine what the salient parameters of contrast /,

German ich (I), Sie (you).

Linguistic graphology is an abstract study (as is are -
curve vs straight line, presence vs absence French ye (I), vous (you).
itscounterpart in the study of speech, phonology, of dot, left-facing vs right-facing curve, and so on. Spanish yo (I), Vd. (=usted),
§28), dealing with the kind of elements used in In French, accents are contrastive (as in (e), (e), Vds. (=ustedes) (singular

a language's writing system, the number of ele- and (e)). In Chinese and Japanese, the contrasts and plural forms of 'you').
ments there are and how they interrelate, and the are carried by the strokes that constitute the charac- Nouns in German all begin
rules governing the way these elements combine ters. However, no general typology of distinctive with a capital letter:

graphological features has yet been established.

English the lamp, a hammer
in written texts.
German die Lampe, ein
Hammer
GRAPHEMES Functional differences
The term 'graphology' was coined on analogy with The analogy between graphology and phonology Days and months
is important, but there is no identity of function
English Monday, Tues-
'phonology', and several of the phonological
day.
(§31). Graphemes may signal phonemes, but they
.

notions used in the study of speech have also been French lundi, mardi . .

applied to the study of written language. In particu- may also signal words or word parts (as with the Spanish lunes, martes . .

lar, the idea of a grapheme has been developed, numerals, where each grapheme (1), (2), etc. is
English January, Febru-
analogous to phoneme (p. 160). Graphemes are spoken as a word that varies from language to lan- ary...
the smallest units in a writing system capable of guage). Graphemes of punctuation show links and German Januar, Februar. .

causing a contrast in meaning. In the English alpha- boundaries between units of grammar that may Spanish enero, febrero . .

have nothing to do with the sound of speech (nota- French Janvier, fevrier.
bet, the switch from cat to bat introduces a meaning
.

change; therefore, c and b represent different gra- bly, the use of the (p. 205)). And several
hyphen
Language names
phemes. It is usual to transcribe graphemes within of the morphological relationships between words English / speak Portu-
angle brackets, to show their special status: (c), (p.90) are conveyed by graphology more clearly guese
(b). The main graphemes of English are the 26 units than phonology: for example, the link between sign Portuguese Falo portugues.
that make up the alphabet. Other graphemes and signature is closer in writing than in speech
Questions and exclama-
include the various marks of punctuation: (.), (;), (where the g is pronounced in the second word,
tions
etc., and such special symbols as (@), (&), and but not in the first), and the same applies to such In Spanish, question marks

<£). sets as telegraph/ telegraphy/ telegraphic, where and exclamation marks are
Graphemes which may adopt
are abstract units, there are several stress and vowel changes in used both at the beginning
and at the end of a sentence,
a variety of forms. The grapheme (a), for example, speech, but none in writing.
the first one being inverted:
may appear as A, a, a, or in other forms, depending
I Como se llama este pueb-
on the handwriting style or typeface chosen. Each lo? 'What is this village
of these possible forms is known as a graph (cf. called?'

phone in speech). There is a vast amount of physical /Que dial What a day!'

variation in the shapes of graphs that does not

affect the underlying identity of the grapheme.
Whether a word is printed cat, CAT, cat, or even SPECIAL WRITING SYSTEMS
caT or cAt, we still recognize it as a sequence of Graphological studies have Orthography The writing Paedography A system
three graphemes (c), (a), (t). a functional as well as a system in standard, devised to help children to
formal dimension. Within a everyday use, which read, as in such alphabets
language, several kinds of consequently attracts most asi.t.a. (p. 217).
Variants and features
system may be invented in study.
When graphs are analysed as variants of a gra- order to perform a
Technography A system
Stenography A system that enables a specialized
pheme, they are known as allographs (analogous specialized set of
that enables writing to take perform its function,
field to
to allophones, p. 160). sometimes possible to
It is
functions. One
place at speed, as in the such as phonetic
classification recognizes
work out the rules governing the use of particular many systems of transcription (p. 158),
five types, differentiated
allographs: in English, for example, we find 'capital according to the purpose
shorthand (p. 206). chemical notation,
letters' at the beginning of a sentence or proper cartography, or computer
for which they were Cryptography A system
coding.
name and in a few other contexts; otherwise, 'small devised (after J. Mountford, devised to keep a written
letters' are used. However, the choice of most allo-
1973): message secret (p. 58).

194 •
V THE MEDIUM OF LANGUAGE: WRITING AND READING
 .

Hiragana
Kar>|i
Punctuation marks
Graphology at breakfast
A comparative study of writ- There are three hiragana
ing systems usually deals symbols in the headline, as
with different languages; but shown.
it is important to remember • Loan words (from lan-
that different systems may guages other than Chinese)
coexist within a single lan- are expressed in the angular
guage. In Yugoslavia, Ro- katakana syllabic script
man and Cyrillic scripts are (p. 201 ). Immediately below
both used for Serbo-Croat. In 'Big Chance', for example,
Chinese, there is now the we find a syllable-by-syllable
use of the romanized alpha- translation' of the English
bet pin-yin alongside the use 'Stoneware Morning Set',
of characters (p. 31 2). In Ja-
pan, however, four writing
systems are in regular daily
use (five, if arabic numbers
are counted as a separate
system), as can be seen
from the back of this 1 985
cornflakes packet:
• The English alphabet is
noticeable in the brand
names, the vitamin letters,
and the gift advertisement
(advertising is a variety in

which English script and loan

words are very common,
p. 390).
• The more complex-looking
characters are kanji logo- Stoneware
grams (p. 200), derived from Morning Set
Chinese. For example, the
large characters at the top
right of the illustration read
respectively (from top to bot-
tom): 'morning' + 'food' +
trms* i

'newspaper' - in other
words, 'breakfast news- 348*4 Ittfriiihrrnn:
J.**.\<r:
paper'. The same two 'break- eik»s* xiM-9i*a«
fast' graphemes occur at the which comes out as (allow-
beginning of the headline ing for long vowels and con-
(reading from left to right): sonants) su-to-o-n-u-e-a m»mm «tteo4iifi3oa'±
the typeface differences are mo-o-ni-n-gu-se-t-to. The
worth noting. translation of Kallogg's is
»«*>.» «i[3»4)U6B30
• Several graphemes from itself interesting: Kellogg
is a

the rounded hiragana sylla- loan word, so it katakana

bic scriptcan be seen on the
page. They are used for var-
ious particles that express
grammatical distinctions.
is in tHf7\sd>Kr>fl&irbr>Xt>-i.
(ke-ro-g-gu), but the 's is a £ST
particle, so it is in hiragana
(the symbol o>no expressing
possession).

The Somali problem

One aims of the
of the stated Ethiopic elements. It is others are arbitrary inven- inthe ways of representing any 1 973 a Latin script was
1969 revolution Somalia in drawn (right) with accom- tions, displaying the general long vowels (not shown finally adopted, and given
was to solve the problem of panying Latin equivalents influence of Ethiopic scripts, here). official status.
which writing system to use (from D. Diringer, 1968). Arabic influence is found in The Osmanian alphabet
for the country's main lan- The role of Latin script, the order of the letters, and was not successful. In Janu-
guage, Somali. mainly deriving from Italian
For several years, the influence in the area, is evi-
question had been highly dent in the way vowels as
controversial, with the merits well as consonants are sym-
c
of Latin, Arabic,
nian scripts all being advo-
and Osma- bolized (unlike
and also
in Arabic),
the left-to-right
in
>
b t 3 li h d n J / d g f
cated. The last of these was direction of the writing. Some
named after its early 20th- letter shapes are taken over
century inventor, Osman Yu- from the Latin alphabet,
suf, and shows an interesting though they are not given the
mixture of Arabic, Italian, and same phonetic values; rrt 7? jv
y u o

33 GRAPHOLOGY 195
 shows the relationship between some of the tokens
The history of writing and the incised characters that appear in the earliest
Myths and legends of the supernatural shroud the Sumerian tablet inscriptions (not all of which can
early history- of writing, as they do of speech (§§3, be interpreted). The similarity between the three-
49). Archaeological discoveries provide enthralling dimensional tokens and the two-dimensional
pinholes of illumination along with frustrating pro- inscriptions is striking. (After D. Schmandt-Bes-
blems of interpretation. An account of the early serat, 1978.)
history of writing has gradually emerged, but it
contains many gaps and ambiguities.
The matter is complicated by the fact that, in
this early period, it is by no means easy to decide
whether a piece of graphic expression should be
counted as an artistic image or as a symbol of primi-
tive writing. In principle, the difference is clear: Primitive picture-writing
and subjective mean- Exploration has revealed (i) Coloured river pebbles from the Azilian
culture of
the former convey personal
ings, and do not combine into a system of recurring many primitive pictures and southern France
signs that resemble writing,
symbols with accepted values; by contrast, the lat-
but that lack the systemati-
ter is conventional and institutionalized, capable zation we expect of a writ-
of being understood in the same way by all who ing system. Human figures,
are using the system. When the product is a rock geometric signs, and other (ii) Geometrical signs on stones found in Spain
shapes have been found
carving or painting of an animal, there is little
carved or painted above
doubt that its purpose is non-linguistic (though and below ground on rocks
whether it has an aesthetic, religious, or other func- ('petroglyphs'), buildings, (iii) Various patterns found in California
tion is debatable). However, when the product is tombs, pottery, and other
a series of apparent geometrical shapes or tiny
characters, the distinction between art
becomes
flect the
less obvious. The languages may even
problem: in early Greek, and in Egyptian,
the same word was used for both 'write' and
'draw'.
and writing
re-
objects in many parts of the
world. Their significance is
generally unknown. (From
D. Diringer, 1968.) X l ^ &>* ^ f%
One point, at least, is fairly clear. It now seems BC, appear in many differ- cised characters in the ear-

most systems evolved indepen-

likely that writing
Clay tokens ent shapes. Some of the liestSumerian inscriptions
These tokens from Susa, commonest shapes are (only some of which have
dently of each other at different times in several dating from around3000 here compared with the in- been interpreted).
pans of the world — in Mesopotamia, China, Meso-
America, and elsewhere. There is nothing to sup-
port a theory of common origin. There are of
course similarities between these systems, but these
are not altogether surprising, given the limited
ways of devising a system of written communica-
tion.

Precursors
The earliest examples of a conventional use of writ-
ten symbols are on clay tablets discovered in vari-
ous parts of the Middle East and south-east Europe
from around 3500 bc. Large numbers of tablets
made by the Sumerians have been found in sites
$ )
around the Rivers Tigris and Euphrates in present-
day Iraq and Iran. For example, on tablets from

D e
the city-state of Uruk, about 1,500 symbols have
been listed, most of them abstract in character.
They seem to have recorded such matters as land
»
NUMERAL 1
t>*

<$
sales, business transactions, and tax accounts.
Several correspondences have been noted be-
tween the symbols used on these tablets and the
clay tokens that were used throughout the area for
d> cow
several thousand years before the advent of writing. OIL GRANARY

These tokens, of several distinctive shapes, seem

to have been used as a system of accounting from
at least the 9th millennium bc. A selection of tokens
from Susa, dating from the end of the 4th millen-
..

HEART.
WOMB
^
ANIMAL'
(UNIDENTIFIED)
ffi m c WOOL
nium bc, is shown right. The adjacent diagram

196 •
V THE MEDIUM OF LANGUAGE: WRITING AND READING
Types of writing system They have been discovered in Egypt and Mesopota-
mia from around 3000 BC, and in China from
It is possible to talk about writing systems on the around 1500 bc.
basis of such graphetic factors (§32) as the size,
style,and configuration of the symbols, or the di-
rection in which they are written; but this does
not help us to understand what the graphemes are
Pictographic symbols
and how they are used. In principle, any of the
Below: Some of the picto- 1 968). Over 1 00 symbols recognizable, showing that
systems to be described below could be written in
graphic symbols used on represent human figures, there has been some deve-
almost any set of graphetic conventions. Some- seals and tablets in the body parts, animals, and lopment towards an ideo-
times, for example, several directions are used dur- early Minoan period in other everyday objects. Not graphic system.
ing the history of a language, as in early Greek, Crete (from D. Diringer, everything is immediately
which was written right-to-left,
at different periods
left-to-right,and even using alternate directions (in **>
boustrophedon writing, p. 185). jj
A more useful approach to writing systems is
to classify them into cases that show a clear rela-
tionship between the symbols and sounds of the £r><>
language {phonological systems) and those that do
Below: A drawing of one of trophedon, p. 185), with the is clear (e.g. birds, fish), but
not {non-phonological systems). The vast majority
the wooden tablets carved alternate rows inverted: the others cannot be inter-
of present-day systems are phonological; the non- with symbols, found on reader has to turn the tablet preted. The script as a
phonological systems are mainly found in the early Easter Island (from D. Dir- upside down at the end of whole has not been deci-
history of writing, which is where we begin. inger, 1968). The direction each line. The pictographic phered.
of writing alternates (bous- character of many symbols

NON-PHONOLOGICAL SYSTEMS
PICTOGRAPHIC
In this system, the graphemes (often referred to
as pictographs or pictograms) provide a recogniz-
able picture of entities as they exist in the world.
For example, a set of wavy might represent
lines
the sea or a river, and outlines of people and ani-
mals represent their living counterparts. There is
no intention to draw the reality artistically or
exactly, but the symbols must be sufficiently clear
and simple to enable them to be immediately recog-
nized and reproduced as occasion demands as part Below: Drawing of a Zapo- being met by a Zapotec of the latter is a place
of a narrative. tec inscription from the old ruler. Their names (indi- name, (e) 'Hill of 1 Jaguar',
To 'read' such a script, it is enough only to recog- Monte Alban in
city of cated by the shaded sec- and (f) the name Za-
of the
south-western Mexico tions) are (a) '13 Knot', (b) potec ruler, which is not
nize the symbols, and the sequence may then be
(from J. Marcus, 1980). '9 Monkey', (c) '1 Owl', and fully decipherable.
verbally described in a variety of ways, in whatever
Four visitors to the city are (d) Treble Scroll'. In front
language one happens to speak. There is thus a
great deal of possible ambiguity when it comes to
reading sequences of pictograms, and many of
these scripts have proved difficult or impossible to
decipher. The problem can be illustrated with a
modern pictogram, such as the road sign (right).
Without knowing the context, the sign could be
'read' in all kinds of ways - someone has been/will
be/is digging/clearing/stopping a landslide - or
Below: Some modern pic- two cultures is instructive. safe for the mountain goat,
even (as was discovered in a competition to find tographic road signs, Inthe one case, the road but not for the rider. The
the most absurd road-sign interpretation) strug- alongside an Indian rock leads to water, and vehi- parallel would have been
gling to put up an umbrella on a windy day! drawing from New Mexico. cles should take care; in even if the modern
closer,
The parallel between the the other, the rocky trail is sign had contained a fish!
Modern drivers know the likely context, so ambi-
guity is uncommon. When we are studying 5,000-
year-old pictograms, the likely context may not be
known. The many undeciphered or partly-deci-
A Forton

A
phered pictographic scripts of ancient Crete illus- Services 2m 1

trate the size of the problem (cf. the Phaistos disc,

p. 183).
Pictograms constitute the earliest system of writ-
ing and are found in many parts of the world where Petrol •• iari c

the remains of early people have been discovered.

33 GRAPHOLOGY- 197
IDEOGRAPHIC Today's ideograms
Ideographic writing usually distinguished as a
is
Early ideograms Married love, with
pillows for head and Modern signs are frequently
later development of pictographic. Ideograms, or
Some Hittite ideograms, feet (a sign of ideographic, as with the dia-
used in the 2nd millennium
gonal lines used to express
ideographs, have an abstract or conventional wealth).
BC prohibition (e.g. no right
meaning, no longer displaying a clear pictorial link Quarrel between
turn).Signs such as 'no dogs
with external reality. Two factors account for this.
^Jox husband and wife (a
allowed' and do not iron' mix
pillow is between
The shape of an ideogram may
so alter that it is pictograms and ideograms.
vthem).
no longer recognizable as a pictorial representation n_. stooi

of an object; and its original meaning may extend A woman with six

to include notions that lack any clear pictorial

a city children, a husband,
and a pillow.
form. In early Sumerian writing, for example, the @god
picture of a starry skycame to mean 'night', 'dark', y A man comes to a
woman who has a
^king Q
came to
,

or 'black'; a foot represent 'go', 'stand', Vj>

t
husband and asks
and other such notions. ^=3 river J her to live with him.
It israre to find a 'pure' ideographic writing sys- "V* Three men seek the
tem - that is, one in which the symbols refer directly I speak
H/
-1
same married wo-
4

to notions or things. Most systems that have been ./p» man.

<f% great
called ideographic in fact contain linguistic ele- (From D. Diringer,
(From D. Diringer, 1968.)
ments. The symbols stand for words in the lan- 1968.)

guage, or parts of the symbols represent sounds. Below: Ideograms from Below: Seal inscriptions
Nsibidi, a system disco- from the Indus Valley,
The Sumerian, Egyptian, Hittite, and other scripts
vered in southern Nigeria in north-west India. The writ-
of the early period were all mixtures of picto-
1 904. Most of the signs ex- ing system has not been
graphic, ideographic, and linguistic elements. press a range of marital sit- deciphered, but it is
uations and relationships, thought to contain a mix-
CUNEIFORM as in the following exam- ture of ideographic and
ples: phonetic graphemes, re-
The cuneiform method of writing dates from the
presenting proper names.
4th millennium bc, and was used to express both (FromD. Diringer, 1968.)
non-phonological and phonological writing sys-
Married love.
tems in several languages. The name derives from
the Latin, meaning 'wedge-shaped', and refers to
the technique used to make the symbols. A stylus
was pressed into a tablet of soft clay to make a Married love, with vmini)
sequence of short straight strokes. In later periods, pillow.

harder materials were used. The strokes are thick-

est at the top and to the left, reflecting the direction
of writing: at first, symbols were written from top
to bottom; later, they were turned onto their sides, (1°
C3 f(7
* U.n
and written from left to right.
The earliest cuneiform was a development of pic-
tographic symbols. Subsequently, the script was
J> &
(7

#^ m Itud
">»vt*>
used to write words and syllables, and to mark
phonetic elements. It was used for over 3,000 years
t «= *T •*» to ipu>

throughout the Near East by such cultures as the

Sumerians, Babylonians, Assyrians, and Hittites,
^ t> f*
too J

finally dying out as the Christian era approached. <p §? WW *% but

The latest cuneiform tablets date from the 1st cen-
tury bc. The script could not be read until the 19th
century, when several of the languages it repre-
^ 1 ff ff
-jttT
in

sented were finally deciphered. <& d" $m «J*

to

The columns right show a series of original picto-

grams, first vertically, then in the altered position ^ & &7 W Co
f
#

used by later cuneiform. Two versions of cuneiform

are shown - an illustration from the early period,
^~ r <H #f bird

and one from the later period, when simplified sym-

bols were introduced bv the Assyrians (after D.
&> 4 * W-
fish

Diringer, 1968).
V x> %> yt- OX

v" 4> <*= COW Part of the Black Obelisk of

Shalmanazer m (858-624
b*rky discovered Nimrud.
f **}- tm®- *•
bc), at
The monument is carved
-O 1
Jun with scenes showing tribute
i *> 1 dty bearers and Assyrian cunei-
form texts.

198 V THE MEDIUM OF LANGUAGE: WRITING AND READING

EGYPTIAN HIEROGLYPHIC gram) R to represent the word beer, or followed Royal names
In Egypt, aform of pictography developed around by K to represent the word beak. In hieroglyphic, Two Egyptian royal names,
3000 bc, which came to be called hieroglyphic this convention was used to express two-consonant in hieroglyphic, showing (a)

(from the Greek 'sacred carving'), because of its sequences as well as single consonants. the royal-divine names, (b)
the personal names.
prominent use in temples, tombs, and other special miiii
/w»vw\ n
i i m-n m-s
places. The term has also come to be used for scripts 111 Amen-em-het (reigned
I

of a similar character from other cultures, such as
the Hittite, Mayan, or Indus Valley; but the most
fully developed system of hieroglyphic writing is
undoubtedly the Egyptian. The system continued
in use for three millennia, until it was finally re-
• Determinative symbols are signs that have no 1991-1 962 bc)
phonetic value but are placed next to other symbols
to tell the reader what kind of meaning a word
has. Words that would otherwise appear to be iden-
tical could thus be differentiated. An analogy might

placed by the Coptic-based script of the early Chris- again be drawn with a word game in English that
tian era. could distinguish the two senses of the word table
The units of the writing system are known as by adding a chair (for the item of furniture) and -& <

hieroglyphs. They tend to be written from right an eye (for the typographical arrangement). Egyp-
to left, with the symbols generally facing the begin- tian symbols that were commonly used as determi-
ning of a row; but vertical rows are also found, natives included the following:
Below: Alexander the Great
following the line of a building. The script gives
sun, sun god, daytime "»
)\ moon, month (356-323 bc).
the general impression of being pictorial, but in
fact it contains three types of symbol that together
^Ifstar, hour, time to pray C3 mountain
represent words: (^ city, town see

• Some symbols are used as ideograms, repre-

senting real-world entities or notions:
Jj
pray, adore, praise ^ weep, grief
.as
A combined example
;2r^ swallow y\so The hieroglyph ^ is composed of two elements:

beetle

eat
^find
fresh
v^" which
and
for
is the ideographic symbol for 'wood',
# and a which are the phonetic symbols
,

h and t respectively. Combined, these would

represent the word htl. However, htl had two
Q v^ a
meanings: 'carve' and 'retreat'. These are therefore
• Some symbols (phonograms) stand for one or distinguished by the addition of a determinative: The names are surrounded
by a ring, or cartouche,
more consonants, in much the same way as the the addition of a knife symbol produces ©^ TS*s
which was intended to pro-
rebus system isused in present-day children's and the addition of a pair of legs walking
'carve', tect thebearer of the name
games (p. 65). For example, in English we might backwards produces !T~r\., 'retreat' (after H. Brun- from harm. (From D. Dir-
use a picture of a bee followed by the letter (phono- ner, 1975). inger, 1968.)

A hieroglyphic inscription
from Thebes, 1 8th Dynasty
The Rosetta stone
(c. 1 490 bc). The statue is Egyptian hieroglyphic re-
of Sennefer,Chancellor mained undeciphered until

and Superintendent of the the 1 9th century. Members

Palace at the time of of Napoleon's Egyptian ex-
Hatshepsut or pedition of 1 799 discovered

Tuthmosis ill. a black basalt stone,

The inscription measuring 1 1 4 x 72 cm, at
consists of a Rashid (Rosetta). The stone
prayer to Osiris, was carved with three differ-
a list of Sen- ent scripts: hieroglyphic, the
nefer's positions derived demotic script, used
and achieve- for everyday purposes, and
ments, and a re- Greek.
quest for a The Greek version could
happy afterlife. of course be translated, and
The statue, of thisprovided the key to the
black granite, is other texts. The stone com-
87 cm high. memorates the accession of
Ptolemy V Epiphanes
(205-1 80 bc). It thus con-
tains several royal names,
whose phonetic values could
be related to those of the
names in Greek. The text
was finally published in 822 1

by the French Egyptologist

Jean-Frangois Champollion
(1790-1832). The Rosetta
stone is now in the British
Museum.

11 GRAPHOLOGY •
199
LOGOGRAPHIC
Logographic writing systems are those where the Chinese characters same phonetic indicator. • chuanchu Characters
The meaning of the ma formed by modifying the
graphemes represent words. The best-known cases Traditionally, Chinese character when used alone shape or orientation of a
are Chinese, and its derivative script, Japanese characters are divided into ('horse') is disregarded. character to produce a
six types (liu shu 'six
kanji (pp. 195, 313). The symbols are variously- word of related meaning,
scripts'). t% mother' ^ scold'
referred to as logographs, logograms, or - in the e.g. the character for
• hsingsheng Most
case of oriental languages — characters. But there corpse' F derives from that
characters are of this type, • chihshih These char- for man'A-
are two terminological complications. First, be- containing two elements. acters represent abstract • chiachieh Characters
cause Chinese writing derives from an ideographic There is a semantic ele- ideas and are closest to were borrowed from
that
ment, known as a radical' ideograms, e.g.
script,with several pictographic elements, the char- others of similar pronuncia-
(similar to the determina-
acters are commonly referred to as ideographs. Character English tion, e.g.^ wan ten thou-
tives' of hieroglyphic,
However, this term is really not appropriate, as p. 199). This is combined ^ 'middle'
'
sand' derives from the use
of this character for wan
the characters refer to linguistic units, and not di- with a phonetic element, X: lar 9 e
scorpion'.
rectly to concepts or things. Secondly, the char- whose function is to remind small'
the reader of how the word
'J> • hsianghsing A small
acters in fact often represent parts of words one' group of characters that re-
is be pronounced.
to
(morphemes, p. 90) as well as whole words, so For example, the word • huii Compound char-
tain a close connection with
original pictograms, e.g.
that even the term 'logographic' is slightly mislead- mother' ma is expressed acters which the ele-
in
the forms for (a) sun', day'
ing; but in the absence of a more appropriate term by the semantic element ments have a semantic
(//), (b) mountain' (shin) and
(such as 'morphographic'), it continues to be used. woman' -j£ followed by a connection, e.g.
(c) 'field' (f/an).
Several thousand graphemes are involved in a
phonetic indicator ma .
ij sun - moon' = bright' b#
The word for scold' is also 'woman' + woman' =
Ancient Modern
logographic svstem. The great Chinese dictionary ma (with a different tone, 'quarrel' $& form form
of K'ang Hsi (1662-1722) contains nearly 50,000 p. 172), and this is ex- 'man' + 'man' + man' = (a)
B
characters, but most of these are archaic or highly
specialized. In the modern language, basic literacy
pressed by the semantic
element mouth' c7 (re-
'crowd' & (b)
^ ti.

peated) followed by the (c) El IS

requires knowledge of some 2,000 characters.
Similarly, in Japanese, 1,850 characters are pres- — l
b 31
ic 61 •^jp 91 /f 12 ' % 161 £ 181 ^t 211 £241 Hi 27)

cribed by the Japanese Ministry of Education and

adopted by law as those most essential for everyday
I » J.
32
yL 02 cfj 92 ^> 122
firr 152 J? 132 ^212 £ 242 5^272

use. Of these, 881 are taught during the six years

» 3 77 S3
ft 03 'H- 93 <tc. 123
171] 153 ifc 183 tU 213
5fc243 ^273
M Tj-w it W >t
of elementary school. ii
-
124 _g_154 £<184 ^214 ^244 K 2T4

Most languages make use of some logograms: r • ^7 30

Tf 65 IB 96 ^125 § 155
X ,85
3. 215
# 245 a
ill*
B7»
a selection of widely used graphemes is given
below.
7^ 36 f~ 00 ^\ 126 g 150 ^"186 g 216 JP, 24C Mi 2:s

Li ^37 S. 97
/j\
127
E 15 7
& "7 3C 247
i. 277

X38 tL m TL »8
Jf 128
f/g
168 c|l 188
[5 2l8 ||] 248 ^278
-.
Right: Chinese and Japanese characters are classified on the
basis of the number of strokes used to write them. The
j X 30
S » 7K 120
J^ 169 Tfc 189
O 219
^ 249 ||1 279

increasing order of graphic complexity can be seen in this

<
"1"
">
P 40
Fl 70
7L 100 ^130 ^•160 yfcm § 220 ^. 250 gf 280

list of 300 primary characters in Chinese, which may be used b « #101 ^131 jAj 161
f£ 191
$fl
221
individually or
1968.)
as part of compound forms. (From D. Diringer -Z.12 Jjj 42
P " ft 102
/ft
132
£ 102 $- 192 £3 222 ^« ^ 282
J_is T« 73
#M 03 ^133 pFj 1C3
^ 193 ^ 2230253 3 233
\u fQ74 /t 104 ^134 7ft 164 £ 194
f§J W4 ^254 Sg2*
Modern logograms A 1&
f-45 A «* [XJ 135
^ I6S IT. 195
^ M6 H.255 ,^ 285
These symbols are widely used in modern written A. 16 -=_46
E" x B 106 13ti
^ 166 g l'JO
®
«• ^ 28C
languages. Their spoken equivalents, of course, vary
from language to language. The most developed
E« ^107 B 137
^w nn 197 ^227 S. 267 5 287

logographic systems are found in scientific notations,

JL 18 <«« 2,78
~7f
108
R l38
7K I 68
rfn 1SW F# 22 8 J| 258 ^ 28?
such as in logic and mathematics (p. 381 ). ^49 T^ro fy)
109 £ 139 ^169 @1W |^229 ^ 259 *^289

+- X /L'20
7L 21 A, 61
±.r 86

'J\ 81
-^
£
t»«
n0

Tfc hi
/£
•^
170

171
[fj 200

g 201
t230
231
fj
3^261
200
R
K7
200

291

-f-a >7] 62 82 ^112 £"142 i^ 172 fSj 202

M 232 -^•262 ^ 2K;

t* A 63 rfl 83 ^ 113 ;T£ 173 [^] 203 ^233 M. 263 ^

^±r
293

n ^6424 Lp *4 ^S 114 Til 144 ^ 174 £3 204

If
234 :§264 ^294
r—»25 ^56 lL 85 ^. 115
JL 145 ^ 175 -fef-
205 ^235 ^.265 ¥«*
u* /? 66
m * -ij- 116 Bg u6 ^ 170 f^J 200
M 23B 3jE 260
ti 2%
a 27 £-67 ±87 jym few "
T 177 ^g 207 ^2.37 7^267 d 297

A a 5.68 T 88
^ 148
7ft 178 ffl 208 ^ 238 ,?! 2 «
**
?ffr
-
293

f{z)=
1 7
z
_ 1 'e V(r)df+ I
'_ e L» ^59 •=p 89
3f 119
FF? «« 179 ^ 209 ^ 239
r=i
16? ^299
|
-^60 ^ 120 -ffl- 160 180 ^210 # 240 ^ 270 ^ 300

200 V THE MEDIUM OF LANGUAGE: WRITING AND READING

 — »

PHONOLOGICAL SYSTEMS
SYLLABIC Katakana
In a system of syllabic writing (a syllabary), each The Japanese katakana The system contains a few used mainly to write foreign
grapheme corresponds spoken syllable, usually
to a syllabary contains 75 gra- phonetic features, such as words which have come to
a consonant-vowel pair. Such systems have been phemes, three of which the regular use of * to be used in Japanese (other
enter into combinations to mark the voiced element in than those of Chinese ori-
found from earliest times (e.g. Mycenean Greek)
produce a further 36 forms, a contrast. The system is gin):
and in modern times can be seen in Amharic, Cher-
okee, and Japanese kana. The number of gra-
phemes in a syllabary varies — from around 50 to :*-fc- V -V V # J
f
x1 r-

several hundred.
America coffee jumbo jet

t-^F7';t y>X
Cypriot Australia jazz television
The clearest example of a syllabic script in classical
times comes from Cyprus, where it was used from
about the 6th to the 3rd century bc. Typical symbols
are shown below, along with an interpretation of the
T 1
*7

ka
if
ga
9-

sa
If
za
*
ta
/
da
-7-

na
/\
ha
/t
ba
/t
pa
v| 7 7 7r >
ma wa
a |
ra fa n
sound values (from 0. Masson, 1961).
The Cypriot (or Cypriote) syllabary was deciphered A * * x v- f- f — b b b 5 'J 7>f
towards the end of the 1 9th century; the inscriptions i ki gi shi ji chi ji ni hi bi pi mi ri fi

are mostly in Greek, though the script seems to have

been designed for a different language. There is no *| *| 7 X X 7
7'
% 7 7 7 A JV
way of indicating vowel length, several Greek sounds u ku 1 gu su zu tsu zu nu fu bu pu mu ru
cannot be distinguished, and syllables containing two
consonants have to be expanded as two syllables (e.g. J. y Y -fe -b T f * 'N "* ^ / 1/ 7x
ptolin — po-to-li-ne), much as modern Japanese has
e ke ge se ze te de ne he be pe me re fe
to do with foreign loan words. The script is mainly
written from right to left. The system may be distantly * 3 =f y / h K / * # * * n 7* 7
related to the early linear script known as Linear B
o ko go so zo to do no ho bo po mo fc fo
(p. 301), which was also largely syllabic in character.
\

*MHr =?> X-V it* f* f* -V b^ b> b-v S* 'J-r

ya kya gya sha ja cha ja nya hya bya pya my a rya

J

a e i U
A 1*3. *x X3. ya fa. fa. :=ol bi h'a bi ^ j. 'Ja

X * X X r yu kyu gyu shu ju chu ju nyu hyu byu pyu myu ryu

3 *3 *3 /3 i>3 + 3 fa —alba baba 5 3 'ja

9 ^* yo kyo gyo sho jo cho jo nyojhyo byo|pyo|myo ryo

1

w X< I >'< f
Cherokee
r 2 fr ^ 9. n This syllabary was in- used by the people and Latin alphabet, but the Latin-
vented in 1821 by a half- missionaries for many ate symbols are not used
Cherokee Indian named years. ts 85 symbols show with their original sounds.
I ^ 8 z. + fa Sequoya, and came to be the strong influence of the (From H. A. Gleason, 1955.)

m
v
>< % ^ CD •X'
J>

i
a

ga
R
h
e

ge
T
y
'
l

1
«*>

A
o

go
CP

J
u

gu
X
E
A

gA

n T 1* 1 V >! & ha
?
he
j\
hi 1
*
ho
r hu
Sir
hA
9 ka

w la
/ le
p 1: 1
(5
lo
M lu
3 1a

* V r <A> \T hna
P s
V ma
Oi
me
H " 1

3 mo •y mu
G nan
na ne
L r ak T F E e A f\
ni 1

Z no
q nu
0- n*
oO 3

T gwa A gwe
?p g^ i
Jvpgwo cD
gwu
& gWA

k 1 * Y n ¥ u sa
4 se
b « 1

* so
V su
K SA
^ ti
A * A s
s V r ± M yk u.
da
f de do du (Ta d*
Ju tla
i
A dla L die G <Jli J* dlo T dlu
P dU
%
z Q dza V dze h» d« K dzo J dzu C= dZA

X )( (H
a wa OS we © w,
t) wo s wu e WA

GO ya 43 ye 4 V R yo (T yu B ya

33 GRAPHOLOGY- 201
 -

ALPHABETIC
With alphabetic writing, there is a direct correspon-
dence between graphemes and phonemes, which
makes it the most economic and adaptable of all
the writing systems. Instead of several thousand
logograms, or several dozen syllables, the system
needs only a relatively small number of units, which
it then proves easy to adapt to a wide range of
languages. Most alphabets contain 20—30 symbols,
but the relative complexity of the sound system
(§28) leads to alphabets of varying size. The smal-
lest alphabet seems to be Rotokas, used in the Solo-
mon Islands, with 11 letters. The largest is Khmer,
with 74 letters.
we and Gaelic, where
find such cases as English
there is a marked degree of The extent
irregularity.
to which there is a lack of correspondence between
graphemes and phonemes is inevitably reflected in
the number of arbitrary 'spelling rules' that child-
ren have to learn (p. 213).
There are also many alphabets where only
certain phonemes are represented graphemically.
These are the 'consonantal' alphabets, such as Ara-
maic, Hebrew, and Arabic, where the marking of
vowels (using diacritics) is optional. There are also
cases, such as the alphabets of India, where diacri-
tics are used for vowels, but the marking is obliga-

In a perfectly regular system, as in some of the
alphabets that have been devised by linguists to
record previously unwritten languages, there is one
grapheme for each phoneme. However, most alpha-
bets in present-day use fail to meet this criterion,
to some degree, either because the writing system
has not kept pace with changes in pronunciation,
or because the language is using an alphabet not
originally designed for it. Languages vary greatly
in their graphemic/phonemic regularity. At one
extreme we find such languages as Spanish and Fin-
nish, which have a very regular system; at the other,
tory, with the diacritics being attached to the
consonantal letters.
The earliest-known alphabet was the North
Semitic, which developed around 1700 bc in Pales-
tine and Syria. It consisted of 22 consonant letters.
The Hebrew, Arabic, and Phoenician alphabets
were based on this model. Then, around 1000 bc,
the Phoenician alphabet was itself used as a model
by the Greeks, who added letters for vowels. Greek
in turn became the model for Etruscan (c. 800 bc),
whence came the letters of the ancient Roman
alphabet, and ultimately all western alphabets.

New alphabets from old

The development of the early alphabet, and the relationship between several modern alphabets.
Classical Modern Greek Hebrew- Arabic-
Phoenician Old Hebrew Early Greek Greek Etruscan Early Latin Roman form name Cyrillic form name form name

< * A A A ^ Aa Aa alpha Aa K
^
aleph. alel i
alif

9 9 8 6 3 Bb B0 beta &6 beth -r'

be
7 1 r > < Cc r> gamma Bb 3 gimel wf la
A. 5 A Dd AS delta Tr 1 daleth w» tha
* »
^ a E 3 <r Ee E€ epsiion Ha n he fim
»
Y t A A /* Ff u zeta Ee *,
vav. waw ha
) A Gg Hr, eta Ee i zaym £ kha
* n B H e u Hh ee theta 5K* n heth J dai
I 1
i i Ii u iota 33 c teth 3 dhai
*r
i J) Kk kappa Mm Hm «
yod. yodh J ra
> 3 * K X K Kk u lambda Kk ="i kaph j zay
i L 1 s\ i I LI Mm- mu .l.i Gj
lamedh Jf sin
i J ~1 n ^1 n Mm Nv nu Mm DD mem J- shin
1 y 1 N m ry Nn =.1 XI Hh J' nun J- sad
O O O o Oo Oo omicron Oo D samekh J- dad
1 1 n TT 1 p PP n* P< nn ay in im ta
} a> Q Q Oq Pp rho Pp e" pe m za
* "\ <\ P 4 P Rr £as Sigma Cc i.

w n *V sade. sadhe 'ayn

w f i > Ss Tt tau Tt P qoph c-
ghayn
ht X. T T X T~ Tt Vv upsilon y> -i resh ,J fa
Y Y V V Uu 4><f phi <j>* & sin i qaf
Vv *-X chi. khi Xx - shin *i kaf
W* ya, psi Uu n lav taw J lam
X — X Xx iliO omega Mm mim
HI in nun
I Z Zz LUui a ha
tb waw
bb j ya
Uu
3i
K)io
Ha

202 •
V THE MEDIUM OF LANGUAGE: WRITING AND READING
Alphabetic scripts Anglo-Saxon times
The range of the world's alphabetic scripts can be seen The earliest English alpha- dern English thin and this) • /a/ (as in modern Eng-
in this selection of extracts from Biblical texts. Other bet was devised by mission- came to be written by a was pronounced
ru- lish hat)
examples are found on pp. 1 88, 303. (From E. aries in Britain, who used nic symbol, known as quite high in the mouth, al-
Gunnemark& D. Kenrick, 1985.) the forms of the Latin
Irish a further
'thorn', p. Later, most with the quality of [e]
alphabet to present the symbol, 8, was devised by (p.1 54). This quality was

sounds of Anglo-Saxon as drawing a line through the represented by using the

Armenian phonetically as possible.
(linjitiJ (/> fayutjiin, miiiwi/i m^i//.1. Latin d; this came to be Latin digraph ae, which
But they ran into difficulties called eth'. However, the came to be called ash',
Balinese when they encountered two new letters were not after the name of the runic
four sounds which had no used to separate the two symbol that represented
counterpart in Latin. sounds, and in the Middle the same sound. By the
• /w/ came to be written with English periods they were Middle English period it

Buginese a runic symbol p, known as replaced by th. Curiously, p had fallen out of use, pro-
'wynn'. It was replaced by has survived to the present bably because sound
uuor win Middle English day, in such artificial forms changes (§ 54) had made it

Burmese and is rarely found after as Ye Olde English Tea no longer needed.
1300. Shoppe', where the V really
uojnoc:coi coxx]33a>TOC)cc[gc:cliieTOogoa5o:
• /9/ and /6/ (as in mo- stands for a badly made p.

Coptic IlkipH-^ Pkp ^ t^ IXCnp. I11HOCUOC

goicTc ntqujHpi uutrtrtj hTtcjTHicj fink
». -\ * r\ rs

W*i ch <MI<
"v
H J|M->Hd Hd ^IN^I Runes
Devanagari
The runic alphabet was The version found in Britain
<HMH ^<^ <*>l mIN ^Ui cflMli i
used in north-west Europe, used extra letters to cope
Ethiopic
mainly Scandinavia and
in with the range of Anglo- \ H
<ni»TWA. :: l^hT h. TA.V ft7A A.HW o*(\r the British Isles, and has Saxon sounds. No-one
K
: .
i i i :

been preserved in about knows where the alphabet

Georgian 4,000 inscriptions and a came from but it is probably
few manuscripts. It dates derived from the Roman t
from around the 3rd cen- alphabet; there were many
tury ad and continued to be trade contacts between
Gujarat!
used on charms and monu- Roman and Germanic k t
ments until the 1 7th cen- peoples in the Rhine area
tury. The common runic al- during the first centuries of x
Javanese phabet (given at right) our era.
consisted of 24 letters, and
P M
is usually known as the Below: One side of the N r\
'futhork' or 'futhark', from Franks Casket, an 8th-cen-
Kannada the names of its first six let- tury ad carved box, with a r
Uj, *^ uijjflt sijrW^ **Wjti); eS^ ters. Several variant narrative inscribed in
shapes of the letters exist. runes.
Khmer
iJtytjjntwnUrjtoJaftqptenBDynin:

Malayalam

Maldivian SS- J si 3 st ? t-
-O-* -» sij-^s-si

ss-s-., s-^' >ss.

Sinhalese

Syriac

Tamil

Ogham ^
c = 1

Telugu The origins of the ogham ments southern Ireland.

in
5=
1' = t
—b
(or ogam) alphabet are un- The alphabet has 20 let- *
Thai miu ri Midi" nn fn Inn, luln'iJnmu mtu«i
known, though links have
been proposed with both
ters, divided into four sets *5
!» = d E— i

of five letters. The letters 5- h

•in minis) mimk, runic and Etruscan. It was were simple strokes ^ng =
inern nu n in 11 lu wituni
used for writing Irish and notches cut into the
or
edge of s\ — n
e

Tibetan Pictish from around the 4th a stone. They are usually ^g E u
s
century ad. There are read from bottom to top, or •-m
=
, -
- :
«^-q«C - -

pfc 5C"C'«}»l oj8Br »»|-'5 »JO|'OI"«I^C

about 500 inscriptions,
mainly on stone monu-
from right to left.
— q
z^ V
— a

33 GRAPHOLOGY • 203
Graphological contrasts The size of the graphemes, for example, is a major
way of conveying the relative importance of parts
Once a writing system has been devised, it can be of a message, such as in advertisements or invi-
used to convey a wide range of graphological con- tations. The switch from Roman to Gothic type
trasts. These are best illustrated from the range of may convey an 'old world' connotation, as in many
possibilities available in alphabetic systems. Christmas cards and shop signs. However, it
should be noted that not all languages have the
Spelling The essential identity of words is con- same set of possibilities — for example, there is no
veyed by the correct selection and sequence of gra- use of italics or capitalization in Hebrew.
phemes - the spelling rules of the language. This
is the main component of any graphological des- mark both
Capitalization Initial capital letters
cription. It is a study that needs to include, not lexical and grammatical units 194), usually
(p.
only the 'normal' rules that have to be learned in sentences or words. A single graphic contrast is
order to read and write, but any dialectal, stylistic, involved: big vs small. The graphic contrast be-
or 'free' variations. Dialectal variation is illustrated tween large and small capital letters (A vs a) con-
by American—British differences such as color/ veys no conventional meaning difference. Also,
colour or the use of thru for through. Stylistic varia- capitalization does not apply to numbers: if 33
tion can be illustrated by the way authors adapt were to appear beginning of a sentence,
at the it
the spelling system to reflect or suggest the pronun- would not be written 33.
ciation of non-standard speech (p. 180). An inter-
esting example is the use of shuvvle for shovel, in
Spatial organization
portrayals of Cockney speech: the two forms have
The general disposition of symbols on a page (or
identical pronunciations, in fact, but the former
other format) can itself convey semantic contrasts.
manages to convey the impression of a non-stan-
This is something newspaper editors are very much
dard accent. Free variations (p. 161) include such
aware of when they juxtapose stories on the same
alternatives as judgment /judgement and -ise/-ize.
page - in one case, a story about the Ethiopian
famine of 1984 was placed next to a story about
Special symbols A large number of symbols are
the mountains of food being stored in Europe. If
available to express frequently occurring meanings
the stories had been on different pages, the effect
inan economical way. Most of these are logograms
would have been lost. Other examples of contras-
(p.200), such as +, @, £; but some do not relate
tive layout include the use of captions under pic-
to individual words, such as ^
used to mark
,
tures (particularly noticeable when the caption is
the place where paper may be cut; -» which ,
placed under the wrong picture), the placing of
indicates a direction; a , marking a telephone
headlines or titles, the layout of headings and sub-
number; and the dagger (f), showing that a person
headings in a script or report, and the layout of
is dead. Special symbols may also be used to help
literary texts, especially poetry (pp. 72, 388).
organize a written text (such as asterisks or super-
script numbers relating to footnotes) or to draw
attention to part of it (such as a large star before
a name in an advertisement). An important use of
the asterisk has been to show omitted letters, espe-
cially in taboo words (p. 61).
Deviant spelling
Abbreviations Shortened forms of words are a The use of abnormal spell-
ing to make a point is more
major feature of written language, as in the use common than we might
of titular contractions and abbreviations such as think. It is sometimes used
Mr, Dr, Ms, Lt, and Capt, or the use of acronyms, as an economical way of
such as COD, VIP, and nato. The abbreviations expressing a contrast in
poetry, or identifying a per-
may even come from a different language, and the
sonality in a story. And it is
full form may not be known, e.g. e.g. {= exempli
a commonly used device in
gratia), i.e. = id est), and etc. Some abbrevia-
( the world of advertising,
tions are spoken as words (e.g. nato is usually where can make the
it

/'neitau/); some are spelled out (e.g. VIP is always name of a product or shop
stand out and be remem-
/'vi: 'ai 'pi:/); some are automatically expanded
bered, or provide the basis
(e.g. Mr is/'mista/); and some permit a choice (e.g. for a legal trade mark. Hyway Inn Resistoyl
viz. spoken as /viz/ or as namely). (After S. Jacobson, 1 966.) Kilzum (insect spray) Rol-it-on
Bar-B-Q Koffee Kake Savmor (discount store)
EZLern (U.S. driving Kwik Koin Wash Strippit
Graphic contrasts Italic, boldface, capitalization,
school) Loc-tite Tini-plugs
colour, and other graphic variations are major Fetherwate Masqit Wundertowl
ways of expressing semantic contrasts, some of Firetuf No-glu
which are illustrated on pp. 180—1 and in §32.

204 V • THE MEDIUM OF LANGUAGE: WRITING AND READING

PUNCTUATION
The punctuation system of a language has two The semantics of layout
functions. primary purpose is to enable stretches
Its
The importance of layout
of written language to be read in a coherent way; for semantic effect can be The Suicide, or Descartes a Rebours
its secondary role is to give an indication of the seen in this poem of Jose

rhythm and colour of speech (though never consis- Paulo Paes (translated by
cogito
Edwin Morgan). The spac-
tently). It roughly corresponds to the use of supra-
ing between the words
segmental features (§29), but it differs from speech gives time for the reader
in that its contrasts are to some extent taught in (who knows Descartes'
schools, and norms of punctuation are convention- famous dictum) to build up
an expectation that the last
ally laid down by publishing houses in their style ergo
word is going to be sum.
manuals. The effect would be lost if
the poem had been printed
Features that separate in a single line.

Punctuation mainly used to separate units of

is

grammar (sentences, clauses, phrases, words, §16)

from each other. The various marks are organized boom
in a broadly hierarchical manner: some identify
large units of writing, such as paragraphs; others JOSE PAULO PAES
translated by Edwin Morgan
identify small units, such as words or word parts;
others identify units of intermediate size orcom-
plexity. The main English-language conventions
• quotation marks (inverted commas): identify the
are as follows:
beginning and end of an extract of speech, a
• space: separates words; identifies paragraphs - title, a citation, or the 'special' use of a word.

the sentence begins a new line, with the first

first The choice of single vs double quotes is variable:
word usually indented; extra space may also be the latter are more common in handwritten and
inserted between paragraphs, especially to mark typed material, and in American printing.
a break in the discourse. • hyphen: marks two kinds of divisions within a
• period (full stop): identifies the end of a sentence, word — to show that a word has been split in
along with question and exclamation marks; two because of the end of a line (a feature that
sometimes followed by a wider space than is has no spoken counterpart), and to relate the
usual between words (printing and typing con- parts of a phrase or compound word to each
ventions differ); also used to mark abbreviations other (as in pickled-herring merchant - vs pick-
(though practice varies); a sequence of (usually led herring-merchant — and washing-machine);
three) periods indicates that the text is incom- practice varies greatly in the latter use, with Bri-
plete. tish English using hyphens in many contexts
• semi-colon: identifies the coordinate parts of a where American English would omit them.
complex sentence, or separates complex points
in a list (as in the previous paragraph). Features that convey meaning
• colon: used mainly to show that what follows Some punctuation features express a meaning in
it is an amplification or explanation of what pre- their own grammatical con-
right, regardless of the
cedes it - as in the present sentence. text in which they occur. (Special symbols of this
• comma: a wide range of uses, such as marking kind are illustrated on the facing page.)
a sequence of grammatical units, or a unit used • Question mark: usually expresses a question,
inside another; displays a great deal of personal but occasionally found with other functions,
variation (such as whether it should be used such as marking silence (p. 180) or uncertainty)
before and in such lists as apples, pears, and (e.g. this is an interesting (?) point).
plums). • Exclamation mark: shows varying degrees of
• parentheses (
) and brackets [ ]
: used as an alter- exclamatory force (e.g. ///); also, some special
native to commas to mark the inclusion of a uses {e.g. John (!) was there).
grammatical unit in the middle or at the end • Apostrophe: most commonly used to mark the
of a sentence. genitive singular or plural (cat's, cats'), and
• dash: used in pairs with the same function as grammatical contractions (I'm, won't); found
parentheses or brackets; used singly to separate also in certain words {o'clock, fish 'n chips);
a comment or afterthought occurring at the end subject to a great deal of usage variation {St
of a sentence or to express an incomplete utter- Johns or St John's} Harrods or Harrod's?) and
ance; in informal writing, often replaces other uncertainty [*ice cream cone's, "'todays bar-
punctuation marks. gains).

33 GRAPHOLOGY 205
 .

Shorthand
Samuel Pepys's diary
Shorthand is a method of writing at speed using A page from Samuel and 265 arbitrary symbols, phered until the beginning
special symbols or abbreviations for the usual let- Pepys's original shorthand such as 2 for to, a larger 2 of the 1 9th century. It was
ters and words of speech. It is a system intended diary. The diary was written for fwo, 5 for because, 6 for first published in 1 825.
for a limited readership (usually only one person, between 1660 and 1669, us, etc. There are also se-
when Pepys was forced to veral empty' symbols, Samuel Pepys (1633-1703).
the writer) and for short-term preservation (apart
stop writing because of fail- used presumably to foster
from the occasional literary or scientific diary). It ing eyesight. It consists of the secrecy of the work,
is therefore prone to idiosyncratic use: it is quite six small volumes, totalling and several of the more
common for secretaries trained in the same system over 3,000 pages, written censorable passages are
in a system devised by the
to be unable to read each other's shorthand. written in various foreign
early 1 7th-century transla- languages.
The practice of shorthand writing is variously Thomas Shelton. The
tor, The diary remained un-
known as stenography ('narrow writing'), tachy- system contains reduced read for several decades,
graphy ('quick writing'), and brachygraphy ('short forms of letters, dots for as Pepys left no key to the
writing'). It is best known from its use in press vowels, abbreviated words, system, and it was not deci
reporting and in clerical and secretarial work -
mainly the verbatim recording of legal proceedings
and the dictation of business correspondence - l1
though in recent years the advent of voice-record- ' So.
ing equipment has somewhat reduced the demand
for professional shorthand skills. L b tj d f £~* c / <s
y
Shorthand was well known in Ancient Greece
and Rome - the earliest recorded instance is the S z f~f c v <* I fJr- /

system used by the historian Xenophon to write

the memoirs of Socrates. In 63 bc, a Roman free-
man, Marcus Tullius Tiro invented a system for
recording the speeches of Cicero — a system that v
continued in use for over 1,000 years. Julius Caesar
*>
1*'
l
v
*<r
/
* 7 i
was one of many in this early period who learned
the use of shorthand.
The use of shorthand died out in the middle ages
y
.

r
tk r V
because of its imagined association with witchcraft.
The 16th century saw a revival and produced the
*
firstprinted
Characterie:
manual Timothy Bright's
in English:
An Arte ofShorte,
and Secrete
Swifte,
<- *r, p 'h

Writing by Character (1588). Shorthand became *. / /W -*

extremely popular 17th-century England, espe-
in
cially in relation to the aims of the movement to
find a universal language (§58). Several systems
were invented - notably, those of John Willis, Tho- <-V»-4 V' t* <*< /;l/
mas Shelton, Jeremiah Rich, and William Mason.
It came to be studied in school. In church, sermons

would be written down and taken home for later

study. In the 18th century, the demands of the
Industrial Revolution promoted the use of short-
hand in business administration, and its popularity
grew in Europe. Finally, the 19th century saw the
invention of the main shorthand systems that are A ->
still in present-day use.

There are a variety of methods of writing short- . nc^ < -HP •• £3 I ' <y> 11
.J
hand. Some abbreviate the normal spelling of .

words; others are based on ways of representing

the sounds of speech; still others require the user
to learn a list of arbitrary symbols; and there are
General Bertrand's diary
several combinations of these approaches. The
result is that over 400 shorthand systems have been N. so. le mat. en cal: il d6j. Saint Helena. The diary is An interpretation of the
devised for the English language alone. bi. se. trv. un. peu fat . . written in such an abbre- passage reads: Napoleon
This is part of an entry (for viated style that it is tanta- sort le matin en caleche: il
20 January 1 821 from the ) mount to a shorthand sys- dejeune bien, se trouve un
diary of General Henri- tem. This kind of private peu fatigue (Napoleon
Gatien Bertrand, who was shorthand is probably quite goes out in the morning in
companion toNapoleon dur- widespread among diarists. a carriage: he lunches well,
ing his exile on the island of finds himself a little tired').

206 •
V THE MEDIUM OF LANGUAGE: WRITING AND READING
Isaac Pitman (1813-1 897) 2000 Pitmanscript Teeline

amateurs

"V
thieves

fastened
L I
> neighbour

^~^\
4
<
security
^
Outlines compared Five and - a relative newcomer - that is becoming increasingly

w
words transcribed in Pitman Teeline, a combination of popular in the 1 980s.
2000, Pitmanscript, Gregg, shorthand and speedwriting

.
Shorthand by machine
K
t4,:Mf
' % A stenotype machine,
inventedin 1906 by W. S.
pear
quences
in the centre. The
of letters are then
se- culty keeping up with nor-
mal conversational speed

i'^H

WW |
Ireland,
reporter.
to record the verbatim pro-
ceedings
an American court-

law courts and

legislative meetings.
It is a small machine,
It is mainly used

of
printed (without noise) on a
roll of paper.

some words are abbre-

viated,
have to
The
looks strange, because
printout

and some letters

be typed using
(p. 269). The system is
standard, so that the output
of different operators
tually intelligible (not al-
ways the case with pen
shorthand). However, the
is mu-

with a keyboard of 22 keys combinations of other expense of the machines,

that the operator strikes us- letters. and the training of opera-
ing both hands simul- An experienced steno- tors, has limited the appli-

taneously. The left-hand type operator has no diffi- cation of the approach.
fingers type consonants
occurring before vowels,
and these are printed on
the left of the paper; the
right hand fingers type con-
sonants occurring after
vowels, and these appear
on the right. The thumbs
type the vowels, which ap-

Pitman's Stenographic Sound Hand, published in s EJ PR s

1 837, was based on the sounds of English. The system
T
uses a combination of straight lines, curves, dots, and TK A PR
dashes, as well as a contrast in positioning and F
shading (heavy vs light). Several of the graphic H EJ s
contrasts relate systematically to the sound system T E N
(§27); for example, all stop consonants are shown by P H A PB
straight lines; all labial consonants slope backwards; H A s
and the distinction between voiced and voiceless ST R t-»
sounds is indicated by line thickness. Most vowels are T O
omitted. ICP M u PB
Pitman's is the main system in use in Britain, and K A EJ T

is widely used in other English-speaking countries. The WM EJ s

phonetic principles of the approach also make it T P E L

HR O E s
relatively easy to adapt for use with other languages.
A P8
T O
R O R

John Robert Gregg (1 867-1 948) S P E RPB s

T H A L
Gregg devised an alterna- more use of loops and cir- T3
tive toPitman which cles, compared with Pit- s
H E
avoided the latter's reliance man's angular system, and e
on shading and positioning. the line of writing more clo- T P R
His approach uses separ- sely resembles a longhand T*PW T
ate symbols for consonants script. It is now the main E PB
and vowels, and all sym- system in use in the USA, F P L T
bols are written on a single and has been adapted to
same thickness.
line in the several other languages.
His symbols also make A sample of stenotype
shorthand

33 GRAPHOLOGY- 207
34 The process of reading and writing
pupil

Reading aqueous humour

It might have been thought desirable, before begin-

ning the account of written language, to present

an anatomical and physiological description of the
visual and manual systems in human beings, in
much the same way as the articulatory/auditory
systems were presented for speech in Part iv. How-
ever, this is not usually done in linguistic discus-
sions of reading and writing because there is so
little that can be said in our present state of know-

ledge. The study of what happens when language

is visually perceived and processed is very recent,

and while a certain amount is now known about

the likely processing operations involved, the
neuroanatomical correlations of these processes
remain obscure. Moreover, there is perhaps little The structure of the eye
in principle that can be said, given that the struc-
tures of the eye and hand do not seem to be biologi-
cally adapted for written language in the way that
the vocal organs are for speech (though, given the and is the area that gives the best visual detail,
relatively recent development of writing, §33, this such as required for identifying graphic forms.
is

is hardly surprising). As a result, the bulk of the
enquiries are carried out by psychologists con-
cerned less with the structure and function of the
eyes, and more with models of the 'deeper' ways
in which the brain works when it processes written
language.
The is from the fovea, the poorer
further a stimulus
our ability to discriminate. The parafoveal area sur-
rounds the fovea, and this in turn is surrounded
by the periphery. These areas are less involved in
the act of reading, but they do have some relevance
in the detection of larger visual patterns in a text.

Eye movements Perceptual span

One physiological topic has attracted considerable How much linguistic material can be seen during
attention, however: the nature of eye movements. a fixation? Most information about visual percep-
These movements can be recorded using various tual span comes from using a tachistoscope: sub-
techniques, such as by attaching a mirror to a con- jects are presented with a briefly flashed sequence
tact lens placed on the cornea; it is then possible of letters or words, and are then tested to see how
to film a beam of light reflected off the mirror. many they recall. In a single exposure of 1/100 sec, A sequence of
(See also thecomputational technique described it is usually possible to recall 3-4 isolated letters eye movements
below.) Using such methods, researchers have or 2—4 short words. Several factors affect subjects' Fixations are shown as
shown that the eyes work together, and that when performances, such as the distance of the stimuli and the order of
circles,
movement is shown by
searching for an object they move in a series of from the eyes, or whether the letters or words are
arrows; information about
rapid jerks, known as saccades (from French, 'the linguistically connected. the duration of each fixa-
flick of a sail'). Between each movement there is However, this approach does not replicate what tion is not given. In the fix-

a period of relative stability, known as a fixation. actually happens in reading, where people make ations for sentence 'The
vehicle almost flattened a
During reading, the eyes do not follow lines of print several fixations a second and do not have to name
pedestrian' given here the
in a smooth linear manner but proceed in a series what they have seen. Accordingly, several other reader spends most of the
of saccades and fixations. We usually make 3—4 methods of studying perceptual span have been time at the beginning of the
fixations a second, though rate and duration can tried. One sophisticated study involved the use of sentence. Note (i) that the
be affected by the content of what is read, and computer technology (K. Rayner &
G. W. McCon- word a does not receive a
separate fixation (and
there are some interlanguage variations. kie, 1977). An eye-movement monitor illuminated
possibly nor does the); (ii)
What happens during a fixation is of particular the eye with invisible infra-red light, and measured the first part of the sen-
importance in studying the process of reading. The tence is looked at three
nerve cells that convert light into electrical pulses times. Effects of this kind

are located in the retina, at the back of the eye. require a complex explan-
which physical
ation, in
The central region of the retina, where these recep-
features of a text (such as
tor cells are packed closely together, is known as word length) interact with
the fovea. It extends for some 2° of visual angle, The vehicle almost flattened a pedestrian its semantic properties.

208 -V THE MEDIUM OF LANGUAGE: WRITING AND READING

the amount of from certain parts of
light reflected we may find out something about what they have Joint reading
the eye's surface. A
computer was then attached read, but not about how they read it. Nor are Smith has taught himself to
to this equipment and programmed to check the experimental situations necessarily convincing, read Russian letters, but he
eye position 60 times a second, keeping a record because they make readers do abnormal things. hasn't had time to learn the
language. Bronski was
of where the person was looking and how long And analysing the behaviour of people with read-
brought up speaking Rus-
each movement or fixation took. The text to be ing handicaps may produce results that do not sian, but he never learned to
read was displayed on a screen, also under the con- apply to healthy readers (p. 259). read. One day, Bronski gets
trol of the computer, thus enabling the researchers Given the difficulties, the field of reading research a letter in Russian from a re-
lative. He cannot read it. He
to make changes in the display during the period would not seem to be a particularly promising or
shows it to Smith. Smith can
of an eye movement. attractive one. It is, however, an area that has
not understand it. But all is
In one experiment, a piece of text was 'mutilated' attracted many investigators, partly by virtue of well: Smith reads the words
by replacing each letter with an x. When subjects its very complexity, and partly because any solu- aloud; Bronski recognizes
looked at this display, the computer automatically tions to the problem of how we read would have them, and interprets them.
immediate application in areas of high social con- He is happy. But who is
replaced the x's within a certain region around their
'reading'?
point of central vision with the letters from the cern. A large number of children have great diffi-
original passage. This created a 'window' of nor- culty in learning to read, and many never read well.
mal text in the subject's foveal region for that fix- Estimates suggest that between 10% and 20% of
ation. When the subject's eyes moved, the old the U.S. population are functionally illiterate (p.
window was replaced by x's and a new window 272). Such figures thus bring a sense of urgency
was created. The size of the window was under to reading research.
the control of the researcher: in the table (see right), 'Reading' in all of this does not mean simply
a window of 17 characters is shown. Subjects had 'reading aloud', which might be done by a suitably
no difficulty reading under these conditions, unless equipped automatic machine that would not know
the window became too small. what it was saying. 'Reading' crucially involves
By using different window sizes and mutilating appreciating the sense of what is written: we read
the text in different ways, it was possible to draw for meaning. It is this link - between graphology
various conclusions about perceptual span. Reduc- (§33) and semantics (§17) - that has to be
ing the window size slowed the subjects' reading explained by any theory of reading.
speeds, but it did not affect their ability to compre-
hend the text (even if all the reader could see was
nine letters — little more than a word at a time). Fixation windows
The study suggested that subjects were using letter A line of text on four suc- technique. Each window to the left of fixation
cessive fixations, using area is 1 7 letters wide - (marked with a dot) and
information no further than 10 or 11 positions
Rayner and McConkie's that is, eight letter positions eight to the right.
from their centre of vision, though information
about word-length and general shape could be
obtained from further away. A follow-up study Fixation
Text
also showed that these regions were not symmetri- number
cal around the centre of vision: on the left side, 1 Xxxxhology means perxxxxxxxx xxxxxxxxx xxxx xxxx xxxxxxx Xxxx XX X
the area used during a fixation was restricted to 2 Xxxxxxxxxxxxxxs personality diaxxxxxx xxxx xxxx xxxxxxx. Xxxx XX X
four letter positions. 3 Xxxxxxxxxxxxxxx xxxxxxxxxxx xiagnosis. from hanx xxxxxxx. Xxxx xx x
It is possible to conclude that, when looking at
4 Xxxxxxxxxxxxxxx xxxxxxxxxxx xxxxxxxxx xxxm hand writing. Xxxx xx x

a text of average type size about 30 cm away,

readers do not usually identify more than two or
three short words (about 10 letters) on each fix-
ation. Larger units cannot be seen 'all at once'. Graphic typography
One of several typographic

THEORIES OF HOW WE READ systems designed for re-

search in visual pattern
Following a fixation during reading, a visual pat- The general
identification. ird
tern of graphic features is conveyed to the retina, aim of such research is to
and then transmitted via the optic nerve for inter- see whether, by varying vi-
sual form, words can be
pretation by the brain (§45). The stages involved
made more discriminable
in this process are not well understood, and several
and pronunciation features
different theories have been proposed to explain more salient. The typo-
Oj^-Ivj pmnip
what happens when fluent readers read. One graphy, which enhances >
reason why the field is so controversial is that it the main graphic features
of letters, makes the visual
is extremely difficult to obtain precise information
shape of a word stand out
about the events that take place when people read. more clearly than does
am'
it

In fact, very little actually seems to happen, apart when printed in a conven- don
from the eye movements — and these do not begin tional way. It should be
to explain how the reader is managing to draw noted that each letter is the are separated with dashes, ple, /and rare shaded, and
same whenever occurs. it and a few pronunciation vowels are darker. (From L.
meaning out of the graphic symbols. Similarly, if Certain morphemes (p. 90) cues are added; for exam- Brooks, 1977.)
people are tested after they have read something,

34 THE PROCESS OF READING AND WRITING 209

READING BY EAR OR BY EYE?
Most people have encountered the struggle that Reading by ear
takes place as a child is learning to read. A major Spoken language Written language

feature of this task is that words and letters are

H V
'sounded out'. It is as if reading is possible only
if the symbols are heard -
reading 'by ear'. One Visual
Acoustic
theor> of reading therefore argues that a phonic
r

analysis analysis
or phonological step is an essential feature of the
process - a theory of 'phonic mediation'. The view U w
implies that reading is a serial or linear process,
taking place letter-by-letter, with larger units Acoustic Lerter-to-sound Graphic
gradually being built up. code conversion code

The alternative view argues that there is a direct

1'
relationship between the graphology- and the se-
mantics, and that a phonological bridge is unne-
Auditor.' word A model
cessary (though it is available for use when reading of the 'reading by approach. The process
ear'
recognition and assigns them to a
of visual analysis identifies letters
aloud). Words are read as wholes, without being
graphic code (graphemes). These are then translated into
broken down into a linear sequence of letters and W an acoustic code (letter-to-sound conversion), at which
sounded out - reading 'by eye'. Readers use their Doint the Drocess links uo with the nrocessinn svstfiim
peripheral vision to guide the eye to the most likely required for speech (§25). The auditory patterns are then
Meaning recognized as words and semantically interpreted.
informative part of the page. Their knowledge of
the language and general experience helps them
to identify critical letters or words in a section of
text. This initial sampling gives them an expec-
Reading by eye \X ritten language
tation about the waythe text should be read, and
they use their background knowledge to 'guess' the
A model of the 'reading by
eye' approach. The pro-
remainder of the text and fill in the gaps. In this cess of visual analysis
view, a text is like a problem that has to be solved and as-
identifies letters Visual
using hypotheses about meaning and structure.
its signs them to a graphic analysis
The arguments for and against these views code (graphemes). The vi-
sual patterns produced are w
are complex and multifaceted, deriving from the
recognized as words and
results of a vast number of experiments on aspects given a semantic interpre- Visual word
of reading behaviour. Some of the points that have tation. The semantic code recognition
been raised are summarized below. stored in the brain may
then be used to activate an w
Support for the ear associated acoustic se-
quence of events, should Meaning
• Associating graphemes and phonemes is a this be needed (as in read-
natural process, which cannot be avoided when ing aloud). If

first learning to read.

Speech
• Letter recognition is very rapid — about
production
10—20 msec per letter - which is enough to account
for average reading speeds (around 250 words per 1 '

minute). These speeds are similar for both silent

Spoken language
and oral reading (though the latter is slightly
slower, presumably for articulatory reasons), and
are close to the norms for spontaneous speech (p.
125).
• Statistical studies of word frequency (§15)
show that most words in a text are of very low
frequency, several occurring only once over long
periods; some will be completely new to a reader.
Readers can therefore have few expectancies about
such material and will need to decode it phonologi-
cally. It is an everyday experience to break new
long words up into phonemes or (more usually) ing. Yet we are able to read these variations quite
syllables: try picomalesefeso, and see. rapidly, even in experimental situations (using such
• When people read difficult material, they often forms as BoAt).
move their lips, as if the phonology is needed in • Reading by eye would be a very complex mat-
order to help comprehension. There may be other ter. Each word would have to be given a separate
sub-vocal movements not so far observed. orthographic representation in the brain, along
• It is difficult to see how the 'eye' theory can with a separate retrieval process. This is not a parsi-
handle the many variations in type and handwrit- monious explanation.

210 •
V THE MEDIUM OF LANGUAGE: WRITING AND READING
Support for the eye Itshould be noted that some of the arguments
• Fluent readers are not confused by such homo- that come to mind, in relation to this issue, do
phones as two and too. Phonology cannot help in not clearly support either theory. For example, it
such cases. Moreover in words like tear, there is has been argued that people who have been pro-
no way of deciding which pronunciation is in- foundly deaf since birth, and who subsequently
volved (/tia/ or /tea/) until after the reader has learn to read, provide clear support for the 'eye'
selected a meaning. theory; in their cases a phonological bridge cannot A combined model
• be available. However, the fact that such people This incorporates some of
In one type of reading disorder ('phonological
the findings from the experi-
dyslexia', p. 272), people lose the ability to convert do have great difficulty in learning to read could
mental work referred to in
isolated letters into sounds; they are unable to pro- be interpreted to show the importance of phonolo- this section. It is based on
nounce even simple nonsense words, e.g. pob). But gical mediation after all. Similarly, the existence the approach of the British
of Chinese and Japanese kanji (pp. 195, 200) is psychologist John Morton
they are able to read real words, showing that a
sometimes proposed in support of the view that (1933- ), but it ignores
non-phonological route from print to meaning several detailed features of
must exist. a phonological stage is unnecessary. But here too
that model (especially on the
• The 'ear' theory does not explain how some the evidence is ambiguous. Logographic systems output side) and does not
people can read at very rapid speeds, which can seem to be difficult to learn, with few users master- use his distinctive termin-
ing more than about 4,000 symbols out of the ology, in which the units of
be in excess of 500 words per minute. The eyes
word recognition and pro-
can take in only so many letters at a time. Rapid 50,000 or so which exist. On the other hand, very
duction are referred to as
reading poses less of a problem for the 'eye' theory, little is known about the orders of difficulty that
logogens.
as it simply requires that readers increase their sam- are encountered in learning kanji symbols, and the 1. Familiar words are vi-

degrees of expertise that exist in using logographic sually analysed, recognized,

pling as they speed up.
and assigned a meaning.
• In brief exposure experiments, people identify systems.
Their spoken form may be
whole words more rapidly than isolated letters. For As with most major theoretical oppositions, ele- retrieved from the speech-
example, if subjects are shown BAG, BIG, A, I, ments of both approaches are required to ex- production system, which
IBG, etc., and asked whether they have just seen plain the experimental findings. As a consequence, may be activated by the
several 'compromise' models have been devised, meaning (reading with
A or I, they perform best with the familiar words. understanding) or by the vi-
This is the 'word superiority' effect. which integrate the main features of both 'ear' and
sual patterns directly (read-
• The fact that different sounds are written iden- 'eye' theories. Some of these models are extremely ing without understanding).
tically, and different letters can have the same pro- complex, postulating a large number of compo- 2. Unfamiliar words are vi-

nunciation, complicates a phonological view. Also, nents and pathways, but this is only to be expected. sually analysed and may
then be analysed phonologi-
some orthographic rules seem totally unrelated to Despite the clear visual signals provided by the
cally (letter-to-sound conver-
the phonology, e.g. skr- is acceptable in English written medium, learning to read is a complex
sion). The resulting sound
speech, but does not occur in normal writing. process, and only an appropriately sophisticated pattern can be referred back
• Some higher-order processing must be involved theory will explain it. to the auditory word-recogni-
tion system, to see whether
in reading, because of several observed effects.
it 'ringsa bell'.
Experiments have shown that it is easier to recog-
nize letters in real words than in nonsense words.
Typographic errors are often not noticed when
reading through a text (the proof-corrector's pro-
blem). Errors made by fluent readers while reading Spoken language Written language

aloud are usually syntactically or semantically

appropriate; they make few phonologically in- \\ 1f

duced errors (cf. the findings of speech perception, Acoustic Visual

analysis analysis
p. 147).

' 1 '
1

Compromise? Acoustic Graphic

It evident that neither approach explains all
is
code code

aspects of reading behaviour; it is likely that people

1 ' 1 '

make use of both strategies at various stages in

learning and in handling different kinds of reading Auditory word Visual word 1 '

problem. The 'ear' approach (sometimes referred recognition recognition

to as a 'bottom-up' or 'Phoenician' theory, because a

of its reliance on basic letter units) is evidently very
w Meaning _
important during the initial stages. Perhaps after '
1 Letter-to-sound
several exposures to a word, a direct print— mean- conversion
i
r

ing pathway comes to be built up. But the 'eye'

approach (sometimes referred to as a 'top-down' Speech
or 'Chinese' theory, because of its reliance on
whole-word units) is certainly needed in order to
explain most of what goes on in fluent adult read-
ing. Spoken language

34 THE PROCESS OF READING AND WRITING- 211

 3

Writing do they correct them? Are the corrections appro-

priate? Many such questions await answers.
It is extremely difficult to discover what happens
Few
A model of written composition must also allow
when people compose a written text. satisfac-
when
for the fact that what people see they write
tory experimental techniques have been devised.
may affect the way
they think. Authors' comments
Direct observation of the handwritten product
are illuminating: doesn't look right now I've 'It
gives very limited information, for it fails to pre-
written it down', 'That's not what I'm trying to
serve the order in which revisions are made. Direct
say.' Meaning does not always exist prior to writ-
observation of people engaged in writing tells us
ing; often the process operates in reverse. A typical
little about what is going on 'beneath the surface'.
comment is Edward Albee's: 'I write to find out
And introspection is of little value, for as we think
what I'm thinking about' (§5). Such remarks
about our writing activity, so we destroy its natur-
emphasize the main lesson to be learned from the
alness. Fluent writers are in any case unaware of
study of the process of writing: it is not a merely
what they do when they put pen to paper.
mechanical task, a simple matter of putting speech
Only the most general of accounts can be given
down on paper. It is an exploration in the use of
of what is involved in the writing process, there-
the graphic potential of a language — a creative
fore. Models of this process recognize at least three
process, an act of discovery.
factors.

• There must be a planning stage, in which

thoughts are organized, and a lexical/grammatical
outline prepared. This involves writers working out
what their readers need to know, in order for their
message to be understood. In particular, they must Time to think
anticipate the effect their words will have (§21). An extract from a video movement. During these currence of mental plan-
• Writers need to be aware of the linguistic and study describing the writing pauses, other kinds of body ning and provide clues to
behaviour of some high activity may be taking the difficulty of the writing
social conventions affecting their use of written lan-
school boys. The length of place: the eyes may scan task. Variations in pause
guage. These include such general considerations the writer's pauses be- the text or look away, and length may thus convey in-
as the need to make handwriting legible, to stay tween words is marked in the hand may move away formation about the pro-
within the constraints of a single writing system, seconds. In line 5, for ex- from the text (presumably cess of writing, especially
and to follow the normal conventions of graphic ample, he paused for reflecting major decision- when considered along
1 6.6 sees; he then changed making) or stay close to it with other temporal as-
expression (such as writing in the expected direc-
the period to a colon, and (suggesting that the writer pects of writing.
tion) as well as the specific requirement to follow paused again before conti- expects to resolve the pro- (From A. Matsuhashi,
the rules of spelling and punctuation (§33). nuing. blem quickly). 1982.)
• Writers need to choose a specific medium of 'Pause' refers only to the The evidence suggests
expression, such as handwriting, typing, or word ceasing of the writer's pen that pauses reflect the oc-

processing, and this requires a consideration of

1 Truly ^successful' 'person -to -person 2 J communi- 5 8

motor-control abilities. Several factors are in- 8

cation 3 5 li9
because 69 people 6 9 9
1
2 - -

is difficult
1
in general' 'are poor
volved, such as hand—eye coordination, hand grip 70
3 '"listeners. They I0 would "rather 1 4
listen
5 9
to themselves
19
speaking
or position, position of the body, and so on. Many 2 4
4 'than someone else s 4,7It 9 is "my that "this 8 occurs
. feeling'
9 9

people have a slow handwriting speed, or have

because' 'of 2 a 8 basic 2 "self-centeredness.' 66 55 people 48 tend' 2 to
6
5 ' -

difficulty holding an implement or working a 7

'- 9
6 be 6 more interested in"their 9
own lives' 5
to'
2
bother' "exposing
machine (most noticeable in the case of physical 3 5
handicap, p. 280). The result is not simply that 7 '
themselves To how others
8
live.

a message takes longer to write. The attention and

memory may be so taken up with controlling the
motor activity that linguistic content and structure Some things that writers do
may be affected. People can forget what they A piece of handwritten text, Rearrangement: separat- Differentiation: adding or
wanted to write, even after they have begun to write showing four of the things ing or bringing together expanding material.
it.
that writers do while com- material; changing logic or
posing: word order. Note that the changes give
Deletion: elimination of Consolidation: making the no information about the
However, these threefactors are not the whole false starts, and unnecess text more compact or sequence of events that
story. For example, they do not allow for the fact ary or wrongly chosen streamlined, while retaining occurred while the text was
words. the content. being written.
that a great deal of written composition is rewrit-
£-«wuwr?_
ing. Any model must take into account the act of
revision - from the first stages of making notes,
jottings, and headings, through various drafts, to
the final version. This is a promising field of
research — the self-corrections and errors intro-
duced while composing written language. How do
writers ensure that their work is legible, readable,
and How do they detect problems in these
lucid?
respects? How do they identify the problems? How
rUw

212 V THE MEDIUM OF LANGUAGE: WRITING AND READING

Spelling
Reading and writing have long been thought of
as complementary skills: to read is to recognize
and interpret language that has been written; to
write is to plan and produce language so that it
can be read. It is therefore widely assumed that
being able to read implies being able to write —
or, at least, being able to spell. Often, children are
taught to read but given no formal tuition in spell-
ing; it is felt that spelling will be 'picked up'. The
attitude has its counterpart in the methods of 200
years ago, when teachers carefully taught spelling,
and assumed that reading would follow automati-
cally.
Recent research into spelling errors and
of the pen' has begun to show that matters are
not so simple. There is no necessary link between
reading and writing: good readers do not always
make good writers. Nor is there any necessary link
between reading and spelling: there are many peo-
ple who have no difficulty in reading, but who have
a major persistent handicap in spelling — some
researchers have estimated that this may be as
many as 2% of the population. There seems more-
over to be a neuro-anatomical basis for the distinc-
tion, as shown by brain-damaged adults who can
read but not spell, and vice-versa (p. 272).
With children, too, there is evidence that know-
ledge of reading does not automatically transfer
to spelling. If there were a close relationship, chil-
dren should be able to read and spell the same
words; but this is not so. It is commonplace to
find children who can read far better than they
can spell. More surprisingly, the reverse happens
with some children in the early stages of reading.
However, the differences between reading and
spelling cannot be explained simply by arguing that
spelling is 'more difficult', for this would not
explain such facts as children who can spell better
than they read. Rather, the two skills seem to
involve different learning strategies. Whereas read-
ing is largely a matter of developing direct links
between graphic expression and meaning (p. 211),
spelling seems to involve an obligatory phonologi-
cal component from the very outset. The study of
spelling errors shows that we learn to spell by mak-
ing associations between graphemes and pho-
nemes, and not simply on the basis of how graph-
eme sequences 'look'. Visual strategies can be
important; for example, with irregular words,
'slips
where a phonological strategy could not work, peo-
ple do sometimes write down alternative spellings
to see which 'looks right'. But for the most part,
it is the signs of phonological activity that are the
most noticeable - as when we see beginners pain-
fully writing C-A-T and saying the letter names
of sounds as they write, or adult writers sounding
out words (especially long words) while writing
them down.
Why is there this preference for phonology? Per-
Spelling mistakes
haps because spelling involves a conscious ability
Spelling mistakes are not
to form linear sequences of letters - an ability that
very common in adult hand-
is routinely required for processing the linear pho-
written texts - descriptive
neme-strings of speech, but that is not found in studies suggest that, on
visual pattern recognition (as is required for whole- average, about 1% of letters
word reading). To be a good speller, we need to are affected, and 1-1 .5% of
words. One study classified
have both this phonological awareness (to cope
errors into four formal types:
with the regular spelling patterns) and a good visual omissions (buton), additions
awareness (to cope with the exceptions). Poor (hopeful!), substitutions (at-
spellers, it seems, lack this double skill. tendence), and inversions
(tabef). Other possibilities

One study gave children the same list of words are rare. The diagram shows
which kind of error occurs in
to read and to spell: several actually spelled more
different letter positions in a
words correctly than they were able to read cor- word. There is very little dif-
rectly. ference in the first letter. In-
sertion errors are much more
likely towards the end of a
Why so difficult? word, whereas the other
types are more likely towards
Why should reading and spelling be so different?
the beginning. Substitutions
It is partly a matter of active, production skills and omissions seem to go
being more difficult than passive, receptive ones. together.
Spelling is a more conscious, deliberate process, (After A.M. Wing & A. D.
which requires awareness of linguistic structure, Baddeley, 1980.)

and a good visual memory, to handle the excep-

tions to the regular patterns. It is possible to read
by attending selectively to the cues in a text, recog-
nizing just a few letters, and guessing the rest. It
is not possible to spell in this way: spellers have

to reproduce all the letters.

Also, more things can go wrong while spelling:
there are far more graphemic alternatives for a pho-
neme than there are phonemic alternatives for a
grapheme. For example, sheep has really only one
possible pronunciation, /Ji:p/; whereas the form
/Ji:p/ could be written in at least three different
ways — sheep, sheap, shepe. One study worked out
that in English there are 13.7 spellings per sound,
but only 3.5 sounds per letter (G. Dewey, 1971). Relative position of letter in word

34 THE PROCESS OF READING AND WRITING •

213
 :

HOW IRREGULAR IS ENGLISH • and early 17th centuries, many

In the late 16th The ghoti
SPELLING? new loan words entered English from such lan- phenomenon
The widespread impression that English spelling guages as French, Latin, Greek, Spanish, Italian,
and 'unpredictable' is based on such
'chaotic'
A famous comment by G. B.
is and Portuguese. In the following list of words from Shaw has promoted the view
famous sentences as 'Though the rough cough and this period, it is not difficult to see some of the that English spelling is highly
hiccough plough me through, I ought to cross the new patterns of spelling (e.g. -que, -zz-, -//-) that irregular. He observed that
lough.' However, descriptive studies show that this would make learning to spell consistently a much fishcould be written as ghoti
kind of thing the exception, not the rule. It is -f asm 'cough', /as in
is
more complex matter, especially in longer words. 'women', and sh as in
difficult to arrive at a firm figure for the amount
anonymous epitome idiosyncrasy 'nation'. But joining together
of spelling irregularity in a language, because peo- exceptional spellings proves
armadillo excrescence inclemency
ple differ overwhich words to include in the study. nothing about the basic sys-
balcony exhilarate intrigue
Should proper names be included, for example? tem of the language. An
bizarre galleon moustache
Should the estimates be based on word types or even more bizarre example
brusque gazette piazza is G. Dewey's (1 971 spelling
tokens (§15)? In one USA study, a computer analy-
)

canoe genteel pneumonia of taken as phtheighchound

sis of 17,000 English words showed that 84% were caustic grotesque potato (as in 'pMnisic', 'weigh',
spelled according to a regular pattern, and that only chaos grotto system 'school', 'glamour', 'hand-

3% were so unpredictable that they would have cocoa harass vogue some').

to be learned totally by rote (P. R. Hanna, et al.,

1971). A widely cited figure is that English is about
75% regular. On the other hand, the 400 or so
among the most fre-
irregular spellings are largely
quently used words in the language, and this pro-
motes a strong impression of irregularity.
Where does the irregularity come from?
The history of the language provides many reasons
The result is is an amalgam of dif-
a system that
ferent traditions:Anglo-Saxon, French, and Classi-
cal spelling patterns are all used. The system is
basically a phonemic one, but the phonemes are
represented by letter patterns as well as single let-
ters. In addition, the spelling preserves a great deal
of information about the relationships between
words (e.g. author/ authority, sign/ signature), and

for the irregularities of English spelling. enables us to see links with other languages, which
• The basic fact is that, in the Anglo-Saxon have many similarly spelled words. Foreigners who
period, an alphabet of 27 graphemes (the 23-letter have only a reading knowledge of English are much
Latin alphabet, plus four other symbols) had to helped by this similarity. But the task facing the
child learner remains considerable.
cope with a sound system of nearly 40 phonemes.
Later, i/j and u/v were distinguished, and w was
added, but many sounds still had to be signalled
by combinations of letters.
• After the Norman conquest, French scribes re-
spelled a great deal of the language, introducing A page from John Hart's A Methode or Comfortable Beginning
such conventions as qu for cw {queen), gh for h for all Unlearned, Whereby They May Bee Taught to Read
(night), and c before e or i in such words as circle English, in a Very Short Time, With Pleasure (1 570)
and cell.
• The printing process caused complications.
Many early printers were foreign (especially from well t) bide tbt fortroyturt wdbyonr finger, orfime Pjprrfitre tt cuter tbt

Holland), and they used their own spelling norms. put xntbt HneytutoktilemeUjtetattji bimtobauetbt more rtgudett the

Also, until the 16th century, line justification fioftf tftbt letter u
(p.
^WWwi the learner doth know, and con name the Tor tray teres rea-
183) was often achieved by abbreviating and con- Aiyjjtm that teatbjnnj} take beede that you teach not your Scbolicr as you
tracting words, and also by adding extra letters •wen taught. Undfirfo Tmderftandyonjbat thtfricf^ which 1 •mitt vnaer
(usually an e) to words, rather than extra space. entry vewellas a& worth aa: and t,et : iji :t,oo: u, nn.andm ttacbiig
• Especially after printing, the writing system did the tetter or naming it,you muft found ae mote of tbt fortrayurt which doth
remember it j ban tbt Tony found tftbt letter*u may be that Gf^tfjf, anf
not keep pace with the sound changes that were
tare, e : tfeele,; : tftlci-leafi, t : andtf wood cxk..u. ^Andfo of toe reft as
affecting the language (§54). The 'Great Vowel
tfUtn(jt» may not name tbt L maunr r. as you baitt bene taught, calling
Shift' of the 15th century was the main reason for tbtm cljcmjnjr : but glut them tbt fame foundes,ytn do find in their for-
the diversity of vowel spellings in such words as trayurts,witboutfounding ofany ytwei before tbem)at may be tlut,l-y»n
name, sweet, ride, way, bouse. Similarly, letters ti—tnl : K— fd!:r\—ing -.and as tbty art foundedwhen they are doubled
that were sounded Anglo-Saxon became 'silent',
in in tbt middtft ofwordes/u m
filling, Miller, founding tbtm s^nc,ai rn.iyt

e.g. the k of know and knight, or the final e in

bttfUjnfa—tl—ing/nd'kii-ll-tr. Soofm*ts of Ha— on : cfn.as mm—
in Ta—mn—er . *fr, as in fu~rr—edjt»ibing totulnng any -until:for tbty
stone, love, etc. tare indifferent fir all vowelt and diptbongt, both before , andalfo after
• In the 16th century, therewas a fashion to tUm, and tbtrefire»nrit not tobauetbenMnefljytif with any cf tbtm, for
make spelling reflect Latin or Greek etymology (p. much as may be : and it tt as reafo-
that utttbe learners btnderatoice^v

330), e.g. the g was added in reign (from regno), nable name t be Ww// with he btlpt ofany ont or other Confonant^s tbt
to t

Cmftnant by tbt bdft of anyone trttber ttwtD. *And intcatbingtle

and the h in debt (from debitum). Unfortunately,
many false forms were concocted: for example, the
s of island was added because the word was
thought to come from Latin insula, whereas in fact
it is Anglo-Saxon in origin.
hrtatbjt.it fgvifietb nttbiug tlft but the putting forth of the fur e breath
wit bout any nuner offound, ar other afjnration: at in a—bb-a, or tab tt—
bb~t& bth,andfo oftbt reft tftbt towels and diphthongs. Undf* the. Jb
which Ifbajx withontfile pgMttbiuS. for that it ctmmetb dearly with a
breath, bardtbrufl tborow the tettb, in drawing the tongue inward t nurds

214 •
y THE MEDIUM OF LANGUAGE: WRITING AND READING
 .

SPELLING REFORM Parliamentary

A desire to eradicate irregular spelling can be traced the Simplified Spelling Society (1908) in Britain. proposals
back to the 16th century. In 1551, John Hart A system of 'Nue Spelling' was devised and widely
(d. 1574) complained of the 'vices' of English writ- promulgated, and this was followed by many other The Simplified Spelling So-
ciety's publication New
ing, which cause it to be 'learned hard and evil proposals in the first half of this century. Spellingwas presented to
to read'. In the following centuries, several experi- Systems of spelling reform are of several kinds. Parliament in 1 949. The new
mental orthographies were published. By the 19th Some, such as Nue Spelling (p. 216), are standard- system was to be introduced
three stages.
century, the view that English needed a more con- izing systems: they aim for a more regular use of in

sistent orthography had attracted widespread Bri- the familiar letters; no new symbols are invented.
1 It would first be intro-

tishand American support. A landmark was the Others, such as i.t.a. (p. 217) are augmenting or duced into the primary
publication in 1844 of an augmented Roman supplementing systems, which add new symbols schools; after five years, old
alphabet known as 'Phonotypy' by Isaac Pitman to those of the regular alphabet. Occasionally, sup- spelling would cease to be
taught.
(1813-97). Soon after, in 1876, the Spelling planting systems are devised, in which all the letters
2. During the next five years,
Reform Association was founded in the USA, fol- are new: an example is Shaw's Proposed British new spelling would be com-
lowed by the Simplified Spelling Board (1906), and Alphabet (p. 216). pulsory in films, advertise-
ments, and public announce-
THE PROS AND CONS OF SPELLING REFORM ments.
3. After ten years, it would
Advantages • All who
have learned old spelling have a vested be compulsory in all legal
• Children would save an enormous amount of interest in it, and few would be willing to learn documents, records, etc.
an alternative system, or wish to have their children New literature would not be
time and emotional effort in learning to read.
granted copyright unless it

• It would be of great help to children with learn- learn one. The problem of inertia and conservatism
was printed in new spelling.
ing difficulties. is probably insuperable.

• Because fewer letters would be used (an esti- • The saving in costs might be outweighed by the The bill was rejected, but by

mated saving of 15%), there would be a great sav- need to reprint important works in new spelling. only 87 votes to 84! A subse-
quent bill,953, in fact
in 1
ing in writer's time, and in the time and costs of • As a phonetic principle came to be intuitively
passed its first stage, though
typing, printing, and associated matters (paper, recognized, differences between accents might pro- later opposition by the Minis-
ink, storage, transport, etc.). mote diversity of spellings. try of Education forced it to

• There would be considerable benefits to foreign • There seems to be no agreement amongst the be withdrawn.
learners of English, and thus to the spread of Eng- various groups of reformers about an optimum sys-
lish throughout the world. tem. Also, the arguments are often presented in
an evangelistic manner, which many find unappeal-
Disadvantages ing.
• There would be a major break in continuity
The history of the spelling reform movement
between old and new spelling, especially in the
indicates that the disadvantages are generally felt
more radical schemes. The period of transition
to outweigh the advantages. The problems, it
would present major problems. It is difficult to see
would seem, are too great to be overcome. But the
how a programme of spelling reform could be
enthusiasm of spelling reform bodies all over the
implemented in a practical or realistic way.
world continues unabated.

Sound-spelling correspondences
A U.S. spelling study (p. Sound Spelling % Example
214) plotted the corres-
pondences between Eng-
M e 72 me
Some irregular English spellings lish vowels and their
ee 10 see
ea 10 meat
all

although
course
debt
listen

move
shoulder
some
spellings. A selection
their results illustrates the
of
M a
ai
80
9
late
sail
among way some of these rela-
do none sugar ay 6 say
answer
are
does
done
of
once
sure
talk
tionships are highly regu-
whereas others are
lar,

much less so. The spel-

N i

y
... e 74
14
side
shy
6 high
aunt
autumn
blood
dough
early
one
only
two
was
ling of /a/
regular
was
in their
entirely

(though a few rare excep-

sample M igh
o...e
oa
87
5
hope
boat
eye own water
tions do exist, e.g. plait). ow 5 low
build

busy
folk

friend
people
pretty
were
what
On the other hand, the
spelling of i is split be-
M u ...e
ew
90
3
rune
few
castle gone where
clerk
quay
who
tween and / y, and there N i 73 bid
great receive are many exceptions y 23 happy
climb
colour
have
hour
rough whole
women
(using
in
all vowel
women, busy, and
letters, as
vil-
M e
ea
93
4
set
stealth
said
comb island salt you
lage).
N a 100 cat
come journey says young N u 88
10
hut
son
cough key scarce
could lamb shoe M a
95
5
cot
wash

54 THE PROCESS OF READING AND WRITING .

215
SOME SPECIFIC PROPOSALS
cnjn SvaS, /h ij\W iukIcI - /h <!./yy1 dwSidDrl - /s d/wl
(71 i\ r

World English Spelling gxo - ?xh. 9 ptf j\\. cif iS \ |\|., 5a SlD7prc|. 510. gvr dwSicbcia M
91S

)r> r\7{ <.n "\vn loo 1 a v tfbJtl. 9 /vki. /ic oVc \o1 \u c\fl 5i/v/(/>
Forskor and seven yeerz agoe our faathers braut
/Yl /s Sc 510, |7l 11 ij\ \\.(n Jn?v.1 /M 9c \\\, gio. n 1? Ju 7S, 9 cip8,
forthh on this kontinent a nue naeshon, konseevd
D/9/o. 1 {s wvtta 510 1 9 7\Ji\i£1 /oc /\i 9c ja Jvl gn $j( 97S Jn So
in liberti, and dedikaeted to the propozishon that
aul men ar kreeaeted eekwal. Nou wee ar en.gaejd
in a graet sivil wor, testing whether that naeshon,
or eni naeshon soe konseevd and soe dedikaeted,
kan long enduer. Wee ar met on a graet batlfeeld
Tall:
*)

peep
1
toe
(f
kick
J
fee
d
thigh
S
So
L
Sure
t
Church
\
yea
I
hung
ov that wor.

Regularized English
I V ? f 9 ? 7 7 / 8
bib dead gag Vow they Zoo neaSure Judge Woe ha-ha

Regularized Inglish iz a system ov spelling which

lays down definit rules ov pronunciation which short c / f v. j r \ v < s
loll mime If egg ash ado On wool out ah
wood make it eazier for aull children to lern to
read and write. In aull probability it wood lead > •*
age Up
to a saving ov at least wun year's wurk for aull
Oil

schoolchildren. It wood aulso contribute very lar- The Shaw alphabet for writers
gely towaurdz abolition ov the existing amount ov One of the most recent enable English
al- ters to to be (letters with descending
illiteracy and backwardness in reading. phabet inventions was in- written without indicating strokes), and compounds.
spired by George Bernard single sounds by groups of Several phonetic princi-
New Spelling Shaw (1856-1950); it is of- letters orby diacritical ples are used; e.g. voice-
ten referred to as Shavian'. marks. He termed this the less and voiced con-
Forskor and seven yeerz agoe our faadherz braut
As a writer, Shaw felt there Proposed British Alpha- sonants (§27) are related
forth on dhis kontinent a nue naeshon, konseevd was an enormous waste of bet'. There was a competi- by reversed shapes. Capi-
in liberti, and dedikaeted to dhe propozishon dhat time and effort involved in tion, and in due course tals are not distinguished.
aul men ar kreeaeted eekwal. Nou we ar en.gaejd English spelling, and he Kingsley Read's design Proper names are identi-
was a vigorous campaigner was adjudged the winner.
in a graet sivil wor, testing whedher dhat naeshon, fied by a raised namer' dot.
for a new alphabet. He him- The alphabet follows the The four words and,
the, of,
or eni naeshon soe konseevd and soe dedikaeted, self always wrote in Pit- normal basic conventions and to are given separate
kan long enduer. We ar met on a graet batlfeeld man's Shorthand, which of English, being read from symbols: \, andl
? ,
f ,

ov dhat wor. was then transcribed by a
secretary.
In his will, Shaw ap-
Simpler Spelling Association
pointed the Public Trustee
forskor and sevn yerz ago cur faharz brot forfi Dn to seek and publish an
alphabet of at least 40 let-
his kDntinsnt 3 niu najan, ksnsevd in libsrti, and
dedikated tu "ra propszijsn hat d1 men ar kreated
ekwsl.
nai we ar engajd in 3 grat sivil wor, testirj
Arguing the case
hwehsr hat najsn or eni najsn so kansevd and
The flavour of the argu-
so dedikated, kan brj endiur. we ar met Dn 3 grat
ments surrounding spelling
batl-feld dv hat war. reform is well captured in
this extract from a letter writ-
ten by Shaw to the periodi-
cal Tit-Bits (22 March 1946).
He is reacting to an earlier
proposal from the reformer,
Mont Follick:
Everything that Dr Follick
says about our spelling is
true: but it was said by Alex-
ander J. Ellis a hundred
years ago, and has been re-
peated again and again by
the most eminent phoneti-
cians without producing the
smallest effect. The reason
is that as so presented the
change has seemed enor-
mously expensive and the
phonetically spelt texts
culuus and even sometimes
obscure.
What is needed is a new
alphabet of not less than 42
left to right, and using word
spaces. Punctuation and
numerals are unchanged.
There are four types of let-
ter: shorts, tails (letters
ascending strokes), deeps
letters, which is the lowest
number sufficient to repre-
sent all the sounds of spok-
en English recognizable by
a single symbol each. Dr
Follick, by confining himself
to 22 letters of the present
alphabet, is compelled to re-
present single sounds by
several and has
letters,
landed himself in such mon-
strosities as ei tscheir' to
spell a chair': 9 letters for
3 new British 42 letter alpha-
sounds! can write a chair'
I
1 2 times in a minute, and ei
tscheir' only 9 times.
number of minutes in
is 1.440. In a year
525,600!!!
To realize the annual
ference in favour of a forty-
ridi- two letter phonetic alphabet
as against Dr Follick's Ooto-
matik alphabet you must
multiply by the number of
minutes in the year, the
respectively. In the dia-
gram above, the letter
names
are given below the
symbols: the sound of each
with letter is given in bold type.
number of people the
in
world who are continuously
writing English words, cast-
ing types, manufacturing
printing and writing ma-
chines, by which time the
have become
total figure will
so astronomical that you will
realize that the cost of spell-
ing even one sound with two
lettershas cost us centuries
of unnecessary labour. A
bet would pay for itself a mil-
lion times over not only in
The hours but in moments.
a day When this is grasped, all the
useless twaddle about
enough and cough and
dif- laugh and simplified spelling
will be dropped, and the
economists and statisticians
will be set to work to gather
in the orthographic Golcon-
da.

George Bernard Shaw (1856-1950)

216 V THE MEDIUM OF LANGUAGE: WRITING AND READING

 -

Traditional < olumn A ugmented column

NAME CHARACTER EXAMPLE NAME CHARACTER EXAMPLE

1 ae *se *xnjel
2 bee b b«tl 25 zess *s houses
3 kee c curly 26 whae v,h whether
dee chae
4 d didn't 27 * rhicken
5 ee ee eeven 28 ith th (hicken
6 ef f fserly 2Q thee Jh rhaerfor
7 gae 9 given 30 ish Jh Jhoodn't
8 hae h hasn't 31 zhee ue3ueal
3
9 le I€ levory 32 ing 3 fhiqkiq
IO jae )
jeneral
i r kae k kwickly 33 ur r absurd
I 2 el 1 lieon
<3 em m mcetor 34 ah a farher
14 en n never 35 aw an aufool
'5 oe oe oenly 36 at a atlas
16 pee P prievxt 37 et e ended
17 rae r rvmm 38 it i idiot
18 ess s sudden 39 ot offis /

19 tee t tueba 40 ut u uther /

20 ue ue uesfool 41 oot CD wcodn't /

21 vee V very 42 00 (SO tattca) /

22 wae w werher 43- ow ou allou /

23 yae y yuqster 44- oi oi annoi /

24 zee z zombi

i.t.a. The Initial Teaching Al- alphabet. The 24 traditional closely resembled traditio- reduced to one. The system the appearance of each let-
phabet, devised in 1 959 by letters are retained. Capitals nal orthography, to ensure is a compromise between ter in only one form helpful,
James Pitman (1901 — are larger versions of the an easy transfer to normal simplification and familiarity. because it would remove
1 a system of 44
985), is lower-case letters. spelling in due course. As a The not a 'method'
i.t.a. is variations the visual pat-
in
lower-case letters, each The i.t.a. is not a proposal result, certain features often for teaching reading terns of words.
corresponding to a single for the permanent reform of eliminated in spelling- (p. 250); but it can be used At its peak, i.t.a. was be-
phoneme (§28). Extra sym- English spelling, but a sys- reform proposals are re- in relation to several such ing used in several coun-
bols are introduced to han- tem intended to assist chil- tained on the grounds that methods. For example, pho- tries, including some where
dle contrasts not systemati- dren in their first encounter they will aid this transition; nic approaches might be English was being taught as
cally represented by with reading. Irregularities forexample, the system aided by the consistent links a second language (§62). In
traditional orthography are not eliminated, simply keeps double letters, as in between graphemes and recent years, however, its
hence its characterization postponed. The main aim appl, and vowels in un- phonemes; and look-and- popularity has declined.
as an 'augmented' Roman was to design a system that stressed syllables are not all say approaches might find

A successful reform
Anahfis of Sounds in the Engli/h-
The differences between and
British U.S. Britain

American spelling show that changes can be program programme (except

introduced if circumstances are right. The in computing) Language.
changes derive from the rules introduced by theater theatre*
Noah Webster ( 1 758-1 843) such as the
, traveler traveller*
use of -orfor -our and -error -re. In his later worshiping worshipping
writings, Webster came to advocate spelling
come to be used in both
Several spellings have
reform.
countries, but th ere are still strong usage
Examples of the spelling differences twenty-Ate (ingle charac-
include the following (* indicates a rule that
preferences:
IN the English alphabet there
that rtand as rcprefimtstitm of certain founds.
ters,
i.ii

U.S. Britain
applies to many other words of the same A, b, e, d, e, f, g, i, j, k, I, in, n, 0, p, q, r, f,

type). encyclopedia encyclopaedia* t, u, r, w, x, y. z. H is not a mark of foaad, but it

qualitiesor gives (o4m to a iucceedir>2 found.

U.S. Britain jail gaol
In order to understand ihife Ulterior rather the found*
leaped leapt*
they teprefent, il is neceftary to dearie the meaning ef the
catalog catalogue omelet omelette woids, vowel, dij-hthong, and confonaot.
crudest cruellest* practice (verb) practise A
vowel is a iiniplc articulate found. fimple found A
gray grey isformed by opening the mouth in a certain niar.ner, with-
honor honour* out any contact of the pans of it. Whenever a found can
license (noun) licence Ue begun and compleaced with die fame poGtion of the organs,
Right: The opening of Webster's success-
liter litre* it it a fimi>!e found.
ful 'Blue-backed Speller' (1783)

34 THE PROCESS OF READING AND WRITING •

217
 .r. -^xy^*^ . i. _,j^^jrw

1
•v

" .

;np a

§^ 1 i^lsJSB
x<
PART VI
The medium of language:
and seeing
In recent years, the study of spoken and written
language has been supplemented by interest in a
third means of linguistic communication - sign lan-
guage, or simply, 'sign'. This mode of behaviour
is of particular importance because, like speech
(and unlike writing, and other codes), several of
its manifestations have a natural, biological basis.
Most of this interest is directed at the various
natural sign languages used by the deaf population
(§46), but any survey must also take into account
the nature of the artificially constructed sign lan-
guages and systems that have been devised to help
hearing people communicate with the deaf. Other
kinds of signing behaviour used by the hearing
population are more restricted in character; these
are described in relation to nonverbal communica-
tion in §64.
Sign is one of the most neglected aspects of com-
M munication study, as is reflected in the regrettably
small extent of this part of the encyclopedia. The
neglect is due to several factors — not least the
widespread popular as well as scholarly reluctance
to accept the possibility that sign could be a real
language, worthy of systematic study. Researchers
A filmed sequence of signs from British Sign Language, used
as part of an experiment into the way deaf people remember
signs. The signs, in sequence, are: true, you, chew, car,
brother, two, blue, through, new, who, few, cold, agree, milk,
worth, wash, shoe, break, paper, make.
«
signing
in the field face several difficulties. The number of
people for whom signing is their natural, everyday
language is relatively small, and it is not always
easy to find people who are able to communicate
in a fluent and unselfconscious way. Moreover, it
is only since the 1970s that film and video tech-
niques have become sufficiently routine to enable
the basic data to be recorded for analysis. But
recording the data is only a first step: special ways
of analysing and transcribing the data have to be
devised, to enable appropriately detailed descrip-
tions to be made. As a result, the true complexity
of this medium of visual communication has begun
to be appreciated only in recent years.
Part vi therefore begins with a consideration of
several popular fallacies about sign language, in
particular stressing the importance of seeing sign
as a language in its own right. The study of sign
language structure is then introduced with refer-
ence to American Sign Language, which has
attracted most of the linguistic research since the
1960s. Part vi concludes by looking at some of
the other signing systems that have been developed
for use with the hearing-impaired population.
 35 Sign language Comparing sign languages
When a comparison is sign for father, but interpret low and separate, but the
made of different sign lan- itas the ASL sign secret, particular CSL combination
guages, structural differ- CSL help is equivalent to is not an ASL sign.

ences clearly emerge. In a ASL push. (After E. Klima & U. Bellugi,

study of Chinese Sign Lan- • Some CSL signs use 1979.)
guage (CSL) and Ameri- shapes or movements
Myths and reality can Sign Language (ASL),
sign
which are not possible for-
for example, it proved pos- mations in the ASL system,
The first step in considering the nature of sign lan-
sible to identify several sys- such as the signs for Wed-
guage is to eradicate traditional misconceptions tematic differences in the nesday and introduce.
about its structure and function. Popular opinions use hand shapes and
of • Some CSL signs have
about the matter are quite plain: sign language is movements. the form of possible signs
• There are signs in both in ASL, but are not in fact
not a real language but little more than a system
languages that have the used in ASL; for
actually
of sophisticated gesturing; signs are simply pic-
same form, but different example, elements of the
torial representations of external reality; and meaning; for example, ASL CSL sign for distracted are
because of this, there is just one sign language, signers recognize the CSL like the ASL signs for yel- CSL distracted

which can be understood all over the world. It is

now clear, from the results of the first research stu-
dies of this subject, dating from the 1960s, that
all of these opinions are wrong.

A clear distinction must be drawn, first of all,

between sign language and gesture. To sign is to
use the hands in a conscious, 'verbal' manner, to
express the same range of meaning as would be
achieved by speech (especially by grammar, §16).
By contrast, gesturing is far less systematic and CSL father, like ASL sec- CSL Wednesday ASL yellow
comprehensive; there are in fact very few hand ges- ret

tures (§64), and these are used in an ad hoc way

to express a small number of basic notions. Every-
one can gesture; but few have learned to sign. (A
similar point can be made about facial expressions
and body movements.)
Some of the hand movements of sign language
can be plausibly interpreted by non-signers because
they reflect properties of the external world (they
are iconic); but the vast majority of signs are not.
CSL help, like ASL push CSL introduce ASL separate
It is possible that many of the signs were iconic
when they were first devised, but little information
is which some
available about this point in the past, Iconic signs?
have speculated may be as early as the origins of The signs of sign language

human language (§49). In any case, whatever the are often deceptively iconic.
After its meaning has been
original situation, the iconicity has been lost in
revealed, a sign may appear
most instances because of the influence of linguistic 'obvious'; but it proves not so
change, which affects sign as it does spoken lan- easy to predict the meaning
guage (§54). from the shape of the sign
alone. This can be seen
As a result of linguistic change, and because
even in a sign language
of independent creation in different parts of the which aims for maximum ico-
world, no single sign language exists. There are nicity. The pictures are of
many such languages (American, French, Danish Gestuno, a system adopted
by the Unification of Signs
. .), and they are not mutually intelligible. They
.

Commission of the World Fe-

use different signs and different rules of sign forma-
deration of the Deaf in 1 975.
tion and sentence structure. Even within an area Once one is told that these
that uses the same spoken language, the differences are animal signs, it becomes
may be so great as to preclude mutual comprehen- possible to make reasonable
guesses at what they mean
sion - as happens, for example, between British
- but even so, the guesses
and American Sign Language (BSL and ASL). are often wrong (glosses are
Sign languages have a structure of comparable given at the foot of the page).
complexity to spoken and written language and
perform a similar range of functions. There are
rules governing the way signs are formed, and how
they are sequenced — rules that have to be learned, 1 : hen 2: horse 3: lion
either as children (e.g. from deaf parents, §46) or 4: monkey

220 VI THE MEDIUM OF LANGUAGE: SIGNING AND SEEING

as adults (e.g. when working with the deaf). There Some ASL sign
are a large number of signs available within a sign symbols
language (around 4,000 have been recorded in An important stage the history of sign language
in
analysis took place the 1960s, when the term
in Tab
ASL), and these are used to convey a considerable cherology was coined on analogy with phonology TT neck
range of meaning. When two fluent signers commu- (§ 28) to refer to the study of the contrastive units
a sign language. The
H trunk
nicate, they provide impressive evidence of the (cheremes) that occur in

structural analyses subsequently made provided a \ upper arm

creative potential of sign, and of its social and psy-
valuable indication of the difficulty researchers face as / elbow, forearm
chological reality as a language.
they try to 'capture' the dynamic, multi-dimensional
properties of sign.
& wrist, arm on its back

MODERN DEVELOPMENTS In this approach, three classes of cheremes are t) wrist, arm face down
Very little information is available about the early W. C. Stokoe etal. 1965):
identified (after
• tab {tabula) the location in the sign space where a Dez
history of sign languages. References to deaf sign- 5 spread hand
sign is made;
ing are found in Greek and Roman writings, but • dez (designator) the active hand configuration used G index finger points from
there are no study of
details. In recent times, the tomake the sign; and fist

signing is dated from the work of the French edu- • sig (signation) the action of the active hand. Signs H index and second finger,
cator, Abbe Charles Michel de l'Epee (1712-89), are described as simultaneously occurring side by side, extended
combinations of tab, dez, and sig.
who in 1775 developed a sign language for use Various constraints affect the use of these contrasts.
I little finger extended from
in a school for the deaf in Paris. The origins of Not all possible combinations occur. Some are
compact hand
his system are obscure: several of his signs were physically impossible. Some are not used by K index finger points from
modifications of those used by the French native convention - for example, signs are not made from and thumb touches
fist,

head to shoulder but from head to chest. There is a middle of second finger
deaf population, but he also made some use of the
strong tendency towards hand symmetry: if a sign L thumb and index finger at
Spanish manual alphabet (p. 225), and he may have requires two active hands, both hands will have right angles, others
incorporated some of the signs used by the Spanish identical shapes and orientations. Several such usually bent into palm
Benedictine monks. Several foreign educators stu- constraints govern the structure of a sign language,
and a major focus of recent research has been to O and
fingers curved
died at his school, and the influence of this system squeezed together over
discover the rules governing sign formation, and the
spread to many parts of the world, including thumb
contexts (such as poetry, irony, or humour) where
Russia, Ireland, and America. For example, the departures from these rules are tolerated. V index and second fingers
American educator Thomas Gallaudet (1787— extended/spread ('vic-
tory')
1851), together with Laurent Clerc (1785-1869),
a teacher of the deaf, brought the signs to the USA, Transcribing signs W thumb and little finger
touch, other fingers
where they came to be used alongside those already A series of signs with their tab and dez elements transcribed extended/spread
in use by the American deaf population. Modern (see below). Sig actions cannot be clearly shown without
American Sign Language (ASL, or Ameslan) der- moving The pictures show the right hand as dez and
film.
Sig
the when used, as tab. Left-handed signers often use
left,
a upward movement
ives from this system.
the left hand as dez. The triple mark above the V shows that
When a sign language becomes widely used, it the sign is made with the fingers bent.
v downward movement
develops the same kind of dialects and varieties > rightward movement
as occur in spoken language. This kind of variation
Tab IT Dez G
a palm-up rotation
can be seen in ASL, which is now used by over
o palm-down rotation
half a million deaf people, many of whom use it
» closing action
as a native language. Some varieties are regional
o circular action
but many are due to the age at which
in origin (§8),
the sign language is learned, and to social factors, o entering action

such as the home environment (whether the parents

are deaf) and the educational background of the
Some ASL words
signer. Afurther important variable is the extent
Tab \ Dez H O 5g°
to which the sign language has been influenced by
the spoken language of the community. A dialect . / % included

continuum (p. 25) seems to exist among the Ameri- Oo F a

I
soul, spirit(ual)
can deaf, and this is probably a universal phenome- 0° F^
non. The continuum ranges from ASL varieties that
m ^M I tea
show no influence of spoken language to those that
have been markedly shaped by properties of Eng-
lish - by word order, in particular. Several pidgin
varieties of signing exist along this continuum
($55).
/
Tab a Dez W,
Note:
0° H£
resign, quit

When two sig

elements are placed hori-
zontally, they are signed
in sequence; when placed
vertically, they are signed to-
gether.

35 SIGN LANGUAGE. 221

36 Sign language structure

A great deal is now known about the structure of
natural signing, following several years of detailed
study of American Sign Language (ASL). It has
become clear that this language has a highly devel-
oped structure which needs to be described in
own terms. And as research continues on other sign
languages, similar conclusions are beginning to
emerge.
Describing sign language 'in its own terms' is,
however, easier said than done. We are so used
to thinking of language in terms of the structures
of speech or writing that it is extremely difficult
to grasp what is going on when a completely differ-
ent medium is involved. This can be seen by consi-
dering the effect of a monologue in sign when it
is given a fairly literal 'translation' into English.
Here is an extract from one study (I. M. Schlesinger
&L.Namir, 1978, p. 100):
Two children. One marry. Two grandchildren. Work
close frat building. One still school. Mother gone. My
aunt true me phone. Sorry can't funeral. Me work. Me
awful cry. Come night.
This kind of transcription inevitably gives the
impression of a reduced or simplified language -
nitely large number of signs, and it is possible to
see several organizational principles operating.
'Locations' can be established that identify differ-
ent sentence elements or semantic functions.
its
• Time relationships can be expressed by dividing
the space into neutral (present), further forward
(future), and further back (past) areas; these
areas can then be used both for tense forms and
for time adverbs {then, now, next, last, etc.).
• Several persons (pronouns) can be distinguished
using different spatial areas: you is front-centre;
one third-person form is signed to the right;
another to the left; and others divide up the
intervening space. Moreover, once a space is
established for a given person, it is normally
'reserved' for that person for the remainder of
the conversation.
• Questions can be signalled by an appropriate
accompanying facial expression, such as raised
eyebrows and backwards head tilt.
• Great use is made of reduplication (p. 175) to
express such notions as plurality, aspect, degree,
or emphasis; for example, such verbal meanings
as continuity, repetition, or habituality can all

the grammar is highly telegraphic, there are no be signed by repeating a verb sign with varying
inflectional endings, and words are omitted. But speed.
this is to judge one language by the criterion of • The use of pause between signs or sign sequences
another - something that breaks a cardinal princi- is available to mark grammatical boundaries.

ple of modern language study (§2). A word-for- • The whole spatial area can be enlarged or con-
word translation from a foreign language (§57) fined to express 'louder' or 'quieter' signing.
would give just as odd a result. In omitting articles
or the copula verb be (as in man happy), for exam-
ple, ASL is no different from Russian. And, conver-
sely, ASL makes use of many conventions that
other languages lack — such as the simultaneous
The normal signing space
use of signals (e.g. hand signs plus facial expres-
sions, eye movements, and shifts of the body), that
literal translations cannot easily convey. Moreover,
the speed of fluent signing - between one and two
signs per second - produces a conversational rate
The time line
that is comparable to that of speech. It usually takes
A line along the vertical
longer to make a sign than to pronounce a word, plane, near the signer's ear
but many signs express a meaning far more suc- and cheek, is regularly
cinctly than the corresponding spoken output. used to express time rela-
tionships.
(From N. Frishberg, 1979.)
THE USE OF SIGN SPACE
The expressive potential of sign can be appreciated
only by looking at signing behaviour in some detail.
There is, first of all, a three-dimensional sign space:
vertically, this consists of the distance just below
the waist to the top of the head (signs are rarely
made above the head, below the waist, or towards
the back of the head or body); laterally, the space
forms a 'bubble' which extends outwards in front
of the signer from extreme right to extreme left.
Within this space, there is room to make an indefi-

222 •
VI THE MEDIUM OF LANGUAGE: SIGNING AND SEEING
Aspect modulations
ASL makes use of a complex system of simultaneous sign Structure contrasts
'modulations', analogous in function to the sequential
The non-manual activities which accompany a sequence of signs can have an
inflections of spoken morphology (p. 90). These can be
important structural function in ASL.
illustratedby the forms used to convey semantic contrasts
of aspect (p. 93). In one study, the variations on the sign for
sick were analysed, and several aspectual modulations were
recognized. Eight of these are illustrated in the figure below.
The horizontal dimension represents the relative length of the
sign,and the vertical divisions show the number of repetitions
of themovement. The blank sections represent the relative
durations of holds' (i.e. points at which the hand is held
steady). Muscular tension, accentuation, and certain other
characteristics of the movements are not represented.
(After E. Klima & U. Bellugi, 1 979.)

Uninfected sign meaning 'be sick'

(a) The sequence of signs 'woman-forget-purse' is used as a statement, The woman

forgot the purse (the articles are not separately signed).

Predisposirional aspect (circular modulation)

'prone to be sick'

Susceptative aspect (thrust modulation)

'get sick easily'

Continuative aspect (elliptical modulation)

(three repetitions)
'sick for a long time'
(b) The same sign sequence is accompanied by a forward movement of the head
and shoulders, and the eyebrows are raised: this would express the yes-no question,
Did the woman forget the purse?

*» Incessant aspect (tremolo modulation)

'never stops being sick'

Frequentative aspect (marcato modulation)

'often sick'

3 4 s

Intensive aspect (tense modulation)

«
'vervsick'
(c) The same sequence is used as part of the sentence 'woman-forget-purse-

I recently-arrive' ( The woman who forgot the purse has just arrived). Here, the relative

clause section The woman who forgot the purse) is signalled by having the brow and
(

upper lip raised, and the head tilted back.

(After S.K. Liddell, 1980.)
Approximative aspect (lax modulation)
'sort of sick'

1 _i_i_l_S 6 7 g 9 10 11 12 13

Resultative aspect (accelerando modulation)

'became sick'

Relative sign duration

36 SIGN LANGUAGE STRUCTURE •

223
37 Types of sign language

Several sign languages may be in regular use within cational domain, and the sincerity and enthusiasm
the boundaries of a particular speech community. of the creators, all of the systems have been put
The most widely used are the concept-based sys- to valuable use in a range of teaching situations.
tems that have developed naturally among the deaf However, objective techniques for evaluating their
communities, and it is these that are most com- relative strengths and weaknesses have not so far
monly referred to as 'sign language' - American been devised.
Sign Language, British Sign Language, Danish Sign Each of these systems aims to reflect the structure
Amer-lnd
Language, etc. In addition, in recent years edu- of English, but they do this in different ways. All
Over the centuries, the North
cators and linguists have devised several new kinds follow English word order, but they differ in the
American Indians have spok-
of signing system. These are mainly taught to deaf way they form signs, and in how much finger spell- en hundreds of languages
children or adults, but they are also sometimes ing (see below) they use. Many arbitrary decisions from several different fami-
found being used with other handicapped popula- have to be made by the system's creator; for exam- lies (p. 320). It is not surpris-
ing, then, that they deve-
tions, such as the mentally retarded. ple, it is not obvious how to allocate signs to such
loped a form of signing as a
The greatest proliferation of new signing systems forms as irregular nouns, verbs, or adjectives (§16). means of communication be-
has been within the English speech community. Should past-tense forms {took, gone, etc.) be signed tween different tribes. Fol-
Most approaches involve modifications of ASL or with the same sign as past participle [taken, went, lowing early descriptions of
this 'hand talk', an adaptation
BSL, with the aim of bringing the signing closer etc.) or with different signs? Should took be signed
was made for use with the
to spoken English. Several of these systems emerged as 'take + past', 'take + -ed\ Hake + e + d\
handicapped by Madge
in the late 1960s in the USA, notably Seeing Essen- or '£ + o + o + &'? There are many such possibili- Skelly(1903- ), an Indian-
tial English (1966), and its two derivatives, Linguis- ties, and different systems go in different directions, born speech pathologist. The
tics of Visual English (1971) and Signing Exact with varying degrees of consistency. system is conceived as a
gestural code, rather than a
English (1972). Other systems were devised that
language. It contains a li-
closely followed the structure of speech, such as mited number of signs, re-
Signed English (1969) and Manual English (1972). presenting concrete mean-
Given the urgent need for progress in the edu- ings, and it has no
grammatical structure apart
from sequence. The signs
are chosen so as to be im-
Paget-Gorman mediately recognizable, so
Sign System (PGSS) that the viewer can interpret
without formal instruction,
The earliest proposal to be
and regardless of language
widely adopted in modern
background. Four signs are
times was based on
illustrated below.
Richard Paget's A System-
atic Sign Language (1 951 ).
After his death in 1 955,
system was developed
this
by his widow,
in Britain

Grace Paget, and Pierre

Gorman, at that time lib-

rarian of the Royal National

Institute for the Deaf. It
contains some 3,000 signs,
representing the words and
morphemes (p. 90) of spok-
en English. Sentences are
signed following English
word order. The system
makes use of a set of
'basic' signs - semantic
fields such as 'action',
'animal', 'colour', 'con-
and 'food'. Different
tainer',
words belonging to each
are identified with
field
reference to the same basic The different colour words, for example, are all derived
sign, plus an identifying sign. from the basic sign above. To sign blue, one hand is
held as for colour, while the first finger of the other hand
is held pointing up, back outwards, in line with the
Six of the PGSS basic signs signer's side (i.e. the colour of the sky). To sign red, the
1 action 4 container same basic sign is used, while the other hand makes
2 animal 5 thing the sign for blood.
3 colour 6 think
hungry

224 • VI THE MEDIUM OF LANGUAGE: SIGNING AND SEEING

Finger spelling
Finger spelling, or dactylo-
logy, is a signing system in
which each letter of the or-
dinary alphabet is given its
own sign. The principle can
be applied to any language
which has developed an al-
phabetic writing system.
However, there are conven-
tional differences: in particu-
lar, the British manual alpha-
bet is formed using two
hands, whereas the Ameri-
can and Swedish systems,
for example, use only one.
The main strength of finger
spelling is its great scope
and flexibility. It is quick to

learn and can then be used

to sign an indefinite number
of words. It is a particularly
useful system for signing
proper names, which are not
given their own signs in other
sign systems. However, it is
a slow system to use, rarely
exceeding 300 letters per
minute (about 60 words).
Moreover, it cannot be used
at all unless one is able to
spell (a problem for young
children, who also have diffi-

culty controlling the hand

shapes required), From the
receiver's point of view, it is SWEDISH
difficult to distinguish the
hand shapes at a distance,
and, even close to, intelligibi-
Cued speech
lity can be a problem if the Normal lip-reading techniques allow only certain sounds
rate of signingspeeds up, - those towards the front of the mouth - to be easily
and the signer begins to omit distinguished; and there are many sentences which lip
letters. readers find difficult to make out, especially when the
Finger spelling is best context is unclear (such as It is in the tin, where the lip
thought of as an auxiliary position is almost identical throughout). Cued speech

signing system, a convenient hopes to eliminate such difficulties by making it possible

bridge between spoken or for a deaf person to see' the sounds of speech as they
writtenlanguage and sign are spoken. It is a system of hand cues that are used
language proper. The use of alongside lip movements to draw attention to the
the method has been docu- phonemic (§ 28) contrasts of speech. The system was
mented from the 1 7th cen- devised in 1966 by the American educator R. Orin
The philosopher
tury. Cornett (1913- ), and has since been adapted for use
it

George Dalgarno, for exam- in over 30 languages.

ple, recommended its use by Below: The system uses 36 cues for the 44 English
all members of a family phonemes. Vowel cues are shown by the position of the
whenever it contained a deaf hand. Four positions are recognized: at the side, throat,
child, arguing that the acqui- chin, and mouth. Each position signals a group of three
sition of spelled language vowels of different lip shapes; vowels with the same lip
would thereby be as natural shape can then be readily distinguished by noting the
as the acquisition of spoken accompanying sign. Consonant cues are shown by the
language. In modern times, shape of the hand. There are eight hand shapes, each
some educational ap- of which is associated with a group of consonants of
proaches make a great deal different lip shapes; as with vowels, consonants with the
of use of it - the Rochester' same lip shape can then be distinguished by noting the
method in the U.S., for ex- accompanying sign.
ample, based on a combi-
is

nation of finger spelling and

speech, and it is reported
that a Cyrillic manual alpha-
>^
bet is widely used in the Side position Throat position Chin position Mouth position
Soviet Union.

AMERICAN

37 TYPES OF SIGN LANGUAGE •

225
PARTVH
Child language acquisition
The study of how children learn to speak has
proved to be one of the most fascinating, impor-
tant, and complicated branches of language study
in recent years. The fascination of the subject stems
from the natural interest people take in the develop-
young children. Its importance lies
ing abilities of
in the way that language acquisition research can
assist our understanding of language as a whole,
and also in the many applications of this research
— especially in the field of child language handicap
(Part vin). The complexity arises from the enor-
mous difficulties that are encountered as soon as
anyone attempts to establish and explain the facts
of language development, especially in the very
young child.
Part vn begins, therefore, with a discussion of
some of the approaches and methods that have
been used to find out about children's language
- diaries, recordings, experiments, tests, profiles,
and other procedures. The result of this enquiry
has been an explosion of information about many
details of language acquisition, and an increased
awareness of such general issues as the nature of
developmental stages, and the relationship between
speech production and comprehension in the
course of early learning. Several major theoretical
accounts of child language acquisition have also
been proposed in recent years, and these are
reviewed in the final part of this section.
An ingenious way of fostering an early interest in
'reading express' in a public Library.
Following these general observations, sections
39-43 examine different aspects of the language
acquisition process in somewhat greater detail. We
begin with the early development of vocalization
in infants during the first year of life, and the asso-
ciated emergence of the skills of speech perception
and speech interaction. Around 1 year of age, a
more clearly defined linguistic ability is apparent,
and it then proves possible to begin analysis in con-
ventional linguistic terms, using the distinctions
recognized in Parts in and iv. We look separately
at phonological, grammatical, semantic, and prag-
matic development, with particular reference to
studies of preschool children.
Once children arrive in school, they meet a com-
pletely fresh range of factors that influence their
language development. The final section of this Part
therefore reviews recent and contemporary educa-
tional approaches to the question of how linguistic
skills should be fostered in school. We begin with
a discussion of the issues that arise in relation to
spoken language (or 'oracy'), proceed to a review
of the corresponding approaches that have been
proposed in relation to the teaching of reading,
and conclude with an account of current thinking
about the most neglected area of all, the child's
developing awareness of written language.
words: a

38 Investigating children's language
For over 200 years, scholars have shown an interest
Ancient questions
in the way children learn to speak and understand
Child language study has
their first language. Several small-scale studies were
exercised its fascination on
carried out, especially towards the end of the 19th rulersand scholars alike
century, using data recorded in parental diaries. over 2,000 years, espe-
But detailed, systematic investigation did not begin cially in relation to such
questions as the origins
until the middle decades of the 20th century, when
and growth of language
the tape recorder came into routine use. This made Many felt that the
(§49).
it possible to keep a permanent record of samples study of linguistic develop-
of child speech, so that analysts could listen re- ment in the child (language
peatedly to obscure extracts, and thus produce a ontogenesis) would pro-
vide clues about the
detailed and accurate description. Since then, the
linguistic development of
subject has attracted enormous multi-disciplinary the human race (language
interest, notably from linguists and psychologists, phylogenesis). Some inter-
who have used a variety of observational and esting similarities have
been noted between the
experimental techniques to study the process of
vocal tracts of infants and
language acquisition in depth. non-human primates (§49),
Central to the success of this rapidly emerging but there is still a great gap
field lies the ability of researchers to devise satisfac- between the emotional ex-
pression of infants and the
tory methods for eliciting linguistic data from chil-
propositional content of
dren. The problems that have to be faced are quite adult language, which stu-
different from those encountered when working dies of acquisition have not
with adults (p. 410). Many of the linguist's routine yet been able to bridge.
techniques of enquiry cannot be used with children. Someone who was re-
markably modern in his
It is not possible to carry out certain kinds of exper-
iments, because aspects of children's cognitive
development — such as their ability to pay attention,
or to remember instructions - may not be suffi-
Parental diaries
ciently advanced. Nor is it easy to get children to
The earliest approach to
make systematic judgments about language — a task the study of child language
that is virtually impossible below the age of 3. And was to keep a written diary
anyone who has tried to obtain even the most basic of observations about
kind of data — a tape recording of a representative one's own child. Several
1 9th-century scholars en-
sample of a child's speech - knows how frustrating
gaged in this task, includ-
this can be. Some children, it seems, are innately ing August Schleicher and
programmed to switch off as soon as they notice Charles Darwin. The ap-
a tape recorder being switched on. proach fell out of favour
with the advent of audio-
Since the 1960s, however, several sophisticated
and video-recording techni
recording techniques and experimental designs ques, which permitted a
have been devised. Children can be observed and more systematic, objective,
recorded through one-way-vision windows or and comprehensive analy-
using radio microphones, so that the effects of hav- sis. It has nonetheless
occasionally been used
ing an investigator in the same room as the child
with good effect in recent
can be eliminated. Large-scale sampling pro- decades - notably in
grammes have been carried out, with children Werner Leopold's four-
sometimes being recorded for several years. Parti- volume study of his
daughter Hildegard, pub-
cular attention has been paid to devising experi-
lished between 1 939 and
mental techniques that fall well within a child's 1949, Speech Develop-
intellectual level and social experience. Even pre- ment of a Bilingual Child.
linguistic infants have been brought into the Below are some extracts
research: acoustic techniques are used to analyse from the earliest known
diary study, by the German
their vocalizations, and their ability to perceive the
philosopher Dietrich Tiede-
world around them is monitored using special mann (1748-1803) about
recording equipment (§39). The result has been a his son, Friedrich, kept be-
growing body of reliable data on the stages of lan- tween 1 782 (when the child
was 6 months) and 1 784:
guage acquisition from birth until puberty.
228 VII CHILD LANGUAGE ACQUISITION
. .
views was the Mogul Em-
peror of India, Akbar the
Great (1542-1 605). He be-
lieved that speech arose
for
from people listening to
others, and
who were isolated from hu-
man contact would not be
able to speak. A contem-
porary Persian account,
the Akbamama of Abu'l-
Fazl, takes up the story:
As some who heard
appeared to deny
order to convince them,
had a serai [mansion] built
in a place which civilized
sounds did not reach. The
newly born were put into
that place of experience,
and honest and active
guards were put over them.
For a time, tongue-tied wet-
nurses were admitted
there. As they had closed
the door of speech, the
place was commonly called
theGang Mahal (the dumb-
house). On the 9th August
On February 10th he
showed the
surprise and approval; so
far his only
pain, anger, impatience,
and pleasure had been
ing, writhing, laughing.
Now, when he saw some-
thing new and delightful, he
greeted it with the exclama-
tion 'ach!' - the
of admiration
After all
cise in the production of
tones, and
sition of some
using the speech organs
variously, he commenced,
on the 1 4th of March, to ar-
ticulate consciously and to
repeat sounds. His mother
said to him the syllable
'Ma'; he gazed attentively
at her mouth, and at-
tempted to imitate the syll-
able . .
A few words he pro-
nounced clearly on No-
vember 27th and knew also
their meanings exactly;
these were Papa' and
Mama' . .
On the 8th of March,
. .
1 582 he went out to hunt.
That night he stayed in Fai-
zabad, and next day he
went with a few special at-
tendants to the house of
that children experiment. No cry came
from that house of silence,
nor was any speech heard
there. In spite of their four
years, they had no part of
the talisman of speech, and
nothing came out except
the noise of the dumb.
this
(From H. Beveridge, 1897-
it, he, in
1910, pp. 581-2.)
Akbar the Great
the sight of an object, he
first signs of would repeat its name if he
had frequently heard it, but
expressions of he still found it hard to pro-
nounce words of several
cry- syllables.
On the 30th of July he
finally succeeded in utter-
ing complete, though short
sentences, for example:
natural sign There he stands, There he
. . lies . .
manner of exer- [February 14, 1784] This
is as far as my observa-
after the acqui- tions go. Other business
skill in prevented me from their
continuation. greatly de-I
sire that others may make
similar ones; it will then be
possible to determine var-
ious things by comparison,
and that important branch
of psychology, too little ex-
ploited as yet, which stu-
dies the development of
human - the foun-
faculties
dation of pedagogy - will
make appreciable progress
thereby. (From C. Murchi-
son & S. K. Langer, 1927.)
at
RESEARCH PARADIGMS Sampling: how much? how often? Research
There is no single way of studying children's Those who do research in child language are always cooperation
language. Linguistics and psychology has each being pulled in two directions, when they have to Children do not always see
brought its own approach to the subject, and many decide questions of sampling. They can choose to the need to cooperate to the
variations have been introduced to cope with the follow a single child, or a small group of children, best of their ability in lan-

which children engage, and guage acquisition research,

variety of activities in in an intensive way, taking relatively large samples
as the following story shows:
the great age range that they present. Two main at frequent intervals. Or
they can select a large
research paradigms are found. number of children and take smaller samples at Another week we noticed
less frequent intervals. Both procedures have their thatAdam would sometimes
Naturalistic sampling A sample of a child's spon- pluralizenouns when they
strengths and limitations. The former enables the
should have been pluralized
taneous use of language is recorded in familiar and researcher to plot the gradual emergence of linguis- and sometimes would not.
comfortable surroundings. One of the best places tic patterns from absence to acquisition; but it is We wondered he could
if

to make the recording is in the child's own home, unable to provide confident generalizations about make grammatical judg-
but it is not always easy to maintain good acoustic ments about the plural, if he
these patterns, given the small number of children
could distinguish a correct
quality, and the presence of the researcher or the examined. The latter permits such generalizations, form from an incorrect form.
recording equipment can be a distraction (espe- but is likely to miss points of significant progress 'Adam', we asked, 'which is
cially if the proceedings are being filmed). Alterna- that fall between the sampling intervals. right, "two shoes" or "two

tively, the recording can be made in a special Depending on the method used, therefore, sam- shoe"?' His answer on that
setting, such as a research centre, where the child occasion, produced with ex-
pling intervals can range from every few days, espe-
plosive enthusiasm, was
is allowed to play freely with toys while talking cially when the children seem to be undergoing a 'Pop goes the weasel!' The
to parents or other children, and the observers and period of rapid progress, to 3 months or more. two-year-old child does not
their equipment are unobtrusive. The major research programme launched by the make a perfectly docile ex-
A good quality, representative, naturalistic sam- perimental subject. (From R.
American psychologist Roger Brown (1925- in )
Brown & U. Bellugi-Klima,
ple is generally considered an ideal datum for child the 1960s sampled three children for at least two 1964, p. 134.)
language study. However, the method has several hours a month — in one case, for half an hour a
limitations. These samples are informative about week. By contrast, a British programme of the
speech production, but they give little guidance 1970s, directed by the psycholinguist Gordon
about the way children understand what they hear Wells (1935- ), involved 128 children, and took
around them. Moreover, samples cannot contain a half-hour sample from each child every three
everything, and they can easily miss some impor- months. Even larger numbers of children are some-
tant features of a child's linguistic ability. They may times used, but this restricts the research to the
also not provide enough instances of a developing study of a very small set of linguistic features. It
feature to enable the analyst to make a decision should be borne in mind, too, that large samples
about the way the child is learning. For such rea- do not guarantee the occurrence of important fea-
sons, the description of samples of child speech tures. In the Wells project, a search for passive
has to be supplemented by other methods. verbs (e.g. was kicked) in 18,000 utterances from
60 children who were recorded three times between
Experimentation The methods of experimental 3 and 2>\ years of age, produced next to nothing:
psychology have been widely applied to child lan- 12 children used such a verb a total of 19 times!
guage research. The investigator formulates a spe- Half-hour samples are a popular measure,
cific hypothesis about children's ability to use or though often people use a sample consisting of a
understand an aspect of language, and devises a fixed number of utterances (e.g. 100 utterances
relevant task for a group of subjects to carry out. taken from some point in a recording session).
A statistical analysis is made of the subjects' beha- Whatever the length, samples need to be as repre-
viour, and the results provide evidence that sup- sentative as possible of the child's language, and
ports or falsifies the original hypothesis — or, at researchers therefore need to anticipate the in-
least, suggests ways in which the experiment might fluence of such factors as time of day, the nature
be better designed next time! of the setting, and the presence of observers
Using this approach, as well as other methods (p. 231).
of controlled observation, researchers have come
up with many detailed findings about the produc-
tion and comprehension of groups of children.
However, it is not easy to generalize the findings Longitudinal vs cross-sectional
of these studies. What may obtain in a carefully Studies that follow the guage emergence, by
controlled setting may not apply in the rush of daily progress of a set of vari- studying a set of vari-
interaction. Different kinds of subjects, experimen- ables over time in the ables in a group of chil-
tal situations, and statistical procedures may same set of children are dren of different ages,
known as longitudinal using different subjects at
produce different results or interpretations. Experi- each age. This is known
studies. Most child lan-
mental research is therefore a slow, painstaking guage research is of this as a cross-sectional
business; it may take years before researchers are form. However, it is also study. Combined designs
convinced that all variables have been considered possible to build up a are also possible.
'composite' picture of lan-
and a finding is genuine.

38 INVESTIGATING CHILDREN'S LANGUAGE 229 •

 Talking dolls
How do we know when
young children are able to
recognize errors in what
people say? One ingenious
research technique made
use of a doll that was able
to 'talk'. A toy panda, about
75 cm tall, was brought into
a nursery where children
(aged between 3 and 5)
were playing. They were
told that this was a very
special kind of panda, be-
cause he was learning to
talk. He wanted the
dren to come and see him
one at a time, and talk to
him so that his speech
would improve. They were
all very willing to help.
In the test sessions, two
experimenters were in-
volved. One stayed in the
room with the panda and
the child, playing with var-
ious materials. The other
was outside the room, ob-
serving the session
through one-way-vision
glass and speaking into a
Task effects
Setting up an experimental
task so that it does not
hinder a child's perfor-
mance is never easy. Even
the simplest tasks can hide
snags that make it difficult
or impossible to interpret a
response correctly.
the child is seated, how the
toys are arranged, and how
the experimenter gives the
instructions can all cause
problems. The apparently
simple instruction to Put
the car behind the lorry', to
test knowledge of behind',
illustrates some of the
culties.
I. The child is sitting
site the experimenter.
Should she put the car be-
hind the lorry from her own
point of view, at X, or from
the experimenter's, at Y, or
should she use her know-
ledge of the real world, and
place the car at the back
end of the lorry, at Z, as it
would appear when travell-
ing along the road? A fail-
ure to respond, or a wrong
placement, may reflect only
her confusion, not her lack
of knowledge of what the
preposition means.
230 VII CHILD LANGUAGE ACQUISITION
microphone linked to a
loudspeaker in the panda's
head. The children were
trained to press a bell when
the panda said something
they thought was right, and
to press a buzzer when he
was wrong. The panda
would also ask the children
why he was wrong, if they
did not spontaneously give
a reason.
The children adapted to
this situation enthusiasti-
chil- cally, and so the technique
was used in several kinds
of study. It proved
very good way of testing
sentence comprehension
and conversational skills.
In one of the comprehen-
sion studies, for example, a
car was placed each of in
four garages, and a fifth car
was left outside. The
then said such sentences
as all the garages have
cars in them' (which was
true) and 'all the cars are in
the garages' (which was
Where
diffi-
oppo-
2. Now the child is alongside
the experimenter, but there
is still a problem. The
of the lorry is facing her.
she is still faced with the
problem of what the experi-
menter intends.
3. A does not have a
ball
front and a back end, so
there should be no difficul-
ties from the real world
here. Unfortunately,
false).The children's reac-
tions thenshowed how far
they were able to grasp the
distinction between the
sentences.
The main reason for de-
veloping this method was
to reduce the extent to
which a child might be in-
Chu-Chu and child
fluenced by an adult exper-
exchanging names at the
imenter, or overawed by an
beginning of a session.
artificial test situation. It
proved to be an extremely
successful technique, and
it has since been used in
studies of speech produc-
to be a tionas well as of compre-
hension and interaction. By
putting the children 'in
charge', researchers are
able to elicit a natural
speaking style, and to ob-
serve several structures
(such as the use of com-
doll mands) that are often
avoided when talking to
adults. (After P. Lloyd & M.
Donaldson, 1976.)
been placed near the back
of the table, so that the
tail end child has difficulty reaching
So behind it. Also, she might
think that the car will fall off
the table if she places it so
far away. Such factors
could once again lead her
to act indecisively, or to put
the car somewhere else,
thus giving a misleading
impression about her
it has linguistic knowledge.
 Individual differences Technological revolutions
There has been a natural emphasis in language The invention of the audio
acquisition research on the universal characteristics of
tape recorder led to the first
development. However, there are many individual language
revolution in child
differences which also need to be taken into account. research methodology.
Children may vary in their rate and strategies of The invention of the video
learning for a variety of reasons, to do with such factors recorder may well prove to
as sex, intelligence, personality, and social be a second. Each tech-
background (§§6-10). There are fast developers' nique has its strengths and
and 'slow developers'. But it has not yet been possible limitations.
to generalize about the way these variables affect the
The audio tape recorder
course of language development. themore widely used
is
Popular notions nonetheless exist. For example, it means of obtaining
child
is widely believed that girls learn to speak more rapidly
language data. Audio tapes
than boys, and several researchers have noticed a and equipment cost less,
trend for girls to be linguistically superior, at early ages. and the technique is relati-
But there is negligible evidence for a definite effect. vely unobtrusive.
If radio
Samples tend to be very small, and measures microphones attached to
selective. Such differences as are found seem to be the child's clothes are
due more to the effect of the different ways in which used, the actual recorder
boys and girls are brought up, rather than to need not even be in the
physiological or genetic factors. Parental style and same room, and recordings
expectations seem to be far more important. of excellent quality can be
made.
However, an audio re-
cording gives no informa-
tion about what a child
doing. Gestures and facial
expressions, which are
often used to supplement
speech or show compre-
hension, are not available.
It may not be possible to in-
terpretsentences clearly:
on an audio tape, Pur that
over there makes very little
sense. It is possible to get
round these problems to
some extent, by having an
observer present who
makes notes on what is
happening. But this is far
inferior to a video record of
the event, which can be
viewed several times by
different researchers.
With video, the tiniest
features of non-verbal be-
haviour, and the role of the
is accompanying context, can
be transcribed and ana-
lysed. is thus a frequently
It
used tool in modern child
language research - espe-
cially in studies of compre-
hension and parent-child
interaction. But video stu-
dies are never straightfor-
ward: lighting, camera an-
gles, sound recording, the
intrusion of the camera,
and other matters need to
be carefully thought out if
an informative picture is to
be obtained.

'Fell down', says the child - a totally obscure utterance without the picture.

Observer effects

The presence of research observers in a recording

session may affect the mother more than the child!
But it has taken some time for this point to be
appreciated by researchers.
One of the first findings about maternal language
concerned the presence of grammatical expansions
when talking to a child. Mothers would often pro-
vide a gloss for their child's utterance which added
elements that were not present:

child: Go car.
mother: Yes, daddy's going in his car.

In Brown's research project 229), it was found

(p.
that expansions appeared in nearly a third of
mothers' interactions, in the early stages of learn-
ing. Their function seemed to be as a teaching aid
for the child, in that the mothers were providing
their children with a target that was slightly ahead
of their performance. However, in Wells's project CHILDES
(p. 229), very few expansions were found. How Modern methods of com- an international computer a raising of standards of
is this discrepancy to be explained? putational analysis and network. This is the main data analysis, because
data processing could well aim of the Child Language errors can be readily
The main factor is thought to be the presence
revolutionize the study of Data Exchange System checked and corrected,
or absence of observers. In Brown's approach, language acquisition. One (childes), which was estab- and extra analytical obser-
there were always researchers present; in Wells's, of the main problems fac- lished in 1 984 by an inter- vations incorporated.
there wereno researchers present, and the mother ing the child language re- national group of language However, a sharing of re-
was alone with the child most of the time. Wells searcher is that the collec- acquisition researchers. sources is possible only if
tion and transcription of It is now possible to researchers can agree on a
made use of radio microphones and a sampling data samples is extremely transcribe tape-recorded set of policies and standard
programme in which 90-second recordings were time-consuming. An hour data directly into computer conventions for obtaining,
made automatically at 20-minute intervals ofrecorded conversational files,where the material transcribing, and storing
throughout the day, so that the parent would be data can take 1 or more can be edited, analysed, child language data in com-
hours to transcribe, check, and duplicated. Files of puterized form. These are
unaware when a recording was taking place. With
edit,and type. It has there- data can thus be shared currently under discussion.
these parents, the frequency of expansions fore been proposed that, between researchers who It will take some years be-
increased only when another adult was present. once scholars have made have computer access to fore all the methodological
This suggests that the main function of expansions data
their transcriptions, the the central database, mak- problems can be solved,
is to act as a gloss for the benefit of an observer,
should be made available ing a considerable saving but the outlook for child lan-
to the wider research com- of time and money. The guage research is extre-
and not, as was first thought, solely to provide the munity through the use of process could also lead to mely promising.
child with extra grammatical information.

38 INVESTIGATING CHILDREN'S LANGUAGE- 231

 . 1

PRODUCTION, COMPREHENSION, use However, there is increasing evidence that

it. Elicited imitation
IMITATION this simple relationship does not always obtain. The technique of 'elicited imi-

'Acquiring a language' involves two distinct skills: Production may precede comprehension, or the tation' can be used to find out

two processes may be so intimately connected that what a knows about

child
the ability to produce speech in a spontaneous way;
language. The experimenter
and the ability to understand the speech of others. they develop in parallel. There is certainly a great
reads out a sentence to be
The former is relatively easy to study: all we have deal of evidence to show that children produce a repeated. If the child makes
to do is turn a tape recorder on, and analyse what word or construction without having a full under- any changes, these can indi-
standing of Doggie, says one young child, point- cate aspects of the language
comes out. Research into speech comprehension it.
which are still being learned
is far more difficult because we need to take into ing to a cat. He got hat on, says another, and then
or not yet acquired, One 25-
account not only what is spoken to the child, but later says Take that hat on off- as if hat on were year-old child, 'Echo', gave
the situation in which it is uttered, and the child's a noun. This kind of thing happens frequently from the following imitations:
prior knowledge of the world. In one study, a 2- around age 2 - and, indeed, it could be argued 1 The owl eats candy and
year-old child was observed to respond correctly that our readiness to use linguistic forms we do the owl runs fast.

not fully understand stays with us throughout life! Echo: Owl eat candy and he
when his mother said, at bedtime, 'Go and get your
run fast.
pyjamas out of the drawer in your bedroom.' But It has also been recognized that imitation is a
2. The owl who eats candy
it is not at all clear, without a careful investigation, distinct skill in language acquisition - many chil- runs fast.
which parts of this sentence the child had under- dren spend a great deal of time imitating what their Echo: Owl eat a candy and
stood — it might simply be that the word pyjamas, parents have just said. This is most noticeable when he run fast.

said at bedtime, and coupled with the knowledge new sounds or vocabulary are being learned, but The first imitation suggests
thatEcho understands the
of where pyjamas are kept, was enough to produce it has been shown that imitation may be important
meaning and structure of the
the appropriate action. in the development of grammar too. Often, chil- coordinate sentence (p. 95).
What is the relationship between production and dren imitate sentence patterns that they are unable She uses the same strategy
comprehension when it comes to language learn- to produce spontaneously, and then stop imitating in the second case, which

suggests that she cannot yet

ing? There are three possibilities. The
commonsense view is that comprehension always
precedes production: children need to understand
a word or grammatical construction before they
traditional, these structures when they start to use them in their
cope with the more difficult
speech - suggesting that imitation is a kind of sentence containing a subor-
'bridge' between comprehension and spontaneous dinate clause introduced by
production. who, though she does follow
its meaning. (After D. Slo- I.

bin& C.A.Welsh, 1967.)

Vocabulary size
The average vocabulary size
PLOTTING THE COURSE OF one child may use cold to mean only 'cold weather',
samples of children
of ten
LANGUAGE DEVELOPMENT whereas the other may use it to apply to water, between the ages of 1 and 6.
A popular metaphor in child development is to talk food, and grim facial expressions. In these circum- (After M. E. Smith, 1 926.) To
of 'milestones' - the age at which a child takes stances, a single score, based on one developmental interpret such totals, a great
deal needs to be known
a significant step forward in behaviour (such as parameter, conceals more than it illuminates: it
about the method for defin-
sitting, crawling, standing). The metaphor does not needs to be supplemented by a wider and more ing 'words' used by the in-
work so well when it comes to language: too much detailed series of measures that take into account vestigator (p. 104). Were go,
happens too quickly. There is simultaneous de- the qualitative range of linguistic features used by goes, going, etc. counted as
velopment of sounds, grammar, meaning, and the child. one word or several? Were
words of radically different
interaction skills; and significant progress can be After several years of acquisition research, in
meaning (e.g. bear 'animal'/
made on several different fronts in a matter of days. which many measures have been investigated, it 'carry')counted separately?
It is thus no easy matter to quantify the amount is possible to isolate certain broad trends with some Decisions of this kind have a
of language learned by a child within a particular confidence, and these are the subject matter of major influence on the totals
arrived at in a word count.
period (as we need to do in deciding what counts §§39—43. It appears that most children do follow
as 'normal' development, and in plotting depar- the same general path as they acquire sounds and i i i i i i i i i i i. i 2,600
tures from this norm (p. 279)). grammatical structures, and several common 2,400
Several attempts have been made to find impor- *
trends are evident in the learning of vocabulary 2,200
o
tant single measures of development, within parti-
cular linguistic levels (§13) - notably the notions
of sentence length and vocabulary size, both of
and pragmatics (§21) also. However, there seems
to be considerable variation in rate of development,
and there are many individual differences in the
/; 2,000 5
1,800

1,600 ?
|

which steadily increase as children grow older. order of acquisition of specific features that have
1,400 J"
Such indices can provide general indications of pro- to be taken into account (p. 231). The study of these
gress, but they have serious limitations. Two sen- variations is a major emphasis of current child lan- 1,200
tences may consist of exactly the same number of guage research. 1,000 |
words, morphemes, or syllables, and yet be very 800
J
different in terms of their syntactic complexity: /
600 a
see a cat and a dog and a cow is much simpler
400
than J see a cat that is next to a dog, though both |
are the same length. Similarly, two children may
both have vocabularies of 100 words, yet differ
i£ I I I I I I I I
200 >

— <N rf. -rf iy~ vC

in therange of words used and in their meanings: Age in years

232 -VII CHILD LANGUAGE ACQUISITION

MLU
Measuring the mean length of utterance (MLU) has
Language tests
been one of the most widely practised indices of A language test takes a se- cess of intervention. But lations in the U.S. and is
grammatical development in young children. The total lection of linguistic features they are of limited value in designed to identify Spa-
number of utterances in a sample is divided by the total - for example, a set of con- fundamental research in nish-speaking children who
number of words (in some procedures) or morphemes sonants, grammatical con- child language, because do not demonstrate native
(in others) (p. 90). structions, or items of voca- they deal with only a frac- syntactic proficiency com-
The best-known measure, which uses morphemes, bulary - and establishes tion of the linguistic fea- mensurate with their age.
was devised by Roger Brown (p. 229) in the 1 960s. whether a child has know- tures being acquired. By In the comprehension part
The diagram shows the way Brown's three subjects ledge of them, in either pro- their nature, testshave to of the test, the examiner
gradually increased their utterance length. Five stages duction or comprehension. be short and highly selec- reads a sentence to the
of development are recognized, based on a division The child's responses are tive. They can give useful child, who has to point to
of the lengthcontinuum into intervals of 0.5 scored, according to some background information the appropriate picture.
morphemes. There is a good correlation between MLU predetermined criterion. As about a child's general le- The illustration shows four
and age, but the relationship between MLU and the language ability pro- vel, but they cannot provide pictures used to test know-
range of constructions found in a sample is less clear. gresses, higher scores are the detailed account of the ledge of a contrast involv-
Predicting the grammatical complexity of a speech obtained. emergence of linguistic ing two prepositions: El
sample from length alone is by no means Tests are widely used in skills that acquisition re- perro esta\ detr&s de la silla
straightforward, especially as length increases. the field of language handi- search requires. and El perro esta debajo de
(R.Brown, 1973.) cap (§46), because they A task from a language la silla (The dog is behind/
provide a convenient test is illustrated below. underneath the chair'). The
/
means of identifying chil- The Screening Test of Spa- remaining two pictures are
;
00 -
4.00 dren who require special nish Grammar is based on 'decoys'. (From A. S. Tor-
: /
teaching or therapy, and a the language of the Mexi- onto, 1973.)
1 way of monitoring the suc- can and Puerto Rican popu-
2-3
o J 50
/
V
: /
E
/•ve Adam 1 ,5arah

00 - 1*
/

/
/
so ;

: / A/Ry
1 *J
2 2 00 v ;
: / :\ -

/ •/
- /

en \! 1 I I 1 1
i i i i i i i i

16 18 20 22 24 26 28 30 32 34 36 38 40 42 44
Age in months

— >o Minor Responses 3O Vocatives Other Problems

Profiles V —
Comm. Quest. Statement
An extract from a profile of grammatical development
55o Major
•V q- v N' g
used the study of language handicap and based on
in fc Other Problems

a synthesis of findings from the study of normal Conn. Clause Phrase Word
language acquisition. The abbreviations in this
So \X QX sv 3 \X ~l DN 1/ w 1

procedure, known as LARSP (Language Assessment,

z so vo 4- AdjN 3 V part 1 -ing
Remediation and Screening Procedure) refer to
different grammatical constructions, e.g. sc / vc NN Int X
Pi

SVO = Subject + Verb + Object', PrN = Neg* Other PrN fe Other 1 X

-ed
'Preposition + Noun' (§16). X + SNP x ~ V:VP X + CNP X + ONP Z X + AAP '

The totals refer to the number of instances of a = * VXY QXY SVC VCA DAdjN Cop
-en

category used by a child in a sample of spontaneous

speech.
H lei XY
vsw
SVO
SVA
I VOA
VO dO,
Adj Adj

Pr DN
N
Am" 3s

Profiles permit a more detailed impression of the

doXY NegA-y Other Pron^ * Other gen
range of structures used than can be obtained from
a test, and enable the analyst to plot emerging XY + SNP XY f VVP XY + CNP XY * ONP XY + AAP
strengths and weaknesses in several areas of + S QVS SVOA AAXY NP Pr NP Neg V n't

grammar simultaneously. In the present case, the child QXY + SVCA Other Pr D Adj N Neg*
cop
has begun to use constructions at Stage II (typically
XXY + VS,Jf + ) svo„o, cX 2 Aux
age 1 8 months-2 years), but there are several gaps,
tag SVOC XcX Other
and he has not yet made much progress in Stage III. 'aux

As the child in the sample had

reached the in fact and Coord Coord. Coord I 1 + Postmod 1 1 +
chronological age of 2 years 3 months, he would seem

SCO
clause -est
c Other Other Subord. A 1 1 +
to be a somewhat slow developer, as far as the -er
s
acquisition of grammar is concerned. Postmod. +
Comparative phrase -K
(D. Crystal era/., 1976.) Other

38 INVESTIGATING CHILDREN'S LANGUAGE 233


Theories of language
acquisition
IMITATION
Language acquisition has long been thought of as
a process of imitation and reinforcement. Children
learn to speak, in the popular view, by copying
the utterances heard around them, and by having
their responses strengthened by the repetitions, cor-
rections, and other reactions that adults provide.
In recent years, it has become clear that this princi-
ple will not explain all the facts of language de-
velopment. Children do imitate a great deal,
especially in learning sounds and vocabulary'; but
little of their grammatical ability can be explained
in thisway. Two kinds of evidence are commonly
used in support of this criticism — one based on
the kind of language children produce, the other
on what they do not produce.
The first piece of evidence derives from the way
children handle irregular grammatical patterns.
When they encounter such irregular past-tense
forms (p. 90) as went and took, or such plural
forms as mice and sheep, there is a stage when
they replace these by forms based on the regular
patterns of the language. They say such things as
wented, taked, mices, mouses, and sheeps. Evi-
dently, children assume that grammatical usage is
regular, and try to work out for themselves what
the forms 'ought' to be — a reasoning process
known as analogy (p. 330). They could not have
learned these forms by a process of imitation.
Adults do not go around saying such things as
wented and sheepsl
The other kind of evidence is based on the way
children seem unable to imitate adult grammatical
constructions exactly, even when invited to do so
('elicited imitation', p. 232). The best-known
demonstration of this principle in action is the dia-
logue reported by the American psycholinguist,
David McNeill (1933- ), where a child proved
unable to use a pattern, even though the parent
presented the correct adult model several times:
child: Nobody don't like me.
mother: No, say 'Nobody likes me.'
child: Nobody don't like me.
(Eight repetitions of this dialogue.)
mother: No, now listen carefully: say 'Nobody
likes me.'
child: Oh! Nobody don't likes me.
The child, at this point in its learning of grammar,
was clearly not ready to use the 'single negative'
pattern found in this dialect of English. Such exam-
ples suggest that language acquisition is more a
matter of maturation than of imitation.
LNNATENESS
The limitations of an imitation/reinforcement view
of acquisition led in the 1960s to an alternative
proposal, arising out of the generative account of
234 VII CHILD LANGUAGE ACQUISITION
language (§65). It was argued that children must
be born with an innate capacity for language de-
velopment: the human brain is 'ready' for lan-
guage, in the sense that when children are exposed
to speech, certain general principles for discovering
or structuring language automatically begin to
operate. These principles constitute a child's 'lan-
guage acquisition device' (LAD).
The child uses its LAD to make sense of the utter-
ances heard around it, deriving from this 'primary
linguistic data' hypotheses about the grammar of
the language — what the sentences are, and how
they are constructed. This knowledge is then used
to produce sentences that, after a process of trial
and error, correspond to those in adult speech: the
child has learned a set of generalizations, or rules,
governing the way in which sentences are formed.
This sequence of events can be summarized in the
following way:
INPUT LAD
Primary General
Grammatical
linguistic
data (adult
—* language
— knowledge
learning
speech) principles
(rules)
There have been many differences of opinion Jean Piaget (1896-1980)
over how best to characterize LAD. Some have
argued that LAD provides children with a know-
ledge of linguistic universals (§14), such as the exis-
tence of word order and word classes; others, that
it provides only general procedures for discovering
how language is to be learned. But all of its sup-
porters are agreed that some such notion is needed
in order to explain the remarkable speed with
which children learn to speak, and the considerable
similarity in the way grammatical patterns are
acquired across different children and languages.
Adult speech, it is felt, cannot of itself provide a
means of enabling children to work out the regular-
ities of language for themselves, because it is too
complex and disorganized (p. 52). However, it has
proved difficult to formulate the detailed properties
of LAD in an uncontroversial manner, in the light
of the changes in generative linguistic theory that
have taken place in recent years; and meanwhile,
alternative accounts of the acquisition process have
evolved.
COGNITION
The main alternative account argues that language
acquisition must be viewed within the context of
a child's intellectual development. Linguistic struc-
tures will emerge only an already-estab-
if there is
—
lished cognitive foundation for example, before
children can use structures of comparison (e.g. This
car bigger than that), they need first to have de-
is
veloped the conceptual ability to make relative
judgments of size. Several early child language
scholars maintained such a relationship exists, but
the most influential account stems from the model
of cognitive development proposed by the Genevan particular use of a y/ie ending, and similar forms Motherese - or
psychologist Jean Piaget (1896-1980). have been noted in several other languages, such otherese?
Several controlled studies have been carried out as Japanese -ko, Gilyak -k/-q, Berber -f/-ftt. Occa-
The term motherese' seems
investigating the link between the stages of cogni- sionally, totally different words will be used, e.g. a natural one, given the im-
tivedevelopment proposed by Piaget and the emer- bunny for 'rabbit'.There may be special use of portant role of mothers in
gence of linguistic skills. The links have been most individual sounds, such as the use of rounded lips early child development.

clearlyshown for the earliest period of language However, it would be more

in English, or special palatal sounds in Latvian and
accurate to refer to 'paren-
learning (up to 18 months), relating to the develop- Marathi. tese',as fathers are also
ment of what Piaget called 'sensori-motor' intelli- Some of these features also seem to function as able to adapt their speech
gence, in which children construct a mental picture ways of holding the child's attention, or of identify- when talking to children, and

of a world of objects that have independent exis- ing particular words and sounds. This may well use very similar strategies.
Motherese and 'fatherese'
tence. For example, during the later part of this be the reason for the very common use of high, are not identical, however.
period, children develop a sense of object perma- wide pitch range in maternal speech. Mothers also Fathers tend to be more in-
nence - they will begin to search for objects that devote a great deal of time to obtaining feedback tense and demanding in their
they have seen hidden — and some scholars have from their children, especially in the first three communication, using more
direct questions and a wider
argued that the ability to name classes of objects years. Their speech contains a very high frequency
range of vocabulary.
(i.e. to give them a comparably 'permanent' linguis- of question forms, and many utterances have a high But even parentese' is too
tic status) depends on the prior development of rising intonation {yes?, all right?). specific a notion. Some of
this cognitive ability. However, it is difficult to the characteristics of mother-
These modifications are evidently important ways ese can be found in other
show precise correlations between specific cogni-
of establishing and maintaining meaningful com- adults too - and even in 4-
tive behaviours and linguistic features at this early
munication with the child, as they can be found year-olds, when they talk
age. The issue is a highly controversial one, which 'down' to younger children.
in the earliest mother-child interactions (§39). It
increases in complexity as children become linguis- Moreover, in some non-
has even been suggested that these features are
tically - and cognitively - more advanced. Western cultures (e.g. West-
universal, but this claim is premature in the absence ern Samoa), the primary care-
of empirical studies, and there is already some givers may not be the par-

INPUT counter-evidence from other cultures — several of ents at all, and the develop-
ing child may receive most of
For many years, in the wake of the innateness these features are lacking in Samoan and Quiche its linguistic stimulation from
hypothesis, the importance of the language used Mayan, for instance. However, the highly struc- siblings, other adult relatives,

by adults (especially mothers) to children was mini- tured character of maternal input is not in doubt, or neighbouring families. A
and its possible influence on the course of language more neutral term, such as
mized. But studies of 'motherese', as it came to
acquisition is now taken very seriously. baby talk', is thus preferred
be called in the 1970s, showed that maternal input by some researchers -
is by no means as complex and fragmentary as pro-
Unfortunately, it is difficult to show correlations though its ambiguity (speech
ponents of innateness theory claimed it to be. Many between the features of motherese and the subse- by children or to children?)

parents do not talk to their children in the same

quent emergence of these features in child speech, limits its usefulness. 'Care-
and even more problematic to move from talk taker speech' is also widely
way as they talk to other adults. Rather, they seem used.
capable of adapting their language to give the child about correlations to talk about causes. Some stu-
Very little study of the na-
dies, searching for such relationships, have found
maximum opportunity to interact and learn. ture of cultural differences

Several of these adaptations have been noted (after very few; others have found occasional correlations has taken place. In the
between specific structures, though often with an Samoan case, for example,
C. A. Ferguson, 1977).
appreciable gap between the use of a feature by many of the features of
Anglo-American motherese
• The utterances are considerably simplified, the mother and its subsequent use by the child;
(such as the use of simplified
especially with respect to their grammar and mean- yet others argue that input structures are very structure, expansions, and
ing. Sentences are shorter: one study showed that closely tailored to the needs of the child (the 'fine diminutives) were found to
the average length of maternal sentences to 2-year- tuning' hypothesis). The use of different research be absent. The turn-taking
pattern was also different,
olds was less than four words - half that found methodologies clouds the picture, but it is now
often taking the form: child
when the mothers talked to other adults. There plain that the nature and frequency of linguistic talks to mother —> mother
is a more restricted range of sentence patterns, and features in maternal input can no longer be neg- talks to older sibling — > sib-
a frequent use of sentence 'frames', such as Where's lected in devising theories of language acquisition. lingattends to child. Such
— ? or That's a —
The meanings are predomi-
.
differences have important
implications for the develop-
nantly 'concrete', relating to the situation in which CONCLUSIONS ment of any theory of lan-
mother and child are acting. It isnot possible, in the present state of knowledge, guage acquisition in which
• There are several features whose purpose seems to choose between these various approaches. The motherese plays a part.
(After E.Ochs, 1982.)
to be clarification. Extra information is provided number of definite, general facts known about
that would be considered unnecessary when talking language acquisition is still very small. In particu-
expanded and para-
to other adults. Sentences are lar, much more information is needed about the
phrased and may be repeated several times. The way children learn languages other than English.
speed of speaking is much slower than that used Doubtless imitative skills, a general language-
to other adults. learning-mechanism, cognitive awareness, and
• There is also an expressive, or affective, element structured input all play their part in guiding the
in motherese, shown by the use of special words course of language acquisition. Unravelling the
or sounds. Diminutive or reduplicative words (e.g. interdependence of these factors constitutes the
doggie, choo-choo) are common. English makes main goal of future child language research.

38 INVESTIGATING CHILDREN'S LANGUAGE 235

 like sound made towards the back of the mouth. Good talkers?
39 The first year Many have a nasal quality.
Later in this period, cooing sounds are strung
Frequent and varied vocal
play and babbling is some-
together - often 10 or more at a time. These strings times thought to be the sign
of a good talker, or an indica-
are not pronounced in a rhythmical way; there are
tion of superior intelligence,
no clear intonational contours. However, some of memory, or personality.
For many parents, a child's first words, uttered at the sequences (such as [ga] and [gu]) do begin to There is no evidence that
around 1 year of age, mark the first real evidence resemble the syllables of later speech. Then, at any such correlations exist.
There have been many quiet
of language development - the child has 'started around 4 months, the first throaty chuckles and
babies whose subsequent
to talk'. But this is to ignore a great deal of early laughs emerge.
language development has
progress during the first year, without which no During the cooing stage, babies seem to be per- been rapid; and, conversely,
first word would emerge at all. This progress has forming the first gross activities required for the there are certain areas of

to be made in three main areas: sound production, production of speech. The tongue begins to move mental subnormality (such
and horizontally, and the vocal folds as Down's syndrome) in
speech perception, and speech interaction. vertically
which a good range of bab-
begin to be used in coordination with it. There is
bling can be heard. No-one
a great deal of lip movement and tongue thrusting, has been able to discover a
which it is thought may be a form of imitation. direct link between early vo-
Sound production calization and later intellec-
tual or linguistic development
Between birth and 12 months, a vast change takes Stage HI (20-30 weeks): Vocal play
(§7).
place in a baby's sound-producing abilities, and The sounds of vocal play are much steadier and
several stages of development have been proposed. longer than those of cooing. Most segments are
over 1 second, and consist of consonant + vowel-
Stage I (0-8 weeks): Basic biological noises like sequences that are frequently repeated. They
Over the first few weeks of life, a baby's vocal are usually at a high pitch level, and involve wide
sounds directly reflect its biological state and activi- glides from high to low. A considerable range of
ties. States of hunger, pain, or discomfort that cause consonant and vowel qualities is apparent, includ- Cry analyser
crying and fussing are known as reflexive noises. ing nasal and fricative sounds made in various parts This instrument was deve-
Breathing, eating, excreting, and other bodily ac- of the mouth (§27). There are many individual dif- loped in Sweden in the
1 970s as a means of auto-
tions concerned with survival cause a wide range ferences in the order of emergence of these sounds,
matically analysing infant
of vegetative noises, such as sucking, swallowing, and several changes in the focus of the activity dur-
cries. It continuously moni-
coughing, and burping. Infant reflexive cries have ing the period - on some days uvular sounds may and
tors cry activity, pitch,
been studied in detail. The normal 'basic' cry con- be the dominant sounds heard; on other days it duration,and the accompa-
sists of a series of 1 -second pulses separated by may be labial sounds. In due course, the sounds nying heart and respiration
rates, printing out numerical
brief pauses. The vocal folds (§22) vibrate strongly, combine into longer sequences, to produce the first
data on a paper strip. It has
and the pitch of the voice falls sharply with each babbled utterances. mainly been used in pediatric
pulse. The quality of the sound is similar to that There seems to be a strong element of practice wards for babies with post-
of an [a] vowel. in the activities of this period, but anyone who has natal complications.

It is not easy to attribute clearly different func-

tions to cries at this age. Hunger and pain cries

tend to merge into a single distress cry, though pain
cries are often much tenser and have a different
rhythm. Discomfort cries are usually much shorter
(5 sec) and occur in brief sequences. Vegetative
noises are even shorter (4 sec) and contain more
consonant-like sounds.
There is nothing language-specific about these
early sounds. However, they do have some features
in common with later speech. An air-stream mech-
anism (§22) is being used to produce noise; there
is rhythmical vocalization; the vocal folds are being

used to produce pitch patterns: all of these are fun-

damental characteristics of later speech.

Stage II (8—20 weeks): Cooing and laughing

Between 6 and 8 weeks, the first cooing sounds
are produced, generally when the baby is in a set-
tled state.These sounds develop alongside crying,
gradually becoming more frequent and more var-
ied, as the child responds to the mother's smiles
and speech. They are quieter, lower pitched, and
more musical than crying, usually consisting of a
short, vowel-like sound preceded by a consonant-

236 •
VII CHILD LANGUAGE ACQUISITION
observed it will recognize that it also provides a Babbling in
great deal of enjoyment for parent and child alike. Early vocalization language
different
Which consonant sounds (§27) do infants use most environments
Stage IV (25-50 weeks): Babbling often in their early vocalizations? In the first six months,
There a considerable simi-
Babbling is much less varied than the sounds of back (velar) consonants predominate. Then, between is

larity in infant babbling pat-

vocal play, in the early part of this period. A smaller sixand nine months, alveolar sounds become
terns, whatever their lan-
dominant. Labial sounds are never the most frequent.
set of sounds is used with greater frequency and guage environment. The
This pattern stabilizes at around nine months. Children
stability, to produce the [bababa] and other table below shows the occur-
seem to have the physical capability to produce the
sequences known as reduplicated babbling (be- rence of English or English-
words of a language long before these sounds are
like consonant segments in
cause of the repeated use of the same consonant used in speech. (After B. L Smith & D. K. Oiler, 1 981 .)
the babbling of infants reared
sound). About half-way through the period, this in 15 language environ-
develops into variegated babbling, in which conso- ments. The consonants have
nants and vowels change from one syllable to the been divided into two groups
of 1 2: consonants which are
next (e.g. [adu]). The rhythm of the utterance and
* Labial heard frequently in English
the syllable length at this point are much closer babbling (accounting in fact
to that found in speech. Babbled utterances seem for about 95% of all conso-

to have no meaning, though some may resemble nants heard); and those
the words of later speech. which are heard infrequently.
It can be seen that there is
It used to be thought that there was no link
a very close correspondence
between babbling and spoken language. The child between the different lan-
was imagined to be trying out every possible sound guages, though there is no
in a random manner, and that babbling would stop identity. Only [m] and [b] turn
- 40 -
up in all language environ-
before speech began. Recent studies have shown
ments.
that this view does not hold. In many cases, bab- The similarities are im-
bling continues long after speech begins - some- pressive. However, it should
times as late as 18 months. Nor are the sounds be noted that the numbers of

of babbling a random selection: most of babbling children involved in these

studies are very small - often
consists of a small set of sounds very similar to
only one child, and rarely
those used in the early language to be spoken by 0-3 3-6 6-9 9-12 12-15 more than five. The compar-
the child. The brain seems to be controlling the Age (in months) ative picture may become
development of babbling and early speech in a simi- more complex as the data-
base increases.
lar way, so that a set of well-practised sounds is
(After J. L Locke, 1983.)
available for use at the time when children become
intellectually capable of using
munication of meaning.
sound for the com-
Environment
Age in
months hdbmtgswnkjp
Frequent English consonants

Afrikaans 11-12
Stage V (9-18 months): Melodic utterance Mayan 9
Variations in melody, rhythm, and tone of voice Luo 12
Thai 10-11
(§29) become a major feature of child utterance Japanese 9-12
towards the end of the first year. Parents begin to Hindi 9-10
sense intentions behind these utterances, with their Chinese 8-11
more well-defined shape, and often attribute mean- Slovenian 11
Dutch 11
ings to them, such as questioning, calling, greeting,
Spanish 9
or wanting. Games and rituals may develop their German 10-12
own melodic contours. Individual syllables come Arabic 6-10
to be used with a fixed melody, producing 'proto- Norwegian 0-12
words', where the sounds are clear, but it is not Latvian 6-12
English 1-15
possible to be sure what they mean. These are the
first real signs of language development, and chil- Age in Infrequent English consonants
dren growing up in different language environ- months vl ezf/6rj3rt/d3
ments begin to sound increasingly unlike each Afrikaans 11-12
other. Mayan 9
Luo 12
Thai 10-11
Japanese 9-12
Hindi 9-10
Chinese 8-11
Slovenian 11
Dutch 11
Spanish 9
German 10-12
Arabic 6-10
Norwegian 0-12
Latvian 6-12
English 1-15

39 THE FIRST YEAR 237

 hypothesized that children's perceptual apparatus Habituation
Speech perception is in some way 'programmed' to discriminate
experiments
Very young babies present an extraordinary range speech sounds - that they are born with special
A widely used technique in
of auditory abilities. There have been several 'feature detectors' that respond to the acoustic pro- infant speech perception is
experiments in which different sounds are played perties of speech. A great deal of research has since to play sounds to the baby
to babies, and their responses monitored. For focussed on this issue, in an attempt to determine through a pair of head-
whether the children are displaying a general audi- phones or a loudspeaker,
example, day-old babies have been played their
and then to monitor the
mother's voice speaking normally, the same voice tory ability (which might be shared by certain other
baby's responses, such as
speaking abnormally (in a monotone), and a species — experiments on both chinchillas and rhe- the speed of its heart-beat. In
stranger's voice: only the first caused them to sus monkeys have shown comparable responses), thisapproach, a sound is
attend. Other studies have shown how babies turn or whether it is a specific ability tuned to phonetic played several times to the
child,and the heart-rate is
their heads towards the source of a sound within distinctive features. The amazingly early age at
monitored. There is an initial
the first few days of life, and prefer human voices which infants begin to make auditory discrimina- 'orienting response' and a
to non-human sounds as early as 2 weeks. Abilities tions is accepted, but the critical issue - how
now period of heart-rate deceler-
of this kind are so apparent that some researchers these basic perceptual capacities come to be affec- ation, as the child gets used
to the stimulus. This effect is
have concluded that auditory training must begin ted by the infants' emerging experience of language
— is not yet resolved. known as habituation'. A
within the womb. second sound then played.
is
The question of when the babies learn to dis- Ifthe child notices the differ-
tinguish the sounds of speech is controversial. An Speech comprehension ence, there will be a new or-
ienting response.
auditory ability to discriminate certain pairs of con- Between 2 and 4 months, babies begin to respond
There are many methodo-
sonants or vowels (e.g. [pa] vs [ba]) is present from to the meaning of different tones of voice, such
logical problems in this kind
around 4 weeks, and this ability to discriminate as angry, soothing, or playful voices. From around of research, especially with
becomes increasingly sophisticated in subsequent 6 months, different utterances begin to be related very young babies. More-
months. An early finding was that infants seem able to their situations, e.g. Bye-bye, Clap hands, or over, there is still some way
to go before the findings
to perceive these distinctions in the same way as pointing in response to questions. Some individual
about the perception of iso-
adults. Adults make a sharp, categorical distinction words may be recognized, such as names of family lated syllables can be related
between such sounds as [pa] and [ba] in perception members, or basic responses (e.g. No). Most chil- to the perception of these
experiments (p. 147). When 1-month-old infants dren understand several words by the end of the syllables in connected
were presented with sets of sounds that also varied first year. In one study of eight children, six showed
speech. But the habituation
paradigm has shown how it
only in the degree of consonant voicing, they too clear evidence of understanding up to 20 words
is possible to make a begin-
made categorical distinctions (P. D. Eimas et al., by the end of the first year, and one child under- ning in answering the difficult
1971). By careful monitoring of the babies' res- stood as many as 60 (H. Benedict, 1979). In all question: when are linguistic
ponses (see below), it was shown that presentations comprehension ability was at least a
cases, this contrasts first perceived?
of two kinds of [pa] or two kinds of [ba] caused month ahead of the appearance of the children's
no reaction, whereas [pa] vs [ba] did. first words.
On the basis of such findings, the investigators
75 r

A -month-old infant sucking on a special nipple while listening to a recording of different

1

syllables.The child's normal sucking rate is first established, and then sequences of sounds
are presented. As the child hears the first sound, the rate of sucking increases. During subse- 60
quent repetitions of the same sound, the sucking rate shows a gradual decrease. A new
sound is then played to the child. If no distinction is perceived, the sucking rate will continue
to decrease; but if a change is perceived, it will show a sudden increase. £45
V-,

.3 30 v
15

J I I I L I I I
I I I I 1 1 ''''
X 5432 101 234 X 5432 101 234
Time (min)
What happens when the difference between a pair of
auditory stimuli is (a) ignored and (b) perceived by a
group The line at represents the point
of children.
when the sound stimulus is changed. The different
presentations are shown at 1 -minute intervals. X
shows the normal sucking rate before the stimuli are
presented. In the first case, two kinds of [ba] or two
kinds of [pa] produce hardly any change in sucking
rate. In the second case, the change from [pa] to [ba]
(or vice versa) is clearly shown. (After P. D. Eimas
era/., 1971.)

238 •
VII CHILD LANGUAGE ACQUISITION
 .

Speech interaction mother's voice becomes more varied in response.

As the child starts to take interest in the environ-
From the moment
a baby is born, a mother holds ment and looks around, the mother speaks more
it and talks to it - despite the fact
in front of her, loudly, drawing attention to different objects. Her
that she knows it does not yet have any language! intonation becomes more exaggerated, and she
Mothers seem to have an instinct to promote com- often repeats her sentences. Simple face-to-face
munication as soon as possible, using the child's games are played (such as peekaboo), promoting
earliest biological noises (p. 236) as stimuli. Cries, a great deal of communication.
burps, sneezes, and other vocalizations are seized After 6 months, the baby's more purposeful
upon and interpreted, as the extract (see box, right) movements and explorations produce more
shows. The mother is very ready to ascribe inten- extended commentaries by the mother. She no
tions to the baby's utterances and to build them longer responds to every vocalization that is pro-
into a conversation — something she does not do duced, but focusses special attention on those that
with its non-vocal activities, such as head move- are more structured in character - in particular,
ments or arm waving. The conversational pressure the first babbled utterances. Between 8 and 10
can be quite intense: in one study, over 100 ques- months, babies attempt to attract the attention of
tions, comments, and other utterances were used others by pointing. They begin to 'follow' adult
by a mother while attempting to elicit a burp from conversations, looking first at one person, then at
a 3-month-old: Where is it?, Come on, come on, the other. By the time their first words appear,
come on, You haven't got any, I don't believe you, babies have learned a great deal, both from obser-
etc. (C. Snow, 1977). vation and from practice, about what a conversa-
It would seem that the foundations of conversa- tion is and how to participate within it.
tion are being laid in these early interactions. The
mother's behaviour not random. She uses a large
is

number of questions, followed by pauses, as if to One-sided conversations

show the baby that a response is expected and to
Michael (3 months): (Loud crying.)
provide an opportunity for it to respond. She con-
Mother. (Enters room) Oh my word, what a noise! What
tinually greets the baby, even after very short per-
a noise! (Picks up baby.)
iods of separation. Moreover, she talks to the child
Michael: (sobs.)
at length only when the child is (in principle) in
a position to reply. While the baby Mother. Oh dear, dear, dear. Didn't anybody come to
is feeding, for
see you? Let's have a look at you. (Looks inside
example, mothers tend to remain silent, taking up nappy.) No, you're all right there, aren't you.
the conversation only when the baby ceases to suck
Michael: (Spluttering noise.)
or needs to be winded. This cyclical pattern of
Mother. Well, what is it, then? Are you hungry, is that
speech and silence anticipates the fundamental it a long time since dinner-time?
it? Is
structure of older conversations.
Michael: (Gurgles.)
There are many changes in conversational style
Mother. (Nuzzles baby.) Oh yes it is, a long long time.
during the first year. At around 5 weeks, the
Michael: (Cooing noise.)
exchanges become more emotive, as smiling de-
velops. The mother's utterances change as the
Mother. Yes, I know. Let's go and get some lovely

baby's vocalizations grow. At around 2 months,

grub, then . .

A conversational
(D. Crystal, 1986, p. 51.)
the emergence of cooing elicits a softer voice. Some cycle at age
time later, the baby begins to laugh, and the 6 weeks
1 . Greeting smiles. Mother
says 'Come on then.'

2,3. 'Oh aaa. Come on.'

Baby coos.
4,5,6. Baby coos Aghoo.'
Mother says, 'Come on. Hoo
hoo.'

7, 8, 9.As before, then with

baby's smile, That's right.
Woo hoc'
(After C. Trevarthen, 1 983.)

39 THE FIRST YEAR- 239

40 Phonological development

By the time children are a year old, they have attempt to use them. These trends include: The 'tis'
learned a great deal about the way adults use • Fricative consonants (p. 157) tend to be re- phenomenon
sounds to express differences in meaning (§28),
placed by stops, e.g. see is pronounced [ti:]. Several studies have re-
but their own ability to produce these sounds lags • Velar consonants (p. 155) tend to be replaced ported intriguing conversa-
some way behind. Some 1 -year-olds can recognize by alveolar consonants, e.g. gone is pronounced
tionsbetween a young child
several dozen words, involving a wide range of and an adult, showing that
[don]. there may be a big difference
vowels and consonants, but their own ability to
• Consonant clusters are avoided, e.g. sky is pro- between what children hear
pronounce these words may be restricted to just nounced [kai]. and what they can say. The
two or three consonants and a single vowel. One • Consonants at the ends of words are often phenomenon was first re-
child at 13 months could use only [b], [d], and ported in the following way:
omitted, e.g. hat is pronounced [ha].
[a], but he used these sounds to express a variety One of us, for instance,
• Unstressed syllables are often dropped, e.g.
of words — for example, [ba] was used for baby, spoke to a child who called
banana becomes [nana].
bath, cup, and Peter. By 15 months, he had added his inflated plastic fish a fis.
• As words become longer, sounds in one part the child's pro-
In imitation of
[m], [p], and [u] to his repertoire, and was thus
of a word canalter the pronunciation of sounds nunciation, the observer
able to distinguish a much larger number of words. in other parts.This tendency for sounds to 'harmo- said: This is your fis?' No,'
He also began to use some of these consonants
nize' (p. 161) is found with both consonants and
said the child, 'my fis.' He
at the ends of words as well as at the beginning; continued to reject the
vowels. Consonant harmony is found in such pro- adult's imitation until he was
for example, was used for a nasty smell,
[pu]
nunciations of dog as [gog] or [dod], with identical told, That is your fish.' 'Yes,'
and [ap] was used for up. By age 2, he was using he said, my fis.'
(or near-identical) consonants. Vowel harmony (J. Berko &
over a dozen consonants and vowels, and was able R.Brown, 1960, 531.)
would be heard if window were pronounced for p.

to pronounce over 200 words in an intelligible

example as [wouwou] or [wada]. The effect has been referred
(though often immature) manner. • There is a preference for [w] and.[j] sounds toas the 'fis phenomenon'
It is not possible, at present, to make precise pre- ever since. Such reports indi-
to be used instead of / and r, e.g. leg as [jeg].
dictions as to the order in which children come cate that childrenknow far
more about adult phonology
to use new sounds. Some children have 'favourite'
than their own pronunciation
sounds, which they introduce into many words, suggests.
whether the sound is in the adult version or not;
others 'avoid' sounds — for example, persistently
dropping certain consonants at the ends of words.
There may also be a great deal of variation in the
way target sounds are produced — one child pro- The emergence of consonants
nounced blanket as [bwati], [bati], [baki], and Most English consonants ing photographs of familiar being produced correctly
[batit], within a few hours of each other. Another begin to be acquired be- objects. The periods shown by the majority of the chil-
tween 2 and 4 years of age. are averages, and the up- dren at age 2; and others
produced ten different forms of pen within a single
This diagram shows the per age-limit is based on a were still not being said
half-hour! order of emergence as correct pronunciation by correctly at age 4.
Nonetheless, as a result of several studies involv- found in a study which eli- 90% of the children. The (After E. M. Prather era/.,
numbers of children, certain general trends
ing large cited pronunciations of diagram also shows that 1975.)
words from 20 children, us- some sounds were already
can be shown. For example, consonant sounds are
more likely to be first used correctly at the begin- Over 4
nings of words; final consonants emerge later 1 r \ tj *3 V z 3 8 e
(though there are exceptions, such as the early use 4;0
of final [f] and [s] in English). A 1971 survey of s

100 English children showed that, during the 3;8

second year, [p], [b], [k], [n], [f], [d], [g], [m], and
[h] were commonly used word-initially; but only 3;4
I w
the first five of these sounds were developing word-
finally (D. Olmsted, 1971). This survey also
B 3;0
showed that at least eight vowels or diphthongs t b B d
•J

were usually in use by the end of the second year: « 2;8

it
[i], [i:], [a], [u], [d], [px], [a:], and [ai]. By age 4, < t k 1

all the vowels and diphthongs were in use, and

2;4
only a few consonants were still posing problems m P h
- [0], [8]> [<&], and [3], and certain uses of [1], [rj],
2
[t], and [z]. n

It is also possible to see trends in the way children Below 2

change the sounds of the language, when they

240 • VII CHILD LANGUAGE ACQUISITION

Reduplication
During the second year, an Jigger tidi

effect known as reduplication scissors didi

(p. 1 75) is an important fea- take keke
T
ture of children's phonolo- Angie naeno
gies: the different syllables of B:
a word are pronounced in the Snoopy supi, nupi
same way. In one child, necklace nekis
water was pronounced
[wowo], bottle as [bubu], and
hungry
chip
h.\rjki

tip
H
window as [mumu]. Even water WDt
monosyllabic words can be chicken t/ik T
reduplicated, as when ball drop dAp
becomes [bobo]. sock Jap
Children do not all redupli- Francie faeti
T
cate to the same extent. With
some children, most words
are affected, and the process
hospital
hair
truck
pit
heir
tAk
i* |B i
ljl 1
B lll
ftl 1 jp,,

can be observed for several kitten kiki T

months. other cases, there powder pav
E H H
In
may be very few words in- pencil peta i;3
(1) (7) 1) 1)
E~3^ a.
may
volved, and the effect burger byga
last only a few days. The dif- outside ausaid T
ference can be seen in the
following word-lists, taken
from two children, A and 0.
boat
Eleanor
bot
ano
The purpose of reduplica-
i;4
(2)
ssai, (2)
a, a,, a,0 J
2]

A is a reduplicator: the list tion has been much dis-

contains ten complete redup- cussed. It may be partly mo-
lications and five items tivated simply by the need to
i>
a, a,a,0„,0 „ (
(6)
e=3
(2)
m„ m„ m„ h... eu m.,
where there is only a small play with sounds or to prac-
change between the sylla- tisethem. But it is more likely Phone trees So many time is shown vertically, cal line; if a phone was be-
bles. B uses just one redupli- that theprocess helps chil- things happen at once in a and the set of phones used ing used in different words
cated word (kitten). (After R. dren as they try to cope with developing phonological by H at any one point is (e.g. mama, milk), they are
G.Schwartz era/., 1980, the pronunciation of more system always dif-
that it is shown horizontally. linked by a dotted line. We
P. 79.) complicated words. It has ficult to what is going
follow The procedure grouped can thus see how Hs
A:
been argued that a word like on. 'Phone trees' are an at- together all Hs versions of phonological inventory in-
tiger, with its changes of con- tempt to visualize the a given word, as well as all creases over time, as well
Christmas d z id z i
necklace neke sonant and vowel, would be emerging relationships be- Hs words that began with as the variation affecting in-
difficult for a 1 1-year-old to tween different types of the same phones. Each dividual words. (From C. A.
hungry hA:n
learn at one go: reduplication phonetic segment, or class of phones is summar- Ferguson & C. B. Farwell,
chip ti

water WDWO would give the child a phone (§27) over a period ized within a box: for exam- 1975.)

chicken kika chance to master the pro- of time. The trees illus- ple, [p] (4 j
means that H Diagrams of this kind
banana mimi, mimi nunciation in stages, by first trated (above) show the in- used 4 tokens beginning have proved to be very
producing the word's syllable consonant system of a with [p] in a given session; useful in the study of
thank you dete: itial

sister sisa: structure and stress, along child, H, between |p~b| (2) means that there phonological acquisition,

belly button beba with the most noticeable 10months(0;10)and were 2 cases of variation as they clearly show the
phonetic features. A more 1 7 months (1 ;5), based on between [p] and [b]. If suc- difficulty of making simple
mouth mamav
precise pronunciation would a series of recorded sam- cessive samples contained generalizations about how
clock kak
come later, after this phon- ples. Development of differ- the same word, the boxes individual phonemes are
candy kei:
mimi etic outline had been ent classes of phones over are joined by a solid verti- learned.
money
house didi learned.

Intonation - early and late

Most have begun to
children
make some use of their lan-
guage's intonation patterns
(§ 29) before the end of the
firstyear. Different tones of
voice are used to express
such meanings as question-
ing, demanding, calling,
greeting, warning, recogni-
trastiveDaddy gone (i.e. mal patterns of intonation youngest children were understanding the differ-
not someone else). quite early on, but their hardly able to do it, and ence, signalled by intonation
As the child's grammati- awareness of the range of even the oldest children did and pause, between such
cal and social abilities de- meanings that these pat- not reach the level of com- sentences as she dressed,
velop, so new uses of intona- terns convey is still develop- petence shown by the and fed the baby (i.e. the
tion emerge. For example, ing as they approach their adults. In fact, only one child person dressed herself, and
the contrast between rising teens. This was first shown out of 28 got all the results then fed the baby) and she
and falling tones differen- in a study of the way British right (A. Cruttenden, 1974). dressed and fed the baby
tiates the two functions of a radio and TV announcers The implications of this (i.e. the baby is both dressed

tion, and surprise. During tag question in English read out football results experiment go well beyond and There can be few
fed).
the second year, as two- ('asking', He's outside,
as in (e.g. Everton3, Liverpool the world of football, for the clearerexamples of the dif-
word sentences develop isn't he?, and 'telling', as in 3). By listening to the into- intonation patterns used are ferences that can exist be-
(p. 242), a wider range of at- He's outside, isn't he!), and nation of the first part of the tobe found in everyday tween production and com-
titudes is expressed, and this islearned during the result, it is possible for speech also. It seems that prehension skills (p. 232).
prosody begins to signal dif- third year, along with the adults to predict whether the aspects of the intonation
ferences in emphasis. At grammar. score is going to be a draw, system are not only the first
this point, it becomes pos- What is surprising is that a home win, or an away win phonological features to be
sible to distinguish such the learning of intonation (p. 56). When this task was learned, but also some of the
general sentences as goes on for so long. Chil- given to children aged 7 to Even teenagers have
last.

Daddy gone from the con- dren seem to master the for- 1 1 it was found that the
, been shown to have difficulty

40 PHONOLOGICAL DEVELOPMENT .
241
 )

41 Grammatical development

SINGLE-WORD UTTERANCES
The earliest stage of grammatical development On wugs, and other things
hardly seems like grammar at all, since only single Languages often make use word in the sentence
words are involved - utterances such as Gone, ofa system of word end- There are two If the— .'

More, Dada, and Bye-bye. Sometimes longer- ings to express grammati- children said wugs, it was
cal meanings (§§1 6, 50). inferred that they had
sounding utterances are heard (such as Allgone or
English has around a do- learned the plural ending. If
All- fall-down), but these are deceptive: they have zen such endings (-ing, -s, they said wug, they had
been learned as whole phrases, and children use -ed, etc.). When do chil- not.
them they were single units.
as if dren learn about these It is important to use non-
Most of the words used at this stage (about 60% morphological aspects of sense words in experi-
grammar? ments of this kind, to guar-
seem to have a naming function and will develop Now there is another one.
One of the best-known antee a genuine response.
into nouns. About 20% express actions. Many of early studies in child lan- If realwords had been
There are two of them.
There are two
these will develop into verbs, though not all. When guage investigated this used (e.g. showing a pic-
a child says In!, holding a brick and gesturing vio- question (J. Berko, 1958). ture of a cow, and asking and plural) first appear at
lently at a container, we have to interpret this as
The experimenter elicited for cows), a correct answer around 18 months, but take
from the children a series by a child would prove very several months to be used
an action utterance, even though the word class of forms which required dif- little. The child might have correctly. Other endings
(p. 91) is a preposition. Other word classes are also ferent grammatical end- learned the form cows by appear at intervals over the
found at this stage (such as adjectives and adverbs), ings, such as plurals and heart at some point, and subsequent two years, and
along with several words that it is difficult to assign past tenses. The instruc- might not really know that show a similarly gradual pat-
to any word class (such as Bye-bye).
tions were all of the type il- it was composed of cow + tern of development. Many
lustrated in the picture -s. Invented words get irregular inflectional patterns
The 'one-word' stage is usually most noticeable (right), which was the first round this problem. (e.g. comparative forms such
between 12 and 18 months. But to talk about it item in the study. The ex- There have now been as worse) may not be cor-
solely in terms of 'words' is misleading. In many perimenter would show the several studies of children's rectly used until the age of 8
child the picture and read learning of English morpho- or 9.
respects, these early utterances function as if they
out the text, leaving the logy. They show that some
were sentences (and they have been given capital
child to supply the missing word endings (usually -ing
lettersabove, to represent this interpretation). For
example, one child used the word dada in three
different ways: as he heard someone approach out-
side, he said Dada?, with a rising intonation; as
he saw that it was indeed daddy, he said Dada,
with a triumphant, falling intonation; and then he
said Da-da!, with an insistent, level, intonation,
with his arms outstretched. At a later stage in de-
velopment, these three functions would be called
'question', and 'command'. At this
'statement',
stage, these utterancesdo not have a distinctive
grammatical form, but the use of prosody and ges-
ture conveys the force of these sentence types
nonetheless. In such cases, many scholars are happy
to talk about 'one-word sentences', or holophrases.

TWO-WORD SENTENCES
Most people think of 'real' grammar as beginning
when children string two or more words together,
which takes place around 18 months. This tends Sample 2
not to happen abruptly. There is usually a tran- Age (months) 27
sitional period, in which words are brought One child's development of ending takes 1 6 months to eral trend is rising. The
together, but the sequence is not uttered as a single, the -ing and plural inflec- get from a 50% level (at curves start to flatten out at
rhythmical unit, as in Daddy. Gone. Lengthy tions between age 27 and sample 1 ) to the 1 00% le- around the 90% level,
sequences of such words can often be heard: one 48 months. The diagram sample 16). There
vel (in which has led several
shows the gradual nature are several dips in the investigators to conclude
child said Daddy. Daddy.
Daddy. Garden. See.
of morphological acquisi- child's performance, prob- that this is the point when
Garden in quick succession. But soon two-word tion. The plural ending ably due to fluctuations in 'acquisition' of a feature
sentences emerge with great confidence — and takes only a short time to the number of instances can be confidently as-
increasing frequency. reach the 90% level of cor- that turned up in the vari- serted.
rect use, whereas the -ing ous samples, but the gen- (After R. Brown, 1973.)
Several studies have been made of the meanings

242 • VII CHILD LANGUAGE ACQUISITION

 . .

expressed by these two-word sentences. They If Father Christmas come down the chimney, and he Saying no
include such sequences as the following: will have presents when he came down, can I stay up Several studies have been
to see him? made of the expression of
an Actor performs an Action Daddy kick. negation by young children.
an Action affects an Object Shut door. The sorting out of grammatical errors is a parti-
1 The first negative words
an Object is given a Location There teddy. cular feature of 4-year-old speech. Many of the emerge in the second year -
an Object or Person is Described She cold. irregularities of syntax and morphology are being usually no or nor as a one-
mastered around this age, though it can take several word sentence.
These sentences could also be described in more years before such errors as the following are elimi- 2. The negative words com-
traditional grammatical terms. Daddy kick, for nated: bine with other words to
example, has the clause structure (p. 95) Subject make two-word sentences:
+ Verb; She cold could be analysed into Pronoun You bettern't do that. No sit, Gone no, Not there.
That's more better. Several different meanings
4- Adjective word classes. However, not all sen- can be expressed at this
tences uttered by children at this stage are capable Are there much toys in the cupboard?
stage- in particular, non-ex-
of a clear grammatical or semantic analysis. One It got brokened.
just istence, e.g. No car (while
child looked at a photograph of her father and said Are we going on the bus home? looking for a toy); rejection,
e.g. No drink (while pushing
Daddy Mummy; another put a car in a garage and The study of errors is important, because they show a drink away); and denial,
said Car want. We may hazard plausible meanings children breaking fresh grammatical ground. They e.g. Not mine (pointing to
to such sequences, but definite interpretations are provide the main evidence of how children go someone else's coat). (L. M.
often out of the question. Bloom, 1970.)
about actively learning new constructions.
3. During the third year,
More advanced grammatical constructions con-
negative words come to be
SENTENCE STRUCTURE tinue to be acquired throughout the early school
used within constructions,
At around 2 years of age, many children produce years. the age of 7, more sophisticated
Around e.g. You no do that, Mummy
sentences that are three or four words in length forms of sentence connection begin to emerge, not got it. At the same time,
such verbs as can and
(p. 233), and combine these words in several differ- using such words as really, though, anyway, and 't

won 't appear.

ent ways to produce a variety of grammatical con- (at later ages) for instance, actually, and of course.
structions. Typical sentences at this stage include Children begin to distinguish different underlying 4. The negative words and
endings come to be used
Man kick ball, Him
got car, Where daddy going?, meanings for sentences that look the same (e.g. Ask more accurately: not re-
and Put that on there. Questions and commands John what time it is vs Tell John what time it is). places no, and n't is used
are being used as well as statements, and different And they consolidate their awareness of the way with more verbs, e.g. You've
clause patterns are now evident. By the end of the sentences may have the same meaning even though not got one, She isn t going.
Double' negatives for em-
third year, clause structures of four or five elements they look very different — for example, the relation-
phasis (§ 1 are a normal de-
You give me my
)

can be heard, as in car now. ship between active and passive sentences (The girl velopment, e.g. Nobody
The 'telegraphic' character of early sentences has chased the boy*— The boy was chased by the girl),
> don't like to go in.

often been noted in many children - an impression which is not thoroughly sorted out until the ninth 5. A few advanced negative
derived from the omission of grammatical words year. A popular impression of grammatical learn- constructions are not ac-
(such as the and is) and word endings (such as ing is that it is complete by age 5; but recent studies quired until the early school
years, e.g. the use of some
-ing). By the end of the third year, this character have shown that the acquisition of several types
vs any (cf I've not got any
.

has largely disappeared, and children's sentences of construction is still taking place as children rather than 'I've got any), or
more closely resemble their adult counterparts. approach 10 or 11. the use of hardly and
the age of 3, there is a major grammati-
Towards scarcely.

cal advance, with the appearance of sentences con-

taining more than one clause. A large proportion
Asking questions
of these sentences are coordinate clauses (p. 95),
One of the first topics to be more complex as the third tus as a question - the
linked mainly by and - a pattern which, once studied by child language year approaches, e.g. word where and the inver-
learned, produces utterances that go on and on: researchers was how chil- Where Daddy going ?, sion is daddy. Children of-
dren learn ask ques-
to What you doing in there ? ten rely on the alone,
Daddy have breaked the spade allup and - and - and first

tions. Three main stages 3. A major advance comes and for a while produce
itbroken - and - he did hurt his hand on it and - and have been proposed for with the learning of the verb such sentences as Where
- and - it's gone all sore and . .
English, and similar de- to be, and such auxiliary daddy is going?. (After R.
velopments have been verbs as have and do. Chil- Browner a/., 1968.)
Sentences involving subordination (p. 95) are also noted in several other lan- dren discover the appar-
increasingly found at this age, using such words guages: ently simple rule that turns The complexity of question
1.The earliest stage statements into questions formation can be seen from
as 'cos, so, if, after, what, and when:
makes use of intonation by changing the order of the following selection of
I let go 'cos it hurted me. (§29), e.g. Daddy there?, the Subject and Verb (e.g. errors, all made by 2-year-
spoken with a high rising That is a car —> Is that a olds:
Tell me what it's called.
tone, in effect asks Is car?), and then learn that it Whose is that is?

A great deal of grammatical knowledge is required Daddy there?' is not so simple after all What are you did?
2. During the second year, (e.g. it cannot be * Went he What did you bought?
before these constructions are used correctly, and
children start to use ques- to town ?, but Did he go to Is it's my car?
common to find errors and non-fluency as chil-
it is
tionwords. What and town?). Sentences that use Don't he wanted it?
dren attempt to handle longer sequences. For where are usually the first question words pose par- Despite this complexity,
example, this child of 3 years, 9 months gets into to be acquired, with why, problems: Where is
ticular most of these difficulties
how, and who coming later. daddy going? has in fact are overcome before the
trouble with his sequence of tenses, as he tries to
These questions become two forms indicating its sta- age of 3.
express a complicated thought:

41 GRAMMATICAL DEVELOPMENT •
243
42 Semantic development

The learning of vocabulary (§17) is the most notice- The first 50 words
able feature of the early months of language acqui- How much do children say in a day? These are the first 50 words
sition. From the point when a child's 'first word' Using radio microphones and tape recorders, it is now used by two American chil-
is identified, there is a steady lexical growth in both possible to make large-scale surveys of children's dren between 1 1 and 16
comprehension and production. An indication of lexical usage. Large portions of a child's day can be months. There are very few
recorded - in some cases, covering everything the items in common, and major
the scope and speed of progress can be obtained
child says between waking up and bedtime. differences in order (e.g.
from a study of American 1-year-olds: the average The table below gives the age of several German mommy is Sarah's second
time it took eight children to get from 10 to 50 children recorded in one study, along with the length word, but Daniel's forty-

words in production was 4.8 months — about 10 of the recording, and the number of word tokens (§ 1 5) third).
used in the recording. As the recording times are not
new words a month. In comprehension, the chil- Daniel Sarah
the same, the right-hand column gives a standardized
dren understood an average of 22 new words each total, based on an assumed 1 2-hour day. 1. light 1. baby
month (H. Benedict, 1979). By 18 months, it is 2. uh-oh 2. mommy
thought that most children can speak about 50 Age 3. what's that 3. doggie
words and understand about five times as many. (years; Time 12-hr 4. wow 4. juice
months) (mins.) Tokens total 5 banana 5. bye-bye
6. kitty 6. daddy
1;5 202 3,881 13,800 7. baby 7. milk
The content of early vocabulary 1;8 241 3,907 1 1 ,700 8. moo 8. cracker
Young children talk about what is going on around 2;1 213 5,978 20,200 9. quack 9. done
them - the here and now' - and rapidly build a 3;6 189 9,891 37,700 10. cookie 10. ball
vocabulary in several semantic fields (p. 1 04). 5;4 152 6,464 30.600 11. nice 11. shoe
• People mainly relatives and house visitors - daddy, 8;7 193 6,630 24,700 12. rock (noun) 12. teddy
baba, grandma, man, postman. 9;2 311 10,524 24,400 13. clock 13. book
• Actions the way things move {give, jump, kiss, 9;6 869 25,401 21 ,000 14. sock 14. kitty
gone), and routine activities in the child's day (bye-bye, 97 804 28,142 25,200 15. woof-woof 15. hi
hello). 16. daddy 16. Alex
• Food occasions as well as products - din-din, milk, These results far exceeded the expectations of the 17. bubble 17. no (-no)
juice, drink, apple. researchers. No-one had imagined that children as 18. hi 18. door
• Body parts usually facial words first (mouth, nose), young as 2 could produce in excess of 20,000 words 19. shoe 19. dolly
then other areas (toes, handie(s)) and body functions in a day, or that a 3i-year-old could produce nearly 20. up 20. what's
(wee-wee). twice that number! that?
• Clothing of all kinds - nappy/diaper, shoes, coat. The number of different words (word types, § 1 5) 21. bye-bye 21. cheese
• Animals whether real, in pictures, or on TV - used during a day was also much larger than had been 22. bottle 22. oh wow
doggie, cat, horse, lion. expected. These ranged from a remarkable 1 ,860 (for 23. no 23. oh
• Vehicles objects and their noises - car, choo-choo, the 1 ;5-year-old) to over 5,000 for an 1 1 -year-old, with 24. rock (verb) 24. button
brrm. an average of 3,000 for the whole group. (After K. R. 25. eye 25. eye
• Toys and games many possibilities - ball, bricks, Wagner, 1985, p. 477.) 26 nose 26. apple
book, dolly, peep-bo. No corresponding survey has yet taken place for 27. fire 27. nose
• Household objects all to do with daily routine - cup, English. But as the children came from a variety of 28. hot 28. bird
spoon, brush, clock, light. social backgrounds, and engaged in many kinds of 29. yogurt 29. alldone
• Locations several general words - there, look, in, up. activity during their day, it is likely that the figures will 30. pee-pee 30. orange
• Social words response noises- m, yes, no, ta. be fairly typical - in which case, traditional impressions 31. juice 31. bottle
• Describing words early adjectives - hot, pretty, big. of children's vocabulary growth (p. 232) will have to be 32 ball 32. coat
• Situational words several pointing' words (deictics, radically revised, in an upward direction. 33. whack 33. hot
p. 1 06) - that, mine, them. 34. frog 34. bib
35. hello 35. hat
36 yuk 36. more
37. apple 37. ear
The meaning of early words 38 Big Bird 38. night-
Children do not learn a word with its meaning 'ready night
2. Underextension In this case, the word is used with
made'. They have to work out for themselves what it must 39. walk 39. paper
a narrower meaning than it has in the adult language. Dog
mean, and in so doing they make errors. Three types of might be applied only to the family dog, or shoes only to 40 Ernie 40. toast
error occur often during the second and third year. a child's own shoes. 41. horse 41. OToole
42 more 42. bath
3. Mismatch Here, there is no apparent basis for the 43. mommy 43. down
1 Overextension A word is extended' to apply to other
. wrong use of a word by the child, as when in one case 44. bunny 44. duck
objects that share a certain feature, such as a common a telephone was referred to as a tractor. There is usually 45. my 45. leaf
property of shape, colour, or size. Dog might be applied no way of tracing back the association of ideas that has 46. nut 46. cookie
to other animals, or moon to other round objects. caused such misidentifications. 47. orange 47. lake
48 block 48. car
49 night-night 49. rock
(noun)
50. milk 50. box

(After C Stoel-Gammon & J.

A.Cooper, 1984, p. 264.)

244 • VII CHILD LANGUAGE ACQUISITION


Cups and glasses
Children can take several
years to learn the meaning
of a word, especially when
the word is used along with
others to refer to objects or
ideas that are not easy to
distinguish. Even everyday
objects may prove
to differentiate and
a consistent way - such as
the distinction between
'cups' and 'glasses'.
The uncertain boundary
between these categories
is well illustrated from the
After the age of 18 months, very little precise
keeping cup for the smal- By age 9, some quite information is available. By age 2, spoken vocabul-
lest items. subtle distinctions were be- ary probably exceeds 200 words. But after this,
Both glass and mug ing made, with the names
estimates become extremely vague. A dramatic
came to be used regularly and showing
definitions
by the 6-year-olds, who that the childrenwere be- increase in the size and diversity of the lexicon takes
also added such labels as ginning to be aware that place during the third year, so much so that it has
dish and vase. A still wider they were dealing with an so far proved impossible to make accurate calcula-
range of labels was used area where boundaries are
difficult
tions (especially about vocabulary comprehen-
by the older children (e.g. vague. This is most clearly
label in eggholder, can, measure). shown from the definitions
sion), or work out any norms of spoken lexical
During this period, also, they gave, which contained frequency. What happens at older ages is largely
there was an increasing qualifying words, such as guesswork, and vocabulary totals cited for these
use made of words reflect- usually or could have; for children should therefore be viewed with great cau-
ing the perceptual proper- example 'a cup holds
- espe- tion.
ties of the objects things to drink, and some-
cially their size, shape, and times has a handle .', .'a .

25 drinking vessels drawn material (e.g. big, round, glass is like a cup only it Other issues
in the diagram below. For paper). However, between could be taller, doesn't
adults, some of these are The study of semantic development takes in far
6 and 9, the children's pre- have a handle, and could
clearly cups, some are ferences showed a clear be plastic or glass'. is age
It
more than vocabulary (p. 107). Grammatical con-
clearly glasses, and some change: the older they be- 12 before competence be- structions also need to be studied from a semantic
require a more complex came, the more they pre- comes close to that of an point of view — for example, the way in which
kind of description. How ferred to use attributes ref- adult - nine years or more children master the complex conditional meaning
long does ittake children to lecting what they thought after the distinction is first
become aware of these were the functional proper- of if constructions, or the causal meaning of
introduced. (From E. S. An-
distinctions? ties of theobjects (e.g. me- dersen, 1975.) because, so, or since. That there are problems here
Children aged between dicine cup, Martini glass). can be readily shown from the errors children
3| and 1 2k were shown make:
this set of objects, and
asked to carry out various
tasks, such as naming, de-
1 D 6
^ 11
[J
16 Q) 21
p The man's fallen off the ladder because he's broken
his leg.
and sorting. The
fining,
2 had one name was
youngest children used
fewer names to describe
far y 7 'D 12 Q 17 O 22 fp I fish left, because its Bill.

Children aged 8 or 9 may have mastered the gram-

the objects, often overex-
tending the word cup to ap-
3
? 8 D 13 23
mar of such constructions very well, but still be
ply to items which the 1 2-
[J 18
Yj g^ having difficulty with the meanings they encode.
year-olds called glass (e.g.
items 4 and 9 in the dia- 4? 14 O 3

24
Auxiliary verbs such as ought, must, and should
provide another problem area, as do subtle pre-
gram). One 3-year-old D I D>
positional forms (e.g. despite) and verb contrasts
went the other way, calling
most things glass, and 5 O 10 g
15
U
rf
20 B? 25
(e.g. ask vs tell, say vs promise). The ability to use
\J) figurative expressions, and to see double meanings
in language, also develops largely after the age of
Four replies
6.
Age 3 Age 6 Age 9 Age 12 One of the most significant developments of this
1. cup glass glass fruit cup later period is the child's emerging ability to inte-
2. big cup glass glass wine glass grate several features of semantic knowledge into
3. cup glass a measure cup for liqueur
little
a single defining statement. Young children cannot
4. little cup glass glass juice glass
define: in response to such questions as 'What's
5. big cup dish little tiny dish fruit cup
X?', they give empty, ambiguous, or idiosyncratic
6. big cup glass glass glass
replies. What's a shoe? asked an adult. That, re-
7. cup glass _ cup glass
8. little cup plastic cup glass medicine cup plied one young child, pointing. And a sock, replied
9. big cup plastic cup glass glass another. Mummy got a shoe replied a third. Gradu-
10. cup plastic cup cup outdoor cup ally, however, definitions become more sophisti-
11. Ron McDonald cup plastic cup glass Ron McDonald happy cup cated. A particular feature is singled out {A knife
12. little cup plastic cup cup cup is sharp) or its function is specified (A knife is when
13. cup glass glass cup you cut with it). not until around age 8
But it is
14. big cup cup cup coffee cup
or 9 that statements with something resembling
15. coffee cup plastic cup cup sippercup
an adult definitional form begin to be produced
16. cup cup cup coffee cup
(A apple is a sort of fruit, and it's round and red,
17. paper cup paper cup cup Dixie cup
18. big cup cup glass glass
and we eat it). (B. Litowitz, 1977.)
19. orange cup metal cup measure for wine measuring cup Semantic development continues throughout the
20. little cup metal cup cup baby's metal cup school years - and, indeed, throughout adult life.
21. coffee cup mug coffee mug coffee cup Unlike phonology and grammar, it is not largely
22. coffee cup glass cup cup beer mug over when children enter their teens. There is
23. coffee cup glass (mug) cup mug always new vocabulary to be learned, and new
24. cup glass cup coffee cup worlds of meaning to explore.
25. cup cup cup coffee cup

42 SEMANTIC DEVELOPMENT .
245
43 Pragmatic development

The task of language acquisition requires that to initiate a dialogue — the various ways of obtain- Talking backwards
much more than patterns of sound,
children learn ing and holding a listener's attention. They can From around age 7, children
grammar, and vocabulary. They must also learn handle several of the conventions of turn-taking. develop a large creative re-
to use these patterns appropriately in a rapidly They know a great deal about how to respond pertoire of interactive linguis-
increasing range of everyday social situations. appropriately - for example, by providing clarifica- as they learn to tell
tic skills,

jokes and riddles, insult each

This developing pragmatic awareness (§21) has tion when requested to do so.
other, maintain group iden-
attracted a great deal of study in recent years, par- These skills develop greatly between 3 and 5. tity, and make up language
ticularly in relation to the way children learn strate- In particular, there is a major development in child games (p. 59). One of the
gies of conversational interaction. not yet It is awareness of the social factors that govern a suc- most remarkable of these
abilities is talking backwards.
possible to talk about definite stages of develop- cessful conversation — such as the correct use of
A study of two 9-year-olds
ment; but the very early age at which these strate- forms of address and markers of politeness (e.g. who were able to talk back-
gies emerge is now clearly established. please, sorry), and how to make requests in an wards showed two quite dis-
indirect way. They also learn to anticipate points tinct styles. One child (A) re-

of potential breakdown (carry out conversational versed the sounds of each

word and ignored the spell-
'repairs', p. 116), such as by repeating utterances
Conversational skills ing. The second (B) reversed
that are unclear, or asking for clarification. In par- the spellings, sounding the
Between the ages of 2 and 4, a remarkable develop- ticular, they develop their ability to cope with situa- letters out. The pronuncia-
ment takes place in the ability of the child to par- tions where they do not have things all their own tions which resulted were
ticipate in a conversation. At the earlier age, very different. Size, for ex-
way. In one study of two 4-year-old children play-
ample, would come out as
conversations are often very erratic and disjointed, ing together, there were 576 sequences in which [zais], using As method, but
with parents doing most of the 'work', and children one child (A) requested the other (B) to perform would come out as [ezis], us-
using sequences of utterances, many of which are an action; in 122 cases, B refused to comply. It ing B's. Here is a selection
not obviously directed to any listener. The effect of their reversed words:
was therefore necessary for A to adopt various per-
is a curious mixture of monologue and dialogue: suasive tactics in order to gain compliance: B
nine nam 'enin
child: Ball. Kick. Kick. Daddy kick. a: Say yes. guy aig jag
mother: That's right, you have to kick it, don't you. B: No. boil bib ljab
child: Mmm.
Um. Um. Kick hard. Only kick hard. A: I'll be your best friend if you say yes. mouse saum 'esuam
Our play that. On floor. Our play that on floor. Now. continue ujs'nitnak utsnik
Our play that. On floor. Our play that on floor. No A: Change lunch boxes. bomb mab bamab
that. Now. B: No. castle la'saek 'eltsaek

MOTHER: All right. axe skae ks£e

A: You'll have a bigger one, so you will.
child: Mummy, come on floor me. bone nob 'enab
(M. McTear, 1985, p. 109.)
auto o'ta 'otuwa
mother: Yes.
child: You tip those out. Some of these exchanges can be very lengthy. In inhale le'ni ela'nae
elevate 'tevale .etae'levet
mother: Mm. All right. thesame study, A's request for a pair of scissors
child: That one broke. was continued for over 60 turns before it was The sentence Please pre-

(P. Fletcher, 1985, p. 64.) (reluctantly) complied with. sent an idea to the class' was
translated by Sas: [eselp
Studies of young children's conversations show tane'zep nae 'aeda 'At ?et
The contrast with a 3-year-old's conversation is that many adult interaction skills are already pre- Vselk]. The words are not
striking, with both parties very much involved with sent well before school-age. There is still a great always perfectly accurate re-
the detail of what each is saying: deal to learn, of course — for example, 5-year-olds versals; but there is clearly a
system of rules governing
child: Hester be fast asleep, mummy. do not make much use of such 'manipulating'
their production.
mother: She was tired. devices as you know or actually, and they must Once someone learns to
child: And why did her have two sweets, mummy? learn the strategies associated with the more formal talk backwards, the ability
mother: Because you each had two, that's why. interactions that are part of educational learning seems to stay. Interviews
She had the same as you. Ooh dear, now what? with27 adults who had been
and discussion (§44). But all of this will build on
child: Daddy didn't give me two in the end. backward talkers as children
a foundation of conversational ability that in many showed that the ability was
MOTHER: Yes, he did.
children is already extremely sophisticated by the still present. Some were only
child: He didn't.
fifth year. able to do it slowly, or on
mother: He did. short words, but three re-
child: Look he given one to -two to Hester, and two tained an impressive facility,
to us. reversing not only the order
mother: Yes, that's right. ofsounds in words, but the
child: Why did he give? order of words in sentences
mother: 'Cos there were six sweets. That's two each. as well - and often at speeds
(P.Fletcher, 1985, p. 91.) very similar to those found in
forwards speech! (After N.
By 3, it is plain that children have learned many Cowan & L. Leavitt, 1 982,
aspects of conversational strategy. They are able pp.491,ff.)

246 VII CHILD LANGUAGE ACQUISITION

TWINS
The language learning environment of twins is keep a conversation going, and to interact with
unique. During their early years, their linguistic
Secret languages
adults. By age 7 or 8, the delay seems to have dis-
Twins have often been ob-
experience differs greatly from that of single chil- appeared. served to talk to each other
dren. Singletons receive most of their language in a way that is unintelligible
stimulation from adults or older children, whose to adults or other children.
utterances provide a more advanced learning 'tar- The phenomenon has been
variously labelled 'crypto-
get'. Twins, however, spend a great deal of time
phasia', 'idioglossia', or au-
together, with each learning from a linguistic set-
tonomous speech'. Esti-
ting in which the other speaker is at the same devel- mates of incidence are
Poto and Cabenga
opmental linguistic level. In such circumstances, it uncertain, but some have
grace Cabenga, padem manibadu peeta.
:

suggested that as many as

is hardly surprising to find many twins developing
Virginia Doan nee bada tengkmatt, Poto.
:

40% of twin pairs develop

a private form of communication.
Reported extracts of this kind from a twin conversation some form of private speech,
One study found a great deal of private language achieved world-wide publicity in the late 1970s. They especially in the second
play in early-morning twin conversations. At 33 came from the Kennedy twins of San Diego, California, year.
months, for example, there were dialogues in which who at the age of 8 were still using their own private There seems to be no
language. They called themselves by different names basis for the view that a com-
each child responded to features of pronunciation
in this language: Grace became 'Poto' and Virginia pletely different 'language' is
it noticed in the other: became Cabenga' - names which were later used as involved. The patterns heard

A: zaeki su the title of a film about their early years. can largely be explained with
Their totally unintelligible speech for a while reference to the children's
B: (laughing) zaeki su zaeki su (both laugh) ae:
promoted the impression that the children were cope with the kind
efforts to
A: api:
mentally retarded, but this proved not to be so. In due oflanguage used around
B: olpolt olt
course, a detailed study of their language came to be them, and to the kind of pro-
a: opi: opi: made. This indicated that their speech was not as alien cesses that take place in nor-
B: api: api: (laughing) api api api as its bizarre sound had led people to believe, it was mal language acquisition.
a: ai ju basically a severely distorted form of English, with The twin situation promotes
B: (laughing) ai ju api (repeated several times) some features of German, several idiosyncratic the continued use of imma-
A: kaki (repeated several times) grammatical characteristics, and a proportion of ture and idiosyncratic pat-
invented vocabulary. What made it so difficult to follow terns of sound, grammar,
B: ai i: o:
(and also to analyse!) was its extremely rapid speed and vocabulary, and a perso-
A: ai i: oo:
of articulation and its staccato rhythm - features that nal style of interaction often
(E. O. Keenan, 1974, p. 171.)
later transferred to their English, when therapists characterized by abnormal
To the outsider, this kind of dialogue might resem- began to work with them. intonation and rhythm.
There are probably special reasons for the late These patterns become par-
ble a 'secret language', but it is no more than a
retention of private speech in this case. The children, ticularly noticeable when the
form of phonetic play. itseems, had very little opportunity to hear good children continue to use
One of the most interesting features of twin lan- models of English speech in their early years. They them past the normal period
guage is the way in which they 'share' the response saw few other children in the area where they lived. of 'baby talk'. In the most
Their parents were both working, and during the day dramatic cases, private
to an adult utterance:
they were cared for by their German grandmother who speech has lasted until age
mother: What can you see in the picture? spoke no English. There was also an expectation that 5 or more, when it often at-
twin A: A cat. they might be retarded (because of a history of tracts a great deal of publi-

twin B: And a dog. convulsions), which affected the style of the parents' city.

interaction. Left to themselves, the twins would have

Observers have been struck by the intuitive way had little alternative but to develop their own medium
in which one twin is able to respond very rapidly of communication.

to what the other has just said, and how the first
twin is able to anticipate when to stop. They very
seldom talk at the same time. Even very short utter-
ances can be split in two:
mother: What do you want me to read?
twin A: Puss.
twin B: In boots.
This kind of skill can only come from the frequent
opportunities the twins have had to interact, in the
early years. They know each other's rhythms, and
each is able to predict a great deal of what the
other is likely to say.
Perhaps because of dependence, twins
this close
are usually somewhat developing their indi-
late in
vidual language skills. When their language is for-
mally assessed, during the third and fourth years,
it is often found to be about 6 months behind the

norm for singletons. On the other hand, there are

certain aspects of their development that may be
ahead of other children — notably, their ability to

43 PRAGMATIC DEVELOPMENT .
247
 . .

44 Language development in school

When children arrive in school, they experience contact with the ability to use spoken language;
a different linguistic world. They meet for the first and oral skills are being supplemented by work Imaginative
time children from unfamiliar regional, social and on listening comprehension. Above all, teachers speech and
ethnic backgrounds, whose linguistic norms differ have begun to stress that children's linguistic ability
writing
greatly from their own. They encounter a social is a major factor influencing their success in the

situation in which levels of formal and informal learning of other subject areas, such as science,
speech are carefully distinguished, and standards mathematics, and history.
of correctness emphasized. The educational setting In the 1970s, this central, integrating role of
presents them with a variety of unfamiliar, subject- language work promoted a host of new language
related styles of language. They have to learn a schemes, materials, and approaches, and a philos-
new range of linguistic skills — reading, writing, ophy which is best summarized in a phrase that
and spelling. And they find themselves having to has since become something of an educational slo-
talk about what they are doing, which requires that gan in Britain, following the UK government's pub-
they learn a special technical vocabulary — a 'lan- lication of the 1975 report on the teaching of
guage for talking about language', or metalan- English in schools, A Language for Life ('The Bul-
guage. lock Report'): 'language across the curriculum'.
In recent years, educationists have begun to Since then, other aspects of the role of language
recognize the complexity of the language demands have come to be better appreciated - not least, the
being made on the young schoolchild, and to realize need for a corresponding emphasis on children's
that progress in many areas of the curriculum is 'vertical' development, as they move between
greatly dependent on a satisfactory foundation of classes and schools and encounter different kinds Look at this strange animal
linguistic skills. The traditional emphasis on liter- of language experience in a variety of subject areas. on the blackboard (hard
and write, has been supple-
acy, the ability to read From the children's point of view, a great deal of luck!/ What sounds would it
make when it is: happy?
mented by an emphasis on oracy, the ability to language work must seem fragmentary and incon-
hungry?. ..
speak and listen. Teachers now pay increasing sistent, as long as no effort is made to ensure a
attention to a child's preschool linguistic experi- coherent frame of reference that will accompany This is a task taken from a

ence, seeing this as a foundation on which they them as they move through the educational system. course designed to foster
children's ability to use
can build. Special efforts are made to relate differ- Research into the best means of achieving a coher-
speech and writing in a more
ent kinds of linguistic learning: the task of writing ent linguistic approach is thus a major aim of the imaginative way, aimed at
is being brought closer to the child's experience slowly developing field of educational linguistics. children aged from around 7
of reading; reading, in turn, is being brought into to 1 2 (R. James & R. G. Gre-
gory, 1966). It introduces
elements of art, music, and
drama, and stresses the en-
joyment that can be obtained
from the use of language.
The various exercises try to
No talking! make children more aware of
the different qualities of
The traditional emphasis in An example of the difficulty The first question is open- The teacher teaches within sounds made by things, ani-
schools on the written lan- comes from a class on reli- ended, and P1 takes this to his frame of reference; the mals, and people, and to sti-
guage (§ 1language
) led gious education, in which mean that an improvised pupils learn in theirs, taking mulate their imagination to
educationists in the 1 970s Treacher) is asking two answer would be appro- in his words, which 'mean' use language themselves in
to emphasize the important F\upils) about life in New priate. He does his best, but something different to them, a more personal and creative
role of speech as part of Testament Palestine. receives only the comment and struggling to incorpor- way.
children's development and T: How did they get the 'You might do it that way', ate this meaning into their
learning. was pointed out
It water from the well? ... Do spoken in a doubtful intona- own frames of reference. What sounds do you make
that talk was often discour- you remember? . . . Yes? tion. P2, perhaps noticing The language which is an when: you are given an ice
aged, and that in almost P1: They ... ran the bucket this, gives Ta different kind essential instrument to him cream? somebody treads on
every setting, teachers did down er . . . . . . and it was of answer, the name of the isa barrier to them. (D. your toe ? you are in a
most of the talking - filling fastened on to this bit of object. This is evidently the Barnes, 1969, pp.29, 30.) haunted house and a door
up 70% of classroom time, string and it [some inau-
. . .
answer Twants, and so P2 opens?...
according to one study. dibility] . . . other end to the gets the praise - even As a result of such studies,
A great deal was learned water. though we might think that the spoken language of Make the sound of a clock.
from the analysis of tape re- PTs response was not un- teachers and pupils became Now, instead of the sound,
T: You might do that way. it

cordings of teacher-pupil . . . Where did they put the worthy. a major focus of interest in make the word that comes
dialogue. One study was Having studied several the 1970s. from the sound ..
water . . . John?
able to draw attention to a P2. In a big . . . er . . . pitcher. cases of communicative dif-
hitherto unrecognized pro- the auth- At more advanced stages,
T: Good ... in a pitcher . .
ficulty of this kind,
blem in the styles of ques- or comments: the children study the power
which they carried on their
tioning used by teachers. of descriptive words and the
...?
P2. Heads.

248 • VII CHILD LANGUAGE ACQUISITION

 STRUCTURE vs USE
There is considerable recognition these days of the
need to develop a child's linguistic skills so that
they will promote educational growth inside school
as well as social and personal development outside
school. However, there is much less agreement
about the best way of achieving a competent and
tivities, the need to have available a more systema-
tic way of analysing and talking about the language
became increasingly apparent. However, the ap-
proach made no provision for a standard descrip-
tive apparatus. Criticism therefore came from
teachers who felt such an apparatus was necessary,
in order to facilitate the understanding of texts,

Jack aruL Jill

confident use of language in these domains. Since and to make comparisons between them. Other
the 1950s, the controversy has focussed on the critics argued, with varying levels of cogency and
competing claims of two approaches to language emotion, that a return to 'older standards' was
teaching- one based on 'structure', the other based prerequisite for progress.
Sta/np. titamn! CLIMB on 'use'.
Current trends
RK Well! Parsing The situation in the middle 1980s is unclear. At
The traditional structural approach stresses the one extreme there are those who advocate a whole-
Clop, oulh! importance of getting children to analyse grammar sale return to the principles and practices of tradi-
in a conscious way, and to learn the appropriate tional parsing. At the other, there are those who
terminology. The technique, as traditionally avoid anything that smacks of grammatical termi-
practised, is known by
such names as 'parsing', nology. Fortunately, several fruitful intermediate
'clause analysis', or 'diagramming' (p. 96). Sen- positions seem also to be emerging. There for
W r^ r,^ tences would be analysed into their constituent
parts, the labels for the different parts of speech
example, a healthy movement arguing for a general
approach based on 'awareness of language', aimed
is,

iM"5fc -JMJXJ} <juuuJ? would be learned, and great stress would be laid
on the mastery of formal rules (§§1,62).
nh iootraao
This is still the most widely practised approach
to mother-tongue teaching in Europe, but during
° the 1960s it fell out of favour in most English-
p
speaking countries. Parsing was felt to develop a
particularly at the middle years of schooling. It
aims to stimulate the child's curiosity about lan-
guage, and to integrate the various elements of lan-
guage training in school — both 'horizontally', in
relation to the different language elements across
the curriculum (English, foreign languages, minor-

^ J0i!6£'D
An 1 1 -year-old's 'poem',
based on some of the work
in the James & Gregory
course
sound effects conveyed by
names:
What taste and what size
would you give to sweets
mechanical, analytic set of skills that bore little
relationship to children's everyday linguistic needs.
ity languages,
moves up through
etc.), and 'vertically', as the child
the school. Structural and func-
Most children, it was argued, found such exercises tional elements are involved, and much of the
dull and irrelevant, unrelated to the problems and appeal of the approach lies in the way in which
practices of living English. pupils are invited to discover the importance of
both these aspects of language work for themselves.
Language in use Contrasting covers of two school books reflect the change
developed various ap- in attitude towards language teaching in recent years. The
In place of parsing, there
first is a grammar book intended for use in Brazilian primary
proaches to the study of language in use. Pupils
schools: the children are shown happily holding a grammar
and their teachers would collect samples of real book, and writing out a list of parts-of-speech labels. The
linguistic situations, both spoken and written, and second is a language awareness text for use with young

with these names? Cholly,
Wumps, Chooce,
Teenies,
Jelloc, Quangles?
They study the stress pat-
terns of phrases and sen-
tences in prose and poetry;
they begin to experiment
with language, in the form of
nonsense verse; and they
look at how the effects are
produced in a range of des-
criptive and atmospheric
poetry. Here, this kind of ap-
proach links up with the tradi-
tional focus on written litera-
ture. But it is hoped that by
grounding the study of liter-
ary language in the early ex-
perience of speech and
sound, children will develop
a more sensitive and crea-
tive approach to language
work, not only in their imagi-
native reading and writing
but also in their everyday
speaking and listening.
discuss the distinctive features of the language. senior-school pupils: it shows some of the varied topics, from
smoke signals to television, that form its content.
Typical situations, used by one leading project of
the 1970s for work with older secondary school
children were: advertising, news reporting, operat-
ing instructions, being tactful, using technical
terms, and projecting an image. Pupils would use
tape recordings, collections of written material,
role play, and other techniques that provided invol-
vement and realism. By choosing situations that
would be meaningful and motivating, educationists
hoped that pupils would explore the nature of their
experience when using or responding to language,
and thus arrive at a fresh understanding of its
meaning for them. In this way, it was felt that they
would develop their awareness of what language
is and how it is used, and by degrees extend their
own competence in handling it. Similar ideas were
introduced at junior levels using more elementary
functional notions, such as asking questions, giving
instructions, and providing explanations.
The 'language in use' approach is still widely
practised, but too has limitations. As teachers
it
and pupils became more involved in language ac-

44 LANGUAGE DEVELOPMENT IN SCHOOL .

249
Learning to read
Literacy has long been considered the main evi-
dence of a child's educational progress. As a result,
more attention has been paid to the nature of the
task facing children as they learn to read than to
any other area of the curriculum. Hundreds of Part of a page from Barnes's
reading schemes and philosophies have been de- New National First Reader,
vised in the past 200 years, and many have achieved 1883). The book gives chil-

a degree of success. However, it is usually an open dren practice recognizing

in
words in whole sentences.
question whether success is due to the properties
Pictures have replaced cer-
of an approach or to the enthusiasm with which tain key words to help reduce
it is promoted by its adherents. Systematic research the reading load and to pro-
into the teaching of reading is relatively recent, and vide cues to meaning.

although a currently fashionable field, there have

as yet been few definitive findings into the nature
of the child's reading process (discussed in §34) Active reading
and the effectiveness of different teaching methods. There are two contemporary preoccupations. First,
Many positions are advocated. Some recom- there is a focus on the need to motivate children
mend the initial use of a particular scheme or to read by providing materials and activities that
method to all children; others argue that there is are interesting. It is pointed out that the content
no 'right way', and that a range of approaches of traditional reading-scheme books is singularly
Breakthrough The active
should be available to suit the needs of individual uninspiring: children often view such reading as approach to reading is
children. For some, reading is essentially the skill a dull, decoding task, and choose very different nowhere better illustrated
of decoding written symbols; for others, it is a kinds of books when they read by themselves for than in the use of the word
means of discovering the meaning 'behind' the sym- cards and sentence maker
enjoyment. Today, this contrast is less apparent,
provided by the scheme
bols. In this deeper view, reading plays a funda- with new schemes placing a greater emphasis on Breakthrough to Literacy (D.
mental role in promoting children's critical and story-telling and more appealing visual design. The Mackay era/., 1970). Words
imaginative thinking, and thus their intellectual world of the child's own experience L also increas- are taken from a holder and
and emotional development. A similar concern ingly represented, through the use of familiar social placed on a sentence frame,
to make sentences of a
motivates the view that the teaching of reading situations and everyday visual language contexts,
child's own choosing. The
should not be restricted to the classroom. In par- such as road signs, shop names, and vehicle labels sentences can build on the
ticular, several recent studies have indicated that (e.g. taxi, police). children's personal
regular parental reading aloud to children, accom- Secondly, there is an emphasis on training the experiences; they use words
panied by informal discussion of what is being cognitive skills that children need in order to read that reflect their spoken
language; and they can be
read, may be the single most important factor in
without the need to
'written'
promoting reading ability. cope with the task of hand-
writing.

Syllabic teaching cards are part of a set designed to

introduce children to reading through the use of syllables (as
opposed to phonemes or whole words). The particular
programme is known as the syllabary curriculum'. At the
stage shown, pictures are being used to motivate a sequence
of words in a sentence. Later stages involve techniques for
the identification of sounds and the blending of syllables.
(From P. Rozin & L. R. Gleitman, 1974.)

250 VII CHILD LANGUAGE ACQUISITION

 .

efficiently. Research has shown the relevance of

such sequencing, and pat-
abilities as classifying,
Some approaches to reading
tern matching; and new 'pre-reading' materials
It is not possible to make a
neat classification of reading schemes into types, because
therefore provide practice in these areas, along with many are basec on a mixture of principles. However, some general comparisons can
opportunities to draw, cut out, colour in, and so be made, as is illustrated below for three kinds of approach.
on. It is also evident that many children find their
Characteristics Advantages Disadvantages
first encounter with the world of print confusing, 'Basal reading' Widely used in the U.S.; a Comprehensive; Expensive; can be
so that a great deal of attention is now being paid programmes large-scale system of pre- graded; carefully used inflexibly;
to ways of providing an opportunity to think about paratory texts, graded planned; children does not promote
what is involved in reading and writing (e.g. how readers, work books, tests, get to know the an exploratory use
and other materials. characters, set- of language out-
are books made? what is writing for?), and a meta-
ting, etc. side the scheme.
language for talking about these activities (e.g.
Language Integrates work in listening, Caters for indivi- Little grading of
page, line, beginning, space).
experience speaking, reading and writ- dual differences; structure; difficult
Much of this work involves an active approach programmes ing, using a variety of ma- promotes creati- to evaluate pro-
to reading. When children encounter a word they terials and activities relat- confidence,
vity, gress; little motiva-

cannot read, emphasis is now often laid on helping ing to the child's own world. and meaningful tion to read out-
activities. side the child's
them to work out for themselves what it must
immediate world.
be, by using such techniques as reading on to the
end of the sentence, reading back to the sentence Individualized Children select their own Maintains interest; Difficult to orga-
programmes reading based on interests fosters indepen- nize book availabi-
beginning, and checking any illustrations. In this
and ability, and read at dence and confi- record-keep-
lity,
approach, the intention is to make them rely less their own pace; each child dence; is flexible; ing,discussion
on the mechanical task of decoding letters, and to has its own programme, makes no public time with teacher;
capitalize more on their linguistic experience and using checklists and distinction be- no systematic de-
charts,and discusses tween good and velopment of
awareness of context so that they can guess what
reading with the teacher. bad readers. skills.
a word might be. Reading, according to the Ameri-
can researcher, Frank Smith (1928- is a 'psycho-
)

linguistic guessing game'. This is not to say that

accuracy in word decoding is unimportant. Rather,
it is stressed that 'getting the words right' is a
Alphabet decoding
gradual process - as indeed it is inspoken language
Several techniques have to highlight certain sound- over; digraphs have a bar
acquisition (§38).
been proposed that try to letter combinations. Diacri- under; schwa (see Appen-
In recent years, attention has also been drawn facilitate the task of decod- tics may also be added to dix I) is marked with a dot;

to the many different kinds of activity that are ing the phonic basis of the letters, as in the following il- and an asterisk is used for

found under the heading of 'fluent reading'.At one writing system. Some, such lustration of a system from exceptions.)
as Unifon or i.t.a, (p. 217), the 1960s. (Silent letters
extreme, there is the careful, complete, and vocal 7

introduce new symbols. are marked with a slash; Once upon a time a hen
technique known as 'reading aloud'; at the other, Some make use of colours long vowels have a bar live'd on a farm. The hen . .

there is the rapid, selective, and silent technique

known as 'scanning' or 'skimming' - something
widely practised by time-pressed adults as they
work quickly through a report or read the morning 'Phonics' vs 'whole word'
newspaper. In between there are many other activi- Since the early 9th cen-
1 these
ties, such as critical reading (e.g. underlining tury, the relative merits of
sections of text, or adding marginal notes), proof- phonic and whole-word ap-
reading (checking one's own or someone else's text proaches have dominated
educational debate about
for errors), and reading for learning (if you sud-
the teaching of reading.
denly discovered that as soon as you had read this Schemes have been de-
page you would be asked questions on it, your read- vised based largely on one
ing strategy would alter immediately). Current principle or the other, and
there have been several
thinking about reading draws attention to the
'mixed' schemes, which
importance of all these real-world skills. tempt to integrate the
strengths of each. In recent
years, the debate has tak-
en on a new dimension, in
the light of the evidence ac-
cumulated by experimental
psychology about the na-
ture of the reading process
(§34).
• Phonic approaches are
based on the principle of
identifying the regular
sound-letter relationships
in a writing system, and
teaching the child to use
to construct or de-
code words. Phonic
schemes have attracted a
great deal of support,
mainly because of the
they give children a ratio-
nale for 'sounding out' new
words. On the other hand,
they have also been
severely criticized. The
at- child's phonetic awareness
is often not up to the task
of phonic decoding. The
task of blending isolated
sounds into whole words is
not easy: to get from c [ka]
+ a [a] + r[ta] to cat [kat],
an actual change of pro-
nunciation is involved, as
the phonetic transcription
shows. And first books
have severe restrictions on
their permitted vocabulary,
which often results in arti-
ficial or bizarre sentences
(e.g. Pat and Dad ran).
• Whole-word or look-and-
say approaches are based
on the principle of recog-
nizing individual words as
way wholes, without breaking
them down into constituent
letters orsounds. The main
aim is to avoid the use of
strings of meaningless
phonic syllables, and to
permit access to longer
and more meaningful sen-
tences, through the use of
frequently occurring words
(the, go, saw, little, my,
etc.) - and even much
longer words, such as
aeroplane and doctor.
Whole-word approaches
have been criticized for
their lack of clear grading
principles, and for the way
words are often arbitrarily
selected, unrelated to the
child's experience.

44 LANGUAGE DEVELOPMENT IN SCHOOL 251

 . -

Line breaks Ready to read?

There is much that is not tences was much easier to as a control. Groups of chil-
known about the factors learn than the second: dren read the stories aloud,
that promote and hinder and their performance was
the process of reading in The very old man measured in terms of non-
the young child. Even basic was always sitting down fluency, reading errors, and (a) Circle each word.
typographical questions re- on one of the big chairs. comprehension.
main unresolved, such as The results, although
The very old
the optimum size of type or limited to given construc-
man was always sitting
the distance between lines tions, were clear. Breaks
down on one of the big chairs.
in the first texts. One such within grammatical phrases
(b) Circle each capital letter.
question relates to the ne- The first set of line breaks caused many more pro-
cessity of line division. Is a occurs at major grammati- blems than breaks be-
child's reading ability af- cal boundaries in the sen- tween phrases. Also, and
fected by the way in which tence, whereas the second at the beginning of a line
lines of text end? A com- set does not. Would this caused more trouble than
pletely arbitrary set of line kind of factor affect the when it appeared at the (c) Circle each thing that is a sentence.
breaks cause
is bound to reading ability of young end of the previous line.
some difficulty - hyphena- children? The small sample size
tion, for example, would To obtain some evidence makes these findings ten- Sandy. Bruce and James ran home
hardly be a help to young on this point, a recent study tative, but they support the
readers! But would their printed a story in several conclusion that line breaks each stop.
(d) Circle full
performance be facilitated different ways, so that the are a possible hazard in
if lines were made to end breaks were altered.
line A early reading - a matter
according to certain linguis- sample is given below: it which should be borne in A series of tasks from one section of the larr Test
tic principles - for example, shows the placement of mind when designing texts (LinguisticAwareness in Reading Readiness, 1 983),
following a major semantic and changed on three oc- for young readers or for which investigates young children's awareness of the
or syntactic boundary? casions, and the changed those with a reading handi- features and functions of written language. The main
In a study of adult recall, once. A text using justified cap (p. 272). (From B. Ra- aim is to see whether children have grasped the
it was found that the first of right-hand margin setting ban, 1981.) concepts that are required in order to recognize and
the following two sen- was also prepared talk about literacy skills. There is an important
(p. 1 83)
connection between children's awareness of the
technical 'metalanguage' of literacy and their
subsequent performance in reading and writing.
Mr. West lived in a house with a monkey. Mr. West lived in a house with a monkey.
That there are 'errors' in this domain is readily
One day, the monkey got up first. One day, the monkey got up first.
illustrated from the replies some children give when
He got up before Mr. West and He got up before Mr. West asked about the form and function of language.
before the sun. and before the sun.
• Tell me a long word', an investigator asks some 3-
He opened the window. The monkey He opened the window. The year-olds. 'Train,' replies one. 'Giraffe,' replies another.
had a look down the street. monkey had a look down the street. Both are thinking of the length of the objects.
There was no one about. There was no one about.
• 'Show me who's reading,' says an investigator,
"Mr. West is in bed and "Mr. West is in bed
pointing to a set of pictures in which different people
he is sleeping," he said. and he is sleeping," he said. are reading, writing, drawing, and looking at
"I'm going down the street "I'm going down the street and photographs. The child points to the photographs.
and into the park into the park.
Other kinds of difficulties are known: many children
Mr. West can't stop me." Mr. West can't stop me." have problems in naming letters, or saying whether
sequences of sounds are the same', or correlating
short words in speech. There is little doubt that children
need to have developed control over several
metalinguistic notions before they are 'ready' to read
and write.
Readability formulae
However, the notion of readiness' has been
Over 50 procedures have of words containing three is no neat correlation be- somewhat controversial in recent years, and must be
been devised that claim to or more syllables and di- tween sentence length and interpreted with caution. There is no single criterion for
be able to compute how vide this total by the difficulty (p. 233); and not saying that a child is 'ready' to read. Several psycho-
difficult a text is to read. number of 1 00-word sam- all long words are difficult to logical factors are involved, such as concept
The 'Fog Index' (1952), for ples. read. Factors such as the development, memory, attention, intelligence, and left-
example, is arrived at in complexity of sentence right orientation. There should be several linguistic
four steps:
4. Add
the results of 2 and
construction and the nature skills present, such as good sound discrimination, the
3, and multiply the total by
of word meaning are far ability to follow instructions, a developing spoken
1 Select several 1 00-word 0.4. The product is the
more important, but these language, as well as an ability to talk about language.
samples from a text. (American) grade level for
the procedures usually ig- Reading seems to presuppose a great deal.
which the text is appro-
2. Calculate the average
priate, in terms of difficulty.
nore. Readability formulae But the exact requirements are by no means clear
sentence length by dividing have thus attracted a great as is shown by parents who have successfully taught
the number of words by the Several such formulae deal of criticism, but in the their baby to read (usually through the use of words
number of sentences. (In- have been proposed, of absence of more sophisti- on cards, which are simultaneously shown and spoken
clude only complete sen- varying levels of complex- cated measures, they con- for regular periods each day) - in some cases, starting
tences.) ity. Most assume that diffi- tinue to attract widespread as early as the first year. In such cases, the idea of
culty can be measured use, as a reasonably con- a 'natural' stage of reading readiness becomes
3. Obtain the percentage simply in terms of the venient way of predicting obscure.
of long words
the entire in words and/or
length of (though not explaining)
sample: count the number (J. Downing era/., 1983, Manual, pp. 12-13)
sentences. However, there reading difficulty.

252 VII CHILD

•
LANGUAGE ACQUISITION
Learning to write
Poor handwriting?
A MOTOR SKILL One of the most wide-
For many people, learning to write is primarily a spread misconceptions is

matter of acquiring the motor skill of forming and that poor handwriting in
older children or adults in-
sequencing letters in a fluent, automatic manner,
dicates a careless or other-
and positioning them clearly on a page. But this wise inadequate persona-
involves far more than the correct formation of lity, or perhaps low

letter shapes: letter sizes, word spaces, spaces intelligence (p. 189). The
view has no basis in fact.
between lines, margins, and other matters of layout
Poor writing may be little
also need to be consistent, if a writing style is to
more than the reflex of a
be acceptable. These skills do not always come busy or rushed life-style. It
easily, and it is therefore necessary to determine may even relate to high in-

the factors that promote or hinder the development telligence, where the writer
is having difficulty keeping
of efficient handwriting. There have been few scien-
up with his thoughts.
tific studies, so that claims about the 'best' kind Even in young children, it

of handwriting to teach, and the 'best' way to teach would not do to dismiss a
it, tend to be impressionistic, subjective, and con- piece of work simply be-

troversial.
cause it looked messy. <z(7 T Her
First impressions of the ex-
Of the many issues raised, most attention has tract (right) cannot be
been paid to the question of 'writing posture' — good, but in fact it is part of
the optimum position of the body for writing. It an extremely impressive
hand linguistic performance - a
includes a consideration of such factors as
34-page epic narrative in
position, finger grip, the angle of the body towards
the paper, and the height of the writer's chair. Too
eight chapters,
events and characters,
full of
-to' feuy
low a
position,
chair, for
which
example, can cause a twisted hand
inhibits finger
prevents the formation of a free cursive (p. 186)
movement, and thus
written by Patrick, aged 5i
(From D. Mackay &
Simo, 1976.)
J. W (*- h
k/ rye.
style. In addition, simple management strategies CHAPTER SEVEN - IT WaS NOT.
need to be taught — such as the need to move the ROY TOLD MaRBELAR TO STOP
TH6 AIRCRAFT a MINIT6 SO TH6T
writing paper upwards as one nears the bottom
of a page (rather than to move oneself, which is
HE CUD6 GET HIS GUN, TO PLAY
WITH IN TH6 PLANE
pu./^VC
what some children do). The type of writing im-
plement and the kind of paper need to be con-
sidered - a child may be unwilling to write with
a certain kind of pen, or find it difficult. The ques- A month's progress
tion of when to introduce lined paper needs careful Although it takes a long (a) Several letter shapes been a notable advance in
thought: lines help the child to control the direction time to control all the fea- are made well, but there is g)\ word spaces are well
and size of script, but they also constrain the spon- tures of handwriting, a great deal of size varia- used. Lines are steadier,
monthly samples of a tion (e, u, some con- but there is still need for im-
taneity of a natural writing style. There may also
f),

child's work will show se- trasts are not clearly provement in the use of line
be difficulty in transferring letter shapes from one veral signs of progress. formed (g, u, c), word and
It is spaces, and in the relative
visual plane (e.g. on a blackboard) to another (the not difficult to see which of line spaces are erratic, and positions of letters on the
page). And there may be problems of coordination these two samples, taken a the lines are not straight. line (e.g. in going, the se-

between eye and hand movements, especially if month apart from a 5-year- cond g is written on a level
old girl, is the more ad- (b) Letters are of a more with the n, instead of des-
there has been little experience of scribbling and
vanced. consistent size, and are cending below it).
drawing. It is easy to see why it can take children formed (there has
better (From D. Mackay & J.
three years or more to develop a reasonably Simo, 1976.)
smooth, automatic writing technique.
e n
n (jrv^ K(? C
\^ r V^°&C\Ko* IW°
tt^te^^
Of)O © "P \ \ ft
-t° N ^h|e
mummy went to the clinic instead of to
school this morning.

Holding a pencil
The normal 'tripod' grip (top), widely
recommended for everyday use,
tb's
contrasted with a less efficient grip
(bottom). Children often use even
tighter grips - holding the pencil in
a clenched fist, for example. This
usually results in too much pressure we are having a new car this afternoon and
on the paper, and an erratic writing my grandma is going to have a ride in it.

style. It is also unnecessarily tiring.

44 LANGUAGE DEVELOPMENT IN SCHOOL 253

 '

A FUNCTIONAL SKILL • Children may address themselves, as in diaries,

There is far more to writing than the automatic notes, and first drafts.

exercise of a motor skill — a point

that has been • They may address their peers, as in writing an
emphasized in recent years, as researchers have account of an event for their class, or writing
begun to study the development of writing in rela- a letter to a friend.
tion to children's emerging cognitive, social, and • They may address a trusted adult, using a very
linguistic abilities, and to the demands being made personal style of writing.
upon them by the curriculum. The unique role of • They may address their teacher, seen as a
written language as a means of enabling children partner in dialogue, in the expectation that they
to formulate their thoughts to themselves, and to will receive help.
reflect on what they mean, is now widely recog- • They may address examiners, whether in rou-
nized. Writing is seen as an integral part of the tine class assessments or in formal examina-
process of learning, and not simply as an ancillary tions.
function — something to be used as a way of check- • They may address an unknown audience, as
ing that learning has actually taken place (as in when they have to produce work for a public
the traditional subject essay). This view requires occasion, or write a letter of application for
Writing blocks
a more sophisticated account of the nature of the a job.
Allchildren (and adults) ex-
writing process than has traditionally been avail-
In the Martin, et al. study, it was found that perience writing blocks at
able — in particular, it requires an appreciation that some time or other. Blocks
half of all school writing had the examiner in mind
writing is used for a wide range of purposes and arise for a variety of reasons:
— writing seemed to be used more as a means of
a variety of audiences (p. 212). the writer may be unhappy
testing than as a means of learning. In many set- about embarking on the writ-
tings, it was not being seen as part of the learning ing task, be unsure how to

Why write? process, but as something that happened after express something, or
learning was supposed to have taken place. Most simply not know what to say
The purpose of writing should never be taken for - the 'mind has gone blank'.
granted. One 5-year-old, returning from a nature of the other audience experiences were conspicuous
In such cases, children need

ramble with his class, was asked by his teacher by their absence. The research stressed the impor- to be helped over the block.

to 'write about it'. 'Why?' he replied. 'It's easier tance of giving children the opportunity of writing The 'Make-a-story' chart il-
for a wide range of audiences, in view of the lustrates one way of helping
to tell you!' Adults tend to forget that the 'obvious' children who are having diffi-
reasons for a community's use of writing (§31), demands that would be placed upon them once
culty with the development of
may be quite obscure to the young child. they had left school, and pointed to the need to a story. (From H. Cowie &
When the point is investigated, it quickly be- develop a balanced writing curriculum. H.Hanrott, 1984.)

comes apparent that writing is used for an indefi-

nitely large number of purposes - to express
feelings, tell stories, report events, complete forms,
keep records, and much more (§4). Children have
to learn about these purposes, and how the func-
tional differences affect the nature of the language
that is used. Several simple classifications of writing
styles have been made, as a means of describing
the nature and development of children's writing °a
J?"» / X— / «*
L°*°
/ -" "Ofhl
a °<*
in school, and in order to give guidelines to teachers /

anxious to develop a balanced writing curriculum.

One approach main styles: an
distinguishes three
and content to the
'expressive' style, close in style
everyday use of speech, which focusses on the
writer's personal feelings; a 'transactional' style,
which focusses on reasoned, logical statement; and £7
a 'poetic' style, which presents the reader with an
imaginative experience. When this system is used
to analyse the nature of traditional writing of older
schoolchildren, it emerges that most of this writing
is transactional, with expressive writing hardly

being used at all. (After N. Martin etai, 1976.)

Who is the child writing for?

The style and content of written language is much
affected by the nature of the recipient (§31), and
an important goal in working with children is thus
to develop their 'sense of audience'. Several pos-
sible kinds of audience for child writing have been
F?-fT7'
^
identified.
'
/^
254 •
VII CHILD LANGUAGE ACQUISITION
New approaches
Based upon such findings, several ways of fostering
children's abilityand enjoyment of writing have
been suggested. New writing programmes encour-
age teachers to provide a variety of real audiences
and functions for their pupils' work, so that chil-
4 ~"
dren can see that their writing has a genuine pur- 0U4OM- It hMA C^vrvrnl^a cttAtX ***- cC*uX . /* *>«* futtnA a
pose, and that it is not being done solely to be
'marked'. In addition to essays and experimental
reports, there are now increasing opportunities to
write in other styles for other audiences — such as
magazine articles, or letters to the press. Many
teachers have begun to keep a 'chart' of different
kinds of writing produced by their pupils. And
*
more attention is now paid to discussing samples
of writing with the children, both in groups and
individually. It has long been appreciated that writ-
ing arises out of talk, in the early years; perhaps
the most important aspect of current thinking is
the realization that the reverse process is just as
important — to give children the opportunity to
talk about what they write. 4
A LINGUISTIC SKILL
In addition to motor ability and functional aware-
ness, young writers need to develop the ability to tJhcthif £ T^W JcaIu jUtoi
use the structures of language in an appropriate
and mature manner. This ability takes several years Marking conventions the strengths of different errors, he must assume
to emerge. There have been few detailed longitudi- Recognizing the kinds of writing, and to that the teacher is giving
importance of writing has provide the best kind of false praise.
nal (p. 229) studies, so that analyses in terms of
many consequences for feedback. One of the In recent years there has
developmental stages tend to be very general or the organization and problems of traditional been a great deal of
anecdotal; but a number of initial distinctions have practice of teaching. It is by marking practice is discussion educational
in

been proposed. One scheme recognizes four stages no means easy to provide illustrated in this example circles about the need to
children with a range of ofa story written by a develop more balanced
of development (after B. M. Kroll, 1981):
audiences and purposes in young secondary-school and comprehensive kinds
1 A preparatory stage, when basic motor skills the institutionalized setting boy. The 20 corrections of assessment, in which
of a school. Links need to give a general impression interaction with the pupil
develop, and the principles of the spelling system
be made with the outside of failure, yet the mark plays a major part. But
are acquired. community (e.g. local given is 70% and the main many theoretical and
2 A consolidation stage, usually from around the government, press, or comment is well done'. practical problems have to
seventh year, when children begin to use the writing employers). But creating Whichever way the writer be faced before alternative
system to express what they can already say in such links is a slow and interprets the marking, philosophies and strategies
time-consuming process. thereis a problem: if he of marking can be
speech. Writing at this stage closely reflects the pat-
More important, new notes the good mark, he successfully implemented
terns of the spoken language. There may be many ways of responding to must assume that the - not least, the constraints
colloquialisms, strings of clauses linked by and, children's work need to be errors cannot be very on the teacher's time.
unfinished sentences, and other features of the found, in order to recognize important; if he notes the (From P. Gannon, 1985.)
child's conversational experience.
3 A from around the ninth
differentiation stage,
year, which writing begins to diverge from
in
speech, and develops its own patterns and organi-
zation. Errors are common at first, as children learn
new standards, and experiment with new struc- and editing ways of obtaining the best
as essential
tures found in their reading. Their written work expression. From point of view, such activities
this
becomes fuller and more diverse, as they encounter as crossing out have to be seen not simply as 'mis-
the need to produce different kinds of writing for takes', to be criticized on grounds of haste or care-
different audiences and situations. lessness, but as an indispensable step in the search
It is at this point that children most need for the best expression of what children are trying
guidance about the structures and functions of to say.
written language. In particular, they must learn 4 The integration phase is found when writers
that writing aids thinking in ways that speech can- have such a good command of language that they
not perform. Writing is a medium where there is can vary their stylistic choices at will and develop
time to reflect, to re-think, to use language as a a personal 'voice' — something which is rare before
way of shaping thought (§§31, 34). They therefore the middle teenage years, and which, in a sense,
need to see the importance of drafting, revising, continues to develop throughout adult life.

44 LANGUAGE DEVELOPMENT IN SCHOOL • 255

 2 *r 9
•
^c

X cm O
-funny
r* ,
PARTVm
Language, brain, and handicap
Ithas often been remarked that we come to appre-
ciate the unique complexity and function of lan-
guage only when it starts to go wrong. This
happens daily in many small ways, when we detect
ambiguity, express ourselves incoherently, or speak
at cross-purposes. Less commonly, it happens in
a dramatic and devastating manner, in the form
of language handicap. Those who find it difficult
or impossible to communicate, on account of some
physical, psychological, or other disability, face a
frustrating, isolated, and uncertain future, in which
their handicap is often not recognized, and commu-
nity support services may be inaccessible or absent.
Drawing attention to the existence and extent of
language handicap is thus an important role for
any encyclopedia of language.
Most language handicaps involve a consider-
ation of areas of the brain that may be impaired.
Little is in fact known about the way the brain
controls and processes language, and this provides
the focus for a great deal of contemporary research,
which reviewed in the first section of this Part.
is
We begin with a general account of brain structure
and function, and look in particular at the main
ideas about hemispheric dominance and localiza-
A physically handicapped child using a head pointer with a
communication board.
tion of function that have influenced thinking
about language in the past 100 years. We then con-
sider recent developments in the neurology of lan-
guage, and refer to some of the fruitful ways in
which we can make deductions about brain func-
tion from the study of human speech behaviour.
The section on language handicap opens with
a general review of issues relating to incidence, cau-
sation, and classification, and then looks at each
major category of handicap individually. We begin
with deafness, paying particular attention to recent
technological advances in this field. This is fol-
lowed by a discussion of the various handicaps of
spoken and written language that can accompany
damage to the language-processing areas of the
brain — aphasia, dyslexia, and dysgraphia. The next
sections consider the main kinds of 'output' pro-
blems - disorders of voice quality, articulation, and
fluency — and the notion of language 'delay'. Part
vin then concludes with an examination of alter-
native communication systems and the rapidly
developing field of communication aids, which
involve the latest advances in information tech-
nology.
45 Language and the brain

The human brain consists of several anatomically
distinct regions. The largest part is the cerebrum,
which is divided into two great lobes of similar
size - the left and right cerebral hemispheres. The
hemispheres are connected to the spinal cord by
the brain stem, which consists of the mid-brain,
the pons, and the medulla oblongata. At the back
of the pons is the cerebellum, which is responsible
for the maintenance of body posture and the
smooth coordination of all movements.
Most research has focussed on the structure and
function of the cerebrum, especially on its surface
information to the right hemisphere, and vice-
versa. Such sophisticated 'wiring', it has been sug-
gested, enables us to make many more qualitative
judgments about sounds and images (e.g. about
their distance and location) than might otherwise
be possible.
DOMINANCE
The functional relationship between the brain's
two hemispheres has for over a century been a
major focus of research in neuropsychology' and
clinical neurology. For some time, it was thought

layer of grey matter (nerve cells), the cerebral cor- that one hemisphere (the left, in most people) was
tex, which is the area primarily involved in the superior to the other in the control of most activi-
control of voluntary movement and intellectual ties. Today, it is recognized that each has its own

functions, and in the decoding of information from role, being more involved in the performance of Below left: A section through
the brain, showing the main
the senses. Beneath the cortex is a body of white some activities and less involved in others. A hemis-
anatomical areas
matter (fibre tracts), which transmits signals be- phere thus said to be the 'dominant' or 'leading'
is
Below: The brain seen from
tween the different parts of each hemisphere, and one for certain mental functions. The development
(top)above and (bottom) the
between the cortex and the brain stem. A notable of these functions within one or the other hemis- side,showing the cerebral
feature is that the surface of the cortex is not phere is known as 'lateralization'. hemispheres and the four
smooth, but has folded in on itself, to produce a Language and handedness have long been the main lobes

series of convolutions, or gyri, which are separated two major factors in any discussion of cerebral (a)
by fissures, or sulci. dominance. The left hemisphere is dominant for Left cerebral Right cerebral
The figure below shows the main anatomical fea- language in most right-handed people (estimates hemisphere hemisphere
tures. Seen from above (a), the main feature is the are usually over 95%). This is most noticeable in Median Central
longitudinal Sulcus
median longitudinal fissure separating the hemis- cases of aphasia (p. 270), where damage to the
fissure (fissure of
pheres. does not extend the whole way through
It left side of the brain may cause both language Frontal >^^7T^N Rolando;
the cerebrum: lower down, the hemispheres are handicap and a right-sided paralysis. However, the lobes
joined by a thick bundle of nerve fibres, the corpus relationship is not a symmetrical one: it does not
callosum. This is the means whereby information automatically follow that the right hemisphere is
can be transmitted from one hemisphere to the dominant for language in left-handed people. Left-
other. Seen from the side (b), the main features handers are by no means a homogeneous group,
are the central sulcus (the fissure of Rolando) "and
the lateral sulcus (the Sylvian fissure), which are
used as criteria for dividing the brain into its four
main lobes: frontal, temporal, parietal, and occipi-
Corpus callosum
tal.
Occipital lobes
One other important anatomical fact needs to
be borne in mind when discussing brain functions: (b)
each hemisphere controls movement in and re- Central Sulcus
ceives sensory input from the opposite side of the (fissure of Rolando)

body. Many nerve fibres from the two hemispheres Parietal \ Frontal
cross each other as they descend through the brain lobe \ \ lobe

stem, so that the left hemisphere controls the move-

ment of the right side of the body, and vice versa.
That is why brain damage to one hemisphere is
usually correlated with bodily effects (such as para-
lysis) on the opposite side. In the case of the ears,
signals from each ear go to both hemispheres, but
most information is transmitted to the opposite Brain
stem
side — a fact that has led to an important technique Temporal
,obe Lateral
for investigating brain function (p. 259). In the case Occipita
sulcus
of the eyes, the situation is yet more complex: the lobe
\ (Sylvian
left half of the visual field of each eye transmits Spinal cord
Cerebellum fissure)

258 • VIII LANGUAGE. BRAIN, AND HANDICAP

 Finding out about dominance
Brain abnormalities ies (the Wada' technique); their activities. The effect on changes in metabolic rate nals, such as music and
• The traditional approach, this paralyses a hemisphere language use is particularly within a hemisphere. In par- environmental sounds.
used since the first syste- for 2-3 minutes, during dramatic. For example, increases in cortical
ticular,
matic work on aphasia which time some of the when an object is presented blood flow can be measured Problems
(p. 270) in the mid-1 9th cen- patient's language or other to the right visual half-field, through radioactive tracers. These approaches present
tury, is to identify the loca- abilities can be tested. (right-handed) patients can There are also ways of many problems of principle
tion of a specific area of • A great deal of informa- talk about it: the visual infor- showing different kinds of and method. Although the
brain damage (a 'lesion'), tion has been obtained from mation is relayed to the left chemical action and tem- imaging techniques now
and see whether behaviour the results of surgery which hemisphere, where speech perature changes (p. 260). used in clinical neurology
has been affected in any may be carried out in cases processing takes place. • An area of the normal are a great step forward,
predictable way. of severe epilepsy. The However, if the same object brain can be electrically sti- they still convey very limited
• It is sometimes necess- hemispheres are separated is presented to the left vis- mulated, to see what effect information about brain
ary to remove the whole at the corpus callosum ual half-field, patients will be this has on behaviour. This function. It is not always
area of a diseased
cortical ('commisurotomy'). This en- unable to talk about it, even approach has mainly been possible to be precise about
hemisphere (hemispherec- ables the role of each side though they have seen it: used to establish areas of the location of a lesion,
tomy). The effects on be- of the brain to be studied in- the visual information has motor and sensory control which in any case is not
haviour can then be ob- dependently -the 'split gone to the right hemis- (P- 260). usually a neatly defined
served. brain' experiments. phere, where no speech area. Moreover, there is al-
• It is possible to observe These studies have processing takes place. Experiments ways a problem in genera-
what happens to behaviour shown that there are no ma- The role of each hemis- lizing from the performance
while one hemisphere is jor changes in intellect, per- Monitoring phere in processing a stimu- of a diseased brain to that
temporarily anaesthetized. sonality, or everyday beha- • An established approach lus can be inferred from the of a normal brain.
The usual technique, often viour following the (electroencephalography different time it takes a per- Studies using healthy
used before brain surgery, operation, but deficits are (EEG)) uses electrodes son to react when stimuli brains also pose difficulties.
is to inject sodium-amytal found the ability of the two placed on the surface of the are presented to each side
in There are problems with
into one of the carotid arter- hemispheres to integrate scalp to monitor continuous of the body. In a 'dichotic lis- reaction-time experiments,
cortical electrical activity - tening' task, for example, where it is necessary to
LEFT SIDE RIGHT SIDE in particular, the amount of different stimuli are simulta- match competing stimuli in
rhythm in the brain
'alpha' neously presented to each a very accurate way, and to
waves, which is reduced ear, and the subject has to control variations in sub-
when an area of the brain is report what is heard. When responses (due to
jects'
in active use. the signals to one ear prove such factors as shifting at-
• A related technique (the to be more accurately or ra- tention). Also, the detailed
Auditory Auditory 'averaged evoked res- pidly reported, con- it is data provided by monitoring
stimuli stimuli
ponse') uses electrodes to cluded that the opposite devices (such as EEG) can
monitor the activity in an hemisphere is more in- as yet be given only a very
area of the brain in res- volved in their processing. general interpretation. How-
ponse to repeated presen- In this way, for instance, a ever, technical progress,
Non-linguistic stimuli (e.g. coughs)
a stimulus.
tations of general right-ear advantage and associated computatio-
^^-^— Linguistic stimuli (e.g. words) • Techniques are also now has been shown for linguis- nal analysis, suggest a pro-
The routes between ears and hemispheres, as shown by available to monitor neuro- ticsignals, and a left-ear ad- mising future in this area.
dichotic listening tasks nal activity by observing vantage for nonverbal sig-

and in over 60% of cases the left hemisphere is orientation, creative sensibility, musical patterns,
either dominant for language or very much in- and emotional expression or recognition.
volved ('mixed' dominance). A pattern of mixed These identifications must be made cautiously,
dominance throughout the body (for example, a avoiding an oversimplified contrast — such as is
person may be right-handed, left-footed, and right- found when people talk about the left hemisphere
eyed) further complicates the investigation. as the 'analytic' or 'intellectual' part of the brain,
The specialized intellectual functions of each and the right as the 'creative' or 'emotional' part.
hemisphere, and their neurophysiological bases, It is now known, for example, that the right hemi-
are only partly understood. There are important sphere can handle certain nonverbal tasks that
anatomical asymmetries between the hemispheres require intellectual capacity (such as spatial judg-
(for example, there are differences in the length ment), and that there is a limited capability for
and orientation of the Sylvian fissure, and there auditory analysis and comprehension. Moreover,
is often a larger left temporal plane (part of the it must not be forgotten that there are several activi-

temporal lobe)); but it is unclear how these relate ties that usually involve both hemispheres (such
However, on the basis
to functional specializations. as face recognition, and the factors involved in
of various kinds of experimental and clinical evi- attention and fatigue) — a fact that is currently
dence, several generalizations have been made. attracting a great deal of research as scholars focus
With right-handed people, the left hemisphere is on the brain's integrating (rather than the lat-
found to be dominant in such activities as analytical As with the studies of localiza-
eralized) abilities.
tasks, categorization, calculation, logical organiza- tion 260), therefore, statements about the
(p.
tion, information sequencing, complex motor relationship between anatomical form and intellec-
functions, and language. The right is said to tual function, given our present state of knowledge,
be dominant for the perception and matching of must remain extremely tentative.
global patterns, part— whole relationships, spatial

45 LANGUAGE AND THE BRAIN 259

LOCALIZATION pite the many exceptions, in the vast majority of Mapping the cortex
The idea that a single area of the brain can be cases where linguistic symptoms have been the The task of mapping the ar-

related to a single behavioural ability, such as result of brain damage, the lesion around
is in or eas of the cortex involved in
vision or speech, is known as the theory of cerebral the areas originally identified by Broca and Wer- body activities was under-
taken by a team of neuro-
'localization'.Support for the theory came from nicke. It seems that there may well be primary areas
surgeons in Montreal during
the work of such neurologists as Paul Pierre Broca after all, but these have to be seen in the context the 1950s. Relationships
(1824-80) and Carl Wernicke (1848-1905), who of the brain as a whole, with other areas making were discovered by electri-
had found that damage to specific areas of the brain a contribution to language skills. Determining the cally stimulating different
parts of the exposed brain in
correlated with the loss of certain kinds of linguistic relationships between these areas is now a major
epileptic patients, in order to
ability in their patients (aphasia, p. 270). Damage goal of neurolinguistic research. find outwhich areas were in-
to 'Broca's area' resulted in a reduced ability to volved in seizures before
speak, though comprehension remained relatively proceeding to surgery. The
brain contains no pain recep-
unimpaired. Damage to 'Wernicke's area' resulted
tors, so the patients were not
in a reduced ability to comprehend speech, though
anaesthetized and could
the ability to speak was relatively unaffected. thus report their mental and
From the outset, the theory was hotly contested physical sensations (such as
by those who felt that other areas of the brain were a tingling sensation or a me-
mory of some event). Muscle
involved in language processing. Several kinds of
contractions at various
evidence were felt to go against a strict localization- points in the body could also
ist theory. Patients were found with apparently be observed, as could sud-
similar lesions, yet with very different linguistic den involuntary vocalizations
abilities; and, conversely, similar linguistic difficul-
and inabilities to speak.
The photograph shows the
ties could apparently result from lesions in widely
left hemisphere of a patient's
different areas. There are now cases on record of brain with the places marked
patients whose Broca's and Wernicke's areas were where responses were ob-
unaffected by lesions, but whose linguistic ability tained. Most of the effects on
language were disruptive.
was nonetheless seriously impaired; and conver-
Stimulation of points 23 and
sely, there are cases of patients who have even had
24 halted the patient's ability
Broca's area surgically removed in both hemis- to articulate or caused
pheres, and who were still able to speak. speech to be slurred. Symp-
New techniques of neuroimaging (p. 259) have toms similar to aphasia re-
sulted when points 26, 27,
brought to light many such counter-examples to
and 28 were stimulated.
the localization hypothesis. When we add to this (FromW.G.Penfield&L
the facts that other symptoms (especially of a psy- Roberts, 1959.)
chological kind) can result from so-called 'linguis-
tic' lesions, and that it has so far been impossible

to define specific brain areas in a precise way, we

can readily understand such comments as that
made by the British neurologist, John Hughlings
Jackson (1835-1911): 'to locate the damage which
destroys speech, and to locate speech, are two dif-
ferent things'. There may be many points within
the neuronal network that, if damaged, could have
the same effect on a person's linguistic processing
ability. It does not much matter whether a tele-
phone fault exists in the hand-set, along the line,
or in the telephone exchange: the resulting deathly
silence in the receiver is the same.
Cortical blood flow These computer-processed
There is now
doubt that several other areas
little images show the pattern of blood flow through the left
of the brain apart from the cortex are involved hemisphere. A tiny amount of a special chemical
133
in linguistic processing. Neurolinguists and neuro- ( Xe) is injected into the carotid artery, and the
psychologists postulate several kinds of subcortical
gamma radiation monitored. The distribution of blood
is related to the activity of the nerve cells in different
connection, as well as connections between the regions of the cortex. Each square corresponds to
hemispheres. The areas marginal to the classically 2
1 cm of cortex. The darker the square, the greater the

located ones are of particular interest, in this res-
pect; but research is also focussing on other parts
of the brain, such as other parts of the frontal lobes,
and the thalamus (p. 258). However, this direction
of research does not support a theory of opposite
extremes - that there is no localization at all in
the brain, and that every region is equally involved
in all activities (a theory of 'equipotentiality'). Des-
blood flow, and thus the greater the activity in that brain
area.
In (a), the subject is reading silently. Four areas of
the cortex are active: part of the visual area, the motor
area, the frontal eye field, and Broca's area. In (b), the
subject is reading aloud. Here, two further areas are
active: the mouth region of the motor and sensory
areas, on either side of the central fissure, and the
auditory cortex in the temporal lobe.

260 • VIII LANGUAGE, BRAIN, AND HANDICAP

 .

The homunculi Many of the

results of this research were
summarized in the shape of
two 'homunculi' - a human
form, drawn against the
shape of the outer surface of
the brain, in which the size of
the parts of the body is made
proportional to the extent of
the brain area involved in
their control. The large area
of the brain devoted to the
motor control of the vocal or-
gans and the hand is clearly
seen in the drawing of the
'motor homunculus' (right). A
drawing was made
similar to ^J I
- Lower lip

show the areas which re- 1^5 ' Teeth, gums, and iaw

ceive input from the senses

- the 'sensory homunculus'
(far right). Once again, the
large area devoted to the
hands and vocal tract should
be noted. (From W. G. Pen-
field &L Roberts, 1959.)
X

Central sulcus
Motor control
(of Rolando)
area

Lateral (Sylvian) Visual

sulcus cortex

Wernicke's
area

Language areas
The areas which have been
proposed for the processing
of speaking, listening, read-
ing, writing,and signing are
mainly located at or around
the Sylvian and Rolandic fis-
sures (p. 258). Several spe-
cific areas have been identi-
fied.
• The front part of the parie-
tallobe, along the fissure of
Rolando, is primarily in-
volved in the processing of
sensation, and may be con-
nected with the speech and
auditory areas at a deeper
level.
Neurolinguistic processing
• The area in front of the fis- • The lower back part of the Some of the neural pathways that are considered to
sure of Rolando is mainly in- frontal lobe is primarily in- be involved in the processing of spoken language.
volved in motor functioning, volved in the encoding of
and is thus relevant to the speech. This is 'Broca's 1. Speech production The basic structure of the
utterance thought to be generated in Wernicke's
is
study of speaking and writ- area'.
• Another area towards the area and is sent to Broca's area for encoding. The
ing.
• An area the upper back
in back of the frontal lobe may motor programme is then passed on to the adjacent
motor area, which governs the articulatory organs.
part of the temporal lobe, ex- be involved in the motor con-
tending upwards into the par- trol of writing. It is known as 2. Reading aloud The written form is first received
ietal lobe, plays a major part 'Exner's centre', after the by the visual cortex, then transmitted via the angular
in the comprehension of German neurologist Sig- gyrus to Wernicke's area, where it is thought to be
speech. This is 'Wernicke's mund Exner ( 1 846-1 926) associated with an auditory representation. The utter-
area'. • Part of the left parietal re- ance structure is then sent on to Broca's area, as in
• In the upper part of the gion, close to Wernicke's d).
temporal lobe is the main area, is involved with the
3. Speech comprehension The signals arrive in
area involved in auditory re- control of manual signing.
the auditory cortex from the ear (§25), and are trans-
ception, known as 'Heschl's • The area at the back of
ferred to the adjacent Wernicke's area, where they are
gyri', after the Austrian the occipital lobe is used
interpreted.
pathologist R. L. Heschl mainly for the processing of
(1824-81). visual input.

45 LANGUAGE AND THE BRAIN .

261
NEUROPSYCHOLOGICAL MODELS OF the body ('proprioceptive' feedback). Other kinds
LANGUAGE of internal monitoring, at 'higher' levels, may also
In real life, a snatch of dialogue {How are you}, take place. An analogous sequence of events would
Fine, thanks) takes place so quickly that it is easy be involved if the structure were to be written or
to forget the complexity of the neurological plan- signed.
ning and execution involved in the process. Any The nature of neurolinguistic programmes has
model of the production and comprehension of lan- attracted a great deal of research in recent years,
guage — whether spoken, written, or signed — especially in relation to speech production (§22).
involves several steps, each of which must have It is evident, for example, that the brain does not

some kind of neural representation. Neuropsycho-
logical models of language attempt to delineate
what these steps are and how they interrelate.
In speech production, for example, an initial
intention to communicate is followed (or perhaps
accompanied) by some kind of conceptualization
of the message. There has also to be a point at
which this conceptualization is encoded into the
semantic and syntactic structure of the language
used by the speaker (though it is not clear how
far this stage can be separated from the preceding
one). If the structure is to be spoken, it must first
be given some sort of phonological representation
(e.g. as syllables, phonemes, or distinctive features,
§28). A motor-control programme must then be
used, to coordinate the multiplicity of signals that
have to be sent to the appropriate muscles controll-
ing the different parts of the vocal tract (§27).
While this activity takes place, it is being constantly
self-monitored: feedback is being received from the
ear, from the sense of touch, and from the internal
sensations generated by the movement of parts of
issue motor commands one segment at a time. A
word such as soup is not neurologically transmitted
as three separate steps- [s] + [u] + [p]. The articu-
lation of [s] is lip-rounded, under the influence of
the following vowel, which shows that the brain
must be 'scanning ahead' while issuing commands
for particular segments (coarticulation, p. 156).
When we consider the whole range of factors that
affect the timing of speech events (such as breathing
rate, the movement and coordination of the articu-
lators, the onset of vocal-fold vibration, the loca-
tion of stress, and the placement and duration of
pauses), it is evident that a highly sophisticated con-
trol system must be employed, otherwise speech
would degenerate into an erratic, disorganized set
of noises. It is now recognized that many areas
of the brain are involved: in particular, the cerebel-
lum and thalamus are known to assist the cortex
in exercising this control (p. 258). But it is not
yet possible to construct a detailed model of neuro-
linguistic operation that takes all speech-produc-
tion variables into account.

A psycholinguistic model of SLIPS OF THE TONGUE - OR BRAIN?

speech production Tongue slips — involuntary departures from the
Intention to speaker's intended production of a sequence of lan-
communicate
— are very common. Sounds, syllables,
guage units
T
Conceptualization
morphemes (p. 90), words, and sometimes larger
units of grammar can be affected. Often, the

T deviant performance is immediately detected by the

speaker (though not always consciously) and cor-
Semantic - syntactic
encoding rected.
Several large collections of tongue-slip data have
+ now been made top right), and
(see facing page,
SPEECH the errors analysedfrom a variety of viewpoints.
PRODUCTION
Sigmund Freud, for example, saw tongue slips as William Archibald Spooner
f
Phonological Internal symptoms of unconscious forces or mental conflict (1844-1930) Spooner, an
representation monitoring Anglican clergyman and
within an individual, which needed careful psycho-
/ Warden of New College Ox-
/ analytic interpretation. They have also been seen ford, had a nervous manner
/
Motor control as providing insights into the mechanisms of lan- that led him to produce many
programme guage change and evolution (§54). But most recent slips of the tongue - typi-
cally, involving reversals
research has studied these errors to see what light
Proprioceptive which led to unintentional
feedback they throw on how the brain or mind works.
Neuromuscular comic effects. Several of the
events The main linguistic finding is that tongue slips spoonerisms' attributed to
are not random, but are by refer-
largely explicable him are famous, such as
i Tactile
ence to certain basic constraints. For example, the queer old dean' (for dear old
Articulatorv ^. feedback queen').
two words involved in a tongue slip (the word con-
events
4 taining the slip and the word that influences it)
/

1 '
Audirorv are often found within the same syntactic consti-
Acoustic ^ feedback tuent or intonation/rhythm unit (§§16, 29). More-
events
over, the influencing word is often the most
strongly stressed within the tone unit. And most

262 VIII LANGUAGE, BRAIN, AND HANDICAP


tongue slips involve the symmetrical substitution Tongue slips classified
within a syllable of one sound by another: for An enormous variety of ton- Consonant addition
example, an initial segment in the influencing word gue slips occurs in everyday optimal number—* moptimal
replaces the initial segment in the slipped word. speech. Here are examples
of some of the most frequent
Combining such constraints, it is possible to
categories found in a corpus
make predictions about the form tongue slips are of over 1 2,000 spontaneous
likely to take when
they occur. Given the intended slips (after V. A. Fromkin,
sentence 'The car missed the bike / but hit the wall' 1 973). The words affected
are given in the second col-
(where / marks an intonation/rhythm boundary,
umn. (For phonetics termino-
and the strongly stressed words are italicized), the logy, see §27.)
likely slips are going to include bar for car or wit
for hit. Most unlikely would be bar for car (show- Initial consonant anticipated
word a reading list —» leading
ing the influence of a less prominent in the
it's a real mystery -> meal
second tone unit) or lit for hit (showing a final
Initial consonant persevera-
consonant replacing an initial one). tion
Tongue slips tell us a great deal about the neuro- black boxes -» bloxes
psychological processes that underlie speech. The gave the boy —> goy
different kinds of errors provide indirect evidence
Consonant reversals
well made-* mell wade
for some of the stages recognized by models of
baked a cake -» caked a
speech production, and suggest the kinds of linguis- bake
tic unit that these models need to take into account. Final consonants
For example, if slips tend to occur inside rather with a brush -» wish
king, queen -> king, quing
than between intonation units, it suggests that this
Consonant deletion
unit has a neuropsychological reality within which below the glottis —> gottis
the events of articulation are serially organized and tumbled -» tubbled
integrated.
A critical period for language?
The notion of a 'critical
iod' was first used by etholo-
gists studying the origin of
species-specific behaviour.
was found that with certain
species (e.g. rats, goslings)
there were periods in which
a particular kind of stimulus
had to be present if the baby
Genie was to develop normal beha-
The tragic case bears directly on the critical
of 'Genie' viour.
period hypothesis. Genie was discovered in 1970, at The question was there-
the age of 1 3z, having been brought up in conditions fore raised whether there
of inhuman neglect and extreme isolation. She was were periods in hu-
critical
severely disturbed and underdeveloped, and had been man maturation also. The
unable to learn language. In the course of her American psycholinguist Eric
treatment and rehabilitation, great efforts were made Lenneberg (1921-75) ar-
to teach her to speak. She had received next to no gued that such a period ex-
linguistic stimulation between the ages of 2 and isted in the case of language
puberty, so the evidence of her language-learning acquisition. The develop-
ability would bear directly on the Lenneberg ment of language was said
to
hypothesis. be the result of brain matu-
Analysis of the way Genie developed her linguistic ration: the hemispheres were
skills showed several abnormalities, such as a marked equipotential at birth, with
gap between production and comprehension, language gradually becom-
variability in using rules, stereotyped speech, gaps in ing lateralized in the left
the acquisition of syntactic skills, and a generally hemisphere (p. 260). The
retarded rate of development. After various process began at around the
psycholinguistic tests, it was concluded that Genie was age of 2 and ended at pu-
using her right hemisphere for language (as well as berty, when the brain was
for several other activities), and that this mighthave fullydeveloped, and laterali-
been the result of herbeginning the task of language zation was complete. At this
learning after the critical period of left-hemisphere point, there was no longer
involvement. The case was thus thought to support any neural 'plasticity' which
Lenneberg's hypothesis, but only in a weak form. would enable the right
Genie was evidently able to acquire some language hemisphere to take over the
from exposure after puberty (she made great progress language function if the left
in vocabulary, for example), but she did not do so in hemisphere was damaged.
a normal way. (For other 'lost' children, see §49.) The argument in favour of
(After S.Curtiss, 1977.) a critical period was based
largely on claims about the
. . .
Stress changes
similarly—* similarily
kitchen sink—> kinchen paying for it —> pay taring it
Consonant movement Word reversals
—
pinch hit * pitch hint a tank of gas -> a gas of tank
bacon and eggs -> aeon and a job for his wife -> a wife for
begs his job
Consonant clusters Telescopic errors
heater switch — sweeter Nixon witness —> nitness
>
hitch parking permit-* parking pit
damage claim -» damage Derivational affixes (p. 90)
dame often -» oftenly
Consonant clusters divided flashing light—* flasher
stick in the mud —> smuck Blends
. .
tid person/people-* perple
->
fish grotto frish gotto draft/breeze -» dreeze
Vowels Word substitution
fill the pool -> fool the pill Idon't sleep very well in a
Bev and Bill -> Biv and Bell single bed -» speak very . . .
Vowel + r well . .
foolish argument-* farlish chamber music-* chamber
fightvery hard -> fart hide . . . maid
Single features Other grammatical errors
spell mother-* smell bother It looks as if —» look... I
bang the nail -+ mang the the day when was born -* I
mail the day where
. .
Errors within words
relevance -> revelance
whisper-* whipser
per- patterns of recovery in brain- ever, most of these skills are
damaged adults and chil- well established before the
dren. If adults with left- age of 5, and some linguistic
It hemisphere damage failed to semantics and prag-
skills (in
recover language within a matics) are still developing in
few months, it was argued, teenage children and young
they would never do so. Chil- adults.
dren, however, showed an The neuropsychological
ability to recover over a evidence generally fails to
longer period - and could support the Lenneberg hy-
make a complete recovery if pothesis, showing lateraliza-
they were very young at the tion to be established long
time of the damage. In such before puberty - some stu-
cases, even total removal of dies suggest this may even
the left hemisphere did not be as early as the third year.
preclude the reacquisition of Cerebral anatomical asym-
language. metries have been found at
Controversial evidence birth, and several functional
The critical-period hypothe- asymmetries have been
has been controversial.
sis noted in infants (e.g. a pre-
The pathological evidence is ference for rightward turning
mixed, because compari- and right-hand grasping).
sons of adult and child cases Certain dichotic listening ad-
are extremely difficult to vantages (p. 259) are also
make, and paths of recovery present from a very early
have not been studied in a age, including some related
detailed linguistic way. It may to speech perception.
be that aspects of child reco- On the other hand, laterali-
very are helped by the invol- zation plainly takes some
vement of the right hemis- years before it is firmly es-
phere; but there are also tablished, and
overlaps this
cases of left-hemisphere da- the main period of language
mage producing severe and acquisition in a way that is
long-lasting aphasia in chil- not yet understood. The rela-
dren. tionship between lateraliza-
The evidence of normal tion and language is thus an
language acquisition (Part extremely complex one, and
VII) is also mixed. Aspects of presents a continuing re-
phonological and grammati- search challenge in develop-
cal acquisition do continue mental neuropsychology and
until around puberty; how- neurolinguistics.
45 LANGUAGE AND THE BRAIN •
263
46 Language handicap

Language handicap refers to any systematic defi- there is no doubt that the cause of the linguistic Terminological
ciency in the way people speak, listen, read, write, handicap lies in a person's abnormal physical con- handicap
or sign that interferes with their ability to commu- dition. The field of language handi-
nicate with their peers. At one extreme, the handi- However, majority of cases, it is not pos-
in the cap is bedevilled by termino-
cap may be quite mild, such as a minor impediment sible to find a clear organic cause, given the present logical difficulties. The pro-
of pronunciation; at the other, there may be an state of medical knowledge. Thousands of children blem of labelling individual
handicaps will emerge at
almost total breakdown of all modes of communi- have a delayed language development, and in most
various places in this sec-
cation, as in severe forms of brain damage. In every of them there is nothing in their medical history tion; but even the most
case, we see language to some degree ceasing to that can account for the problem. There are many general of notions attracts a
function in a natural, spontaneous, and unselfcons- thousands of stutterers whose handicap, likewise, diversity of labels. Thus the

cious way, and drawing attention to itself, thus heading to this section, 'han-
cannot be explained in any simple physical way.
might have used dis-
dicap',
becoming a barrier rather than a means to commu- And a large number of people develop problems order, disability, defect, dis-
nication. in the use of their voice that have no physical expla- function, or impairment.
Because handicap exists in a continuum from nation. In such instances, we can search for 'func- There is little to choose be-
mild to severe, it is very difficult to obtain accurate tional' causes in a person's psychological, social, tween such terms in a gener-
al account.
estimates of its prevalence (the number of cases or linguistic background. A particular life-style, for
The professionals whose
in a population at any one time) or incidence (the example, may be the ultimate cause of a poor voice job it is to diagnose, assess,
number of new cases within a particular period). quality. A child's weak memory may explain a case and treat language handi-
A British government survey of the 1970s indicated of language delay. But in very many cases, even caps also vary in name from
country to country, and even
that about \% of the population were sufficiently these dimensions of enquiry result in no clear cause
within countries. In the U.S.,
seriously handicapped as to require the services of being discovered. they are usually known as
a speech therapist (see below), but accepted that Assessment of language handicap must also allow speech pathologists',
this figure was vague and probably far too low. for the fact that many conditions have multiple though 'language patholo-
If other categories of the population are included, causes. For example, the level of language achieve- gist' isalso quite wide-
spread. In Britain, the gener-
such as less seriously impaired people, or those who ment reached by a deaf child cannot be explained
al term is 'speech therapist'.
have an abnormal degree of difficulty with reading, solely with reference to the child's degree of hearing In Europe, the same tasks
writing, or spelling, the figure must approach loss: many other factors contribute - such as the are carried out by orthopho-
2-3%. And if a functional notion of handicap is child's personality and family background, and the nists' (e.g. in France), logo-
paedists' (e.g. in Germany),
used, to include the language problems faced by kind and amount of exposure to spoken or signed
and 'phoniatrists' (e.g. in
immigrants and other minority groups, the total language. Or again, an adult's voice disorder might Czechoslovakia). All at-
increases dramatically to perhaps as many as 5% begin as hoarseness arising out of a straightfor- tempts to standardize the
of the population. What is plain is that, however ward disease, such as laryngitis; but anxiety over name of the profession have
the problem is identified, several million people in so far failed.
The range of professional
the world suffer from an inability to communicate
skills and qualifications
that limits their personal development, their social found under these headings
relationships, and their effective contribution to also varies - reflecting the
society. The main aim of research in this area is Need for speech therapy fact that courses of study can
thus to understand the physical and linguistic basis range from just a few months

of language handicap, and to devise therapeutic

% Numbers to over four years. Some
needing needing countries make a sharp dis-
ways of alleviating the condition so that a handi- Total speech speech tinction between those who
capped person can achieve as full a life as possible. population therapy therapy specialize in language pro-
Adults blems and those who look
more general kinds
THE CAUSES OF LANGUAGE Geriatric patients
Hospitalized stroke
40,000 9 3,600 after
learning handicap. Some
of

HANDICAP patients 16,000 33 5,000 make a distinction between

In about 40% of cases, a language handicap can Others (eg stutterers, problems of speech and
be related to a clear physical cause. For example, voice disorders) 30,000 100 30,000 those of hearing. Some have
different kinds of specialists
many children are born with brain damage that Children
Preschool 2 million 3 60,000 to deal with adults and chil-
causes a degree of mental or physical handicap,
Ordinary school 9 million 2 180,000 dren.
and linguistic skills are usually seriously impaired Educationally sub-normal But one thing all countries
as a consequence. Deafness can have a crippling (moderate) 60,000 20 12,000 share: a lack of financial re-
impact on the normal development of spoken lan- Educationally sub-normal sources being devoted to the
(severe) 35,000 50 17,500 study and treatment of what
guage. Parts of the brain can be destroyed by illness,
Physically handicapped 12,000 25 3,000 is probably the most neg-
strokes, accidents, or acts of violence, to produce lected of all human handi-
the many forms of aphasia (p. 270). Various kinds TOTAL 311,100 caps.
of abnormal growth may affect the functioning of
Estimates of numbers needing speech therapy in Britain in
the vocal folds, or may lead to the larynx having the early 1970s. (Department of Education and Science,
to be surgically removed. In such cases as these, 1972.)

264 •
VIII LANGUAGE. BRAIN, AND HANDICAP
the continued use of the voice (if the person is a
singer, for example) might promote the develop-
ment of excessive strain while talking, with the
result that the hoarseness continues long after the
disease has disappeared. Similar combinations of
organic and functional causative factors underlie
most if not all language handicaps.
THE CLASSIFICATION OF LANGUAGE
HANDICAPS
Studies of language handicap traditionally
use of a number of basic two-way
a means of imposing an initial organization on this
complex field. Some of these distinctions provide
a useful introductory perspective, but they must
be used with great caution if their inevitable simpli-
fication is not to mislead. The common classifica-
There are however many kinds of handicap
where problems of production and reception are
simultaneously encountered. For example, aphasia
(p. 270) is often classified into 'expressive' vs
'receptive' types, whereas it is usual to find both
kinds of difficulty within the same patient, in vary-
ing proportions. Clinically, it is more accurate to
talk of a patient being 'predominantly' expressive
or receptive. And even the apparently 'straightfor-
ward' handicaps raise problems for the distinction.
make Deafness is patently a receptive handicap, but it
distinctions as often manifests itself in a highly disordered lan- The communication chain
guage production (p. 267). Cleft palate and stutter- This is a widely used model
ing are patently production handicaps, but in both for studying language handi-
cap. Different steps within
cases problems of reception can be found: a signifi-
the processes of production
cant proportion of cleft palate children have an and reception can be identi-
additional hearing loss(p. 277); and stutterers may fied, and used to plot the

tion of handicaps into 'organic' and 'functional' be so involved in the problems of fluency control primary 'location' of deficits,
as when handicaps are said
types (p. 264), for example, is well motivated in that their ability to listen and comprehend may
to be neurological ('central'
many cases, but ultimately breaks down when we deteriorate. In all these cases, the simultaneous
vs 'peripheral'), physiological
encounter handicaps in which both elements play involvement of problems of production and recep- (e.g. muscular), or anatomi-
an important defining role. And a similar problem tion warrants a more complex kind of analysis than cal (e.g. cleft palate). How-
faces anyone wishing to make use of the most the traditional binary account provides. Such ever, the model does not
show the importance of the
widely used primary classification of language handicaps are in fact more properly called 'syn-
various kinds of feedback
handicap: production vs reception. dromes', with several different elements contribut- (§§20,25), nor the psycho-
This way of looking at communicative break- ing to their identification. and other fac-
logical, social,

down, ultimately derived from information theory, can determine the

tors that
seriousness of a language
is referred to in several different ways, such as
handicap.
motor/sensory, encoding/decoding, or executive/
evaluative. A being drawn
basic distinction is

between handicaps of language production and PRODUCTION TRANSMISSION RECEPTION

handicaps of language reception. Production refers
to the whole sequence of neurological, physiologi-
cal, and anatomical steps required to encode a
_ j Neurological
linguistic message and make it ready for transmis-
sion (p. 262). Any disruption to the normal chain Neurological
of events would thus result in 'expressive' handi-
caps: a clear example would be a neurological com-
Physiological
plaint that led to slurred speech. Reception refers
to the sequence of anatomical, physiological, and Anatomical
neurological steps required to decode such a mes-
sage as it is being received. Disruptions here would
result in 'receptive' handicaps: the clearest case is
deafness.

marily handicaps of grammar municative handicap, such

Language vs speech and semantics (Part The
ill). as those due to hearing and Spoken Written Sign
The term language', as used latter refer to the 'non-sym- the visual modalities. Above language language language
in this section, is a cover aspects, that is, those
bolic' all, the distinction leaves un-

term for all modes of linguis- concerned only with the use clear the status of phonology
Phonetic Graphetic 'Cheretic'
tic communication - whether of sounds seen as a set of (§ 28), which is concerned
by speech, hearing, reading, meaningless phonetic enti- with both semantic and pho- handicaps (§27) handicaps (§32) handicaps (§36)
writing, or signing. (This, in- ties - as found in problems netic properties of language.
deed, is its use throughout of voice quality, fluency, and For example, it is not obvious Phonological Graphological 'Cherological'
the encyclopedia.) In the tra- articulation (Part iv). whether children who have handicaps (§28) handicaps (§33) handicaps (§36)
ditional study of handicap, Although this distinction is failed to master the phonolo-
however, it is used in a more still widely used, it has begun gical system of their lan-
restricted way. A distinction to attract criticism in recent
Grammatical handicaps
guage should be classified
is drawn between handicaps years. The term 'speech' is as displaying a 'speech dis-
of 'language' and handicaps ambiguous, as it is often order' or a language dis- Semantic handicaps
of 'speech'. The former refer used in the sense of 'spoken order'. Alternative systems of
only to the 'symbolic' aspects language', which includes classification are therefore A linguistic classification of communication handicaps,
ofcommunication, i.e. those grammar and meaning required to solve the pro- using the model of structural levels (§13). Pragmatic
concerned with the formula- (§ 31 ) Also, the focus on
. blems presented by the lan- handicaps are not shown. The terminology for signing
tion and structuring of mean- speech neglects the impor- guage/speech division. handicaps is not an established usage.
ing - in modern terms, pri- tance of other forms of com-

46 LANGUAGE HANDICAP • 265

Deafness brain. A case of sensorineural deafness is illustrated Who's who
in the audiogram below of a child who was unable Otologists are doctors who
About 1,000 children have a hearing loss that
1 in to respond to high frequency sounds such as [s] have specialized in diseases
is present at birth, or acquired soon after, caused ('high tone' deafness). of the ear.
Audiologists are clinicians
by pathology of the inner ear and its relationship Other forms of hearing impairment have been
who assess the nature and
to the auditory nerve (§25). Maternal rubella (Ger- identified, such as tinnitus — a range of noises in degree of hearing loss and
man measles), meningitis, and several other dis- the ear (ringing, hissing, pulsating, etc.) that can conservation, and who ad-
eases are known to be causative factors. Many occur in acute, debilitating form. Around 2 million vise on the rehabilitation of
more children have a hearing loss that they acquire people with hearing impair-
people have this problem, to some degree, in the
ment.
in the preschool or early school period, because USA. There is also the major problem of 'fluctuat-
of pathology of the middle ear. Several middle-ear ing' hearing loss, commonly related to persistent
problems get better without intervention; but middle-ear infections, which can seriously affect
others recur, become chronic,and do not respond young children's ability and motivation to attend
well to treatment. In such chronic cases, and in to speech sounds, especially in noisy environments,
all cases of inner-ear deafness, there can be serious and thus promote language delay and learning pro-
consequences for the development of speech com- blems in school. There is also the little-understood
prehension and production. central (or cortical) deafness, where there is loss
Many adults - perhaps as many as a third of of hearing sensitivity due to damage of the auditory
the population over 60 years of age - have an nerve in the brain stem or in the hearing centres
acquired hearing loss, which can noticeably affect of the cortex. In this last case, of course, there is
their ability to comprehend and speak (the latter, a problem of diagnosis, as if a person fails to re-
because they are unable to use hearing to monitor spond, despite normal peripheral hearing, other
what they are saying). Regular exposure to loud factors (such as aphasia or mental handicap) may Audiometric tests
noise (at work, in discos, etc.) is a common cause. be involved. Pure-tone audiometry
However, because language has been acquired A specially calibrated ma-
chine generates pure tones
before the onset of the deafness, these disorders
(§ 23) at different frequencies
are usually less serious. and intensities. The tones
The high incidence of the handicap is often not are presented to one ear at
appreciated. In 1975, it was estimated that there a time through headphones
(in an air-conduction test) or
were 170,000 children (under 16) and 2,360,000
through the bones behind
adults suffering from some degree of hearing loss the ear (in a bone-conduc-
in Britain. In theUSA, the figure is thought to be tion test). Any response to a
over Depending on the criterion of deaf-
1 6 million.
The audiogram sound is noted, and plotted
ness used, estimates vary from 2% to 10% of the This is the most widely used measure of hearing im- as an audiogram.
pairment. The horizontal dimension of the chart shows The results of pure tone
population. As many as 15 people in every 1,000
a range of sound frequencies up to 8,000 Hz; the audiometry do not always
have a hearing loss in one ear. vertical dimension shows a range of intensities of correlate clearly with a per-
hearing loss up to 1 20 decibels. A person's ability to son's ability to hear the com-
hear pure tones, presented at different frequencies and plex tones of speech. Two
TYPES OF DEAFNESS intensities, is plotted on the audiogram, and the curve people can have the same
There is no single, simple phenomenon of 'deaf- compared with the normal minimal audibility curve audiogram, yet display very
ness' but a wide range of kinds and degree of hear- (shown here as a straight line at 0). The sensitivity of different linguistic skills. It is

ing impairment. The loss may affect only one ear each ear is separately measured using headphones, therefore important to sup-

{unilateral) or both ears {bilateral). At one ex- and marked by different symbols ( o is here used for plement such findings with
the right ear, and x for the left). the results of alternative
treme, there may be a slight inability to hear a few
The audiogram below shows the responses of a 1 22- tests, using speech stimuli.
low-intensity frequencies (§23), which interferes year-old girl with a high-frequency loss in both ears,
only occasionally with normal communication; at the loss being more severe in the left ear. A sample Speech audiometry
the other extreme, a person may have no detectable of her free writing is given on p. 267. The ability to respond to the
sounds of speech often as-
is
response to any frequency, no matter how intense Frequency 125 250 500 1000 2000 4000 8000 Hz
sessed impressionistically.
the sound. The latter is uncommon: most deaf peo- i
with the audiologist speaking
ple have some degree of 'residual' hearing. But a at different levels and dis-
10
residual ability to hear amplified sound up to i
tances from the listener. A
20 more precise estimate can
500 Hz, for example, is of very limited value: at <

5 30 C* X i

be achieved by recording a
this level, people receive only about 10% of the
information conveyed by the speech waveform.
'8

.5
40
50
VN K x
series of words or sentences
which have been carefully
The main classification of hearing loss is based constructed so as to repre-
on where the interference lies in the auditory path- I
60 AP
rv X^i
i

i
sent the different kinds of

way. Conductive deafness arises when there is M ~0

\ speech sound, and playing

interference with the transmission of sound to the

inner ear, as when the middle ear becomes inflamed
[otitis media) or the ear drum or ossicles are affec-
|
X
100
80
90 — T- — T- — •^

" !S Q_± _
i these to the listener at known
intensities under various
conditions. The kind and le-
vel of background noise, the

110 characteristics of the room,

ted by disease or trauma. Sensorineural deafness
and even the position of the
arises when the source of interference lies within 120
x listener's head can affect the
the inner ear, or along the auditory nerve to the °Right Left
response.

266 • VIII LANGUAGE. BRAIN. AND HANDICAP

ORALISM vs MANUALISM sign. Studies have begun to show stages of develop- Deaf- and dumb?
The whether deaf children should be taught
issue of ment in the way
these 'native signers' learn to sign This widely used phrase is
to sign (§35) has been hotly debated for over a that can be related to the stages of language acqui- extremely misleading, and
century. There is strong opposition between those sition found in hearing children (Part vn). This should be banned. It does
not follow that, if someone is
educators who support the primary role of signing change of perspective is quite crucial: for these chil-
deaf, they will be unable to
fmanualists') and those who support the exclusive dren, signing is not a handicap but a natural means
speak. Many deaf people
teaching of speech ('oralists'). There are also many of expression quite comparable to the expressive achieve excellent levels of
who support a combined approach, in which potential of spoken language. For them, learning oral ability. Everything de-

speech and some kind of signing system are used to speak is the handicapping condition. pends on such factors as fa-
mily background, age of on-
simultaneously ('total communication') - an ap- Contemporary gradually coming to
society is
set of deafness, and the kind
proach that has become particularly influential in give signing the recognition it requires, and deaf of language education pro-
recent years. The arguments are complex, and often signers the opportunities they deserve — but it is gramme followed.
emotional, because they raise questions of the iden- an extremely slow process. In this encyclopedia,
tity of deaf people and the quality of their lives. the linguistic status of signing is symbolized by

The main argument against manualism is that treating itmajor section (Part vi), even though
as a
it is setting deaf people apart from all but their research has not yet advanced to a stage permitting
own small community, labelling them as 'deaf and a detailed treatment comparable to the sections on
'different', and making it difficult to communicate speaking and writing. In social, educational, and
with the hearing world. The main argument against political life, too, progress is being made, with the
oralism is that its methods are often unsuccessful, image of the deaf signer as a handicapped person
with the deaf person becoming just as isolated, who displays bizarre behaviour slowly being
being left with speech that is limited and difficult eroded. In the USA, in particular, there is a strong
to understand. movement to obtain recognition for the needs of
In support of manualism, it is argued that signing the deaf, such as by providing interpreters in
enables a deaf person to enjoy a wide range of university classes, local government meetings, and
communicative experience in the social and crea- television programmes.
tive life of the deaf community. In support of oral-
ism is the evidence that some methods have proved Written language
to be very successful, especially if a natural speak- The following two samples atlow frequencies, and no guinea pig got pink ears.
ing environment is part of a child's early upbring- of the free writing of deaf response at all at higher The guinea pig got four leg.
ing. children have been chosen frequencies. The highly

However, there are schools of thought within to illustrate the linguistic deviant syntax results in a B
which can de-
difficulties very low level of intelligi- The Star Wars was the two
both approaches, which themselves give rise to
velop in deaf children's bility. spaceship a fighting
fierce argument. In relation to oral approaches, writing. The first (A) is from opened door was coming
there is controversy over the kind of oral language the 1 2|-year-old whose the Men and Storm trooper
to teach the child — whether it should be natural audiogram is given on There is a guinea pig. The guns carry on to Artoo De-
p.266. The story shows a guinea pig name is Funny. too and threepio at go the
conversation or a simplified input. Several 'struc-
very limited range of gram- The guinea pig got black space. The Earth was not
tured' approaches are available under the latter matical ability, with stereo- and white. The guinea pig grass and tree but to the
heading. In relation to manual approaches, some typed sentence openings. got pink nose. The guinea sand, R2D2 and C3PO at
educators think that a 'pure' sign language, unin- The second (B) was written pig is standing. The guinea going look for R2D2 walk
fluenced by speech, should be learned first, with by a 1 6i-year-old boy with pig is waiting the food. The the sand people carry away
a loss of 90 dB in both ears Artoo Detoo sleep.
other varieties coming later. Others think that this
approach is sensible only for the small number of
deaf children born to deaf parents. In cases where
the parents have normal hearing, they argue,
spoken language will be in use, and so a signing
system related to speech might as well be used from
the outset.
These controversies have been around, in some
form, for over a century, promoted largely by the
lack of objective data about the way deaf children
learn. It is likely that, as research findings accumu-
late, some of these issues will be resolved. But there
is little sign of this at present.

DEAF SIGNING - A NEW

PERSPECTIVE
A deaf or hearing child born to deaf parents learns
sign language as a 'mother tongue', producing a
level of manual awareness and sophistication that Gallaudet College, Washington, U.S.
is from the deaf children of hearing par-
different A unique college of higher education in the English-speaking world, Gallaudet provides
ents or from hearing people who have learned to courses for the adult deaf.

46 LANGUAGE HANDICAP •
267
TECHNOLOGICAL ADVANCES all the alternatives, and arrive at the one that is Auditory implants
There has been a considerable advance in the design best suited to an individual's hearing problem and Insome cases of almost total
of communication aids for the deaf in recent years, life-style. Nor
choosing an aid a once-and-for-all
is deafness, thedamage
decision: technical advances are being made all the seems to be located in the
following progress in acoustics and instrumental
hair cells of the cochlea, with
phonetic research (§§23-4). Several approaches time, so that easy for a type of aid to become
it is
the auditory nerve to the
have been investigated in relation to the main sen- out of date. Also important is the need to maintain brain remaining intact. In
sory modalities. There have been experiments with the aid so that it performs at peak efficiency — some- such cases, it proves pos-
vibro-tactile aids, for example, which represent thing that may seem obvious, but that is often disre- sible to provide a sensation
garded, especially by old people. of hearing by direct electrical
speech frequencies by means of spatial vibration stimulation of the nerve end-
patterns applied to the fingers — a kind of audible ings, using tiny wires that are
Visual aids
braille (p. 280). But most technological aids fall surgically implanted in or

into the two main categories of auditory ('hearing The idea of a device that would make all the sounds near the cochlea, and that
of speech immediately visible to the deaf over pick up signals transmitted
aids') and visual. is
viaa device placed in the
a century old, but only limited progress has been
outer ear.
Hearing aids made towards this goal. Modern approaches to the Several kinds of implant
It is thought that over 95% of babies born deaf task use techniques of acoustic analysis, and invoke have been devised. The ba-
have some degree of residual hearing, and therefore the powerful display possibilitiesmade available sends a single elec-
sic type

through computational electronics. Most of these trical impulse to the nerve

the earlier some kind of auditory training can be - based on
fibres either an
established, the better. A basic way of providing approaches are an early experimental stage, re-
at
amplified speech waveform
early help is to amplify sound to the child's ears flecting the state of the art in speech analysis and or some kind of modified
through the use of a hearing aid. Similarly, hearing synthesis (§26), and commercial application is still waveform (as in the case of
a long way off; but the potential value of the hearing aids). More complex
aids can be of great value with adults where hearing
systems make use of a ser-
is deteriorating because of the natural process of research can be seen even in the basic displays that
ies of wires, each of which is
aging (§6), as well as with those who have acquired can currently be produced. capable of acting as an inde-
a degree of hearing loss through disease or trauma. Most visual pattern displays aim to improve pendent information channel
Hearing aids are therefore worn by many people: speech production as well as speech reception. To carrying just one aspect of
the speech signal (e.g. a for-
for example, it is estimated that nearly 2 million achieve this, the approaches present a speech target
mant, §23). These can be
are currently in use in the USA. visually and store it on a display screen. The deaf
linked to the auditory nerve
However, hearing aids have their limitations — person then attempts to pronounce the speech in or directly into the cochlea
a fact that needs to be borne in mind when one such a way that the pattern produced matches the ('cochlear implants' - see
one already on the screen. The main advantage of diagram below).
faces the contemporary proliferation of firms offer-
It is too soon to say how

ing a bewildering variety of aids. Hearing aids
amplify, but do not necessarily clarify, speech. All
sound in the environment of the aid is amplified
— including background noise. Also, raising a lis-
tener's sensitivity may lead to discomfort with cer-
tain speech sounds and background noises — [s],
for example, can become unpleasantly harsh.
It is of course possible to reduce the output of an
aid, but this can lead to the speech sounds becom-
ing distorted. For such reasons, many hearing-
impaired people have been known to reject their
aids after a limited period of use: one recent report
suggested that as many as a half of
purchased end up not being worn.
The problemmost marked when someone has
is
which may remove
a high-frequency hearing loss,
this technique is that there is no delay in providing
successful these implants
feedback about the success of the pronunciation. can be. The basic sensations
However, very little research has yet taken place of sound that are produced
into the relative efficiency of the various devices using current techniques are
apparently a long way from
now available in promoting intelligible speech in
the sounds of normal
young children.
speech, and the deaf person
Visual display devices that are routinely used in has to learn to decode these
speech training programmes include the laryngo- sensations. But on the princi-
ple that any auditory informa-
graph (p. 141), which is used mainly for work on
tion is better than none, a
intonation, and several systems that train indivi- great deal of interest is cur-
dual speech elements, such as fricatives and nasals, rently being shown in the de-
or distinctive features of sound, such as voicing velopment of the approach. It
is however an operation that
all hearing aids (p. 128). Many are based on spectrographic dis-
is used only on people who
plays (p. 136) or make use of some simpler form
are permanently and pro-
of spectral analysis. foundly deaf, as it is a com-
Several other visual methods have been devised plete substitute for normal

many consonants and environmental noises (such to assist the deaf, including the use of a palantypist hearing.

as running water or a telephone bell). Here, it is (p. 207) to provide a simultaneous transcript of
essential to find an aid that selectively amplifies a conversation, and the provision of specially pre-
the frequencies that are most affected. However, pared subtitles for television programmes using the
if the lossvery severe, this approach will not
is teletext service offered by some major networks
help. An alternative is to use a type of aid that (p. 193).
converts the high-frequency information in the
sound wave into low frequencies that fall within
the range of residual hearing - a procedure that
is often beneficial, though the speech can end up
very distorted in the process. Several other tech-
niques have been devised in an attempt to cope
with the many individual variations that exist.
The audio signal is fed to an antenna coil (A) in the outer
Specialized audiological advice is essential in ear; picked up by the receiving coil (R) and sent to the
it is

choosing a hearing aid. It is not easy to evaluate electrode (E) in the cochlea. (From M. M. Merzenich, 1 975.

268 VIII LANGUAGE, BRAIN, AND HANDICAP

Speech sound patterns Subtitles for the deaf
Six speech sound patterns, as produced on the Voice The invention of teletext as
Visualizer (from W. Pronovost etal., 1968). an additional service on the
television signal allowed the
broadcasting authorities in
Britain (BBC and IBA) to
introduce regular subtitling
services from 1 980 onwards.
The aim is to help viewers
who are deaf or hard of hear-
ing to watch, at their own
choice, subtitles that convey
a soundtrack in written form.
The viewer with a teletext
receiver calls up the appro-
priate page number, and the
subtitles automatically
appear on the screen. By
1987, each authority was
providing up to 25 hours of
teletext subtitling each week.
A sophisticated code has
been developed so that the
subtitles can best convey the
information and feeling of the
soundtrack. Speakers are
identified either by the use of
coloured text, or by placing
the subtitle next to the per-
Speech traces
In thisexperimental speech training system, features
of place and type of articulation, and voicing, are
shown as different kinds of traces varying in height,
width, and colour, moving from left to right with time.
The colour of the original is here shown using shading.
(After G. J. Bristow& F. Fallside, 1979.)
SAY
a s e
A subtitled exchange in the
BBC word game 'Blankety
Blank'. The word 'Bicycle' in
the first line is spoken by the
competitor, and subtitled in
green. The remaining text,
spoken by the host, is in yel-
low.
son talking. Conventions grammes, such as news bul-
have also been established sports coverage, or
letins,
to communicate any accent broadcasts of public events.
or tone of voice which the In both Europe and the USA,
subtitler feels is germane to systems have been de-
a viewer's grasp of the pro- veloped using stenography
gramme. Similarly, sound or palantyping (p. 207). The
effects and music are des- output from the operator's
cribed when relevant. machine is recognized by a
To allow viewers adequate computer containing a large
time to read the subtitles and phonetic dictionary, and this
follow the pictures, the verbal generates a written version
soundtrack must be quite as a subtitle. This system
heavily edited. Most tele- was first used to subtitle
Pre-
vision programmes are pre- sident Reagan's inaugural
recorded on videotape or film speech in 1981 However, in
.
before transmission. A sub- its early experimental days it
titler, working from a video was not error-free, and the
cassette of the programme, BBC has recently developed
and sometimes helped by a an alternative system, com-
script, writes the subtitles bining a number of different
onto a computer disk. This techniques, which depends
disk is then played into the on a closer relationship be-
television transmission sys- tween the subtitling and the
tem in synchrony with the methods of television pro-
programme. duction. This system is
Different techniques must called RECAP.
be used with live pro-
46 LANGUAGE HANDICAP .
269
 . . . .

islikely to be aphasia. The other causes of aphasia

Aphasia include certain kinds of cerebral tumour, brain dis-
When an area of the brain involved in language ease, and traumatic damage (head injuries due to
processing is damaged (p. 261), the language dis- traffic accidents, falls, acts of violence, etc.). About
order that results is known as aphasia or (especially a quarter of penetrating head injuries lead to all

in Britain) dysphasia. This terminological choice aphasia. Altogether, the annual incidence of the
arises from a literal interpretation of the two pre- handicap is about 0.6% of the population (1 in
fixes: a-phasia suggests a 'total' lack of language; 200), with males more at risk.
dys-phasia implies a 'partial' lack. However, the About a quarter of all patients recover within
distinction has no clinical significance: all aphasic three months. The rate of recovery then decreases,
people have some residual language ability, even with full recovery increasingly unlikely after six
if this is only a minimal level of comprehension. months. A further 25% of patients are still severely
It therefore makes no difference which prefix is affected after a year, with little subsequent im-
used (as long as usage is consistent): we are dealing provement expected. The different communication
with a continuum of disability from very mild to modalities usually recover at different rates: gener-
very severe. The a- prefix is now more widespread, ally, comprehension improves more rapidly than
especially in the USA, and it has come to be used production. However, the process of recovery is
in the name of the research field, aphasiology. little understood. It may be that cells close by the
A more important question relates to the nature damaged area regain some of their function after
of the behaviour affected. Aphasia is usually defined a while, or perhaps other parts of the brain (such
as a handicap of language comprehension and/or as the right hemisphere) may come to be used.
production caused by specific brain damage. It
therefore clearly excludes language handicaps asso-
ciated with other conditions, such as peripheral
deafness (where there is no brain damage, p. 266)
or senile dementia (where there is a more general
deterioration of mental faculties). But it is more The effects of aphasia
difficult to exclude handicaps that involve other Samuel Johnson Walter Scott say and was embarrassed I

aspects of symbolic expression and the associated From a letter written on 19 From his diary, 5 January because could not say I

June 1 783 three days after 1826:

cognitive skills - as when aphasic people display anything. I read all the
a stroke robbed him of Much alarmed. had I spare time Ihad. In the am-
problems with understanding gestures, the symbo- speech: walked till 12 with Skene bulance, where used to I

lism of colours (as in traffic lights), performing Iwent to bed, and in a short and Colonel Russell, and spend upwards two of
arithmetical operations, remembering, or paying time waked and sate up as then sat down to my work. hours daily with four and
has long been my custom, To my horror and surprise people week after
five
attention. Should these difficulties be considered
when felt a confusion and
I
Icould neither write nor week and where was less
as part of the disorder or separate from it? The
I

indistinctness in my head spell, but put down one embarrassed, I used some-
focus of aphasic handicap is undoubtedly on pro- which lasted, supposed I word for another, and wrote times to try different words.
blems of expression and comprehension in gram- about half a minute: was I nonsense. was much I One week was optimistic I

mar and semantics, whether in speaking, listening, alarmed and prayed God, overpower'd at the same and the next there was no-
that however he might af- time and could not con- thing
reading, writing, or signing; but these problems . .

flict my body, he would ceive the reason. On wak- My writing was more de-
relate closely to difficulties of a pragmatic, cogni- spare my understanding. ing my head was clearer . . pressing. had only written I

or perceptual kind, and a sharp boundary line

tive, This prayer, that might try I (W. E. K. Anderson, 1972, 'Good luck, Clif or a mes-
cannot always be drawn. the integrity of my faculties p. 55.) sage like cigarretes' (spelt
I made in Latin verse. The (22 April 1830) wrong - this might have ar-
lines were not very good, Anne would tell you of an oused my suspicions, but it
CAUSES OF APHASIA but know them not to be
I awkward sort of fit had on I did not), and for the rest
The brain is totally dependent on the oxygen con- very good. made them I Monday last; lasted about it made the excuse that did I

veyed by its blood supply; brain cells will die if easily, and concluded my- five minutes, during which not write with my left hand.
deprived of oxygen for more than a few minutes. self to be unimpaired in my I lost the power of articula- But it was my mother's
There are many cerebrovascular accidents (CVAs, faculties. tion, or rather of speaking birthday in May and I felt
Soon perceived
after, I
what wished to say. re- that should write her a let-
commonly known as 'strokes') that can cause this
I I I

that I had suffered a paraly- vived but submitted to be ter. no sooner had the
I

to happen, and these account for about 85% of and that my

tick stroke, bled. paper in front of me when
all cases of aphasia. In adult western people, arter- Speech was taken from (S. G. Lockhart, 1900, every single word galloped
ies can become 'furred up' with fatty cholesterol
me. had no pain and so
I
p. 262.) out of sight. was left star- I

little dejection in that dread- ing at the blank sheet.

deposits, associated with such factors as smoking,
ful state that I wondered at A recent account Nearly half an hour passed;
diet, and lack of exercise: the deposits cause nar- my own apathy . .
Douglas Ritchie wrote a panic grew; this was no-
rowing and obstruction of the arteries, and this order to rouse the vo-
In diary of his recovery from thing to do with my left
may cause a stroke. Another possibility is for the calorgans took two I
stroke in 1960. One year hand. At length my wife
drams. Wine has been after the stroke he felt like came in and she dictated
arteries to become blocked by foreign matter that
celebrated for the produc- this: slowly, letter by letter,
has entered the blood stream. Or they may hae- tion of eloquence; put my- I
My speech? might have I 'many happy returns . .
.'.
I

morrhage in various ways. Whatever the reason, self into violent motion, had two or three stray managed to forget my
if these events take place in the areas of the brain and, think, repeated it. But
I
words but could not tell. In
I panic for a time.
that deal with language processes (something that all was vain . . the Centre rarely spoke to
I (D. Ritchie, 1960,
anyone. had nothing to
I pp. 96-7.)
happens in about a third of all strokes), the result

270 VIII LANGUAGE. BRAIN. AND HANDICAP

TYPES OF APHASIA
There have been many different classifications of Expressive aphasia
aphasia, reflecting the difficulty aphasiologists find Several of the symptoms of /dy/euh, Docteur Durand. finally,examinations, uh, fi-
in grouping patients together so that their medical Broca's aphasia can be Les reins. Euh, kinesithera- nally, a coma uh, a little bit.
and theirbehavioural symptoms coincide. A seen from this French pa- peute. Marche euh, euh, Uh, a month/ a month, uh,
tient's description of the tres bien, enfin, un peu, un
based on the site of the lesion(s) will
classification pavillion F-3/dy/uh, Doctor
evolution of his disease peu. Euh, premier no- Durand. My kidneys. Uh,
make neurological sense, but may not result in a (abnormal drops in pitch vembre, medica/ Le/ Gis- physiotherapist. Walk uh,
neat description in linguistic or psychological are marked by/): card/ Docteur Giscard euh, uh, very well, finally, a little,
terms. Correspondingly, a behavioural classifica- Euh, hemiplegie, euh, ful- reeducation. Euh, euh, oui, a little. Uh, November first,
tion usually cuts across some of the traditionally gurant, euh, Hopital Pas- euh, kine/non, huit heures, medica' The/ Giscard/ Doc-
teur, Nice, Nice. Euh, Doc- kine, euh, un quart d'heure torGiscard uh, therapy.
recognized neurological distinctions. There also
teur Dupont. Euh, euh, Uh, uh, yes, uh physio/ no,
may be some change in the aphasic symptoms, as examens/enfin, examen, (Uh, hemiplegia, uh, fulgur- eight o'clock, physio, uh, a
the recovery period progresses. A few major cate- euh, enfin, un coma euh, ant, uh Pasteur Hospital, quarter of an hour...)
gories have sufficient homogeneity, both medically un petit peu. Euh, un mois/ Nice, Nice. Uh, Doctor Du- (A. R. Lecours etal., 1983,
un mois, euh, pavilion F-3 pont, uh, uh, examinations/
and behaviourally, to stand the test of time, and p. 86.)

these continue to be cited as 'classical' aphasic syn-

dromes. These patients, however, may well be out- Receptive aphasia
numbered by the many cases where the aphasic Several of the symptoms of puisqu elle se presente les/ died. And then, she is there
symptoms are 'mixed', to some degree, and where Wernicke's aphasia can be Je I'avais envoyee a I'ecole now. She will soon be /syz/
a classical diagnosis is unclear. seen French pat-
in this puisque, moi, je travallais years old. She is still going
ient's response to a ques- bien dans les /syz//- euh to school since she pre-
tion about his family (strong -a la/faRmid/ de/ de/ de sents herself the/ had sent
1

stresses are italicized): /syz/, n'est-ce pas, de deux her to school since myself I

Broca's aphasia The lesion is classically located

/etmiR/ was indeed working in the
. Et alors, je/ Cette
in and around Broca's area, typically extending Oui, j'ai une autre femme /mwaze/- la - euh, Gi- /syz//- uh - at the/faRmid/
some way back along the Sylvian fissure (p. 260). qui est restee depuis la nette, elle s'appelle - elle/ of/ of /syz/ isn't it, of two

The nature of the symptoms has led to its also being /bcetRe/ de I'enfant de ma elle/abil/... /etmiR/. And then, I/ This
fits. // a elle avait dix ans (Yes have another woman /mwaze/- there uh, Ginette
called expressive or motor aphasia. The language
I

quand mon/fes/est mort. who has remained since is her name - she/ she
is usually characterized as markedly non-fluent -
Et alors, elle est la mainte- the /bcetRe/ of the child of /abi|/...)
slow, laboured, hesitant, often one syllable at a nant. Elle va sur/syz/ ans. my son. He is/ she was ten (A. R. Lecours etal., 1983,
time, with great difficulty in articulation, and dis- Elle va toujours a I'ecole, years old when my/fes/ P-94.)
turbed suprasegmental features (§29). Sentences
are short and reduced to a 'telegrammatic' style,
with little use of the normal processes of grammati- Other symptoms
cal construction (§16). Individual words are often
Aphasia is often accompa- • Apraxia (or dyspraxia): nying weakness or paraly-
repeated. Comprehension of everyday language is nied by other symptoms an often severe difficulty in sis in the side of the body
near-normal. which need to be taken into controlling voluntary move- opposite the hemisphere
account when assessing ments of limbs or vocal or- which has been damaged
the communication impair- gans. there
In particular, (p. 258). When this affects

Wernicke's aphasia The lesion is classically loca- ment as a whole. may be an inability to con- the face or neck, the func-
• Agnosia: a difficulty in trol sequences of sounds organs
tioning of the vocal
ted in Wernicke's area (p. 260), though there is
recognizing familiar sen- (articulatory or verbal can be impaired, to pro-
some variability. The nature of the symptoms has sory stimuli. When the dis- apraxia) or gestures. The duce a poorer quality of ar-
led to its also being called receptive or sensory ability relates tosounds, it intention to communicate is ticulation. The effects

aphasia. The language is characterized as fluent, is known as auditory agno- present, but the patient range from mild to severe
sia; when it relates to pic- cannot carry it out. - from a slight slurring to
often excessively so, with no articulatory difficulty,
tures or shapes, it is known • Anarthria (or dysarthria): total unintelligibility.
though there may be several erratic pauses. There as visual agnosia. there is often an accompa-
is usually a severe disturbance of comprehension,

though this is obscured by a normal intonation.

The speech illustrates many stereotyped patterns,
circumlocutions, unintelligible sequences (known
as 'jargon'), errors in choosing words and pho-
nemes (§28), and problems in retrieving words
from memory.

Global aphasia The symptoms are those of severe

Broca's and Wernicke's aphasia combined. There
is an almost total reduction of all aspects of spoken

and written language. The patient's expressive

abilities are minimal, and in most cases do not
much improve over time. Comprehension of
spoken language, initially very poor, shows limited
recovery. The disorder is sometimes known as 'irre-
versible aphasia syndrome'.
A group enjoying the atmosphere of a stroke club - one of many voluntary groups that have
been up to aid the process of rehabilitation in people who are impaired by the range
set
of handicaps that follow a stroke - notably paralysis and aphasia.

46 LANGUAGE HANDICAP •
271
 )

Dyslexia and dysgraphia Several other types have been proposed. There Deep dyslexia
is, for example, a visually based dyslexia, in which
The onset of brain damage in adult life frequently
symptoms
people fail to read the parts of a word correctly
leads to a disorder of reading or writing in people (e.g. one patient read 'night' when shown near + The first patient providing
who have previously been literate. The handicap light), or confuse words of similar appearance (as
evidence of a deep dyslexia
syndrome was studied by a
is usually accompanied by aphasic symptoms affect-
when met was misread by one patient as 'meat', Medical Research Council
ing spoken language (p. 270); occasionally, it is and rib as 'ride'). In such cases, the patient can team in Oxford in the 1 960s
the only, or predominant, symptom. In all cases, often name the letters of the word correctly, but - a person who had been a
the reading disorder is referred to as (acquired) remains unable to identify the whole word. There highly literate adult before

dyslexia and the writing disorder as (acquired) dys- his left-hemisphere injury.
are also several disorders of a neurologically more His reading errors were clas-
graphia. The a- prefix is also used, especially in 'peripheral' kind, such as letter-by-letter reading, sified into five types (in each
Europe and North America (alexia, agraphia). The in which patients find it necessary to name all the case the target word is on
handicap from the
label 'acquired' distinguishes the
letters of a word (aloud or subvocally) before they the left and the patient's ver-
more widely known developmental kinds of dys- can identify it.
sion on the right).

lexia and dysgraphia that occur in young children

The search
for 'pure' types of dyslexia is compli- Semantic errors
where there is no evidence of any brain damage cated by the occurrence of individual differences act -» play
(see p. 273). close -> shut
between patients, and by the existence of cases
Neuropsychological studies of these handicaps where symptoms are 'mixed'. Problems of interpre- dinner -> food
have generally proceeded by classifying patients afternoon -» tonight
tation are therefore considerable. Are deep dyslexic
into types, based on a detailed description of the
errors due to a partial impairment of the left hemis- Derivational errors (p. 90)
kinds of errors made. The process is a slow and phere alone, or is the right hemisphere involved wise -» wisdom
difficult one, partly because of the large amounts strange -» stranger
in some way? And, within the first of these possibi-
of vocabulary that have to be analysed before an pray -* prayers
lities, is the disorder the result of an impaired
birth -> bom
error pattern emerges, and partly because there are
semantic system, or is that system intact, with the
usually associated language symptoms that also problems arising out of an impaired ability to make Visual errors
need to be taken into account. Nonetheless, since correspondences between graphemes and pho- stock -> shock
the 1970s several types of acquired dyslexia and quiz-* queue
nemes? Answers to such questions will only emerge
dysgraphia have been proposed, based on a small crocus -» crocodile
once the database is enlarged by in-depth linguistic saucer -> sausage
number of case studies.
descriptions of many more cases.
Function words (p. 91
TYPES OF ACQUIRED DYSLEXIA for -» and
his -» she
Phonological dyslexia People with this problem TYPES OF ACQUIRED DYSGRAPHIA the -» yes
are unable to read on the basis of the 'phonic' rules Most work in this field has studied the disruption in -> those
that relate graphemes to phonemes (§34). This caused to spelling ability (p. 213). Three syndromes
have been proposed, analogous to those proposed Non-words
means that they can manage to read familiar words, wux -> ('don't know')
but they have great difficulty with new words (such for acquired dyslexia.
wep -» wet
as technical terms) or with simple nonsense words dup-»damp
(such as lak). Phonological dysgraphia People with this prob- nol-> ('no idea')

lem can spell real words but not nonsense words (J. C. Marshall & F. New-
combe, 1980, pp. 1-3.)
Deep dyslexia Here too people are unable to read (though they can sometimes read many of them,
new or nonsense words, but in addition they make and speak them aloud).
many semantic errors (e.g. reading forest as 'trees').
There are also several other types of difficulty, Deep dysgraphia Here too there is no ability to
including visual errors reading signal as 'sin-
(e.g. spell on a phonetic basis; asked to write a dictated
gle'), and errors that combine visual and semantic nonsense word, for example, it is often replaced
properties (e.g. reading sympathy as 'orchestra', by a real word that is similar in sound (e.g. blom
presumably because of the link via symphony). is written flower, presumably because of the word

Words with concrete (as opposed to abstract)
meanings are easier to read. The table (right) gives
further examples of this unusual syndrome.
Surface dyslexia People with this problem are
very poor at recognizing words as wholes, and rely
greatly on a process of 'sounding out' the possible
relationship between graphemes and phonemes.
Irregular words (such as yacht) pose particular dif-
ficulty. A wrongly pronounced word will be given
a meaning on the basis of how it sounds, not how
it looks (e.g. one person read begin as 'beggin',
then added 'collecting money'). There is a problem
with homophones (see Glossary, Appendix I) (e.g.
one person understood bury as 'a kind of
bloom). Errors seem to be semantically related (e.g.
one person, asked to write bun, wrote cake). The
spelling of words with concrete meaning is better
than that of words with abstract meaning. The rela-
tionship to reading ability is unclear: one patient
studied had normal reading ability, but most seem
to have some deep dyslexic symptoms also.
Surface dysgraphia People with this problem can
spell spoken nonsense words in a plausible way,
but cannot spell irregular real words (e.g. one per-
son wrote biscuit as bisket) - and even regular
words may be affected. They seem dependent on
using grapheme-phoneme conversion rules; whole-
hat'). word spelling is impaired, though not entirely lost

272 •
VIII LANGUAGE, BRAIN, AND HANDICAP
 . ,

Spatial dysgraphia
Deep dysgraphic errors
Responses of one deep dysgraphic patient to part of
This patient has had an operation to remove a tumour
a single-word dictation test.
from the parietal lobe of the right hemisphere (p. 258).
Function words are particularly poor: some are not
One of the symptoms that resulted was a spatial attempted; some bear little resemblance to the
dysgraphia that is clearly seen in the patient's writing
stimulus word. In three cases, he added content word
abilities. The left-hand side of the page is neglected;
homophones (1 b, 4b, 9b), and was able to spell twc )

the lines are at an angle; some letters are spaced

of them. The content word list shows several visual
abnormally; and there are several unnecessary
errors (e.g. wAiyfor way), but none of the semantic
repetitions of letter strokes and letters. (From H.
errors that were also a feature of this person's
Hecaen & P. Marcie, 1974, p. 359.)
handicap (e.g. writing smallior little). (From F. M.
Hatfield & K. E. Patterson, 1984, p. 189.)

Content words Function words

\ v*k • use ),«. «\\ too

my
String inside
J; <u\ ^* say \ 1 r„„ •] [two]

tA
^ 5/

</
o>i

M«
•
"1°"
old

day
*,

\
\AJVw

-
how
off
head
It was easy to talk about

°6<x ~."*> s/ ^ TT,<ir\ s men k* WWtV v»>Vnc him

what had seen in the park,
I

or to sort out the ballet

4X u. ^o^e see [fcfc """V u>Sj [hymn]
shoes, or to put books away
1 y ia, way S o.'f out neatly according to size, but
« <-> put I, - has to decipher the alphabet, or

wo war
recognize OAT. and say
«\, % UUCXY why
what it spelt was almost im-
v co set - yet
• \o, v.
* possible When was re-
. . . I

••- - _ our quired to write, a strange

I** C> y
n ^
*-'d (hour]
who
feeling
felt there
came over me, and
was a long piece of
I

string in my head.
My mother would say,
'CAT. spells cat. Susan,
(e.g. one person spelled yacht as yhagt, showing there are around 40 different terms
ability'). In fact
what does CAT. spell?'
some visual recall). used for problems in this area, some of which retain don't know, don't know,
'I I

Acquired dysgraphic patients are usually also a medical bias, such as 'minimal brain dysfunction' Mrs Hampshire (as called I

dyslexic to some degree. Moreover, classification and (in parts of Europe) 'legasthenia'. Because the her at school), I don't know
must allow for cases where there are specific motor handicap is viewed as a problem with 'written lan- what it spells.'
The string inside my head
or sensory impairments. For example, there are guage' in all its forms, the term 'dyslexia' usually
stopped me from answering.
people who can speak, read, spell aloud, and type, subsumes the kind of difficulties referred to as 'dys- It actually felt as though my
yet who cannot produce the letter shapes or move- graphic' in the brain-damaged adult. skull housed a whole ball of
ments required for writing by hand. Letters are The blighted school career of such children, string, with an end sticking

badly formed, misplaced, repeated, or omitted. In when no-one recognizes their handicap, has been
out of my crown. I thought
that if I pulled at this, couldI

such cases, it is graphetic rather than graphological well documented. Their inability to read, whether get the string out, empty my
ability that is affected (p. 185). for information or pleasure, and their daily failure head of it, unravel the tangle
in theirattempts at written work, has a devastating in my brain . .

effect upon their ability and motivation to learn.

'Mummy, can I feel my
string.'
There are often associated problems in coping with
Developmental dyslexia number symbols (in arithmetic), and in tasks
'Don't be ridiculous, Su-
san.'
Since the early years of this century, it has come requiring short-term memory, such as following The page, the pencil, my
to be widely recognized that there are children who, instructions. Their poor writing and spelling tends mother's face, her slightly
oily skin - not a line on it -
after a few years at school, are consistently seen to be viewed as a symptom of educational subnor- her dark brown eyes com-
to fail at the tasks of reading, writing, and spelling, mality or lack of intelligence - or, if the child is pelling me to answer cor-
despite normal intelligence, instruction, and oppor- known to be intelligent, leads to a charge of laziness rectly, her nail polish half

tunity to learn. No
medical, cultural, or emotional or 'not trying', with subsequent punishment in erased by the washing-up,
all this could see and re-
reason available to explain the discrepancy school and increased family tension at home. As
I

is
member-but could not re- I

between their general intellectual and linguistic a result, it is not surprising to find that many such member CAT. Probably the
abilities and their level of achievement in handling children become anxious, withdrawn, or aggressive most difficult word in the

written language. There is often a history of early - with deteriorating behaviour in some cases lead- world, CAT. If only the other
language delay, but by age 9 or so, spoken language ing to them being described as maladjusted. Career children couldn't spell CAT.
'Stop looking in the mirror
ability is apparently normal, whereas written lan- prospects, in such cases, are minimal.
and think about how you
guage skills may remain at the level of a 5- or 6- Questions of incidence and causation are dis- spell cat.'
year-old. cussed on pp. 274—5, along with a more detailed I couldn't. I just could not.
These are the children who have been called 'dys- illustration of the range of dyslexic symptoms.
I tried, but I couldn't. My
lexic', though alternative labels have been devised
head was empty - except for
the string.
for the condition in an attempt to escape the origi-
(From Susan's Story (1 981
nally medical connotations of this term (notably pp. 26-7), the autobiography
'specific reading disability' and 'learning dis- of Susan Hampshire.)

46 LANGUAGE HANDICAP • 273

Incidence
The dyslexia problem is becoming increasingly Copying
recognized, with many countries now setting up Copying by three children, the younger child (c). (c) Normal reader, aged 6
which shows some of the (a) Normal reader, aged 9 years, 6 months, with a
organizations to draw attention to the handicap
problems of the backward years, 8 months. reading age of 6 years, 9
and to provide special help. In a very few countries, reader. The style of (b) is (b) Backward reader, aged months. (From L. Bradley,
this help is guaranteed by legislation. It is however well behind that of (a), who 9 years, 8 months, with a 1983, p. 238.)
extremely difficult to arrive at an accurate estimate is the same
age, and in reading age of 6 years, 9
of incidence because there are no internationally many respects it is not as months.
well organized as that of
accepted reading tests and criteria of handicap. In
one survey of 16 countries, the mean percentage Wa^^tiAt Wttwy r^peA**,"** <&
hi went h> jUf tm. V*
of non-retarded children with reading difficulties (a) Th*
was 8% — but this covered a range that went from
1% (China) to 33% (Venezuela). Some estimates if / »M h* YaA. J^t KoA. tirr* to sh**e Kumdj-
suggest that dyslexic boys outnumber girls in a bJhm he jam/ r/*m conuAjy icrtb it* AnL-
ratio of around 3:1, others that it may be as many
as 10:1.
The uncertainty derives from the fact that read- (b)TU^^wLK to 5U*wrtK<.4^d*i^t. M -5
ing difficulty is a continuum from normal to abnor-
mal, with the only criterion of handicap being that »"
the children's ability is well below their age and

intelligence. Everything therefore depends on how

VMwtn K ^nw ir^ruo ftfr^W tvZ
tVx^tr^
intelligence and reading achievement is measured,
and what is considered to be 'well' below normal.
For example, if the definition of dyslexia includes
only those children who are retarded by at least
two years in reading ability, the numbers affected
will be appreciably greater than one that requires
that they be retarded by at least three years. Such
differences of method, even within a single country,
make it virtually impossible to arrive at an agreed
statement of incidence.

Causation
The question of causation has promoted great also
^ cn css the SaOd 3

controversy. Until recently, there was a widespread

assumption that all dyslexics were fundamentally
alike, and that a single cause of the handicap could
Dyslexic progress
be found. A large number of candidate 'causes' (c) The same child's spell-
(a) A sample of a dyslexic months, after specialized ing at age 8. The words
were therefore proposed, postulating any of several
boy's free writing at 8 years help. The sentence reads: are: see, cut, mat, in, ran,
medical or psychological factors, such as visual per- of age. The sentence 'I was walking down the led, lot, hat, pen.
ception, intersensory integration, memory, atten- reads: 'My favourite hobby and heard a scream (d) The same child's spell-
street I

tion, eye movement, verbal processing, and is art work and maths.' and went into the house
I ing of these words at age

hemispheric dominance (p. 258). There could be (b) The same child's free and saw a man with a
I 9. (From M. Thomson,
age 9 years, 8 1984, pp. 41

^ tM
ing at
writing knife.
knife.' 41-2.)
-2.)
several possible approaches within any one of these
headings. For example, under dominance it has
been argued that dyslexia is the result of (a) a lack wl fck** W<^ u<LWv^
of dominance, (b) a lag in dominance development,
(c) a specific left-hemisphere deficit, (d) right-
hemisphere interference, or (e) a disintegration of
functioning between the two hemispheres. The role
of the left hemisphere is strongly implicated (as
is suggested by associated spoken-language delays

and errors, and problems of motor coordination),

but its exact influence is unclear.
Recent reviews of what is now a vast experimen-
tal literature indicate that a unitary explanation
for dyslexia is illusory. The modern focus on
individual case studies (as opposed to the tradi-
tional use of group studies) is bringing to light the
existence of a variety of dyslexic syndromes, re-
flecting several possible causes. A popular contem-
W f e* Cut *** i* r*r\ li^ dat
hot {**
porary view is that there is a large set of factors

274 •
VIII LANGUAGE, BRAIN, AND HANDICAP
implicated in dyslexia, some sub-set of which turn There is, moreover, always the possibility that the

up in individual cases. For example, in one group brain mechanisms that underlie reading acquisition
of children, therewas clear evidence of an unstable are different from those used to maintain reading
eye dominance: the children had not established skills in later life. The unity view thus provides
a stable 'leading' eye in their reading. Another us with a set of intriguing but at present largely
group showed with making perceptual
difficulties speculative hypotheses.
distinctions (e.g. distinguishing same/different
letters). A further group displayed problems with
short-term memory.
The main methodological problem in such re-
search is to determine whether the weakness shown
by dyslexics is the cause or the result of the handi-
cap. For example, many of these children have
faulty eye movements (shorter saccades, longer fix-
Common features of dyslexia
A wide range of factors has No serious lack of school- • A history of late lan-
ations, more regressions, §34), but it is an open been implicated in the ing. guage development.
question whether these form a constitutional pro- search for a definition of • Some pronunciation dif-
blem that made it difficult for them to learn to dyslexia. The following fea- Psycholinguistic features ficulty, especially with long
read, or whether the poor movements began as a tures commonly recur, but • Reading, spelling, and words.
it must be stressed that writing all below that ex- • Non-fluency in speech.
result of their difficulties with reading, or whether
there is great variation be- pected for age and IQ. • Poor auditory discrimi-
there is no functional relationship between them tween dyslexic children • Persistent and often bi- nation of speech sounds.
at all. If information is not available on what the (and, of course, between zarre reading and spelling • Problems of visual per-
children were like before they began to read, and those whose symptoms errors, e.g. letters reversed ception.

on how they perform with non-reading tasks, it have continued into adult or out of order (confusion of • A history of motor clum-
life). Probably no dyslexic b/d, was/saw, etc.). siness.
is difficult to interpret the results of such experi-
child would display all of • Confusion when labell- • Problems of finger differ-
ments. these features, but most ing left and right, and gener- entiation.
The conflicting and ambiguous research find- display several. poor directional ability.
ally • Mixed handedness or
ings, linked with ambitious claims about 'the' cause • coding
Difficulties in confused laterality.
Background features symbols and sounds, e.g. • Poor concept of self.
of dyslexia, have led to a great deal of scepticism
Sight normal. naming letters of the alpha- • Sometimes good spatial
about the condition, especially when the possibility Hearing normal. bet. model making.
skills, e.g.
of an underlying medical cause is being stressed. IQ near-average or above. • sequenc-
Difficulties in

It will be some time before these doubts are Health normal. a

ing, e.g. putting things in
Adequate first teaching. series, remembering days
resolved, and new research initiatives will need to
No previous emotional dis- of the week, keeping one's
be forthcoming - in particular, devising individual turbance. place.
developmental profiles, along the lines of the No gross brain damage. • Poor short-term mem-
acquired dyslexia research, and relating findings No socio-cultural depriva- ory, e.g. remembering
more to the nature of reading development in nor- tion. tables or instructions.

mal children, in order to establish what counts as

an 'abnormal' error. In such ways, it will be pos-
sible to devise better developmental classifications Hyperlexia
based on behavioural symptoms. Reading-retarded children year-old. He had learned well beyond their other
sometimes develop a sur- nursery rhymes and televi- cognitive abilities. They
prising ability to read aloud sion commercials as early have great difficulty, for ex-
Acquired vs developmental?
- including the accurate as age 2, and learned to ample, associating the
The several similaritiesbetween the symptoms pre- production of quite ad- read soon after 4 with little words they read with ob-
sented by the two kinds of dyslexia have led some vanced vocabulary, well help from his parents. By On the
jects or pictures.

scholars to argue that there is an underlying iden- beyond their level of com- the time he was 5 years other hand, they have great
prehension. In one study, old, he was fluently reading facility in sounding out non-
tity. Parallels have been proposed between develop-
thisremarkable skill was aloud material that would sense words. (After P. R.
mental dyslexics and acquired deep dyslexics (p. observed in a 7-year-old be appropriate for a normal Huttenlocher & J. Hutten-
272) - for example, both groups have trouble in boy with an IQ of only 77, 10-year-old. locher, 1973.)
reading nonsense words, and are better at reading and a level of motor deve- The ability such children
lopment equivalent to a 3^- have to read aloud goes
concrete words. However, so far there is little clear
evidence that children display the kinds of semantic
error that are crucial to the identity of the deep
dyslexia syndrome. Similarly, there have been pro- A serious directional pro-
blem illustrated in the
posals that developmental dyslexia displays a par-
writing of a 7-year-old
allel with acquired surface dyslexia (e.g. because girl. (From L. Tarnopol &
of similarities in phonic reading ability) and with M. Tarnopol, 1976,
phonological dyslexia (e.g. because of similarities p. 283.)

in direct visual word recognition). None of these

positions has yet produced a substantial child data-
base, however, and several differences between the
adult and child populations remain - in particular,
the greater variability of children's performance.

46 LANGUAGE HANDICAP 275

Voice disorders Laryngeal
Many people develop an expressive handicap in abnormalities
which their voice has a markedly abnormal quality. The illustration shows
some abnormal laryngeal
The pitch, loudness, and timbre (§29) may be so conditions, all of which can
inefficient that the message carried by the spoken cause dysphonia.
language may be largely or wholly unintelligible. (a) Vocal nodules - inspi-
But even if the speech can be understood, the voice ration followed by phona-
quality interferes with communication by calling
tion. The vocal folds are
prevented from closing pro-
attention to itself. The sound can be highly unplea- perly along their mid-line.
sant - for example, the harsh hoarseness or highly (b) A large, broad-based
nasal qualities that can be heard in some disorders. polyp, which will interfere
Alternatively, the voice quality may simply be inap- with the normal movement
and closure of the vocal
propriate to the speaker or the needs of the situ-
folds.
ation - as when an older male teenager retains the (c) Contact ulcers, formed
high-pitched voice of a younger child. at the point of maximum
Voice handicaps are classified into disorders of closure of the vocal folds.

phonation (an abnormal kind of vibration in the

vocal tract, as when the vocal folds fail to function
normally) and disorders of resonance (abnormal
modifications of the sound vibration as it passes
through the cavities of the vocal tract, §22). The
Laryngectomy
Malignant growths in the itiate vibration, resulting in emits a buzzing sound,
firsttype manifests itself mainly in abnormal quali-
throat, in the region of the a throaty 'esophageal' and, when placed against
ties of pitch and loudness - such as very monoto- vocal folds, can be treated voice quality. Alternatively, the neck while they are
nous, high-pitched, or weak voices - and in a range with radiotherapy, but if this they may use an artificial 'mouthing' speech, pro-
of breathy, husky, and hoarse effects that are cumu- fails, it may be necessary larynx to provide a source vides a source of phona-
latively labelled dysphoria. The second type is best to remove the larynx surgi- of vibration - a device that tion.
cally, in an operation
illustrated from the many abnormal nasal res-
known as a laryngectomy.
onances that can - some excessively
affect the voice After this operation, the tra-
nasal (or 'twangy'), some with reduced nasality chea (§22) cannot be re-
('blocked nose' effects). joined to the pharynx, as
food would spill into the
About a third of all voice handicaps have a
lungs. The defect in the
clearly physical cause - an anatomical or neuro-
pharynx is therefore closed
physiological abnormality in the vocal tract. Exces- during the operation, and
sive friction between the vocal folds can cause an alternative opening to

nodules or nodes to form at their margins. Other the trachea is made at the

contact front of the neck (a trache-

interfering formations include polyps,
ostomy).
ulcers, and various kinds of cancerous growth. Patients who have had
Vocal-fold movement may become weak because this operation are laryngec-
of disease affecting the main nerve leading to the tomees. Many learn to use
the upper part of their phar-
The upper respiratory tract, before (left) and after (right)
larynx. External damage, such as a blow to the laryngectomy.
ynx and esophagus to in-
neck, can easily affect larynx functioning.
The majority of voice disorders, however, have
a non-physical, or 'functional' cause. For example,
emotional stress can itself be sufficient for people Artificial larynx
to 'lose their voice', resulting in a range of psycho- An electronic artificial lar-

logical conditions that require lengthy and sym- ynx being used by a laryn-
gectomee. The device
pathetic investigation and therapy if they are to
weighs 200 grams (about 7
be resolved. Factors of this kind may also have oz) without its battery, is
physical consequences. Nodules and ulcers result 12icms (5") long and 4
from 'vocal abuse' — an excessive use of the voice, cms (1 i") wide. offers the It

user both volume and tone

which in time causes chronic dysphonia. But the
controls, and by rotating
reasons for the abuse are functional, arising out the head of the unit it will
of the life-style of the speakers: in particular, it give a softer or more stri-
is very common for nodules to form in those who dent tone.
live by their voice, such as singers and teachers,
and who are regularly faced with vocally demand-
ing situations.

276 • VIII LANGUAGE. BRAIN. AND HANDICAP

 Cleft lip and palate
Between one and two out of deve-
tion affects not only the
every 1 ,000 children is born lopment of speech, but also
with cleft palate syndrome. A the child's ability to eat. It is
cleft palate is a congenital also extremely disfiguring,
fissure along the midline of and a source of great emo-
the palate (§22). It may ex- tionaltrauma to parents.
tend throughout the whole Early surgical intervention is
palate (a) or affect only part thus normal (lip operations
of it (b). Cleft lip (an older la- are often within the first three
bel, 'hare lip', is nowadays months). Special prosthetic
felt to be demeaning)
is the devices are sometimes used
associated condition in to cover the palatal gap, until
which the upper lip is split. the operation is performed,
The lip may be only slightly to aid the development of
notched, or the division may normal movement within the
be complete (c), and include mouth.
the upper teeth ridge behind Because of the early inter-
(d). The split may be in the vention, many cleft palate
middle, on one side, or on children develop fairly nor-
both sides (e). Lip and palate mal speech. However, pro-
may be simultaneously af- blems of voice quality (often
fected (f). There are also very nasal) and articulation
'sub-mucous' clefts of the can persist for several years,
palate, in which the surface and there may be associated
tissues have united, but the problems of language delay
underlying structures have and hearing loss. A child who
not. is making poor progress can erning lack of progress are the extent to which the vocal movement. The avai-
flexible
Clefts of the lip or palate still be largely unintelligible to not well understood, though organs have been able to intensive speech
lability of
have very serious conse- all but his family at the age a great deal must depend on grow in a normal way, after therapy is also a crucial fac-
quences because the condi- of 3 or 4. The reasons gov- the severity of the clefts and surgical closure, permitting tor.

that cannot be explained by such factors as lan-

Articulation handicaps
guage delay or poor motor coordination. In some
Traditionally, problems of articulation were stu- cases, there is a difficulty in the perception of
died solely from a phonetic point of view; today, sounds (§25) — in particular, a failure to discrimi-
there is also a phonological dimension to the analy- nate differences between related sounds (such as
sis (§§27-8). An extremely wide range of difficul- [p] vs [b]). In others, it proves impossible to detect
ties is subsumed under the term. At one extreme, any problems of an anatomical, neurophysio-
there are slight difficulties with pronunciation that logical, or sensory kind, and yet pronunciation
hardly interfere with communication, but that may be considerably disordered. The explanation
cause some anxiety to the speaker (such as a lisp, of such 'functional' articulation problems is
or a 'weak' r). At the other, there are sound systems unknown, though one must assume that something
which are so misarticulated or disorganized that is wrong in the area of the brain involved in the

the person is largely unintelligible - something that control of phonological skills (both segmental and
is quite common in adults following acquired brain suprasegmental, §§28-9).
damage, when the motor control of speech can be
severely disrupted (p. 271).
In children, many of the pronunciation problems
that cause parental concern are due to a general Breaking the code
delay in the ability to control movements of the One 42-year-old girl pro- trends, which would need child mixes up her place of

vocal organs. Some children at age 4 or 5 are still duced the following pro- to be confirmed by a much articulation, substituting [t]

nunciations: larger sample before a for/k/and[d]for/g/. She

pronouncing words in ways typical of a child of bees/bi:t/ horse /hart/ therapist would begin treat- seems to have no pro-
2 or 3 - making immature omissions, substitutions, car/ta:/ little/jiV ment. blems with bilabial plo-
additions, or transpositions of sounds. should (It feather /tedV pen /pen/ • Vowels are generally sives, /p/ and /b/. Note that
finger /pinnV in all cases the contrast of
be added that the children's hearing may be per- scissors /hkAd/ correct, or nearly so.
five/pait/ seven /hebin/ • Fricatives (p. 1 57) are voicing (p. 128) is main-
fectly normal, in such cases.) Others have more
girl/dau/ spoon /pu:n/ always replaced by plo- tained.
serious problems of incoordination, such as being cake/ta^ blue/bwu:/ sives, except for two cases • There are evidently pro-
unable to control the speed and direction of tongue train /brain/ boy/bo^/ where [h] is involved, one blems with M
which is
movements, or to maintain consistent pressure flower/pfaua/ warm/woim/ case where [s] is omitted omitted or replaced, and
four/po:/ bird/b3:d/ (inthe consonant cluster, which turns up inexplicably
between articulators. It takes only a slight lack of
Her speech was generally spoon), and one 'odd' case in scissors.
control to turn a plosive into an affricate or fricative difficult to understand, but (scissors). • In no case is a conso-
f
(e.g. [p] becoming [p or [f]), or a fricative into
] her errors were usually not • Plosives are always arti- nant cluster used correctly,
a plosive (e.g. [s] becoming ft]). random. A detailed analy- culated as plosives, but the as in blue, train and flower.

There are other articulation problems, however, sis shows some interesting

46 LANGUAGE HANDICAP •
277
 speaker's personality or emotional state (some trac- Cluttering
Fluency disorders
ing the problem back as far as infancy). 'Anxiety' The main symptom of this
A disorder of fluency, in the context of language, theories stress the role that adverse listener reac- less well-known form of non-
fluencyis the excessive rapi-
refers to a major lack of ability to communicate tions can play in promoting a stutter. A typical
speech. Clutterers
dity of
easily, rapidly, and continuously. The problem is example is when
parents prematurely correct their
seem unable to control their
most noticeable when people have difficulty in con- children for non-fluency, or become impatient speech rate, and as a result
trolling the rhythm and timing of their speech, to when their child is non-fluent; this causes insecurity introduce distortions of
produce the phenomenon of stuttering or (as it is and anxiety, which in turn causes further growth rhythm and articulation into
speech. Sounds be-
their
more widely known in Britain) stammering. in the non-fluency. It is difficult to choose between
come displaced, mispro-
Stuttering is difficult to summarize because it such theories, as so much of the relevant evidence nounced, or omitted. Syll-
involves several kinds of non-fluency that vary con- islacking (e.g. what happened when the stutterer ables telescope into each
siderably from speaker to speaker. was young?). other. The utterance comes
out in relatively short bursts,
• The most widely recognized symptom is the ab-
Many treatment methods and programmes are
often interfering with syntax.
now available, all of which, in the absence of agreed The speed may increase as
normal amount of repetition of sounds, syll-
theory, have had their successes and failures. Some the utterance proceeds - a
ables, words, or phrases, e.g. p-p-p-please, he's
methods focus on the feedback problem, such as phenomenon known as fes-
got a — got a - got a — car. Very often the re-
by taking stutterers' attention away from their non- tination'.
• Sounds may be abnormally lengthened, e.g. sulting speech is largely un-
fluent speech (by playing specially generated noise
sssee, where the initial [s] can last several intelligible. What is surprising
into their ears while they speak). Others focus on is that clutterers (unlike stut-
seconds, often with an uncertain rhythm.
altering the stutterer's breath control, or develop terers) are often unaware of
• The speaker prepares to articulate a sound, but
techniques in which speech comes more slowly or the problems caused by their
is unable to release it. In severe cases, facial
evenly than normal. Learning to relax is an essen- non-fluency.
spasms and sudden body movements may be There is no clear explana-
tial feature of many methods, as is learning to inter-
used in an effort to get over the 'block'. tion for the handicap, though
act with others (especially in the situations that many researchers accept
• Extra words are introduced at points of diffi-
cause particular tension). These days, particular that it may well have a physi-
culty, e.g. oh, gosh.
attention paid to helping the stutterer develop
is cal basis. Electrical record-
• Words show erratic stress patterns, and there ings of brain activity often
a style of behaviour that more closely resembles
is an abnormal intonation and speed of speech show significant abnormali-
that of fluent people; many stutterers, it must be ties, for example. A possible
(§29).
remembered, have become so isolated and with- theoretical explanation is an
• Words and phrases may be left unfinished.
drawn on account of their stutter, that they may inadequate neural 'pace-
• Speakers may avoid words and phrases that con- maker' at the neurophysiolo-
need to be taught a new way of life (or at least,
tain the sounds they find difficult, and replace gical level in the motor con-
way of looking at life) as therapy proceeds.
speech (p. 262); but
these by circumlocutions. One stutterer, who trol of
there as yet no empirical
is
had great difficulty with [p], would always re-
support for this hypothesis.
place policeman by 'officer of the law'.

A amount of 'normal non-fluency' is

certain
found young children (especially around the age
in
of 3), and indeed everyone is prone to hesitation,
especially in situations where they have to speak
under pressure. Stutterers too vary greatly in the
DAF
control they have over their speech, and clinicians
The equipment used in a being forced to speak more stuttering severity, delay
therefore look closely at the contexts that most pro- 'delayed auditory feedback' slowly, which itself helps time, feedback intensity)
mote a stutter when they are investigating the pro- (DAF) system. The control a stutter. A great before clinicians will be in

blem. very difficult to draw a clear line between

It is speaker talks into a micro- deal more research needs a position say why some
to
phone attached to a tape to be carried out into the stutterers benefit, and
normal speech and stuttering, though there is no
recorder. A device in the variables involved (e.g. others do not.
mistaking the handicap in its severe form, with its
recorder delays the voice
uncontrolled, tense, and irregular speech, and the for a fraction of a second,
anxiety and embarrassment (for listener as well as and then plays it back
speaker) that is invariably present. through the headphones
sufficiently loudly that it be-
comes the dominant sound
Theories and treatment
to be heard. When this is
Many theories of the origins of stuttering have been done to normally fluent
proposed, but there is no current consensus. people, many of them start
to stutter. The technique is
Doubtless the many variations between stutterers
sometimes used with stut-
obscure the existence of several contributing causa- terers, and in some cases
tive factors. Physical factors have often been impli- it has a positive (but tem-

cated, such as mixed cerebral dominance (p. 258), porary) effect.

a specific left-hemisphere deficit, or a defect in the Why it should help is un-

clear: the stutterers may be
feedback mechanism between ear and brain (which
benefitting from having
reduces a person's ability to monitor output effi- normal speech
their
ciently). There have been several 'psychoneurotic' masked, or they may be
theories, which attempt to relate stuttering to the

278 VIII LANGUAGE, BRAIN, AND HANDICAP


Language delay
Most of the children seen in a speech therapy clinic
show some kind of delay in their development of
spoken language. In school, where there is usually
an accompanying - and often more dramatic -
delay in the acquisition of the written language (p.
274), estimates of prevalence vary from 2% to as
much as 15%. Delays range from a barely notice-
able few months to one of several years. In some
conditions, teenage children and young adults may
stillbe using a kind of spoken language equivalent
to that found in 1- or 2-year-olds.
In about a third of cases, the reason for the
delayed language development is known. Mentally
handicapped children, for example, display some
of the most marked delays. In a study of 1,381
severely sub-normal children between the ages of
3 and 16, over 40% of the 16-year-olds were
As with normal language acquisition, develop-
mental language handicap needs to be investigated
in terms of both production and comprehension:
the children are regularly referred to as having pre-
dominantly 'expressive' or 'receptive' problems
(pp. 232, 265), and of course there are many
'mixed' cases. Within each category, also, the pro-
blems may be located under any of the headings
recognized in models of language structure and use
(§13). Phonological handicaps manifest themselves
insuch areas as poor rhythmical ability, the persis-
tence of immature processes of sound formation,
and difficulties in auditory discrimination. Gram-
matical handicaps are shown by a restricted range
of sentence constructions, uncertain control of word
order, and the avoidance of particular features in
morphology and syntax (e.g. omitting word end-
ings to be in such
on nouns and verbs, or the verb
sentences as The man
happy). Semantic handi-
is

unable to use grammatical constructions; and caps are mainly noticeable in a limited vocabulary.
17i% had not even reached the one- word stage And there are several kinds of pragmatic handicap
(§41) (W. Swann &
P. Mittler, 1976). Other (§21), whereby children fail to make use of norms
groups where the language delay is part of a more of interaction appropriate to their age — for exam-
general problem include those who are deaf, psy- ple, an inappropriate use of questions, or an inabi-
chologically disturbed, autistic, or physically lity to 'keep to the point'.

handicapped. Accidents, strokes, and other inci-

dents result in acquired language problems (p. 270)
for a small number of children.
In the majority of cases, however, there is no
clear physical reason for the language delay. The
children have no relevant medical history, are of
normal intelligence, and are not socially deprived Grammatical handicap
This 32-year-old boy (C) C: - A: stops the fire.
or emotionally disturbed. Nonetheless, their lan- It

guage is well behind that of their peers. In several

had several phonological A: Hm? C: Yes.
problems, as can be seen C: 2 syllables A: Mhm. And what do the
instances, there are accompanying difficulties of a from his unintelligible utter- A: Where does it go? Does men do?
cognitive or social kind — such as a general auditory ances (marked as syll- it go to a fire? C: On ladder here.
imperception, poor memory, poor concentration, ables), but his main difficul- C: Yes. A: On the ladder.
ties were a failure to A: What does do, when C: Yes.
or a reluctance to cooperate with others. However,
it

develop a normal gramma- itgets to the fire? A: What do they do on the

not all delayed children display such problems: for tical system. His short sen- ladder?
C: Ladder.
many, the language difficulty is the primary or only tences are more typical of A: There's a ladder, yes, C. Go up.
symptom. And even when other problems are pre- a 1 \- to 2-year-old. He also and what does the ladder A: They go up, yes. And
had some difficulty compre- do? when they get to the top
sent, it is never easy to determine whether one fac-
hending some of the thera- C: Go -up. what do they do?
tor (such as memory) is the 'cause' of the language -
pist's (A) questions, and he A: Goes up, yes. And then, C:
delay. hardly ever produced a when the ladder's up, what A: Hm?
sentence spontaneously. do the men do? C: -
'Aphasic' children A: What does the fire en- C: Water -out. A: Do they jump off?
gine do? A: Oh. And what does the C: Yes. No. Come down.
These are the children who are sometimes called C: syllable water do? (D. Crystal et al., 1976,
'developmental^ aphasic' (or 'dysphasic') - labels A: Where does it go? C: 4 syllables Stop -fire. p. 142.)

that are somewhat controversial. The terms derive

from an analogy with the linguistic symptoms of Pragmatic handicap
people who have suffered brain damage (p. 270), This 10-year-old boy (C) speech therapist (A) about X?
but they have been criticized on the grounds that has no real problems with a forthcoming sports day at C: There is not. There is a
segmental phonology or his school. sports day in Y (= his own
the developmental condition raises quite different
grammar, and he has a A: Which race would you school).
issues. Here, there is no evident brain damage, and
fairly good vocabulary. His like to be in? A: Then what's X got to do
the associated problems are best defined in psycho- main handicap an inabi-
is C: Ilike to be in X(= a with it?

logical, social, and educational terms, and not lity to organize and main- town several miles from the C: Nothing.
neurologically. It has also been argued that the tain a conversation; his school) in the sports day. A: Then why did you men-
answers are often said to A: InX? tion it?
linguistic symptoms are different: language delay,
be 'bizarre' and his thinking C: Yes. C: Indeed I did mention it.

typically, displays few usages that fall outside the confused', as illustrated by A: What do you mean? A: Why did you mention it?

range of normal language development; aphasic the following extract from a C: I mean something. C: I don't know.

disturbance, typically, contains a great deal of ab- conversation with his A; Is there a sports day in (M. McTear, 1985, p.246.)

normal, 'deviant' construction.

46 LANGUAGE HANDICAP- 279

Alternative communication Braille puter-assisted systems a considerable amount of
systems Braille is a system de-
the use
that turn text into braille. training beforeit can be ef-

signed to facilitate The typed on a stan-

text is fectively used.However,
For many disabled people unable to make use of language by blind
of written dard typewriter keyboard, there are now over 1 0,000
normal modes of communication, alternative sys- people. consists of a se-
It and is made available for Optacons throughout the
tems have been devised. Some are designed to sup- quence of cells, each of editing before the final ver- world, used on many writ-
which contains a 3 x 2 ma- sion is embossed. Other ing systems - even includ-
plement normal language use ('augmentative'
trix embossed dots. In
of possibilities for speeding ing Japanese kanji (p. 200).
systems); others completely replace it ('alternative' Grade 1 braille, the cells re- up the process of text pre- Talking books' - texts
systems). In the simplest cases, the letters of the present letters, numbers, paration include the deve- recorded on audio cas-
written language are coded into a different med- punctuation marks, and se- lopment of optical char- settes - have been avail-
veral short words and con- acter recognition, which able to the blind for some
ium, such as the use of braille by the blind. In other
tractions. In Grade 2 'reads' an orthographic time. A more recent deve-
cases, the complex grammar and vocabulary of braille, words are abbre- text, and converts it into a lopment is the Kurzweil
everyday language is simplified to meet the needs viated using the symbols digital form capable of be- Reading Machine - a de-
of the disabled person. Several new systems of available from Grade 1 to ing used as input to a vice that reads printed
manual signs (e.g. Amerind, §37), tokens, or pic- produce a kind of shorthand braille transcription system. characters and converts
system (p. 207). There are also several them into synthesized
torialsymbols have now been invented with this
Because of the time it non-braille devices. With speech using a system of
consideration in mind. takes to transcribe texts the Optacon, the fingers phonological rules (§28).
The proliferation of systems, techniques, and into braille, and because of rest on a set of vibrating Many typefaces can be
aids reflects the range of handicapping conditions, the size of brailled books rods that provide a tactile read at speeds of over 200
(often 20 times the size of 'image' of the text. The sys-
and the existence of many individual differences. words a minute, which is
a normal printed text), ma- tem permits a more faithful only a little below average
At one extreme, there are physically handicapped medium have representation of the text
terials in this human reading speed
people with high intelligence and minimal motor been very limited. This sit- than does braille, and can 210).
(P.
skills (as with many of the cerebral palsied); at the uation is already changing be used for all kinds of
other, there are mentally handicapped people with with the advent of com- printed material. requires

RBCDEFGHIJKLMNOPQR
It

very low cognitive levels and normal motor skills.

Most equipment has been designed with the needs a •
of the former group in mind — people who may
be able to control the movements of only one part
of their body, such as their eyes. But far more is
STUUWXVZ and for of the with
involved than the development of 'hardware'. The
content of the system has to be sufficiently simple
to permit easy learning, yet have enough potential
to enable linguistic skills to develop. A difficult task Symbol systems
is to decide exactly what vocabulary will be most Several communication 1970s by Charles Bliss tions, including the cerebral

systems have been de- (1897-1985). Bliss, a che- palsied, mentally handi-
useful to people at the outset (some selections are
vised using pictures or ob- mical engineer, aimed to capped, and autistic.
shown below), and how to display it (a particular jects of varying size, devise a set of symbols Currently, a great deal of
problem with large communication boards, which shape, and colour as logo- that could be translated' research is being devoted
may contain over 1,000 items). graphic symbols (p. 200). into any language (as could to investigating the

Two basic techniques are employed. In 'direct • Rebus symbols, used in the symbols of chemistry). strengths and weaknesses
selection', users type, touch, or point to the element
some early writing sys- Some of these symbols are of the various systems, in

tems, and widely known illustrated below on a stan- relation to the needs of in-
of the message they wish to communicate, using through their use in chil- dard chart, along with the dividual children at different
any mobile body part to operate a variety of key- dren's comics, are found in written words. The ap- stages of their communica-
boards, keypads, switches, joysticks, and pointing some approaches to early proach has been used with tive development.

devices (e.g. a stick attached to the head). For those reading in mentally handi- a variety of clinical popula-
capped children. For exam-
who have some speech, voice-activated systems are ple, a picture of a child ^J* ~lo
also becoming available. The element is then dis- feeding himself ('eat')

played on a screen, printed out, or speech synthe- might be combined with the 3 szc
sized. In 'scanning', they wait while someone or produce the
letter 'h' to
symbol for heat (p. 65).
something systematically scans a set of possible M cc
• A system known as
message elements (e.g. letters of the alphabet), and 'Non-sup' (= Non-Speech
h \
they indicate when the right element is reached. Language Initiation Pro-
':
This can be a very slow process, but it can be gram'), based on the work
tY
speeded up in various ways, such as by using a of David Premack (p. 398),
was devised in the 1 970s
two-stage scan, in which a group of elements is for use with the mentally
Gc;* *
first identified, and the specific element in that handicapped. This made
~
~*
group subsequently singled out (as with etran use of plastic symbols in
-

frames, p. 281). various abstract shapes as

a means of teaching syn- h
The new 'information technology' promises to
tactic sequences.
change the lives of many people who previously
had negligible communicative ability. Microelec-
tronic techniques make the new systems faster,
more powerful, easier to use, and more portable.
• Blissymbolics
supplement to speech de-
veloped in Canada in the
is a visual

ra
280 • VIII LANGUAGE, BRAIN, AND HANDICAP
 .

Possums ETRAN frames

Several communication aids can be controlled by a An etran eye-gaze chart
simple switching arrangement, activated by a mobile for use by severely handi-
part of the body, such as the arm, foot, eyebrow, chin, capped children. This is a
or mouth. The picture shows one of the scanner two-movement system: the
systems devised by Possum Controls Ltd. first eye movement indi-

cates the group where the

required element is to be
found, and the second indi-
cates its specific location
within the group.

The Canon Communicator

This device is a very compact, portable communication aid that straps to the wrist.
The messages are printed onto a paper strip.

The Phonic Ear

Inrecent years, speech synthesis has added to the
communication boards. The first
capabilities of
portable system in this field was the Phonic Mirror
Handivoice, a 128-square direct selection aid
introduced in 1 978. The picture shows one of the new
generation of Phonic Ear products, the Vois 130. The
user can make four selections of preprogrammed
words, or can select to communicate in phonemes
(§28) or letters (thus allowing the compilation of an
indefinitely large vocabulary). A few affixes and
phrases are also built in to the system (e.g. un-, -ing,
My name is .). The three-part system consists of a
.

synthesizer, a speaker, and a touch-sensitive display

board. There are also smaller versions, with fewer keys
and a numeric keyboard, intended for use by people
with poorer motor ability and higher cognitive skills. The Portaprinter
This was the first fully port-
able scanning aid. The de-
vO>s vice scans rows and col-
umns, enabling the user to
make a selection when an
appropriate symbol is
reached. The letters are ar-
ranged to common
form
patterns, with some re-
peated. The output is
printed on a paper strip.

46 LANGUAGE HANDICAP •
281
I
mmss^sBw-oh

WKfCESd

I RE7RIJ

<_ .^.i:<^^'>>>.. <y^,?-v-* ::-::.^3s^:$^

 .

PARTEK
The languages of the world

Aakwo, Anus, Bella Coola, Blood, Bok, Deerie, Comparative philology is the branch of language
Gold, Grawadungalung, I, Kukukuku, Lule, Mar- study which first looked as far back as possible
raawarree, Mimika, Ngeq, Nupe, Ok, Ron, Santa, into the history of language. Out of this study
Shiriana, Tiini, Tzotzil, U, Yangman, Zyrian . . comes the basic classification of the world's lan-
The litany of the world's lesser-known languages guages into families. We describe the way in which
reads like a mixture of mad invention and poetic families of languages are set up, and the alternative
inspiration. Over 22,000 names of languages, dia- methods of classification that have been devised.
lects and have been collected, as part of the
tribes We then outline the languages of the world, family
task of linguistic identification and classification. by family, beginning with Indo-European, and end-
In the first sections of this part, we look at the ing with some of the languages which do not fit
problems which would face anyone embarking on neatly into any of the families - the so-called 'iso-
this task. How many languages are there? How lates'.
many speakers do these languages have? Such ques- When we account for linguistic diversity,
try to
tions are not easy to answer, and it is important it is the phenomenonof language change which
to know why. is the central fact to be identified and explained.

The multiplicity of languages leads naturally to The remaining sections therefore describe the many
a consideration of the reasons for the diversity. ways in which language can change, and discuss
What are the origins of human language? In the the various reasons which have been proposed to
next sections, we review the ancient speculations account for this process. Finally, we look at some
concerning the origins of language, and the special cases which illustrate the ways in which
attempts by scholars in the 20th century to say social forces can promote new language growth
something sensible about the matter, using tech- - the world's pidgins and Creoles.
niques borrowed from palaeontology and other
sciences.

The Tower of Babel, 1 3th century, in the vaults of the Basilica

di S. Marco, Venice.
47 How many languages?
There is no agreed total for the number of lan- there were thought to be over 1,000 Indian lan-
guages spoken in the world today. Most reference guages in Brazil; today, there are fewer than 200.
books give a figure of 4,000 to 5,000, but estimates
have varied from 3,000 to 10,000. To see why there LANGUAGE - OR DIALECT?
is such uncertainty, we need to consider the many For most languages, the distinction between lan-
problems facing those who wish to obtain accur- guage and dialect is fairly clear-cut (p. 25). In the
ate information, and also the reasons (linguistic, case of English, for example, even though regional
historical which preclude a simple
and cultural) vocabulary and local differences of pronunciation
answer to the question 'What counts as a lan- can make communication difficult at times, no-one
guage?' disputes the existence of an underlying linguistic
unity that all speakers identify as English, and
which isconfirmed by the use of a standard written
language and a common literary heritage. But in
Discoveries hundreds of cases, considerations of this kind are
in conflict with each other, or do not clearly apply.
An obvious reason for the uncertainty over
The best-known conflicts occur when the criteria
numbers is that even today new peoples, and there- of national identity and mutual intelligibility do
fore languages, continue to be discovered in the
not coincide. The most common situation is one
unexplored regions of the world - especially in the
where two spoken varieties are mutually intellig-
Amazon basin (as the Transamazonica road system
ible, but for political and historical reasons, they
is extended), Central Africa, and New Guinea.
are referred to as different languages. For example,
However, only a few languages are likely to be using just the intelligibility criterion, there are
encountered in this way; and it is much more usual
really only two Scandinavian languages: Continen-
to find parts of the world where the people are
tal (Swedish, Danish, and two standard varieties
known, but the languages spoken in their area are of Norwegian) and Insular (Icelandic, Faeroese).
not. There are in fact many countries where linguis-
Swedes, Danes, and Norwegians can understand
tic surveys are incomplete or have not even begun.
each other's speech, to a greater or lesser extent. Number of speakers
It is often assumed that the people speak one of
But as soon as non-linguistic criteria are taken into Number of speakers of the
the known languages in their area; or that they world's languages, based on
account, we have to recognize at least five lan-
speak a dialect of one of these languages; but upon Voegelin and Voegelin's
guages. To be Norwegian is to speak Norwegian;
investigation their speech is found to be so different Classification and Index of
to be Danish is to speak Danish; and so on. In the World's Languages. Lan-
that it has to be recognized as a separate language.
such cases, political and linguistic identity merge. guages which they classify
And there are many other similar cases where poli- as of uncertain existence
(marked by a question-mark,
ALIVE OR DEAD? tical, ethnic, religious, literary, or other identities
in theirbook) have been
Against steady increase in the world language
this force a division where linguistically there is little excluded. In about three-
total, there is a major factor which decreases it. difference — Hindi vs Urdu, Bengali vs Assamese, quarters of the cases where
For a language to count as 'living', there obviously Flemish vs Dutch, Serbian vs Croatian, Twi vs no numerical estimates are
available, the numbers of
have to be native speakers alive who use it. But Fante, Xhosa vs Zulu.
speakers are extremely
in many parts of the world, it is by no means an The opposite situation is also quite common. small. The total number of
easy matter to determine whether native speakers Here we find cases where spoken varieties are languages (including extinct
are still living - or, if they are, whether they still mutually unintelligible, but for political, historical ones) is 4,522.
use their mother tongue regularly.
The speed with which a language can die in the 2400 2269
smaller communities of the world is truly remark- 2200
able. The Amazonian explorations led to the dis- 8 2000
covery of many new languages, but they also led <f
1800
3i 1600
to their rapid death, as the Indians became swal-
jj 1400
lowed up by the dominant western culture. Within Cg 1200
a generation, all traces of a language can disappear. Z 1000
Political decisions force tribes to move or be split -£ 800 59-
a 600
up. Economic prospects attract younger members z 400 409
258
away from the villages. New diseases take their 200 138 [43
toll. Trumai, spoken in a single village
In 1962,
on the lower Culuene River in Venezuela, was More than More than More than More than More than Known to No
1,000,000 100,000 10,000 1,000 100 be extinct estimate
reduced by an influenza epidemic to a population
of fewer than 10 speakers. In the 19th century, Number of speakers

284 •
IX THE LANGUAGES OF THE WORLD
or cultural reasons they are nonetheless called vari-
eties of the same language. The three main 'dialects'
of Lapp fall into this category, for example. Chi-
nese is a case where linguistic criteria alone are
in conflict with each other. From the viewpoint
of the spoken language, the many hundreds of dia-
lects in China can be grouped into eight main types
(p. 312), which are mutually unintelligible to vari-
ous degrees. But speakers of all these dialects share
the same written language tradition, and those who
have learned the system of Chinese characters are
able to communicate with each other. Despite the
linguistic differences, therefore, Chinese is consi-
dered by its speakers to be a single language.
In the above cases, the languages in question
have been well studied, and many speakers are
involved. When languages have been little studied,
or have very few speakers, it is much more difficult
for linguists to interpret all the factors correctly.
For example, when two languages are in close prox-
imity, they often borrow words from each other
- sometimes even sounds and grammar. On first
acquaintance, therefore, the languages may seem
more alike than they really are, and analysts may
believe them to be dialects of the same language.
This has proved to be a real problem in such parts
of the world as South America, Africa, and South-
east Asia, where whole groups of languages may
be affected in this way. Similarly, decisions about
how to analyse all cases of dialect continua (p. 25)
will affect our final total of languages.
LANGUAGE NAMES
A big problem, in working on lesser-known lan-
guage areas, is deciding what credence to give to
a language name. This issue does not arise when
discussing the main languages of the world, which
are usually known by a single name that translates
neatly into other languages - as in the case of
Deutsch, German, Tedesco, Nemetskiy, and Alle-
mand, for instance. But in many cases the situation
is not so straight-forward.
At one extreme, many communities have no
specific name for their language. The name they
use is the same as a common word or phrase in
the language, such as the word for 'our language'
or 'our people'. This is often so in Africa (where
the name Bantu, which is given to a whole family
of languages, means simply 'people'), and also in
Meso- and South America. In the latter areas, we
find such examples as Carib = 'people', Tapuya
= 'enemy', and Macu = 'forest tribes'. Some tribes
were called chichimecatl ( = 'lineage of dogs'),
chontalli (='foreigners') or popoloca (=
'barbarians'), and these labels led to the modern
language names Chichimeca, Chontal, and Popo-
loca. Frequently, the name is the same as a river
on which a tribe has been observed to live, as with
the many groups of Land Dayak, in the West Indo-
nesian family. In several Australian aboriginal lan-
guages, the name for the language is the word for
'this': for example, the nine languages within the
Yuulngu family are known as Dhuwala, Dhuwal,
Dhiyakuy, Dhangu, Dhay'yi, Djangu, Djinang,
Djining, and Nhangu. Asking native speakers what
language they speak is of little practical help, in
such circumstances, if they only answer 'this'!
At the other extreme, it is quite common to find
a community whose language has too many names.
A South American Indian tribe, for instance, may
have several names. A tribe, first of all, will have
a name for itself (see above). But adjacent tribes
may give the people a different name (e.g. Puelche
means 'people from the east' in Araucanian). The
Spanish or Portuguese explorers may have given
them a third name - perhaps a characteristic of
their appearance (e.g. Coroado means 'crowned'
in Portuguese). More recently, anthropologists and
other investigators may have used another name,
often based on the geographical location of the
tribe (e.g. 'up-river' vs 'down-river'). And lastly,
the same language may be spelled differently in
Spanish, Portuguese, English, or in its own writing
system (if one has been devised). For example,
Machacali, spoken in Minas Gerais, Brazil, is
sometimes spelled Maxakali, sometimes Maxa-
kari. When the initial letters vary (as when the Peru-
vian language Candoshi is spelled Kandoshi),
indexing is especially awkward.
There are further complications. Sometimes, the
same name is applied to two different languages,
as when mexicano is used in Mexico to refer to
Spanish (otherwise known as espanol or castellano)
and to the main Indian language (nahuatl). Some-
times, speakers from different backgrounds may
disagree about whether their ways of speaking
should be related at all. Speakers of Luri, spoken
in south-west Iran, say that their speech is a dialect
of Persian; speakers of Persian disagree. Asking the
native speakers is evidently no solution, for their
perceptions will be governed by non-linguistic con-
siderations, especially of a religious, nationalistic,
or socioeconomic kind.
TO CONCLUDE
When all these factors are taken into account, it
is plain that there will be no single answer to the
question 'How many languages?' In some parts of
the world, there has been a tendency to over-esti-
mate, by taking names too literally and not group-
ing dialects together sufficiently — the Malayo-
Polynesian languages are often cited in this connec-
tion. In other places, the totals are likely to have
been underestimated — Indonesian languages, for
example. There are over 20,000 language or dialect
names listed in the Voegelins' great Classification
and Index of the World's Languages (1977), and
these have been grouped into around 4,500 living
languages. Since the publication of that work, the
total must have become somewhat less, in view
of the trend indicated in the table of population
estimates on p. 284; but it seems unlikely that it
should be less than 4,000.
47 HOW MANY LANGUAGES? •
285
48 How many speakers?
Estimating the number of speakers of a living lan- immigrant family from Pakistan, where
in a British
guage is, if anything, more complicated than esti- in addition to theirmother tongue, the mother's
An extract from the 1 986
mating the number of languages. Any language English might be poor, the father's moderate, and
Canadian census form. The
which continues to be spoken in an area counts the teenage children's good, or even fluent. How two language questions
as 'one', in our language total. But the number of does one cater for the situation in Scotland, where focus on speaking, but listen-
its speakers may vary wildly, from one decade to there is enormous variation in the fluency with ing comprehension is implicit
the way question 1 9 is
the next. A contemporary example is Vietnamese, which people command Gaelic? How does one
in

phrased. The Census Guide

where the movements of population since the Viet- cater for the situation in countries where English that is sent out along with the
nam War have led to dramatic differences, in both is an official language (e.g. India, Nigeria), but the form stresses the issue of
and the places they are to
the totals of speakers, population has achieved various levels of profi- ability: 'Do not report French

be found. An even more recent example is the mas-
sive changes that have affected people in Ethiopia
and the surrounding regions during the famine of
the early 1980s. The combined effects of large-scale
loss of life, refugee migrations, and local govern-
ment relocation policies mean that all estimates are
now seriously out of date. It will be
before accurate census information becomes avail-
able, and in the meantime, numerical data from
ciency? How does one allow for people who have or English learned at school
unless you can conduct a
only a reading knowledge of a language, or who
conversation of some length
use a language only for special purposes (§63)? on various topics in that lan-
There are innumerable complications of this kind guage.'
facing the linguistic demographer. By contrast, the 1981 cen-
sus form used in England
Moreover, how trustworthy is the information
had no language question at
many years about language ability obtained through a census all: respondents were asked
or other official report? There are many parts of to state their country of birth
the world where language estimates may be only.

the area must be viewed with extreme caution.

Trends in the world demographic situation will 1* M ;'

^anaca
a 3M M
Canada
* Canada
obviously be a major factor in any speaker esti-
mates. The world population is currently increas-
ing at a rate of nearly 2% per annum. In mid 1987,
the number of people was estimated to be over I T]J 1986 Census of Canada
complete Please your questionnaire on
5,000 million. A UN
projection to the year 2000 UJ June Tuesday. 3. 1986

indicates a world population of over 6,000 million.

In these circumstances, any estimates of speakers,
especially of the languages of the less-developed
1* Canada Canada Canada
countries, are going to be well out of date as soon
as they are printed. In the case of the many lan-
guages spoken within the Indian sub-continent,
most of the available estimates were made in the
6 Jj Recensement du Canada de 1986
1970s; but as the population of India increased
\ A remplir le mardi 3 |uin 1986

by 25% between 1971 and 1981, the figures need

to be interpreted with a great deal of caution. All
18. Quelle langue pari 18. What language do you yourself speak al home now?
the estimates in Appendix HI are subject to this • (If more than one language, which language do you speak
(Si vous en parte z
proviso. (Consultez le Guk. most often?) (See Guide)
Even if a population is stable, it is rarely easy
to obtain accurate information about number of ' O Anglais English

speakers. This obviously a problem in the less

is U Francais C French
Italian
accessible parts of the world. Less obviously, it is •»< Italien -

a problem which can affect any country, especially tt [D Chinois [j Chtnese

where minority groups are involved. The informa- n Alemand D German

tion might be obtained from a census, but not all

census forms contain questions about linguistic Other (spacify)
background. Some countries do not think the mat-
ter important; others would like to know, but find 1 9. Can you speak English or French weM enough to conduct a
19. Connaissez-vous E • conversation? (See Guide)
it difficult to phrase the questions in such a way conversation? (Ct
that they do not offend minority groups. Cochm une seute Mark one box only
It is in any case difficult to use a questionnaire
• )D L anglais m D English only
to establish the facts of language use (p. 26). In
O Le francais •
Q French only
many of the more multilingual communities in the
D Both English and French
'
CD L' angiais et
•

world, not even easy to answer the simple ques-

it is

tion, 'What is your mother tongue?' Nor is it easy

'. i
angiais '.
D Neither English nor French

to allow for such diverse cases as might be found

286 IX THE LANGUAGES OF THE WORLD

inflated, because people who may have relatively
little command of a language claim they speak it, Fivet ypes of rel< ationship between dialect and language
in order to support the cultural or political cause
with which the language is associated (e.g. Breton, Typel
Community Mutually intelligible Community
Welsh). Equally, there are many places where it
A Common cultural history B
is unspoken government policy to underestimate

the number of speakers of a minority language, 1

in order to play down the political significance of
e.g. British „ American
Same language ,,

English English
that section of the community (e.g. Breton, Xhosa).
Official figures, in such circumstances, should not
be taken too literally. Type 2
Uncertainty over the name of a language does Community Mutually unintelligible Community
not help, and this is very common (p. 285). Often, A Different cultural history B

the name of a dialect, the name of a language, and

r
the name of a language family become confused * 1 Una
— as if someone were to say 'He doesn't speak Eng-
lish; he speaks Cockney', or '... he speaks Ger-
manic'. In a little-known area of the world, the
Type 3
reality of the situation may not be at all obvious. Community Mutually intelligible Community
For example, Kru, Chin, Kachin, Dayak, Teso, A Different cultural history B
Nuer, and Mongo-Nkundu are names that have 1
been used both for single languages and for whole e.g. Norwegian - *- Danish
groups of languages, and speaker numbers are dra-
matically altered depending on which perspective
one adopts. Mandara is a Chadic name which is
Type 4
sometimes used for a single language and some- Community Mutually unintelligible Community
times for a group of related languages. If the former A Same cultural historv B
interpretation is taken, the speakers number some
1
30-40,000; if the latter, the total is nearly 400,000. e.g. Cantonese - 9 * Hakka
Approximations and uncertainties are thus the (Chinese) (Chinese )

norm in language estimates, especially for those

languages which are rapidly expanding, and those
which are in a state of serious decline. In Appendix TypeS
Community Partially (un)intelligible Community
III, where speaker estimates are given for 1,000
A Overlapping cultural history B
languages, it has often not been possible to choose
between conflicting totals; and in these cases, both
i :.g. Turkish * — 1
> .... _ .
- Uzbek
upper and lower totals are listed.

Speaker estimates for the world's

Indo-European 2,000,000,000
Sino-Tibetan 1 ,040,000,000
The top twenty languages top 20 languages (given in millions).
The first column lists the languages
Niger-Congo 260,000,000 on the basis of the number of mother-
Official language
Afro-Asiatic 230,000,000 tongue (first-language) speakers they
Mother-tongue speakers populations
Austronesian 200,000,000 have. When conflicting estimates
Dravidian 140,000,000 1. Chinese (1,000) 1. English (1,400) have been given (cf. Appendix III), the
Japanese 120,000,000 2. English (350) 2. Chinese (1,000) higher one has been used here. The
Altaic 90,000,000 3. Spanish (250) 3. Hindi (700) second column gives population esti-
Austro-Asiatic 60,000,000 4. Hindi (200) 4. Spanish (280) mates for those countries where the
Korean 50,000,000 5. Arabic (150) 5. Russian (270) language has official status. Note that
Tai 50,000,000 6. Bengali (150) 6. French (220) the totals do not always coincide,
Nilo-Saharan 30,000,000 7. Russian (150) 7. Arabic (170) since some major languages (such as
Amerindian 8. Portuguese (135) 8. Portuguese (160) Javanese and Telugu) are not official
(North, Central, South America) 25,000,000 9. Japanese (120) 9. Malay (160) languages of whole countries, and
Uralic 23,000,000 10. German (100) 10. Bengali (150) some languages (such as Malay and
Miao-Yao 7,000,000 11. French (70) 11. Japanese (120) Tagalog) are official languages of
Caucasian 6,000,000 12. Panjabi(70) 12. German (100) multilingual countries. The second-
Indo-Pacific 3,000,000 13. Javanese (65) 13. Urdu (85) column figures are often over-
Khoisan 50,000 14. Bihari(65) 14. Italian (60) estimates, as by no means everyone
Australian aborigine 50,000 15. Italian (60) 15. Korean (60) in the countries where a second
Palaeosiberian 25,000 16. Korean (60) 16. Vietnamese (60) language is recognized (e.g. India)
17. Telugu(55) 17. Persian (55) will be fluent in it; on the other hand,
Family statistics Estimated numbers of speakers in
18. Tamil (55) 18. Tagalog(50) the figures are of some interest as
the main language families of the world in the early 19. Marathi (50) 19. Thai (50) indicators of the way languages are
1 980s, when the world population was well over 4,000 20. Vietnamese 20. Turkish (50)
(50) moving.
million.

48 HOW MANY SPEAKERS? 287


49 The origins of language
For centuries, people have speculated over the ori-
gins of human language. What is the world's oldest
spoken language? Have all languages developed
from a single source? What was the language
spoken in the Garden of Eden? How did words
come to be, in the very beginning? These questions
are fascinating, and have provoked experiments
and discussion whose history dates back 3,000
years. The irony is that the quest is a fruitless one.
Each generation asks the same questions, and
reaches the same impasse — the absence of any evi-
dence relating to the matter, given the vast, distant
time-scale involved. We have no direct knowledge
of the origins and early development of language,
nor is it easy to imagine how such knowledge might
ever be obtained. We can only speculate, arrive
at our own conclusions, and remain dissatisfied.
Indeed, so dissatisfied was one group of 19th-cen-
tury scholars that they took drastic action: in 1866,
the Linguistic Society of Pans published an edict
banning discussion of the topic at their meetings.
But the theorizing continues, and these days there
is a resurgence of interest, as new archaeological
finds and modern techniques of analysis provide
fresh hints of what mav once have been.
Early 'experiments'
The lengths to which some people have gone in
order to throw light on the question are truly
remarkable - if the accounts are to be believed.
One of the best-known reports concerns the Egyp-
tian king, Psamtik I, who reigned in the 7th century
bc. According to the Greek historian, Herodotus,
Psamtik wished to find out which of all the peoples
of the world was the most ancient. His way of
determining this was to discover the oldest lan-
guage which, he thought, would be evidence of the
oldest race. This is how Herodotus tells the story.
He gave two new-born babies of ordinary men to a shep-
herd, to nurture among his flocks after this manner. He
charged him that none should utter any speech before
them, but they should live by themselves in a solitary
habitation; and at the due hours the shepherd should
bring goats to them, and give them their fill of milk,
and perform the other things needful. Thus Psammeti-
chus did and commanded because he desired, when the
babes should be past meaningless whimperings, to hear
what tongue they would utter first.
And these things came to pass; for after the shepherd
had wrought thus for a space of two years, when he
opened the door and entered in, both the babes fell down
before him, and cried becos, and stretched out their
hands. Now when the shepherd heard it for the first
time, he held his peace; but when this word was often-
times spoken as he came to care for them, then he told
288 IX THE LANGUAGES OF THE WORLD
!
his lord, and brought the children into his presence when
he commanded. And when Psammetichus had also heard
it, he enquired which nation called anything becos; and
enquiring, he found that the Phrygians call bread by this
name. Thus the Egyptians, guided by this sign, confessed
(f
mmm
that the Phrygians were elder than they. That so it came
to pass I heard of the priests of Hephaestus in Memphis.
Phrygian is now extinct, but at the time it was
spoken in an area corresponding to the north-
western part of modern Turkey.
Psamtik's conclusion was wrong, for we know
from philological studies that Phrygian is but one
of several languages which had developed in that
period of history. So why did the children say
Psamtik of Egypt (663-
I
becos? Doubtless they had begun to babble natur- 610 BC).
ally and repetitively to each other, in a similar way
to twins (see p. 247), and this was one of the
'snatches' that the shepherd recognized. Some com-
mentators have even suggested that they were
imitating the sound of the goats
Whether the Psamtik experiment ever took place
is open to question. Possibly the origins of the story
lie in a fiction invented by someone to discredit
the Egyptians. But whatever the reality, the initia-
tive credited to Psamtik has apparently had its par-
allels in several later times and places. At least two
similar experiments have been reported — though
again, there are doubts as to their authenticity (see
also p. 228).
The Holy Roman Emperor, Frederick II of Hohenstau-
fen 94-1 250) also carried out an experiment with chil-
(1 1
dren. According to the chronicle of a Franciscan friar,
Brother Salimbene:
He made linguistic experiments on the vile bodies of hap-
less infants, bidding foster-mothers and nurses to suckle
and bathe and wash the children, but in no wise to prattle
or speak with them; for he would have learnt whether they
would speak the Hebrew language (which had been the
first), or Greek, or Latin, or Arabic, or perchance the ton-
gue of their parents of whom they had been born. But he
laboured in vain, for the children could not live without
clappings of the hands, and gestures, and gladness of
countenance, and blandishments.
James IV of Scotland 473-1 51 3) is said to have car-
(1
ried outa similar experiment. The account given in the
History of Robert Lindesay of Pitscottie runs as follows:
The king took a dumb woman and put her in Inchkieth,
and gave her two young children in company with her, ' '
1
and furnished them of all necessary things pertaining to
their nourishment, that is to say food, drink, fire and can-
dle, clothes, with all other kinds of necessaries which is
required to man or woman, desiring the effect hereof to
come to know what language the children would speak
when they came to lawful age. Some say they spoke good
Hebrew, but as to myself I know not but by hearsay.
Five theories of the origins of language
The Danish linguist, Otto natural sounds, the theory The 'ding-dong' theory approaches the breast, and nation for the latter would still

Jespersen (1860-1943), has little support. Speech arose because peo- bye-bye or ta-ta show the have to be found.
grouped commonly held ple reacted to the stimuli in lips and tongue respectively
theories about the origins of The 'pooh-pooh' theory the world around them, and 'waving' good-bye. The 'la-la' theory

language into four types, and Speech arose through peo- spontaneously produced Jespersen himself felt that, if
added a fifth of his own. They ple making instinctive sounds ('oral gestures') The 'yo-he-ho' theory any single factor was going
are often referred to by nick- sounds, caused by pain, which in some way reflected Speech arose because, as to initiate human language, it

names. anger, or other emotions. or were harmony with the

in people worked together, would arise from the roman-
The main evidence would be environment. The main evi- their physical efforts pro- tic side of life - sounds asso-

The 'bow-wow' theory
Speech arose through peo-
ple imitating the sounds of
the environment, especially
animalcalls. The main evi-
dence would be the use of
onomatopoeic words (p.
the universal use of sounds
as interjections (p. 91 ), but
no language contains many
of these, and in any case the
clicks, intakes of breath, and
other noises which are used
in this way bear little
dence would be the universal duced communal, rhythmical ciated with love, play, poetic
use of sounds for words of a grunts, which in due course feeling, perhaps even song.
certain meaning, but apart developed into chants, and But again, the gap between
from a few cases of apparent thus language. The main evi- the emotional and the ratio-
sound symbolism (p. 174), dence would be the universal nal aspects of speech ex-
the theory has nothing to use of prosodic features (p. pression would still have to
relation- commend it. Several fanciful 1 69), especially of rhythm; be accounted for.

1 as few of these
74), but
exist in a language, and as
languages vary so much in
the way they represent
ship to the vowels and con- examples have nonetheless but the gap between this kind
sonants found inphonology. been cited - mama is sup- of expression and what we
The spelling is never a satis- posed to reflect the move- find in language as a whole
factory guide. ment of the lips as the mouth is so immense that an expla-

CHILDREN OF THE WILD Most attempts them to speak failed. The

to teach
For several hundred years, cases have been reported cases of 1694, 1731, and 1767 (Fraumark) are said
of children who have been reared in the wild by to have learned some speech, and Tomko of Hung-
animals or kept isolated from all social contact. ary (also 1767) is reputed to have learned both

These cases are listed below, adapted from Lucien Slovak and German. The 1717 girl and the 19th-
Malson's Wolf Children (1972). Sometimes the century Bankipur child are both said to have
information is based on little more than a brief learned some sign language. But of the well-attested
press report. At other times, the cases have been cases, the results are not impressive. Victor, the
studied in detail - in particular, the stories of Vic- 'Wild Boy of Aveyron', remained unable to speak,
tor, Kaspar Hauser, Amala and Kamala, and though he could understand and read to some
Genie. extent. Kamala of Midnapore learned some speech
The ideas of Psamtik I receive no support at all and sign. The two most successful cases on record
from these children. Only some of the reports say are Kaspar Hauser, whose speech became quite
anything about the children's language abilities, advanced, and Genie (§46), who learned a few
and the picture is quite clear: none could speak words immediately after discovery, and whose sub-
at all, and most had no comprehension of speech. sequent progress in speech was considerable.

Recorded cases of child isolation

Date of Age at Date of Age at
discovery discovery discovery discovery

Wolf-child of Hesse 1344 7 Wolf-child of Sekandra 1872 6

Wolf-child of Wetteravia 1344 12 Child of Sekandra 1874 10
Bear-child of Lithuania 1661 12 Wolf-child of Kronstadt ? 23
Sheep-child of Ireland 1672 16 Child of Lucknow 1876 ?
Calf-child of Bamberg C1680 ? Child of Jalpaigun 1892 8
Bear-child of Lithuania 1694 10 Child of Batsipur 1893 14
Bear-child of Lithuania ? 12 Child of Sultanpur 1895 ?
Kidnapped Dutch girl 1717 19 Snow-hen of Justedal ? 12
Two boys of Pyrenees 1719 ? Amala of Midnapore 1920 2
Peter of Hanover 1724 13 Kamala of Midnapore 1920 8
from Sogny
Girl 1731 10 Leopard-child of India 1920 ?
Jean of Liege ? 21 Wolf-child of Maiwana 1927 ?
Tomko of Hungary 1767 ? Wolf-child of Jhansi 1933 9
Bear-girl of Fraumark 1767 18 Leopard-child of Dihungi ? 8
Victor of Aveyron 1799 11 Child of Casamance 1930s 16
Kaspar Hauser of Nuremberg 1828 17 Assicia of Liberia 1930s ?
Sow-girl of Salzburg ? 22 Confined child of Pennsylvania 1938 6
Child of Husanpur 1843 ? Confined child of Ohio 1940 ?
Child of Sultanpur 1843 ? Gazelle-child of Syria 1946 ?
Child of Sultanpur 1848 ? Child of New Delhi 1954 12
Child of Chupra ? ? Gazelle-child of Mauritania 1960 ?
Child of Bankipur ? ? Ape-child of Teheran 1961 14
Pig-boy of Holland 9 ? Genie, U.S.A. 1970 13J
Wolf-child of Holland ? ?

49 THE ORIGINS OF LANGUAGE •

289
 Another way of looking at the problem is to ask Primate vocal tracts
Scientific approaches whether primitive man had the physiological capa- The vocal tracts of primates

By contrast with the informal discussion and specu- city to speak, and this has led to a great deal of are very different from that of
interesting research. The problem is that only the modern man. They have
lation of preceding centuries, serious attempts have
long, flat, thin tongues, which
been made in recent years to see if modern science shape of the jaws and the oral cavity are preserved have less room move. The
to
can throw any light on the question of the origins in fossils; there is no direct information about the larynx is higher, and there is
of language. The study is sometimes called glosso- size and shape of the soft tissues of tongue, pharynx little sign of a pharynx. There
genetics — the study of the formation and develop- or larynx, nor about the ability to move these is no evidence of ability to
change the configurations of
ment of human language, in both the child and organs (§22). Most of the reasoning has therefore
the vocal tract, to produce
the race. The main sciences involved are biology had to be based on reconstruction using plaster the range of sounds required
(especially sociobiology), anthropology, psycho- casts, and comparison with the physiology and in speech. In the course of

logy, semiotics (p. 399), neurology (for the study
of brain evolution), primatology, and linguistics.
THE EVIDENCE FROM
PALAEONTOLOGY
Might it be possible to deduce, from the fossil
record of early man, the point at which speech
began? The matter has been well investigated, but
the results are not conclusive.
It is possible to make plaster casts of the bony
cavities within the skulls which have been found.
It can be shown, for example, that both Nean-
derthal man and Cro-Magnon man (pre-30,000
bc) had similar brain sizes to that of modern man.
But this information is of limited value. The relative
size and shape of the brain can be established, but
none of the more relevant detail (such as the orien-
tation of the various furrows, or sulci (§45)). In
any case, there is no direct correlation between the
size of a brain and the use of language: in modern
man, language is found in people with small brains,
such as nanocephalic dwarfs, or children who have
had large areas of brain removed — and some goril-
vocalization of present-day primates and human evolution, posture becomes
erect, and the head moves
infants.
forward. The larynx des-
It ispossible to say with some conviction, using
cends and a long, flexible
this kind of argument, that the older hominids did pharyngeal cavity develops.
not possess speech; but the position of the more (From V. E. Negus, 1949.)
recent remains is unclear. It is unlikely that Austra-
lopithecus (who appeared around 4—5 million bc)
could speak, but the evidence is ambiguous for
Neanderthal man (70-35,000 bc). Linguists and
anatomists have compared the reconstructed vocal
tract of a Neanderthal skull with those of a new- Capuchin
born and an adult modern man. The newborn and
Neanderthal vocal tracts are remarkably similar.
Neanderthal man would have been able to utter
only a few front consonant-like sounds and centra-
lized vowel-like sounds, and may have been unable
to make a contrast between nasal and oral sounds.
\ Orangutan
This is well below what is found in the phonologies
of the world's languages today (p. 165). It would
have been possible to construct a linguistic code
out of these limited sounds, but it would have
required a level of intellectual ability apparently
lacking at that evolutionary stage. On the other
hand, these phonetic abilities are far ahead of
modern primates. It has thus been concluded that ^ Gorilla

las have a brain size close to these. It is plausible Neanderthal man represents an intermediate stage
that an increase in the number of brain cells in the gradual evolution of speech. Cro-Magnon
increases intellectual or linguistic capacity, but no man (35,000 bc), by contrast, had a skeletal struc-
correlation has been established. ture much more like that of modern man.

Casts of the nasal, oral, and pharyngeal air passages of (1 ) dimensions can be clearly seen (below right) when the four
a newborn baby, (2) an adult chimpanzee, (3) a Neanderthal tracts are drawn so that they are nearly equal in size. (After
reconstruction, and (4) an adult man. The differences in P. Lieberman, 1 972.)
Rhodesian

Neanderthal

% Modern man

290 IX •
THE LANGUAGES OF THE WORLD
HOMO LOQUENS
It would seem that the human vocal tract evolved Was there ever an
from a non-human primate form to enable fast and original language? Original language
efficient communication to take place. Speech is Right: The view that all

not merely the incidental result of a system de- languages have diverged
signed for breathing and eating. The changes that from a common source, the
result of cultural evolution or
took place in the larynx, pharynx, and mouth came known
divine intervention, is
about at the cost of less efficient breathing, chew- as monogenesis. The exist-
ing, and swallowing. Modern man can choke from ence of differences between
food lodged in the larynx; monkeys cannot. The languages is then explained
as a result of people moving
survival value of speech must be considerable to
apart, in waves of migration
compensate for such deficiencies. The human around the world. In this
being, in short, seems to have evolved as a speaking view, language universals
animal - homo loquens. (§1 4) would be interpreted

Some hominids had a human-like vocal tract as as evidence of common

origin.
far back as 200,000 bc, but they probably did not
have a sufficiently developed nervous system to Centre: The opposite view,
control it. There is general agreement on a time- that language emerged more
scale from 100,000 to 20,000 bc for the develop- or less simultaneously in Original languages
several places, known as
is
ment of speech. If the Neanderthal evidence is ac- L2 L3
polygenesis. Language
cepted, this scale narrows to 50—30,000 bc, in the universals, and other similar-
latter part of the Upper Palaeolithic period. ities between languages, are
This is the conclusion regarding speech. But the then explained by pointing to
the similar constraints which
lack of physical similarities with modern man does
must have operated upon
not prove that there was no language in an abstract the early speakers (in terms
sense, or other modes of communication. At the of both their physiology and
time indicated above, cultural development was their environment), and by

relatively advanced, and there must have been some the likelihood that, as groups
came into contact, their
efficient way of transmitting information about
languages would influence
skills from one generation to the next. Any degree each other - a process Original languages
of social interdependence — as found in tribal known as convergence.
L,
grouping, religious activity, or group hunting tech-
Bottom: There is also a third
niques - would seem to require a communication
given the vast
possibility,
system. Cave drawings of the period also suggest time-scale involved. All of the
the existence of an intellectual capability such as languages that now exist
would be required for language. may indeed have diverged
An elaborate gesture system is one possibility. from a common source,
but this may have been just Source of all
The early development of language may well have
one line of descent from an extant
been assisted by some kind of signing, which would earlier era when several languages
have been the simplest way of communicating basic independent languages
meanings — such as how to use tools. Hands were emerged.
no longer necessary for locomotion, so they could Extinct

be used for other activities. Perhaps primitive peo-

ple who were skilful in using signs stood a better
chance of survival. Natural selection could then
have led to the development of the intellectual
faculties prerequisite for speech.
Learning to use tools, and to pass the skills on,
would be most efficiently done through language.
It has even been suggested that learning to use tools
and learning language are interrelated skills. They
are localized in the same general area of the brain
(§45); and both tool using and gesture require
sophisticated use of the hands. On the other hand,
some non-human primates can use tools, and
is unlikely that the hands could have been used
for two such different purposes for long. However,
in an indirect way, tools could have promoted the
development of speech. Sounds made at the same
time as the gestures might have come to be asso-
ciated with various activities. The idea has been
proposed that, as tools came to be used for more
advanced purposes, food would be stored, so that
there would be intervals between meals, and thus
more time available for the mouth to be put to
other uses - such as the development of spoken
language.
We can only speculate about the link between
oral and gestural language. Similarly, the gap be-
tween human language and the communication
it systems of the nearest primates remains vast, and
there is no sign of a language-like increase in com-
municative skills as one moves from lower to higher
mammals. Human language seems to have emerged
within a relatively short space of time, perhaps as
recently as 30,000 years ago. But that still leaves
a gap of over 20,000 years before the first unequi-
vocal evidence of written language (p. 196).

49 THE ORIGINS OF LANGUAGE •

291
50 Families of languages

The first scientific attempts to discover the history is the way the term is used in this encyclopedia).

of the world's languages were made at the end of But in some classifications, a distinction is drawn
the 18th century. Scholars began to compare in terms of how definite the relationship is. If there
groups of languages in a systematic and detailed is clear linguistic evidence of a close relationship,
way, to see whether there were correspondences the term family continues to be used; but where
between them. If these could be demonstrated, it the relationship is less definite, or more remote,
could be assumed that the languages were related the grouping is referred to as a phylum. Sometimes
— in other words, that they developed from a com- the term is used for yet more general
macro-phylum
mon source, even though this might no longer exist. and groupings. It is evident, for exam-
less definite
Evidence of a common origin for groups of lan- ple, that all the aboriginal languages of Australia
guages was readily available in Europe, in that (p. 324) are related, but as there is no clear-cut his-
French, Spanish, Italian, and other Romance lan- torical evidence which bears on the matter, and
guages (p. 301) were clearly descended from Latin little typological work, scholars often refer to the
— which in this case is known to have existed. The Australian '(macro)phylum' rather than to the Aus-
same reasoning was applied to larger groups of tralian 'family'.
languages, and by the beginning of the 19th cen-
tury, there was convincing evidence to support the
hypothesis that there was once a language from
which many of the languages of Eurasia have
The comparative method
derived. This proto-language came to be called In historical linguistics, the comparative method
Proto-Indo-European (p. 296). Very quickly, other is a way
of systematically comparing a series of
groups of languages were examined using the same languages in order to prove a historical relationship
techniques. between them. Scholars begin by identifying a set
The main metaphor that is used to explain the of formal similarities and differences between the
historical relationships is that of the language languages, and try to work out (or 'reconstruct')
family, or family tree. Within the Romance family, an earlier stage of development from which all the
Latin is the 'parent' language, and French, Spanish, forms could have derived. The process is known
etc. are 'daughter' languages; French would then as internal reconstruction. When languages have
be called a 'sister' language to Spanish and the been shown to have a common ancestor, they are
others. The same approach
is used with larger said to be cognate.
groups. Within the Indo-European family, Proto- The clearest cases are those where the parent
Indo-European is the parent language, and Latin, language is known to exist. For example,on the
Greek, Sanskrit, and others are the daughter lan- basis of the various words for 'father' in the
guages. In a large family, it will be necessary to Romance languages, given below, it is possible to August Schleicher (1 821-
distinguish various 'branches', each of which may see how they all derived from the Latin word pater. 1868) The family tree' theory
(Stammbaumtheorie) was in-
contain several languages, or 'sub-families' of lan- If Latin no longer existed, it would be possible to
troduced by the German
guages. reconstruct a great deal of its form, by comparing linguist Schleicher, who
This way of talking must not be taken too liter- large numbers of words in this way. Exactly the thought of language as an
ally. A 'parent' language does not live on after a same reasoning is used for cases where the parent organism which could grow
and decay, and whose
'daughter' language nor do languages
is 'born', language does not exist, as when the forms in Latin,
changes could be analysed
suddenly appear in the way implied by the meta- Greek, Sanskrit, Welsh, etc. are compared to recon- using the methods of the
phor of birth. Nor is it true that, once branches struct the Indo-European form, * pater. The aster- natural sciences.
of a family begin to emerge, they develop quite
independently, and are never afterwards in contact
with each other. Languages converge as well as pater
diverge. Furthermore, stages of linguistic develop-
ment are not as clear-cut as the labels on a family
tree suggest, with change operating smoothly and
uniformly throughout. Linguistic change, we now
know, is much more uneven, with different social
groups responding to change in different ways
(p.328).
Since the 19th century, other classificatory terms
have come into use. Family is still used as a general
term for any group of languages where there is
Italian Spanish
a likelihood of a historical relationship (and this padre padre

292 • IX THE LANGUAGES OF THE WORLD

isk in front of aform, in historical linguistics, shows use of word order. Chinese, Vietnamese, and Philologist -or
that the form is a reconstruction which
in question Samoan are clear cases. For example, 'I bought linguist?
has not been attested in written records. Exactly some oranges to eat' in Beijing Chinese would be: People who study the history
how such reconstructed forms were pronounced of languages are sometimes
is a matter of (at times fierce) debate: some scholars
Wo mdi jtizi chT
called comparative philo-
literally, I buy orange eat
are happy to assign phonetic values to the forms, logists (or just 'philologists')
and sometimes historical
and pronounce them as if they were part of a real
Inflecting, synthetic, or fusional languages linguists. The difference lies
language; others argue that the forms are little partly in the training, partly in
more than abstract formulae, summarizing the sets Grammatical relationships are expressed by chang- the subject matter. The philo-
of correspondences which have been noted (§54). ing the internal structure of the words - typically logical tradition is one of

by the use of inflectional endings (p. 90) which painstaking textual analysis,
often related to literary his-
express several grammatical meanings at once.
tory, and using a fairly tradi-
Latin, Greek, and Arabic are clear cases. For exam-
Types of linguistic ple, the -o ending of Latin amo 'I love' simulta-
tional descriptive
The newer,
framework.
linguistic
classification neously expresses that the form is in the first person approach tends to study his-

singular, present tense, active, and indicative. toricaldata more selectively,

There are two main ways of classifying languages: as part of the discussion of
the genetic (or genealogical) and the typological broader issues in linguistic
Agglutinative or agglutinating languages.
theory, in the process using
(§14). Both are used in contemporary language
Words are built up out of a long sequence of units, the conceptual apparatus of
work, but the former has received far more investi- modern linguistics. Needless
with each unit expressing a particular grammatical
gation, and has the better developed procedures to say,proponents of the two
meaning, in a clear one-to-one way. A sequence
and frame of reference. A further approach (an approaches do not always
of five affixes might express the meaning of amo, see eye to eye. Philologists
areal classification) is reviewed in §8.
for example — one for each category of person, are often still sceptical of the
number, tense, voice, and mood. Turkish, Finnish, new science, remembering
GENETIC CLASSIFICATION Japanese, and Swahili form words in this way. 'I the days when linguists con-
This is a historical classification, based on the sidered historical topics to be
love you' in Swahili is mimi ninakupenda wewe,
assumption that languages have diverged from a of secondary importance
which can be analysed as: (§65). Historical linguists,
common ancestor. It uses early written remains as similarly, are often impatient
evidence, and when this is lacking, deductions are mimi ni — na —ku —penda wewe with the philologist's reluc-
made using the comparative method to enable the me I present you love you tance to develop general
form of the parent language to be reconstructed. TENSE explanatory theories of
The approach has been widely used, since its intro- language change. But nowa-
days there are many signs
duction at the end of the 18th century, and provides Polysynthetic or incorporating languages that the skills of these two
the framework within which all world-wide lin- categories of scholar are
Words and complex, containing
are often very long
guistic surveys to date have been carried out. The being seen as complemen-
a mixture of agglutinating and inflectional features,
tary, not in opposition.
success of the approach in Eurasia, where copious
as in Eskimo, Mohawk, and Australian languages.
written remains exist, is not matched in most other
For example, the aboriginal language Tiwi ex-
parts of the world, where a classification into
presses'I kept on eating' as ngirruunthingapukani,
families is usually highly tentative. which is analysable as:

TYPOLOGICAL CLASSIFICATION ngi — rru -unthing —apu — kani

This based on a comparison of the formal similar-
is I past for some eat repeatedly
ities which exist between languages. It is an attempt tense time
to group languages into structural types, on the Some linguists, however, do not regard this as a
basis of phonology, grammar, or vocabulary, separate typological category.
rather than in terms of any real or assumed histori-
cal relationship. For example, it is possible to group
languages in terms of how they use sounds - how
many and what kinds of vowels they have, whether What sort of language is English?
they use clicks, whether they use tones, and so on.
English a Germanic
is chicken supreme) and pho- tional endings, and word-
Languages can also be classified in terms of whether language, according to the nology (e.g. the use of final order changes are the
their word order is fixed or free, and which order genetic method of classi- in words like garage). basis of the grammar.
Izl
is favoured (p. 98). The earliest typologies, how- fication. But from other Ifwe consider the place
ever, were in the field of morphology (p. 90). These, points of view, the picture names of North America, Three-ln-one
alters. Culturally, it displays then we have to allow a Isolating .The boy will ask
propounded by August von Schlegel (1767-1845) many similarities with relationship with Amerin- the girl.
and others in the early 19th century, recognized Romance, in view of the dian languages (Chappa- The girl will ask the boy.
three main linguistic types, on the basis of the way number of loan words
large quiddick, Susquehanna). The biggesr boys
Inflecting:

a language constructs its words. (p.330) it has taken in from From a typological view- have been asking.
French and Italian, and the point, English is in fact Agglutinating: anti-dis-
way these languages have more similar toan isolating establish-ment-arian-ism.
Isolating, analytic, or root languages even exercised some in- language like Chinese than
fluence on grammar (e.g. Latin: there are few inflec-
words are invariable: there are no endings.
All the
Grammatical relationships are shown through the

50 FAMILIES OF LANGUAGES 293

THE PROBLEM OF CLASSIFICATION
These days, typological questions ?ri of undoubted
interest — especially in relation ' the search for
language universals (§14). But .»me of the early
have been severel
classifications criticized because
of the way they were interpreted. No one would
now follow the early tendency of typologists, under
the influence of Darwin, to evaluate languages as
if they were points on an evolutionary scale — that
isolating languages are 'not as well developed' as
inflecting languages, for example. Nor is there any
evidence that languages of a particular type are
inevitably associated with particular geographical
areas, or with people of a particular ethnic or cul-
tural group. It must also be appreciated that there
is no such thing as a 'pure' instance of one of the

above types. Languages seem to have these charac-

teristics to various degrees.
Is a typological classification possible, therefore?
Everything depends on how we evaluate the vari-
ables which provide the basis of the classification.
Morphology is only one variable. When we take
into account all the features of language - syntax,
phonology-, discourse, and language use (§13) —
the nature of the problem is evident. There are a
vast number of possible classifications, and how
should we decide which criteria are the most impor-
tant? If two languages are 90% similar in phono-
logy and 50% similar in grammar, are they more
or less closely related than two languages which
are 50% similar in phonology and 90% in gram-
mar? Linguistic theory has hardly begun to answer
such questions.
Both typological and genetic classifications
ignore the relevance of cultural links between
languages — the fact that languages influence each
other by contact, such as by borrowing words from
each other. Sometimes languages that have no
historical relationship can converge so that they
seem to be members of the same family. Con-
versely, related languages can be influenced by
other languages so much that the differences
become more striking than the similarities. The role
of cultural contactis a real problem in studying

many language families, where it is often totally

unclear whether two languages are similar because
they share a common origin, or because they have
borrowed from each other (p. 330).
Some linguists have tried to move away from
a classification into general types, proposing
instead to rank languages in terms of individual
structural criteria. One criterion could be the
number of morphemes (p. 90) per word in a
language (an 'index of synthesis'). In the sentence,
'The boys saw the girl', there are 5 words but 8
morphemes, producing a synthetic index of 1.6.
Using this criterion, according to one study, the
average for English was 1.68, compared with 1.06
for Annamese and 3.72 for Eskimo. There are
several other grammatical ratios which could be
investigated in this way. The language families of the world

294 • IX THE LANGUAGES OF THE WORLD

 ySi Afro-Asiatic

Jj Algonquian

{ 1
Altaic

|%%j Andean-Equatorial

_J Australian Aboriginal

| Austro- Asiatic

Remaining areas contain isolated, unclassified, or unknown languages - or no speakers at all.

50 FAMILIES OF LANGUAGES •
295
 . 9

51 The Indo-European family

'Indo-European' is the name scholars have given of the people. There are no words for 'palm tree'
to the family of languages that first spread through- or 'vine', for example, which suggests, indepen-
out Europe and many parts of southern Asia, and dently of any archaeological evidence, that the mig-
which are. now found, as a result of colonialism, rations did not begin in the Mediterranean area.
in every part of the world. The parent language, But other clues often seem contradictory. The word
generally known as 'Proto-Indo-European', is for 'beech tree' is widely attested, and, as this tree
thought to have been spoken before 3000 bc, and does not grow in Asia, it has been suggested that
to have split up into different languages during the the Indo-Europeans must have originated in north-
subsequent millennium. The differences were well central Europe. On the other hand, there is little
established between 2000 and 1000 bc, when the evidence of a common word for 'oak', which is
Greek, Anatolian, and Indo-Iranian languages are also a European tree, and if this word was not
first attested. known to the Indo-Europeans, the view is sup-
ported that their migration must have begun in Asia
after all. Indo-European philology (§50) raises
many fascinating questions of this kind.
Who were the Indo-Europeans?
Archaeological evidence shows the existence of a
semi-nomadic population living in the steppe The discovery of Proto-Indo-European
region of southern Russia around 4000 bc, who It was
not possible to deduce foundation of Indo-European cation of the first major Indo-
began to spread into the Danube area of Europe the existence of this family of philology. In 1 81 6, the Ger- European grammar: Com-
languages until scholars be- man philologist Franz Bopp parative Grammar of San-
and beyond from around 3500 bc. The people are came aware of the systema- published a study, whose skrit, Zend, Greek, Latin,
known as the Kurgans, because of their burial prac- ticresemblances which can scope is well illustrated by its Lithuanian, Old Slavic, Goth-
tices {kurgan being the Russian for 'burial mound'). be found between European title (translated): On the con- ic and German. It took 1
Kurgan culture seems to have arrived in the Adria- languages and Sanskrit, the jugation system of the years to complete, and by its
oldest-attested language of Sanskrit language, in com- third edition incorporated Old
tic region before 2000 bc, and this coincides well
the Indian sub-continent. parison with those of the Slavic, Celtic and Albanian.
with the kind of time-scale needed to produce large When these were first Greek, Latin, Persian and In due course, this work and
amounts of linguistic change. The ancestors of the noticed, in the 1 6th century, Germanic languages. The its contemporaries became

Kurgans are not known, though there are several many people thought that relationship of Germanic to out of date, as a result of the
Sanskrit was the parent of Latin, Greek, Slavic and Bal- vast amount of philological
similarities between Proto-Indo-European and the
the European languages; but tic was demonstrated in a study undertaken in the
Uralic family of languages (p. 304), spoken further towards the end of the 1 8th work written in 1 81 4 by the second half of the 1 9th cen-
east, and these may well have had a common par- century the systematic Danish linguist, Rasmus tury. A further publishing
thousand years before.
ent, several studies began which showed Rask, but not published until landmark was Karl Brug-
similar vocabulary of the conclusively that this was not 1818, Investigation on the mann's Outline of Compara-
By comparing the
the case Origin of the Old Norse or tive Indo-European Gram-
extant Indo-European languages, it is possible to
Following an early state- Icelandic Language. Further mar 1 897-1 9 1 6) A new
( .

draw some conclusions about the geographical ment of the common origin philological treatises fol- Indo-European Grammar,
origins and life-style of the people. For instance, hypothesis in 1 786, by Sir lowed, mainly written by Ger- the outcome of a project di-
many family words (such as 'mother', 'husband', William Jones, the early 1 9th mans, such as Jakob Grimm rected by the Polish linguist,
century produced several and August Schleicher. In Jerzy Kurytowicz, com-
'brother') can be reconstructed for Proto-Indo-
major works which laid the 1 833. Bopp began the publi- menced publication in 1968.
European. These include several words for 'in-
laws', which seem to have been used solely with
referenc o the bride. Evidence of this kind sug-
gests thatit was the wife who was given a position
William Jones (1746-94) both of them a stronger aff in-
within the husband's family, rather than the other British orientalist and jurist, ity, both in the roots of verbs,

way round, and that the society must therefore whose presidential address and in the forms of grammar,
have been patriarchal in character. to the Bengal Asiatic Society than could possibly have
786 contained the follow- been produced by accident;
The reconstructed language has words for in 1

ing observation, generally so strong, indeed, that no

horses, dogs, sheep, pigs, and other animals; there MR - JH philologer could examine
quoted as the first clear
is a wordsome kind of vehicle, and this vehicle
for statement asserting the them all three, without be-

definitely had wheels; there are many words for existence of Indo-European : lieving them to have sprung
parts of the body; there are several words relating
_ _
The Sanskrit language
from some common source,
which, perhaps, no longer
to farming, and a few words relating to tools and whatever be its antiquity, is
exists
weapons; many abstract notions are attested, relat- of a wonderful structure;
more perfect than the Greek,
ing to such fields as law, religious belief and social
more copious than the Latin,
status; numerals went to at least 100. Words relat- and more exquisitely refined
ing to fauna and flora are of particular interest, than either, yet bearing to
for they can provide clues as to the place of origin

296 • IX THE LANGUAGES OF THE WORLD

 What did Proto-Indo-European came to be called laryngeals, a term taken
the study of Semitic languages (p. 316), where con-
from

sound like? sonants in the region of the larynx were known

There are no written records relating to this period. to occur. Laryngeal consonants did not occur in
The Kurgans must have been illiterate - unlike the any Indo-European language known at the time,
peoples of Egypt and Mesopotamia of the time. but the previous existence of some kind of sound,
So the entire character of Proto-Indo-European has it was argued, was indicated by the way they had

been the result of painstaking reconstruction on caused the changes to take place in adjacent vowels
the part of philologists, using the methods outlined (altering their length and quality) that had long
on p. 292. been noticed in the extant languages.
Jacob Grimm (right) There
is general agreement about the number The laryngeal theory was immediately contro-
(1785-1863) of contrasts in the consonant system (p. 165), versial, and received little support for many years.
Well known to children though the status of some of the less well-attested But attitudes changed after 1927, when it was
everywhere for the collection
sounds (such as/b/) is disputed. This system seems found that Hittite (discovered several years after
of fairy tales and songs
which he compiled with his largely to have been composed of plosives (p. 157), the theory was postulated) had a sound, repre-
brother. To linguists and organized into three series: voiceless, voiced, and sented by h, that occurred in some of the places
philologists, he is also re- (less definitely) voiced aspirate. Four main places where Saussure had predicted the laryngeals should
membered for his major
of articulation were used: labial, dental, palatal or be. However, the phonetic character of these laryn-
works in Germanic philology,
velar, and labio-velar. There was a single fricative, geals is still quite unclear, and philologists disagree
especially his explanation of
how the consonants of differ- which was voiced or voiceless according to context. on how many laryngeal sounds there were, whether
ent Indo-European In addition, there were probably one or more laryn- their phonetic properties can (or should) be
languages relate to each geal consonants (see below). There were two defined, and whether better analyses can be found.
other. There is, for example,
nasals, two continuants, and two semi-consonants It is generally recognised that there must have been
a regular relationship be-
tween words beginning with (p. 152), all of which could occur at the centres three (some say four) types, pronounced some-
pin Latin and fin Germanic of syllables as well as at syllable edges. This system where in the back part of the mouth, probably as
languages (as in pater and may be summarized as follows: fricatives or glottal stops (p. 157). They are usually
father), or between initial fin
Greek and initial th- in Eng-
symbolized by H or schwa (a), and numbered with
lish (as in treis and three).
Labial Dental Palatal/ Labio- subscripts(H l5 2 , H etc. or a
l5 a 2 , etc.) Alternative
Velar velar
The rules governing these analyses which postulate an earlier vowel, rather
sound shifts became known Plosives than a laryngeal, have also been proposed.
as 'Grimm's law' (p. 328). Voiceless P t k kw Laryngeal theory can be illustrated in this way.
Voiced b d w
g g Most Proto-Indo-European basic forms (or 'roots')
Voiced aspirate bh dh gh gj-b
Fricatives s(z)
had a structure of Consonant—Vowel— Consonant
Nasals m n (CVC, which is often written as CeC, when discuss-
Continuants 1 r ing this language), e.g. *bher- 'bring', *med-, 'mea-
Semi-consonants w i sure'. But several forms had only one consonant,
e.g. It is argued that these
*es- 'be', *do- 'give'.
There is more disagreement over the vowel sys- roots can be reconstructed as having the regular
tem - vowels, as always (p. 167), being more diffi- CVC structure, by postulating a laryngeal as the
cult to analyse. Four main contrasts are generally
'missing' consonant, e.g. *Hes-, *doH-. In roots
recognized: mid-front, mid-back, open and central, such as *doH-, with a preceding vowel, when the
the first three occurring both in long and short laryngeal finally disappeared, it caused the vowel
forms (though how far these were independent con- to lengthen, as is attested in Latin donum 'gift',
trasts, as opposed to laryngeally controlled var-
and elsewhere. Using these techniques, it is possible
iants, is unclear). In addition, some scholars toshow that almost all the roots of the proto-lan-
recognize two further contrasts in close position, guage (there are still a few exceptions, such as
/i/ and /u/, but the overlap with the use of these numerals) had a CVC structure.
sounds as semi-consonants makes this analysis less
certain also. The possible vowel system can thus Some grammatical features
be summarized as follows: People often think that the oldest languages must have
been simpler than their modern counterparts (§49).
(1) (u) The noun inflections of Proto-Indo-European clearly
e/e: 3 o/o: show this not to be so. It is possible to reconstruct three
genders (masculine, feminine, and neuter) and up to
a/a:
eight cases (nominative, vocative, accusative, genitive,
dative, ablative, locative, instrumental). Adjectives
THE LARYNGEAL THEORY agreed in case, number, and gender with the noun.
Towards the end of the 19th century, the Swiss The verb system was also rich in inflections, used for
linguist Ferdinand de Saussure 407) put forward aspect, mood, tense, voice, person, and number (p.
(p.
93). Differentgrammatical forms of a word were often
the view that, in order to explain various anomalies
related by the feature of ablaut, or vowel gradation:
in early Indo-European forms, an extra set of the root vowel would change systematically to express
sounds would have to be postulated as occurring such differences as singular and plural, or past and
in Proto-Indo-European. Saussure did not suggest present tense, as is still the case in English foot/feet
or take/took.
any phonetic details for these sounds, but later they

51 THE INDO-EUROPEAN FAMILY •

297
 .

55 Indo-European varieties
Our father, who art in Greek Iranian Geographical distribution of the
Heaven . . Pater 'emon. 'o en tofs Max fyd, kaecy dae aervty Indo-European family of languages
ouranofs (New Testament) midaeg (Ossetic)
Celtic Patera mas, pou efsai stous Ya bawk-T ema, ka la
Ein Tad, yr hwn wyt yn y ouranoiis (Modern) asman-a-y (Kurdish)
nefoedd (Welsh) Ei pedar-e-ma, ke dar
Ar n-atheir, ata ar neamh Baltic asman ast (Persian)
(Irish Gaelic) Teve musy, kurs esi danguje
Phith manf, ki bihishta asti
Ar n-athair a tha air neamh (Lithuanian)
(Baluchi)
(Scottish Gaelic) Musu tevs debesfs (Latvian) Ajjmug plara, ce pa asman
Ayr ain, t' ayns niau (Manx) Tawa nouson, kas tu essei kxe ye (Pashto)
Agan tas ny, us yn nef en dangon (Old Prussian)
(Cornish)
Slavic Indo-Aryan
Germanic Otice nasf ize jesi na Bho asmakham svargastha
Unser Vater, der Du bist im nebesichu (Old Church pitah (Sanskrit)
Himmel (German) Slavonic) Saggatha no pita (Pali)
Undzer voter, vos bist im Otce nas, siiscij na He hamare svargbasf pita
himl (Yiddish) nebesach (Russian) (Hindi)
Fasder ure, t>u be eart on Ojca nas, katory jesc u
heofonum (Old English) He sade pita, jihra surg vie
nebe (Belorussian)
Onze vader, hai (Panjabi)
die in de Otce nas, sco na nebi
hemelen zijt (Dutch) E asan-ja piu, jo asmana
(Ukrainian)
Fader var, du som er i
men Ihe (Sindhi)

himmelen (Norwegian) Ojcze nasz, ktorys jest w Ai sani mali, yus asmanas
Fader var, som ar i
niebiesiech (Polish) path chu (Kashmiri)
himmelen (Swedish) Otce nas, kteryz jsi v He hamra svargavasf pita
Vor Fader, du som er nebesich (Czech) (Nepali)
i

himlene (Danish) Otce nas, ktory si v

akasamanna amara bapa
nebesiach (Slovak)
(Gujarati)
Italic
Oce nas, sto si na neboto He amacya svargatila pitya
Pater noster, qui es in caelis (Macedonian) (Marathi)
(Latin) Oce nas, koji si na nebe- Svargayehi vadasitina
Notre pere, qui es aux cieux sima (Serbo-Croat) apage piyaneni (Sinhalese)
(French) Otce nas, kojto si na
Padre nuestro, que estas en nebesata (Bulgarian) He amar svargat thaka pitri
los cielos (Spanish)
Oce nas, ki si na nebesih (Assamese)
Pai nosso, que estas nos (Slovene) He amader svargastha pita
ceus (Portuguese)
(Bengali)
Pare nostre, que estau en lo Armenian He ambhamananka
eel (Catalan) Mer hayr or erknk'umn svargasha pita (Oriya)
(East)
Albanian Ov hayr mer or erkink'n es Dade amare, kaj isien k'o
Atiyneqejeneqiell (West) devle (Romany)

The Indo-European family tree, reflecting geographical distribution

PROTO
INDO-EUROPEAN
O <?
TOCH.4RIAN"
^
French
(Xatjn
Catalan
Spanish ITALIC
Portuguese

Sardinian

Maldivian Sinhalese
51 THE INDO-EUROPEAN FAMILY 299
The Indo-European family
ALBANIAN
This language forms a single branch of the Indo-
European family, spoken by nearly 3 million people
in Albania, and parts of Yugoslavia, Greece and
Italy. Albanian has two main dialects, known as
Gheg (in the north) and Tosk (in the south), but
these contain many further dialect divisions, not
all of which are mutually intelligible. The history
of the language is obscure, and it is not possible
to demonstrate a clear relationship with any other
Indo-European group. This is partly because of the
many loan words which have shaped the modern
language, and partly because so few written re-
mains of earlier times exist, dating only from the
15th century, largely on religious themes. An offi-
cial alphabet was not introduced until 1909, using
roman characters. Since the Second World War,
the official language has been based on the Tosk
dialect.
ANATOLIAN
A group of languages, now extinct, spoken from
around 2000 bc in parts of present-day Turkey
and Syria. The main Anatolian language is Hittite,
shown to be Indo-European only as recently as
1915. written remains, consisting of tablets
Its
inscribed with cuneiform writing (p. 198), date
from the 17th century bc. The earliest forms of
Hittite ('Old Hittite') are the oldest Indo-European
texts so far discovered. Most of the texts have reli-
gious themes, but they also contain a great deal
of historical and social information. Other lan-
guages of the group are Palaic, Lydian, Lycian, and
Luwian (represented in cuneiform and hieroglyphic
systems). Also grouped under this heading are cer-
tain languageswhich do not belong to the Indo-
European family (Hurrian, Urartian) or where the
relationship is not certain (Phrygian).
ARMENIAN
This branch of Indo-European consists of a single
language, spoken in many dialects by between 5
and 6 million people in the Armenian republic of
the USSR and Turkey, and (through emigration)
in parts of the Middle East, Europe, and the United
States. The spoken language may have been estab-
lished soon after 1000 BC, but there was no written
form until after the introduction of Christianity.
Classical Armenian, or Grabar, is the language of
the older literature, and the liturgical language of
the Armenian Church today. The oldest writings
date from the 5th century, and the 38-letter alpha-
bet, invented by St Mesrop, is still widely used.
Modern literary Armenian exists in two standard
varieties: East Armenian is the official language of
the Armenian SSR; West Armenian is the dominant
variety elsewhere. Because of the large numbers
of loan words (see p. 330) which have come into
the language, its basic Indo-European character is
often obscured.
300 IX THE LANGUAGES OF THE WORLD
BALTO-SLAVIC
Baltic languages and Slavonic languages are often
placed together as a single branch of Indo-Euro-
pean, because of their similarities, though there is
some dispute over whether these constitute evi-
dence of common origin rather than of more recent
mutual influence. Taken together, these languages
are spoken by about 300 million people, more than
half of whom speak Russian.
The main Baltic languages are Latvian (also
known as Lettish) and Lithuanian, with written
texts dating from the 14th century. There are
around 4 million speakers in the Baltic area, with
a further million abroad, mainly in the United
A bilingual tablet in Hittite
and Luwian, dating from
States. Both languages have standard forms, and
around 1400 bc, on which
many dialects. Several other languages of this is written a ritual against
group are now extinct, though there are a few writ- the plague. The tablet was
ten remains of Old Prussian. found in Hattusas, modern
Bogazkdy, Turkey. Inscrip-
The Slavonic (or Slavic) languages are more
tions from this area provided
numerous, and are usually divided into three some of the earliest
groups: South Slavonic, found in Bulgaria, Yugo- evidence for the classifica-
slavia, and parts of Greece, includes Bulgarian, tion of Hittite as an Indo-
Macedonian, Serbo-Croat, and Slovene; West Sla- European language.
vonic, found in Czechoslovakia, Poland, and East
Germany, includes Czech, Slovak, Sorbian, and
Polish; East Slavonic, found in the USSR, includes
Russian, Belorussian, and Ukrainian. Each of the
main Slavonic languages has an official status as
a standard (pp.38, 364); but there are numerous
dialect differences within these groupings. Old
Church Slavonic is evidenced in texts dating from
the 9th century, and its later form (Church Slavo-
nic) is still used as a liturgical language in the East-
ern Orthodox Church. The distinctive Cyrillic
alphabet (p. 202), attributed to Saints Cyril and
A page from the Codex
Methodius in the late 9th century, is still used for
Argenteus, a 5th- or 6th-
writing Bulgarian, Serbian, Macedonian, and all century copy of the Bible of
the East Slavonic languages. In modified forms, it Ulfilas; its name derives from
isalso used for about 100 non-Slavonic minority the lettering, which is in gold
languages of the USSR. and silver on a purple parch-
ment. It is kept at Uppsala,
in Sweden, not far from the
GERMANIC Goths' homeland.
The various branches of the Germanic family of
languages derive from the migrations of the Ger-
manic tribes who lived in northern Europe during
the 1st millennium bc. Some Germanic words are
recorded by Latin authors, and Scandinavian
inscriptions in the runic alphabet (p. 203) are
recorded from the 3rd century ad. The earliest
main text is the Gothic Bible of Bishop Ulfilas (or
Wulfila), translated around ad 350, using an
alphabet of his own devising (the Gothic alphabet:
p. 186). Anglo-Saxon and Old High German are
recorded from the 8th century, and the oldest forms
of Scandinavian languages from the 12th century.
Germanic languages are used as a first language
by over 500 million people, largely because of the
world-wide role of English (§59). They are usually
classified into three groups. East Germanic lan-
guages are all extinct, and only Gothic is preserved
in manuscript to any extent - most recently, in
a few words recorded in the Crimea in the 16th

&T PE tARBjk
Ion. opt fcyU ferpnj
m moni^u m.<^imi
Up: pun W lie \j;~
fc«)i paicnum
fXI
ii J^hnvtc p.i'ui >-m(> ("irco-i
ion iiu>«- Wfuui fcolie ^oiiiktrr
I p" y.q' s^ cvmn;j.
folce cop)iap(ie- pi-no. ^al[lpt-
Jimi^ef upp^iftj. puU
e- pOJt^tl :.
:
A page from the oldest
epicpoem in English,
BeowulfThe work was prob-
ablycomposed in the 8th
century, but the only surviv-
ing manuscript dates from
around ad 1 000.
story of a Scandinavian hero,
Beowulf, who fights and kills
a monster, Grendel, in Den-
mark. He is later made king
of the Geats, in southern
Sweden, where, as an old
man, he kills a dragon, in a
fight that leads to his own
death.
This inscription, carved in
an almost impossible posi-
tion at the top of a steep cliff
in Behistun (modern Bisitun),
Iran, recounts the feats of
King Darius the Great of Per-
sia (522-486 bc ). It is in
three languages, Old Per-
sian, Akkadian, and Elamite,
and proved to be of particular
value in deciphering the
cuneiform writing system.
century. North Germanic includes the Scandina-
vian languages of Swedish and Danish (East Scan-
dinavian), Norwegian, Icelandic, and Faeroese
(West Scandinavian), and the older states of these
fcftijie-
languages, most notably the literary variety of Old
m<t> oo feclk
Icelandic known as Old Norse - the language of
the Icelandic sagas. West Germanic comprises Eng-
:|iop|i<- s;eU lish and Frisian (often grouped as Anglo-Frisian),
and German, Yiddish, Netherlandic, or Dutch
t tuviiA^ju
Sna
(including local, Flemish dialects in Belgium), and
Afrikaans (often grouped as Netherlandic-Ger-
^«n tttp.
man). Dialect similarities often blur the distinctions
suggested by these labels (§§8, 47).
on
GREEK
This branch of Indo-European consists of a single
language, represented in many dialects, and
attested from around the 14th century bc. The
earliest evidence of the language is found in the
inscriptions discovered at Knossos and other
centres in Crete, written mainly on clay tablets in
a syllabic script known as Linear B, and discovered
to be Greek only as recently as 1952 (p. 201). This
period of the language is referred to as Mycenaean
Greek, to be distinguished from the later, classical
Greek, dating from the 8th century bc, when texts
came to be written in the Greek alphabet (p. 202)
It tells the
— notably the epic poems, Iliad and Odyssey. The
great period of classical drama, history, philoso-
phy, and poetry ended in the 4th century bc. A
later variety of Greek, known as koine (or 'com-
mon') Greek, was spoken throughout the eastern
Mediterranean from around the 4th century bc for
nearly a thousand years. In its written form, it was
the language of the New Testament (p. 384). The
modern varieties of Greek, spoken in Greece,
Cyprus, Turkey, the United States, and other locali-
ties, derive from this koine (p. 43).
INDO-IRANIAN
This branch of Indo-European comprises two large
groups, known as Indo-Aryan (or Indie) and Iran-
ian. There are over 500 Indo-Aryan languages,
spoken by around 500 million people in the north-
ern and central parts of the Indian subcontinent.
They may be divided into several groups, on a
broadly geographical basis: the Midland group
mainly includes Hindi/Urdu (p. 284), the Bihari
languages, and the Rajasthani languages; the East-
ern group includes Assamese, Bengali, and Oriya;
the West and South-west groups include Gujarati,
Konkani, Maldivian, Marathi, and Sinhalese; and
the North-west group includes Panjabi, Sindhi,
Lahnda, the Dardic languages, and the Pahari lan-
guages.The Romany language of the gypsies is also
a memberof this family. The early forms of Indo-
Aryan, dating from around 1000 BC, are collec-
tively referred to as Sanskrit — the language in
which the Vedas, the oldest sacred texts, are written
(p. 384). Later forms, the Prakrits, lasted 1,000
years, and were the medium of Buddhist and Jain
literature.
During the same period, the Iranian languages
were being spoken in an area centred on modern
Afghanistan and Iran — especially Old Persian and
Avestan (the sacred language of the Zoroastrians),
both of which have texts dating from the 6th cen-
tury bc. The main modern Iranian languages,
spoken by over 60 million people, are the closely
related Persian (or Farsi) and Tadzhik, as well as
Pashto, Ossetic, Kurdish, and Baluchi, but there
are many other languages, and innumerable dia-
lects, which have so far received no definite classifi-
cation.
ITALIC
The main language of this family is Latin, the lan-
guage of Rome and of its surrounding provinces,
preserved in inscriptions from the 6th century BC,
and most systematically in literature from the 3rd
century bc. Other languages of the period include
Faliscan, Oscan, Umbrian, and Venetic, spoken in
and to the north-east of modern Italy. From the
spoken, or 'vulgar' form of Latin, used throughout
the Roman Empire, developed the Romance lan-
guages —
French, Spanish, Portuguese, Italian, and
Romanian, along with Sardinian, Occitan (in
southern France), Rhaetian (various dialects in
northern Italy and Switzerland), and Catalan (pre-
dominantly in north-east Spain). A Romance lan-
guage known as Dalmatian, spoken along the
Yugoslavian coast, became extinct when its last-
known speaker died in 1898. But the main
Romance languages have spread, as a result of col-
onialism, throughout the world, so that today over
500 million people speak a Romance language, or
one of the Creoles based on French, Spanish, or
Portuguese (pp. 334—9).
TOCHARIAN
This language, now extinct, was spoken in the
northern part of Chinese Turkistan during the 1st
millennium AD. The first evidence of Tocharian
was discovered only in the 1890s, in the form of
various commercial and Buddhist religious docu-
ments, dating from around the 7th century, and
on the basis of these discoveries, two dialects were
established - an eastern variety, from the Turfan
region, which was labelled Tocharian A, and a
western variety, from the Kucha region, which was
labelled Tocharian B. The functions of these dia-
lects, and the identity of their speakers, have been
sources of controversy in comparative philology,
as has the very name of the language (based on
that of the Tochari people, who lived further east,
and who were probably speakers of an Iranian lan-
guage). But the status of Tocharian as an indepen-
dent Indo-European language is not in doubt.
For Celtic, see pp. 302-3.
51 THE INDO-EUROPEAN FAMILY 301
 and two languages were mutually
said that the
The history of one Indo- it is

intelligible until the 15th century.

European family: Celtic The movement of Goidelic peoples continued
from Ireland into the Isle of Man, and by the 10th
EARLY CELTIC century they were also found throughout Scotland.
The Celts were the first Indo-European people to The geographical boundaries led quickly to distinct
spread across Europe. Known to the Greeks as Kel- dialects of the original languages, known as (Com-
toi, they emerged in south-central Europe around mon) Gaelic. From about the 10th century, there
the 5th century BC, speaking a language which has is evidence of the distinction between Irish and

been reconstructed under the name of Common Scottish Gaelic (until recently often called Erse),
(or Proto-) Celtic. In a series of waves they spread and doubtless Manx Gaelic began to diverge at
throughout most of Europe, reaching as far as the this time too. It is not clear how long these dialects
Black Sea and Asia Minor, south-west Spain, cen- stayed mutually intelligible, but the development
tral Italy, and the whole of Britain. Their culture of different cultural and literary traditions in the
is known as La Tene (after the Swiss archaeological middle ages persuades most writers to take them
site of that name). as different languages from this time.
The main migration was by the Galli, or Gauls, In recent times, Celtic languages have spread out-
into France, northern Italy, and the north of side Britain on two occasions.
In 1865, 150 Welsh
Europe. Evidence of the Gaulish language is found settled in PatagoniaArgentina) and by the early
(in
throughout this area in place names and inscrip- 1900s their numbers had increased so that there
tions. In other places, the language goes under dif- were nearly 3,000 speakers of Patagonian Welsh.
ferent names. The Celts who went into the Balkans Nowadays, the language has largely disappeared,
and Asia Minor were called Galatae by the Greeks, under the influence of Spanish. In the 18th century,
and Galatian remained in use until around the 5th many Gaels emigrated to Cape Breton Island, Nova
This bronze shield is a pro-
century ad. The Celts who went into Spain were Scotia; there were an estimated 30,000 speakers
duct of the La Tene culture
known as Celtiberi, and Celtiberian is found in of Cape Breton Gaelic in the 1930s, but only a of the 1 st century bc. Its
inscriptions (only partly decipherable), especially few are left today. height is 77.5 cm.
in the north and east. Some 2nd-century bc inscrip-
tions in Switzerland are often referred to as Lepon- Celtic movements
tic.
in Europe, and their
The range of Celtic dialects spoken on the Conti- main linguistic
nent of Europe has been labelled Continental Cel-
consequences
tic. which came
Insular Celtic refers to the dialects
Cape Breton
and Brittany, and
to be spoken in the British Isles Gaelic

almost all our information about the Celtic lan-

guages comes from this area. There seem to have
been two waves of invasion: the first, into Ireland
in the4th century bc, led to a type of Celtic known
as Goidelic (or Gaelic) which later reached Scot-
land and the Isle of Man; the second, into southern
England and Wales, and later over to Brittany, pro-
duced a type of Celtic known as Brythonic (or,
simply, British). Linguistically, the first language
group is known as Q-Celtic, because it retained
the /kw-/ sound of Proto-Indo-European, writing
itas q, later c; the second group is referred to as
P-Celtic, because /kw-/ developed into /p-/. The
contrast can be seen in such pairs of words as
modern Irish Gaelic ceathair, Welsh pedwar 'four'.

THE FORTUNES OF INSULAR CELTIC

The Anglo-Saxon invasions in the 5th century ad
pushed the British Celts westward and northward,
so that the various dialects quickly became distinct.
In the area now known as Cornwall and Devon,
the language developed into Cornish; in Wales,
into Welsh; and in Cumbria and parts of Scotland
intoCumbric. There was a movement into Brittany
from southern England, around the 5th century
ad, which led to the development of Breton. In
its early period, Breton was very similar to Cornish,

302 IX THE LANGUAGES OF THE WORLD

 InEurope, the most noticeable modern charac-
teristic of thislanguage family is its dramatic dec-
line, under the influence of its powerful linguistic
neighbours, English and French. But equally dra-
matic is the 20th-century revival of interest in Celtic
languages, as symbols of nationalistic unity, and
as keys to earlier periods of cultural and literary
brilliance.
PERIODS OF DEVELOPMENT
Early (5th-9th c)
A few names and inscriptions in Irish Gaelic,
Welsh, Breton, and Cornish. The language of the
early Irish period is preserved in a writing system
known as Ogam (p. 203).
Old (9th-12th c)
Old Welsh, Cornish, and Breton can be dis-
tinguished, on the basis of glosses and vocabulary
lists. Because of the work of Christian missionaries,
there is far more available for Old Irish, and it
is difficult to draw a line between Irish literature
of the Old and Middle periods.
Middle (12th-15thc)
An extensive literature in Welsh and Irish, several
plays in Cornish, and verse fragments and plays
in Breton. There is no Scots Gaelic literature, as
the earliest Scots Gaels wrote in Irish.
Modern (16th-20thc)
Cornish Little further development, the language
dying out at the beginning of the 19th century.
A recent revival of interest is based on the Middle
Cornish period.
Manx A largely religious literature (the Book of
Common Prayer was translated in the early 17th
century), some ballads and carols. The Isle of Man
was wholly Manx-speaking until the 18th century,
and the laws of the island are still promulgated
in Manx. There were still some 5,000 speakers at
the beginning of the 20th century, but the
mother-tongue speakers died in the late 1940s.
Breton A major period of growth, from the mid-
17th century, led to a new orthography, several
grammars, and a large literature of plays, legends
and ballads. A strong nationalistic movement
began in the 1890s, and Breton was recognized
as a school subject in the 1950s. There are no offi-
cial figures, but it has been suggested that there
were around a million speakers in the
that this figure is now more than halved.
Irish Gaelic Ireland was wholly Gaelic-speaking
until the 17th century, but the dominance of Eng-
lish, and the effects of 19th-century famine and
emigration, led to a sharp decline. In the 1981 cen-
sus, over 5,000 people claimed to be monolingual
in Irish, and over a million to be speakers of Irish
ceuiD Traditional Gaelic script A
semi-uncial form of medieval
writing (p. 186). The sign ex-
presses the traditional greet-
roiLe ing: a hundred thousand wel-
comes'.
.poriLte
and English (without specifying level of pro-
ficiency). The constitution makes Irish the first offi-
ciallanguage. Gaelic has been taught in schools
since 1922, and a standard grammar has been de-
veloped, along with a movement for reform of the
complex spelling system. There is now a marked
resurgence of interest in the language and its litera-
ture.
Scottish Gaelic There was a major period of poe-
tic literature in the 18th century, but a standard
written language did not develop until the Bible
translation of 1801. It is still spoken in the west,
especially in the Isles and parts of the Highlands,
and it now attracts a strong nationalistic interest;
but the decline in numbers has been steady in the
present century (from over 250,000 in 1891 to
fewer than 80,000 in the 1980s).
Welsh Wales was monoglot until the 16th cen-
tury, when
the Act of Union with England (1536)
led to a rapid decline in numbers of Welsh speakers.
Revivals in the 1 8th and 1 9th centuries led to Welsh
being taught in schools, and the present century
has seen this revival continue on an unprecedented
scale, with the language now given official status.
It istoo soon to say whether this interest has come
too late to stem the steady decline in the number
of speakers — from just under a million in 1900
to around half a million in the 1981 census - but
fresh factors are now operative, not least in the
last form of a new Welsh-speaking television channel,
and these may well alter the census trend.
1940s and
Sign of the times Road signs in Wales are these days
printed in both Welsh and English - a policy which does not
satisfy some nationalists.
51 THE INDO-EUROPEAN FAMILY •
303
52 Other families
Uralic
The Uralic family consists of those languages which
have descended from an ancestor, called Proto-
Uralic, spoken in the region of the north Ural
Mountains in the USSR over 7,000 years ago. Ura-
lic languages are attested in written form from the
13th century. The most noticeable trend in the 20th
century has been the decline of many of the lan-
guages, under the influence of dominant neigh-
bours, especially Russian. Also, several of these
languages have more than one name - an earlier
derogatory name, used by outsiders, has often been
replaced by a form more acceptable to the people.
Two main branches of the family are represented
today: Finno-Ugric and Samoyedic. The Finno-
Ugric group of languages is found in one part of
central Europe, and in those northern territories
where Europe and Asia meet. In the north, the 'Fin-
nic' branch of the family is located in the region
between northern Norway and the White Sea, the
whole of Finland, and parts of adjacent Soviet terri-
tory. The main language of the group is Finnish,
with over 5 million speakers in Finland, Sweden,
the USSR, and (through emigration) the USA.
Estonian has around 1 million, mainly in the Esto-
nian SSR. There are only around 30,000 speakers
of Lapp (or Lappish], but they are spread through-
out the whole of the north.
Curiously isolated from the rest of the family
is the main language of the 'Ugric' branch — Hun-
garian (or Magyar). This is spoken by around 11
million people as a national language in Hungary',
Where Uralic languages are spoken
I Finnic
Samoyedic
Ugnc
304 IX THE LANGUAGES OF THE WORLD
and by a further 3 million in surrounding areas,
and through emigration in many parts of the world.
Two other Ugric languages are found to the east
of the Urals, around the River Ob, and are known
as Ob-Ugric. They are Khanty (or Ostyak), with
around 15,000 speakers, and Mansi (or Vogul),
with around 4,000.
The remaining Finno-Ugric languages are
spoken within the Soviet Union. One group is
found in the Kola Peninsula in the north, and south-
wards towards the Gulf of Riga. Some of these
languages (Ingrian, Livonian, and Votic) have very
few speakers, and may not survive for long. Kare-
lian, the most widespread, has over 100,000
speakers. Veps has fewer than 10,000. A second
group is found further into the Soviet Union, scat-
tered around the central Volga. The most widely
used languages are Mordvin, with around 1 million
speakers; Mari (or Cheremis) and Udmurt (or
Votyak), both with around 500,000; and Komi (or
Zyryan), with around 400,000.
The other branch of the Uralic family is spoken
by the Samoyeds — fewer than 30,000 people scat-
tered throughout a vast area in Siberia and the Arc-
tic USSR, whose economy is largely based on
reindeer hunting and breeding. The most widely
spoken language is Nenets (or Yurak), with around
25,000 speakers. Selkup (or Ostyak Samoyed) has
around 3,000. The other languages still spoken are
Nganasan (Tavgi, or Aram) and Enets (or Yenisey),
with only a few hundred speakers each. The last
of a group of languages once spoken in the Sayan
Mountain area seems to have recently died out.
^*X
SAYAN
MOUNTAINS
 Where Caucasian languages are spoken Caucasian
Union of Soviet
The area between the Black Sea and the Caspian
Caucasus Mountains, is rela-
Sea, surrounding the
Socialist Republics
and compact — not quite twice the size
tively small
of the United Kingdom — but it contains one of
Caspian Sea the highest concentrations of languages in the
---&*._ world. Leaving aside the Indo-European, Semitic,
\
and Altaic languages which have infiltrated the area
in the past 3,000 years, there are about 40 lan-
Georgian S.S.R. <J /, \ X^ /
guages which are recognized as belonging to a
- / *, ^ — -^ -^ Azerbaidzhan S.S.R.
single Caucasian family. They are classified into
Abkhazo-Adyghian group, found
three types: the
"> .
Armenian
. \V s in the north-west of the region; the Nakho-
r )

, S.S.R. :
Dagestanian group, found in the north-east; and

Turkey
.J
v
-
v x
the Kartvelian (or Iberian) group, found in the
south.
The family as a whole is represented by around
i
5 million speakers, almost all to be found in the
Iran
*J Caucasus region. Over 3 million of these live in
the Georgian SSR, speaking a Kartvelian language
— mainly one of the dialects of Georgian, which
The inscription on this frag-
is the state language, used throughout the area.
ment of an engraved cross is
written in Khutsuri, an eccle- —. ..
Other languages of the south are Zan (including
siastical script of 38 letters Mingrelian and Laz) and Svan. Only Georgian has
used in Old Georgian. The -.-. a written form, which dates from the 5th century
palaeographic evidence (p.
AD.
1 87) suggests that the date
In the north-west, the main languages are Kabar-
is 1 1 th century. The cross is

preserved in the grounds of dian (or Circassian), with over 300,000 speakers,
the new Church of the Trans- Adyghe (around 100,000) and Abkhaz (around
figuration in Akhalaki, some
r*?. Jr
80,000). There are around 25,000 speakers of
30 miles north west of Tiflis.
* ._" Abaza. In the 1970s, only a handful of people were
known to speak Ubykh — a language whose large
number of consonants (80) assures it a special place
in phonological studies (§28).
Most of the languages of the north-east belong
to the Dagestanian group. The main languages are
Avar (around 400,000), Lezghian or Kuri (around
300,000), Dargwa or Khjurkili (over 200,000),
>>v Lakk (over 80,000), and Tabasaran (over 50,000).
Several other languages are spoken by 10,000 or
fewer. Also in the north-east, the Nakh group of
languages comprises Chechen (over 600,000),
Ingush (around 150,000) and Bats (around 3,000),
found Georgian SSR. The
in a single village in the
linguistic profile of this area complicated by the
is

difficulty of drawing a clear line between language

and dialect (p. 284), and this has led to several
different estimates of the number of languages in
the Caucasian family.
Several of the northern languages have a written
form, based on the Cyrillic alphabet (p. 202), and
are used as state languages. There is much evidence
*
of the influence of previous periods of contact with
adjacent families (such as Arabic and Persian).
Today, the most noticeable influence on the family,
** especially in the area of vocabulary, is Russian.

52 OTHER FAMILIES •
305
Palaeosiberian
The once-extensive Palaeosiberian culture is now
represented by only a few thousand people scat-
tered throughout north-eastern Siberia. The lan-
guages they speak have been classified into four
groups, and since the 19th century these have been
studied together under the 'Palaeosiberian' head-
ing; but the groups are not genetically related to
each other, and therefore they do not constitute
a family in the linguistic sense. Nor are the links
with other families any clearer, though several
attempts have been made to trace connections with
other families found in the region.
The Luorawetlan group is the best-represented,
in the far north-east, consisting of Chukchi (about
12,000), Koryak (or Nymylan, 8,000), and Kam-
chadal (or Itelmen), Aliutor and Kerek, with only
a few hundred speakers between them. To the west,
the Yukaghir group is now represented by just a
single language (Yukaghir, or Odul), spoken by
around 500 people. Further west again, along the
Yenisey River, about 1,000 people speak the only
surviving member of the Yeniseian group — Ket
(or Yenisey-Ostyak). And to the south, about 3,000
speak Gilyak (or Nivkhi), which has no known
relatives. Since the earlier part of this century, each
of these languages has been given a written form,
based on the Cyrillic alphabet (p. 202).

Korean and Japanese

Korean There are evident similarities between the Kor-
ean language and the Altaic family, but is not clear
it

whether these can best be explained by a hypothesis

of common descent or one of influence through con-
tact. Thus in some classifications the language is
placed within the Altaic family, and in others it is left
isolated. A relationship with Japanese has also been
suggested.
Korean is spoken by well over 50 million people in
North and South Korea (where it is an official lan-
guage), China, Japan, and the USSR. The language
has been much influenced by Chinese: more than half
its vocabulary is of Chinese origin, and the earliest

records of the language, dating from before the 1 2th

century, are writtenin Chinese characters.

Japanese The genetic relationship between Japan-

ese and other languages has not been clearly estab-
lished. It is most often considered to be a member of
the Altaic family, but resemblances with other lan-
guage families of the region have also been noted.
There are several dialects, those in the south (and
especially in the Ryukyu Islands) displaying major
differences from the standard language based on the
Tokyo dialect.
Japanese is spoken by around 118 million people
on the islands of Japan, and by a further 2 million in
other parts of the world, especially in Brazil and the
United States. Apart from a few isolated forms, the first
written records of Japanese date from the early 8th
century, using Chinese characters, or /can/7 (p. 313).

306 • IX THE LANGUAGES OF THE WORLD

 there Uzbek (over 10 million), spoken mainly
is

in theUzbek SSR, and Uighur (around 6 million),

mainly found in China (Xinjiang) and nearby
USSR. In the north-west, the main languages are
Tatar and Kazakh (both around 6 million), Kirghiz
(around 2 million), and Bashkir (around 1 million),
found largely in the USSR, with some speakers in
China and nearby territories. In the north-east, lan-
guages are spoken by smaller numbers. There are
around 300,000 speakers of the geographically iso-
lated language, Yakut, and 150,000 speakers of
Tuvinian, but other languages number only tens
of thousands (including one named Altai - for-
merly, Oirot - which should not be confused with
Altaic, the name of the family as a whole). Chu-
vash, spoken by over H million in the middle Volga
region, is usually listed along with other Turkic
Chukchi
Peninsula ^^
f-'* languages, but many consider it to be a separate
branch within the Altaic family.
The main Mongolian language is known as
Mongol (or Khalka), spoken by around 4 million
people in the Mongolian People's Republic and
nearby China. Related languages in the same
region are Buryat (around 300,000), Santa (around
Luorawetlan V 200,000), Dagur, and Monguor (both fewer than
Kamchatka 100,000). Further west, the group is represented
Peninsula
Yukaghir 1 by Oirat (or Oyrat), Kalmyk (or Kalmuk) (both
around 140,000), and Mogol (around 50,000).
Yeniseian
||
There are many uncertainties of classification in
Gilyak ^$ this area, due principally to problems of applying

lorea

s—J
r
/ J

J J a P an
Korean

Japanese

Altaic
||s|j|

•Ryukyu islands
Japan is one of the world's
leading publishers of books,
magazines and, especially,
newspapers. The major
papers appear in morning
and evening editions. In
1984, the largest paper, pub-
lished by Yomiuri, had a
morning circulation of nearly
9 million, and an evening cir-
culation of nearly 5 million -
the highest in the world. It
has been estimated that on
average each Japanese
household reads two papers
a day.
\_
[ merly Tungus) may have as many as 30,000
Where Japanese, Korean,
Palaeosiberian, and Altaic are spoken
Altaic
The Altaic family of languages cover a vast area,
from the Balkan peninsula \o the north-east of Asia
— an area which includes the Altai mountain region
of central Asia, from which the family receives its
name. It comprises about 40 languages, classified
into three groups: Turkic, Mongolian, and Man-
chu-Tungus. The common ancestry of these groups
is maintained by many scholars; but this hypothesis
is contested by those who feel that the linguistic
similarities could be explained in other ways - such
as the mutual influences displayed when languages
are in contact with each other (p. 33).
As the name 'Turkic' suggests, the best-known
language of this group Turkish, spoken by
is
around 45 million people in Turkey and surround-
ing territories. Other main languages of the south-
west are Azerbaijani (12 million) and Turkmen
(around 3 million), both spoken mainly in the
USSR, Iran, and Afghanistan. In the south-east,
the distinction between language and dialect (p.
284).
The Manchu-Tungus group is spoken in a large
number of dialects over a wide area. Evenki (for-
speakers; but the other languages have fewer than
10,000 - Lamut (or Even), Nanai, and Manchu.
The Manchu people of north-east China number
over 3 million, but very few now speak the once-
important Manchu language - a lingua franca
between China and the outside world for over 200
years.
There is little evidence of the early development
of the Altaic family. Written remains of Turkic are
found in a runic script dating from the 8th century
ad; but Mongolian script dates only from the 13th
century; and the earliest Manchu records are even
more recent - mid- 17th century. Several writing
systems seem to have been used throughout the
early period.
In the 20th century, the most notable develop-
ments have come from the major political changes
which have affected the area since the First World
War. There has been a considerable effort to
modernize the languages, especially by promoting
fresh vocabulary. Several new literary languages
have emerged, based on local languages (as in the
case of Uzbek), and some of the older written lan-
guages have been reformed (seen most dramatically
in the case of Turkish, which in 1929 replaced Ara-
bic by Latin script).

52 OTHER FAMILIES- 307

Dravidian Where Dravidian languages are spoken
The Dravidian family is a group of over 20 lan-
guages, most of which are found close together in
the southern and eastern areas of India - though
one language (Brahui) is curiously isolated, being
spoken 1,000 miles away from the main family,
in the north of Pakistan. Through emigration,
speakers of the main Dravidian languages are today
found throughout South-east Asia, in eastern and
southern regions of Africa, and in cities in many
parts of the world.
The name given to the family comes from a
Sanskrit word, drdvida, which is used in an early
text with apparent reference to one of the lan-
guages, Tamil. Tamil has the oldest written records
of this family, dating from the 3rd century bc, and
scholars believe it to be close to the ancestor lan-
guage, known as Proto-Dravidian. But, despite the
historical records and associated reconstruction,
there is little agreement about the origins of the
language, or its speakers. One tradition speaks of
migration from lands to the south, now submerged;
other views suggest a movement from Asia, via the
north-west, perhaps around 4000 bc. A relation-
ship has been proposed with both the Uralic and Arabian
the Altaic language families, but the hypothesis is Sea The fight for independence
controversial. There is, however, strong support Soon after Indian indepen-
for the view that Dravidian languages were once dence, Dravidian language
spoken in the north of India, and were gradually militants began to fight for a
political structure in which
displaced by the arrival of the Indo-European
their separate languages
invaders (§51). Below: would be represented. In
The four main languages of the family are Demonstrations outside 1948, the report of the
Telugu, Tamil (both with around 50 million the Indian parliament, Linguistic Provinces Com-
speakers), Kannada (also known as Kanarese), New Delhi, in 1963 mission opposed any
protesting against a bill change, largely on the
and Malayalam (both with around 25 million
giving the English grounds that sub-nations'
speakers). Each language can be identified with a language associate status would be a major obstacle to
Sn Lanka
southern India - Andhra Pradesh, Tamil
state in with Hindi. The banners the spread of national con-
Nadu, Mysore, and Kerala, respectively. Of the read Death to English' sciousness in the new India.
four, Tamil has the greatest geographical spread, and 'Long live the Further pressure led to a se-
mother tongue'. cond committee of enquiry,
including several million speakers in Sri Lanka,
which reached the same
Malaysia, Indonesia, Vietnam, parts of East and conclusion, but accepted
South Africa, and many islands in the Indian and that, if the demand continued
South Pacific Oceans. The other languages are not to be matter
insistent, the
should be reopened. Many
so widely used outside India, though both Telugu
people saw in this the first
and Malayalam have some currency. Written real chance of success, and
records date from the 5th century AD for Kannada, from 1 949 the campaign in-
the 7th century for Telugu, and the 9th century tensified, especially among

for Malayalam. the Telugu speakers in Mad-

ras. The climax came in De-
Other languages with over a million speakers
cember 1 952, with the death
include Gondi, Kurukhi (or Oraoni), and Tulu. of Potti Sriramulu, who had
Brahui and Kui may have as many as half a million. chosen a hunger strike, the
Malto, isolated from the other languages in the time-honoured method of
Gandhi, to make his point.
north-east, is spoken by around 90,000 people. The
Prime Minister Nehru backed
remaining languages of the family have many fewer down, and the first of the lan-
speakers, sometimes numbering only a few thou- guage-based states, Andhra
sand — but it is not always obvious how to draw Pradesh (for Telugu) was in-
the line between language and dialect (p. 284). New augurated in October 1953.
Three years later, the whole
languages continue to be reported - Naiki, Pengo,
of south Indiawas reorga-
and Manda have been identified onlv since the early nized on the basis of linguis-
1960s. tic regions.

308 IX THE LANGUAGES OF THE WORLD


A group of Vietnamese chil-
dren learning English.
onhoWoy
Austro-Asiatic
Most of the languages of this family are spoken
inSouth-east Asia, in the countries between China
and Indonesia; but a few are found further west,
in theNicobar Islands and in parts of India. The
membership of the family, and its main subdivi-
sions, are not entirely clear.Few of the languages
have a written history, and classification has been
based on other methods (p. 293). Links between
this and other families (in particular, the Austrone-
sian family) have been proposed, but are uncertain.
Three main branches of the family are generally
recognized. The largest branch is the Mon-Khmer
group of languages, spoken throughout the south-
eastern mainland, mainly in North and South Viet-
nam, Laos, Kampuchea (before 1976, Cambodia),
and parts of Burma and Malaysia. It has three main
languages. Mon (or Talaing) is spoken in Burma
and Thailand by over half a million; Khmer (or
Cambodian), the official language of Kampuchea,
is spoken by over 5 million people. Inscriptions
in both languages date from the 6th-7th centuries
AD.
The main language of the group, Vietnamese,
poses something of a problem. This language is
spoken by around 50 million people in North and
South Vietnam, Laos, and Kampuchea, and in
recent years small groups of emigrants have taken
it to many parts of the world. Its status in the Mon-
Khmer group has, however, been disputed: some
scholars see it as a marginal member, while some
J Mon-Khmer
^ Munda
I Nicobarese
Where Austro-Asiatic languages
are spoken
relate it to the Tai family (p. 310). Its early history
is obscured by the use of Chinese throughout
South-east Asia — the result of over 1,000 years
of rule by China, which lasted until the 10th cen-
tury ad. The modern Latin-based alphabet, known
as Quoc-ngu ('national language'), was introduced
only in the 17th century.
The other two language groups are clearly separ-
ated geographically from the Mon-Khmer. The
Munda group of languages is spoken in several
parts of India, mainly in the north-east, but also
in a few central areas. Mundari (around li million)
and Santali (perhaps 5 million) are the most widely
used languages. Lastly, a tiny group of languages
is spoken by around 10,000 people on the Nicobar
Islands in the Bay of Bengal. These constitute a
separate, Nicobarese branch of the Austro-Asiatic
family.
There are well over 100 Austro-Asiatic lan-
guages. Exactly how many depends on the distinc-
tion drawn between language and dialect (p. 284),
and on the criteria used to demonstrate structural
similarities (p. 293). A few other languages spoken
in Malaysia and India have at times been proposed
as members of the family. Nothing is known about
the early movements of the peoples involved. It
is possible that the various groups of languages
which make up the Mon-Khmer branch began to
split up in the second millennium bc, but where
the Austro-Asiatic peoples came from, and when
they migrated, remains pure guesswork.
52 OTHER FAMILIES •
309
Tai
Where Tai languages are spoken China
The Tai family of languages are all found
South- in
east Asia, in an area centred on Thailand, and
extending north-eastwards into Laos, North Viet-
nam, and China, and north-westwards into Burma
and India. The spelling 'Tai' is used to avoid confu-
sion with the main language of the family, Thai
(or Siamese), which is the official language of Thai-
land.
The 40 or so Tai languages are usually divided
into threegroups: south-western, central, and
northern. Most speakers belong to the south-west-
ern group, which includes Thai, spoken by around
30 million people in a wide range of dialects, and
Lao (or Laotian), widely spoken in Thailand, and
the official language of Laos (10 million). Shan
(mainly in Burma), and Yuan (Thailand), both have
over 2 million speakers. Other languages of this
group have relatively few: of the central and north-
ern groups, only Nung and Tho have over 100,000
speakers. But in this part of the world, such esti-
mates are very approximate.
The relationship between the Tai family and
other languages is unclear. Written remains of the
south-western group date from around the 13th
century. Links have been proposed both with the
Sino-Tibetan and the Austronesian families. In par-
ticular, several languages of south-west China,
belonging to the Kadai and Kam-Sui groups, dis-
play interesting similarities to Tai.

Sino-Tibetan
The membership and classification of the Sino-
Tibetan family of languages
Where Sino-Tibetan languages are
is highly controversial.
The 'Sinitic' part of the name refers to the various
spoken (including Miao-Yao)
Chinese languages (often referred to as 'dialects');
the 'Tibetan' part refers to several languages found
mainly in Tibet, Burma, and nearby territories. But
as there are notable similarities with many other
languages of the region, some scholars adopt a
much broader view of the family, so as to include
the Tai and Miao-Yao groups (p. 311).
The Sinitic languages (see p. 312) are spoken
by over 1,000 million people. The vast majority
of these are in China (over 980 million) and Taiwan
(19 million), but substantial numbers are to be
found throughout the whole of South-east Asia,
especially in Hong Kong, Indonesia, Malaysia,
Thailand, and Singapore. Important Chinese-
speaking communities are also found in many other
parts of the world, especially in the USA.
South China Sea
There are nearly 300 languages in the Tibeto-
Burman family, and these have been classified in Burmic languages f|
several different ways. It is possible to identify 'clus-
ters' of languages which have certain features in
Miao-Yao languages I
common, such as the 50 or so Lolo languages, Sinitic languages
[_
spoken by around 3 million people in parts of
Burma, Thailand, Vietnam, Laos, and China. The Tiberic languages
[_
Indian Ocean
80 or so Naga, Kuki, and Chin languages, spoken

310 IX THE LANGUAGES OF THE WORLD

Biggest book The whole
Buddhist canon is carved on
stone slabs housed in 729
stupas in the Kuthodaw Pa-
goda in Burma. The achieve-
ment, sometimes called the
'world's largest book', was
created by King Mindon in
1872, with the help of 2,400
monks.
Facing page
Thai traders in Bangkok's
Floating Market.
in Burma and India, comprise another group. But
groupings of kind display many differences as
this
well as similarities, and it has not yet proved pos-
sible to find a neat way of classifying these, and
the other groups thought to belong to the same
family, into two or three types. It is by no means
clear, for example, whether the small group of
Karen languages, spoken by around 2 million peo-
ple in Burma, should be included or excluded from
the Sino-Tibetan family.
After Chinese, Burmese and Tibetan are the two
main languages of this family. Burmese is spoken
by over 25 million people in Burma as a mother
tongue, and several million more use it as a second
language throughout the region. It has written
records dating from the 11th century. Speaker esti-
mates for Tibetan are very uncertain, largely
because of the influence of Chinese in recent years;
but a figure of 3—4 million seems likely. There are
several major dialects, which are sometimes viewed
as separate languages. Written records date from
the 8th century AD, treating largely of Buddhist
religious subjects. The alphabet of this period,
which pronunciation of the time, is still
reflects the
in use today, with the result that there is consider-
able divergence between spelling and modern Tibe-
tan speech.
Miao-Yao
This is a small group of languages spoken in southern
China and adjacent parts of South-east Asia -
especially northern Laos, Thailand, and Vietnam. The
two chief languages, which give the group its name,
are Miao (also called Hmong, or Meo), spoken by over
2i million people, and Yao (also called Man), spoken
by nearly a million. The sub-classification of the group
into languages and dialects is controversial, as indeed
is its status as a separate language family. Links with
Tai, Mon-Khmer, and Sino-Tibetan have been sug-
gested, and it is within the latter family that Miao-Yao
languages are most often placed.
52 OTHER FAMILIES- 311
 An oracfe bone This rem-
The languages of China nant of China's Bronze Age,
from the Shang dynasty of
Because there has long been a single method for the 2nd millennium bc, was
writing Chinese, and a common literary and cul- found in the Anyang district
tural history, a tradition has grown up of referring of Henan. Shang kings had
questions inscribed on
to the eight main varieties of speech in China as
polished ox bones or tortoise
'dialects'. But they are as different from each
in fact
shells. A pattern of holes was
other (mainly in pronunciation and vocabulary) as drilled in them, and they
French or Spanish is from Italian, the dialects of were then heated until they
the south-east being linguistically the furthest cracked. The pattern of
cracks produced by this pro-
apart. The mutual unintelligibility of the varieties
cess was interpreted as the
is main ground for referring to them as separate
the ancestral spirits' way of ans-
languages. However, it must also be recognized wering the questions. Thou-
that each variety consists of a large number of dia- sands of oracle bones have
lects, many of which may themselves be referred now been discovered, and
There have been moves to reform the language the study of their inscriptions
to as languages. The boundaries between one so-
from as early as the 2nd century bc, but there has has now become a new
called language and the next are not always easy branch of Chinese linguistic
been nothing to equal the complexity of the pre-
to define. studies, known as jiagushu.
sent-day programme, in which frequent reference
The Chinese refer to themselves and their lan-
is made to the names of several different varieties
guage, in any of the forms below, as Han — a name
of the Chinese language.
which derives from the Han dynasty (202 bc-
ad 220). Han Chinese is thus to be distinguished
Wen-yan ('literary speech' or 'body of classical
from the non-Han minority languages used in
writing'). The cultivated literary language, re-
China. There are over 50 of these languages (such
corded from around 1,500 BC, and the traditional
as Tibetan, Russian, Uighur, Kazakh, Mongolian,
unifying medium for all varieties of Chinese. Its
and Korean), spoken by around 6% of the popula-
complex system of characters is explained on
tion.
p. 200. It differs greatly from everyday speech,
especially in its terse grammatical style and specia-
THE CHINESE LINGUISTIC lized literary vocabulary.It is now less widely used,
REVOLUTION because of the success of the current reform move-
The 20th-century movement for language reform
ment for written Chinese.
in China has resulted
in the most ambitious pro-
gramme of language planning (§61) the world has
ever seen. The programme has three aims:
(i) to simplify the characters of classical written
Where the main Chinese 'dialects'
Chinese, by cutting down on their number, and are spoken i J
i
reducing the number of strokes it takes to write Heilongiiang
-H
a character;
-<"
(ii) to provide a single means of spoken commu-

nication throughout the whole of China, by popu- _.J

larizing the Beijing-based variety, which has been
M o n g o 1 i

_ S < J
chosen as a standard;
(iii) to introduce a phonetic alphabet, which

would gradually replace the Chinese characters in

everyday use.
Xinjiang A --_.
C H 1 I
Hebei
Vy*"-> rZ /
<
n, X.f X/
A-^~ (^J Y
y < I

)
Shan*'( ^Shandong/

Q'"^' > fmui

Gansu
ijansu Cci_ *Z Henan f "^ J

i
(
C Jshaanxi^
/Shaanxi*" / f*

'Dialect' Where spoken i

Cantonese (Yuen) In the south, mainly Guangdong,

Hakka
southern Guangxi, Macau, Hong Kong.
Widespread, especially between Fujian bet \ Sichuan
"*
K
A ^ XZheiiang
V"* ^.Jiangxi A_-
and Guangxi.
J „ N-MIN
Hsiang (Hunan) South central region, in Hunan. Hunan /
,
% (.
Kan Shanxi and south-west Hebei.
Mandarin A wide range of dialects in the northern, 1
V i r'Guizhou i «"\ ^•ka 4
central and western regions. North t v_.Guangdon
,
Mandarin, as found in Beijing, is the / -VGuangxi %*- Taiwan
basis of the modern standard language. Yunnan
1
«=» C*J
) jO
Northern Min
-\ .i^yrf Hong Kong
(Min Pei)
North-west Fujian.
. —f ^SMacau
Southern Min The south-east, mainly in parts of
(Min Nan) Zhejiang, Fujian, Hainan Island and
Hainan
Taiwan.
Wu Parts of Anhui, Zhejiang, and Jiangsu.

312 IX THE LANGUAGES OF THE WORLD

Bai-hua ('colloquial language'). A simplified, ver- Romanizing
nacular style of writing, introduced by the literary Chinese
reformer Hu Shih in 1917, to make the language
Several systems of romani-
more widely known to the public, and to permit zation for Chinese have
the expression of new ideas. A style of writing been invented. The oldest in
which reflected everyday speech had developed as current use is known as
Wade-Giles, introduced by
early as the Sung dynasty (ad 960-1279), but had
Sir Thomas Wade in 1859,
made little impact on the dominant wen-ydn. How- and developed by his suc-
ever, the 'May Fourth Movement' (which origi- cessor in Chinese Studies
nated in political demonstrations on 4 May 1919 at Cambridge University,

after the Paris Peace Conference) adopted Hu Herbert Giles. This is the
system which is most familiar
Shih's ideas, and bai-hua was recognized as the
to western eyes. In the
national language in 1922. 1 930s, a system known as

gwoyeu romatzyh ('national

Putonghua ('common language'). The variety romanization') was devised
by Lin Yu-t'ang and Chao
later chosen as a standard for the whole of China,
Yuen-ren. During the
and widely promulgated under this name after the Second World War, Yale
establishment of the People's Republic of China University introduced an
in 1949. (In Taiwan, it goes under the name of intensive programme of Chin-
guo yii, or 'national speech'; in the West, it is gener- ese training for Air Force
pilots,and introduced a new
ally referred to simply as 'Mandarin'.) It embodies
system, related more clearly
the pronunciation of Beijing, the grammar of the to American pronunciation.
Mandarin dialects, and the vocabulary of collo- But pin-yin has now become
quial Chinese literature. In 1956, it became the the dominant system. The

medium all schools, and a policy

of instruction in
name for China illustrates
some of the differences be-
of promoting use began (p. 365). It is now the
its
tween these systems:
most widely used form of spoken Chinese, and is Chinese typewriter
the normal written medium for almost all kinds The complexity of classical writing is well illustrated
device - a Chinese typewriter. The tray contains over 2,000
by this t IS
of publication.
characters, with several thousand more being available on Chinese characters
other trays. The typist first aligns the tray, then presses a key, Wade-Giles Chungkuo
Pin-yin ('phonetic spelling'). After several pre- which makes an arm pick up the required character and strike Gwoyeu
vious attempts to write Chinese using the letters it against the paper. The machine can type vertically and hori- romatzyh Jonhhwo
roman alphabet, this 58-symbol writing sys-
of the zontally. It is a slow process, with good typists averaging at Yale Junggwo
most 20 characters a minute. Pin-yin ZhongguP
tem was finally adopted in 1958. Its main aims
are to facilitate the spread of putonghua, and the Some familiar spellings, with
learning of Chinese characters. Pin-yin is now in instance, Mao Zedong with a marked
spoke it their pin-yin equivalents:

widespread use. In the 1970s, for example, a new
map of China was published using the alphabet,
and a list of standard spellings for Chinese place
names was compiled. New codes were devised for
such diverse uses as telegraphy, flag signals, braille,
and deaf finger-spelling (p. 225).
The programme is not en-
future of the reform
tirely clear. It may
be that pin-yin will ultimately
supplant the general use of characters, or there may
be a reaction to preserve the traditional written
language. With putonghua, new varieties of re-
gional pronunciation are certain to develop (for
Hunan accent), which may lead to problems of
Peking Beijing
intelligibility. And if putonghua is to succeed as
Canton Guangzhou
a popular means of communication, it needs to Mao Tse-tung Mao Zedong
anticipate the potential conflict with local regional
dialects (for example, whether local words should
be used). Much will depend on how flexibly the
authorities interpret the notion of standard, and
whether they are able to achieve a balance between
the competing pressures of respecting popular
usage (where there is a strong case for variety) and
the need for national communication (which could
lead to a form of centralized laying down of pres-
criptive linguistic rules).

Tigers do not breed dogs' 'Calamities do not Dccur singly'

<

These phrases, usually of four characters, illustrate The equivalent phrase in

the telegraphic literary style of w6n-y&n. The nearest English is 'It never rains but
equivalent to this proverb in English is perhaps 'Like it pours.'
father, like son.' Mao Zedong was particularly adept
at incorporating classical features of this kind into his
political speeches.

JL
hu
(tiger)
5C
fu
(father)
* * * ^ ^
wu quan z! hud bu dan
n xing
(no) (dog) (son) (calamity) (no) (single) (act)

52 OTHER FAMILIES -313

The languages of Africa
Africa contains more languages than any other con-
- around
tinent 1,300, spoken by over 400 million
people. The language total is uncertain, because
many areas are inaccessible, and many dialect
groups have not been well investigated, but it is
probably an underestimate. Very few of these lan-
guages are spoken by large numbers: less than 5%
have more than a million speakers. As a conse-
quence, Africa is a continent of lingua francas. Ara-
bic is used throughout the north and north-east;
Swahili is used throughout most of East Africa;
English and French are widespread, often as official
languages, in former colonial territories; and, espe-
cially in West and Central Africa, several languages
have come to be used as ways of fostering commu-
nication between different tribes (such as Hausa,
Bambara-Malinka, Wolof, Kongo, Lingala, and
various pidgins and Creoles, such as Pidgin English,
Krio, and Sango).
The most widely accepted classification of
African languages recognizes four main families,
though there is considerable difference of opinion
about the boundaries between them, and about
several of the language groups which they subsume.
There is little historical evidence available to aid
classification. Written records of most African lan-
guages have existed only since missionary activities
began on the continent, less than 150 years ago.
As a consequence, the field of African languages
has proved to be one of the most controversial areas
within the domain of comparative linguistics.
NIGER-CONGO
This is the largest African family, with around
1,000 languages, and several thousand dialects,
whose status is often difficult to determine. The
family spreads across the whole of sub-Saharan
Africa, west of the River Nile, and extends along
the eastern half of the continent as far north as
the Horn of Africa. It is usually divided into six
groups of languages, which are estimated to have
diverged well over 5,000 years ago.
The largest group is the Benue-Congo — around
700 languages spoken throughout central and
southern Africa, over 500 of them belonging to
the Bantu sub-group. The main Bantu languages
are Swahili (4 million native speakers, but used
by about 30 million speakers as a lingua franca),
Kongo (7 million), Rwanda (15 million), Makua
(6 million),Xhosa and Zulu (sometimes considered
dialects of the same language, but considered by
their speakers to be different languages - both over
5 million). The largest non-Bantu languages are
found in Nigeria, Efik (4 million), and Tiv (2 mil-
lion).
The Adamawa-Eastern group of around 90 lan-
guages is spokenremote, northern part of
in the
central Africa. Its main members are Sango, a pid-
ginized language used throughout the Central Afri-
can Republic (H million) and Gbaya (300,000).
Several other pidgins are spoken in the area.
314 • IX THE LANGUAGES OF THE WORLD
The Kwa group of over 800 languages is spoken
in the southern part of the bulge of West Africa.
It contains several important members, especially
Yoruba (17 million), Igbo (13 million), Akan (8
million, Ewe (2 million), and Ijo (H million). Eng-
lishor French are official languages in the area.
The Voltaic (or Gur) group consists of over 70
languages spoken in a broad area around the Upper
Volta River. Its main member is Mossi, spoken by
around 4 million.
The West Atlantic group, as its name suggests,
consists of over 40 languages spoken extreme
in the
western part of the African bulge. Fulani is the most
widespread language, spoken by around 15 mil-
lion.
The Mande group of over 20 languages is also
spoken in the western part of the bulge. Its main
members are Bambara (2 million), Malinka (3 mil-
lion), Dyula, and Mende (each over 1 million).
NILO-SAHARAN
The major group of languages in this family is
spoken in two areas around the upper parts of the
Chari and Nile rivers, and is generally referred to
by the name Chari-Nile. It contains around 100
languages, whose sub-classification has given rise
to much controversy. In particular, scholars have
argued for over 100 years about the best way of
classifying the languages spoken along the Nile,
in Sudan, Uganda, and nearby territories (the so-
called Nilotic group). A postulated western branch
includes such languages as Luo [1\ million), Dinka
(2 million), Acholi, and Lango (both with over half
a million speakers). The larger languages of the
eastern, or Nilo-Hamitic branch, are Nandi, Bari,
and Masai, all with around half a million speakers.
But the closeness of the relationship between the
two branches is disputed, as is the implication of
relatedness to the Hamitic group of languages.
Other languages in the area with over 100,000
speakers are Lendu, Mangbetu, Lugbara, Madi,
and Gambai. Nubian, spoken in Sudan and Egypt,
has around a million speakers. It is the only lan-
guage of this group to have a long written history,
with manuscripts in a modified Coptic alphabet
dating from the 8th century ad (Old Nubian).
Argument also surrounds the relationship
between the Chari-Nile languages and the 20 or
so other languages which have been grouped into
the Nilo-Saharan family. Particularly unclear is the
status of Songhai, spoken by over a million people
in a wide area between Mali and Nigeria. Fur
is another isolated language, spoken by around
400,000 in the Sudan and Chad. The remaining
languages have been classified into small groups:
Saharan, Maban, and Koman. The Saharan lan-
guage, Kanuri, is the largest, with over 3 million
speakers.
There is often no clear corre-

between language
lation
groups and racial or cultural
groups in Africa. The Pyg-
mies, who live in the central
forests,have no single lan-
guage, but use the lan-
guages of neighbouring peo-
ples. By contrast, the
Khoisan of South Africa do
form a homogeneous linguis-
tic and racial group. The pic-
ture shows a group of Khoi-
san tribesmen cutting up a
hartebeest.

KHOISAN
This is the smallest language family in Africa, con-
sisting of fewer than 50 languages; but they are
a well-known group because of their use of click
consonants (p. 126). They are spoken in the south-
ern part of Africa, in an area around the Kalahari
Desert extending from Angola to South Africa,
though two click languages are spoken as far north
as Tanzania (Sandawe and Hatsa). 'Khoisan' de-
rives from the name of the largest Hottentot group
(the Khoi-Khoin) and that of the Bushmen in the
* *
Nama region of Namibia (San). Few of these lan-
guages have more than 1,000 speakers — only
Kwadi and Sandawe (spoken in Tanzania) have
over 10,000. The numbers are diminishing, and
several languages are known to have become
extinct. About half of the languages have been
given a written form by missionaries and others.

52 OTHER FAMILIES -315

AFRO-ASIATIC
This family, also known as Hamito-Semitic, is the
major family to be found in North Africa, the east-
ern horn of Africa and south-west Asia. It contains
over 200 languages, spoken by nearly 200 million
people. There are six major divisions which are
thought to have derived from a parent language
that existed around the 7th millennium bc.
The Semitic languages have the longest history
and the largest number of speakers. They are found
throughout south-west Asia, including the whole
of the Saudi Arabian peninsula, and across the
whole of North Africa, from the Atlantic
Red Sea. The oldest languages of the group,
extinct, date from the 3rd millennium bc; they
include Akkadian, Amorite, Moabite, and Phoeni-
cian, all once spoken in and around the Middle
East. There was a vast literature in Akkadian, writ-
ten in cuneiform script (p. 198). Hebrew dates from
the 2nd millennium bc; its classical form was pre-
served as the written language of Judaism;
modern spoken and written form used by around
is
4 million people in Israel and throughout the
world. Old Aramaic, the language of Jesus and the
Apostles, also dates from this period. Aramaic dia-
lects are still spoken by tiny groups in the Middle
East. form of one of these dialects —
A modern
— is spoken by as many as
Syriac (or Assyrian)
a million people in and around the Middle East,
and in the USA.
The major language of the group is Arabic,
spoken by over 150 million people as a mother
tongue, and used by several million more as a
second language. It exists in both classical and col-
loquial forms. Classical (or literary) Arabic
sacred language of Islam, and is used as a lingua
franca of educated people throughout the Arabic-
speaking world. Colloquial Arabic exists in many
modern dialects, not all of which are mutually intel-
ligible - they include Algerian, Moroccan, Egypt-
ian, Syrian, Iraqian, and several dialects of Arabia
and the Sahara. Maltese, spoken by over 300,000
people on the island of Malta, is also a development
from Arabic.
316 • IX THE LANGUAGES OF THE WORLD
In the south of the region, in Ethiopia, there are
several Semitic languages, notably Tigrinya (3| mil-
lion),Amharic (the official language of Ethiopia,
used by around 13 million), and Gurage (around
1 million).
The remaining branches are less widespread.
Egyptian is now extinct: its history dates from
before the 3rd millennium bc, preserved in many
hieroglyphic inscriptions and papyrus manuscripts
(p. 199). Around the 2nd century ad, it developed
into a language known as Coptic. Coptic may still
have been used as late as the early 19th century,
to the and is still used as a religious language by Mono-
now physite Christians in Egypt.
There are over 20 Berber languages spoken
throughout North Africa by around 12 million
people, mainly in Algeria and Morocco. They
include Riff, Kabyle, Shluh, and Tamashek, the
widely scattered language of the Tuareg nomads.
There are about 30 Cushitic languages, spoken by
its around 13 million people. The largest is Oromo
(or Galla), several dialects of which are spoken in
Ethiopia and Kenya by over 10 million people; and
Somali, spoken in Ethiopia, Somalia and Kenya
by nearly 5 million. There are over 20 Omotic lan-
guages, spoken by nearly 2 million people in west-
ern Ethiopia and northern Kenya. About half of Tifinagh Below left: The
Tuareg tribesmen have pre-
these are speakers of Walamo. Omotic is some-
served a unique form of writ-
times classified as a western branch of Cushitic. ing, known as tifinagh ('char-
Lastly, there are over 100 Chadic languages, whose acters'). It derives from
status within the Afro-Asiatic family is less clear. Numidian, an ancient Libyan
These languages are spoken by over 25 million peo- script used in Roman times.
The system consisted only of
ple in an area extending from northern Ghana to
consonants, usually written
the Central African Republic. Hausa is undoubtedly right to left.
is the the most important language of this group, spoken
by around 25 million people as a mother tongue, Below: Maltese is the only
form of Arabic to be written
and by several million more as a second language
in the Latin alphabet. It is re-
throughout the region. It is the only Chadic lan- lated to the western Arabic
guage to have a written form — a roman alphabet dialects, but it also shows
now being used in place of the Arabic script intro- the marked influence of the
duced in the 16th century. Romance languages (via
This holiday advertise-
Sicily).
ment shows some of the
symbols needed to cope with
the Semitic sounds.
ISSA TISTA TIEHU
MAYFAIR HOLIDAY
fl'Londra — York — Birmingham — Glasgow —
Edinburgh — Ruma — Parigi — Amsterdam —
Zurich — Geneva — u Yugoslavia
MAYFAIR HOLIDAYS jmkludu l-vjagg bir-ntorn. Ijkanda u breakfast kollox
Imnallas
Ipprofitta ruhek mill-esperjenza taghna u ffranka l-inkwiet u xognol bla' bzonn
L'UNTOURS joffrulek ukoll APEX FARES ghall-lngilterra. I'America u l-CANADA
bl'cnsurance gnalihom
Fares Specjali ghall-Awstralja
u koicessjom kbira onra fl'insurance gnall-vjaggi —
Nassigurawk ghal kull ma jista'
jignkit waqt il-vjagg. minn qabel ma' titlaq sa wara tasal lura. anke jekk ga tbati
li
minn xi mard jew difett fiziku Bizzejjed it-tabib jikkonferma tista ssiefer
li
Progrcmmi, bookings u nformozijoni minghand:
UNTOURS LIMITED
Workers' Memorial Building. 66 Tnq it-Torn. tas-Sliema
Triq l-lfran. il-Belt. Branch Office.
Tel 23640. 21966. 24451 Tel 36477. 36478
Austronesian
The Austronesian language family covers a vast
geographical area, from Madagascar to Easter
Island, and from Taiwan and Hawaii to New Zea-
land - a territorial range which is reflected in an
alternative name often given to the family: Malayo-
Polynesian. It is one of the largest families, in terms
of bothnumber of speakers (around 200 million)
Indo-Pacific
There are over 600 languages spoken in New
Guinea, and about a further 100 spoken in the
islands to the immediate east and west, which do
not belong to the Austronesian family. Two small
language groups lie much further away from those
spoken in the New Guinea region, and some scho-
lars believe that there enough evidence to justify
is

and number of languages (at least 500, and perhaps their placement withinthis group: Andamanese,
as many as 700). from the Andaman Islands in the Bay of Bengal;
In this part of the world, it is particularly difficult and Tasmanian, now extinct, from the island of
to establish language identities. Apart from the Tasmania, to the south of Australia (p. 319). Fewer The Easter Island Inscrip-
tions (illustrated here on a
usual problems of distinguishing dialects from lan- than 3 million speakers are involved. carved wooden gorget) are
guages (p. 284), several different names may be Over half of the Indo-Pacific languages have been pictures with magical signifi-
used with reference to the speakers in an area, and shown to be related, especially many of those in cance, and not a form of writ-
it is never obvious whether these names refer to central New Guinea. But the linguistic picture is ing. The literature of Oceania
is oral, consisting of many
different languages, or are simply alternative names by no means certain: in the more inaccessible parts memorized passages for
for the same languages. For example, over 70 of New Guinea, there are still tribes who have not use in social events, and
names have been recorded for the various dialects been contacted, and whose languages are not also lengthy myths about
of the Dayak language of north-western Borneo known; and data are sparse on many others, which events and places, which
would be formally recited.
and south Sarawak, but it is possible that research may have only tens or hundreds of speakers. Many
The language was held in
will show several of these to be so different that different classifications have been proposed, some
special esteem: the reci-
they could legitimately be counted as distinct lan- of which recognize over 100 sub-families. Other tation of a myth belonging to
guages. The linguistic picture is also complicated names for the family, such as Papuan, are also in some other clan was consi-
by the existence of many pidgins and Creoles (§55) use. dered theft.

which have grown up as the result of trade contacts There is nowhere compare with the multi-
to
in the area. Moreover, several languages have come lingual diversity of New Guinea — so many lan-
to be extensively used as lingua francas — notably guages crammed into an area of only 300,000
Bahasa Indonesian, Bazaar Malay, Chinese, Eng- square miles, and containing a total population of
lish, and French. only around 6 million. A sense of this complexity
The Austronesian family is usually divided into can be obtained by 'translating' these figures into
two main groups, the boundary falling within the British terms: Britain, one third of the size, would
island of New Guinea. The Western group contains 200 languages,
find itself containing nearly separ-
perhaps 400 languages, spoken in Madagascar, ated from each other by distances of only 20 miles.
Malaysia, the Indonesian Islands, the Philippines,
Taiwan, parts of Vietnam and Kampuchea, and
the western end of New Guinea. Two languages
of Micronesia (Chamorro and Palauan) are also
included. The Eastern group, usually referred to
as Oceanic, contains around 300 languages,
spoken over most of New Guinea, and throughout
the 10,000 or more islands of Melanesia, Microne-
sia, and Polynesia. Despite its geographical and
linguistic diversity, only a small minority of
speakers (under 2 million) belongs to the eastern
group.
Because of the many structural differences
between the languages, it is estimated that the Aus-
tronesian family has a history of development of
over 4,000 years, with archaeological and linguistic
evidence suggesting a probable geographical origin
in the New Guinea area. But despite extensive
research into Austronesian languages in recent
years, the early history of this family remains ob-
scure and controversial, and several competing
linguistic sub-classifications have been proposed.

A map of the area covered by the Indo-Pacific and Austronesian

families, and more detailed information about some of the lan-
guages, is provided on pp. 318-319.

52 OTHER FAMILIES -317

Andamanese A tiny number of people speak vari- Tok Pisin This pidgin language is widely used
ous dialects of Andamanese, on the Andaman within Papua New Guinea, especially in the north
Islands in the Bay of Bengal. There were around of the country. It is spoken by over 750,000 people
500 speakers in the 1950s, but this figure had been — by many, as a mother tongue ($55).
more than halved by the early 1980s. The language
is not clearly related to any of the others in the Sundanese There are around 20 million speakers
region, but a tentative placement alongside other of Sundanese, found throughout the western part
Indo-Pacific languages has been proposed. of Java. Written records date from the 14th cen-
tury.
Pilipino Pilipino is the name given to the national
language of the Philippines, when the country Javanese Javanese has the largest number of
became independent in 1946. It is a standardized mother-tongue speakers in the area - well over 60
form of Tagalog, an indigenous language spoken million. It is spoken throughout the island of Java,
in central and south-western parts of the island and to some extent in Malaysia. It has a strong
of Luzon, which includes the capital, Manila. There written literary tradition, dating from the 8th cen-
are around 12 million native speakers of Tagalog, tury, which continues to flourish, although some-
but over 30 million now speak Pilipino, which what eclipsed these days by the influence of the
along with English is the medium of instruction standard language, Bahasa Indonesian.
in schools. There is an extensive literature of folk-
lore and mythology in Tagalog, and also in the
other major indigenous languages of the Philip-
Shadow-puppet plays are
pines, Cebuano (over 10 million speakers through- found in many parts of
out the south) and Ilocano (over 5 million speakers South-east Asia, originating
in the north). A movement for a new national lan- in Java 1 ,000 years ago.
Carved or leather puppets,
guage, Filipino, began in the 1970s.
fixedon sticks, represent
mythological figures. Light
Malay Malay is extensively spoken throughout from a flame lamp passes
Malaysia and Indonesia as a second language, and through holes in the puppets
onto a cloth screen, to pro-
is the mother tongue for around 10 million people.
duce a spectacle of flickering
The dialect of the south Malay Peninsula has shadows which symbolize a
become the standard language, and under the name mystical world. The narrator
of Bahasa Indonesian (bahasa = 'language') has follows a basic scenario, us-
ing stock phases to introduce
been the official language of Indonesia since 1949;
heroes and events; but there
often referred to simply as Indonesian. A pidgi-
it is
ismuch improvization, ad-
nized form of Malay, known as Bazaar Malay, is ding music, satire, and local
widely used as a lingua franca throughout the Indo- details, and introducing
nesian archipelago. Its use long predates the time many voice qualities. The
plays often last all night long.
of contact with European languages, in the 17th
century. Another form, known as Baba Malay, is
used by Chinese communities in Malaysia. Written
records of Malay date from the 7th century ad,
consisting of various inscriptions found on Suma-
tra. The modern standard language alphabet is now
different from that of the older, literary Malay,
because of the introduction of spelling reforms.
^^maiaysia m

Malagasy Indonesian migrated to the

traders
uninhabited island of Madagascar during the 1st Sumatra

millennium ad, the linguistic differences with other ^W_ lava r--
1
(

Austronesian languages suggesting that the sepa- Nd nG

ration took place towards the beginning of the o ;Ts^
period. Despite its closeness to Africa, the language
shows only the occasional influence of African lan-
guages and Arabic. There are now around 9 million
speakers. The standard language is based on the
dialect of the largest ethnic group, the Merina ('ele-
o Madagascar

Indian Ocean
vated people' — those who lived on the plateau),
who were dominant in the 19th century. Various Eastern Austronesian languages
dialects of Malagasy are also used on several of
J
the islands in the region around Madagascar. J Western Austronesian languages Tasmania
'%%/fy Indo-Pacific languages

318 IX THE LANGUAGES OF THE WORLD

 '

Hanuman the monkey god,

comes to the aid of Sita in Motu Motu spoken by around 10,000 in the
is Chamorro One of the two languages in Microne-
the Ketyak dance, in Ubud, central part of Papua New Guinea. A pidginized other being Palauan) which belong to the
sia (the
Bali. This dance makes a
variety (once known as Police Motu, because it western branch of Austronesian. It is spoken by
ritual use of male voices. A

large number con-

of men sit
became the language of the multilingual police around 50,000 people, mainly on the island of
centrically around a lamp, force) developed as a trade language between Guam, where the official language is English.
chanting rhythmical sets of speakers of Austronesian and Indo-Pacific lan-
heavily emphasized syl- guages; it is now called Hiri Motu. It is spoken Tasmanian There were five main dialects spoken
lables. The men in their
by around 150,000 throughout the country, where in Tasmania, and these have now been classified
trance-like states are thought
to be possessed by monkey
it has official status. as two languages. There is little information avail-
spirits. able about them: the last known speakers died to-
wards the end of the 19th century. Tasmanian is
not clearly related to any other language, but some
scholars feel that a placement within the Indo-Paci-
fic family is justified.

Maori The Maori population has been steadily

increasing during the 20th century, and there has
been a revival in the language and culture. Maori
is now an optional second language in schools.

There are around 100,000 speakers, all of whom

are bilingual in English. Maori is still to be heard,
in the form of songs, speeches, and ritual chal-
and when official visit-
lenges, at special gatherings,
ors to New Zealand are being formally received
(p. 49).

Gilbertese This has the largest number of

speakers of all the languages of the eastern Austro-
nesian branch in Micronesia. It is spoken on the
16 coral atolls which constitute the Gilbert Islands,
named after the Englishman Thomas Gilbert, who
arrived there in 1788.

Tahitian Tahitian is widely used as a lingua

franca throughout French Polynesia, and is the
native language of the Society Islands. There are
perhaps 70,000 speakers in all. It is an official
regional language in Tahiti.

Tongan This is the national language of the

Kingdom of Tonga, also known as the Friendly
Islands. It is spoken by around 80,000 people.

Samoan There are around 200,000 speakers of

Samoan Western Samoa (where it is an offi-
in
cial language, along with English) and American
Samoa. Several sizeable communities now live in
New Zealand and the USA.
Fijian Over 200,000 people speak Fijian as a first
MELANESIA \ c or second language of the Fiji Islands. This is less
i- , \ w
*. Samoa ,
Is
Fin Is. ^
"• Society Is.
than half the population of the islands, the re-
'/ •Tonga mainder being Indian, Chinese, and other immi-
^ Is.

grants. The standard form, based on the Bauan

'
T-
Easter Islands dialect, is used in broadcasting and in the press,
along with Hindi and English.

Easter Island One of the first islands to be settled

in Polynesia. Only a small number of its tiny popu-

J New Zealand Where Austronesian and

Pacific languages are spoken
Indo-
lation of 1,800 speak the language,
known as Rapanui or Pascuanese.
which is also

52 OTHER FAMILIES • 319

The languages of the Americas

NORTH AMERICAN INDIAN

LANGUAGES
There were originally around 300 languages
spoken by the indigenous American Indian (or
Amerindian) tribes, but this number had more than
halved by the 1970s. Many of the languages are
now spoken by only a few old people. Only about
50 of the languages have more than 1,000 speakers;
only a handful have more than 10,000. In the mid-
1970s, the total number of speakers was estimated
at around 300,000.
The Amerindian languages have been classi-
fied into over 50 families, showing many kinds
and degrees of interrelationship. However, this
allows a great deal of scope for further classifica-
tion, and Amerindian linguistics has thus proved
to be a controversial field, generating many propo-
sals about the links between and within families
(see further, p. 322). It is not known whether the
languages have a common origin. The peoples are
The Macro-Siouan family consists of 26 lan- A Jicarilla Apache chief and
thought to have migrated from Asia across the Ber- from the North
guages, spoken in a broad swathe from Canada his family,
ing Strait, perhaps in a series of waves, but the American Plains.
down through central USA, and in two areas
only North American languages which show any
further east. The best-known members are Cher-
clear links with Asian languages are those belong-
okee, Dakota (Sioux), Crow, Mohawk, and Paw-
ing to the Eskimo-Aleut family.
nee, but only the first two have over 10,000
Eskimo-Aleut is the name given to a small group
speakers.
of languages spoken in the far north, in Alaska,
Canada, and Greenland, and stretching along the
Aleutian Islands into Siberia. Eskimo is the main Isolates
language, spoken in many dialects by around There are over 30 languages whose relationship to the
90,000. Its two main branches - Yupik in Alaska main language groups in North America has not so far
and Siberia, Inupiaq (Inuit, or Inuktitut) elsewhere been determined. Over 20 of these are the Salish lan-
— are sometimes classified as separate languages. guages, spoken along the Canadian/USA Pacific
coastline, and some way inland. They include Bella
Greenlandic Eskimo has official status in Green- Coola, Okanogan, Shuswap, and Squamish. These
land, alongside Danish. A standard written form days, the numbers of speakers are very small - mostly
dates from the mid- 19th century. There are also fewer than 1 ,000 and in several cases fewer than 1 0.
a few hundred speakers remaining of Aleut. Pentlach, spoken on Vancouver Island, was already
extinct in the 1 970s. The six languages of the Waka-
Further south, the Na-Dene group consists of
shan family, spoken on the British Columbia coast
over 30 languages, spoken in two main areas: (notably, Nootka and Kwakiutl) constitute another iso-
Alaska and north-west Canada, and south-west- lated group.
central USA. Most of the languages belong to the
Athabaskan family, whose best-known member is
Navaho, with around 120,000 speakers — one of
the few Amerindian languages which has actually FROM NORTH TO SOUTH
increased in size in recent years. The various dia- The main linguistic bridge between North and
lects of Apache are closely related to Navaho. South America is formed by the (Macro-) Penutian
The Algonquian family is geographically the group, which in its broadest interpretation consists
most widespread, with over 30 languages covering of over 60 languages (many of these grouped into
a broad area across central and eastern Canada, smaller families), spoken from south-west Canada
and down through central and southern USA. down through the western states of the USA,
Many well-known tribes are represented — the Ara- throughout Mexico and Central America, and into
paho, Blackfoot, Cheyenne, Cree, Fox, Micmac, south-west South America. In a narrower interpre-
Mohican, Ojibwa, Potawatomi, and Shawnee - tation, only the 20 or so North American lan-
though only Cree and Ojibwa have substantial guages, none of which has many speakers, are
numbers of speakers (around 50,000). Several subsumed under this heading.
other languages spoken mainly in the south-east The languages with most speakers belong to the
USA have now been grouped along with Algon- Mayan family, spoken in Mexico and Central
quian into a Macro-Algonquian family — most America - notably Maya (or Yucatan), Mam, Kek-
notably the Muskogean group, which includes chi, Cakchiquel, and Quiche, all of which have over
Choctaw and Muskogee. a quarter of a million speakers. In South America,

320 • IX THE LANGUAGES OF THE WORLD

 The Pueblo Indians of Ari-
themain candidate for membership is Araucanian
zona and New Mexico are
(orMapuche), spoken mainly in Chile by around linguistically very diverse -
200,000. Chipayan and Uru (spoken by a few about 25,000 people speak
hundred people in Bolivia) have also been proposed languages belonging to no
fewer than four families. In
as belonging to Penutian.
the east, they mainly speak
The Hokan group of around 30 languages is Tewa (a member of the Ta-
spoken by small numbers in parts of western and noan family within the Aztec-
south-west USA, and eastern Mexico. Tlapanec is Tanoan group) and Keresan
the only language with over 20,000 speakers. Simi- (a language isolate); in the
west, they speak Keresan,
larly, most of the 30 or so languages which belong
Zuni (a Penutian language)
to the Aztec-Tanoan group have few speakers and Hopi (a Utc—Aztecan
today. The group includes the languages of such language, within the Aztec-
well-known tribes as the Comanche, Paiute, Shos- Tanoan group). The picture
hone - and also the Hopi (p. 15). Three Mexican shows part of a Hopi cere-
mony from 1 91 6: a group of
languages are still widely spoken: Aztec, or Nahua
women, dressed as men, ar-
(around a million speakers), Tarahumar (over rive bearing fruit.

50,000), and Pima-Papago (nearly 20,000).

51 OTHER FAMILIES 321

CENTRAL AMERICA
The indigenous languages of Central America are
generally known as Meso-American (or Middle
American) Indian languages. In an area extending
from Mexico to Nicaragua, about 70 languages
are spoken by around 6 million people. Several of
the languages belong to one of the North American
families (Penutian, Hokan, Aztec-Tanoan); some
belong to South American families (grouped under
Macro-Chibchan). The only group which is res-
tricted to this region is Oto-Manguean. Almost all
Oto-Manguean languages are spoken within a
small area centred on the state of Oaxaca, Mexico.
The main languages are Otomi, Mixtec, and Zapo-
tec, each spoken by around a quarter of a million
people.
SOUTH AMERICAN INDIAN
LANGUAGES
Indigenous Indian languages are used throughout
the whole of the continent of South America,
including the southern part of Central America and
the Antilles group of islands. They are spoken by
over 11 million people. In former times, as many
as 2,000 languages may have been spoken in the
area, but fewer than 600 of these have been
attested. Despite the considerable efforts of ethno-
graphers and missionaries, especially in the present
century, few languages have been completely de-
scribed. Many tribes consist of small numbers liv-
ing in extremely remote jungle areas. Even in the
more accessible cases, there is considerable uncer-
tainty over the identity of the languages, and what
kind of language/dialect boundaries operate (p.
284). Many are under threat of extinction as wes-
tern civilization (in the linguistic shape of Spanish
and Portuguese) opens up the area. It seems likely
that over 1,000 tribes have become extinct before
their languages could be recorded.
In spite of this decline, South America remains
one of the most linguistically diversified areas of
the world. Some accounts suggest that there are
more than 100 language families on the continent.
However, because of the difficulties in obtaining
accurate information, classifications of the lan-
guages have tended to be very general, and there
are many differences among the sub-groupings
which have been proposed. At the most general
level, three major groups have been suggested.
The Macro-Chibchan group is one of the most
widespread, being found in Central America, Col-
umbia, Venezuela, and south into Bolivia and
Brazil. There are around 50 languages in the group,
but only five (Guaymi, Cuna, Waica, Epera, Paez)
have as many as 20,000 speakers, and several are
on the verge of extinction.
The Ge-Pano-Carib group of nearly 200 lan-
guages is spoken east of the Andes along most of
the length of the continent and along the Brazilian
Amazon basin. It has a very small number of
speakers (perhaps a million) for such a vast area.
The Carib family, within this group, is one of the
322 •
IX THE LANGUAGES OE THE WORLD
largest in South America, containing over 80 lan-
guages spoken by tiny numbers throughout the
whole northern region. Only Carib itself has as
many as 5,000 speakers. Macro-Panoan, also
within this group, is a family of about 70 languages
spoken from Peru and Bolivia eastward to Brazil,
and southward to Paraguay and Argentina.
Mataco, spoken mainly in northern Argentina and
Paraguay, is the only language with more than
10,000 speakers.
The Andean-Equatorial group consists of about
250 languages, and contains many sub-divisions.
Within the Equatorial division, for example, there
is the Arawakan group, which once extended into
North America, and is still widespread, being
spoken from Central America to southern Brazil.
Goajiro (over 40,000) is its main member. Within
the Andean division, the Quechumaran group is
preeminent in the Andes highlands between Col-
ombia and Argentina. Aymara was once a major
language throughout the central Andes, but is now
restricted to around 600,000 speakers in Bolivia
and Peru. Quechua, the official language of the
Incas, is now spoken by over 6 million from Colom-
bia to Chile. It is widely used as a lingua franca,
and its literary history dates from the 17th century.
In the south, in Paraguay, the Indian language of
Guarani (a member of the Tupi family) is spoken
by perhaps three million people (mainly non-
Indians), and is the majority language of that
country - the only Indian language to achieve such
a status. By contrast, over a dozen Tupian lan-
guages have become extinct in the first half of this
century.
Tne South American Indians migrated from the
north, but hardly any of the languages of the area
are plausibly related to the language families of
North and Middle America. The only links which
have attracted support are under the heading of
Penutian, where some scholars have placed Arau-
canian, Chipayan, and Uru. Others, however, see
these languages as part of the Andean-Equatorial
group.
A NEW CLASSIFICATION
In a fresh classification presented in1985 by the
American Joseph Greenberg (1915- ),
linguist
all the languages of the New World are brought
together, and grouped into three main families: Na
Dene, Eskimo-Aleut, and Amerind. Eskimo-Aleut
is seen as part of a 'Euroasiatic' family, whose other
members include Indo-European, Altaic, Japanese,
Korean, and several other languages. Amerind is
an extremely large family, comprising 1 1 sub-fami-
lies and, at a lower level of classification, nearly
200 groups of languages (a reanalysis of the lan-
guages included in the families listed on pp. 320-3).
It covers the whole of North, Central, and South
America, and incorporates several languages pre-
viously thought to be isolates ($53).
Where South American Indian
languages are spoken

Atlantic Ocean

Atlantic Ocean

Andean-Equatorial [JatjSj

Ge-Pano-Carib [jjl

Macro-Chibchan | |

Penutian |
Unknown or unclassified Pi

52 OTHER FAMILIES- 323

Australian aboriginal
languages
No clear relationship has yet been found between
aboriginal languages and the rest of the world's
languages. With no written records, historical dis-
cussion is largely speculation. In the 18th century,
there may have been over 500 aboriginal languages
in spoken by over 300,000 people.
Australia,
Today, about 250 languages are documented from
many parts of Australia (but excluding Tasmania:
see p. 319). Only five languages have more than
1,000 speakers; most have very few; and at least
half are nearly extinct.
A frequently cited estimate is that fewer than
50,000 people speak the languages today, with
different levels of ability. But, for many reasons,
population estimates are difficult. Aboriginal
people often live in isolated areas; most are bi-
lingual to differing degrees; and it is not always
easy to obtain accurate information from the
speakers themselves, many of whom overestimate
or underestimate their ability to use the language,
for social or political reasons. By the same token,
scepticism is sometimes expressed about the results
of national surveys, which might be used as evi-
dence for or against the provision of educational
or social facilities for aboriginal groups.
Aboriginal languages have been grouped into 28
families, all of which are thought to be related.
All but one of these are found in the northern parts
324 • IX THE LANGUAGES OF THE WORLD
of Western Australia, Northern Territory, and
Queensland, in an area comprising no more than
one-eighth of the continent. Arnhem Land in par-
ticular shows a high concentration of these
languages. By contrast, a single family, Pama-
Nyungan, covers the remainder of the continent;
about 175 languages once belonged to this family,
but fewer than 50 are spoken today, most of these
surviving in the north-west.
The languages with the largest number of
speakers are Tiwi, Walmatjari, Warlpiri, Aranda,
Mabuyag, and Western Desert - all but the first
belonging to the Pama-Nyungan family. Several of
the languages have come to be used as lingua fran-
cas. Gunwinygu is used in this way in much of
north-west Arnhem Land, and Pitjantjatjara in
much of northern Western Australia, partly as a
church language. Warlpiri is one of the most vigor-
ous of these languages, spoken in many central and
southern parts of Northern Territory. Several pid-
gins and Creoles ($55), related to English, have also
developed in northern areas.
The future of aboriginal languages is uncertain,
but several of the languages now have a written
form, and bilingual school programmes have been
devised. Organizations such as the Australian Insti-
tute of Aboriginal Studies promote the study of
these languages, their history, and their contempor-
ary social and political status. It therefore seems
likely that a small number of these languages will
remain vigorous for some time to come.
Below: A set of engraved tally
marks found in Koonalda Cave
on the Nullarbor Plain,
Western Australia.

Above: An attempt to stop

the destruction of a sacred
site in Kimberley, Western
Australia, in 1978.
It is a

symbolic reminder of the

daily threat facing the abori-
ginal heritage, and thus its

linguistic identity.

The Dreaming places in the landscape.On the other hand, several

No-one knows how long aborigines have been in
Australia, but estimates of at least 40,000 years
have been made. According to traditional aborigi-
nal belief, life began during a mythological period
known as the 'Dreaming' - a period which had
a beginning but no end. During that beginning,
the aboriginal world was formed by spirit beings,
both animal and human in form. They moved
across the land, which was believed to have been
soft and malleable, creating the shape of the terri-
tory as they went, and thus leaving signs of their
spiritual presence. Many physical features of the
landscape were thought to be formed by these
beings, and the sites are consequently held to be
sacred.
Several of these sites are rocks, caves, and trees,
often marked by carvings or paintings. Motifs
include animal tracks, artefacts, humans, kanga-
roos, and other mammals. There are also many
kinds of geometrical shapes, often in combinations.
These are not an attempt at written language, but
seem to be symbols representing the significant
have been found which seem
sets of parallel lines
to have been made by a finger dipped in an ochre
paste. They are sometimes referred to as 'tally
marks', because of the possibility that they may
have been a register of people, periods of time, or
events.
Aborigines thus see themselves as having a very
close spiritual (as well as economic) relationship
to the land, and it. But today,
to particular places in
many of the sacred danger — partly from
sites are in
natural erosion, but more particularly from indus-
trial development and vandalism. Legislation is
now available to protect them, but conflicts of
interest can still occur. As one Nyungar aborigine,
Ken Colbung, put it: 'I am sure that people who
are not of aboriginal descent are unaware of the
strong emotional feeling we have for a particular
place. We see it as part of our spiritual background;
and that is what is being consistently undermined.
The problem is not simply one of mining and the
royalties which, at least in part, should come to
us. It is one of land.''

52 OTHER FAMILIES •
325
53 Language isolates ^
A language isolate is a language which has no
known structural or historical relationship to any
other language (p. 293). Most of the world's lan-
guages can be grouped into families using compara-
tive linguistic techniques. But occasionally one
encounters a language where resemblances to other
languages are few or non-existent. Sometimes, the
few points of contact are sufficient to motivate a
tentative classification - thus some scholars place
the Scots language Pictish within Celtic, the African
languages Fur and Songhai within the Nilo-
Saharan group, the Mexican language Huave mates of the number of
speakers vary, from 500,000
within Penutian, and Tasmanian and Andamanese
to over 700,000. Most Bas-
within Indo-Pacific. However, others see the differ- ques live in a 4,000-square-
ences as more important than the points of similar- mile area of northern Spain
ity, and list these languages as isolates. and south-west France, but
Many languages have been classified as isolates many went into exile in the
USA after the Spanish Civil
simply because little is known about them, linguis-
War. Attempts have been
tically or historically. For example, preliminary made to show a relationship when it was realized that this/

research into South American Indian languages has with Caucasian languages language was quite different \
brought to light several possible isolates, but (p.305), with North African from others written in the <

languages, and also with same script. Sumerian was

further study may well indicate relationships with
Iberian, the now extinct lan- spoken in southern Mesopo-
other languages — provided the cultures survive many inscriptions
guage of tamia (part of modern Iraq)
long enough for these studies to be carried out (p. found along the Mediterra- until the 2nd millennium bc.
322). Examples are Callahuaya in Bolivia, and Ari- nean coasts; but none has It was then supplanted by a

capu, Baenna, Hixkaryana, Juma, and Natu, in been convincing. The written
history of the language can
Brazil. Then, from a historical point of view, there
be traced to Roman times,
are several languages of ancient Asia Minor which through various inscriptions.
are known only from passing references to them There is now intensive local
in classical Greek literature, or occasional place concern to develop the lan-
names and inscriptions — examples include Bithy- guage, and introduce it into
primary education; but for
nian, Cappadocian, Carian, Cataonian, Cilician,
many abroad, the language
Gergito-Salymean, Hattic, Isaurian, Lyconian, and culture are more associ-
Myriandynian, Ordek-Burnu, Paphlagonian, Pisi- ated with the violence of the
has been made using philo-
logicalmethods, because
Etruscan seems to bear no
relationship to any other lan-
guage. There is no extant
literature or historical record
of the civilization. Why this
should be so remains one of
the great unanswered ques-
tions of classical studies.
Semitic language (Akkadian)
- though the written form of
Sumerian continued to be
used for nearly 2,000 years.
There are many records of
the language - business, le-
gal, religious, administrative,
and private texts and inscrip-
tions. Literary work is pre-
served from the later period,
in a range of forms including

dian, Pontic, and Sidetic. It is unlikely that their political separatist move- 4) Linear A This is the name hymns, rituals, proverbs, and
ment, Euzkadi ta Azkatasu- given to a Cretan script used myths. Several dialect forms
affiliations will ever be known.
na (ETA). (Euskara is the in the middle of the 2nd cen- are known. Attempts have
The diagram some information about
gives Basque word for their lan- tury bc. It has still not been been made to relate the lan-
several of the languageswhich have been proposed guage (p. 34).) deciphered, and the lan- guage to many other fami-
as isolates. It includes languages which remain guage it represents is there- lies, including Altaic and Dra-
3) Etruscan The area of fore not known, though some vidian, but none has been
undeciphered, languages where there is insufficient
Tuscany in modern Italy is believe it to be Minoan (or successful.
material available to establish a family relation- the site of the ancient Eteocretan). The name
ship, and languages where, despite a great deal of country of Etruria, where the refers to the way the script is
A Sumerian account listing
data, the relationship is undetermined. Two of the Etruscan civilization was at written in lines, probably
livestock in the charge of a
its height in the 6th century from left to right - a contrast
best-known isolated languages, Korean and Japa- herdsman. The tablet comes
bc. The language is known with previous hieroglyphic
nese, are discussed on p. 306. from the ancient city of Ur,
from about 10,000 inscrip- writing. The label A' dis-
tions, mainly short epitaphs and is dated in the ninth year
tinguishes the script from
and dedications, of King Shu-Sin, 2029 bc.
written in an Linear B, which was used to
1) Iberian This language 28-letter alphabet shows the alphabet probably derived write Greek later in the same
was spoken in parts of south- influence of both the Greek from the Greek, and from millennium (p. 301).
ern and eastern Spain, espe- and the Phoenician alpha- which in due course came
cially around the Ebro River, bets, but for the most part its the Latin alphabet. The lan- Sumerlan This is the ol-
5)
in pre-Roman times. It may history is unclear. guage may still have been dest known language to be
formerly have been used spoken as late as the 4th preserved in written form. In-
throughout a much wider 2) Basque is the only lan- century ad. scriptions date from around
area of western Europe. It is guage remaining of those Only a few words of the 3100bc, written in cuneiform
known mainly through in- which must have been spok- language have been deci- script (p. 198). The existence
scriptions on stones and ar- en in south-west Europe be- phered: no contemporary of Sumerian was not recog-
tefacts of the period, few of fore the advent of the Indo- translations seem to have nized until cuneiform was de-
which can be interpreted. Its European invasions. Esti- survived, and little progress ciphered in the 19th century,

326 •
IX THE LANGUAGES OF THE WORLD
 1-
r 60,000
- epecha, and Mechoacan.
Mohenjo-Daro
^ 7)
> Pakistan. Excavations at the
<?
6) Elamite This extinct lan-
guage was spoken in the an-
cient country of Elam - an
area now corresponding to
Khuzistan in south-west Iran.
The oldest writings are in the
form of pictographic inscrip-
tions from the 3rd millennium
bc. Later writing is in cunei-
form script. The language
was still in use in the 1 st mil-
lennium ad. A relationship
with Dravidium has been
proposed.
An Elamite inscription
stamped on a baked clay
brick, dating from the 12th
century bc. It describes the
rebuilding of the temple of
the 'Great King' by King
Shilhak-lnshushinak i.
Mohenjo-Daro The
name (which means the
mound of the dead') refers to
a group of mounds on the
bank of the Indus River in
site since the 1 920s have
brought to light the remains
of a major city, dating from
the 3rd millennium bc. The
many finds contain evidence
of a script, which so far is un-
deciphered.
8) Burushaski This lan-
guage is spoken in north-
west Kashmir, India, and in a
small part of adjoining Pakis-
tan,by over 20,000 people
belonging to the Burusho
has no written form.
tribe. It
Fewer than 1 ,000
9) Nahali
people speak this language,
in a small area in south-west
Madhya Pradesh, in India.
Some scholars have related
the language to the Munda
group of Austro-Asiatic
(p. 309), but most view it
independent.
10) Gilyak This language
spoken by 3,000 or more
people inthe north-east
USSR, on the island of Sak-
halin and on parts of the
mainland opposite. Gilyak
(or Nivkhi, the name used by
the people themselves) is of-
ten listed along with the
neighbouring Palaeosiberian
languages (p. 306), but pro-
posed links with these and
other languages of the area
(especially Korean and the
Altaic languages (p. 307))
have not been accepted.
11) Ainu About 16,000 Ainu
tribespeople live in
Hokkaido, Japan, and in the
Sakhalin and Kuril Islands,
but in recent years, the
culture as well as the
language has lost ground to
the Japanese, and there
must now be few native
speakers left. The traditional
Ainu are unlike the Japanese
in physical appearance, and
it is thought that they may be
descendants of Caucasoid
peoples who were once
as
is
An Ainu tribesman
spread throughout north
Asia.
12)Kutenai There are
many spellings and names
for this language - Koote-
nay, Cootenais, Skalzi, Are-
a-plat,and Flatbow are some
of those recorded. It
en by a North American
dian tribe, mainly in south-
east British Columbia and Al-
berta, but also in northern
parts of Idaho, Washington,
and Montana. Their numbers
are decreasing-fewer than
500 in the late 1 970s. Some
scholars have postulated re-
lationships with other Amer-
indian languages (p. 320),
but none of the proposals is
generally accepted.
1 3) Also known as
Keres
Keresan, or Queres, this lan-
guage is spoken in different
dialects by several groups of
Pueblo Indians in the New
Mexico area. There were
around 7,000 speakers in
the late 1960s. Originally
thought to be a member of
the Hokan-Siouan family,
now considered an isolate.
14)Tarasca This language
was spoken by around
Mexico.
in the late 1960s,
parts of south-west-central
It goes under sev-
in
eral names and spellings,
including Tarascan, Por-
A relationship to other lan-
guages of the area has been
proposed (under the general
heading of Penutian, p. 320),
but is unclear. In recent
years, the number of
speakers has been decreas-
ing, with many Tarascan In-
dians becoming assimilated
within the mixed European
culture dominant in Mexico.
1 5) Het This South Ameri-
can Indian language, also
known as Chechehet, be-
came extinct at the end of
the 18th century. It was
spoken in Argentina, and is
known from only a few words
and place names.
l6)Karankawa This lan-
guage, also known as Clam-
coets, was spoken by Indian
tribes living along the Texan
coastline in the 1 8th century.
They seem have died out
to
by the mid- 1 9th century, with
the influx of white settlers
into the area.
17)Calusa An extinct tribe
is spok-
In- ofAmerican Indians who
lived in the south-west part of
Florida until the end of the
18th century, and perhaps
later. Many families emi-
grated to Cuba, to escape
from the invasions of other
tribes, and, ultimately, the
British.
18)Beothuk This lan-
guage, spoken by an Indian
tribe on the island of New-
foundland, is now extinct. Its
last known speaker died in
1829. Some scholars have
argued that it should be clas-
sified as an Algonquian lan-
guage, but the opinion is
controversial. The Beothuk
rubbed red ochre on their
bodies - a practice which
may well be the reason for
it is the European name Red
Indians'.
53 LANGUAGE ISOLATES •
327
54 Language change

Languages are always in a state of flux. Change cluded that these shifts operated in such a regular Grimm's law
affects theway people speak as inevitably as it does manner that they could be seen as sound 'laws'. Indo-European Germanic
any other area of human life. Language purists do The first and most famous of these laws was Voiced
not welcome it (p. 5), but they can do very little worked out by Jakob Grimm in 1822, in his Ger- Voiced plosives
aspirates* (unaspirated)
about it. Language would stand still only if society manic grammar, and it has since become known
did. A world of unchanging linguistic excellence, as 'Grimm's law'. bh i b
dh d
Grimm noticed that where a word in Sanskrit,
1

based on the brilliance of earlier literary forms,

gh 1
g
exists only in fantasy. Latin, or Greek began with [p], the Germanic lan-
guages usually used Voiced
During the greater part of the 19th century, [f], as in:
plosives Voiceless
linguistic scholarshipused the comparative method (unaspirated)
Latin pater English father plosives
(p. 292) to establish the facts of language change. b 1
P
piscis fish
What features of language have changed in the d . t

past? When did they change? How did they In a similar way, words beginning with [t] usually 9 k

change? During the present century, especially as had [9], as in: Voiceless Voiceless
a consequence of recent trends in sociolinguistics plosives aspirates*
Latin tres English three
(p. 333), the emphasis has shifted towards a search P
for explanations. Why do languages change?
t
Altogether, nine correspondences were
sets of
k
noted, which fell into a clear phonetic pattern, sug-
The 'circular' relationship be-
gesting that Germanic languages had diverged from
What changes? Indo-European in a regular way. It was not possible
tween the correspondences
is its major feature:
and use are subject
All aspects of language structure to say exactly when the changes took place, but
to changes, but the most noticeable and frequent they were complete by the time the earliest Ger-
changes affect pronunciation and vocabulary, and manic texts came to be written (4th century ad
it is these which have attracted most study. The Gothic).
science of comparative philology was at first en- However, Grimm's law did not explain all the
tirely devoted to the study of sound change, and differences between Germanic and the other lan-
contemporary sociolinguistic studies have a similar guages. There seemed to be several exceptions. For
emphasis. example, the word for 'daughter' was dauhtar in
Gothic and duhita in Sanskrit; but according to
SOUND CHANGE Grimm's law, the Sanskrit form should have been
From the earliest days of comparative philology, * dhuhita. Or again, the word for 'father' was
fadar
it was noted that the sounds of related languages in Gothic and pita in Sanskrit. The change from
*
Grimm's term includes both
aspirated plosives and frica-
corresponded to each other in apparently systema- [p] to [f was regular, but why did the [t] become
]
tives (p. 157).
tic ways — what were referred to as 'sound shifts'. [d], when according to Grimm's law it should have
Verner's law
Later, on the basis of several studies, it was con- been [9]?
Germanic Germanic Later
Voiceless Voiced forms
fricatives fricatives
Types of sound change (arising from
The processes which affect (potato), earlier Tartufleln, Loss. A sound disappears Example: Latin domina be- Grimm's law)
sound change are many and where the [k] dissimilated from the language. coming Italian donna (lady). f > v > b
various. This list illustrates from [t] in the 17th century. Example: Old English velar 9 >5 >d
some of the types which fricative [x], which was a var- Apocope. The loss of final x » y >g
Merger or coalescence.
regularly occur. iant of /h/, as in eahta (eight); sounds.
Two sounds become one.
Assimilation. Probably the
this sound had disappeared Example: Modern English
Example: Old English /ev
by early Modern English. help, from Old English helpe. (i) the consonant is not initial
most important type of and/ae:/, which became
Haplology. The loss of a (ii) the preceding and follow-
change, in which one sound Modern English /i:/, as in
sound, because of its similar- Prothesls. The introduction ing sounds are voiced
is influenced by the pronun- sweef (OE swete) and clean
ity to a neighbouring sound. of an extra initial sound. the preceding vowel
(iii) is
ciation of a neighbouring (OE claene).
Example: Modern English Example: Latin schola unstressed
sound (p. 164).
Split. One sound becomes England, from Old English (school) becoming Spanish
Example: Latin noctem
two. Englalond (land of the escuela, Old French escole,
(night), which became noffe
Example: Old English /s/,
the
in Italian, M
being assi-
which was realized as [z]
Angles). etc.

milated to the following /t/. Metathesis. Two sounds

only between voiced sounds, Epenthesis. The introduc-
change places.
Dissimilation. A sound as in thousand (OE thu- tion ofan extra medial
Example: English third, from
moves away from the pro- send); in Modern English, /z/ sound.
Anglo-Saxon dridda.
nunciation of a neighbouring has split off from /si, becom- Example: Old Icelandic ofn,
sound. ing a phoneme in its own Syncope. The loss of medial alongside Old English ofen,
Example: German Kartoffel right. sounds. Modern English oven.

328 • IX THE LANGUAGES OF THE WORLD

 There were several discrepancies of this kind. the form of a satirical comment or passage about
Many of them came to be resolved by examining a certain style of speech. For example, in The Merry
the phonetic contexts in which the sounds occur. Wives of Windsor, Falstaff reacts to the speech of
The German philologist, Hermann Grassmann the parson, Sir Hugh Evans, who
has tried to pro-
(1809-77), worked out an explanation for the kind
*
nounce the words cheese and "Seese" and
butter: '
Why does Jaques
of case illustrated by dhuhita. He argued that "putter"? Have I lived to stand at the taunt of one
laugh for an hour?
these forms had a sequence of two aspirates
all that makes fritters of English?'. But more direct
In As You Like It, the courtier
in Indo-European, and that this was enough to evidence comes from the detailed accounts made
Jaques reports meeting the
block the application of Grimm's law. His formula- by orthoepists, specialists in the study of pro-
court jester, Touchstone, in
tion became known as 'Grassmann's law'. The nunciation. In 17th-century England, there were the forest.
Danish linguist, Karl Verner (1846-96), discovered many such writers, including Bishop John Wilkins And then he drew a dial from
a reason for cases such as pita. He found that (1614-72) and the mathematician John Wallis his poke,
Grimm's law worked well whenever the stress fell (1616-1703). In the Orthographie of John Hart And, looking on it with lack-
on the root syllable of the Sanskrit word; but when (d. 1574), for instance, which was published in lustre eye,
Says very wisely is ten
it fell on another syllable, the consonants behaved 1569, we find detailed descriptions of the organs 'It

o'clock;
differently. A
further change took place: [p, t, k] of speech and of sounds of 16th-century English. Thus we may see' quoth he
did not stay as [f, 0, x], but became [b, d, g]. The 'how the world wags;
Poetic evidence
precise formulation of this regularity became Tis but an hour ago since it
The way in which early poets made words rhyme, was nine;
known as 'Verner's law'.
or gave their lines a particular metrical pattern, And after one hour more 'twill
As a result of such explanations, philologists in
be eleven;
provides a great deal of evidence about where the
the later part of the 19th century began to feel And so, from hour to hour,
stress fell in a word, and the way vowels were pro-
that all exceptions to sound laws could be we ripe and ripe,
nounced. Puns, too, draw our attention to points And
explained, as long as proper attention was paid then, from hour to hour,
of similarity between sounds. For example, we can we rot and rot;
to the phonetic environment and to such matters
infer from the way Chaucer rhymes was with the And thereby hangs a tale.'
as stress. The view was highly controversial, but When did hear
French par cas that its pronunciation must have I

it exercised a great deal of influence on the subse- The motley fool thus moral
been something like 'wass'. Such comparisons do
quent development of the comparative method, on the time,
not tell us exactly how the words were pronounced, My lungs began to crow like
and of linguistic theory.
nor whether these were normal pronunciations of chanticleer
the time; but they do provide the historical linguist That fools should be so deep
HOW DO WE KNOW WHAT OLD with valuable clues.
contemplative;

LANGUAGES SOUNDED LIKE? And did laugh sans

I

intermission
Alphabetic evidence
An hour by this dial.
Contemporary accounts When European languages first came to be written
What was it that made
Writers sometimes give an account of contempor- down, those who devised the alphabets borrowed
Jaques laugh a whole hour?
ary pronunciation features - often indirectly, in symbols from alphabets already in use elsewhere The bawdy pun involved can
— usually Latin.They would then modify or add be appreciated only when we
to these symbols whenever they came across sounds realize that hour and whore

They have been at a great feast of which the older alphabet could not cope with. It
were pronounced alike, at
that time.
languages and stol'n the scraps' is thus possible to use these changes as evidence
This is Moth, in Love's Labour's Lost, talking about a way the two languages were
of differences in the
conversation between Holofernes, the schoolmaster, spoken. For example, when the Latin alphabet was The
and Sir Nathaniel, the curate. Between the lines of used for Anglo-Saxon, the symbol x was added
Holofernes' description of Don Armado's
Neogrammarians
— presumably because the missionaries felt that the name given to the
pronunciation, valuable hints about contemporary This is

speech styles and attitudes can be gleaned. Latin a and e were inadequate to represent the group of German philologists
sound they heard. who claimed in the 1 870s
He draweth out the thread of his verbosity finer than
that 'sound laws have no ex-
the staple of his argument. abhor such fanatical
I
Comparative reconstruction ceptions'. They were called
phantasimes, such insociable and point-devise
This procedure works backwards from languages Junggrammatiker, somewhat
companions; such rackers of orthography, as to speak
scornfully, by their older col-
'dout' fine, when he should say 'doubt'; det' when he whose pronunciations are known, using the com-
leagues, and this name,
should pronounce debt' - d, e, b, t, not d, e, t. He parative method to reconstruct earlier forms (p. translated as neogrammar-
clepeth a calf 'cauf, half 'hauf; neighbour vocatur
292). Most of our information about the oldest ians',has continued to be
'nebour'; neigh abbreviated 'ne'. This is abhominable
states of languages derives from this method. But used for them. The group in-
- which he would call abbominable'. It insinuateth me
cluded such men as August
of insanie: ne intellegis, domine? to make frantic, the further back in history we travel, the less certain
Leskien(1840-1916), who
lunatic. our phonetic deductions are likely to be. It may coined the above slogan,
be clear that an early language contrasted two and Karl Brugmann (1849-
sounds, but quite unclear as to how this contrast 1919). Brugmann's five-
was actually realized in speech. volume comparative gram-
mar of Indo-Germanic lan-
Tape recordings guages (published in an en-
larged second edition
Historical linguists of the future will be able to
between 1 897 and 1916) re-
rely on the records and tapes made in the present
mains unsurpassed in its
century, as the clearest evidence of all for linguistic comprehensive coverage of
change. the field.

54 LANGUAGE CHANGE • 329

GRAMMATICAL CHANGE the best source of information is a dictionary which Some surprising
The most noticeable way in which grammatical sys- has been written on historical principles, such as etymologies
tems change is known as analogy. In this process, the Oxford English Dictionary. Many languages The words in the left-hand
irregular grammatical patterns are changed in ac- also have specialized etymological dictionaries. column once had the mean-
cordance with the regular patterns which already ing given on the right.

exist in the language. New words and old treacle «- wild animal
A well-studied case is the verb system in the his- The two most obvious factors in semantic change villain <- farm labourer
tory of English. Several of the irregular verbs of taxation «- fault finding
are the arrival of new words and the loss of old
Anglo-Saxon have under the influence of the bonnet <- a man's hat
fallen ones. In most languages, the vast majority of new furniture *- equipment
regular verbs in the 1,000 years. For example, hel- words are in fact borrowings from other languages <- ingenious
pretty
pan (help) had healp as a past tense and bolpen — though this term is not a very appropriate one, cheater «- rent collector
as a past participle; but by the 14th century, the as new words are not given back at a later stage! naughty «- worth nothing
verb had become regular, using the normal -ed end- vulgar <- ordinary
Borrowing proceeds in all directions. Weekend and
sly <- wise
ing — helped. During the early Middle English per- parking have been borrowed by French from Eng- publican «- public servant
iod, over 40 other verbs (including walk, climb, lish; chic and savoirfaire have been borrowed by orchard *- garden (without
burn, and step) were influenced in the same way. English from French. Some languages have bor- fruit trees)
Social factors, such as the development of the stan- rowed so extensively that native words are in a
dard language, and the growth of printing, slowed minority.
the change down, so that present-day English still A special type of borrowing is known as a loan Some English
has many irregular verbs. But the force of analogy translation or caique. In this process, a word is sources
can still be heard, when people use non-standard not borrowed whole, but its parts are translated ballot Italian
forms (such as knowed), or when children, learning separately and a new word formed — as when Ger- banshee Scots Gaelic
the language, experiment with such forms as chow mein Chinese
man produced the equivalent of English telephone
garage French
goned. The tension between regular and irregular in Fernsprecher (literally, fern 'distant' -I- sprecher gong Javanese
forms is also illustrated by problems of modern 'speaker'). goulash Hungarian
usage, such as the choice of strove vs strived, chid When word
or sense ceases to be used, it is
a junta Spanish
vs chided, or sown vs sowed. said to be obsolescent or obsolete. This often hap- kiosk Turkish
Analogy does not operate only in word forms. llama Quechua
pens because an object or concept is no longer of
marmalade Portuguese
Syntactic constructions can also be affected. In value to a community (other than to the historian robot Czech
Anglo-Saxon, for example, the Subject—Verb— Ob- or literary scholar); but a word or sense may schmaltz Yiddish
ject pattern applied only to main clauses; in subor- become obsolescent if it develops unpleasant asso- slim Dutch
dinate clauses, the object preceded the verb. In sofa Arabic
ciations, or is replaced by another word which is
Modern English, both clause types show the same tomato Nahuatl
felt to be more modern. Wight (person), leman
tycoon Japanese
order (§14). sweetheart), and hie (hasten), are examples from veranda Hindi
Analog) does not create new grammatical pat-
-

Elizabethan English which are now no longer used; window Old Icelandic
terns: it simply extends the range of a pattern which humour (= 'temperament') and conceit (= 'idea') yen Chinese
already exists in the language. Other processes of (= 'desire']
illustrate obsolete senses from the same period.
change have a more radical role, creating new pat-
terns and eliminating old ones. For example, in Some types of semantic change
Latin, the relationship between subject and object
Extension. A word widens its meaning.
Boyfriends and
was shown by inflectional endings, and the order Example: In Latin, virtue was a male quality (cf. vir girlfriends
of the elements was not important; but in the 'man'); today, it applies to both sexes. Whether a language will bor-
modern Romance languages, these relations are Narrowing. A word becomes more specialized in row a word whole, or trans-
expressed by word order. In early Indo-European, meaning. late its parts, is never predic-
Example: In Old English, mete referred to food in table. As the words girlfriend
there were three grammatical genders for nouns
general (a sense which is retained in sweetmeat), and boyfriend spread from
— masculine, feminine, and neuter; these have been the west to the east, they
today, it refers to only one kind of food.
retained in modern German and Greek, but are were handled differently. The
reduced to two in modern Dutch common vs Shift. A word moves from one set of circumstances Chinese loan-translated the
to another.
neuter) and French (masculine words as nan pengyu (male
vs feminine), and Example: Navigator once applied only to ships, but it
friend) and nu pengyu (fe-
have been completely lost in modern English. now applies to
planes, and even to cars. male friend). The Japanese,
SEMANTIC CHANGE Figurative use. A shift in meaning based on an however, borrowed the
words as wholes, adapting
analogy or likeness between things.
This is perhaps the most obvious area of linguistic them to their sound system:
Example: Crane, a bird with a long neck, has led to
change, and the one which many people find the the use of crane as a piece of equipment for lifting the result was boifurendo
most fascinating. Semantic change is profoundly weights. and garufurendo.
connected with the life, literature, and culture of Amelioration. A word loses an original sense of
a community. Innumerable examples can be found disapproval.
in the pages of any old book, or simply by careful Example: Mischievous has lost its strong sense of
watching and listening to everyday usage. But plot- disastrous', and now means the milder playfully
annoying'.
ting the history of the changes in the form, meaning
and use of words and morphemes is difficult work, Pejoration. A word develops a sense of disapproval.
Example: Notorious once meant widely known', and
because the evidence is often lacking.
now means widely and unfavourably known'.
To find out about lexical history, or etymology,

330 • IX THE LANGUAGES OF THE WORLD

Lexicostatistical Working backwards, on this basis, they con- The basic word-list
structed a table of historical divergence, extracts horn
glottochronology from which are given below. Using this kind of
1

you tail
lie

sit

table, estimates have been given for the possible

we feather stand
This isan approach, devised by the American
this hair give
linguists Morris Swadesh (1909-67) and Robert point of divergence of the languages in many of head
that say
Lees (1922- in the late 1940s, which determines
)
the world's families. The Caucasian family (p. 305) who ear sun
the rate at which a language has changed, over the is outlined, as an illustration. what eye moon
centuries. It aims to work out the length of time not nose star

CRITICISMS all mouth water

which haselapsedsince two related languages (ortwo many tooth rain
languages thought to be related) began to diverge. The approach is a controversial one for several rea- one tongue 1
stone
Glottochronology is the name of the study; lexico- sons. The method itself has been attacked on the two claw sand
statistics is the name of the technique it uses (but ground that it is impossible to construct a word-list big foot earth

some authors use the two terms synonymously). that shows no cultural bias — sun and moon, for long knee cloud

example, have great religious significance in some small hand smoke

A sample of vocabulary is taken from the lan- womar belly fire
cultures. It is also argued that the rate of change man
guages, using the basic word-list given right, and neck ash
the number of similar words between the languages may not be the same for all languages, and that person breasts burn

is counted, allowing for the effect of phonetic

far more known language histories would need to fish heart path
be analysed before the 86% figure was truly con- bird liver mountain
change. Thus, Italian padre and Portuguese pai dog drink red
would be accepted vincing. The method becomes less definite the
as equivalent, or cognate, words louse eat green
for 'father', because the relationship is explicable further back in history it goes, and the slightest tree bite yellow

292), whereas there is no reasonable phonetic of errors in the compilation of the word sample seed see white
(p.
could result in great inaccuracy: for instance, after leaf hear black
explanation which could relate either of these to,
root know night
say, the Eskimo word for 'father', ataataq. The 70 centuries of divergence, there would be only
bark sleep hot
word-list tries to avoid geographically or culturally 12% of cognates left, so that if just one cognate skin die cold
biased words, such as the names of plants or ani- was misanalysed, the result would be three cen- flesh kill full

mals, which would vary greatly from one part of turies in error. There are all kinds of problems blood swim new
which arise relating to whether words from differ- bone fly good
the world to another. grease walk round
Glottochronologists assume that the lower the ent languages are indeed 'the same' — in meaning
egg come dry
as well as in form. And often, not enough informa- name
number of vocabulary agreements between the two
tion is available about a language (especially for
samples, the longer the languages have been separ-
ated. Two languages which have 60% vocabulary
older states) for a complete sample to be drawn up.
Glottochronological
in common would be thought to have diverged Swadesh was aware of the limitations of
fully
estimates of the
longer ago than two languages which have 80% the procedure. But he argued that there must be
time depth
in common. Swadesh and Lees took several lan- a balance between the forces which maintain
of languages
guages where the period of time-change is known, uniformity in language and those which encourage
For example, if two lan-
and worked out a correlation between the percen- fluctuation, and pointed out that it is possible to
guages spoken today have
tage of common vocabulary and the interval of time obtain ancillary evidence from the dating methods 60% cognates, they di-
(or 'time-depth') which has elapsed since they used in archaeology. Certainly, the approach has verged 1 6.9 centuries ago -
diverged (as in the case of the Romance languages, generated many interesting hypotheses about early around 300 ad. The scale
language states and the relative chronology of stops at 25,000 years ago.
which have diverged from Latin since the early
Christian era). They found that on average two modern languages, and several scholars still use Percentage Minimum
it in their work - if only because no alternative
of divergence number of
languages would have 86% in common after 1,000
between two centuries of
years of separation. technique has been devised. languages divergence

A family tree of the Caucasian languages, with lexicostatistical 100

Proto-Caucasian (?)
95 1.7
percentages added (after J. C. Catford)
North Caucasian 90 3.5
-6- 85 5.4
80 7.4
75 9.5
70 11.8
65 14.3
60 16.9
55 19.8
50 22.9
45 26.5
40 30.3
35 34.8
30 39.9
25 45.9
20 56.6
15 75.6
10 102.6
5 148.4
Abkhaz Abaza Ubykh Kabard Adyghe Chechen Ingush Batsbiy Avar 8 Andi 5 Dido Lak Dargi Arci 5 Samur 3 Shah Udi Georgian Zan Svan
1 255.0
languages languages languages Dagh
languages

54 LANGUAGE CHANGE •
331

How does language change?
During the later part of the 19th century, it was
believed that a sound change affected the whole
of a language simultaneously: one sound system
would smoothly develop into the next, and all
words which contained a particular sound would
be affected in the same way.
We now know that linguistic change does not
operate in such an 'across-the-board' manner.
Some speakers introduce the change into their
speech before others; some use it more frequently
and consistently than others; and some words are
affected before others. A more accurate view is to
think of a change gradually spreading through the
words of a language - a view that is known as
lexical diffusion. At first just a few people use the
change sporadically in a few words (commonly
occurring words are influenced very quickly); then
a large number of words are affected, with the
sound gradually being used more consistently; then
the majority of the words take up the change.
Change from above
Three New York department stores were the setting
for the firstmajor sociolinguistic investigation of lan-
guage change in English, carried out by William Labov
in the early 1 970s. Labov wanted to find out why New
Yorkers do not always pronounce the final /r/ in words
like car- a fluctuation which had previously been con-
sidered a case of free' (i.e. random) variation. Using
data obtained from the salespeople at these stores,
he was able to show that the variation was systematic
- a matter of social status.
The stores represented the top, middle, and bottom
of the shopping range. Saks is on Fifth Avenue, a
fashionable shopping area. Macy's is a middle-class
store, in the middle of the price range. Klein's is near
the Lower East Side, dealing in low-priced goods.
Labov worked on the assumption that the sales-
people's accents would reflect those of their cus-
tomers, and that the use of ,'r/ would vary from store
to store.
The research technique involved visiting each store,
and asking staff questions about where certain goods
were located - always requiring the answer that they
were on the 'fourth' floor. If the interviewer were on
the fourth floor, he would ask simply, 'Which floor is
this?' By pretending not to hear the answer, he was
able to obtain a second response, more emphatic than
the first. All pronunciations of fourth were noted, along
with background data about the interviewees (sex,
approximate age, etc.). In each store, there were 264
interviews.
that the use of ,/r/ was highest in Saks,
Labov found
next highest Macy's, and lowest in Klein's. In other
in
words, the higher the socioeconomic group, the more
likely speakers are to use/r/. Moreover, speakers in
Klein's used /r/ much more often in their emphatic res-
ponses, which were more carefully pronounced than
their first, casual reply. This was confirmed in other
studies, which showed the incidence of/r/to increase
still further in yet more careful speech, during reading
aloud. Labov interpreted this to mean that these
speakers were manifesting a linguistic change in pro-
gress: /r/ had been unconsciously recognized as a
marker of high prestige, and was beginning to be used
in careful lower-class speech. The direction of change
was 'from above'.
332 •
IX THE LANGUAGES OF THE WORLD
100%
The evidence for this kind of process has largely
come from sociolinguistic studies of contemporary
*
linguistic variation. Pioneered by William Labov
(1927- ), on the assump-
these studies proceed
tion that the variation in language use which is
found in any community (and which fuels the
debates over linguistic identity and acceptability
(pp. 2-5)) is evidence of change in progress in a
language. Detailed observations are made of the Time
way in which different kinds of people speak in
Rates of change The rate at
different social settings. The parameters along which a change spreads
which these differences can be plotted are known through a language can be
as linguistic variables (p. 32). By examining the drawn as a curve with a char-
acteristic S-shape. There is a
frequency with which different people used a vari-
slow start, with few words af-
able, Labov was able to draw conclusions about
fected; a period of rapid ex-
the motivation, direction and rate of linguistic pansion, with most words af-
change. fected; and a slow
These are small-scale studies, but they have conclusion (after M. Chen,
1972).
large-scale implications. It is likely that the same
gradual process of change affects whole languages
as well as dialects. This process is not most accu-
rately represented using the family-tree model
(§50), with its clean splitting off of branches, and
several proposals have been made for a more dyna-
mic and sociolinguistically real alternative. The
metaphor of a 'wave' has proved particularly
attractive since the late 19th century: a change
spreads through a language in much the same way X Geographical space
as a stone sends ripples across a pool. But even
movement A wave model of linguistic
this implies too regular a to account
change A change starts at
for the reality of sociolinguistic variation. point X, and moves simul-
taneously through a geo-
Change from below graphical area (shown hori-
Martha's Vineyard is a small island off the east coast
zontally) and through the
of Massachusetts. Here, Labov studied the way in
strata of society (shown ver-
which the diphthongs (p. 1 54) [ai] and [au] seemed to
tically). The further it travels,
be changing. Traditionally, the first element of these
the less effect the change
diphthongs was a vowel sound which resembled the
has. If many changes were
a in far. But now, many people were pronouncing this taking place at once, the
vowel with a central quality, resembling the erof butter
speakers furthest away from
([ai] and [su]). Could this change be explained?
X would gradually lose their
To establish the nature of the variation, Labov inter- linguistic identity with those
viewed several islanders, and obtained examples of
at X. They would become
words containing these diphthongs. The results were
different dialects, and in due
analysed in terms of such factors as age, location, and
course perhaps even differ-
occupation. He found that the change was most notice-
ent languages (after C.-J.
able in the speech of people aged 30—45, and was
Bailey, 1973).
least in evidence in the oldest group (over 75). The
change was also more common in the western part
of the island, especially in the Chilmark area, where
the fishing trade is centred. The fishermen used more
of the centralized diphthongs than did people of any
other occupation - indeed, the pronunciation had long
been around in their speech, but in a less marked form.
This suggested that the change had begun with the
fishermen in the west, and spread from there.
But why should the fishermen have initiated this
change? Labov's explanation relates to the way the
island is used during the summer as a tourist centre:
its 6,000 permanent residents are increased by over
40,000 [ai]- or [au]-using visitors. The local fishermen
(especially the oldest) reacted against this invasion in
many ways, one of which was the subconscious exag-
geration of speech features which made them sound
different. Other islanders came to imitate the way the
fishermen were speaking, because (again subcon-
sciously) they admired their traditional character and
way of life. In due course, the change spread through-
out the island - a change from below'.
Why does language change? understood the causes of change, we could begin
to make predictions about when a change was
Sputnik
Only rarely can we give an
It is easy to recognize a change in language - but likely to take place, and observe it while it was exact date of entry for a new
only after it has taken place. It is not difficult to happening. There has long been imaginative specu- word into a language - a re-
reflect on how people spoke several years ago, to cent example being the inter-
lation on the matter, with suggested causes coming
national use of the word
point to a new word which has recently entered from fields as far apart as theology (that change sputnik, following the launch
the language, or to express an opinion about the is a consequence of man's arrogance, as manifested of the first artificial satellite
emergence of a cliche. What is almost impossible in the Tower of Babel) and climatology (that on 4 October 1 957. On 3 Oc-
is to predict a language change. Which sounds, change tober, the word was un-
is the consequence of man's physical loca-
known. By the evening of 4
words or grammatical constructions will change tion — the mountain-dweller having a physiologi-
October, it had entered
in the next 10, 20 years?
. . . cally different capacity for speech compared with hundreds of languages.
It is just as difficult to be precise about the origins the valley-dweller). Some scholars have adopted a
of a change in language. Who first used the new highly pessimistic view, feeling that the causes can
form? Where was it used? And when, exactly? His- never be found.
torical dictionaries always give an approximate These days, the speculation and pessimism are
date of entry for a new word or meaning - but being replaced by an increasing amount of scientific
these dates invariably reflect the earliest known use research, which has shown that there is no single
of that word in the written language. The first use reason for language change. Several factors turn
of the word in speech is always an unknown out to be implicated, some to do with the nature
number of years previously. of language structure, and some with the nature
To obtain answers to these questions, we need of society.
to know more about why language changes. If we

The nature of society

Geography When people fall out of daily use, and the the levels of bilingualism where people are not aware ish English is usually in the
move away from each language becomes obsolete. found in contact areas of the direction in which their direction of those linguistic
other, their language will di-
(§60). The minority lan- speech is moving, is less forms which are widely and
verge. The two groups will Imperfect learning Ac- guage forms a 'substratum' noticeable, but far more openly recognized as presti-
have different experiences, cording to one view, chil- which in the long term in- common. The movement gious (they have 'overt pres-
and at the very least their dren could be the initiators fluences majority usage. may be towards a favoured tige', as in the case of stan-
vocabulary will change. oflanguage change. They For example several varie- accent or dialect (one which dard English). This kind of
Similarly, when people might learn the adult forms ties of American English has 'positive prestige'), or change is often initiated by
come into contact with each imperfectly, and a new stan- display the influence of the away from one which is held people from the lower mid-
language will
other, their dard could gradually West African linguistic back- in low esteem (negative dle-class or upper working-
converge. The sounds, emerge. There are indeed ground of its black popula- prestige'). The speakers are class - especially women,
grammar, and vocabulary of similaritiesbetween the pro- tion. usually aware of the exis- who seem to be more aware
one group are likely to exer- cesses of language acqui- tence of linguistic differences of these factors than men. By
cise some influence on the sition (§38) and historical Social prestige People (e.g. 'I don't like the way contrast, subconscious
other. These days, the in- change, but it is unlikely that come to talk like those they those people talk), but un- change is usually in the op-
creased mobility of people, there is a causal connec- identify with or admire (p. aware of any trend in their posite direction, away from
within and between coun- tion. Children imitate 51 ). The process may be own speech related to their overt prestige. It is often in-
tries (p. 36), makes this a society, rather than the re- conscious or subconscious. attitude. itiated by working-class men,
major factor. verse. Conscious change can be Recent research in socio- and associated with such at-
According to another observed in those cases has shown the
linguistics tributes as toughness and
New vs old New objects view, change is the result of where people go out of their way in which patterns of virility, which carry 'covert

and ideas are continually one population imperfectly way to use or avoid a certain change relate to social pres- prestige'.
being created, and lan- learning the language of feature in their spoken or tige.For example, the work
guage changes to take ac- another. This is a common written language - such as of Labov and others has
count of them. At the same occurrence, as illustrated by whom or intrusive/r/ (§1 ). shown that conscious
time, old objects and ideas many immigrant groups, or Subconscious change, change in American and Brit-

The nature of language

Social factors can motivate systems should become change - and even cases ness' has a parallel in pho- chance errors in articula-

people to change their lan- simpler over long periods of
guage, but is there anything time. Some types of sound
in language itself which, so change do provide evidence
to speak, 'welcomes' a for this view - such as the
change? trend in many languages to
weaken or drop final conso-
Ease of articulation In the nants, or to allow adjacent
1 9th century, was widely
it sounds to influence each
where seems to have in-
it
creased. Only a small part
of language change can be
explained by a principle of
least articulatory effort.
Analogy Irregular features
in the grammar of a lan-
nology, where many tion. Certainly, many
changes have led to the de- changes in vocabulary are
velopment of a more sym- isolated and arbitrary; but
metrical sound system (p. there is no strong case for
165). randomness in phonology
or grammar. On the con-
Randomness Might lan- trary, similar processes of
guage change have no sys- change have been found in

felt that sounds changed other, as the case of as-

in guage are often influenced tematic explanation? has It unrelated languages all

because speakers would similation (p. 1 64). But there by its regular patterns: the been proposed that change over the world.
want to speak using as little are also many cases where exceptions are made to might be essentially unpre-
effort as possible. On this articulatory complexity is conform to the rule (p. 234). dictable - the result of ar-
basis, sounds and sound unaffected by a sound This trend towards 'neat- bitrary changes in fashion or

54 LANGUAGE CHANGE •
333
55 Pidgins and Creoles

Pidgin languages which appeared during the American Vietnam Possible origins of
campaign virtually disappeared as soon as the war the word pidgin
A pidgin is a system of communication which has
was over. But there are exceptions. The pidgin
grown up among people who do not share a com- the following have been
All of
known as Mediterranean Lingua Franca, or Sabir, suggested as sources for the
mon language, but who want to talk to each other,
began in the middle ages and lasted until the 20th word pidgin, which is first at-
for trading or other reasons. Pidgins have been vari-
century. tested in print in 1850:
ously called 'makeshift', 'marginal', or 'mixed' lan-
Some pidgins have become so useful as a means • A Chinese mispronuncia-
guages. They have a limited vocabulary', a reduced
of communication between languages that they tion of the English word
grammatical structure, and a much narrower range business.
have developed a more formal role, as regular auxi-
of functions, compared to the languages which
liary languages. They may even be given official • The Portuguese word
gave rise to them. They are the native language
status by a community, as lingua francas. These ocupagao (business).
of no-one, but they are nonetheless a main means
cases are known as 'expanded pidgins', because • The Hebrew word pidjom
of communication for millions of people, and a
of the way in which they have added extra forms (barter).
major focus of interest to those who study the way
to cope with the needs of their users, and have • A Yayo word pidians
languages change.
come to be used in a much wider range of situations meaning people.
It is essential to avoid the stereotype of a pidgin
than previously. In time, these languages may come • Portuguese pequeno
language, as perpetrated over the years in gener-
to be used on the radio, in the press, and may even child)— cf. 'baby-
ations of children's comics and films. The 'Me Tar- (little,

develop a literature of their own. Some of the most talk'.

zan, you Jane' image is far from the reality. A pidgin
widely used expanded pidgins are Krio (in Sierra • English p/geon-suitable
is not a language which has broken down; nor is
Leone), Nigerian Pidgin English, and Bislama (in for carrying simple
it the result of baby talk, laziness, corruption, pri-
Vanuatu). In Papua New Guinea, the local pidgin messages.
mitive thought processes, or mental deficiency. On
(Tok Pisin) is the most widely used language in
the contrary: pidgins are demonstrably creative
the country.
adaptations of natural languages, with a structure
and rules of their own. Along with Creoles (p. 336),
they are evidence of a fundamental process of Below: A road sign in Tok Pisin: Kainantu must once again
linguistic change, as languages come into contact become the gateway to the highlands. Don't throw rubbish Right: Part of the front page
with each other, producing new varieties whose around.' Kainantu is the first major government centre on the ofWantok (Friend), a
Highlands Highway from Lae on the coast to Mt Hagen in Papua New Guinea weekly
structures and uses contract and expand. They pro-
the interior. The sign refers to a slogan in English that was newspaper written entirely in
vide the clearest evidence of language being created used before independence, when Kainantu had an airstrip pidgin (Tok Pisin) (with an
and shaped by society for its own ends, as people (now closed): Kainantu: Gateway to the Highlands'. English sports section). The
adapt to new social circumstances. This emphasis headline of this issue (July,

on processes of change is reflected in the terms pid- Right, below: A street poster from Freetown, Sierra Leone, 1985) reads: More refugees
written in Krio: Electricity has no legs: it's Kabelmetal cable cross border but local people
ginization and creolization. that carries it.' short of food'.
Most pidgins are based on European languages
- English, French, Spanish, Dutch, and Portuguese
— reflecting the history of colonialism. However,
thisobservation may be the result only of our ignor-
ance of the languages used in parts of Africa, South
America or South-east Asia, where situations of
language contact are frequent. One of the best-
known non-European pidgins is Chinook Jargon,
once used for trading by American Indians in
north-west USA. Another is Sango, a pidginized
variety of Ngbandi, spoken widely in west-central
Africa.
Because of their limited function, pidgin lan-
guages usually do not last for very long — sometimes
for only a few years, and rarely for more than a
century. They die when the original reason for com-
munication diminishes or disappears, as communi-
ties move apart, or one community learns the
language of the other. (Alternatively, the pidgin
may develop into a creole.) The pidgin French
which was used in Vietnam all but disappeared
when the French left; similarly, the pidgin English

334 •
IX THE LANGUAGES OF THE WORLD
 Shakespeare in pidgin
The range of pidgin English is well illustrated by the
translations which have been made of such works as
the Bible and Shakespeare. Here is an extract from
Julius Caesar (Act III, Scene 2), translated into Krio
pidgin and Tok Pisin.

Friends, Romans, countrymen, lend me your ears;

I come to bury Caesar, not to praise him.
The evil that men do lives after them;
The good is oft interred with their bones;
So let be with Caesar. The noble Brutus
it

Hath told you Caesar was ambitious.

Samba
Nar, 5*0 — Wik i-,taC long 20 Julai. !«•<
were so, was a grievous fault;
It it it

And grievously hath Caesar answer'd it.

Here, under leave of Brutus and the rest -

I*K> "T..
Neks For Brutus is an honourable man;
So are they all, all honourable men -
Come to speak in Caesar's funeral
I

wik Krio
Padi dem, kohntri, una ohl wey dey na Rom. Meyk una
ohl kak una yeys. A kam ber Siza, a noh kam preyz
am. Dem kin memba bad wey pohsin kin du lohng tern
afta di pohsin kin dohn dai. Boht plenti tern di gud wey
pohsin du kin ber wit im bon dem. Meyk bi so wit i

Siza. Bra Brutohs dohn tel una sey Siza na bin man
• as* t f» cm* > long w«fi
wey want pas mak. It tohk tru, na badbad ting dis
i

ya. En Siza dohn get im bad pey foh dat. A tayk pamis-
hohn frohm Bra Brutohs dem foh kam tohk na Bra Siza
im berin. En Bra Brutohs na ohnareybul O! Dem ohda
wan sef na ohnareybul.
(From T. Decker, 1965, p. 74.)

Moa refuji i
Tok
Pren,
Pisin
man bolong Rom, Wantok, harim nau. Mi
tasol long plantim Kaesar. Mi noken beiten longen.
Sopos sampela wok bolong wampela man stret; sam- i
kam

pela no stret; na man dai; ol wailis long wok no

i i i i

stret tasol. Gutpela wok bolongen slip; lus nating i i

long giraun wantaim long Kalopa. Fesin bolong yumi

kalapim boda man. Maski Kaesar tu, gutpela wok slip.

Brutus ia tokim yu long Kaesar mangal. Sopos olo-
sem, bikpela pekato tru. Tasol Kaesar Kalopa bekim
pinis long virua belongen. Tru, Brutus, na ol pren
i
i

bolongen, gutpela man. orait. Ol gipim mi orait long

I i

mi toktok sore hia long Kaesar.

tasol ol asples i sot long kaikai (From J. J. Murphy, 1 966, pp. 1 9-20.)

in a«a
A page from a New Guinea road safety handbook RotSefti
Long Niugini (1972), with instructions in English (top), Tok
Pisin (middle), and Hiri Motu (bottom).

If you have an accident, get the

other driver's number, if possible
k kaik* his name and address and re-

A^^
port to the police. Do not fight
it

E him or abuse him.

Sapos yu
namba
kisim bagarap kisim
bilong narapela draiva,
sapos yu ken, kisim naim bilong
em na adres tu, na tokim polis
longen. Noken paitem em o tok
nogut long em.
'AM Bema kerere davaria neganai,
tebelmetal taraka o motuka taria tauna ena
ladana oi abia bona ena noho o
gaukara gabuna danu abia. Tar-

^CARRYAM aka o motuka ena naba danu

abia vadaeni Police hamaoroa.
Oi heai bona hereva dika lasi.

kabelmetal
55 PIDGINS AND CREOLES •
335

Creole languages
A is a pidgin language which has become
Creole
themother tongue of a community - a definition
which emphasizes that pidgins and Creoles are two
stages in a single process of linguistic development.
First, within a community, increasing numbers of
people begin to use pidgin as their principal means
and so on - though the genetic relationship of a
creole to its dominant linguistic ancestor is never
straightforward, as the creole may display the
influences of several contact languages in its
sounds, vocabulary, and structure.
Today, the study of creole languages, and of the
pidgins which gave rise to them, attracts consider-
able interest among linguists and social historians.

of communication. As a consequence, their chil- To the former, the cycle of linguistic reduction and
dren hear it more than any other language, and expansion which they demonstrate, within such a
gradually it takes on the status of a mother tongue short time-scale, provides fascinating evidence of
for them. Within a generation or two, native lan- the nature of language change. To the latter, their
guage use becomes consolidated and widespread. development is seen to reflect the process of explor-
The result is a Creole, or 'creolized' language. ation, trade, and conquest which has played such
The switch from pidgin to Creole involves a major part inEuropean history over the past 400
a major expansion in the structural linguistic years.
resources available - especially in vocabulary,
Guvanese Creoles compared
grammar, and style, which now have to cope with French Creole Kno English The similarities between Euro-
the everyday demands made upon a mother tongue
pean-based Creoles are
speakers. There also a highly significant Mangez Ma3e Chop Eat
by its is
striking, as can be seen from
mange
J'ai Mo ma3e A chop I ate
shift in the overall patterns of language use found U/Elle a this comparison of the verb

in the their nature auxi-

community. Pidgins are by mange Li ma3e I chap He She ate phrase in the French-based
Je mange Je creole of Guyana and the
liary languages (§58), learned alongside vernacular
suisen train English-based Krio of Sierra
languages which are much more developed in struc- de manger Mo ka ma3e A de chop I am eating
Leone (after L. Todd, 1984,
ture and use. Creoles, by contrast, are vernaculars J'avais mange Mo te ma3e A bin chop I ate/ had eaten
p. 24).
Je mangeais Mo te ka ma3e A bin de chop I was eating
in their own right. When a creole language devel- Mo ke ma3e A go chap
Je mangerai I shall eat
ops, it is usually at the expense of other languages U/ tile est Li gros pas u I big pas yu He She It is

plus grand bigger than

spoken in the area. But then it too can come under
que vous you
attack.
The main source of conflict is likely to be with
the standard form of the language from which it Where do pidgins and Creoles
derives, and with which it usually co-exists. The
standard language has the status which comes with
come from?
social prestige, education, and wealth; the creole The world's pidgins and Creoles display many
has no such status, its roots lying in a history obvious differences in sounds, grammar, and voca-
of subservience and slavery. Inevitably, creole bulary, but they have a remarkable amount in com-
speakers find themselves under great pressure to mon. Two opposed theories have attempted to
change their speech in the direction of the standard explain these differences.
- a process known as decreolization.
One consequence of this is the emergence of a MANY SOURCES?
continuum of several varieties of creole speech, at A long-standing view is that every Creole is a
varying degrees of linguistic 'distance' from the unique, independent development, the product of
standard - what has been called the 'post-creole a fortuitous contact between two languages. On
continuum'. Another consequence is an aggressive the surface, this 'polygenetic' view is quite plau-
reaction against the standard language on the part sible. It seems unlikely that the pidgins which devel-
of Creole speakers, who assert the superior status oped in South-east Asia should have anything in
of their creole, and the need to recognize the ethnic common with those which developed in the Carib-
identity of their community. Such a reaction can bean. And it is a general experience that these varie-
lead to a marked change in speech habits, as the come into use in an apparently spontaneous
ties
speakers focus on what they see to be the 'pure' way — as any tourist knows who has faced a sou-
form of creole — a process known as hypercreoliza- venir seller. Would not the restricted features of
tion. This whole movement, from creolization to the contact situations (such as the basic sentence
decreolization to hypercreoiization, can be seen at patterns and vocabulary needed in order to trade)
work in the recent historv of black English in the be enough to explain the linguistic similarities
USA. around the world?
The term creole comes from Portuguese crioulo, The view is tempting, but there are several
and originally meant a person of European descent grounds for criticism. In particular, it does not
who had been born and brought up in a colonial explain the extent of the similarities between these
territory. Later, it came to be applied to other peo- varieties. Common features such as the reduction
ple who were native to these areas, and then to of noun and pronoun inflections, the use of parti-
the kind of language they spoke. Creoles are now cles to replace tenses, and the use of repeated forms
usually classified as 'English based', 'French based', to intensify adjectives and adverbs are too great

336 • IX THE LANGUAGES OF THE WORLD

to be the result of coincidence. Why, then, should
the pidginized forms of French, Dutch, German,
Italian, and other languages all display the same
kind of modifications? Why, for example, should
the English-based Creoles of the Caribbean have
so much in common with the Spanish-based Creoles
of the Philippines? How could uniformity come
from such diversity?
ONE SOURCE?
The opposite view argues that the similarities
between the world's pidgins and Creoles can be
explained only by postulating that they had a com-
mon origin (i.e. are 'monogenetic'), notwithstand-
ing the distance which exists between them.
Moreover, a clear candidate for a 'proto'-language
has been found — a 15th-century Portuguese pidgin,
which may in turn have descended from the Medi-
terranean lingua franca known as Sabir (p. 338).
The Portuguese are thought to have used this pidgin
during their explorations in Africa, Asia, and the
Americas. Later, it is argued, as other nations came
Meanwhile, other theories have been proposed,
in an attempt to explain these similarities and dif-
ferences. Other forms of simplified speech have
been noted, such as that used by children (§41),
in telegrams and headlines, and in talking to for-
eigners (p. 373). It is possible that the processes
underlying pidgins and Creoles reflect certain basic
preferences in human language (such as fixed word
order, or the avoidance of inflections). In this con-
nection, these languages provide fresh and intrigu-
ing evidence in the search for linguistic universals
(§14).
Pidgins compared
Lexical similarities and differences between pidgins are
list of items collected by F. G. Cassidy
clearly illustrated in this
in the 1960s, taken from the set of basic words' used in glotto-
chronology (p. 331 ). The English element predominates in
Tok Pisin and Chinese Pidgin; in Sango, the vast majority
of the words are African; in Chinook, most words are from
Chinook or other Amerindian languages (but note the
influence of both French and English). French names for parts
of the body have emerged in Sango and Chinook. Though

to these areas, the simple grammar of this pidgin
came to be retained, but the original Portuguese
vocabulary was replaced by words taken from their
own languages. This view is known as the relexifi-
cation hypothesis.
there is no historical connection between the languages, note
also the coincidences of thought which have produced the
figurative phrases for feaf/7er(grass-of-bird (Tok Pisin), hair-
of-bird (Sango), and leaf-of-bird (Chinook)), and the words
for heart \n Tok Pisin and Chinook, both of which stress the
notion of heartbeat.

There is a great deal of evidence to support the

Tok Chinese Chinook
theory, deriving from historical accounts of the English Pisin Pidgin Sango Jargon
Portuguese explorations, and from modern analy-
bell bel bell ngberena tintin
ses of the languages. For instance, every English- big bigfela big kota hyas
based pidgin and creole has a few Portuguese bird pigin bird(ee) ndeke kalakala
words, such as savi 'know', pikin 'child', and bite kajkajim bitee te muckamuck
black blaekfeb black (zo)vok5 klale
palava 'trouble'. In Saramaccan, an English-based
blood blut blood mini pilpil
creole of Suriname, 38% of the core vocabulary
cold kilfeb colo de cole, tshis
is from Portuguese. Early accounts of Chinese pid- come kam li ga chahko
gin refer to a mixed dialect of English and Portu- die daj dielo kui memaloost
guese. And on general grounds, relexification of dog dog doggee mbo kamooks
drink drirjk dlinkee, yc. mucka-
a single 'proto'-pidgin seems a more plausible
haw muck
hypothesis than one which insists on a radical par- ear ir ear mi kwolann
allel restructuring of several languages. earth grawn glound sese illahie

The shift in approach, implicit in the relexifica- eat kajkaj chowchow kobe, te mucka-
muck
tion theory, is fundamental: it is not the case that
fat gris fat, glease mafuta glease
English, and the other languages, were 'creolized', koati kalakala
feather gras bitarj redder
but that an original (Portuguese) creole was 'Angli- pigin ndeke yaka
cized'. However, not all the facts can be explained tup so
fishee susu pish
in this way. Pitcairnese creole has no Portuguese fish fis

give givim pay fu potlatch

influence, and yet has much in common with other
green grinfeta gleen, lu voko kete pechugh
varieties. What accounts for those similarities? hair gras bibrj hair k5a yakso
Then there are several pidgins and Creoles which hed
have developed with little or no historical contact hand haen hand, sho mab5ko le mah
head hed headee
with European languages — Sango and Chinook,
li la tet

heart klak heart coeur tumtum

for instance. And seem to be many structural
there
know save savvy hinga kumtuks
differences between European and non-European man masn man kali man
pidgins and Creoles, which the common origin no no na non wake
hypothesis finds difficult to explain. nose nos peedza h5 nose
one wanfela one piecee 5ko ikt
The evidence is mixed. Disentangling the struc-
small hkhk likki kete tenas
tural similarities and differences between these sun san sun la sun,
varieties is a difficult task, and the evidence could otelagh
be taken to support either a monogenetic or a poly- talk tok talkee tene wauwau
two tufeb two ose mokst
genetic theory. Far more descriptive studies are
warm hotfeta warm wa waum
needed before we rule out one view or the other.

55 PIDGINS AND CREOLES •

337

100 pidgins and
Creoles
(After Ian F. Hancock, 1971)
1 Hawaiian Pidgin Creole
English-based, with influence of
Chinese. Japanese. Hawaiian,
Portuguese, and Philippine lan-
guages, c. 500,000 speakers,
many as a first language.
2 Pitcairnese Creole English
Descendents of the mutineers
from HMS Bounty, who settled
here in 1790. English-based
with a little Tahitian influence,
150 speakers.
3 Chinook Jargon
Chinook-based, with influence
of English. French. Nootka. and
Salishan dialects. Spoken by
1 00.000 in late 1 9th century, but
now nearly extinct.
4 Pidgin Eskimo
A series of Eskimo-based
gins, used in trading with
whites, and with Athabaskan
dians.
5 Pachuco (Pochismo)
Spanish-English contact lan-
guage in limited use in Arizona
and parts of southern California.
6 Trader Navaho
Navaho-based. used by traders
to the Indians, but not conver-
sely.
7 Franco-Amerindian
Used in the 1 7th century be-
tween French settlers and
dians, around Montreal.
8 Souriquoien
Used in the 1 7th century be-
tween French fishermen and
cal people in Nova Scotia.
9 New Jersey Amerindian
Once used between
and Dutch traders and
sey Indians. Showed the
fluence of English grammar, but
with a largely Algonquian voca-
bulary.
10 Mogilian
Choctaw-based pidgin, formerly
used by many Indian tribes
along the Gulf Coast and Mis-
sissippi River.
11 Amerindian Pidgin
English
English-based pidgin once used
widely in the USA between
traders and Indians. First re-
corded in 1641.
12 Gullah
English-based Creole used
along the USA south-eastern
coast. Many features in com-
mon with West African
Estimates vary. 150-300.000.
13 Louisiana Creole
French
A French-based Creole still
used in parts of eastern Louis-
iana, but diminishing in
of speakers.
338 •
IX THE LANGUAGES OF THE WORLD
<£
14 Belize Creole English 27 Trinidad and Tobago 32*.
Used as a lingua franca in rural Creole
areas; as a first language in Both Spanish- and English-
mainly urban areas. based varieties are used in
these islands, as well as an indi-
15 Meskito Coast Creole genous patois. The influence of
A pidginized form of some of the immigrant languages, such as
%
Caribbean creole dialects used Hindi, is apparent.
along the Meskito Coast area of
28 Guyanese
Nicaragua.
Creole Creolese
16 NahuatkSpanish Creole English-based Creole used in
Used in Nicaragua from the Guyana, with influences from
1 6th century, and now probably
other Creoles, such as Barba-
extinct. dos and Sierra Leone.
c. "****&. 22
17 Papiamentu 29 Nikari Karu Pidgin '
20»>£* J
...20
(Papiamento) A Portuguese-based pidgin, • 23
5
A Spanish Creole, derived from used in Guyana near the u ^Y 17
a Portuguese pidgin, with Dutch Brazilian border. ^5y •«is I
•* **
vocabulary influence. Used in 29 •
) 19
"
Curacao. Bonaire, and Aruba 30 Guyana Dutch Pidgin 31
by c.200,000 speakers. This variety is reported to have
been used on inland rivers in
18 Pidgin Spanish Guyana, in the 1 9th century. A
A Spanish-based trading lan- creolized form is still used in S.
pid- guage used mainly by two In- Guyana.
dian tribes in west Venezuela.
In- 31 Portuguese Bush Negro
19 Spanish Creole Portuguese-based dialects
Several creolized varieties used used in Suriname, with English
by Indians in north Colombia. and African vocabulary in-
fluence.
20 Caribbean Creole English 48 Km English
Around 30 English-based 32 Brazilian Creole An English-based
Creoles are found throughout Portuguese 41 Sabir pidgin used by
the islands of the Caribbean, Used by Brazilians of African A pidginized variety of Proven- Kru fishermen in
some represented by several ancestry in rural An Ita-
areas. cal, used in many Mediterra- Liberia, and along
The largest is Jamai-
varieties. lian-Negro variety known as nean ports (and in the Middle the West African
can Creole, with over 2 million Fazandeiro exists in Sao Paulo. East, during the Crusades), and coast.
speakers (and cf. Trinidad and influenced by the vocabulary of
49 Petit-Negre
Tobago Creole). 33 LingoaGeral
other languages in the area.
ATupi-Guarani-based pidgin A French-based pidgin used
In-
21 Haitian French Creole
Now extinct.
used in Brazil; now losing mainly by soldiers in the Ivory
Used in three main varieties in ground to Portuguese. 42 Cape Verde Creole Coast and other former French
Haiti by over 4 million speakers. A Portuguese-based Creole, possessions along the West
34 Cocoliche African coast.
used in two main dialects in the
22 Virgin Islands Dutch A variety of
Italianized Spanish
Cape Verde Islands. It is re-
lo- 50 West African Pidgin
Creole used around Buenos Aires. ported to be spoken by groups English
Widely used in the 1 9th century descended from 1 9th-century
(a New Testament translation 35 Franco-Spanish Pidgin A mutually intelligible chain of
immigrants in Massachusetts English-based pidgins and
English was produced in 1818), but now A contact language with limited
and California. Creoles is found throughout
New Jer- nearly extinct. use in Buenos Aires - in some
in- studies referred to as Fragnol'. 43 Kryol
West Africa - such as Ghana.
Togo. Nigeria, and Fernando
23 Antilles Creole
36 Russenorsk
A Portuguese-based creole,
French-based varieties spoken
Po (and cf. Gambia, Krio, Meri-
A contact language, used Senegal,
in c. 57,000
derived co. Kru. and Cameroon).
in such islands as Grenada. speakers.
from Russian and Norwegian,
Guadaloupe, Dominica. Martini- now 51 Gulf of Guinea
nearly extinct. 44 Gambian Krio (Aku)
que. Saint Lucia, Trinidad, and Portuguese
Tobago. An English-based creole used
37 Anglo-Romani A group of Creoles used on the
inthe Gambia; widely used as
A creolized variety, islands of Annobon, Sao Tome,
24 Sranan a second language, with limited
derived from Romany and Principe.
English-based creole of coastal use as a first language. Gra-
and English, used by
Suriname and other coastal dually being ousted by English 52 Cameroon Pidgin English
Gypsies in Britain.
ports. A widely used lingua and pidginized Wolof. An English-based pidgin, creo-
lized in some urban areas, used
franca, and a first language of 38 Sheldru (Shelta)
45 Crioulo
c. 80.000 speakers. in Cameroon as a second lan-
An Anglo-Irish creole. used
A Portuguese-based creole guage by about 2 million
mainly by Irish travellers and used in Guinea, widely used as
25 English Bush Negro speakers. Related varieties are
their descendants in England,
English-based creole dialects of a lingua franca. used in eastern Nigeria and
USA, and elsewhere: based on
Boni (or Aluku), Djuka (or Au- 46 Krio Fernando Po.
English grammar with (often
can), used in Suriname. Djuka An English-based creole of an
modified) Irish vocabulary. 53 Ewondo Populaire
is the only creole to have deve-
area based on Freetown in Sier- An African-based pidgin used
loped a syllabic writing system. 39 Ingles de Escalerilla ra Leone, c. 50,000 speakers as as a trade language in the
varieties. Pidgin Djuka is used between A Spanish-English-based pid- a first language, and widely Yaounde area of Cameroon.
the Bush Negroes and the Trio gin used in some Mediterra- used as a second An old-
Indians. nean ports, such as Malaga and fashioned variety is also found 54 Barikanci
A Hausa-based pidgin, used as
La Linea. in Liberia.
26 French Guyana Creole a lingua franca among the
French-based creole. with 40 North African Pidgin 47 (A)Merico armed forces in northern Niger-
some Portuguese influence, French (Settler English) ia. Pidginized Hausa is also
number used in Cayenne and along the A French-based pidgin, also An English-based creole used found in Cameroon, and on
N.
coast, c. 50.000 speakers. known as Petit Mauresque'. along the Liberian coast. the desert route to Mecca.
 88 Chabacano
A Spanish-based creole, with
influences from Tagalog and
Cebuano, used in Zamboanga,
the Philippines.
89 Davaueno
A Spanish-based Creole, used
in Davao, the Philippines.

90 Bamboo Spanish
A Spanish-based pidgin, used
first by the Japanese and later

by the Chinese in Davao, the

Philippines.

91 Ternateno
A language once used in Ter-
between
nate, the Moluccas,
Spanish Mexican soldiers and
the Portuguese-speaking local
community.
92 Tok Pisin (Neo-
Melanesian)
An English-based pidgin, in-
fluenced by local Papuan lan-
guages, widely used in Papua
New Guinea by around 1 million
people; it has been creolized in

some areas.
93 Hiri Motu
A pidginized variety of Motu,
formerly often called Police
Motu', much influenced by Eng-
lishvocabulary, widely used as
a trade language for over 1 00
55 Tekrur 63 Cape Dutch (Taal Dutch) 71 Sri Lanka Portuguese 80 Bamboo English
years in the Port Moresby area
An Arabic-based pidgin, widely Afrikaans-based varieties used A Portuguese-based creole for- An English-based pidgin used
of Papua New Guinea.
used as a lingua franca to the in the South African Cape; the merly used by Indo-Portuguese inKorea, especially during the
east of Lake Chad, and in the label Cape Coloured Afrikaans' Christian immigrants, and still Korean War; almost extinct. 94 Solomon Islands Pidgin
Bodele depression in northern is also used. spoken in parts of Sri Lanka.
An English-based pidgin used
81 Japanese pidgin in and around the Solomon Is-
central Africa.
64 Fanagalo 72 Goanese lands; becoming creolized in
An English-based pidgin widely
56 Juba Arabic A Zulu-based English pidgin, A Portuguese-based Creole used Japanese ports in the
in urban centres.
An Arabic-based pidgin which with some lexical influence, used in Goa; probably now ex- late 1 9th century, and also in 95 Bislama (Beach-la-Mar)
developed in the southern Su- used mainly by African mine- tinct. areas which saw American An English-based pidgin, with
dan around the end of the 1 9th workers around Johannesburg,
occupation in the 1 940s; no locallanguage influences,
century. and in Zimbabwe and Namibia. 73 Indian Portuguese
longer used. widely used in Vanuatu, Fiji,
It is known by many names, Several pidgin and creole var-
57 Galgaliya and surrounding areas as a
such as Mine Kaffir and Kit- ieties used in ports along the In-
An Arabic-based pidgin used by 82 Vietnam Pidgin lingua franca; some first-lan-
chen Kaffir. dian coast, now largely extinct.
An English-based pidgin used
the Kalamafi tribe in north- guage use. Bislama derives
eastern Nigeria. 65 Zambia Pidgins inVietnam between local peo- from beche-de-mer', a local
74 St Helena Creole
Several African-language- ple and American servicemen; variety of sea-slug.
58 Sango
An English-based Creole used
based pidgins are used in the now largely extinct.
A pidginized variety of
Ngbandi, on the island of St Helena; 96 Bagot Creole English
Zambian copper belt, such as some use as a pidgin. An English-based variety deriv-
with French vocabulary in- 83 Malacca Portuguese
the Cen-
Town Bemba. ing from Australian Pidgin, used
fluence, used widely in A Portuguese-based creole
75 Bazaar Hindustani on the Bagot Aboriginal Re-
tral African Republic, and spor- 66 Reunionnais used in western Malaysia by
A Hindi-based pidgin, used as serve near Darwin, northern
adically in Cameroon and Chad. A French-based creole used in about 3,000 people.
a lingua franca in urban centres Australia.
Reunion by over 200,000 peo-
59 Congo Pidgins in northern India.
84 Singapore Portuguese 97 Australian Pidgin
ple.
Many pidginized varieties of lo-
An English-based
76 Madras Pidgin A Portuguese-based creole, pidgin which
cal African languages in the 67 Barracoon may have developed
much with some Malay and English out of
Congo, such as Kituba (derived A language used in the 19th An English-based pidgin,
influence, used in parts of Neo-Melanesian; it is possible
from Kikongo) and Lingala (pid- century in the Mozambique influenced by Dravidian lan-
Singapore. that the influence was in the
ginized Ngala). Kituba has ports, containing elements from guages, used during British rule
and
found in some other direction, with New Gui-
around 2 million speakers as a many languages, such as Ara- in India, still
85 Jakarta Portuguese nea plantation workers carrying
second language. bic, Swahili, Portuguese, and urban communities.
A Portuguese-based creole, for- the pidgin back home.
Malagasy. merly spoken in Jakarta and
60 Asmara Pidgin Italian 77 Tay Boy 98 New Caledonia Pidgin
An Italian-based pidgin still 68 Mauritian French Creole A French-based pidgin, widely nearby, and now probably ex- A French-based pidgin used in
used in Eritrea, in Ethiopia. Used by most of the population used in Vietnam during the per- tinct.
New Caledonia, in the south-
of Mauritius, and by some in iod of French control; now al- west Pacific.
61 Swahili Pidgins 86 Bazaar Malay
Madagascar and the Comoros most extinct.
Several varieties of pidginized A pidginized variety of standard 99 Norfolkese
Islands; increasingly influenced
Swahili are used in eastern Afri- Malay, widely used in Malaysia An English-based creole, which
by English. 78 Makista (Macauenho)
ca,such as Kisettla, spoken in and Indonesia. Also in this developed from Pitcairnese,
Kenya between Europeans and 69 Rodrigues Creole
A Portuguese-based Creole, in- used by settlers who moved to
fluenced by Chinese vocabu- area, Baba Malay, a pidginized
Africans. Creolized in some A French-based creole, used on variety strongly influenced by Norfolk Island from Pitcairn Is-
lary, used mainly in Macau.
areas. Rodrigues Island, near Mauri- Chinese. land, in the 19th century.
tius, by around 17,000 people. 100 Maori Pidgin
62 Afrikaans Pidgin 79 China Coast Pidgin
Used in the Namaland region of 70 Seychellois An English-based pidgin, for- 87 Caviteho and Ermitano An English-based pidgin used
south-west Africa between A French-based creole, used in merly widespread use in
in Spanish-based Creoles used in during the early years of coloni-
tribesmen and Afrikaners. the Seychelles and other is- coastal China and Hong Kong, the area around Manila, in the zation in New Zealand; no
Creolized in some areas. lands by around 40,000 people. but now almost extinct. Philippines. longer spoken.

55 PIDGINS AND CREOLES • 339

 .

M*.l«.ll\Jtsl».]
>^*> I w international r**^ * j
T
Published *ilh The Ne* ^ork Time* and The «a.<-hin£1»n Post
LOMMiV MuNIlO
^* OMM Arnft** jtX>» "V*H -VM wl 0*1 \0**1
M\Rl_H 23.1987 *-T\HIJ-llri< IK

( x> r ft to )
France, Emphasizing U.S. Ties.
Will Sign Pact on Disney Park \»n 112 - N h* - I Tt«i 19 ~-J

^|:\ •
"-

Mr
—
part
Ttep* —
N> tXw I*
' .

hi. I ml
-i
f««rt«;^\ar-jr;nr
' - t~ • « = .

MilS
TV
.
J

1
CORRIERE BELLA SE
=1
La diraione socialista si e pronunciata sulla lettera dei presidente incaricato La C roce Rossa invita gli americani a so ttoi
FmHrilSite^S* ai-li.i^m,.
|<<ac>a4l*(W f ™fj **> ". pef.r- 4NrT| »^

dOCUmentO Aids, e allarme negli Sta

a.' -.

CWKiht^riwiMV*^
II PSi bOCCia il

»*H^m J* J mm at frr»
yff it aaaaaaai J ;**»* «hg »*> Oarav j
Ripropone ad Andreotti: facciamo governo e referendum P6f 16 tfilSfUSIOIII ddl 7J
•-uno dai 30 milioni
25 ai le persone interessate
ontagio dovuti a plasma infetto sarebbero solo I

Nouvelle-Caledonie : la France retrouvee (

Mi^c
MONTANT DU JOUR
30 OOO F a gagner
- 38
LE FIGARO
premier quotidien national francais
» I J tnOtOi II HAAS 1M7 f '3 233, - COITION DE 5 MEuBES - PMX «»F rrancaco La A&octt.
A Ping)
uno yieftzuto n
I fa at*. Meluua In
a inietU
Reunion de Mati g non le g ouvernement
« uni et solidaire »
:

sur les objectifs fixes Les Besse :

cadres la preuve
Chirac : poursuivre mec<
Z^ president c

Taction engagee que fe

ment
potnoir
soit

- Er nnunni - ,
******
j
'-
I
]
Str k pita koooauqoe. pus de relax*, mtis an serieux co§p •^ar-^naaa •»*• 3C
.*»..?•
L£S ttlCLUXQIS de pooce tux unestissemei>ts des entreprises. '.I M 9—i pi- e [ J«^JI
*J. AjBLAM-LVrULNAnOVU.
ar.m.f-.-c g c
coups 1 i- M »- q*g>9*W B*r 4M ' iWITI ';-•• M i *. r. .

^ » iww «»it\*i*|riii org tr •*• imMmM cvn^*y i*eii»« : pe^f

M HH o* 4
.
*) Ci a
*<»P»4?r« 2P*^ w»fl'0' • Sag** :-*->• rar^*- • *•*»-• a tcrM om -.«:•*

-goc-
mti : ac . •«
ULi Uii^J «M jJI >J--ij J-W' y> tU.1
LE r, L A B ,O.Cr l t
IT^
f
JI^JJl Hispanic 28C En Miuicnuttttl

F©RUM
SScFutri Dtl ElUtft
News 47$ C Sum. Uum
Miomo
Journal tk ! <t: ~>. ;

Hr»l Bilingual N»»sp»p»r In Nf» F.ngltnd

JOulNAL BIIINGUE
j--ji i
Alcalde While. Presidenle Carter...

"Religion et Nationali Pondra Hispanos Ordena La Marina

s-4
A Cooperar Con
LOU I SI/!
Mensuel publie par la Fondation Louisiana)!
Cargo Organismos
De La
Alcaldl'a!
» ^Lh
.Bn^'f^
I

Refugiados
"N«dk Seri DevucMa
A Cuba!"...
Uli/MH. LA
NO. 33, AVR1L 1980 Mayor I* hile President Carter.

Hispanics w-H m llr2T Orders Navy

NORDISKSIiSTIDEI Will Be I Cooperate With
MOOWTGIAM 11*11 INOHSKI MVXtTI*
Appointed To j Refugees
"Sobody "-ill Be Returned
Lakkasj* vtd kjt City Posts! TAG* I FAGl* To Cuba.'"...
MihH ».lii.ik i,,,n
 .

PARTX
• Kll 10 '»

HttHWiRJAli- Sf Language in the world

Language is the main means whereby people com- fourth is to foster the growth of multilingualism
municate. It is also, ironically, the main means in individuals and societies, either through the
whereby people fail to communicate. These simple natural course of events, as people come into con-
facts motivate the content and organization of Part tact, or through the promotion of special educa-
x of the encyclopedia. We need to examine the pro- tional procedures for teaching languages. Each of
blem from both international and intranational these approaches is given separate discussion and
points of view. In the former case, we are dealing illustration. A focus is provided by the branch of
with the difficulties posed by the existence of so sociolinguistics known as 'language planning' - an
many languages in the world, and the solutions area that has attracted increasing attention in
that have been proposed to alleviate them. In the recent years.
latter case, we need to consider the consequences Within a language, the co-existence of many
for mutual understanding that stem from the exis- specialized varieties presents a further area of
tence of so many specialized varieties within a lan- enquiry. Fields such as science, medicine, law, reli-

guage. gion, and mass communications have developed

We begin by considering the way foreign lan- styles of language that require careful study if they
guages can act as a barrier to international commu- are to be understood. Popular attitudes to specia-
nication and the various methods that have been lized language also need to be taken into account
proposed to reduce or eliminate the problem. One - notably, in the work of the various campaigns
possible solution lies in and interpreting
translating for 'plain English'. We examine the background
— a field whose future will be much affected by to several of these varieties, illustrating the kinds
progress in computer applications. Another is to of linguistic features they display. Part x then con-
create an international auxiliary language (such as cludes by considering the impact of the 'informa-
Esperanto) or to simplify an existing language for tion explosion', which presents us daily with a mass
international use (such as Basic English). A third of linguistic data, only a fraction of which can be
solution promote the development of an exist-
is to assimilated, and which, as it accumulates, makes
ing language as a world language — something that for increasing difficulties of organization and retrie-
currently seems to be happening to English. A val.

The linguistic variety in the modern world can be seen on

sale daily at any international news-stand.

je^-uj (JiijJl! ^wl l'J

{-. oui_ f^ji > ^sv add uiiiU j*i

. «JUj- >* ijUJ J »j*lill JJ j-4l J-oj JjJI

UULI Mi JjUj J-^JjJ 1 jSI j ijJJJJ 1 JJJJJI

. ij^jjUi ui)l Cj\j*iyt lfjj^J.1 jili . aljljill

.- -ni l u 1 -• . ,
.-t- .an , » m 1 .... .11

j 4jLk^*i>3 4JUI jnSl *i«j jijhiiiloll _^JjJ»JI

Lijll'i. tut V4A4 J-ijU jtt O j'UaJI -A-** J i ... 1

tJL_li
*' J*s~J •*>- -rw" j*' > .'* ' " •8-'
—
> ''jj
ji»
»j«, J j J' /*-**' JJ*- *-**» J*
1

fcjl*"- • '

Cry*I ijjyu-Jl ,

- -/- J z-1' -'jj^"

,^J -*. j -*,., J J J ^-' i^t— V4AJ' -«—
J*
^-'
1

W*.>juu
1- j,»'
*--*
w— _ -— ' -r-

J*.j toi ^/t' >i*J'

JJ- ^— -^.J 1

f -C'l-
*fj» .

'') J>- j"»j« •*• >>

JJ- J*t J-jii- -.- w*J 1

Ji ----- -~ ,- -^«JJ j
56 The language barrier

The discovery that language can be a barrier to EEC Babel

communication is quickly made by all who travel, The EEC building in Brus-
study, govern, or sell. Whether the activity is tour- sels. Nowhere does the for-
eign language barrier exist
ism, research, government, policing, business, or
so markedly as in the offices
data dissemination, the lack of a common language
of the European Economic
can severely impede progress and can halt it al- Community. The member
together. 'Common language' here usually means states use nine official lan-
a foreign language; but the same point applies in guages in their work, and this
presents a massive transla-
principle to any encounter with unfamiliar dialects
tion problem involving 72
or styles within a single language. 'They don't talk different translation
the same language' has a major metaphorical situations. In 1 980. a third of
meaning alongside its literal one (§63). allCommission staff, and
Although communication problems of this kind half of all European
Parliament staff, were fully
must happen thousands of times each day, very
engaged in translation and
few become public knowledge. Publicity comes interpreting, using up over
only when a failure to communicate has major con- been published only in journals written in Euro- 50% of the EEC's total
sequences, such as strikes, lost orders, legal pro- pean languages - other than English. (After D. A. administrative expenditure.

blems, or fatal accidents — even, at times, war. A E. Shephard, 1973.)

reported instance of communication failure took
place in 1970, when several Americans ate a species
of poisonousmushroom. No remedy was known,
and two of the people died within days. A radio
report of the case was heard by a chemist who
knew of a treatment that had been successfully used
in 1959 and published in 1963. Why had the
American doctors not heard of it, seven years later?
Presumably because the report of the treatment had
Several comparable cases have been reported.
But isolated examples do not give an impression
of the size of the problem — something that can
come only from studies of the use or avoidance
of foreign-language materials and contacts in dif-
ferent communicative situations. In the English-
speaking world, for example, surveys of
scientific
books and documents consulted in libraries and
other information agencies have shown that very

Critical languages
In 1985, the U.S. Depart- Azerbaijani French Korean Norwegian Swedish
ment of Education pub- Bahasa Indonesia Fulani Kpelle Nyanja Tagalog
lished a list of 169 lan- Balinese Ga Krio Oriya Tajik
guages which the U.S. Baluchi Ganda Kumauni Oromo Tamil
Government considered to Bamileke Gbaya Kurdish Panjabi Tatar
be critical', in the sense Bashkir Georgian Lahnda Papiamento Telugu
thatknowledge of them Basa (Kru) German Lamani Pashto Temen
would promote important Belorussian Greek (Modern) Latvian Persian Thai-Lao
scientific research or se- Bemba Guaranf Lithuanian Polish Tibetan
curity interests of a national Bengali Gujarati Luba Polynesian Tigrinya
or economic kind. The aim, Berber Haitian Creole Macedonian Portuguese Tiv
supported by special fund- Bhojpuri Hausa Madurese Quechua Toba Batak
ing, was to help counter Bikol Hebrew (Modern) Rappang (Buginese)
Maithili Tsonga
what the Washington Post Bulgarian Hindi Malagasy Romanian Tungus (Evenki,
called the nation's 'lan- Burmese Hmong Malayalam Romany Lamut)
guage illiteracy'. (Spellings Buryat Hungarian Manchu Rundi Turkish
have been made to con- Cambodian Iban (Malay) Mandekan (Bambara) Russian Turkmen
form to those used in Ap- Catalan Icelandic Manipuri Rwanda Tuvinian
pendix III.) Chinese Igbo Marathi Sango Uighur
Chuvash llocano Maya Santali Ukrainian
Ciokwe Irish Mende Serbo-Croatian Urdu
Czech Italian Minangkabau Shona Uzbek
Achinese Danish Japanese Mixtec Sindhi Vietnamese
Acholi Dari (Afghan Persian) Javanese Mongolian Sinhalese Visayan (Sebuano)
Afrikaans Dinka Kamba Mordvin Slovak Wolof
Akan Dutch Kannada More Slovene Yakut
Albanian Efik Kanuri Mundari-Ho Somali Yao
Amharic Eskimo Kashmiri Nahuatl Songhai Yiddish
Arabic Estonian Kazakh Neo-Melanesian Sotho Yoruba
Armenian Ewe-Fon Kikuyu Nepali Spanish Yucatec
Assamese Fijian Kirghiz Newari Sundanese Zapotec
Aymara Finnish Kongo Ngala (Losengo) Swahili Zulu-Xhosa

342 •
X LANGUAGE IN THE WORLD
little foreign-language material is ever consulted.
In one study, an analysis of over 60,000 British
Library requests in the field of science and tech-
nology showed that only 13% were for foreign-
language periodicals (C. A. Bower, 1976). Studies
of the sources cited in publications ('citation analy-
sis') leads to a similar conclusion: the use of for-
eign-language sources is often found to be as little
as 10%. Likewise, in the non-English-speaking
world, there is also a marked reliance on native-
language material - though here the special
influence of English must be considered (sometimes
accounting for over half the requests made or
sources used (§59)).
more readily appreciated that marketing efforts can
be delayed, damaged, or disrupted by a failure to
take account of the linguistic needs of the customer,
or to look after one's own linguistic interests
abroad (in such areas as patenting and trade-mark
control).
The changes in awareness have been most
marked in English-speaking countries, where the
realization has gradually dawned that by no means An advanced class of foreign
everyone in the world knows English well enough students at an English lan-
to negotiate in it (§59), and that this is especially guage school in London.
a problem when English is not an official language There are over 600 such
schools in Britain - though
of public administration, as in most parts of the
less than a third of these are
Far East, the Soviet Union, Eastern Europe, the officially recognized by a pro-

There are several ways of getting around the for- Arabic world, Latin America, and French-speaking fessional body. In the early
1 980s, they were dealing
eign-language barrier, but none is simple, nor has Africa (p. 357). Even in cases where foreign cus-
with over 20,000 students
any as yet been entirely successful. tomers can speak English quite well, it is often for-
each year, and generating
gotten that they may not be able to understand an annual turnover of £300
1. Increase the number and availability of translat-
it to the required level — bearing in mind the re- million. Many schools all
ing and interpreting services (§57).
gional and social variation which permeates speech over the world now put on
2. Develop an auxiliary language that everyone courses geared to the spe-
(Part n) and which can cause major problems of
will understand (§58). cial needs of specific groups
listening comprehension. In securing understand-
3. Develop an existing language as a world lan- of foreigners, such as doc-
ing, how 'we' speak to 'them' is just as important, tors, bankers, lawyers, or in-
guage that everyone will understand (§59).
it appears, as how 'they' speak to 'us'. dustrialists.
4. Provide increased motivation and opportunity
to learn foreign languages (§§60, 62).

THE BUSINESS WORLD

The language barrier presents itself in stark form
to firms who wish to market their products in other
countries. British industry in particular has in
recent decades often been criticized for its linguistic
insularity — assumption that foreign buyers
for its

will be happy to communicate in English, and that

awareness of other languages is therefore not a
priority. In the 1960s, over two-thirds of British
firms dealing with non-English-speaking customers
were using English for outgoing correspondence;
many had their sales literature only in English; and
as many as 40% employed no-one qualified to com-
municate in the customers' languages. A similar
problem was identified in other English-speaking
countries, notably the USA, Australia, and New
Zealand. And non-English-speaking countries were
by no means exempt — though the widespread use
of English as an alternative language made them
less open to the charge of insularity.
The criticism and publicity given to this problem
since the 1960s seems to have greatly improved Expolangues This annual
the situation. Industrial training schemes have pro- linguistic 'motor show', with
around 500 languages repre-
moted an increase in linguistic and cultural aware-
sented, has been held in
ness - for example, programmes of industrial visits Paris since 1 983. The exhi-
abroad, temporary exchanges of junior managers, bits reflect ongoing lan-

or the appointment of overseas agents. Many firms guage-related activities in

now have their own translation service or interna- commerce, technology, cul-
ture, education, publishing,
tional telephone service; to take just one example
and several other fields. In
from Britain, Rowntree Mackintosh now publish 1985, the exhibition attracted
their documents in six languages (English, French, over 40,000 visitors.
German, Dutch, Italian, and Xhosa). Some firms
run part-time courses in the language with which
they are most involved; some produce their own
glossaries of technical terms, to ensure consistency
when material is being translated. It is now much

56 THh LANGUAGE BARRIKR •

343
57 Translating and interpreting

When people are faced with a foreign-language statement, all to do with what standards of 'equiva- Levels of
barrier, the usual way round it is to find someone lence' should be expected and accepted. translation
to interpret or translate for them. The term 'transla- Exact equivalence is of course impossible: no • Word-for-word Each word
tion' is all tasks where
the neutral term used for translator could provide a translation that was a (or occasionally morpheme,
the meaning of expressions one language (the
in perfect parallel to the source text, in such respects p. 90) in the source language
translated by a word (or
as rhythm, sound symbolism (§30), puns, and cul-
is
'source' language) is turned into the meaning of
morpheme) in the target lan-
another (the 'target' language), whether the tural allusions. Such a parallel is not even possible
guage. The result often
medium is spoken, written, or signed. In specific when paraphrasing within a single language: there makes no sense, especially
professional contexts, however, a distinction is is always some loss of information. when idiomatic constructions
drawn between people who work with the spoken On the other hand, there are many kinds of in- are used:

or signed language (interpreters), and those who exact equivalence, any of which can be successful It's raining cats and dogs.
work with the written language (translators) There . at a certain level of practical functioning. It there- II est pleuvant chats et

when chiens.
are certain tasks that blur this distinction, as fore follows that there is no such thing as a 'best'
source speech turned into target writing (for
is translation. The success of a translation depends • Literal translation The
was made, which linguistic structure of the
example, in monitoring foreign-language broad- on the purpose for which it in
source text is followed, but is
casts, or in writing sub-titles for foreign films). But turn reflects the needs of the people for whom it
normalized according to the
usually the two roles are seen as quite distinct, and was made. An inelegant, rough-and-ready transla- rules of the target language:
it is unusual to find one person who is equally tion of a letter can suffice to inform a firm of the
It's raining cats and dogs.
happy with both occupations. Some writers on nature of an enquiry. A translation of a scientific
translation, indeed, consider the interpreting task meaning, but II pleut des chats et des
article requires careful attention to
chiens.
to be more suitable for extrovert personalities, and little attention to aesthetic form. The provision of
• Free translation The
the translating task for introverts! a dubbed film script will warrant scrupulous care
linguistic structure of the
over the synchronization of lip movements, often source language is ignored,
at the expense of content (p. 392). Literary work and an equivalent is found
requires a sensitive consideration of form as well based on the meaning it con-
Translating
as content, and may prompt several translations, veys:

It is sometimes said that there is no task more com- each of which emphasizes a different aspect of the It's raining cats and dogs.
plex than translation — a claim that can be readily original. It is easy to see that what might be 'best' II pleut a verse.

believed when all the variables involved are taken

into account. Translators not only need to know
their source language well; they must also have
a thorough understanding of the field of knowledge
covered by the source text, and of any social, cul-
tural, or emotional connotations that need to be
specified in the target language if the intended effect
NO BILL
is to be conveyed. The same special awareness

needs to be present for the target language, so that

points of special phrasing, contemporary fashions
POSTING
or taboos in expression, local (e.g. regional) expec-
tations, and so on, can all be taken into account.
On the whole, translators work
tongue (or language of habitual use), to ensure a
into their mother Ziehen Verboten
result that sounds as natural as possible - though
some translators have argued that, for certain types
of text (e.g. scientific material) where translation
accuracy
more
is more crucial than naturalness,
sense for translators to be more
it makes

fluent in the
^
source language.

THE PROBLEM OF TRANSLATION

The aim of translation is to provide semantic equi-
valence between source and target language. This
is what makes translation different from other
kinds of linguistic activity, such as adapting, precis
writing, and abstracting. However, there are many
problems hidden within this apparently simple Free translation in practice

344 X LANGUAGE IN THE WORLD

for one set of circumstances may be entirely unsuit- and legal rights undervalued. Some countries view The barrier in
able for another. translation as a menial, clerical task, and pay their
operation
Several different kinds of translation have been translators accordingly. Others (such as the Japa-
The translation literature is
proposed, to allow for this range of possibilities. nese) regard it as a major intellectual discipline in
full of anecdotes about errors
In a pragmatic translation, the emphasis is entirely its own right. The question of status is currently which illustrate the foreign-
on accuracy and knowledge of the subject, as much debated, especially in Europe, where demand language barrier in operation
required for instructional manuals and much scien- for translators is rocketing, especially in relation (§ 56).Some errors are
simply funny; others can pro-
tificresearch (§21). In an aesthetic translation, to the EEC.
voke a diplomatic incident.
important for literary material, the focus is on pre- Since the 19th century the important role of the • L 'Afrique n '6rige plus des
serving the emotional as well as the cognitive con- professional translator has come increasingly to be autels aux dieux ('Africa no
tent of the work, and on maintaining some level recognized. Some hold full-time jobs in translation longer erects altars to the
of stylistic equivalence (§12). Ethnographic or agencies or in government or commercial organiza- gods'), said one UN dele-
gate. The sentence was mis-
sociolinguistic translations aim to pay full attention tions, where they provide an in-house service; but
heard as . . . hdtels odieux
to the cultural backgrounds of the authors and the there is also a large cadre of free-lance translators, and translated as 'Africa no
recipients, and to take into account differences usually working from home. The field now has its longer builds horrible hotels.'
between source and target language, as when own training courses, examinations, career struc- • During a television inter-
view in the United States,
Christian religious traditions based in the Middle ture,and professional organizations — such as the
Soviet premier Khrushchev
East are 'translated' into the cultural norms of Cen- American Translators' Association, the Transla- was told he was 'barking up
tral Africa or modern-day America (§§9-10). And tors'Guild of the (British) Institute of Linguists, the wrong tree'. However,
there are various kinds of linguistic translation, and the Federation Internationale des Traducteurs. this was translated into Rus-
sian as 'baying like a hound'
where the aim is to convey the structural flavour The number of translations made is certainly on
- a highly insulting expres-
of the original text, often in a quite literal manner, the increase, fuelled by a growing number of specia- sion.
emphasizing such features as archaisms, dialecti- lized multilingual publications (such as journals • Many problems occur dur-
cisms, and levels of formality (§§8,10). Most trans- that publish editions in more than one language ing Bible translation. In a

lations, of course, are mixtures of these theoretical or provide issues specially devoted to translations translation intoan Indian lan-
guage America, ass
of Latin
types, reflecting the complex
reality of language of foreign scientific material). As a consequence,
was translated as 'a small
in use, where 'pure' varieties are conspicuously ab- several central organizations have now developed long-eared animal'. The
sent. to coordinate information about the availability of effect was to suggest that

translations and to facilitate their accessibility, Jesus entered Jerusalem

riding on something which
PROFESSIONALISM once they are made - notably the International
closely resembled a rabbit.
Translators aim to produce a text that is as faithful Translations Centre at Delft (The Netherlands) and • In tone languages (§ 29), it
to the original as circumstances require or permit, the National Translations Center in Chicago. In is almost impossible to adapt

and yet that reads as if it were written originally this domain, of course, the advent of computa- the words to a western me-
in the target language. They aim to be 'invisible tional techniques of information storage and retrie- lody and preserve the mean-

- ing. In one Latin American

people' transferring content without drawing val has been a blessing - although one that is not
tone language, as a conse-
attention to the considerable artistic and technical yet as widely shared as it might be. In a 1971 study, quence, the missionaries
skills involved in the process. The complexity of over 90% of a sample of academic staff had never found that a sung translation
the task is apparent, but its importance is often used any translation indexes; and other reports about 'sinners' was in fact
about 'fat people'.
underestimated, and its practitioners' social status indicate that perhaps as many as 80% of scientists
• The slogan 'Come alive
are not even aware of their existence (J. A. Large, with Pepsi' was once trans-
1983). Chinese news-
lated in a
How fast can translators work? Great claims have been made for translation. It paper (in Taiwan) as 'Pepsi
The only satisfactory answer is: it depends - upon has been called the key to international understand- brings your ancestors back
such factors as the translators' experience, their from the grave'!
ing. The Japanese see it as a key to learning. West-
familiarity with the subject matter, and whether they
dictate the translation or have to type themselves.
it
ern Europe, it has been said, 'owes its civilization
The difficulty of the text is a crucial a translator
factor: to translators' (L. G. Kelly, 1979). It is all probably
may be able to achieve 1 ,000 words per hour for so.
popular writing, but only 400 for technical material. The
linguistic relationshipbetween source and target Language ol .anauaaeof citations %
language is also relevant: languages which share publication English German French Other
Language sources Scientists tend to

structural and cultural patterns will be easier to relate

use material published in their own lan-
PSYCHOLOGY
than those which are widely different. And the direction guage before they use foreign-language
English 92 5 52 07 1.6 sources. This analysis of citations is taken
of translationcan also affect speed: some translators
are almost twice as fast in working from language A German 27 91 1 62 00 from an early study of English, French, and
into B than from B into A. French 25 6 74 64 30
German journals in three fields during
Assuming a continuous text, with familiar subject 1952. German authors cited German
CHEMISTRY sources almost as much as English
matter, self-typing, and an aim of producing a
English 79 4 11 9 47 40 authors cited English sources. French
'polished' translation, estimates of translator output
vary from between 2,000 and 8,000 words a day, with German 22 8 64 62 70 authors used a wider range of sources -
most people producing about 3,000. Estimates of the perhaps because there are fewer scientific
French 36 7 27 9 29 1 6.3
publications in this language - but they still
gain in speed from using machine translation (p. 350)
PHYSICS cited French work more frequently than did
are not yet clear: the need for human editing of the
English 86.2 59 20 59 either of the other groups. Similar results
machine-translated text has to be taken into account,
and this can still be considerable at present. German 35 3 58 1 1.4 52 have since been found in citation analyses
of several other languages. (After C. M.
French 49 8 153 30 7 42
Louttit, 1957.)

57 TRANSLATING AND INTERPRETING 345

 . .

THREE WAYS OF TESTING THE Transliteration

QUALITY OF A TRANSLATION
• In back-translation, one translates a text from When the source language is written in a different Transliterating
language A into language B; a different translator script from the target language (§33), it is often Russian
then turns the B text back into A, and the resulting necessary to provide a transliteration of an original One system of English equi-
A text is compared with the original A text. If the word, rather than a translation - something com- valents for Russian letters.
Over a third have alterna-
texts are virtually identical, it is strong evidence monly done with the names of people, places, insti-
tives.
that the original translation was of high quality tutions, and inventions. Here, each character of
(though not incontrovertible evidence, because the the source language is converted into a character
Russian English
second translator might have improved upon the of the target language; for example, Russian alphabet transliteration
work of the first in the reverse process). cnyTHHK 'companion, satellite' becomes sputnik. a a
• In knowledge testing, speakers of language B Transliteration needs to be distinguished from 6 b
are tested about the content of the translation (e.g. transcription, in which the sounds of the source B V
using a questionnaire), and the same questions are words are conveyed by letters in the target lan- r
9
put to speakers of A. If the results correspond, the guage. For example, an English transcription of .]
d
translation must be efficient. Soviet premier Gorbachev's name would have to e(e) e(e)
• In performance testing, speakers of language be Gorbachoff, to reflect the way it is pronounced )K z or zh
B are asked to carry out actions based on the text in Russian. This approach is often unavoidable 3 z
(e.g. in a repair manual), as are speakers of A. The with languages that use partial alphabetic scripts U j

results can then be compared to determine transla- (e.g. Arabic, p. 202), where transliteration would fi j or\
tion efficiency. This is a very time-consuming pro- be very difficult, or logographic scripts (e.g. Japan- K k
cess, however, and requires expert supervision if ese kanji, p. 200), where it would be impossible. -1

clear results are to be obtained. Both approaches have their problems. With M m
transcriptions, the target equivalents are likely to H n
A good back-translation . .
differ when the words are converted into different
Original: Leaks occurring beyond relief valve could languages, as in English Tchaikovsky, Dutch Tsjai- n P
cause some indication of low oil pressure. kowskij, Hungarian Csajkovszkij. With transliter- P r

Back-translation: If oil is leaking at the outside of ations, there is often the problem of there being c s
the pressure relief valve, it can activate the warning insufficient symbols in the target language (so that i
t

of oil low pressure. diacritics have to be added) or too many symbols y u

(in which case arbitrary choices have to be made). * f

. . . and a bad one Arbitrariness is most noticeable when there is no X h orkh

Original: Troubleshooting precautions. close correspondence between the sounds of the Ll c or ts
Back-translation: Preventions while repairing. (R. source and target languages. 4 torch
W.Brislin, 1976, p. 10.) As a result, there are often several transliteration HI s or sh
systems available, as has happened with Chinese uu si or shch
(p. 312) and Russian (where, for example, the fami- b "or"
liar form Khrushchev could appear as Xruscev, or bl
y
in several other ways). In the absence of an interna- b 'or'
tionally agreed scheme, it is often very difficult to 3 e ore
Twain's back-translation trace terms and names in international indexes. A K) juoryu
Mark Twain once complained that his sketch The Russian name beginning with a might be transli- JI ja orya
Celebrated Jumping Frog of Calaveras County' had terated into ia, ja or ya, with major retrieval pro-
been unsuccessful in France because of a bad
blems unless the conversion system is known.
translation. He therefore carried out his own back-
translation of the French text to prove his point.

Twain's original
"'Now, if you're ready, set him alongside of Dan'l, with Only a name . .

his forepaws just even with DanTs, and give the

I'll
Transliteration makes
word." Then he says, "One - two - three - git!" and pecialsense with foreign
him and the feller touched up the frogs from behind, personal or brand names,
and the new frog hopped off lively, but Dan'l give a where unfortunate conse-
heave, and hysted up his shoulders - so - like a quences can result if a
Frenchman, but it warn't no use.' translation is used. For ex-
ample, it would be risky to
Twain's back-translation translate the name of the
"Now if you be ready, put him all against Daniel, with Britishchain of chemists.
their before-feet upon the same line, and give the
I
Boots, into some equiva-
signal" - then he added: One, two, three - advance!" lent word for footwear in a
Him and the individual touched their frogs by behind, foreign language, in case
and the frog new put to jump smartly, but Daniel the connotations were
himself lifted ponderously, exalted the shoulders thus, harmful. The same risk ap-
like a Frenchman - to what good?' plies in reverse,as when
we encounter iceberg' as
es- the name for certain tex-
tiles, or east wind' as the
name of farm machinery -
both examples of transla-
tions from Chinese.
The problem is further
complicated when western
names are turned into logo-
graphic languages (p. 200).
name has to be gi-
First the
ven a phonetic equivalent
in the language. Crystal, for
example, has been turned
into Japanese as Kuri-su-
taru. Each syllable, how-
ever, can be rendered by
several characters, each
with a different meaning.
Several translations' of the
name are therefore pos-
sible - some of which might
be flattering, others insult-
ing. (One meaning of Crys-
tal, it seems, is chestnut-
celebration-barrel!) There
are dictionaries which give
lists of safe' equivalents for
western names - usually
by choosing rare char-
acters, which prompt a
phonetic rather than a se-
mantic interpretation.

346 • X LANGUAGE IN THE WORLD

 . . .

TERM BANKS
One of the most urgent contemporary needs is the DO TERP!
international unification of terminology. people If % POOi -DLL U-26
are not to be perpetually at cross purposes, the PRESS L FOR TERMINOLOGY OR X FOR RBBREUIfiTION
terms used in instructional manuals, codes of prac- *L
tice, scientific research, government meetings, and TYPE CODE OF SOURCE LRNGURGf
cc zzzuru
many other situations need to be correlated and DG GERMAN M BfifilSH EG ENGLISH I ".lIi'jI;
I

standardized. Scientific research, especially in IT ITRLIflN NG DUTCH PT PORTUGUESE

rapidly developing areas such as medicine and com- *EG
puting, is particularly at risk, where work could TYPE CODE(S) OF TRRGET LftHGUAQEvS) WITH SINGLE SPACE BETWEEN
easily be duplicated. Several efforts have therefore (FOR EXftHPLE: DG NG) OR A ^OR RNY LRNGUBGES
been made to organize data banks of terminology
in various fields, using computational techniques CnilPTF
•.'OjrlC
!

i. I
LI
ni
H
'J
I i
.^
u n >}
Ci C
l.
=
.
Cf;
i_'j

of information storage and retrieval. For example, TRRGET LRNGuRGE(S) :DG IT FG NG li'i SP PT
the EEC term bank, eurodicautom (European SUBJECT CODE :

Automatic Dictionary), is now publicly accessible PRESS Q OR ANOTHER COHnfiND

on-line via Euronet, and contains over 180,000
items in the official languages of the Community. .'
t rt * '.•
u tf «u15 1 j UN
Each term or abbreviation is listed along with a
contextual example, equivalent items in other EEC
languages, a definition, and bibliographical refer- := Tj TC - 7 S ii-V
Bffl iii
H
ences.
EC4 ECE
inflation
price inflation is OST LIKELY TO OCCUR «HEW DEKfiND
increases while the LABOUR SUPPLY IS TIGHT AND THE
False friends industrial capacity IS FULLY UTILIZES. .WHEN SOURCES OF
Words same in two languages often do
that look the supply dry up.
not mean same thing. They are known as faux amis
the r b inflation
('false friends'). Here are some French-English
exces de pouvgir d'achat 0u exces des hotens de paiehen
examples (* = wrong meaning).
Oh la confond souvent. .avec la simple hausse bes fri". .

abusif 'abusive/ incorrect, excessive Or CELLE-CI EST LA CONSEeUENCE de l' INFLATION ET NGN DE
achever 'achieve/finish (off)
l'inflation elle-meme.
avertissement "advertisement/warning
PRESS C TO CONTINUE OR GIUE fiNOTHER COSRRND
bande *band/gang
carpette * carpet/ rug The result of an on-line EEC languages were other gives various details of
demander 'demand/ request search for the term inflation requested as targets.) sources, and the last occa-
eventuel 'eventual/ possible sion when the record was
in eurodicautom. Several 3. Is there a subject code,
fastidieux ' fastidious/tiresome
steps are involved. e.g. medicine, nuclear updated (date 750220, i.e.
idiome "idiom/ language (of a group) 1 Does the enquirer require science? (None, in this 20 February 1 975). There is
.

incoherent 'incoherent/inconsistent a term or an abbreviation (L also a reliability code (CF)

case.)
inconvenient 'inconvenient/drawback or X)? 4. The term is then typed in, using a scale from to 5
information 'information/news 2. Which source and target and any listings are printed
lard *
lard/ bacon
languages are involved? (In out (in this case, only an
libeller *libel/make out a cheque the example, the source Ian- English and a French listing
partition 'partition/ musical score
guage was English, and all are available). The printout
petulant 'petulant/lively
phrase 'phrase/sentence
pourpre "purple/ crimson
pretendre "pretend/claim
Translation shifts
prune "prune/plum
resumer "resume/summarize The French definite article, 'a' in // a lajambe cassee shows that le is equivalent

rogue "roguish/arrogant le/la/l'/les is usually He has a broken leg' and to the in only 64.6% of all

sensible "sensible/ sensitive thought of as the equiva- /'can become zero in I'a- cases. Du of the' is trans-

sommaire "summary/ contents list lent of the English definite mour'love'. In a study of lated by zero (no article at
starter (car) "starter/choke article the. However, there French articles found in all) more than half the time.

sympathique "sympathetic/agreeable are many exceptions to this French texts with English Un a' has the best rating,

truculent "truculent/ realistic rule, where the translation translations, the various with 70%. (After R. Hud-
verbe "verb/ word 'shifts' from one form to possible equivalences dleston, in J. C. Catford,
veste "vest/ jacket another. La is translated by were worked out. The table 1965, p. 81.)

vivace "vivacious/hard-wearing
(P. Thody & H. Evans, 1 985.) French English

And a few others . .

zero the some a (other)

German also = therefore zero 67.7 6.1 0.3 11.2 4.6

Spanish constipado = having a head cold le 14.2 64.6 — 2.4 18.9
Danish 0/= beer du 51.3 9.5 11.0 5.9 22.4
Italian caldo = warm
un 6.7 5.8 2.2 70.2 15.1
Polish karawan = funeral procession

57 TRANSLATING AND INTERPRETING .

347
 . . . .

BEST TRANSLATIONS? indolent companions on the voyage, follow the ship Translating names
A great deal can be learned about the art and craft gliding over the bitter depths.
Mari, Mary, Marie, Marenka,
(I. F. Fmlay)
of translating by comparing cases where different Marinka, Marienka, Marus-
people have translated the same text. The following Sometimes, sailors to amuse themselves ka, Mara, Mana, Maruse,

versions of the opening verse of Baudelaire's L'Al- Marka, Marena . .

catch albatrosses, great birds of the sea,

batros shows the different attention paid to form which as companions follow indolently
The problem facing the liter-
and content in the varying treatment of rhyme, the vessel gliding over bitter gulfs. ary translator is well illus-
rhythm, word order, and lexical choice. (C. F. Madnryre trated by this list of names,
men of loafing crews
which are just someof the
Souvent, pour s'amuser, les hommes d' equipage Sometimes for sport the
ways Mary' in
of translating
Prennent des albatros, vastes oiseaux des mers, Snare the great albatrosses of the deep,
Czech, each expressing its
Qui suivent, indolents compagnons de voyage, The indolent companions of their cruise own mood and level of inti-
Le navire glissant sur les gouffres amers. As through the bitter vastitudes they sweep. macy. English has no such
'R.Campbell; range of expressions; the
Often to amuse themselves, the men of the crew trap nearest. Marie and Maria,
albatrosses, the great sea-birds, that follow the ship
In sport a vessel's crew will often take
are thought of as different
slipping over the bitter deeps, like idle travelling The mighty albatross, who on the breeze
names. For everyday trans-
companions. Doth idly sail and follow in the wake lational purposes, it may not
(A. Hartley) Of ships that glide upon the bitter seas. be important convey such
to
A. Conder; exact distinctions in an Eng-
Often, for their amusement, sailors catch albatross, lish version; but in translating
those vast birds of the sea, indolent companion of their Often, as an amusement, crewmen
nuances of this
literary texts,
voyages, that follow the ship gliding across the bitter Catch albatrosses, huge birds of the sea,
kind often need close atten-
depths. Who follow, indolent companions of the voyage, tion.
(F. Scarfe)
The ship gliding over the saltv deeps.
(W. Fowlie,
In order to amuse themselves, the members of the crew
often catch albatrosses, those huge sea-birds which, as (From I. F. Finlay, 1971, pp. 129-32.)

WHEN FOREIGN IS BEST understand them, but the connotations of prestige Some Japanese
Sometimes it pays not to translate, as the business associated with these languages are enough to war- car names
world has long known. Sales can benefit if a pro- rant their use. The purpose of the language (§4) Daihatsu
duct is given a foreign name. In 1960 a Finnish is not to communicate ideas, but to appeal to the Domino
firm distributed tinned coffee for the home market sensibilities of the Japanese viewer, who the manu- Charade
using Finnish labels. Sales were poor. The firm then facturers believe is much influenced by the values Charm ant
had new labels made with a text in English on the of modern cosmopolitan society.
Nissan
same tins, and sales rocketed. Similarly, English Bluebird
marketing firms and other businesses make use of Cherry
foreign languages to convey special effects - such Foreign elements in Japanese Laurel
as the use of French for the names of restaurants, commercials Micra
Patrol
night-clubs, and perfumes. In one page of a British English
Prairie
telephone directory, under 'Restaurants', nearly Terrific,everybody, new, life, now, healthy, power, big,
Silvia
sale, open, happy, nice, beautiful, night, extra.
half the names were in a language other than Eng- Stanza
lish - La Bella Napoli, Le Patron, Les Deux R, French Sunny
P/a/s/r pleasure', image image', cafe coffee Violet
Les Quatre Saisons, Mamma Mia, Maison .

printemps spring', cnoco/af chocolate', accessoire

Romano, Que Pasa, Rendez-vous, Roma, Santa accessory'. Toyota
Lucia, Shangri-La . . Camry
German
The culture that seems to make most use of for- Auslese selection', schick chic . schon beautiful',
Carina
eign languages as a part of business enterprise is Celica
Sahne cream', Wagen car
Corolla
Japanese. Here, a wide variety of foreign names Italian Corona
is used, depending on the particular quality of the Carina sweetheart', manifesto manifest', buongiomo Cressida
product the manufacturer wishes to stress. In the good morning'. Starlet
field of car names, for example, English is used Spanish Tercel

in order to convey an impression of good quality Ole hey domingo Sunday', bonita
, pretty'.

and reliability- (e.g. 'Crown'). If elegance is to be Combinations words are also common, such as high
of
stressed, a French name is chosen (e.g. 'Ballade'). quality, happy smoking time, light and smooth, quick
A and overnight service, la mesure d'&e'gance the
sports car often has an Italian name (e.g.
standard of elegance', nouvel colons new shade
'Leone').
Whole sentences may be used, such as Je suis une
The linguistic effects are most noticeable in tele- femme quiaime la vie am a woman who loves life'
I

vision commercials, where appropriate American,
French, etc. settings are used along with the foreign
language (without translation). Japan is the only
monolingual country to make frequent use of for-
(spoken by a Japanese woman), We want you to win,
Get action on your car. There may even be
combinations of English and French (e.g. bonshop,
sante soft) or a foreign language plus Japanese (e.g.
soft kapuseru soft capsule auto wakkusu auto wax',
,

eign languages (primarily English and French) in p. 195). (After H. Haarman, 1984.)
its commercials. The viewer usuallv does not

348 • X LANGUAGE IN THE WORLD

Interpreting
Interpreting is today widely known from its use How it is all done is not at all clear. That it The ear-voice
in international political life. When senior ministers is done at all is a source of some wonder, given span
from different languagebackgrounds meet, the the often lengthy periods of interpreting required,
When the input language is
television record invariably shows a pair of inter- the confined environment of an interpreting booth, straightforward, the inter-
preters hovering in the background. At major con- the presence of background noise, and the aware- preterwill be a regular 2 or

ferences, such as the United Nations General ness that major decisions may depend upon the 3 seconds behind the
speaker. But when an unex-
Assembly, the presence of headphones is a clear accuracy of the work. Research projects have now
pected textual difficulty
indication that a major linguistic exercise is taking begun to look at these factors — to determine, for emerges, or the speaker
place. In everyday circumstances, too, interpreters example, how far successful interpreting is affected suddenly speeds up, the de-
are frequently needed, especially in cosmopolitan by poor listening conditions, or the speed at which lay may increase, with a con-
sequential 'knock-on' effect
societies formed by new generations of immigrants the source language is spoken. It seems that an
that may take some time to
and Gastarbeiter (§9). Often, the business of law input speed of between 100 and 120 words per resolve. During this period,
courts, hospitals, local health clinics, classrooms, minute is a comfortable rate for interpreting, with when the ear-voice span is
or industrial tribunals cannot be carried on without an upper limit of around 200 w.p.m. But even small several seconds long, the in-
the presence of an interpreter. Given the impor- increases in speed can dramatically affect the accur- terpreter's cognitive linguistic
processing abilities are
tance and frequency of this task, therefore, it is acy of output. In one controlled study, when speeds
under great pressure.
remarkable that so little study has been made of were gradually increased in a series of stages from A common linguistic rea-
what happens when interpreting takes
actually 95 to 164 w.p.m., the ear-voice span also increased son for interpreting delay oc-
place, and of how successful an exercise it is. with each stage, and the amount correctly inter- curs when translating from a
language where the verb oc-
Doubtless the recency of developments in the preted showed a clear decline (D. Gerver, 1969).
curs at the end of the sen-
field partly procedure,
explains this neglect. One Also, as the translating load increases, not only tence (SOV languages, §14)
known as consecutive interpreting, is very old - are there more errors of commission (mistrans- into a language where the
and presumably dates from the Tower of Babel! lations, cases of vagueness replacing precision), verb occurs in the middle
Here, the interpreter translates after the speaker there are also more errors of omission, as words (SVO languages). The inter-
preter often has to wait until
has finished speaking (either in short bursts, or at and segments of meaning are filtered out. These
the speaker gets to the end
the very end of a discourse). This approach is are important findings, given the need for accuracy of the sentence before it is
widely practised in informal situations, as well as in international communication. What is needed possible to translate the in-
in committees and small conferences. In larger and is a more detailed identification of the problem tervening material. The pro-
blem occurs in the case of
more formal settings, however, it has been gener- areas, and of the strategies speakers, listeners, and
translating Japanese into
ally replaced by simultaneous interpreting — a interpreters use to solve them. There is urgent need English, and, notoriously, in
recent development that arose from the availability to expand what has so far been one of the most translating German subordi-
of modern audiological equipment and the advent neglected fields of communication research. nate clauses.
of increased international interaction following the
Second World War.
Of the two procedures, it is the second that has
attracted most interest, because of the complexity
of the task and the remarkable skills required. In
no other context of human communication is
anyone routinely required to listen and speak at
the same time, preserving an exact semantic corres- Linguistic A B C D E
units
pondence between the two modes. Moreover, there
is invariably a delay of a few words between the

stimulus and the response, because of the time it

takes to assimilate what is being said in the source
Interpreter Hears A Hears B Hears C Hears D Hears E
activity
language and to translate it into an acceptable form Speaks A Speaks B Speaks C, D
in the target language. This 'ear—voice span' (p.
147) is usually about 2 or 3 seconds, but it may
Point of
be as much as 10 seconds or so, if the text is com- extra complexity'
plex. The brain has to remember what has just been
said, attend to what is currently being said, and
anticipate the construction of what is about to be
said. As one writer has put it:

As you start a sentence you are taking a leap in the dark,

you are mortgaging your grammatical future; the
original sentence may suddenly be turned in such a way
that your translation of its end cannot easily be
reconciled with your translation of its start. Great
nimbleness is called for to guide the mind through this
syntactical maze, whilst at the same time it is engaged
upon the work of word-translation.
(R. Glemet, in D. Gerver, 1976, p. 168)

57 TRANSLATING AND INTERPRETING •

349
 -

Machine translation ignored the different ways in which languages A pre-edited translatio
structure meaning: word-for-word translation is The below speaks for
extract
The idea of using machines to provide translations often not possible and usually not desirable (p. itself. has been pre-edited
It

between natural languages has been recognized 344). Nor was there any way of distinguishing in Hong Kong on the Chi-

between the different senses of words or deciding nese University Language

since the 1930s, but an appropriate climate for
Translator (cult), a system
development did not arise until the years following whether a group of words was an idiom. Many used for the automatic trans-
the Second World War. At that time, the rise of ambiguities can be resolved only by using an analy- lation of Chinese journal
information theory, the success of advanced code- sis in terms of semantics (§17) or of real-world papers in mathematics and
breaking techniques (p. 58), and the invention of the knowledge, and such analyses were not available physics into English. There
are very few places where
electroniccomputer all indicated that machine trans- at that time. There was evidently a great deal more
further editing is necessary
lation (MT) could be a reality. Warren Weaver, to MTthan 'code breaking'. but this has to be balanced
a founder of the field, caught the optimism in a The dissatisfaction was summarized in a US against the amount of pre-
1947 memorandum: 'One naturally wonders if the report of 1966 by the Automatic Language Process- editing time involved. There
is also the point that the
problem of translation could conceivably be ing Advisory Committee (alpac), which con-
amount of pre-editing will
treated as a problem of cryptography. When I look cluded that human translating more
was faster,
vary, depending on the struc-
at an article in Russian, I say: "This is really written accurate, and less expensive than MT, and that tural similarity between the
in English, but it has been coded in some strange no further support for the latter should be pro- languages.
symbols. I will now proceed to decode."' As a vided. As a consequence, only a minimal amount
A fixed
result, several groups began research programmes of MTresearch was carried on in subsequent years,
rules
set of pre-editing
must therefore be
into MT
during the 1950s, and great claims were either in the USA or in Europe (though continued mulated to enable inexper-
for-

made for the future of the subject (W. Weaver, support was provided in the Soviet Union). ienced and even mono-
1955). lingual people to transform

However, initial results were not encouraging. A NEW MOOD quickly the input into
chine-translatable form. With
ma-
The systems proved to be very limited in the kind The pendulum has begun to swing back again in
this arrangement, post-edit-
of data they could handle. Translations were crude, recent years, following the major intellectual and ing can be kept to a mini-
full of errors, and required so much human post- technological developments of the 1970s in linguis- mum, if not all together elimi-

editing that they proved to be more expensive than tics($65) and computing. A new mood is abroad, nated. Given time and better
programming techniques,
having a human translator carry out the whole task promoted by the promising practical achievements
these pre-editing rules will
in the first place. The main reason was the lack of new commercial MT
projects, by the great gradually be reduced so that
of a sufficiently sophisticated linguistic theory to potential of the new
research programmes in artifi- the computer will eventually

provide a frame of reference for the tasks that MT and by an increased theoretical
cial intelligence, take up this routine work.
needed to undertake. The earliest systems did MT awareness of the translation task which has come Pre-editing can therefore
solve many of the present
little more than look for equivalences between the from progress in linguistics. There is also a greater
linguistic problems that are
words in each language — in effect, they acted as realism concerning what MT
can and cannot do, otherwise dependent on
an automatic bilingual dictionary. After several and a recognition of the need to devise techniques further research in natural
decades of linguistic research, it is easy to see why of human/machine collaboration, in order to get language, computational
linguistics and transforma-
these approaches could not have worked. They the best resultsfrom both.
tion mathematics. In other
ignored the problem posed by the grammatical The main developments have been to provide words, models that are much
dimension of language analysis — the different levels systems of analysis that allow for grammatical and more comprehensive and
of syntactic organization (§16), and the absence semantic complexity. The first steps were in devis- sophisticated than the pre-
of straightforward formal correspondences ing automatic procedures ('algorithms') for parsing sent ones have to be de-
signed. These models may
between units of grammar (such as is illustrated the syntactic structure of a sentence, and for carry-
take years to perfect and, at
by the use of the definite article, p. 347). They also ing out an analysis of word-structure. An automa- present, pre-editing is abso-
lutely essential in order to
achieve the goal. In present
stage of our development,
SYSTRAN very complex sentences can
One of the best-known MT Russian-English jenil , texniceskix sredstv, cow Airport Domodedovo, be translated with the aid of
systems, Systran, was de- An showing
extract follows, dl4 razme5eni4 kotorogo Kennedy's New York Air- pre-editing. A sentence
veloped in the U.S. with par- the kind of output produced trebuets4 territori4, izmer- port). which has a complicated
ticular reference to by Systran, and the kind of 4ema4 v otdel6nyx sluca4x structure can be analysed by
Russian-English translation post-editing required to pro- tys4cami gektarov (na- Revised output the existing program if is it

- for example, it was used
to translate Russian into
English during the Apollo-
Soyuz space project.
also now used by the EEC
fora limited range of pur-
poses (such as abstract
searches) with certain lan-
guages (mainly French).
is unclear, as
future status
it be replaced by
is likely to
a more powerful system
(EUROTRA) in due course.
duce an acceptable transla-
tion. The numbers in the
Russian transliteration have
It is been inserted by the com-
puter to represent Cyrillic
letters not found in the com-
puter's (Latin-based) char-
acter set.
Its
Original
Sovremennyl a3roport
predstavWet sobol slojnyl
kompleks injenernyx sooru-
primermoskovskM a3roport The modern airport is
Domodedovo, H6h-1orkski1 elaborate complex of engi-
a3roport Kennedi). neering structures and tech-
nical devices requiring a
Raw SYSTRAN OUtpUt large territory, which,
A contemporary airport is some cases, measures
the involved complex of en- thousands of hectares (for
gineer constructions and instance, Domodedovo Air-
techniques, for arrange- port in Moscow or Kennedy
ment of which the territory, Airport in New York).
measured sometimes is re- (F. Knowles, 1979, p.
quired by thousands of hec-
tares (for example the Mos-
an broken up into simpler sen-
tences which are then readily
translated by the computer.
Sentences in Chinese are of-
in ten without verbs or subjects
and pre-editing can add the
verbs or subjects so that
these sentences can thus be
analysed and translated.
(S.C. Loh, 1976, p. 108.)
130.)

350 • X LANGUAGE IN THE WORLD

 ' - 5

tic morphological analysis (p. 90) was particularly a highly polished style. If the most basic editing Telephone
necessary to enable the computer to find words steps only are undertaken, the gains in translation translation
in its dictionary memory: cats, for example, would speed can be very great. But the choice, of speed
The automatic translation of
have to have its ending removed, in order to locate vs quality, is not always an easy one to make. telephone conversations, us-
the base form cat. The problem is not very great ing fifth-generation com-
for English, which uses few endings, but it is a THE FUTURE puters, should be a
reality by
ad 2000, according to propo-
major issue for languages that rely more on inflec- It is unlikely that machines will ever replace human sals made by the Japanese
tions. translators; but they can undoubtedly help to take Ministry of Posts and Tele-
Later developments have begun to introduce a great deal of the drudgery out of routine transla- communications in 1985. If
semantic information into the procedure, most tion work, and enable far more material to be pro- Japan were to undertake the
development alone, it was
recently using artificial intelligence techniques to cessed than would otherwise be the case. Firms
estimated to take about 1
simulate human thought processes. If the computer such as Automated Language Processing Systems years at a cost of 400 million
is given enough data on the meaning of words and (alps) or Weidner Communications now have sys- dollars. Japan has therefore

about the context of a sentence, it is argued, it
should be able to work out for itself what analysis
to make in cases where individual words or sen-
tences are ambiguous. The computer, like the
human reader, should be able to look back at the
preceding discourse in order to check its interpre-
tation of a point. The principle is undeniable, but
it has proved extremely difficult so far to write pro-
grams that can handle more than fragments of dis-
course. The advent of special programming
languages, designed to handle the properties of
natural language in a more direct way, will facili-
tems that can process quantities of scientific text suggested joint development
proposals with several coun-
in certain areas at rates of up to 14,000 words
tries, in the interests of sav-
per hour — which, even after editing, produces a ing timeand money, the first
rate of over 1,000 w.p.h., several times faster than agreement (with France) be-
is possible 'by hand'. There is also the rapidly devel- ing reached in 1985.
oping world of 'machine-aided translation' - the
use of computationally organized data banks and
all kinds of peripheral equipment to help transla-
tors in their work. A word processor can save enor-
mously on the production of translation drafts, for
example; and a great deal of time can be saved
if the translator has on-line access to a term bank

tate the task; but it will take many years before (p. 347) to discover the 'best' equivalent for a
the pure research results in routine commercial source-language word (instead of having to engage
applications. in a slow hunt for it through dated dictionaries). MT output These extracts
In the meantime, increasing use is made
being More and more people are finding that the benefits use a German-English sys-
of 'interactive' systems of MT, in which human outweigh the disadvantages, and this in turn adds tem marketed by the Logos
beings pre-edit or postedit the text that the com- to the mood of optimism that pervades current MT Corporation in the Federal
Republic of Germany. The
puter processes. In pre-editing, a natural language debate. At present the MT
world is still quite a original, taken from a con-
source text is rewritten, using a controlled syntax small one, with few research programmes and com- struction contract, is shown
and vocabulary, to produce a version that the com- mercial organizations involved. This situation is along with the raw MT output
puter can handle with relative ease. This procedure likely to have changed dramatically by the end of and the post-edited output.
is practicable only in restricted situations (such as the century.
scientific abstracting), because of the large expendi-

|
ture of time and training required to prepare the
input; but it has produced successful translations
in several cases. In postediting, 'raw' machine-pro-
Treppenreparaw /
jra be
S
„ en
e
nGebaudenw-.e1olgt

^
SE-— —
Hauptemgangsst
duced data in the target language is edited into reparieren .
:I _ ahmen sorgf a««9
*c
an error-free text - also a time-consuming proce- 1

dure, and one that is often tedious and time wasting

(because the human translator may have to dupli-
cate a great deal of the computer's work in order
to eliminate an error). However, there are several
levels of quality possible in postediting, ranging
3 Bss TzerS.«™=rtel
?
s«S ica 1 cm I*

from a basic 'threshold' level of intelligibility to

Zeme

post-edrted
outpUt: 9S as fol-
Rav*WT
-.-work
S* reparSances.epa>a«w'*

enlace step
lows remove cor
Mai" rtli00 , m atlrame.
Careful* from
resulted
re
'
have trnortar .

frames. maaeS which

i

in
"SSi «*-»•'*•
rents and
W««9 3 Scrape ,

, step an«

57 TRANSLATING AND INTERPRETING •

351
 ,

58 Artificial languages

Many people believe that the foreign-language bar- Europe. Most ALs that have been devised in the
rier can be surmounted through the use of an 'artifi- last century belong to this second type. A further
cial' language (AL) — a language which has been division is also sometimes made, within the latter
specially invented to facilitate international com- category, between those ALs that do not allow for
munication. The term 'auxiliary language' is often exceptions (e.g. Esperanto) and those that are
preferred, especially by AL supporters who find closer to existing natural languages, permitting
the connotations of 'artificial' too limiting. In parti- some degree of irregularity (e.g. Occidental).
cular, Esperantists (p. 354) feel that their language
is highly developed in the range of functions it can HISTORICAL BACKGROUND
perform, and is also natural, to the extent that these Several hundred ALs have been recorded, and new
days several children have had it introduced to ones continue to be devised. The desire for a univer-
them as a mother tongue. However, 'auxiliary' is sal language can be traced back to Classical times,
by no means restricted to ALs; it is also used to but the idea began to flourish only in the 17th cen-
refer to natural languages that have been chosen tury, when world exploration brought to light
to aid communication within a special domain (e.g. many new languages, and Latin began to fall from
the use of English or French at international confer- favour as a universal medium. Francis Bacon
ences). 'Artificial' thus continues to be a widely (1561-1626) was one of the first to ask whether An extract from An essay to-
wards a real character and a
used designation. things could be represented by 'real characters' that
philosophical language
ALs are commonly divided into two types would be understood regardless of language, (1668) by Bishop John Wil-
(though several projects in fact display features of instead of the usual letters or sounds. Detailed sys- kins (1614-72). Wilkins's ap-
both). The first type uses an invented set of elements tems were devised by Wilkins, Dalgarno, Leibniz, proach classified all ideas
— often numbers or special symbols — that stand Comenius, and several others. It was an age when and things into 40 basic
types, or genera (e.g. herbs,
for basic concepts; these are then grouped into a symbol systems of all kinds stimulated curiosity, fish, sickness). Each genus
logical (and supposedly universal) classification, especially those that seemed to express concepts was symbolized by a conso-
based on scientific or philosophical principles. directly - music, numerals, ideograms (§33), short- nant and vowel (for speech)
These a priori languages were particularly common hand (p. 206), and cryptography (p. 58). However, and by a written sign. Differ-
ent items within a genus
in the 17th century, but proposals along these lines all the invented systems, though simple, logical,
were then identified through
have continued to be made at intervals ever since. and plausible at the outset, became extremely the use of extra consonants
By contrast, a posteriori languages are based on arbitrary and complicated as they were developed. or vowels or by additional
elements drawn from natural languages — invaria- As the illustrations show, the learner was presented written strokes.

bly from one or more of the languages of Western with a virtually impossible memory load, and a

The Lords Prayer.

An extract from the first universal language scheme,
Francis Lodwick's A Common Writing: whereby two,
although not understanding one the others language,
Hat coba •da ta ril dad, ha bObi to foymta, ha (alba to velcft, ha

yet by the helpe thereof, may communicate their minds tttlbt to vemgtt, my ril dady me rtl dad 10 velpt rctl cXi ril ( poto hat
one to another (1 647). A basic set of radical' words faba vaty, na K> ftfcldytfi letl at hat balgas mcctrta i«eldy«s lal
was selected (e.g. drink), and each assigned a el Vti rex vctlgas rti at oimt to vclco cu, reel bedodlti oil to cttalbo
character. Related words were shown by the addition fti lal vOgaste, nor cXlfalba, oacXl tado, na al tadalct »a ha pittbytf
*
of special marks.
QJ m« to.
The figr.eof the fix forts of NounesSubftani.v: :Af
pellitive, is this,£-a) augmented, as under

For the i.thus.

a. thus,
And
ed to
ipply-
tbe^
YfiT'tyt dttnktt

\fi=>
drihk.
1 3345 67 89 10 11

3. thus, thedruki/i?.
HcXt coba tttf ta ril dad, ha bcXbt to feyratcX ha
right fid.-
4. thus
thus
otihcRa drunkdrd. Our Father who art in Heaven, Thy Name be Hallowed, Thy
j. dx, thus
6. thus drunkewiefje.

^3 dri»kwghouf{„ 13 13 14 15 l6 17 l8 I930 3I 33 33 34 35 36
The 7. Seflion. (alba to TclccXjha tcXlbi io vemgU^mtf ril dady me ril dad, to velpt

The diftin&ionall figne for the proper names (hall be

this, (1) joyncd to the left fide of the Radix, toward
the head thereof, thus. (°Q-%
The figne ofthe feminine gender Anil be this, (tj
Kingdome come, Thy Will be done,
- "

3738393,03133
« »
pf-,
tt
*gj
33
^
34
fo in

3*3637
«
Earth as in Heaven, Give
Ul*. -

38 39 40
- ° W4«
placed as following. rS^O
The fubordinall figne ror the young ones flu 11 he
tCLlOU ril t poto hcXi faba vaty, na Joflrfclditfs lal at ha» balgas
thiSjfg^ placed as following. (efr>)
to us 00 this day our bread expedient and forgive to us our trefpaflcs

352 •
X LANGUAGE IN THE WORLD
rigid classification within which it proved difficult of common vocabulary between the various sys-
to incorporate new knowledge. tems. Most of the proposals have had very short
For such reasons, a priori schemes fell out of lives, but some, such as Esperanto, have achieved

favour at the end of the 17th century. They had an impressive international use.
a revival a century later, with the rise of the 'general Since the turn of the century, an enormous
grammar' movement (p. 84), which aimed to dis- amount of time, energy, money, and ingenuity has
cover universal principles of thought behind the been expended on the invention and dissemination Solresol
variety of grammatical forms in language. There of AL proposals. In the early years they were This remarkable language
was much more at stake here than international usually seen by their supporters as a key to a world was the invention of a
communication: it was felt that a good philosophy of mutual understanding, clearer thinking, and French music master, Jean
Frangois Sudre, at the begin-
of signs, or 'ideology', would help to eliminate peaceful coexistence. The use of a common lan-
ning of the 1 9th century. Sol-
vagueness and ambiguity from language, provide guage does not guarantee peace, however, as is resol, or 'Langue Musicale
a better vehicle for thought, and be a more efficient plain from terrorist activity in many parts of the Universelle', was based on
means of spreading knowledge. But the resurgence world. In recent years, therefore, there has been the principle that the tones of
music, as named in the
of interest was short-lived. By the mid- 19th cen- a tendency to adopt less ambitious goals, and there
seven-note diatonic scale
tury, there was little active support, and since then have been several local successes, with different (do, re, mi, fa, sol, la, si),

the vast majority of AL proposals have been a pos- countries and organizations (especially on the Con- could be used as the ele-
teriori in character. tinent of Europe) being persuaded to introduce an mental syllables of a univer-
In the last quarter of the 19th century, there was AL dimension into aspects of their daily life - in sal language.
Two-note combinations
another flurry of enthusiasm, with several AL pro- hotels, telephone boxes, telegrams, advertisements,
were used for grammatical
posals competing with each other for public sup- and timetables. Even with the most successful of words, e.g. s/'yes', do 'no',
port. The first large-scale movement was Volapiik, these movements, however, very limited progress re and', dore T, domi 'you'.

followed closely by Esperanto, Idiom Neutral, Ido, has been made towards the goal of an internation- Common words used three-
note combinations, e.g. dor-
and several dozen other systems. In 1924, the Inter- ally recognized and universally used auxiliary lan-
edo'time', doremi"day', dor-
national Auxiliary Language Association was guage. efa 'week', doresol 'month',
formed in New York, focussing on the promotion dorela 'year', doresi 'cen-
tury'.Four-note combina-
tions were divided into differ-
ent classes (or 'keys'), each
A CENTURY OF ARTIFICIAL LANGUAGES one being based on a parti-
cular note. 'La', for example,
Language Inventor Date Comment
was used for the field of in-
dustry and commerce. Over
Volapiik ('World Johann Martin Schleyer 1880 8 vowels, 20 consonants; based largely on
9,000 five-note combinations
language') English and German (p. 300). were used for the names of
Esperanto ('Lingvo Ludwig Lazarus 1887 5 vowels, 23 consonants; mainly West animals, vegetables, and mi-
nerals. Semantic opposites
Internacia') Zamenhof European lexicon; Slavonic influence on
were often expressed by re-
syntax and spelling (p. 300). versing the order of syll-
Idiom Neutral V. K. Rosenberger 1902 A former supporter of Volapiik; strongly ables, e.g. misol 'good' vs
solmi 'evil'.
influenced by Romance (p. 301).
The unique feature of this
Latino Sine Flexione Giuseppe Peano 1903 Latin without inflections; vocabulary AL is that it could be played,
(Interlingua) mainly from Latin words. whistled, or sung, as well as
spoken! It became very pop-
Ido Louis de Beaufront or 1907 A modified version of Esperanto (the
ular in the mid-1 9th century,
Louis Couturat name means 'derived from' in Esperanto).
and won several prizes. As
Occidental Edgar von Wahl 1922 Devised for the western world only; with all a priori languages, it

largely based on Romance.

was difficult to learn, and,
with so fewsounds permit-
Novial Otto Jespersen 1928 Mainly Ido vocabulary and Occidental ted, must have sounded
it

grammar. Novial = New + international extremely monotonous.

However, still had some
it
Auxiliary Language.
supporters at the beginning
Interglossa Lancelot Hogben 1943 Published only in draft form. of the present century, and
thus proved to be one of the
Interlingua International Auxiliary 1951 A Romance-based grammar, with a longest surviving artificial
Language Association standardized vocabulary based on the languages.
main Western European languages.
Glosa W. Ashby and R. Clark 1981 A modified version of Interglossa.
Contains a basic 1,000-word vocabulary,
derived from Latin and Greek roots.

And a selection of other projects Latin-Esperanto (1911), Europeo (1914), Nepo

Perio (1904), Lingua Internacional (1905), Eksel- (1915), Horn Idyomo (1921), Espido (1923), Neo
sioro (1906), Ulla (1906), Mondlingvo (1906), (1937), Esperantuisho (1955), Globaqo (1956),
Lingwo Internaciona (Antido) (1907), Mez-Voio Modern Esperanto (1958), Delmondo (1960),
(1908),Romanizat (1908), Romanal (1909), Utoki (1962), Eurolengo (1972), Uropi (1986).

58 ARTIFICIAL LANGUAGES 353

Esperanto
The best-known of all ALs, Esperanto, was the

invention of a Polish oculist, Ludwig Lazarus

Zamenhof (1859-1917). The scheme was first pub-
tl"»
lished in Russian in 1887 under the title Mezb-
dunarodny yazyk ('An international language') PEKING
using the pseudonym 'Doktoro Esperanto' ('Doc-
tor Hopeful'). The language was called 'Lingvo
Internacia', but the name 'Esperanto' quickly Ludwig Zamenhof Zamen-
caught on, and in due course became the official hof's drafts of an interna-
first

title. The first Esperanto journal (La Esperantisto) tionallanguage were made
was published in 1889, and the First Universal when he was 1 5. His own
language background was
Congress of Esperanto was held in 1905, bringing very mixed: Russian was
together nearly 700 delegates from 20 countries. used at home, with Yiddish,
That year also saw the publication of the Funda- Polish, and Hebrew (in the

mento de Esperanto, an authoritative statement of synagogue) outside, and

French, German, Latin,
the language's structure and vocabulary, which
Greek, and English taught in
was to be the 'obligatory basis for Esperantists of school. In his later life, he ad-
all time'. vocated a world religion,
Today Esperanto is frequently encountered at which he called Homaranis-
international conferences. Several journalsand news- mo ('member of the human
race'), dedicated to peace,
papers are published language, and there
in the
tolerance, and the unity of
is a large translated literature, including the Bible peoples. Ironically, this ideal
and the Qur'an. There is extensive original work was the cause of a split with-
in the Esperantist move-
in the language, and several countries transmit
ment, many members wish-
radio broadcasts in Esperanto. In the 1970s, it was
ing to stress the practical
said to be taught in over 600 schools and 31 The front of a tourist information brochure about China,
value of the language rather
universities around the world. Estimates vary published by the China Esperanto League
than its religious signifi-
greatly about the number of fluent speakers (as cance.
with any language (§48)), from less than 1 million
The structure of Esperanto
to over 15 million, almost all being second-lan-
Sixteen grammatical rules word in a clause is allowed. eral formations have been
guage speakers, at various levels of proficiency. By (§16) have been explicitly 1 The objective case
3. is criticized for not using the
1972, there were 60 states with a national associa- laiddown. used when nouns reply to form common in modern
tion, and over 1,250 local societies. However, 1. La is the only article. where? languages. Some words
membership is often low, and usually does not 2. All nouns end in -o; plur- 14. An indefinite preposi- are thus not easy to recog-
als add -/. There are two tion, ye, is used when the nize; for example, hospital
provide a clear guide to the number of speakers.
cases, nominative and ob- choice of another preposi- is malsanulejo (the root
In 1979, the World Esperanto Association had jective - the latter ending in tion would be unclear. sana 'health' plus affixes).
around 31,000 members, most being in Eastern -n. Other meanings are ex- 1 5. Loan words follow the There are also several faux
Europe (especially Bulgaria, Poland, Czechoslo- pressed by prepositions. system of orthography. amis (p. 347), such as for-
3. All adjectives end in -a, 16. The a of the article, esto = absence'.
vakia, and Hungary). Japan is the main non-Euro-
and agree with the noun. and the o of nouns, may be The Slavonic bias to the
pean country with Esperanto speakers. 4. Numerals do not dropped for reasons of eu- pronunciation system has
Esperanto has still to achieve official status as change their forms. phony. been a further source of
an international language. A proposal to the United 5. The pronouns are mi, w, There are only about contention: not all lan-
li/si/gi, ni, w, and ///('I, you, 15,000 roots in the lan- guages share Esperanto's
Nations in 1966 was signed by nearly a million
he/she/it, we, you, they' guage, but these can be reliance on diphthongs and
people from 74 countries, but was not accepted.
respectively). combined in various ways sibilants (§27). Similarly,
There is a great deal of opposition from those who 6. Verbs have the same to produce a large vocabu- use of circumflexed letters
favour English as a world language, and from sup- ending throughout a tense great use
lary. In particular, is often considered
form. There are present, is made of prefixes and cumbersome, especially in
porters of other ALs. There are also objections on
past, and future tenses; im- suffixes to form complex typing and computational
politicalgrounds, because of the language's East
and sub-
perative, infinitive, words. setting (though the substi-
European background and the present-day trend junctive moods, and five Several criticisms have tution of h for " is per-
by many organizations (of varying political or reli- participle forms. been made of the lan- mitted).
gious views) to use Esperanto as a language of inter- 7. Adverbs add an -e end- guage. Within the gram- The language is rela-
ing. mar, the use of the objec- tively straightforward to
national propaganda.
8. Prepositions govern the tive case (not found in learn to read, as can be
nominative case. several modern languages) judged from a glance at
9. There are no silent let- has been particularly con- any illustrative text. As al-
ters. troversial, as has the use of ways with language learn-
10. The accent falls on the agreement between noun ing,however, passive com-
penultimate syllable. and adjective. Languages petence is much easier to
11. Compound words can that do not fit the latinate achieve than active use,
be formed by combining use of tenses or preposi- and a great deal of memory
roots. tions find particular prob- work is still needed before
12. Only one negative lems. In vocabulary, sev- fluency is acquired.

354 •
X LANGl AGE IN THE WORLD
An ideal artificial language ductory book that read: 'I, the undersigned, pro-
mise to learn the international language proposed
Five examples
English
Several criteria have been proposed for an 'ideal' by Dr Esperanto, if it appears that 10 million peo- For, of the things that man-
international artificial language, some of which are ple have publicly given the same promise.' He kind possesses in common,

more achievable than nothing is so truly universal

rather others. planned to publish a book that would contain all
and international as science.
their names and addresses — something that it has
Easy to learn This is usually taken to mean that Esperanto (using h for ac-
never been possible to do (though there is a register
the grammar has to be regular and simple, com- cents)
of Esperanto translators).
pared to that of natural languages; the semantic Char el la komunaj posedaj-
hoj de la homaro, neniu es-
formation of words has to be based on clear princi- Identity One of the chief functions of language
tas tiel vere ghenerala kaj
ples; the spelling should be phonetic; and there is to express identity, and this explains a large
internacia kiel la scienco.
should be no difficult sounds. number of linguistic differences (Part n). ALs do
Ido
not permit these differences. They are therefore in
Relatable to mother tongues It should be possible Nam del kozi, quin la homaro
conflict with the aspirations of movements where posedas komune, nula es
to translate into and out of any natural language
there is a desire to retain and express national, tarn vere universala ed inter-
with comparable ease; its structure would be flex-
regional, or social identities. The growth of nation- naciona kam la cienco.
ible, capable of reflecting the idiom of the speaker's
alism, in particular, stopped the expansion of ALs Novial
own language; would display many universal fea-
it
—
in the late 18th century a time when linguistic Den ek coses kel homaro
li li

tures of language, and use word roots that have posese comunim, nuli is tarn
divisions were seen as an asset and source of pride
a history of international use. verim general e international
— and it has had a marked effect on the AL move-
kam li scientie.
A rich range of functions It must be able to fulfil ment since the First World War.
Occidental
the ordinary needs of everyday speech and writing,
Linguistic bias It is not as easy to develop a simple Nam de omni comun posse-
as well as the specialized needs of science, religion, dages del homanite niun es
and common language as is claimed. Most ALs
trade, sport, politics, etc. It must also be capable tarn vermen general e inter-
are based on western Indo-European languages,
of being used in international communications national quam scientie.
and this acts as a barrier to speakers of other lan-
media, such as telegrams, radio, and television. Latino Sine Flexione
guages. There is still a marked linguistic parochial-
Nam, de commune posses-
Standardized There should be no dialectal vari- ism among ALs, which tend to underestimate the iones de genere humano,
ation, in order to avoid the risk of reduced intelligi- diversity of the world's languages (Part ix). nihil es tarn generale et inter-

bility. An authoritative body would monitor all nationale quam scientia.

Semantic differences Insufficient attention is paid
proposals for new forms. And a test in Interlingua
between lan-
to the semantic differences that exist
Translate!
Neutral It must be politically and linguistically guages: words do not have meanings that can be Tote le membros del com-
unaligned, and therefore equally acceptable to all neatly broken down into components (as some munitate de linguas occiden-
countries. Many AL supporters see this as an indis- 17th-century systems tried to do), or that are tal son in un certe senso dia-

lectos individual que devia

pensable step towards the unity of mankind and exactly equivalent to each other. Speakers of differ-
plus o minus de un patrono
a world of peace. ent languages may translate their mother-tongue commun. De iste facto son
words an AL, but this does not necessarily
into derivate tote principios
Providing insight Several AL supporters see inter- le

national communication as only one aim. They also
hope that the greater regularity and
languages will enable people to think more logi-
cally or rationally, and thus establish a deeper
understanding of the nature of reality. This search
for a philosophical language, in which words and
mean that they understand each other any better. methodic supportante le
compilation del Dictionario
The figurative, idiomatic, and connotative (p. 103)
clarity of these Interlingua-Anglese. Le ter-
uses of words will differ: for example, American
mine interlingua es sol-
and Soviet attitudes to a word like capitalism will mente un synonymo plus
not alter simply because both sides agree to use technic de lingua de patrono.
the same AL label. Illo representa un lingua que

es international proque su
ideas would be logically linked without ambiguity Antagonism Many people who are sympathetic elementos existe de facto o
or irregularity, was particularly common in 17th- to the general idea of a universal language are put potentialmente in un gruppo
de linguas national.
century ALs, which were generally in the form of off by the great fervour with which proponents of
taxonomies of concepts expressed in a complex ALs present their causes. Several AL organizations Crib (if required)
All the members of the wes-
notation. arrange public occasions with songs and hymns,
tern community of languages
each displaying a faith in the efficacy of its own are a sense individual dia-
in
AL which reaches evangelical proportions. Some lects which deviate more or
THE PROBLEMS ALs have been censured because the authorities less from a common pattern.
The international acceptance of an AL is a long associate them with political movements. Esper- From are derived all
this fact

uphill battle for its supporters, whohave to over- the working principles under-
anto, in particular, has been frequently persecuted
lying the compilation of the
come problems of a social, linguistic, and political — notably in Germany and the Soviet Union in the Interlingua-English Diction-
kind. 1930s, when the organization was suppressed and ary.The term interlingua is
many of its members arrested or shot. merely a more technical syn-
Motivation How is the inventor of an AL to per-
onym of pattern language. It
suade people to learn it, when no-one else knows stands for a language that is
it? To avoid this problem, there has to be a massive international because its ele-

period of simultaneous learning, which is extre- ments exist actually or po-

a group of national
tentially in
mely difficult to organize. Zamenhof was one who
languages.
saw this as a crucial factor. His solution was to
include promissory forms at the end of his intro-

58 ARTIFICIAL LANGUAGES •
355
Modified natural languages A section from the
BASIC dictionary
The language barrier has also been attacked by radio, geography, radium, names of countries and
(C. K. Ogden, 1932.)
several proposals to simplify the structure of a currencies). The operation of the system is illus-
natural language, usually by reducing the complex- trated below.
ity of its grammar or the size of its vocabulary. The system was strongly supported in the 1940s
All themain Western European languages have by such people as Churchill and Roosevelt, but
been modified in this way, the most famous there were also many criticisms. The simplification
approach being that of Charles Kay Ogden of the vocabulary is achieved at the expense of a
(1889-1957), known as Basic English (1930). more complex grammar and a greater reliance on
basic is an acronym for 'British American Scienti- idiomatic construction. The replacement forms are
fic International Commercial'. It consists of 850 often unwieldy, involving lengthy circumlocutions.
words selected to cover everyday needs: 400 And although basic proved easy to learn to read,
general nouns, 200 picturable objects, 100 general itproved very difficult to write in the language in cou, °* (dressy,
t
(wire,. T», st
qualities, 50 opposites, and 100 operations such a way that meaning was clearly preserved.
•
(round, a
(adverbs, particles, etc.). The working principle is The system is now largely of historical interest,
fB ' ,0
'>^ laoney .

that all words not on this list can be replaced by though analogous principles can be found in
those that are (permitting several inflectional varia- several spheres of foreign and remedial language
tions). The basic vocabulary is supplemented by teaching (e.g. the concept of a restricted defining
several international and scientific words (e.g. vocabulary in some contemporary dictionaries).

The 850 words of BASIC English

OPERATIONS QUALITIES
A translation into
WO lOOGeneral 50 Opposite
Picturable
BASIC
Below is a section of an eco-
nomics text, followed by Og-
complete den's own translation of it

into basic:

Narrow dispersions, skewed

negatively, signify deliberate
«.TTE*
human restriction of output.
ITMOIIIY
Skewed positively, after the
introduction of selection of
employees by test or exami-
nation, a narrow dispersion
indicates a successful sys-
tem of selection.
START RUSH PLANE PULL OPPOSITE
BUCKET GENERAL PUBLIC
STEAM
BUTTON
PLOUGH
POCKET
GOOD
GREAT The tendency to a common
STEP
STITCH
CAKE
CAMERA POTATO
GREY
HANGING
un
SECRET level of output being more
CARD SHORT
CART HARD frequent, is a sign that output
STORY CARRIAGE HEALTHY SIMPLE
CAT SLOW being consciously kept in-
is
CHAIN RECEIPT SMALL
CHEESB SOFT side a certain limit. When the
CHEST SOLID
SUGGESTION CHIM ROOP LIKE SPECIAL lowest outputs are most fre-
quent and the output of
workers not widely different
and generally high, after se-
lection of workers by test has

PROPERTY THEORY
come into use, the tendency
CUSHION SKIN OPEN
PKOTEST
DOC PARALLEI may be taken as a sign of the
SNAKE PAST
THUNOBB PHYSICAL efficiency of the system of
DRAWER POLITICAL
PURPOSR TIN DRESS POOR
PUSH SPOON
selection.
P -SSI- LI
QUALITY TOUCH EAR
QUESTION TRADE ICC SQUARE PRIVATE
(C. K. Ogden, 1938, p. 146.)

PINGEB STOCKINO REGi LA A

EASON USE STOMACH RESPONSIBLE
ECORD PLAG RIGHT
FLOOR STREET HOUND
ELATION VE5S1L PLY

EPRESENTAT1V FORK
EQUBST FOWL THREAD SERIOUS
FRAME SHARP
SMOOTH
STICKY
TOE STIFF
COAT TONGUE STRAIGHT
IVER STRONG
SUDDEN
WEATHER BWkTKf
TALL
WEIGHT HAT
DU
WINE HEART UMBRELLA
WINTER
TOMORROW OISTEIBU AND WOMAN HORN WATCH
YESTERDAY DIVISION
VIOLENT
DOUBT
SOUTH DRINK
BAST
WUT The original Basic English vocabulary, as
»LI*1I EARTH it was printed in
the early publications.

356 • X LANGUAGE IN THE WORLD

59 World languages
Many people feel that the only realistic chance of
breaking the foreign-language barrier is to use a Official languages
natural language as a world lingua franca. The The below gives the languages which have achieved special status in most of the
list
countries of the world. English is cited 45 times, French 30. About a quarter of the coun-
history of ideas already provides precedents, with
tries have more than one language with official or semi-official status.
Latin used as a medium of education in western Afghanistan Dari Persian,
Europe throughout the middle ages, and French Pashto
Albania Albanian
used as the language of international diplomacy
Algeria Arabic, French
from the 17th to the 20th centuries. Today, English Angola Portuguese
is the main contender for the position of world Antigua and Barbuda
lingua franca (p. 358). English
Argentina Spanish
There are few competitors. Several other lan- Australia English
guages have an important local role as a lingua Austria German
franca but no comparable international level of use, Bahamas English
Bahrain Arabic
such as Russian in eastern Europe, or Spanish in
Bangladesh Bengali
South and Central America. More people in the Barbados English
world speak Chinese than any other language Belgium Dutch,
French (German)
(§48), but in the West Chinese is too unfamiliar
Belize English
to be a serious contender. French is still widely Benin French
used, but far less than it was a century ago. Bermuda English
Many factors contribute to the gradual spread Bhutan Dzongkha
Bolivia Spanish
of a language - chiefly political and military might,
Botswana English, Tswana Jordan Arabic
economic power, and religious influence (all of Brazil Portuguese
which artificial languages lack, §58). These same Brunei'Malay, English
Bulgaria Bulgarian
factors mean that the development of a world lan-
Burkina Faso French
guage is not viewed with enthusiasm by those who Burma Burmese
would have to learn it. Such a language, it can Burundi French, Rundi
be argued, would give its originating culture an Cameroon English, French Lesotho Sotho, English
Canada English, French
unprecedented influence in world affairs and scien- Central African Republic
tificresearch. For example, scientists who used it French
as a mother tongue would be in a privileged posi- Chad French
Chile Spanish
tion: they would not have to spend time learning
China Mandarin Chinese
it and would more easily assimilate ideas expressed Colombia Spanish
in Furthermore, it is thought, a world language
it. Congo French
would inevitably erode the status of minority lan- Costa Rica Spanish
Cuba Spanish
guages and pose a threat to the identity of nations Cyprus Greek, Turkish
(§9). Many people thus view the current progress Czechoslovakia Czech,
of English towards world-language status with Slovak
Denmark Danish
concern and often with antagonism (p. 359). Djibouti French, Arabic
Ironically, the main danger to the growth of a Dominica English
world language comes from within. As the lan- Dominican Republic
guage becomes used in all corners of the world, Spanish
Ecuador Spanish
by people from all walks of life, so it begins to Egypt Arabic
develop new spoken varieties which are used by El Salvador Spanish
local people as symbols of their identity (Part n). Equatorial Guinea Spanish
Ethiopia Amharic
In the course of time, these new varieties might English
Fiji
become mutually unintelligible. How far this diver- Finland Finnish, Swedish
sification will affect English cannot be predicted France French
French Guiana French
(p. 359). It is not easy to weigh the trend towards
Gabon French
diversity against the trend towards unity that Gambia English
results from increased modern contacts through Germany (FRG, GDR)
travel and communications. A hundred years ago, German
Ghana English
predictions were being made that British and Greece Greek
American English would by now be mutually Greenland Danish, Green-
unintelligible. Linguistic predictions have a habit landic
Grenada English
of being wrong.
Guadeloupe French
Guatemala Spanish
Guinea French
Guinea-Bissau Portuguese
Guyana English
Haiti French
Honduras Spanish
Hungary Hungarian
(Magyar)
Iceland Icelandic
India Hindi, English, 14
regional languages
Indonesia Bahasa
Indonesia
Iran Farsi (Persian)
Iraq Arabic
Irish Republic Irish, English
Israel Hebrew, Arabic
Italy Italian
Ivory Coast French
Jamaica English
Japan Japanese
Kampuchea Khmer
Kenya Swahili, English
Korea (N. and S.) Korean
Kuwait Arabic
Laos Lao
Lebanon Arabic
Liberia English
Libya Arabic
Luxembourg French,
German, Letzebuergesch
Madagascar French,
Malagasy
Malawi English, Nyanja
(Chewa)
Malaysia Malay
Maldives Maldivian
Mali French
Malta Maltese, English
Martinique French
Mauritania Arabic, French
Mauritius English
Mexico Spanish
Mongolian People's
Republic Khalka
Morocco Arabic
Mozambique Portuguese
Nepal Nepali
Netherlands Dutch
New Zealand English
Nicaragua Spanish
Niger French
Nigeria English
Norway Norwegian
Oman Arabic
Pakistan Urdu, English
Panama Spanish
Papua New Guinea Tok
Pisin, Hiri Motu, English
Paraguay Spanish,
Guarani
Peru Spanish
Philippines Pilipino,
English
59
Poland Polish
Portugal Portuguese
Puerto Rico Spanish,
English
Qatar Arabic
Reunion French
Romania Romanian
Rwanda Kinyarwanda
(Rwanda), French
St Christopher and Nevis
English
St Lucia English
St Vincent and the
Grenadines English
Sao Tom6 and Principe
Portuguese
Saudi Arabia Arabic
Senegal French
Seychelles English, French
Sierra Leone English
Singapore Chinese, Malay,
Tamil, English
Somalia Somali, Arabic
South Africa Afrikaans,
English
Soviet Union Russian
(local national languages)
Spain Spanish (Catalan,
Basque)
Sri Lanka Sinhala
Sudan Arabic
Suriname Dutch, English
Swaziland Swazi, English
Sweden Swedish
Switzerland French,
German, Italian,
Romansch
Syria Arabic
Taiwan Mandarin Chinese
Tanzania Swahili, English
Thailand Thai
Togo French
Trinidad and Tobago
English
Tunisia Arabic
Turkey Turkish
Uganda English
United Arab Emirates
Arabic
United Kingdom English
United States of America
English
Uruguay Spanish
Vatican City Italian, Latin
Venezuela Spanish
Vietnam Vietnamese
Yemen (Arab Republic and
Democratic) Arabic
Yugoslavia Macedonian,
Serbo-Croat, Slovene
Zaire French
Zambia English
Zimbabwe English
WORLD LANGUAGES •
357
 • 'When finish learn English, my pay as secretary will Opposition to English The
World English I

be increase by nearly ten times.' influence of English within a

(Egyptian trainee secretary.) country has often been bit-
In the minds of many people, there is no longer terly condemned. France
an issue. They argue that English has already • 'My company plans big deals with Arabic world. has issued laws banning its
become a world language, by virtue of the political None of us speak Arabic, and they do not know use in certain public domains
and economic progress made by English-speaking Japanese. All our plans and meetings are in English.' (p. 4). In French-speaking
(Japanese businessman.) Quebec, English advertise-
nations in the past 200 years, and is likely to remain
ments, shop names, and
so,gradually consolidating its position. • After I learned English, I felt I was in touch with traffic signs have been
An impressive variety' of facts about usage sup- the international world for the very first time.' changed, in an effort to stop
port this view. According to conservative estimates, Nigerian teacher.) the advance of English in the
province. There have been
mother-tongue speakers have now reached around • 'If I want to keep up to date with the latest techniques
anti-English movements in
300 million; a further 300 million use English as and products. I must certainly maintain my English very Spain, Germany, Mexico,
a second language (p. 368); and a further 100 mil- strongly.' Burma, India, and several
lion use it fluently as a foreign language. This is (Indian doctor.) other countries; but for the
an increase of around 40% since the 1950s. More • 'Nearly everyone in Denmark speaks English. If we most part they have had little
effect.
radical estimates, which include speakers with a wouldn't be anyone to talk to.'
didn't, there
lower level of language fluency and awareness, (Danish university student.)
have suggested that the overall total is these days
well in excess of 1,000 million. The variation
results largely from a lack of precise data about
English language use in such areas as the Indian
sub-continent, where the historical impact of the
language exercises a continuing influence on many
of its 900 million people, and China, where there
has been a burst of enthusiasm for English language
studies in recent years, with over 100 million peo-
A signpost in Wales. The
ple watching the BBC television English series Fol-
English names have been
low Me. One visitor, returning to China in 1979
defaced by Welsh national-
after a gap of 20 years, wrote: 'In 1959, everyone ists.

was carrying a book of the thoughts of Chairman

Mao; today, everyone is carrying a book of elemen- International varieties
tary English.' Even if only 10% of these learners Some people think they al- particular are faced with a (isn't it?, aren'tthey?. etc.),

become fluent, the effect on totals is dramatic: the ready see signs of the conflict of aims:
should which might be reduced to
break-up of the language they teach British or Ameri- a single form (such as (isn't
number of foreign learners is immediately doubled.
when they find difficulty in can English? both? or that) right?); or one of the
Surveys of range of use carried out by unesco neither, focussing on the
understanding the English indirect object construc-
and other world organizations reinforce the general used in India, West Africa, variety found in their own tions (English has both /

statistical impression. English is used as an official or other parts of the world country? What effect will gave the man a book and
or semi-official language in over 60 countries, and - even, at times, within their decision have on the / gave a book to the man).

Great Britain and North ability of their students to When all such options are
has a prominent place in a further 20. It is either
America (§8). Variation can communicate at an inter- removed, in both grammar
dominant or well established in all six continents. also be seen in the written national level? and vocabulary, we are left
It is the main language of books, newspapers, air- language, mainly in a dis- with the obligatory mini-
ports and air-traffic control, international business tinctive regional lexicon. An Nuclear English mum of the language, its

and academic conferences, science, technology, English edition of an Indian These problems are of re- communicative nucleus.
newspaper, for example, cent origin, and are only To be successful. Nuc-
medicine, diplomacy, sports, international com- beginning to be discussed.
might refer routinely to roti. lear English would have to
petitions, pop music, and advertising. Over two- kapra, and makan (food, One proposal, made by the be easier to learn than any
thirds of the world's scientists write in English. clothing, and housing'), a British linguist Randolph variety of full' English; it

Three-quarters of the world's mail is written in ra/7 roko (railway strike), Quirk (1920- ), argues that would have to meet the
and to such quantities as a the problem of variety communicative needs of its
English. Of all the information in the world's elec-
crore ("10 million') or lakh would be avoided if the lan- users; and it would have to
tronic retrieval systems, 80% is stored in English. ('100,000'). In some inter- guage were specially be capable of development
English radio programmes are received by over 150 national varieties of Eng- adapted to produce a nuc- into an expanded' form, for
million in 120 countries. Over 50 million children lish, the local standard vo- lear' English for inter- more advanced uses. It is

study English as an additional language at primary cabulary (including words national use. Nuclear Eng- too soon to say how far the
for local food, fauna, or would provide a core
lish' proposal can be taken; but
level; over 80 million study it at secondary level
flora) may run to thousands of structureand vocabulary thereis no doubt that the
(these figures exclude China). In any one year, the of items. from within the range of ac- possibilities of linguistic
British Council helps a quarter of a million foreign This variation raises a ceptable English. It would adaptation provide an inter-

students to learn English, in various parts of the question mark against the eliminate all features that esting theoretical alterna-
notion of 'world' English. are dispensable', in the tive to proposals of outright
world. In the USA alone, 337,000 foreign students
With so many varieties, sense that the language adoption, which have so far
were registered in 1983. which one should be used has an alternative means been the focus of attention,
It would be possible to continue with such statis- as the international me- available for their expres- and the source of much
tics for severalpages, but their significance can be dium? Should it be Ameri- sion. Examples of omis- controversy. (After R.
can, British, Indian. Austra- sible structures include the Quirk, 1982.)
illustrated more succinctly by comments made by
lian ...? Teachers in range of tag questions
foreign learners themselves:

358 •
X LANGUAGE IN THE WORLD
 +

Estimates of the number of English speakers in the world

The first column gives figures for those who speak English as a mother tongue or
first language. The second column gives recent total population figures

(usually 981 for those countries where English has official status as a medium of communication, and where people learn it, usually in school, as a second
1 )

language. These totals bear little correlation with the real use of English in the area. There are no figures available for people who have learned English as
a foreign language, in countries where the language has no official status. The plus sign means more than'; the minus sign means 'less than'; and the question
mark is used for cases where no-one knows how many first language speakers there are (though in these cases the figures are likely to be very small).

Country First Country Country First Country Country First Country

language population language population language population
speakers of speakers of speakers of
English English English

Anguilla 10.000- 10.000 Kenya 7,000,000 Seychelles

1 60.000
Antigua and Kiribati 60,000+ Sierra Leone 3,600,000
Barbuda 100.000 100.000 Lesotho 1,400,000 Singapore ? 2.500,000
Australia 14,000,000 15.000,000- Liberia 2,000,000 Solomon Islands 200.000+
Bahamas 250.000 250.000 Malawi 6,400,000 South Africa 2.000.000+ 30,000.000
Bangladesh ? 92.600,000 + Malaysia (East) 14,300,000 SriLanka ? 15.200.000+
Barbados 250,000+ 250,000+ Malta 350,000 Suriname 395.000
Belize 1 00.000 150.000+ Mauritius ,000,000 Swaziland
1 600,000
Bermuda 50,000+ 50.000+ Montserrat 15.000 15,000 Tanzania 18,500,000
Bhutan ? 1,200.000+ Namibia 1,000,000 Tonga 100,000+
Botswana 1,000.000- Nauru 80,000+ Trinidad and
Brunei 200,000+ Nepal ? 15,800.000+ Tobago 1.200,000 1 ,200,000
Cameroon 8,000,000+ New Zealand 3,000,000 3.200,000 Tuvalu 8.000+
Canada 17,000,000+ 24.000,000+ Nigeria ? 90,000,000+ Uganda 13,000.000
Dominica 50,000+ 100.000- Pakistan ? 87,100,000+ U.S. 21 5.000.000 230,000,000+
Fiji 600.000+ Papua New U.S. territories in
Ghana 12.000.000 Guinea 3,500,000 Pacific 300,000-
Gibraltar 30.000* Philippines 50,000,000 Vanuatu 100,000
Great Britain 56,000,000+ 57.000.000 St Christopher and Western Samoa 150,000+
Grenada 100,000+ 100,000- Nevis 60,000 60,000 Zambia 6,000,000
Guyana 900,000+ 900.000 St Lucia ? 100,000+ Zimbabwe 200,000+ 7,600.000
Hong Kong ? 6,000,000- St Vincent and the Other British
India ? 700,000.000- Grenadines 100,000+ 100,000+ territories 30.000+ 30.000+
Irish Republic 3,300.000 3,300.000 Senegambia 600.000 TOTALS 316,015,000+ 1 .606,850.000
Jamaica 2.300.000 2,300,000

The spread of English as a world language This map shows the growing use of English, both in those countries where it is a mother tongue, and in those
where has official or semi-official status. The main countries of the world have been shown larger or smaller than their actual size, to reflect their
it

relative share of the world's population. The role of the Indian sub-continent in the population estimates for English is obvious. There are around 900 million
people in that area, but estimates of those who are fluent in the language have been as low as 3%. (From R. W. Bailey & M. Gorlach, 1982.)

['/
EAST

i^ETHER^
iLANDST
I"*
7
\\POLAND, '
USSR ^MONGOLIA S"
""
"7 WEST\ V / '

IGERMANYV^ . . /JJ^czechqslOVAKIA
\
-^vBELGIUM ST>
>
y-^j!

SWITZERLAND^) ^./'-T^i*
(

^\S
^V P / vT / f > CHINA
FRANCE
/"
A VV--</
x- YUGO-V- A
^,
\
_„.... V
I .JsLAVlA » —
ILBAN1A.\_.JHQ B3V
^ SPAIN Wy^'^^\~~< V_\ ' .-AFGHANISTAN

{
7 Jv TURKEY / /•• •••.•..•:*-r^.
# }

MAUMTA
.oVS^VlRAQ 1 X £.-•• \HL;..

^
.;}:. :
;;;

R*NCXAt>£saj_
i
:

^^^ J aO~\\ (

WA&BADOS

1
I GUYANAM \GU
T\_')\ SUIUNA 1
,...,,
,

7 BRAZIL
SCIM i

rPERU'.
NAMIBIA/
/madaoasca*
ZIMIAJWE'
LANKA
URUGUAY

59 WORLD LANGUAGES .
359
 60 Multilingualism

People brought up within a western society often country, with focal points in those rural areas •What causes
think that the monolingualism that forms a routine where languages are in contact. The bilingualism multilingualism?
part of their existence is the normal way of life may be due to a long-standing co-existence of dif- A multilingual situation can
for all but a few 'special' people. They are wrong. ferent groups (as in Belgium) or to a more recent develop for reasons which
Multilingualism is the natural way of life for and shifting co-existence (as with the many Gast- may be difficult to dis-
hundreds of millions all over the world. There are arbeiter groups in Europe, p. 36). need to We entangle because of their
obscure historical origins.
no but with around 5,000 lan-
official statistics, examine many such situations in individual detail,
Often the situation is of the
guages co-existing in fewer than 200 countries in order to arrive at valid conclusions about the people's own choosing; but it

(§§47, 59) it is obvious that an enormous amount social antecedents and consequences of multi- may also be forced upon
of language contact must be taking place; and the lingualism. them by other circum-
stances.
inevitable result of languages in contact is multi- An important characteristic of these situations
• Politics Annexation,
lingualism, which is most commonly found in an is their fluidity. It is where
rare to find a setting
resettlement, and other poli-
individual speaker as bilingualism. the languages are stable and balanced, and where tical or military acts can have
The widespread impression that multilingualism social controversy over government policy is not immediate linguistic effects.
is uncommon is promoted by government policies: a major issue (Paraguay is one such exception). People may become refu-
gees, and have to learn the
less than a quarter of the world's nations give offi- Usually the language balance is changing, either
language of their new
cial two languages, according to the
recognition to spontaneously or because of government pressure. homes. After a successful
list on p. 357, and only six recognize three or more, In some areas, the level of bilingualism is increasing military invasion, the indige-
however, when we look at what is taking place (e.g. Sweden Second World War); in
since the nous population may have to
learn the invader's language
v :f hin each country, studying the speakers rather others it is decreasing, with second- and third-
in order to prosper.
.nan the national policies, a very different picture generation immigrants becoming increasingly
• Religion People may wish
emerges. It has been argued, in fact, that there is monolingual (e.g. the USA). to live in a country because
no such thing as a totally monolingual country. A distinction is commonly drawn between cases of its religious significance,
Even in countries that have a single official lan- where one language is holding its own despite the or to leave a country be-
cause of its religious oppres-
guage used by the majority of the population (e.g. influence of powerful neighbours (language main-
sion. In either case, a new
Britain, USA, France, Germany, Japan), there exist tenance) and cases where a language has yielded language may have to be
sizeable groups that use other languages. In the to this influence, and speakers have assimilated to learned.
USA, around 7% of the population regularly speak the dominant culture (language shift). Other possi- • Culture A desire to iden-
a language other than English. In Britain, over 100 tifywith a particular ethnic
bilities include extensive vocabulary borrowing by
culture or social group
minority languages are in routine use. In Japan, one of the languages, or the emergence of a new usually means learning the
one of the most monolingual of countries, there 'hybrid' as a result of the contact, as with pidgins language of that group. Na-
are substantial groups of Chinese and Korean and Creoles (§55). Lastly, as shown by the history tionalistic factors are particu-
speakers. In Ghana, Nigeria, and many other Afri- of the Celtic languages (p. 302), the contact can larly important (§9).
• Education Learning
can countries that have a single official language, lead to a language being completely eliminated
another language may be
as many as 90% of the population may be regularly (language death). the only means of obtaining
using more than one language. It is an interesting access to knowledge. This
irony that there may be more bilingual people in factor led to the universal
an monolingual country than in an offi-
officially use of Latin in the middle
ages, and today motivates
cially bilingual one, because in the latter case there
the international use of Eng-
tends to be territorial separation between the lish (§59).
groups (as in Belgium, Switzerland, or Yugoslavia). • Economy Very large
numbers of people have mi-
grated to find work and to im-
prove their standard of living.
Maintenance, Language death
shift, and death It unusual to see the death
is
This factor alone accounts
for most of the linguistic di-
impossible to generalize about the way multi-
It is ofa language commemor- versity of the U.S., and an in-

lingualism manifests itself around the world; there ated - but in this particular creasing proportion of the bi-
case the judgment may have lingualism in present-day
are vast differences in social and cultural situations. been premature. In recent Europe.
Often the majority of a population is bilingual (e.g. years an enthusiastic reviva- • Natural disaster Floods,
the widespread use of Spanish and Guarani in Para- list campaign has been OOfcarHY PFfcTRFATH volcanic eruptions, famine,
guay, where the former is used as the official lan- launched to breathe new life ; O in and other such events can
into the Cornish language (p. «! TO » be the cause of major move-
guage, and the latter as the national language);
303). k iMt Haci ments of population. New
often only a small minority may be affected (e.g. **.t or
language contact situations
this iOwwi » 'H(« r*t
Gaelic speakers in Scotland). The majority of the fA*l>r*T »HOKOJ then emerge as people are
« ftfcf

bilingual speakers may be concentrated in the - 1 yȴ resettled.

cities, or they may be found throughout the

360 • X LANGUAGE IN THE WORLD

 The estimated percentage of
people with a non-English
background in the U.S. in the
mid-1970s, and the main mi-
nority languages in each
state. The highest figures are
in the north-east (around
New York and its hinterland),
in the north-central states,
and in the south-west (where
the main influx of Spanish
speakers has taken place).
The figures are based on the
1 976 U.S. Survey of Income

and Education, and do not

reflect the marked increase
inthe non-English-speaking
population which has since
taken place in certain States
(such as Texas and Califor-
DE7% Grosjean,
nia). (After F.
1982.)
MD7% G,I,V,K

DC 8% G,S

Languages States
A Apache H Hawaiian P Polish AK Alaska Hawaii ME Maine NJ New Jersey SD South Dakota
Ar Arabic HC Haitian Creole Po Portuguese AR Arkansas Iowa Ml Michigan NM New Mexico TN Tennessee
C Chinese I Italian S Spanish AZ Arizona Idaho MN Minnesota NV Nevada TX Texas
Ch Cherokee In Inupiac SI Slovak CA California Illinois MO Missouri NY New York UT Utah
Cr Crow J Japanese Sw Swedish CO Colorado Indiana MS Mississippi OH Ohio VA Virginia
Cz Czech K Korean T Tewa CT Connecticut Kansas MT Montana OK Oklahoma VT Vermont
F French L Lakota V Vietnamese DC District of Columbia KY Kentucky NC North Carolina OP. Oregon WA Washington
Fi Filipino M Mohawk Y Yiddish DE Delaware LA Louisiana ND North Dakota PA Pennsylvania Wl Wisconsin
G German Norwegian Yu Yupik FL Florida MA Massachusetts NE Nebraska Rl Rhode Island WV West Virginia
Gr Greek Na Navaho GA Georgia MD Maryland NH New Hampshire SC South Carolina WY Wyoming

A journey to work
'My French friend had been
brought up with the erro-
neous idea that the United
States was a monolingual
English-speaking country
with a few, fast-disappear-
ing linguistic minorities.
One day took him to work
I were
with me by a roundabout
route to show him the great
linguistic diversity that can
be found in an American
city - in this case, Boston.
As we walked to the bus
stop, we passed a group of
Haitian children playing ball
and shouting at each other
in On the
Haitian Creole.
bus to Cambridge we sat
next to an Armenian Ameri-
can from Waltham reading
one of the two daily Arme-
nian newspapers published
in the Boston area. Walking
down Cambridge Street,
we found ourselves for a
short while in a little Portu-
gal - the people around us
spoke Portuguese, the
stores sold Portuguese
goods, the restaurants
offered Portuguese specia-
and the children in
lities,
the area were off to their bi-
lingual programs in school.
We continued on our tour
and went to Boston's North
End for breakfast.
in Italy. A procession
in honor of a saint was get-
ting under way, a group of
elderly Italian-speaking
people was playing cards
in the shade of a tree, and
storekeepers were setting
up their displays of Italian
cold cuts and cheeses. No-
ticesposted on the walls
were in Italian, and as we
entered a pastry shop the
customers were all speak-
ing Italian to one another.
From the North End we
walked a few blocks to
Chinatown, with its Chi-
nese-speaking inhabitants,
street signs in Chinese and
English, bilingual school,
and Chinese stores, clubs,
and temples. Because time
was getting short we
decided not to visit
Dorchester with its large
Creole-speaking Haitian
population, and we quickly
passed through South Bos-
ton, where many Hispanic
Americans live. We did
Now we have time, however, to buy
the local bilingual Spanish-
English paper and check
the times at the local cine-
ma that shows only Spa-
nish films.
We then arrived at my
university,which wel-
comes, in addition to its
American student popula-
tion, students from thirty
foreign countries. In the la-
boratory we set about our
day's work on a research
project concerning yet one
more language actively
used in the United States,
Amprican Sign Language,
the manual-visual lan-
guage of many deaf Ameri-
cans.'
(F. Grosjean. 1982, pp. 42-
3.)
A corner of 'Little Italy'

60 MULTILINGUAL1SM •
361
Being bilingual Some achieve it in one (their 'preferred' or 'domi- Dormant languages
nant' language), but not the other. For such rea- There is no clear indication
Research into bilingualism usually distinguishes sons, scholars now tend to think of bilingual ability as to whether there is a limit
between large-scale analyses of multilingual so- as a continuum: bilingual people will find them- to human multilingual ability.
Cardinal Giuseppe Mezzo-
cieties ('societal' bilingualism, p. 360) and small- selves at different pointson this continuum, with
fanti (1774-1849), librarian
scale analyses of the settings in which bilingual a minority approaching the theoretical ideal of per- at the Vatican, is reputed to

speakers interact ('individual' bilingualism). fect, balanced control of both languages, but most have been able to speak 50
Several fundamental questions have to be dealt being some way from it, and some having very languages (most with great
with under the latter heading — in particular, how limited ability indeed. However, the notion is a fluency), to understand a
further 20, and to translate
bilingualism is and defined, and
to be identified difficult one to make precise, because so many
114. The Victorian diplomat
what its purpose is within the speech community. different abilities are involved - in speaking, SirJohn Bowring (1792-
Both questions have 'obvious' answers, neither of listening, reading, and writing, as well as in phono- 1 was said to have spo-
872)
which is adequate. logy, grammar, vocabulary, and pragmatics (Parts ken 1 00 languages and read
another 100. Unfortunately,
m-v).
there no way knowing
WHAT IS A BILINGUAL? is of
exactly what proficiency level
The obvious answer is: someone who speaks two WHY USE TWO LANGUAGES? was achieved by these re-
languages. But this answer will not suffice. It does Here, the 'obvious' answer is: to communicate with markable language learners.
not allow for those who make irregular use of one people of different language backgrounds. And It is in fact highly unusual

to maintain proficiency in
or other language, or those who have not used the once again, the obvious answer will not account
more than two or three lan-
language at all for many years (so-called 'dormant' for the remarkable range of linguistic behaviour guages at a time. Most multi-
bilinguals). Nor does it allow for the many people that can be observed in adult bilinguals. The 'easy' linguals have a single domi-
who have developed a considerable skill in compre- cases are those where a bilingual meets different nant language, others being
'dormant' to varying degrees.
hending a foreign language, but who do not speak monolingual people within a multilingual society,
The typical situation can be
it; or those who have learned to read in another and changes from one language to the other in illustrated by a case study
language, but who cannot speak or write It it. order to communicate with them. Somewhat more that was made in the field of
leaves unclear the relationship between different complex are cases where a bilingual chooses to use aphasia (p. 270). It emerged
languages and different dialects, styles, or levels one language knowing that the listener would that the person had learned
seven languages during his
of the same language (as in the case of diglossia, prefer the other (for example, electing to be tried
life, but five had become dor-
p. 43). And above all, this definition says nothing in the language of a minority' group, in order to mant. His mother tongue had
about the level of proficiency that has to be attained embarrass the authorities). Here, language choice been Hungarian. At the age
before speakers can legitimately claim to be bi- is a symbol of national identity. of 4, he moved to Poland,

lingual. But such bilingual/monolingual interactions and learned Polish, and stopped
using Hungarian. When he
The notion of proficiency raises some very com- confrontations account for only a minority of
was 6 he returned to Hung-
plex issues. Again, the 'obvious' answer is to say cases. More often, in a multilingual society, bi- ary, and had to relearn Hun-
that people are bilingual when they achieve native- linguals interact with other bilinguals, and opt to garian. At the age of 1 0, he
like fluency in each language. But this criterion is use their different languages in a complex network moved to Romania, using
far too strong. People who have 'perfect' fluency of interaction that proves extremely difficult to des-
Romanian in school and
Yiddish socially. Two years
in two languages do exist, but they are the excep- cribeand explain. The choice of language will vary later he returned to Hungary,
tion, not the rule. The vast majority of bilinguals depending on the type of person addressed (e.g. where in school he learned
do not have an equal command of their two lan- members of the family, schoolmates, colleagues, German, English, and He-
guages: one language is more fluent than the other, superiors, friends, shopkeepers, officials, transport brew. This was followed by
six years in Germany, during
interferes with the other, imposes its accent on the personnel, neighbours), and on the location or
which time German became
other, or simply is the preferred language in certain social setting (e.g. a family may vary their language his dominant language. At
situations. For example, a child of French/English use depending on whether they are at home, in 25, he moved to the U.S.,
parents went to school and university in France. the street, or in church; at the office, someone may where English became domi-
nant. At the time of the study,
She became a geography teacher, married a British talk to a colleague in language X, but over lunch
only English and Hungarian
doctor, and came to live in England, where she talk to the same person using language Y). Even were regularly used (his wife
had her first child. In general conversation, she more complex, and not well understood, are the isHungarian). The others
could cope with ease in either language; but she many cases when a bilingual talks to another bi- were dormant, and in some
found herself unable to teach geography in English, lingual with the same language background, and cases almost forgotten. (L.
Galloway, 1978.)
and she was extremely reluctant to discuss baby yet changes from one language to another in the
care in French. In each case she knew the slang, course of the conversation — a phenomenon known
jargon, and phrasing which is naturally assimilated variously as 'language mixing', 'language switch-
when learning a new skill - but this was available ing', or simply 'code switching'.
inonly one of her languages. Her linguistic compe-
tence certainly did not resemble that of mono-
lingual teacher-mothers.
This situation seems to be typical. Studies of bi-
lingual interaction have brought to light several
differences in linguistic proficiency, both within
and between individuals. Many bilinguals fail to
achieve a native-like fluency in either language.

362 •
X LANGUAGE IN THE WORLD
Language switching
Bilingual verbal strategies
Switching between languages is extremely common
Language switching is a refused] Tu no fumas, ver- los (unclear speech) no
and takes many forms. A long narrative may switch major feature of this con- dad? Yo tampoco. Deje de habia que no traia cigarros
from one language to the other. Sentences may versation between two na- fumar (You don't smoke, Camille, no traia Helen, no
alternate. A sentence may begin in one language, tive Americans of Mexican do you? don't either.
I I de Leon
traia yo, el Sr.
ancestry. E
a university stopped smoking') and I'm (Yesterday the - there
and finish in another. Or phrases from both lan- is

teacher, who is working as back to it again. weren't any. Camille didn't

guages may succeed each other in apparently ran- a volunteer in a day care have any, Helen, I, Mr. de
dom order (though in fact grammatical constraints centre where Mis a social M: An' - an' - an' they tell Leon have any) and
didn't
are frequently involved). Such behaviour can be worker. The Spanish pas- me, How did you quit, I saw bag crumpled
Dixie's

explained only by postulating a range of linguistic sages are translated in Mary?' din' quit. - just
I I 1 up, so figures she didn't
I

parentheses. stopped. mean it wasn't

I have any, y ahi ando en los
or social factors such as the following.
an effort that made. Que I ceniceros buscando a ver
• The speaker cannot express himself adequately E: What do you dream in? voy a dejar de fumar por- ondeestabala-('And
M: don't think ever have que me hace daho o (That there am in the ashtrays
I

in one language, and switches to the other to make

I I

any conversations in my I'm going to stop smoking looking to see where there
good the deficiency. This may trigger the speaker dreams. just dream. Ha.
I I because it's harmful to me, was the-') din' care I

to continue in the other language for a while. An don' hear people talking: I or) thisortha', uh-uh. It whose they were.
example from a Spanish/English study (G. Valdes jus' see pictures. just - that - eh - used to 1

E: Oh, they're old-fash- pull butts out of the - the The authors of this study
Fallis, 1976): Porque alii hay cashews. You don't - wastepaper basket.
ioned, then. They're not tal- point out that Ms language
like them? (Because here are some cashews ...').
huh?
kies yet, Yeah. (Laughter) I used to switching is not random. M
This tends to happen a great deal when the speaker M: They're old-fashioned. go look in the (unclear is ambivalent about her
is upset, tired, or otherwise distracted. No, they're not talkies yet. speech). Se me acababan smoking, and she signals
• Switching to a minority language is very com- No, I'm tryin' to think. Yeah, los cigarros en la noche this through her choice of
there too have been tal- (My cigarettes would run language. Spanish sen-
mon as a means of expressing solidarity with a kies. Different. In Spanish out on me at night'). I'd get tences in this conversation
social group. The language change signals to the and English both. An' I desperate, y ahi voy al ba- reflect her embarrassment
listener that the speaker is from a certain back- wouldn't be too surprised if surero a buscar, a sacar, and personal involvement;
ground; if the listener responds with a similar I even had some in Chi- you know? (And there go I English is used for more
nese. (Laughter) Yeah, Ed. to the wastebasket to look general or detached state-
switch, a degree of rapport is established. The same
Deveras. ('Really') (M for some, to get some, you ments.
switch may of course also be used to exclude other offers E a cigarette which is know?') (Laughter) Ayer (J.Gumperz, 1970.)
people, who do not know the language, from the
group.
• The switch between languages can signal the
Bilingual acquisition
speaker's attitude towards the listener - friendly,
There is a widespread pop- other. to the other. Indeed, if one
irritated, distant, ironic, jocular, and so on. Mono-
ular impression that the 2. Whensentences begin parent uses the language
linguals can communicate these effects to some children of bilingual parents to contain two or more ele- of the other to the child,
extent by varying the level of formality of their are linguistically at risk. It is ments, words from both there may be quite a reac-
speech; bilinguals can do
by language switching.
it said that their brains will languages are used within tion. The child may be sur-

If two bilinguals normally talk to each other in

not be able to cope, and the same sentence, e.g. prised, embarrassed, fail to
that they will grow up 'semi- (from a 2-year-old German/ understand, think it funny,
language X, the choice of Y is bound to create lingual', confused, or re- English child) ein ('a') big or become upset. An ex-
a special effect. A common example is for a mother tarded. There isno justifi- cow, from up Himmel in tractfrom a recent bilin-
to tell her child to do something in one language, cation for this pessimism, ('sky'). The amount of mix- gual-acquisition study illus-

and then, if the child fails to obey, to switch to as evident from the confi-
is ing rapidly declines. In one trates this last reaction.
dent fluency displayed by study, at the beginning of Lisa (nearly 4 years old)
another language, thereby showing her stronger
millions of bilingual and tri- the third year, nearly 30% has an and a
Italian father
emphasis or displeasure. lingual children over the all of the sentences contained German mother. The father
world. By the time these mixed vocabulary; by the uses a short German sen-
These are but some of the sociolinguistic func- tence to her, to which she
children arrive in school, end of the year, it was less
tions that language switching can perform. The the vast majority have than 5%. replies:
phenomenon is complex and subtle one,
evidently a reached the same stage of 3. As vocabulary grows in Lisa: No, non puoi. ('No,
with speakers usually being totally unaware of the linguistic development as each language, translation you can't')
have their monolingual equivalents develop. But Father: Ich auch - spreche
extent to which they have been switching in a con-
peers. the acquisition of separate Deutsch. (I also speak
versation. If interrupted, they may even be unable But the process of learn- sets of grammatical rules German.')
to say which language they were using in their last ing two languages is not takes longer. For a while, a Lisa: No, tu non puoi! (No,
sentence. Monolinguals often dismiss or satirize exactly the same as the single system of rules you cannot.')
language switching, using such pejorative labels as process of learning one seems to be used both for (V. Volterra & T.
(Part vm). Three main languages, until finally the Taeschner, 1978.)
'Franglais', 'Spanglish', or 'Tex-Mex'. Perhaps
stages of development two grammars diverge. Not surprisingly, it is at
because of this kind of criticism, many bilingual have been noted: When bilingual children this age that children try to
people come to be very self-conscious about their 1 The child builds up a list
. reach this stage, usually in play their parents off
switching, and try to avoid it in talking to strangers of words, as does a mono- the fourth year, they have against each other. One

or on formal occasions. But in informal speech, lingual child, but the list become aware that the two child would always switch
contains words from both languages are not the into French when he saw
it is a natural and powerful communicative feature
languages. It is rare for same. They typically use his English father approach
of bilingual interaction, which presents linguists these words to be transla- each language to the par- him purposefully at bed-
with one of their most intriguing analytical chal- tion equivalents of each ent who speaks it, and not time!

lenges.

60 MULTILINGUALISM 363
61 Language planning

Language, sooner or later, proves to be a thorn assist those (politicians, educators, lawyers, etc.) Two kinds of
in the flesh of all who govern, whether at national whose responsibility it is to make decisions about language planning
or local level. Different social groups wish to see the development of languages in society, many of Many analysts recognize a
their linguistic identities and interests maintained, whom have no specialized knowledge of linguistic binary classification of lan-
and may actively — and often violently - campaign issues. But it also presents a fresh perspective guage-planning activities,
for recognition (§9). Governments have to react for our understanding of linguistic change (§54). based on whether the
changes affect primarily
to these differences, officially or unofficially: they Many linguists have held the view that language linguistic structure or linguis-
may wish them, or try to eliminate
to reconcile change is a natural, spontaneous phenomenon, the ticuse (§1 3). In corpus plan-
them. With the pace of change increasing, and result of underlying social and/or linguistic forces ning, the changes are intro-
countries becoming more heterogeneous, cosmo- that it is impossible or undesirable to tamper with. duced into the structure (or

politan, and internationally aware, it is not possible We should 'leave our language alone' (p. 178). corpus) of a language/
variety - as when changes
to rely on the slow course of natural linguistic evo- However, language planning studies have shown are proposed in spelling, pro-
lution to resolve the many pressures and conflicts that it is quite possible for social groups to alter nunciation, grammar, or vo-
that arise. Many governments, accordingly, try to the course of a language, and that the question cabulary. In status planning,
solve their problems by engaging in conscious, of desirability is a highly controversial one. It is changes are proposed in the
way a language variety is to
principled 'language planning', or 'linguistic engi- still unclear how far languages can be permanently
be used in society (thus al-
neering'. influenced by social manipulation, but there is now tering its status) - as when
Language planning involves the creation and strong evidence that such factors must be taken it is permitted for the first

implementation of an official policy about how the seriously when considering historical linguistic time in law courts or in official
languages and linguistic varieties of a country are matters. publications. The distinction
is not clear-cut, because not
to be used. Decisions of a fundamental nature may
all kinds of planning activity
need to be made, especially in the developing coun- can be neatly classified in
tries. But planning issues are to be found in all this way, but is widely en-
it

countries, as people debate such topics as the place countered in language plan-
ning research.
of minority languages, the role of an academy in
safeguarding standards (§1), the influence of the
media on usage (p. 392), the value of spelling
reform (p. 215), the avoidance of sexist language
Planning in practice
(p. 46), the modernization of religious language
Selecting the norm medium of national or inter- veloped, for use on radio or
(p. 384), the need for plain English (p. 378), stylistic If several languages are in the press. Decisions will
national communication. If

standards in publishing (p. 388), and the mainte- spoken within a country, it
nance of oracy and literacy levels in school (§44). is usually necessary to
choose a single language
Language planning is carried out by a variety
as a norm for official, edu-
of government departments and agencies, aca-
cational, and other pur-
demies, committees, popular societies, and indivi- poses. It may prove
pos-
duals. Activities range from the political and sible to use one of the
judicial, at one extreme, to the unofficial and illegal, indigenous languages, but
intergroup rivalry may
at the other. Popular attitudes towards planning
make it necessary to intro-
proposals include everything from complete sup-
duce a language from else-
port, through partial approval, general indiffer- where as a lingua franca
ence, and mild antagonism, to total antipathy. (e.g. Hindi in India, English
the language has pre-
viously existed only in
spoken form, or in an un-
usual writing system, an al-
phabet will have to be de-
vised, along with rules of
spelling and punctuation.
An early aim will be the
codification of the pronun-
ciation, grammar, and vo-
cabulary to provide a set of
norms for standard use, es-
pecially if there is a great
need to be made about
new or uncertain usages,
especially in technical con-
texts (e.g. how to abbrevi-
ate scientific terms).
Implementation
The chosen standard will
need to be officially imple-
mented, by using it for gov-
ernment publications, in
the media, and in schools.
Inevitably, will come to be
it

Historical, political, economic, religious, educa- in Ghana), which case

in deal of local variation. viewed as the best' form of
the relative merits of these language in the speech
tional, judicial, and social factors all have to be
languages will need to be community (§1 ), because it

disentangled. As a consequence, it is hardly surpris- debated. In addition, it may Modernization will be associated with edu-
ing that those who study this subject have not yet be necessary to choose a The vocabulary will need to cational progress and so-
reached the stage when they can explain why some particular variety of a lan- be modernized, toenable cial status.It will also pro-

guage (Part m), or to con- foreign material (in such vide the norm for literary
planning proposals succeed, whereas others fail.
struct a new variety, taking areas as science, medi- style, and may be associ-
The field of language planning, which dates only into account such factors ated with factors of a na-
cine, or the consumer
from the 1960s, is still largely at the stage of des- as formality, social class, society) to be translated in tionalistic, cultural, or reli-
criptive enquiry, with
continuing need for
a regional dialect, and pre- a consistent way. Princi- gious kind. due course,
In

detailed case studies of the widely differing situa- vious literary use. ples will have to be agreed it is likelybe promul-
to
for the introduction of new gated as a norm through an
tions in individual countries; few general theoreti- official body, such as an
Codification terms; for example, should
cal principles have been proposed. However, the If an indigenous language they be loan words, or coin- academy, or through pre-
area continues to attract a great deal of interest, is chosen, it will need to be ages based on native scriptive grammars, dic-
for both applied and theoretical reasons. developed to meet the de- roots? New styles of dis- tionaries, and manuals of

Most obviously, its findings and analyses may mands placed upon it as a course may need to be de- usage.

364 X LANGUAGE IN THE WORLD

 Inventing an alphabet
One of the first tasks facing Political, religious, and about whether to invent
explorers, missionaries and other considerations may combine letters into
letters,
administrators, when they affect the choice of which digraphs, or go in for dia-
encounter a new language, kind of alphabet to adopt. A critics (such as accents). If
is to devise a means of writ- community may wish to the first path is taken, there
ing down. The basic
it 'align' with countries that use is still the question of
ensure
linguistic task is to Roman, Cyrillic, Arabic, or whether the new forms
thateach phoneme is repre- other alphabets. It may also should be adaptations of
sented by a grapheme be important to choose a set familiar letters, or totally
(§§28, 33). But there are of characters that can be fresh inventions (as in the
hundreds of possible used by all the languages use of some phonetic sym-
graphemic shapes: /tf/, for throughout an area (as in the bols).
example, could be written as case of the All-India Alpha- Many other questions
c, 6, 6, ch, ts, tch, and in bet). Written uniformity is need be considered. For
to
many other ways. The often a powerful political example, if some features of rdsumd,
choice between them in- symbol. It is also an econo- a language are only occasion-
volves factors of a psycho- mical measure, as it reduces ally used to contrast mean-
logical, historical, social, and the costs of printing and ings (as often happens with
educational kind. Language- word processing. the tones of a tone language,
planning principles thus For a language where p. 1 72), should they be sys-
need to be borne in mind there are many new sounds, tematically represented by
from the outset. a decision has to be made some form of symbol, or
Alphabets in conflict
The Roman alphabet has
been so successful that it
has begun to threaten the
status of other alphabets.
question mark hangs over
the future of Chinese char-
acters, now that the roman-
ized system known as
yinhas been brought into
use (p. 313). And in India,
there is a body known as
Roman Lipi Parishad
(RLP) campaigning
adoption of the Roman al-
phabet for the main lan-
guages of the country.
The arguments are com-
plex ones, as can be seen
from the situation
• The RLP argue that the
country cannot afford the
luxury of making machines
for each of the alphabetic
scripts used in India
(p. 203). Already,
70% of mechanical type-
writers in India are
for English, and the
all the other scripts. Elec-
Chinese language planning Some of the most ambitious
programmes of language planning ever conceived have taken tronic typewriters are
place China since the 1 950s, with hundreds of millions of
in
people affected. The two main developments have been the
provision of a romanized alphabet {pin-yin), and the
promotion of a common spoken language, putdnghua, to Capitals in Frisian?
provide a means of communication between the various
Frisian, spoken in
regional languages (p. 312). Reports of early progress in the
dialects in the northern part
campaign are illustrated by Datian county in Fujian province,
of Schleswig-Holstein, pro-
which has over a dozen dialects, and where it was said that
vides a good example of
'people separated by a blade of grass could not understand
the way the invention of
each other'. A group of officials from the north on one
spelling rules can
occasion needed as many as seven interpreters to make a
social forces. In devising
speech to the people in this area. But after an active teaching
an orthography for the lan-
campaign, officials using putdnghua were able to address
guage, the question arose
large crowds without any interpreter being needed. The
as to whether nouns should
picture shows a pin-yin class taking place in an experimental
school in Ningwu County, Shanxi.
new they be ignored? Should
grammatical differences be
represented in the spelling
(as the case of English find
in
vs fined)? And how should
loan words, with their distinc-
tive phonology, be written
down? Even a well-estab-
lished writing system can be
faced with problems of this
kind, as in the continuing
debate over whether French
loans in English should keep
their accents {role, cliche',
etc.).
can
only for English.
• The RLP point to the
need to anticipate the fu-
A ture use of computers, in
relation to the country's
economy. The Roman
script is easier to adapt
pin- electronic screens and key-
boards than the various
Devanagari scripts. A
larger dot-matrix system
(p. 193) would be needed,
for the to cope with the diacritics
that are used above, be-
low, preceding, and follow-
ing the Devanagari letters.
• It is claimed that there
is a greater demand for
in India. material in the Roman
alphabet. In Bombay,
example, there was an
experiment in which tele-
phone were
directories
printed using both the Eng-
lish and Devanagari alpha-
some bets. There was a demand
of 300,000 for the former;
made but less than 50% of the
rest for 5,000 Devanagari copies
were sold.
made • On the other hand,
several be written with a capital
ter,as in German, or with
a lower-case letter, as in
other languages. Support
for the capital letter propo-
reflect sal came from those who
wished to see Frisian's ties
withGermany streng-
thened. Opposition came
from those who wished to
A planning myth
Probably the best-known
myth in the history of lan-
guage planning is the story
that German nearly became
the national language of the
U.S. in the 18th century, los-
ing to English by only one
vote in the legislature (the
'Muhlenberg' legend). In fact,
all that was involved was a
request, made by a group of
Virginia Germans, to have
certain laws issued in Ger-
man as well as in English.
The proposal was rejected
by one vote, apparently cast
by a German-speaking Luth-
eran clergyman, Frederick
Muhlenberg (1750-1801).
But the general status of
English as the major lan-
guage was never in doubt.
(After S. B. Heath & F.
Mandabach, 1983.)
Roman script is not ac-
cepted as an alternative
by any of the 22 Indian
languages recognized by
the Sahitya Akadami, the
highest body devoted to
to literature.
• The cultural identity of
the main groups in India is
very much bound up with
the use of an individual al-
phabet.
Opponents therefore argue
that the adoption of Roman
script would diminish one
of the most important sym-
bols of identity (§§9-10),
and perhaps be the thin
for end of the wedge towards
the eventual supplanting of
indigenous scripts. These
are highly emotive issues,
and it remains to be seen
whether the economic
arguments will be able
tomake much progress,
given the highly charged
atmosphere of linguistic
debate in present-day
the India.
let- see a more autonomous fu-
ture for Frisian. The issue
remains unresolved, with
both groups arguing the
relative merits of each posi-
tion, and producing publi-
cations that follow their
favoured orthographic prin-
ciple. (After A. Walker,
1984.)
61 LANGUAGE PLANNING •
365
Educational policy BILINGUAL PROGRAMMES Positive policies
There has been an extremely rapid growth in bi- In 1 982, the Commission of

One of the most important ways in which a lingual education programmes, with reference to the European Communities

country's language policy manifests itself is in the minority languages, in recent years. Over SI, 000 to the Council of Europe
asked each EC member
kind of provision it makes for the linguistic educa- million have been spent in this area in the USA
country to report on the pro-
tion of children. Which languages and language alone. However, the reasons for introducing such gress it had made in provid-
varieties are to be taught in schools, from what programmes vary greatly. In some countries, the ing education for the children
age, and for how long? These questions are only aim is to find a single language capable of unifying of migrant workers. Several
signs of progress were
partly answerable with reference to the field of for- the nation (e.g. Bahasa Indonesian). In the USSR,
apparent. Special reception
eign-language learning and teaching (§62); far the teaching of Russian to speakers of regional lan-
classes have now been
more fundamental are factors arising out of guages promotes ideological assimilation and created in many countries,
government policy and popular opinion, where a national solidarity. The teaching of English in and a number of types of
wide range of positions is found. Languages can many African countries ensures greater access to provision exist at primary
level (there is much less at
be actively promoted, passively tolerated, delibera- world opportunities. The teaching of local African
secondary level). In France,
tely ignored, positively discouraged, and even languages to minority groups in South Africa dur- for example, the language
banned. ing the 1950s and 1960s was felt to be a way of and culture of origin may be
The results of active promotion are most clearly consolidating the divisions within that society in taught to immigrant primary
shown by the progress of English towards world- relation to the Homelands policy of that time. In school children for three
hours each week, under the
language status (§59). Many countries encourage the USA, the primary concern is to guarantee the
heading of environmental
the teaching of English in school, often at the civil rights and equal opportunities of minority- studies', as long as they
expense of other languages: a recent case is Spain, groups - have been confirmed several
rights that come from countries with
where the early 1980s saw the widespread replace- times in the US Supreme Court since the 1970s. which France has a bilateral
agreement. In Belgium,
ment of French by English as the first foreign lan- In all cases, it should be stressed, bilingual educa-
when the number of immi-
guage. At the other extreme, there are many tion is not simply a matter of language learning: grant children in a nursery
examples of languages receiving official disappro- it involves the acquisition of all the knowledge and exceeds 30%, the number of
val: a clear case is the reluctance of several coun- skills that identify the minority culture. these children is multiplied
tries to teach German since the Second World War by two in order to establish
Maintenance vs transitional the quota of nursery
(e.g. in France, Italy, Israel).
teachers required. (See also
The fortunes of minority languages are closely Bilingual programmes have always attracted con-
P- 37.)
bound up with the political aspirations of their troversy. Two main views are argued (with many
speakers, and the extent to which the government variant positions). On the one hand, maintaining The 7.44 from Mons to Brussels
Belgium has also taken
of the day perceives these to be a threat. Again, the mother tongue is said to develop a desirable
up an idea first tried out
the whole gamut of official attitudes can be found. cultural diversity, foster ethnic identity, permit on a commuter train from
There may be a strong local government policy of social adaptability, add to the psychological secur- Brighton to London - lan-
language maintenance (p. 360), as happens with of the child, and promote linguistic (and perhaps guage learning by train. In
ity
May 1 984 the last carriage of
the teaching of Welsh in Wales. On the other hand, even cognitive) sensitivity. To achieve this, bi-
the 7.44 a.m. Intercity train
there are many instances of languages being dis- lingual instruction needs to be retained throughout from Mons to Brussels (a
couraged (e.g. Gaelic in 16th-century Ireland) or the whole of a child's school career. On the other French-speaking area) was
banned Catalan in Franco's Spain).
(e.g. hand, it is pointed out that a permanent dual- reserved for passengers who
wished to learn Dutch or
Official attitudes today are generally sympathe- language policy may foster social divisions
English. They paid their nor-
tic, with an increasing number of countries sup- and narrowness of outlook (through ethnocentric mal fare plus a small fee for
porting 'at least in principle) a bilingual or multi- churches, media, schools, etc.); the children may the tuition, given by teachers
lingual educational policy. In Europe, an EEC become 'trapped' in their mother tongue, and fail trained by the Belgian Centre
Council Directive has asked member states to pro- to achieve in the majority language, thus reducing d'Animation en Langues.
The venture has proved to
mote the 'teaching of the mother tongue and cul- their access to prosperity; and where there is inade-
be extremely successful. By
ture of the country of origin' to immigrants (p.
all quate teacher preparation, timetabling, and mater- the beginning of 1987, there
370). Progress varies greatly from country to ials, they may fail to achieve in their mother tongue were four language trains' in
country, however, with some countries (such as also. They should therefore be educated in their operation, with more routes

Britain) providing immigrants with relatively little being considered.

by way of mother-tongue education, and others
(such as Sweden) providing a great deal (p. 37).
Conflict is never far away, as progress towards
linguistic recognition inevitably proves to be too
slow for some people, and too rapid for others.
Vocal and vigorous objections to educational
linguistic policies are thus commonly encountered
all over the world. Regrettably, only the most vio-
lent tend to attract international attention.
mother tongue only until they are able to continue
in the majority language. Although many bilingual
programmes subscribe in principle to the former,
'maintenance' view of bilingual education, in prac-
tice the majority (in the USA and Britain, at least)
— though often
are of the latter, 'transitional' type
accompanied by maintenance elements (e.g. in
dance) in a continuing parallel
literature, music,
teaching programme.

366 •
X LANGUAGE IN THE WORLD
 These views about the nature of bilingual educa-
tion continue to be emotionally debated, for they
reflect fundamentally different conceptions of the
kind of nation people want to see around them.
Maintenance views anticipate a society that is char-
acterized by cultural pluralism and linguistic diver-
sity; transitional views look towards a culturally
homogeneous society, characterized by minority
assimilation and language 'shift' (p. 360). How-
ever, the issues are more complex than this simple
opposition suggests, for there are many kinds and
degrees of support for both positions, and com-
promise views have also been proposed.
Case studies show that the notion of 'language
loyalty' is never a simple one. For example, within
an immigrant group, some members may wish
(with varying degrees of conviction) to have their
children retain their linguistic identity; others may
wish them to 'shift' to the majority language as
quickly as possible in order to participate fully in
the new society; and yet others may wish to have
them use their new language in public, but to retain
their mother tongue for a range of private occasions
(e.g. home, church, club). There are many further
immersion' teaching programme which follows the
proportions of the primary school.
The approach continues to attract support. The
children acquire a much higher level of competence
than they would through traditional teaching
methods (though this is still a long way from native-
like proficiency, and doubts have been expressed
over how effective the children's linguistic skills
are outside of school). Their attitudes towards
French-Canadian people tend to be more positive
(though evidence is mixed on this point). And their
mother-tongue abilities do not seem to suffer from
the experience, but may even improve in certain
respects. However, for this last outcome to be cer-
tain, there needs to be a supportive and strong first-
language environment in the community. With
speakers of minority languages where the home
linguistic situation is weak or unstable, an immer-
sion programme would be unlikely to result in
maintenance, but would probably hasten the pro-
cess of assimilation to the majority language.

possible positions, reflecting the different in-

fluences of racial, geographical, political, cultural,
economic, religious, and other factors. Given the
emotional and volatile state of mind with which
people approach the problem, it is impossible to A French immersion class taking place at a junior school in
say whether a genuine 'integrated pluralism' can Montreal.
be achieved in modern educational programmes.
But the need to counter this inherent subjectivity
makes it even more desirable to carry out detailed
and objective studies of linguistic attitudes, within
the field of language planning.

LANGUAGE IMMERSION
In Quebec, in the 1960s, a new kind of bilingual
education programme was introduced, which has
since proved to be popular and successful. The pro-
posal came from the English-speaking minority,
who wished to make their children proficient in
a second language, French, in order to cope with
a society where the role of French was becoming
increasingly dominant. The idea was to arrange
for the whole of their children's first encounter with
schooling to be in the second language (§62) - a
programme of 'immersion'. The children would
speak in their mother tongue to a bilingual teacher,
who would reply in the second language. Gradu-
ally, the children would come to use the second
language themselves. Then, at a later stage, English
would be introduced into the classroom.
After several years of experimentation, two pat-
terns have come to be established. 'Primary' immer-
sion starts at kindergarten, entirely in French.
Gradually an element of English is introduced, until
by mid-primary level the children are taught 60%
in English and 40% in French. 'Secondary' immer-
sion usually starts in the first year of secondary
school with a booster year in which all teaching
is done in French. This is followed by a 'post-

61 LANGUAGE PLANNING 367

 62 Foreign language learning and teaching

To many people, the most obvious way of reducing important individual differences among learners, Language - and
some of the power of the language barrier ($56) especially in personality and motivation, that can
culture
is to promote the teaching and learning of foreign directly influence the teaching outcome. In this
Today, learning a foreign
languages in a variety of child and adult educa- view, people are seen to be largely responsible for language is likely to mean
tional settings. This widely practised approach is their own progress. Research is therefore now di- learning a great deal about
undoubtedly very successful, as can be judged by rected not only at the way teachers teach, but also the foreign civilization and
culture at the same time.
the millions who succeed in mastering a foreign at the way learners learn.
Books and materials increas-
language, even to levels that are comparable to The term 'acquisition' (Part vn) is sometimes ingly incorporate information
those achieved by 'natural' bilinguals (§60). Eng- used to replace 'learning' in this context, when the about such matters as the
lish-speaking monoglots often express amazement emphasis is on the natural, unconscious way in physical geography, econ-
at the linguistic proficiency displayed by foreigners which a learner can assimilate a foreign language omy, history, politics, reli-
gion, social institutions, edu-
— not least, the standards routinely achieved in Eng- (as in bilingual contexts, or when using one of the
cational system, literature,
lish — and conclude that foreigners must have a 'natural' approaches to FLT, p. 373). In several
art,music, science, tech-
'gift' for language learning, which they lack, or that approaches, however, 'acquisition' and 'learning' nology, media, and sport, as
English must be a particularly easy language to are carefully distinguished: the former is then res- well as about daily life-style,

learn. There is no basis for these suggestions. A tricted to what takes place in 'natural' learning popular beliefs, folk customs,
and social values. The
few gifted language learners do exist (p. 362), but situations; the latter to what takes place in class-
material is inevitably very
most people arrive at their fluency only as a result rooms when following a structured course with a selective; but it helps the
of hard work, expended over a considerable period teacher. learner tobecome more fully
of time. aware of differing ways of
On the other hand, there is also a great deal FOREIGN AND SECOND behaviour, and reduces the
risks of culture shock,
of educational failure and lack of achievement in Several terminological distinctions are drawn with- foreigner stereotyping, and
the language-learning field, which also requires A person's 'mother tongue' or 'first
in this field. intolerance.
explanation. Many from a variety of
people, language' (LI) is distinguished from any further A cultural frame of refer-

linguistic backgrounds, are acutely embarrassed by languages that may be acquired (L2, L3, etc.). The ence becomes increasingly
important the greater the
their linguistic inadequacy when travelling abroad, term 'foreign language' is popularly used to refer
'distance' between lan-
and wish to overcome it. Many have tried to learn to any language that is not a native language in
guages. To succeed in an or-
a foreign language, but have made little progress a country; and 'second language' is also commonly iental language, for example,
in it. 'I was never very good at languages in school' used in this way. But many linguists distinguish a westerner needs the sup-
is a widely heard complaint. It is therefore impor- between 'foreign' and 'second' language use, recog- port of several of the above
studies. But a cultural per-
tant to study the factors that govern success or fail- nizing major differences in the learning aims, teach-
spective is needed even with
ure in this field - such
soundness of teaching
as the ing methods, and achievement levels involved. nearby' languages, in order
methods, the attitudes and motivation of the A foreign language (FL), in this more restricted to grasp the social signifi-
learner, the availability of time and opportunities sense, is a non-native language taught in school cance of a linguistic feature
(e.g. slang, accents, or terms
to learn, the adequacy of resources, and the chance that hasno status as a routine medium of communi-
of address, p. 44) or to follow
to put the language to active use. It is evidently cation in that country. A second language (SL) is
the subject matter of daily
a complex situation which, in view of the enormous a non-native language that is widely used for pur- conversation. For example,
amounts of time and money expended within the poses of communication, usually as a medium of in every country, knowing the
foreign language 'industry' all over the world, war- education, government, or business. English, for names of the most famous
men and women of a culture,
rants careful investigation. And in recent years, the example, has foreign language status in Japan, but
whether they are political
subject of foreign language teaching and learning second language status in Nigeria. The latter term figures, folk heroes, or media
has in fact developed to become today the largest is also used with reference to immigrants and indi- stars, a major factor in
is

domain of enquiry within applied linguistics (§65). genous groups whose LI is a minority language: really understanding the

in the USA, for example, English is a second lan-

meaning of a newspaper re-
port, a debate on television,
guage for millions of immigrants from a wide range
TEACHING AND LEARNING of language backgrounds (p. 361) as well as for
or the course of conversa-
tion.
The use of two headings for this section, foreign
speakers of American Indian languages.
language teaching (FLT) and foreign language
learning (FLL), reflects an important development
in the modern study of the subject. FLT was at
one time thought to be exclusively a matter of
teaching techniques; it was felt that, if teaching
Boulangerie = bakery?
was above a certain minimum level of efficiency, At a purely linguistic level, French boulangerie trans-
learning would automatically follow. Teaching lates into English as bakery. However, there is no cul-
was the active skill; learning, the passive one. equivalence between the two notions: in many
tural

Today, the active role of the learner French villages, the bakery acts as a social centre in
is an estab-
ways its English counterpart does not.
lished principle. It is recognized that there are

368 X LANGUAGE IN THE WORLD

Why learn foreign languages? the figure
are much
is around 85%. In Britain, the figures

lower, but the climate is slowly changing.

The question requires an answer,
we frequently find indifference or hostility ex-
pressed towards foreign languages and foreign
people, where teaching resources are limited, and
where other subjects clamour for extra slots within
the school timetable. The criticisms come mainly
from within the English-speaking world, where
in a world where In the 1960s, only 20% of schoolchildren took a
foreign language; in the early 1980s, well over 80%
were taking one for up to three years. In the USA,
a Commission on Foreign Language and Interna-
tional Studies was set up in 1978 to consider the Which language?
FL situation: it concluded that American incompe- In most non-English-speak-
tence in FLL had reached the stage where it threat- ing countries, English is first

FLT has often been attacked on the grounds that ened national security and economic development choice (§59). In English-
speaking countries, this posi-
the time would be better spent on science, mathe- (e.g. only 2% of American scientists could under-
tion is usually taken by
matics, or the mother tongue. Many people think stand material published in Russian). Several French, which has a highly
that FLT is unnecessary in a world where an recommendations have since been made to improve and cul-
prestigious literature
increasing number of people understand English the status and facilities for FL work, at both school ture,and which has been
(§59). and college levels. Extra funding has been allo- used as an international
lingua franca since the 1 8th
Moreover, even where FL instruction
in places cated, both federally and privately. Some states
century (e.g. the official lan-
is provided, the use of traditional teaching methods (e.g. Michigan) have already mandated an FL com- guages of the Council of Eur-
has meant that many pupils find FL work boring ponent as part of high school certification. ope are English and French).
and difficult. In British secondary schools, for In Europe, the Committee of Ministers of the In Britain, the proximity of
example, 60% drop their FL after three years, and France and numerous points
Council of Europe has recommended that FLT in
of historical connection have
even those who pass their exams are often unable schools should be increased and diversified, that led to the concentration on
to use the language for everyday purposes. Such children should learn more than one foreign lan- French in schools - and,
facts fuel the arguments of those who
think that guage if possible, that they should start as early once a subject is estab-
FLT should become a minority subject or even be as possible, and that facilities should be made lished, the need to provide
continuity of teaching makes
dropped from the curriculum altogether. longer-term. There should be a single language
it difficult to displace.
Arguments of kind are rarely encountered
this policy for a school, in which all language work After French, Spanish and
in non-English-speaking countries, where there is (LI and L2) should be integrated. In the 1980s, German both have substan-
a great demand for FL courses. In West German several language associations and committees have tial followings in schools, the

former especially in the U.S.

secondary schools, for example, all pupils take a reiterated this plea, though limited funding has led
because of the proximity of
foreign language to an advanced level. In France to limited implementation. Spanish-speaking countries
and the high level of immig-
ration from these areas.
Each language has consider-
able international status.
Spanish is the world's third
international language, being
used over 20 countries.
in

German is an important

lingua franca throughout

much of Eastern Europe.
It is extremely difficult to

Answering the critics predict which languages chil-

The FLT world has been
slow meet the challenge
to
of the critics. An enormous
outpouring of intellectual
and practical effort
not idly changing society - not thinking about things. Lan- working hours add to dren will need most in adult
life. Patterns of language
only abroad, but within their guage learning, as well as people's leisure time. Tour-
own community. It can help broadens the mind.
travel, ist travel is a major motiva- choice in the various settings
overcome their insecurity • Success in the inter- tion, but many have come to of adult education are quite
has been and develop their confi- national world of commerce find FLL a satisfying leisure different from those found in

devoted to overhauling the dence as they face up to the and industry is becoming activity in its own right, school. A recent BBC TV Ita-
traditional machinery of lan- demands of social and per- more and more dependent enabling them to have direct lian series was watched by
guage teaching. At the 1 .8 million people, and a
sonal relationships not on FLL (§56). Young people access to the world of for-
same time, the rationale for usually encountered in a now find they have more eign cinema, radio and tele- Russian series by about 1
FLL has come to be publicly mother-tongue context. career opportunities when vision, vocal music, litera- million people, though

defended. • There is no doubt that they know a foreign lan- ture, and the history of neither language is much
• FLL is no longer a luxury, language is prerequisite for guage and are increasingly ideas. taught in British schools.
in an international world. It full mutual understanding moving to localities where • FLL provides a valuable Trade and tourism seem to
is a necessity, a country is
if and cooperation between some degree of FL com- perspective for those whose be particularly influential fac-
to exercise a role in world nations. FLL promotes petence is required of them. interest is primarily in the tors.

affairs. Especially in Europe, understanding, tolerance, This mobility is no longer mother tongue. Ultimately,
it is seen as a criterion of and respect for the cultural something that affects only the only way to appreciate
responsible international identity, rights, and values executives but is found with the unique identity and
citizenship, it is a strength to of others, whether abroad, all grades and categories of power of a language is by
be able to meet people from or at home in minority personnel, such as market- contrasting it with others.
other countries on equal groups. People become ing staff, legal specialists, • FLL a primary educa-
is

linguistic terms. less ethnocentric, as they secretaries, and tech- tional right,which should be
• FLL has an essential role come to see themselves nicians. made available to all peo-
in preparing children to and their society in the eyes • FLL is becoming increas- ple, whether they avail
cope with the new perspec- of the rest of the world, and ingly important as un- themselves of it or not.
tives brought about by a rap- encounter other ways of employment and reduced

62 FOREIGN LANGUAGE LEARNING AND TEACHING •

369
 .

THE COUNCIL OF EUROPE specific recommendations to do with language

Recommendation R (82) 18 of the Committee of learning in schools and higher education, language
Ministers of the Council of Europe, adopted in Sep- learning by migrants and their families, initial and
tember 1982, is a clear statement of the issues further teacher training, and measures of interna-
involved in the teaching and learning of modern tional cooperation. The premisses, and the state-
languages. The statement recognizes three general ment of general measures to be implemented, are
premisses, following these by a set of general and given below.

The Committee of Ministers . .
Considering that the rich heritage of diverse languages
and cultures Europe is a valuable common resource
in
to be protected and developed, and that a major
educational effort is needed to convert that diversity
from a barrier to communication into a source of
mutual enrichment and understanding;
Considering that it is only through a better knowledge
of European modern languages that it will be possible
to facilitate communication and interaction among
Europeans of different mother tongues in order to
promote European mobility, mutual understanding and
co-operation, and overcome prejudice and
discrimination;
Considering that member states, when adopting or
own country) as well as the skills in the use of those
"
languages that will enable them to satisfy their
communicative needs and in particular:
1.1. to deal with the business of everyday life in
another country, and to help foreigners staying in their
own country to do so:
1.2. to exchange information and ideas with young
people and adults who speak a different language and
to communicate their thoughts and feelings to them;
1.3. to achieve a wider and deeper understanding
of the way of life and forms of thought of other peoples
and of their cultural heritage.
2. To promote, encourage and support the efforts of
teachers and learners at all levels to apply in their own

developing national policies in the field of modern situation the principles of the construction of language-
language learning and teaching, may achieve greater learning systems (as these are progressively developed
convergence at the European level, by means of within the Council of Europe 'Modern languages'
appropriate arrangements for ongoing co-operation and programme):
co-ordination of policies: 2.1. by basing language teaching and learning on
Recommends the governments of member states, in the the needs, motivations, and characteristics and
framework of their national educational policies and resources of learners;
systems, and national cultural development policies, to 2.2. by defining worthwhile and realistic objectives
implement by all available means and within the limits as explicitly as possible;
of available resources, the measures set out in the 2.3. by developing appropriate methods and
appendix to the present recommendation. materials;

Measures to be implemented concerning the learning 2.4. by developing suitable forms and instruments
and teaching of modern languages for the evaluation of learning programmes.

General measures 3. To promote research and development programmes

1. To ensure, as far as possible, that all
their populationshave access to effective means of
acquiring a knowledge of the languages of other
member states (or of other communities within their
sections of leading to the introduction, at all educational levels, of
methods and materials best suited to enabling different
and types of student to acquire a communicative
classes
proficiency appropriate to their specific needs.

The Committee of Ministers of

the Council of Europe
meeting in Strasbourg

370 • X LANGUAGE IN THE WORLD

 When should L2s
Successful language learning be taught?
There is as yet no single techniques of memoriza- • Teaching objectives tinue withthem after school. Traditionally, L2s have been
theory that can account for tion, and alternative ways of need to be carefully se- Special arrangements may introduced at a relatively late
the diversity of FLL be- expressing what they want lected and graded, to permit need be made, involving
to stage of development -
haviour, and explain why to say (paraphrase). They progress with
realistic interschool and local gov- usually around the age of 10
some learners succeed in may also benefit from train- underachievers, as well as ernment cooperation. New or 1 1 In recent decades, the
.

their task, whereas others ing in the kinds of basic with the gifted. Different combinations of subjects, benefits of an early start
fail. A hint of the complexity skills involved in FLL, such kinds of objectives should more suited to the needs of have been urged, given the
of the task facing re- as those identified above. be explored. Are all four modern society, should be natural way in which young
searchers can be obtained • Exposure to the foreign linguistic modes to be intro- introduced, such as FL + children learn language (cf.
from the resume of relevant language needs to be regu- duced (speaking, listening, science, FL + economics. critical periods, p. 263), the
factors on this page. lar - a problem which par- reading, writing), and if so, • Motivation is a central positive results of immersion
• It is unclear how far there ticularly affects FLT in in which order? Might lim- factor. Students need to see programmes and
(p. 367),
may be a genuine aptitude schools, where timetable ited competence in two lan- that foreign languages are the likelihood that they can
for FLL. Given sufficient pressure, examinations, guages be better than an taken seriously by those devote more time to the task.
motivation, intelligence, and and holidays may lead to excellent command of one? whom they respect, espec- Several experimental FL
opportunity, anyone can discontinuities. Whenever Should the learners be ex- ially in the community at programmes have been tried
learn a language; but the possible, the aim should be posed to only certain var- large (encouragement from out in primary school, and
task is likely to be less oner- to teach and often'.
little ieties (§1 1 ) of the foreign local employers, civic inter- their effectiveness evalu-
ous if certain general perso- Too much exposure at any language? Should the focus est intown twinning, etc.). It ated.
nal qualities are present. one time can be as ineffec- be on communicative skills is critical to take the lan- The results have been
Among these, it has been tive as too little, readily lead- or on formal techniques guage out of the classroom, mixed. FLT with young chil-
suggested, are empathy ing to fatigue and superficial (such as translation) so that students see its use dren can work well, but only
and adaptability, assertive- assimilation ('quickly (p. 374)? in a native community. if learning conditions are op-

ness and independence,
with good drive and powers
of application. People need
to be capable of assimilat-
ing knowledge in difficult
conditions. They should
have a good memory, and
be good at finding patterns
in samples of data (non-
linguistic as well as
tic). Of particular impor-
tance is an ability to detect
phonetic differences (e.g. of
stress, melody, vowel qual-
ity) - something which can
manifest itself in other do-
mains, such as drama or
music.
• Students can benefit
from being taught to learn
how to learn' foreign lan-
guages - useful strategies,
learned is quickly forgotten').
• Exposure to native users
of the foreign language is a
real benefit, through the use
of authentic materials (e.g.
audio tape, video tape,
newspaper library) and, in
school, foreign language
teaching assistants. An im-
linguis- portant dimension is
of educational visits
- but these need to
perly prepared and fol-
lowed up in class, and the
experience should enable
children to be genuinely
integrated within the
vironment. Out-of-school
activities should be encour-
aged, such as pen friends,
private exchanges, and
weekend culture simulation
• Teaching methods need
to be flexible to suit the
needs of individual children
(e.g. their interests and cog-
nitive skills) and to make
best use of classroom de-
sign and resources (e.g. the
availability of audio-visual
aids). There is
the use mula' for successful
abroad (p. 374). There should be
be pro- opportunities for teachers to
interact with children in
groups, pairs, and individu-
ally. If classes are too big,
it will be impossible to ob-
FL en- taingenuine participation
and practice.
• There should be an op-
portunity to take
one foreign language
school, to follow
Moreover, motivation ap- timal. The teaching objec-
plies to teacher as well as tives need to be limited,
student: it is difficult to teach graded, and clearly defined.
enthusiastically if it is known Specialist teachers need to
thatmost of the class are be available. Methods need
going to drop their language to be devised that are appro-
at the earliest opportunity, priate to the interests and
or that society places little cognitive level of the chil-
no single 'for- store by it. dren. And the transition to
FLT • Teacher training needs to secondary school needs to
continue at in-service as be borne in mind, because a
well as initial levels. lack of continuity can negate
Teachers need to be tech- previous work. Unfortunate-
nicallycompetent, that is ly, these conditions do not
they are able to teach in the often obtain, and many early
foreign language, if re-
FLT projects have achieved
quired. They need to keep disappointing results (the
themselves up-to-date with greatest success coming in
the latest research into their second language situations,
more than language and society, as such as in Africa).
in well as in FLT techniques. However, even if formal
them to an FLT is not introduced, it is

such as silent rehearsal, courses. advanced level, and to con- possible to develop
still

young children's general lan-

guage awareness - to
sensitize them towards the
existence and variety of the
Pronunciation languages of others as well
j 5 as of their own language -
Interlingual distance The structural closeness of and to foster the enjoyment
F- -4 languages each other has often been thought to be an
to that can come from being in
important factor in FLL. If the L2 is structurally similar to contact with foreign lan-
the L1 is claimed, learning should be easier than in cases guages. Children can learn
R- i
, it

where the L2 is very different. However, is not possible it FL games, songs, rhymes,
Spelling Grammar
to correlate linguistic difference and learning difficulty in any sayings, everyday greetings,
G --2S straightforward way (p. 372), and even the basic task of and many basic notions (e.g.
quantifying linguistic difference proves to be highly complex, counting, parts of the body,
1 I -i
because of the many variables involved. telling the time). In particular,

The diagram shows one analyst's informal estimate of the pupils from other language
if

-
S structural distance between English and several other backgrounds are present,
E English
languages. Pronunciation, spelling, choice of alphabet the multilingual setting can
F French q F l

(orthography), grammar, and vocabulary are each rated be used to generate a mutual
G German /
separately, using a scale from 1 (least distance) to 5 linguistic and cultural inter-
z
1 Italian -/
(greatest distance). On comes closest to
this basis, Italian est. The experience can pro-
R Russian / English (scoring 6), followed by Spanish (7), German (10), vide a valuable foundation
S Spanish /-(G) French (12), and Russian (15). German would score higher for the systematic study of
4 / if Gothic script were taken into account (3 rather than 1 foreign languages at later
R
/ along the orthography scale). (After C. V. James, 1 979.) ages.

Orthography Vocabulary

62 FOREIGN LANGUAGE LEARNING AND TEACHING •

371
Theories of language learning
As with the study of language acquisition (p.
first
234), several theories of the nature of the FLL pro-
cess have been propounded, with similar issues
being addressed. Indeed, comparisons are fre-
quently made with the way children learn their first
language (LI), as a means of providing hypotheses
to guide FL research.
THE BEHAVIOURIST VIEW
A great deal of language learning and teaching in
the 1950s and 1960s was influenced by the tenets
of behaviourism (pp. 234, 408). In this view, FLL
is seen as a process of imitation and reinforcement:
learners attempt to copy what they hear, and by
regular practice they establish a set of acceptable
habits in the new language. Properties of the LI
controversial (largely because of difficulties in vali-
dating the error analysis - see below), but their
general conclusion is widely supported. The syste-
matic comparison of LI and L2, in order to predict
areas of greatest learning difficulty - a procedure
known as contrastive analysis — explains only a
small part of what goes on in FLL.
THE COGNITIVE VIEW
The main alternative to the behaviourist approach
sees as central the role of cognitive factors in lan-
guage learning (pp. 234—5). In this view, learners
are credited with using their cognitive abilities in
a creative way to work out hypotheses about the
structure of the FL. They construct rules, try them
out, and alter them if they prove to be inadequate.
Language learning, in this account, proceeds in a
series of transitional stages, as learners acquire

are thought to exercise an influence on the course more knowledge of the L2. At each stage, they are

of L2 learning: learners 'transfer' sounds, struc-

in control of alanguage system that is equivalent
to neither the LI nor the L2 - an interlanguage The monitor model
tures, and usages from one language to the other.
A widely used typology distinguishes two kinds of (L. Selinker, 1972). In the 1970s, an influential

transfer. Similarities between the two languages

Error analysis plays a central role in this view of the relationship be-

cause 'positive transfer': it proves acceptable to use approach. Errors are likely to emerge when learners tween acquisition and learn-
make the wrong deductions about the nature of ing was propounded by the
the LI habits in the L2 setting (e.g. the assumption American linguist, Stephen
that the subject goes before the verb satisfactorily the L2, such as assuming that a pattern is general,
Krashen (1941—). This ac-
transfers from English to French). Differences cause
when in fact there are exceptions. The errors pro- count recognizes a sub-
'negative transfer', generally known as 'interfer-
vide positive evidence about the nature of the learn- conscious, natural process

ence': the LI habits cause errors in the L2 (e.g. ing process, as the learner gradually works out ('acquisition'), which is the
primary force behind FL
the same assumption about subject—verb order what the FL system is. For example, learners who
fluency. Learning' is seen as
does not satisfactorily transfer into Welsh). Typical say vous disez instead of vous dites 'you say' have a conscious process that
interference errors include: I wait here since 3 assumed, wrongly, that the -ez ending found after monitors, or edits, the pro-

hours (from French) and How long must my hand vous in most other French verbs (marchez, donnez, gress of acquisition and
etc.)also applies to dire 'say'. The error in this guides the performance of
in plaster stay? (from German). Problems of nega-
the speaker. Its role is - or
tive transfer are thought to provide a major source case indicates that a faulty generalization (or ana-
should be - minor, being
of FLL difficulty. The main aim of behaviourist logy, p. 234) has been made. used only to correct errors in
teaching thus to form new, correct linguistic
is
Since the 1970s, cognitive approaches to FLL speech or to give speech a
have been in the ascendant, and error analysis in more polished' appearance.
habits through intensive practice, eliminating inter-
particular has attracted a great deal of attention. In fact, traditional FLT pro-
ference errors in the process. vides learners with a great
There are several problems presented by this ac-
However, the analysis of errors turns out to be deal of conscious knowledge
a highly complex matter, involving other factors of linguistic rules. As a result,
count of FLL. Imitation alone does not provide
who than the cognitive. Some errors are due to the they may come to rely too
a means of identifying the task facing learners,
influence of the mother tongue, as contrastive ana- much on this knowledge, so
are continually confronted with the need to create that it actually gets in the way
lysis claims. Some come from external influences,
and recognize novel utterances that go beyond the of their ability to communi-
such as inadequate teaching or materials. Some
limitations of the model sentences they may have cate. People who worry too
arise out of the need to make oneself understood much about making a mis-
practised. Nor does imitation suffice as an explana-
by whatever means possible (e.g. replacing words take, and who thus are reluc-
tion of the way learners behave: not many of the
by gestures). Moreover, not all errors are equally tant to use their FL ability,
errors that are theoretically predicted by the differ- are in this view overusing'
ences between LI and L2 in fact occur in the lan- their monitor.

guage of learners; and conversely, other errors are Theories of this kind are
inevitably controversial, given
found that seem unrelated to the LI. In a fre- Two models of foreign language learning our limited knowledge of the
quently-cited early study (H. C. Dulay &c M. K. psychological processes in-
Behaviourist Cognitive
Burt, 1973), 145 Spanish-speaking children aged volved in speech production.
L2 input obtained from Exposure to authentic use of
5 to 8 were observed while learning English. Six There is plainly a need to
controlled, formal, instruction L2 in near-natural situations
take into account the distinc-
structures were selected and the error patterns ana- 1 1
tionbetween conscious and
lysed. emerged that interference errors (such as
It Imitation and Input processed using
subconscious awareness in
They have hunger from Ellos tienen hambre) reinforcement (conscious) natural (universal,
language processing, and
strategies unconscious) strategies
accounted for only 3% of the errors made. The between formal and informal
1 1
majority of the errors (85%, with a further 12% settings, but the way these
L2 habits Transitional stages of
variables interact, it has
unclear) were thought to resemble those that established learning (interlanguage)
been argued, is more com-
appear in the course of LI acquisition (e.g. They 1 1
plex than anything which has
L2 output L2 output
hungry). Analyses of this kind have proved to be so far been proposed.

372 •
X LANGUAGE IN THE WORLD
systematic, disruptive, or unacceptable. Errors of
vocabulary, for example, are less general and pre- The natural order hypothesis
dictable than errors of grammar, but they are During the 1970s, several
usually more disruptive of communication. Some studies drew attention to
the fact that different FL
errors, indeed,become so acceptable that they do learners make similar
not disappear: they become 'fossilized' — tolerated errors, regardless of their
by learners (insofar as they are conscious of them) language background.
because they do not cause major problems of com- Such errors as / going and
this a book were observed
munication (e.g. the pronunciation errors that con-
in Spanish, Russian, Japa-
stitute a foreign accent). nese, and several other
Above all, error analysis is complicated by the learners of English. The
fact that often unclear what the learner
it is conclusion was drawn that
there must be a universal
intended to say, and thus how to identify the error
creative process at work;
that has been made. For example, does The lady
learners were said to be
eat it display an error of the or verb noun {ladies) following a natural internal
- and if the latter, should the correct form be eats, syllabus' (as opposed to
is eating, ate, or some other variant? And even if the external' syllabus of
the classroom). Several of
we assume that the speaker intended to say eats, the errors closely resem-
we are with the question of whether the
still left bled those made by chil-
error is one of pronunciation (the speaker having dren learning their mother
difficulty with the [ts] cluster) or grammar — and, tongue. Analogies were
within the latter heading, whether the difficulty is drawn with the
therefore
language acquisition de-
one of morphology (lack of awareness of the end- vice' postulated by some
ing) or syntax (lack of awareness of number agree- child-language analysts
ment between subject and verb) (§16). 234),and a parallel was
Despite the difficulties, research into errors conti- proposed between the
nues to provide a fruitful way of investigating the
processes underlying FL acquisition. However, as
with contrastive analysis, the approach cannot pro-
vide a complete explanation. Most FLL settings
do not constitute the kind of 'pure', natural linguis- Child vs adult acquisition
tic situation that is presupposed by the cognitive
The similarities between L1
approach, but contains elements of formal teach- and L2 acquisition errors
ing, in which learners are systematically introduced are striking, but there are
to fragments of the L2 (e.g. one tense at a time). many differences between
the two kinds of learning
To understand the way languages come to be situation, which makes it
learned in these 'mixed' settings, it is thus proving difficult to see a parallel be-
necessary to devise more sophisticated models, tween adult foreign lan-
which focus on the relationship between the pro- guage learners and young
children acquiring their
cesses of natural acquisition and those of formal
mother tongue.
learning, and which pay adequate attention to the • The adult has a set of
needs and aims of the students, and to the nature formed cognitive skills and
of the social setting in which FLL interaction takes strategies that should
make the FLL task easier
place.
(e.g. the ability to memor-
ize, imitate, and use dic-
tionaries). A major asset
There is something in the poetry of the ability of most adults
Mill read and write.
Wordsworth which y& always ><5 live. He • Adults already have a
language, and this inevit-
ably reduces their motiva-
Errors in language learning tion to learn another
The error in this sentence, written by a Swedish beyond minimal levels.
student, seems straightforward, but it is not easy to Migrants, for example,
say exactly what the error is, why it was made, and generally learn only
whether the teacher has made the best correction. Is enough to enable them to
the student confusing be to and will? Or has he learned survive in their new
the past tense use of be to in this context (as in There country.
was something in the poetry which was to live forever), • There are several emo-
and assumed that the present tense would work in the tional differences between
same way? If so, is there not an additional error in the adults and children when
position of a/ways?And would notforeverbe a more comes to learning. In parti-
idiomatic word? The corresponding construction in cular, adults are more self-
Swedish is som alltid skall leva, but this will not explain conscious about FLL, and
all that is going on. are less able to assimilate
62 FOREIGN LANGUAGE LEARNING AND TEACHING
natural order of L1 acqui-
sition and the way people
acquired a foreign lan-
guage.
Particular attention was
focussed on the way in
which foreign learners of
English used a set of gram-
matical morphemes (§16),
such as -ing, ed, and plural
-s, which L1 studies had al-
ready found to be acquired
in a certain order (p. 242).
The errors learners made
witheach item were
counted, and the mor-
phemes were ranked on
the basis of how accurately
they were used. This rank-
ing was then assumed to
reflect the order in which
the learners were acquiring
these morphemes. Similar
orders were found in sever-
al different FLL contexts,
both spoken and written
(p. language, thus supporting
the idea of a natural, uni-
versal sequence of acqui-
cultural differences.
• Adults meet a greater
variety of L2 situations than
do children learning their
L1 . Children's needs are
also very different (e.g.
they need language for
play and emotional expres-
sion). Accordingly, the
range of teaching objec-
tives will differ in each
case.
• The adult has less time
and opportunity than the
child forFLL. Some esti-
mates suggest that it takes
well over a year to accumu-
is late as much l_2 experience
to as a young child gets from
theL1 in a month.
• Adults invariably find
themselves in a less natur-
al learning environment
than children. It is rarely
possible to devise a teach-
ing situation which closely
resembles that encoun-
tered by the L1 child, with
its one-to-one interaction
and strong emotional
(caregiver) support.
• There is an uncertain
it parallelbetween the way in
which mothers talk to their
children and the way in
which people talk to adults
using a foreign language
sition that was independent
of the influence of the
learner's first language.
However, criticisms have
been made of this kind of
approach. Order of acqui-
sition as based on a cross-
sectional study of speech
samples may not corres-
pond to the order of acqui-
sition that would emerge
from a longitudinal study
The findings are of
(p. 229).
a
limited generality: only
very small number of gram-
matical items have been
analysed, and there have
been very few studies
(most of which to date have
focussed on English, so that
it is unclear how genuine
the claimed universals
are). And differences in ac-
quisition orderhave al-
in ready begun to emerge,
casting doubt on the uni-
versality of the natural
order hypothesis.
(foreigner talk). Certainly,
adult L1 speakers adapt to
learners, and (often un-
consciously) try to help
them by speaking slower
and louder, repeating
words, simplifying their
grammar, and using
stereotyped expressions
(of which pidgin savvy is
probably the most famous).
They also ignore many er-
rors. But it is unclear how
universal or how systema-
tic these input strategies
are.
• Similarly, it is unclear
how far teacher language
displays correspondences
with motherese (p. 235);
the differences, at present,
are more striking than the
similarities. To facilitate
learning, in the early
stages, teachers need to
keep their input relatively
simple, interesting, com-
prehensible, relevant to the
learning task, sufficiently
repetitive to enable pat-
terns to be perceived, and
capable of providing appro-
priate feedback. Generali-
zation proves difficult, given
the great variation that
exists among teaching
methods (p. 374).
• 373
Teaching methods The grammar translation method This method Communicative
derives from the traditional approach to the teach- teaching
In the long search for the best way of teaching ing of Latin and Greek, which was particularly
During the 1 970s, there was
a foreign language, hundreds of different ap- influential in the 19th century. It is based on the a widespread reaction, in
proaches, or methods, have been devised. Each meticulous analysis of the written language, in both L1 and L2 teaching
method is based on a particular view of language which translation exercises, reading comprehen- (§44), against methods that
learning, and usually recommends the use of a spe- sion, and the written imitation of texts play a pri-
stressed the teaching of
grammatical forms and paid
cific set of techniques and materials, which may mary role. Learning mainly involves the mastery little or no attention to the
have to be implemented in a fixed sequence. Ambi- of grammatical rules and memorization of long lists way language is used in
tious claims are often made for a new teaching of literary vocabulary, related to texts which are everyday situations. A con-
method, but none has yet been shown to be intrinsi- chosen more for their prestigious content than for cern developed to make FLT
cally superior. The contemporary attitude is flex- communicative', by focus-
their interest or level of linguistic difficulty. There
sing on learners' knowledge
ible and utilitarian: it is recognized that there are is little emphasis laid on the activities of listening of the functions of language,
several ways of reaching the goal of FL compe- or speaking. and on their ability to select
tence, and that teachers need to be aware of a range This approach dominated early work in modern appropriate kinds of lan-
of methods, in order to find the one most appro- guage for use in specific
language teaching. A minority still find its intellec- sit-

uations.
priate to the learner's needs and circumstances, and tual discipline appealing; but the vast majority of Increased interest was
to the objectives of the course. It is frequently teachers now recognize that the approach does lit- shown in the situations them-
necessary to introduce an eclectic approach, in tle to meet the spoken language needs and interests selves, and in the kind of lan-
which aspects of different methods are selected to of today's language students. guage the learner would be
meet the demands of particular teaching situations. likely to meet (e.g. at a bank,
eating out). Situational sylla-
Several classifications of teaching methods have
The direct methodThis approach, also known buses' aimed to recreate
been made, in an attempt to impose some degree these situations, and to
as the oral or natural method, is based on the active
of order on what is a highly diverse and idio- teach the various linguistic
involvement of the learner in speaking and listening
syncratic field. Some analysts make use of the fun-
to the foreign language in realistic everyday situa-
activities involved, such as

damental distinction between language structure requesting, thanking, com-

tions. No use is made of the learner's mother ton- plaining, and instructing.
(form) and language use (function) (§13). Under
gue; learners are encouraged to think in the foreign Notional' (or functional')
the first heading, they include those methods that syllabuses provided a major
language, and not to translate into or out of it.
focus on the teaching of formal rules and categor- alternative to the emphases
A great deal of emphasis is placed on good pronun- of formal language teaching.
ies, and that emphasize the importance of accurate
ciation, often introducing students to phonetic Here, the content of a course
written translation and the understanding of litera-
transcription (§27) before they see the standard is organized in terms of the
ture. Under the second heading, they include meanings
orthography. Formal grammatical rules and termi- (notions') learners
methods that lay stress on the teaching of active require in order to communi-
nology are avoided.
participation in natural and realistic spoken lan- cate in particular functional
The direct method continues to attract interest contexts. Major communica-
guage settings, and where the emphasis is on com-
and enthusiasm, but it is not an easy approach to tive notions include the
municative success rather than on formal accuracy.
use in school. In the artificial environment of the linguistic expression of time,
Many approaches are biased in one or the other classroom, it is difficult to generate natural learning duration, frequency, se-
direction, though it is also common to find ap- quence, quantity, location,
situations and to provide everyone with sufficient
proaches that claim to integrate the strengths of and motion. Major communi-
method have thus
practice. Several variants of the cative functions include
both positions.
some
evolved. In particular, teachers often permit evaluation, persuasion, emo-
Certain methods are widely recognized because and the
degree of mother-tongue explanation and gramma- tional expression,
of their influential role in the history of ideas sur- marking of social relations.
tical statement to avoid learners developing inac-
rounding this subject. Communicative methods
curate fluency ('school pidgin'). have attracted universal in-
Easy listening New FLT methodsare invented every day. terest, and much influenced
Many claim to provide remarkable progress - at a price. This the practice of modern FLT.
The audio-lingual method Also known as the
advertisement, taken from a South American newspaper in But there has also been a cri-
aural—oral method, approach derives from the
this
as
1 984. is typical of its genre. What makes it especially tical reaction, linguists
intriguing is its proposed integration of behaviourist and
intensive training in spoken languages given to and teachers encounter prob-
mentalist linguistic theories (p. 408)! American military personnel during the Second lems in providing a principled
World War, which resulted in a high degree of lis- basis for interrelating the
proposed notions and func-
tening and speaking skill being achieved in a rela-
tions. Of particular impor-
tively short time-span. The emphasis is on every- tance is the need to provide
MASTER IT WITH EUEVABLE? day spoken conversation, with particular attention learners with principles that
being paid to natural pronunciation. Language is willenable them to make a
BUT TRUE ana bridge between functional
R Skinner s seen as a process of habit formation (p. 372): struc-
off « F aspects of language and the
A successfuUppUcation resuued ^g system tural patterns in dialogues about everyday situa-
correct use of formal struc-
tions are imitated and drilled (first in choral speech, tures. Proponents of the ap-
then individually) until the responses
learner's proach have recognized
become automatic. There is a special focus on areas these problems, and there

By appointment
Bv * VV^
only. Please
for
c
Mrs.
^ ; of structural contrast between LI and L2. There
has been considerable dis-
cussion of the way communi-
and ask cQSt is little discussion of grammatical rules. Language

We'll gladly
^show^owj^ work is first heard, then practised orally, before
cative teaching might de-
velop in the future. (After K.
being seen and used in written form. Johnson & D. Porter, 1983.)

374 •
X LANGUAGE IN THE WORLD

The approach can instil considerable conversa-
tional fluency in a learner, and was widely used, Humanistic approaches
especially in the 1950s and 1960s.Its reliance on Suggestopedia (Georgi
drills and habit-formation makes it less popular Lozanov) An approach
today, especially with learners who wish for a based on suggestology,
the science of suggestion.
wider range of linguistic experience, and who feel
Devised by a Bulgarian
the need for more creative work in speech produc- psychiatrist, it was origi-
tion. nallyused as a general
teaching method in that
New methods country's primary schools;
elsewhere, it has mainly
Since the 1960s, several fresh approaches to FLL been applied in the field of
have been devised, aiming to provide a radical adult FLT. The method is
alternative to traditional methods, which their pro- based on the view that the
ponents believe have failed. They draw attention brain (especially the right
hemisphere, §45) has
to the success with which people acquire more than
great unused potential,
one language all over the world (§60), and contrast which can be exploited
this with the limited achievements of the classroom through the power of sug-
situation, and the partial accounts of learning pre- gestion. Language learning
can be promoted by draw-
sented by the various theories (p. 372). If FLL can
ing on the reserve capaci-
be made more natural, and the learner made more ties of the unconscious
receptive to the task, it is argued, more efficient mind. Blocks to learning
learning will result. are removed (using desug-
The effectiveness of the different methods re- gestion' techniques), and a
positive attitude towards
mains to be thoroughly evaluated; but each has
language learning devel-
itsreported successes, and some (especially the first oped ('resuggestion').
two in the list (right) have come to be widely prac- In their opening lesson
tised. The following outlines (with the originator's (or concert), learners are
presented with large
name in parentheses) indicate the thrust of each
amounts of the foreign lan-
approach, but they convey nothing of the emo- guage. The text is trans-
tional atmosphere and sense of involvement pro- lated, then read aloud
it is
moted by these methods, which are central to their in a dramatic way against
a background of classical
claims of success.
music. The aim is to pro-
vide an atmosphere of total
relaxation and enjoyment,
in which learning is inciden-
tal. After a session, there
should be a sense of eu-
phoria, reminiscent of the
feelings that follow a
to a health spa (the 'spa
effect'). By using a large
amount of linguistic mater-
ial, the suggestion is con-
Learning by hypnosis An unorthodox method of foreign veyed that language learn-
language learning - am learning English while sleeping,'
I
ing is easy and natural. In
says the notice. a later session, students
use the material in various
communicative activities.
The emphasis, then, is
wholly on informal commu-
nication; no attention is
drawn to grammatical
errors. Learners, it is
claimed, assimilate far
more from such an immer-
JAPPREND sion' than
ally
would
be expected.
tradition-
The silent way (Caleb
Gattegno) This approach
aims to provide an environ-
L'ANGLAIS ment which keeps the
amount of teaching to a
minimum and encourages
EN DORMANT
62
learners to develop their
own ways of using the lan-
guage elements intro-
duced. In the first lesson,
the teacher introduces a
small L2 vocabulary to talk
about a set of coloured
rods, using a few verbs
(equivalent to take', give',
'pick up', and put), adjec-
tives, pronouns, etc., and
gradually extending the
length of the sentence (e.g.
Take the green rod and
give it to Michael). The aim
is to help the learners to
become - to se-
self-reliant
lect their own sentences
and be in control of them,
with good intonation and
rhythm. The teacher does
not repeat the material or
provide sentences for stu-
dents to imitate; and no use
is made of the learners' L1
Charts containing vocabu-
lary and colour-coded
guides to pronunciation are
made available to enable
the teacher to guide the
student's learning while
saying as little as possible.
As students say more to
each other, so the teacher
says less - hence the 'si-
lent' way.
Community language
learning (Charles A. Cur-
ran) This approach builds
on the kind of whole per-
son' relationship found
counselling therapy.
main aim is to foster strong
visit personal links between the
teacher/counsellor and the
learners, and thus to elimi-
nate whatever is found
threatening in the FLL situ-
ation. There is no prepared
material. The learners talk
and
naturally in their L1 , (James J. Asher) This
seek from the teacher FL
equivalents for what they
want to say. The teacher
provides the translation,
and the students repeat it.
Each session is tape re-
corded, and is followed by
a discussion with the
teacher of what went on.
Natural approach (Tracy
D. Terrell) This method em-
phasizes the role of natur-
al'language acquisition,
and underscores the paral-
lels between L2 and L1 It
stresses the importance of
emotional rather than cog-
FOREIGN LANGUAGE LEARNING AND TEACHING 375
.
nitive factors in learning,
and of mastering vocabu-
lary rather than grammati-
cal rules. There is no for-
mal correction. The aim is
to establish an ability to
understand the basic con-
tent of a communication in
informal settings. Learners
use their L1 while their L2
comprehension is develop-
ing.
Language from within
(Beverly Galyean) This
method encourages
learners to be introspective
about their own needs, in-
terests, values, and here
and now' activities, and to
talk about these emotional
responses to others. All
material comes from the
students, as they become
more self-aware, and build
up a close relationship with
each other. The aim is to
enable the cognitive, affec-
tive, and interactive ele-
ments in learning to 'flow
together' - hence the alter-
native name for this
approach, confluent'
teaching.
Delayed oral practice
(Valerian A. Postovsky)
This approach, often
known as the comprehen-
sion approach', is based on
the principle that it is far
easier for learners to
in achieve competence in
The recognizing language,
whether in speech or writ-
ing, than in producing it. A
basic receptive com-
petence is established, and
this isused as a foundation
for work involving retrieval
skills.
Total physical response
method stresses the impor-
tance of aural comprehen-
sion as an exclusive aim in
the early months of learn-
ing. The name derives from
the emphasis on the ac-
tions that learners have to
make, as they are given
simple commands (e.g.
stand', More
sit',
advanced language is in-
'stop').
troduced by building up
chains of actions, using
either spoken or written
. commands.
•
 ,

A selection of kana cards (left), providing

MATERIALS ra _. Japanese learners with practice in recognizing
The days are long gone when FLT materials con- "7 the symbols of the katakana and hiragana
sisted only of a
Today, there is
grammar book and a dictionary.
a vast variety of printed materials
•7 THragana -
C?
alphabets (p. 201 ). On one side, the symbol is
given with three examples of its use; on the other
- course books, workbooks, readers, programmed side, there is the name of the symbol, a

courses, collections of facsimile material, simplified

h y * romanized version of each word, a translation,
arrows showing how to write the symbol, and the
cue cards, charts, newspapers, maga-
y rai*p* lamp
literature,
v y y torakku truck
equivalent symbol in the other kana system.
zines, posters, picture cards, cut-outs, and much
more. These are supplemented by a range of ma- 9 7 4
terials using other media, such as records, audio
tapes, slides, transparencies, filmstrips, video tapes, ™ »

toys,games, and puppets. The advent of computer

technology introduces a further potentially inex-
haustible domain of ancillary equipment. In
modern foreign language teaching and learning, fi »T Das
U 'as
'st

ist
die
das
T,
3

materials design and implementation is a major

— the area where the principles of applied
% x >v rustu me-iafeef
one's absence;
Das ist die Tal

enterprise D *s ist die TUf

linguistic theory, the demands of classroom prac- U '« ist das } |
't Das ist
tice, and the realities of commercial production lie die TtfJ

uneasily together. °as ist das Fend

"'as ist das?
Right: A page from Book 1 Lesson , 1 Das ist das FenJ
of a popular 930s German course,
1

Deutsches Leber. The exercise drills

DasistderFussbo,
questions and answers, and introduces "as ist das ?

—BSHT™
MECAHIQUE
the three forms of the definite article,
der, die, and das.
35 ist
das >
DasistderFussboq

D « ist die
Declce
1

P ROCHE Jas ist die \

Below: An ingenious way of providing U asistdas?Vaild
IE MASS practice in present-tense sentence pat- Das «t die Uand.
miMrthe pow SERVICE terns - Part of a unit from Book 1 of
"
METHODES The Cambridge English Course (M. £as ist das Pttft.
Swan & C.Walter, 1984). " as ist das r

INGEHIEUR NOTES AND

VOCABULARY
The Swan-Walter Universal Holiday Postcard Machi
Abbreviated words are
' given in full on their first
Inrttcw* 9*'
occurrence. I It's easy to write holiday postcards! Write
one now and send it to a friend
l-TJli'r""- !^"" petite mecanique light
engineering
POSSIBLES LOCEMENT Vt I* •* r
~

la formation training
la rentabilite profitability &"*
CV. MANUSCRIT.
„..
or-
»*,.««' I. ,

a vantages (m) sociaux

fringe benefits
(restaurant, expenses.
\r--^-.-*-
etc.)
Curriculum Vitae
IMPTES0C1ETE pretentions (/) salary
^ wot uaU.
un,c it

TELECOMM ™ns required

Publicite (/) advertising
rechi'che
transmettra forward Loot,
ADJOINT un adjoint assistant
MARKETING commerciaux
organisation (/)
un fichier card index, Mary Mantfmlw
-

record
A„ : JS-30 m» Some authentic written
FORMATION language materials. The
W
ECOLE COMMIE
(place)
materials are accompa- my room
mew room
ou 1ngenieuk
nied by lexical glosses for ;
bee
the less familiar items -
!r»e bar
tuebeacr touo;
"'saw?.
•

one of the items in a text- a ca'e

prodo1t"inoustriels book anthology for ad- 3 DadOy
-

Brno rttfcl George and Soe

ANGLWS writ et etrlt vanced students of etc
French (D. E. Ager,
1970).

376 •
X LANGUAGE IN THE WORLD
The language laboratory
The best-known technological aid in FLT is undoubtedly the
language laboratory - a room, usually divided into booths,
where students can listen individually to tape recordings of
FL material, and where they may record and play back their
own responses, while being monitored by a teacher.
When these laboratories were first introduced, they were
heralded as a technique that would vastly improve the rate
and quality of FLL. They would take the burden of repetitive
drills away from the teacher, provide more opportunities for
learners to practise listening and speaking, and enable them
to develop at their own rates and monitor their own progress.
Many schools were quick to install expensive laboratory
equipment. However, within a few years, it became apparent
that there would be no breakthrough. The expected
improvements did not emerge, and the popularity of the
'language lab' showed a marked decline.
There were several reasons for the failure to live up to
expectations. The taped materials were often poorly
designed, leading to student frustration and boredom. The
published programmes failed to reflect the kind of work the
student was doing in class. Few modern languages staff had
received training in materials design or laboratory use. And
it proved difficult to maintain the equipment once it had been
A language laboratory that incorporates computer-assisted language learning (CALL)
installed.
work stations. Microcomputers, used as word processors, complement the audio
facilities, enabling interactive teaching of written language skills. Several kinds of FLT
Today, the strengths and limitations of the laboratory are
exercise can be computationally controlled, such as sentence restructuring, checking
better realized, and the vastly increased potential of modern
of translation or dictation tasks, and cloze testing (see below) using texts displayed
electronic hardware has led to a certain revival. There is now
considerable interest in language learning laboratories, which
on the screen.
contain much more than the traditional systems - in particular,
the introduction of interactive computational aids and video
materials has proved to be extremely popular.
It is now clear that, when used properly, laboratories can

provide a valuable extra dimension to FLT. For example, the

taped material can provide a variety of authentic and well- Master The man is running.
The men are running.
recorded models for the training of listening comprehension. track The men?
And laboratories can be used as resource centres, or
libraries, giving learners extra opportunities to practise at their Learner The men is running. The men are running
chosen level. track
At the same time, the limitations of laboratories must be
borne in mind. Their value will always depend on the
B D
development of appropriate teaching materials which A typical sequence of events on a language laboratory tape. In this illustration, the
reinforce what has been taught in class and provide tape double-track, enabling the foreign model to be recorded on one track, and the
is
opportunities for creative use; and here there is an urgent learner's voice on the other. An example of one kind of drill is given. Learners first
need for research into the efficacy of the different approaches hear a sample sentence, and are given a stimulus to respond to (A). They then record
which have been devised. Laboratory software, seems, has it theirresponse (B), hear the correct version (C), and have an opportunity to repeat
some way to go before it can compare in sophistication with it(D).
the hardware.

Testing Anyway, Susie said that there were no (1) things as fairies,
elves, this that and (2) other. Well, the night she put her
Teachers need to test stu- mastered. These are com- vidual difficulties.
(3) under the pillow we forgot to put (4) money there
dent performance, or enter monly used at the end of Testing can focus on any and take it away —
we (5) all about it. (C laughs) So she
them for formal examina- school terms. linguistic skill (speaking, lis- got up (6) the morning, 'My tooth's all gone and (7)
tions. Test results are criti- • Diagnostic tests aim to tening, reading, writing) and no money.' Dave said, 'Well, there you (8) you see. You
cal, not only because they find out what a student still on any linguistic component said you didn't believe (9) fairies so how can you expect the

affect careers, but because has to learn in a language. (e.g. vocabulary, pronuncia- (10) to come and see you if. .'.
Oh (11) . 1 do believe

in fairies (D laughs), you know (12) really do.' (C laughs).

of the influence they exer- The results provide feed- tion, grammar, spelling).
So Dave said. 'Well, try (13) tonight.' So that night, thank
cise on motivation to learn. back for the teacher, by dis- Many ways of eliciting infor-
goodness, we (14) (C laughs) So the next morning she gets
It is now widely appre- playing the learner's mation have been devised; up (15) happy. 'Oh, they've been, they've been. I've (16)
ciated that tests perform a strengths and weaknesses. such as asking students to my money." And Dave said, 'Well there (17) are. That
variety of functions. Four Although they are very dif- make a translation, carry
main types are usually ferent from achievement out an action, or give a
Cloze testing In this kind of language test, every nth
recognized. tests (e.g. they are not al- paraphrase. Anticipating the word is omitted from a passage (the gaps are usually
• Proficiency tests deter- ways given marks or effects of a task is not easy,
between five and seven words apart), and the student has
mine how much of the L2 a to complete (or 'close') the gaps. This form of testing is
graded), students often fail however; a question may be
learner has mastered, re- see the difference. widely used: it is good at establishing whether a student
to unnecessarily difficult, or an
gardless of the course of • Prognostic (or aptitude) answer may require infor- has a 'feel' for the language, and for testing awareness
of points of detail. Several types of cloze test exist; for
study followed, as in na- tests try to predict how well mation that has not been
tional examinations, and the a person will succeed in taught. It is therefore always
example, students can be given the omitted words below
international Test of English learning an L2. These tests the passage, or they can be given an initial letter as a
necessary to look critically
clue. In the example above, taken from a workbook on
as a Foreign Language. focus on specific kinds of at test procedures, and to
• Achievement tests deter- sound conversational English, there are no clues: the student
activity (e.g. imi- aim for improvements in test
mine how much of a particu- tation, pattern matching), design, selection, and ad- must fill in any appropriate word, bearing in mind the
lar course of study has been and provide data about indi- ministration. informal style of the passage. (K. Morrow, 1 978, p. 49.)

62 FOREIGN LANGUAGE LEARNING AND TEACHING 377 •

 .

63 Language for special purposes

They don't seem to be talking the same language.' Doublespeaking

Plain English campaigns
This common observation acts as a reminder that negative patient care out-
barriers to communication exist, and indeed are Popular anxiety over special uses of language is come (=death)
commonplace, even within a language. Linguistic most markedly seen in the campaigns to promote
portable handheld communi-
difficulties are inevitable when there is interaction 'plain' speaking and writing - notably, the Plain
cation inscriber (=pencil)
between people from different racial, regional, cul- Englishmovements of Britain and the USA. The
tural, social, or occupational backgrounds — some- main aim of these campaigns is to attack the use unlawful or arbitrary depriva-
thing that is increasingly common
modern in of unnecessarily complicated language ('gobblede- tion of life (=killing)

society as people become more mobile and come
into contact with diverse forms of linguistic be-
haviour. A major aim of linguistic studies, accord-
ingly, is to investigate the factors that
maintain the existence of varieties within a lan-
guage (Part n), and to provide descriptions of their
use. These studies have an intrinsic intellectual
interest, as they provide a means of observing
change in contemporary culture and civilization.
But they can also be of practical assistance, by clari-
fying the reasons for the use of unfamiliar language,
and thus providing a perspective that may help to
resolve cases of linguistic conflict.
There is no theoretical limit to the
special purposes to which language can be
society develops new facets, so
to express them. In recent times,
of expression have emerged, in relation to such
domains as computing, broadcasting, commercial
advertising, and popular music. Over a longer time-
scale, special styles have developed associated with
religion, law, politics, commerce, the press, medi-
cine, and science. A detailed linguistic account of
any one of these areas would itself require an ency-
clopedia, as the analysis of the language used would
require an exposition of the conceptual system that
gave rise to it. The following illustrations of linguis-
tic varieties and attitudes are inevitably highly
selective, therefore, in both range and depth of
treatment. But they do provide a hint of the exten-
sive resources that language makes available to
meet the special needs of developing societies, and
of the complications that arise as people slowly
come to terms with them.
gook') by governments, businesses, and other auth-
permanent pre-hostility
orities whose them in linguistic contact
role puts
(= peace)
with the general public. The campaigners argue
promote and that such language, whether spoken or written, revenue enhancement tax-
should be replaced by clearer forms of expression. base erosion control (=tax
increase)
The movements took shape only in the late
1970s, so it is too soon to ascertain their long-term collateral damage (=civilians
influence on the characteristics of language vari- killed in a war)
eties. But they have certainly played a major part
in promoting public awareness of the existence of
communication problems, and have influenced
many organizations to do something about it. In
Britain, the campaign was launched in 1979 by
number of a ritual shredding of government forms in Parlia-
put. As ment Square, London. By 1982, the government
language is devised had published a report telling departments to im-
whole new areas prove the design of forms, and to abolish those
that were unnecessary. By 1985, around 15,700
forms had disappeared and 21,300 had been
revised. In the USA, President Carter's Executive
Order of March 1978 required regulations to be
written in plain English, and although this was
revoked by President Reagan in 1981, it promoted
a great deal of legislation throughout the country,
and an increase in plain English usage among cor-
porations and consumers.
Today the Plain English campaigns continue to
grow, focussing especially on such everyday con-
sumer products as forms, official letters, licences,
leases, contracts, insurance policies, and guaran-
tees. In Britain, annual publicity is given to the Plain
English Awards competition, which gives trophies
to organizations that have produced the clearest
documents, and booby prizes (the Golden Bull

A Golden Bull winner

One of the winners of the medical board varied the week from 23. 11. 83 to 24.3.83 to 7.6.83 and a dis-
British Golden Bull awards assessment of disablement 6.12.83 based upon an ablement pension of £10.72
in 1 982 was a written deci- resulting from the relevant assessment 20% for the
of per week from 8.6.83 to
sion made by a government loss of faculty and this con- period 8.6.83 to 1 .1 2.83 and and £11. 12 per
22.1 1.83
insurance officer, which stitutes a revision of a deci- a disablement gratuity of week from 23. 11. 83 to
opened in the following way: sion on a special question. £52.75 based upon an 13.12.83 (total £290.64)
My revised decision is as assessment of 3% from was awarded and paid . .

As insurance officer I have follows: A disablement gra- 2.1 2.83 to 1 .6.84 are award- (Quoted in C. Maher&M.
decided to review the deci- tuity of £21 .50 based on an ed for the same accident for Cutts, 1986, p. 12.)
sion dated 19.9.83 for the assessment of 3% per which a disablement gra-
following reasons: that by week from 8.6.83 to £21 .50 based upon
tuity of
its decision of 31 . 1 .84 the 22.1 1.83 and £11. 12 per an assessment of 3% from

378 X LANGUAGE IN THE WORLD

Awards) to those whose materials are least intellig-
ible. In the USA, similar interest is shown in the
What is plain English?
annual Doublespeak Awards, awarded by the not easy to devise precise, consistent,
It is

National Council of Teachers of English to 'Ameri- and acceptable guidelines for those who wish
can public figures who have perpetrated language One proponent of
to write in plain English.
that is grossly unfactual, deceptive, evasive, the Plain Englishcampaign in the USA
euphemistic, confusing, or self-contradictory'. makes the following recommendations
In these cost-conscious days, it is stressed that concerning readability and design. (For other
clear language not only avoids anxiety on the part views, see pp. 2, 252.)
of the recipient, it and money. The
also saves time
Readability
campaigns have large dossiers of problem cases.
• Prefer the shorter word to the longer one. Use
In one case, an official government letter provoked
simple, everyday words rather than fancy ones.
so many complaints and questions that a second
Prefer verbs over nouns and adjectives. Prefer the
letter had to be sent to explain the first. In another,
specific word to the general.
an application form was wrongly filled in by 50% • Write short sentences with an average of no
of the applicants, which resulted in a considerable more than 20 words. Use the active voice rather
outlay of effort in returning and reprocessing the than the passive. Be a miser with compound and
form. On the positive side, there are cases of busi- complex sentences and a spendthrift with simple
nesses revising their literature to avoid legal jargon, sentences.
and benefiting from increased sales. • Write short paragraphs with an average of
about 75 words. Avoid paragraphs that exceed
Particular concern expressed about the ambi-
is
five typed lines for business letters and ten lines
guities and omissions found in medical labels. For
for longer compositions.
example, in one pharmaceutical survey, the instruc- • Write with the ear. A sentence may look
tion to 'use sparingly' was found to be misunder-
correct on paper but its cadence may be jarring.
stood by 33% of patients. The instruction to 'take Listen to your sentences in your head as you
two tablets four hourly' received a variety of inter- write, and do not write anything that you could
pretations (e.g. to take eight tablets an hour). not comfortably say.
Related areas of concern include the use of warning
Design
labelson household goods (such as disinfectants) Write for the eye as well as the mind. Prepare an
and on toys for children. overall design, positioning understandable
The instructions accompanying do-it-yourself headings, subheadings, and captions for each
products are also regularly cited as a source of segment, showing the organization of the text.
unnecessary expense or frustration. Few companies Make the whole document visually appealing.
seem to test their instructions by having them fol- • Use appropriate underlining, ink colour that
contrasts sharply with the paper, lists, boxes or
lowed by a first-time user. Often, essential back-
panels, bold or other typefaces to emphasize key
ground information is omitted, steps in the
points.
construction process are taken for granted, and
• Use 'white space' in margins, between sections,
some degree of special knowledge is assumed. This paragraphs and lines to make the document look
is especially worrying in fields where any failure good.
to follow correct procedures can be dangerous.
(J. Y. Dayananda, 1986, p. 13.)
Objections to material in plain English have
come mainly from the legal profession. Lawyers
point to the risk of ambiguity inherent in the use
of everyday language for legal or official docu-
ments, and draw attention to the need for confi-
dence in legal formulations, which can come only Before and after
from using language that has been tested in courts An illustration of plain Eng- delivered, removed or des- the goods until you have
over the course of centuries (p. 386). The cam- lish in action can be seen patched or sold the general paid all our charges and
from the British campaign's lien shall apply in respect of other payments due under
paigners point out that there has been no sudden
version of a contract for such goods as remain in this contract. These include
increase in litigation as a consequence of the moving goods from one the Contractor's posses- charges, taxes or levies
increase in plain English materials. Similarly, pro- house to another. sion. The Contractor shall thatwe have paid to any
fessionals in several specialized fields have de- be entitled to charge a stor- other removal or storage

fended their use of technical and complex language Original version
GENERAL LIEN -The
as being the most precise means of expressing tech-
contractor shall have a
nical and complex ideas. This is undoubtedly true: general lien upon all
scientists, doctors, bankers, and others need their in his possession for
jargon, in order to communicate with each other monies due to him from the
customer or for liabilities
succinctly and unambiguously. But when it comes
incurred by him and for
to addressing the non-specialist consumer, the monies paid on behalf of
plain English campaigners argue, different criteria the customer, and if part of
must apply. the goods shall have been
age charge and all other business, carrier or official
expenses during which a body. While we hold the
lien on the goods is being goods and wait for pay-
goods asserted and all these con- ment you will have to pay
all ditions shall continue to storage charges and all
apply thereto. other necessary expenses.
This contract will apply to
Plain English version the goods held in this way.
OUR RIGHT TO HOLD (Quoted in C.Maher&M.
THE GOODS -We have a Cutts, 1986, p. 12.)
right to hold some or all of

63 LANGUAGE FOR SPECIAL PURPOSES •

379
The language of science than 750,000 species of insect have been dis- Scientific
vocabulary
covered, for example; and if all their names were
The aim of science is to determine the principles incorporated into the largest available dictionaries, (Dates given are of the first
governing the physical universe. Progress towards the books would immediately double in size. recorded appearance in the
this end, however, is to a large extent dependent Oxford English Dictionary)
The grammar of scientific language also contains
on the use of language. The knowledge base of several distinctive features. Sentences are often long 6th century
a subject, upon which all scientists depend, is ac-
and have a complex internal structure. The com- 533 catarrh
cessible only if previous generations have managed 527 cornea
plexity is centred on the noun phrases (§16) rather
to express their findings in a precise and unambi- 543 cranium
than the verb phrases, as can be seen from the
guous manner. Similarly, present-day scientists, 551 genus
illustration below. But probably the best-known
hoping to make their own contribution to this 578 glottis
grammatical feature is the use of the passive con- 598 mumps
knowledge base, must satisfy the same linguistic
struction: The mixture was poured instead of
. . . 551 species
constraints if their work is to be correctly inter-
We poured the mixture The intention behind
. . .
531 temperature
preted and accepted by their peers. Research find- 548 tibia
thisusage is straightforward: as in this example,
ings are of limited value, until they are written up 550 vacuum
itenables a description to be made impersonally,
and published.
without an agent being expressed (contrasting with 7th century
The methodology of science, with its demand pronoun in the corresponding 626 acid
the use of a personal
for objectivity, systematic investigation, and exact 605 acoustic
active construction). However, the overuse of com-
measurement, has several linguistic consequences. 628 apparatus
plex passive sentences in scientific writing has 694 axis
There is an overriding concern for impersonal
attracted considerable criticism in recent years, and 601 cardiac
statement, logical exposition, and precise descrip-
there is now a marked tendency to avoid them. 638 formula
tion. Emotional comment, humour, figurative 641
The Handbook for Chemical Society Authors gravity
expression, and other aspects of personal language 605 laboratory
(1961), for example, recommends: 'Before the final
are avoided (except in writing intended for a lay 615 logarithm
typing every paper should be scrutinised to see 656 microscope
audience). The mathematical expression of rela-
whether it cannot be improved by eliminating 626 pedicle
tionships promotes an extensive use of numerals,
abstract words and passive voices.' But the weight 693 pharynx
operators, letters, and other special symbols, which 601 rheumatism
of traditional usage is not easy to throw off, as
are frequently used in word-like and sentence- 668 stamen
illustrated by this very recommendation, which
like combinations (as formulae, equations, etc.).
itself uses two passives! 8th century
Lengthy sequences of text can be written in logo-
751 antiseptic
graphic form (p. 200), thus giving the language 791 carbonic
of science its highly valued status as a universal 791 etiolate
medium of expression. 771 fauna
791 hydrogen
794 molecule
Vocabulary and grammar 759 nectary
776 neurosis
In addition to this distinctive graphology (§33), 794 nitrogen
scientificlanguage illustrates several important fea- An example 704 nucleus
tures of vocabulary and grammar. The large tech- This short extract (G. H. Williams, 1960, pp. 252-3) 790 oxygen
nical vocabulary is undoubtedly its most illustrates well the nominal bias of scientific language.
726 thyroid
characteristic feature, reflecting the specialized sub- In the final sentence of the paragraph, for example,
1 6 of the 24 words are part of noun phrases: Smaller
9th century
ject matter of domains of enquiry. Every-
scientific 849 allotropy
quantities of tarry products ...the reported yields of
day words are too vague for many scientific pur- diaryls ...the diazo- and azo-compounds discussed 867 aphasia
poses, so new ones have to be invented. This novel above. By contrast, verb phrases are extremely short, 865 barograph
vocabulary is largely based on borrowings from and use a restricted range of items: in the whole extract 812 centigrade
we have provide, be (four times), add, effect, obtain, 890 chromosome
Latin and Greek, showing the influence of Classical
and discuss -typical of the manipulatory' meanings 839 cirrhosis
languages during the period of scientific discovery 832 cretaceous
that the majority of the verbs in scientific language
following the Renaissance. It contains many com- express (adjust, align, arrange, begin, boil, bring, 882 dynamo
pound expressions, some of which (in such fields continue, etc.). Note also the inverted commas, to 856 gyroscope
identify the use of a vague word from everyday 822 laryngitis
as chemistry) can be extremely long and unpro-
discourse. 878 metabolism
nounceable, requiring abbreviation for practical 892 micron
use (a familiar example is TNT, short for trinitro-
(a) Preparative Use of Diacyl Peroxides
The thermal decomposition of diacyl peroxides
toluene). At the other extreme, some fields delight 20th century
provides what is undoubtedly the "cleanest", and, 913 allergy
in using everyday words to identify new hypotheses provided the required peroxide is readily 946 cybernetics
and discoveries - notably, in contemporary particle available, most convenient source of aryl radicals 901 genetics
physics, where we find such technical terms as for the arylation of aromatic substrates. The 951 laser
strangeness, flavour, colour, and charm. purified peroxide, which is generally crystalline, 912 millibar

Moreover, scientific vocabulary requires con- is added to the aromatic solvent, and the 929 penicillin

decomposition is effected by heating, usually to 934 positron

tinual updating in the light of the process of dis-
910 quantum
covery. Science is in fact the main birthplace for about 70-80°. Smaller quantities of tarry products
942 radar
new words in a language: in a comprehensive Eng- are obtained, and the reported yields of diaryls
912 vitamin
are generally higher than with the diazo- and azo-
lish dictionary (§18), the vast majority of the words
compounds discussed above.
would be scientific (or technological) terms. More

380 •
X LANGUAGE IN THE WORLD
Intelligibility gap
The gap between scientific and everyday language
is a large one, which it is difficult to bridge. Scien-
tists are often unable to express themselves in terms
the lay person can understand, or are too busy to
bother. There frequently a need to maintain
is
secrecy, in such areas as national security or indus-
trial invention. Not surprisingly, therefore, there
is a widespread mistrust of scientific language,
which is only partly alleviated through popular
science publications and radio or television pro-
grammes. It is still the exception to find populariza-
tions of science that maintain intelligibility while Explain this!
avoiding oversimplification, and that come to be
acclaimed by specialist and lay person alike.

NONVERBAL EXPRESSION or tree diagram, for example, and verbal descrip-

A large part of scientific expression consists of
representations that are wholly or partly non-
linguistic in character - such as physical models,
charts, pictures, maps, graphs, and diagrams. The
immediacy and economy of presentation achieved
by these methods is self-evident. It would be impos-
sible to provide a coherent account in words of
all the interrelationships found on a map, graph,
tions of formulae and equations are often highly
complex and ambiguous (§33). On the other hand,
linguistic and non-linguistic modes of expression
are never totally independent of each other: verbal
language is always needed in order to interpret and
amplify the meaning or use of nonverbal represen-
tations.
It would be possible to pro-

vide a verbal description for

this electronic circuit, but it

would be so long and com-

plex as to be unintelligible.
On the other hand, the verbal
description of the elements
of such circuits is an essen-
tial feature of training pro-
grammes in the subject.

Mathematical
expressions

250
+7
This simple mathematical
expression is unambiguous,
in its non-verbal form, with all
elements simultaneously
present. But as soon as we
attempt to read it aloud, in a
serial way, complications ar-
Men of science In previous centuries, scientists from different subject areas were keen to follow each other's progress. ise. The verbal version would
Ideas and findings were exchanged and debated in scientific societies, as depicted in this early 19th century painting. Such be: 'the square root of two
'communicative' meetings are rare events today. Scientists find it difficult enough keeping up with the specialized literature hundred and fifty divided by
within their individual fields. three plus seven'. But this
written formulation could be
interpreted in several ways,
such as:

Alternative 'languages'
+7
Chemists have four ways in which unusable. Formulae are shorter; but Label: METHANE
they can describe the substances whereas these can display the ele-
Formula: CH 4
they study: they can use a verbal ments of a substance, they do not
or
label, write a formula, draw a struc- show their physical The Structural diagram:
relationship. 3 +7
tural diagram, or build a structural internal structure of a substance can
model. Each has its strengths and be shown using a structural diagram; 50
weaknesses. Verbal labels are basic, but this is restricted to two dimen- or
.

V200 + — +7
3
but their length (often several dozen sions. Only structural models provide Structural model:
morphemes, p. 90) can make them a three-dimensional picture.

3 + 7

In speech, careful use of in-

tonation and rhythm could
distinguish some of these in-
terpretations, but the risk of
ambiguity would always be
present.

63 LANGUAGE FOR SPECIAL PURPOSES •

381
The language of medicine linguistic difficulty (especially in the case of immi- Symptoms of
grant patients). Even age can intervene. In one breakdown
The field of medicine, more than any other, forces American study, the problem was so serious that
a confrontation between scientific and everyday Several studies have col-
it was found helpful to devise a questionnaire
lected data that indicate the
language. Outside the world of the research labora- phrased in appropriate slang to enable older doc- kinds of communication pro-
tory and clinic, there exists the daily routine of tors to communicate with inner city teenage blem that frequently arise in
medical practice — a communication situation in patients (C. C. Levine, 1970). hospitals. Typical quotations
which a doctor attempts to understand the prob- The careful analysis of medical interviews, using include:

lems of a patient, and the patient attempts to under- • I didn't like to ask; perhaps
audio- or video-recorded samples, has brought to it's just me; I they might
felt
stand the doctor's diagnosis. Language is involved light many instances of these difficulties. Some think I was prying or being
at all points in the medical consultation. The initial people are naturally taciturn in formal situations, nosey.
statement of symptoms is of critical significance, because their social or cultural background has • was strung up when the
I

as it guides the doctor's search for the clinical signs doctors were there, and for-
developed in them a sense of 'knowing their place'.
got things.
of the condition. Similarly, the doctor's explana- Some necessary to talk at length about unre-
find it • They would give you pills
tion of a problem, and the recommendations for lated topics as a preliminary to introducing their and if you asked what they
treatment, need to be clear and complete if the symptoms. Some play down the importance of were you were told to
for
patient is to understand and follow the correct these symptoms, because they have been brought
take and never mind. You
it

course of action (p. 379). were treated like a child, as

up 'not to make a fuss'. Each type of case presents if it was nothing to do with

doctors with a problem of communication analysis. you if the medicine was

Communication problems
But linguistic problems continue to occur even changed. There was no
The need for careful listening and expression by when doctor and patient share the same social reason given.
both parties should be self-evident in a field as • They leave you in the dark
background. Doctors need to be alert to pick up
and serious as health. In practice, many too much. If only they treated
sensitive the linguistic cues that may express the patient's you as if you could under-
problems worried about their health
arise. Patients
real reason for coming to the surgery ('By the way, stand something.
are often uncertain or confused in their accounts. doctor .') or the issue that is causing most sub-
. .
• feel better knowing. You
I

Busy doctors will not have the time to take up every conscious worry (such as repeatedly referring to always imagine things are
point the patient has alluded to. Moreover, there worse than they are.
the heart during the conversation). They also need
is a tradition of medical interviewing which hinders
Many hospitals now give pa-
to anticipate points of potential misunderstanding tients an information booklet
the development of a genuine communicative inter- — such as the common patient interpretation of the telling them what they need
action. One study of ten major medical journals word growth ('You have a small growth here') to to know in advance. Several
in the 1960s found general agreement about the researchers believe that a
mean 'cancer', or thrombosis to mean 'heart dis-
following characteristics of doctor-patient com- paragraph about communi-
ease'. cation should be a routine
munication (after S. B. Heath, 1979):
Medical communication researchers have also feature of such literature.
drawn attention to several areas where medical One suggestion reads as fol-
1 Topics of conversation should be restricted to
lows:
staff could promote their own communicative skills
those dealing with the patient's body and condi- Don't be afraid to ask the
tions contributing to disease.
— for example, by providing explanations of what doctors or the nursing staff
2 Conversation should only be with the patients, they are doing to a patient while they are doing any questions you want to
it, by welcoming questions from patients (rather ask about your illness or its
not with relatives or friends.
than fostering the 'Doctor knows best' attitude), treatment or about anything
3 It is the task of the doctor, not the patient, to
else that worries you. Some
ask questions. and by avoiding patronising or discourteous lan-
people find it difficult to re-

4 The doctor should avoid telling all the truth, and guage (e.g. 'Drink it down like a good girl', said member all they want to ask
discourage the elaboration of information from to an older female hospital patient). when the doctor or sister
Above all, medical staff need to be aware of the comes round. You may find
other sources.
many functions that language can perform (§4), it helpful, and it will help
5 Patients should be told that they are ultimately them, if you make a list of
responsible for the improvement of their own and in particular that language may be used to sig- your questions to show them
health, which will occur only if they follow the nal the desire for social contact, and need not be when they visit you.
doctor's advice. taken literally. For example, in one study, 40 hospi- (CM. Fletcher, 1973.)
tal patients who asked for relief from pain were
This tradition is still widely encountered, though The final breakdown
given either routine nursing attention or a visit
Nowhere has this communi-
it has attracted criticism from both within and out- from a nurse specially trained in communication cation gap been greater than
side the medical profession in recent years. skills: only two of the former group obtained in the management of the
Studies of medical communication have brought immediate relief from their 'pain', whereas 14 of terminally ill patient. Accord-
to light several types of situation in which there the latter group did so; and all of the former group ing to one
970) estimate,
(1

has been a breakdown of communication, and

80% of dying patients know
required analgesics, compared with only six of the
that they are dying and want
where the consultation has had an unsatisfactory latter (M. B. Tarasuk et al., 1965). Such findings to talk about it; whereas 80%
outcome. Regional, social, and cultural differences illustrate the need for a perspective on communica- of doctors deny this, and be-
between doctor and patient can all be sources of tion to be a routine part o_f medical training. lieve that the patient should
not be told. Fortunately, the
present-day climate of medi-
cal opinion seems to be
changing, slowly, for the bet-
ter. (W. A. Cramond, 1970.)

382 • X LANGUAGE IN THE WORLD

 STYLO-GLOSSUS.
Consultation styles
Naming of parts STVLO-HYOIO. Analyses of the language
OBLIQUUS CAPITIS SUPERIOR used in consultations have
A diagram of the muscles of
brought to light great differ-
the neck, taken from one of
ences in styles of interaction.
the classical accounts of hu-
In one study, patient com-
man anatomy, Gray's Ana-
ments of the type 'I'm feeling
tomy, and showing the tradi-
run down. I've got a pain in
tional use of Latin
nomenclature. Most people
my back, and feel tired all
I

day' were found to elicit a

identify medical language
wide variety of responses,
with this kind of terminology,
such as:
or vernacular equivalent;
its

and indeed, the labels of

anatomy and physiology do
• Mmmm. Right, just go into
the next room and get un-
form the core of the subject.
dressed. be along in a
I'll
However, the kinds of con-
minute.
versation that take place in
• Tell me. Just where is this
hospitals and clinics intro-
pain?
duce a wide range of ad-
• When do you feel tired? In
ditional terminology. People,
the morning when you get up
locations, routine objects,
or in the afternoon?
and daily activities all have
• Do you have headaches
their special labels, idioms,
and pains behind the eyes?
or abbreviations, e.g. intern,
• I think you're depressed.
registrar, SRN, ENT, path
lab, sluice, day room,
How do you feel about that?
• What sort of pain is it?
theatre, medical records, op,
• What do you mean by I

scrub up, drawsheet, sam-

'

feel tired all day'?

ple, drainage tube. It is not
• Yes, go on.
difficult to hear sentences
• You look very pale.
that are unintelligible, save to
(P. S. Byrne & B. E. L. Long,
the initiated (e.g. Staff wants
1976, pp. 22-3.)
you to do the TPRs on the
four hourlies (i.e. The staff
The responses vary enor-
nurse wants you to take the
mously. At one extreme the
temperatures, pulses, and
doctor's mind seems already
respirations of those patients
made up; at the other, the
who need this information re-
doctor is ready to enter into
corded every four hours')).
a long discussion about
further symptoms.

Computer-aided diagnosis
The range of medical condi-
tions is now so vast, and the
doctor's time so limited, that
several clinical centres have
begun to introduce com-
puter-assisted diagnostic
systems. In an extension of
this procedure, it is even
possible in some centres for
patients to begin the diag-
nostic process by them-
selves. They are presented
with a series of questions
about their medical history
and symptoms on a com-
puter screen. They are given
sets of possible answers,
and they make their selec-
tion by pressing buttons on
their console. There is no
time constraint, and they
may change their answers at
any time, or simply leave
questions blank. The
answers are then stored in
the computer, to be printed
out when the doctor is ready
to analyse them.

63 LANGUAGE FOR SPECIAL PURPOSES • 383

 .

Language and religion Hinduism The Vedas, a wide range of texts, writ-
ten in Sanskrit, and preserved largely through a
IN THE BEGINNING meticulous oral tradition, which takes particular
The close relationship between language and reli- care over accuracy of pronunciation (p. 405).
gious belief pervades cultural history. Often, a
divine being is said to have invented speech, or
Islam The Qur'an, or Koran, which Muslims
believe was dictated to the Prophet Mohammed
writing, and given it as a gift to mankind. One
of the first things Adam has to do, according to
by Allah, during the month of Ramadan. It is writ-
ten in classical Arabic, in a style which is considered
the Book of Genesis, is name the acts of creation:
miraculous, beyond ability to imitate. The memori-
And the Lord God having formed out of the ground zation of the text in childhood acts simultaneously
all the beasts of the earth, and all the fowls of the air, as an introduction to literacy.
brought them to Adam to see what he would call them:
for whatsoever Adam called any living creature the same
Judaism The Hebrew Bible, or Old Testament,
is its name . .
especially as found in its first five books, tradition-
ally said to be written by Moses. Later varieties
Many other cultures have a similar story. In Egyp- of Hebrew, and some Aramaic, form the language
tian mythology, the god Thoth is the creator of of the large collection of oral and written commen-
speech and writing. It is Brahma who gives the taries on the Bible, known as the Talmud.
knowledge of writing to the Hindu people. Odin Literacy is often introduced into a community by
is the inventor of runic script, according to the Ice-
the spread of a religion. As a result, the distribution The Egyptian god, Thoth,
landic sagas. A heaven-sent water turtle, with
of writing systems in the world today reflects the represented with the head of
marks on its back, brings writing to the Chinese. distribution of world religions far more clearly than an ibis, and here surmounted
All over the world, the supernatural provides a by a crescent moon. As in-
it does the distribution of language families.
powerful set of beliefs about the origins of language ventor of hieroglyphs, he
was named 'lord of holy
(§49).
Religious associations are particularly strong in
TRANSLATING THE WORD words'. He is said to have ac-
Not all religions favour the translation of their complished the work of crea-
relation to written language, because writing is an tion by the sound of his voice
sacred books. Judaism, Hinduism, and Islam stress
effective means of guarding and transmitting sacred alone. The Egyptian god-
the sacredness of the language itself and resist dess of writing, Seshat, was
knowledge. Literacy was available only to an elite,
translation, whereas Buddhism, and especially Thoth's principal spouse.
in which priests figured prominently. Echoes of this
Christianity, actively promote it. But ultimately,
link reverberate in English vocabulary still, through
allmajor religious works are translated - either
such connections as scripture and script, or the
from one language into another, or from an older
reference to scripture as Holy Writ. And there are
variety of language into a modern variety.
widespread sanctions for human action expressed
The formal process of religious translation is
authoritatively in phrases of the form: 'for it is writ-
a long-term, painstaking and frustrating task,
ten'.
usually carried out by committee. Translators have
two criteria, which are always incompat-
to satisfy
Sacred writings ible,because one looks backwards and the other
At the centre of all the world's main religions lies forwards. First, the translation must be historically
a body of sacred writing, revered by believers. Scru- accurate, faithfully representing the meaning of the
pulous attention paid to identifying or preserving
is source, insofar as this can be known, and integrated
the linguistic features of the original texts. Often, within the religious tradition of which it is a part.
the texts are accompanied by a long tradition of Secondly, it must be acceptable to the intended
commentary, which may itself take on special reli- users of the translation — which, in practice, means
gious significance. that it must be intelligible, aesthetically pleasing,
and capable of relating to current trends in religious
thought, social pressures, and language change. No
Buddhism The
Canon, based on oral tradi-
Pali
translation can ever satisfy the demands of all these
tion, containing the teaching of the Buddha. Pali factors, and all translations are thus to some extent
became the canonical language for Buddhists from controversial (§57).
many countries, but comparable texts came to exist The linguistic issues involved may relate to major
in other languages, such as Chinese and Japanese,
conflicts of cultural or historical interpretation, or
as the religion evolved.
be localized problems of style. A phrase such as
'Give us this day our daily bread' is not easy to
Christianity The Bible, consisting of the 39 books translate into a language such as Eskimo, where
of theOld Testament, written in Hebrew, and the the staple food is not bread; nor is it easy to handle
27 books of the New Testament, written in Greek. the Biblical parable of the fig-tree, which refers to
Several other writings, known collectively as the seasonal change, in a language where there are no
Apocrypha, and preserved only in Greek, have con- words for seasons, such as Yucatec. But even sub-
troversial status. A
Latin translation of the Bible, tle, apparently minor, linguistic differences can

known as the Vulgate, is prominent in the Roman become major points of controversy. To address
Catholic tradition. the divinity as thou will satisfy those who feel that

384 •
X LANGUAGE IN THE WORLD
 religious language should adhere closely to tradi- VARIETIES OF RELIGIOUS LANGUAGE
tion and be from the everyday
special, set apart The most striking feature of a religious language
language; but you will be preferred by those who is its heterogeneity, deriving from the wide range
wish religious language to have an immediate con- of activities involved in public and private worship,
temporary meaning and application. The thou/you and the involvement of religion in all aspects of
debate has rumbled on for decades in English - daily living and thinking. Several varieties, all dis-
with echoes in other languages that use a second- tinctive in their use of linguistic structure, are
person-pronoun distinction. widely encountered.
From the
invention of printing until the mid-
Liturgical forms both spoken and sung, produced
1980s, some
or all of the Bible has been published
by individuals and in unison, as monologue and
in about 2,000 languages, with several hundred
dialogue, and including such acts as invocations,
new translation projects ongoing; but a major lan-
petitions, doxologies (statements of glory or
guage with a long literary tradition will itself con-
praise), intercessions, thanksgivings, rosaries, lita-
tain many translations. Some English translations
nies, chants, hymns, psalms, and canticles.
have proved to be specially influential, such as the
Great Bible (Cranmer's Bible) of 1539, the Author- Preaching from formal written sermon to sponta-
ized Version (or King James Bible) of 1611, and neous monologues or even dialogues (as the con-
the Douai Bible of 1609-10. The chief points of gregation reacts), and sometimes involving
controversy then were partly doctrinal, partly sty- elements of song or chant (as in black American
listic, but both stemmed from the major changes preaching, or the hwyl heard in Welsh).
in Christian belief taking place at the time in
Ritual forms used in relation to cultural or social
Europe - changes which led William Tyndale, for
practices, such as baptisms, funerals, confessions,
example, to claim that his translation, unlike pre-
meal-times, remembrance services, weddings,
vious ones, would enable the Bible to be under-
initiation ceremonies, circumcisions, invocations,
stood by all, even by 'a boy that driveth the plough'.
meditations, cleansing rites, oaths, vows, exor-
Questions of level and accessibility remain cen-
cisms, and the blessing of people, objects, or places.
tral today. Nor can the religious translators ever
rest, for at their shoulder is the demon of language Readings from sacred texts, in an original lan-
change (§54). What might seem a 'safe' word today guage or and with varying degrees
in translation,

may be loaded with irrelevant meaning tomorrow. of literalness, formality, and modernity. The texts
A contemporary example is soul, which has deve- themselves will contain a wide range of varieties,
loped fresh meanings in black-American English, such as parables, psalms, historical narrative,
and which would require translators to look care- apocalyptic description, poetry, and paradox.
fully at any version they produced in which the
Doctrinal statements as expounded in official
traditional senses of soul might seem misleading 'canonical' documents, creeds, articles of faith,
or laughable. sutras, expository pamphlets, courses of instruc-
tion, catechisms, and, these days, teach-ins on cable
television. Closely associatedwith this, there is the
Before printing, Bibles had to dimension of theological language, as expounded
be copied by hand - a vast by theologians, biblical scholars, and other relig-
undertaking for a work
ious professionals (p. 51).
whose English translations
generally contain over 2 mil- supernatural
Private affirmations of belief in
lion words and 3* million let-
ters. This is a page from an
beings, expressions of mystical power (such as the

8th century Bible, showing Hindu mantras), expressions of identity and con-
the care with which monks version (such as glossolalia), ecstatic prayers,
'illuminated' the text. In this prophesying, oracles, spirit possession, and testi-
case, the page is largely tak-
en up with the initial phrase,
mony giving.

In principio.

Many communities use a totally different lan-

guage for religious purposes — such as Ge'ez in the
Ethiopian Church, or the traditional use of Latin
in western Christianity. At the opposite extreme,
there are those communities who recommend plain
speaking, or even no speech at all. The distrust
of earthly languages, and the power of silence in
spiritual matters, is especially evident in 17th-cen-
tury Quakerism: 'All languages are to me no more
than dust', wrote George Fox in 1660, and repeated
use was made of the exhortation from Ecclesiastes
to 'let your words be few'.

63 LANGUAGE FOR SPECIAL PURPOSES- 385

 .

The language of the law view, the need for consistency in legal interpre- Legal 'mannerisms'
tation, and for confidence in judgments (which, Wordiness
'The law is a profession of words.' This dictum
they argue, can save much time, anxiety, and annul and set aside = annul
opens David Mellinkoff's classic study The Lan- entirely and completely re-
money), far outweigh the gain that would come
guage of the Law (1963), and it is not possible move = remove
from an increase in popular understanding. totally null and void = void
to find a more succinct way of introducing the pre-
While many lawyers these days accept the desira- without let or hindrance =
sent section. Whatever the legal domain — govern-
bility of some degree of simplification, there is a without hindrance
ment legislation, courtroom activities, or the
natural caution about leaving the safe, charted
documentation that constrains our daily lives (con- Lack of clarity
- domain of traditional legal language, and entering
tracts, conveyances, regulations, by-laws, etc.) The use of lengthy sen-
into a world that may be hiding a host of undisco-
tences containing obscure
we are faced with this fundamental principle: the
vered linguistic pitfalls. Doubtless a certain amount words and awkward con-
words of the law are, in fact, the law. There is
of planned change could be introduced without structions:
no other variety where the users place such store Although the will itself was
harm (simplified documents have in fact increased
on the nuances of meaning conveyed by language, silent as to who would take
in currency in recent years); but this is inevitably
where unstated intentions are so disregarded, and ifthe son predeceased the
a slow process, given the need to think through mother, she not having at the
where the history of previous usage counts for so
carefully the consequences of every change. time of the son 's death re-
much. married, and the son leaving
The overriding concern for precise and consis- issue at his death, which
tent linguistic interpretation has, over the centuries, event occurred, this omis-
produced a highly distinctive style whose complex- sion by itself, in the will only,
cannot aid the son and de-
ity is particularly apparent in the written language.
feat the testator's clear inten-
This style has frequently been criticized by the lay tion that the son should take
public, on the ground that much of legal language only in the event he survived
is unnecessarily complex, and could be simplified the death or remarriage of
his mother.
without loss (p. 378). The point is, indeed, often ..

made by lawyers themselves. Melhnkoff, a profes- Pomposity

sor of law, considers his profession to display four
linguistic 'mannerisms' that warrant criticism:
wordiness, lack of clarity, pomposity, and dullness
(see right).
Historical factors explain the character of pre-
sent-day legal language. Stylistic tradition is a
major influence here, as with several other varieties
(e.g. religious language, p. 384): each and every,
have and hold, null and void, rest residue and
remainder, and many other phrases can be traced
back to Anglo-Saxon, Old French, or Medieval
Latin. The repetition, alliteration, and rhythm of
many expressions (e.g. the truth, the whole truth
. .) reflect the need, in an age before printing and
.
general literacy, for the law to be remembered
clearly, and passed on consistently. The use of tau-
tologous expressions in English documents is often
due to the influence of different languages: for
example, a French or Latin term used alongside
an Anglo-Saxon one (e.g. made and signed, break-
ing and entering, will and testament)
uncertainty of early draftsmen as to whether the
two terms had the same meaning. In such cases,
the safest course of action was to include both.
Historical explanation, however, does not stop
the people in their wisdom; in
the discharge of that impor-
7. Moneys to be invested under this Settle-
(1)
tant duty; trifles with justice;
ment may be invested or otherwise applied on the the result be to weaken
will
security of or in the purchase or acquisition of real or subvert what it conceives
or personal property (including the purchase or ac- to be a principle of the funda-
quisition of chattels and the effecting or maintaining mental law of the land.
of policies of insurance or assurance) rights or inter-
Dullness
ests of whatsoever kind and wheresoever situate Partly due to the above fea-
including any stocks funds shares securities or other tures, but alsoa consequence
investments of whatsoever nature and wheresoever of the tendency to go into the
whether producing income or not and whether
minutiae of a procedure, as
part of its justification:
involving liability or not or on personal loan with
The reason for denying an
or without interest and with or without security to appeal in the latter case is
any person (other than the Settlor or any Spouse not because the order on the
motion to vacate is not within
of the Settlor) anywhere in the world including loans
the terms of section 963 of
to any member of the Specified Class and the Trus-
the code allowing appeals,
tees may grant indulgence to or release any debtor for it may be, and indeed, an
(other than as aforesaid) with or without consider- order refusing to vacate a fi-
ation and may enter into profit sharing agreements nal judgment is in its very na-
and give and take options with or without consider-
ture a special order made
reflects the after the judgment, but be-
ation and accept subsitution of any security . .
cause it would be virtually al-
lowing two appeals from the
same ruling, and would, in
This extract illustrates the traditional style of legal
some cases, have the effect
language. An important feature is the distinctive
graphology (§33). Certain conventions of layout and of extending the time for ap-

the persistent call for change in legal language by typography are present, but there is a total lack of pealing contrary to the intent
punctuation. of the statute.
eliminating archaic or Latinate expressions, simpli-
(After D. Mellinkoff, 1963,
The reasons for this distinctive feature of legal style
fying grammatical structure, and adding punctua- Ch.3.)
are not entirely clear, but they probably have to do with
tion.Those in favour of change argue that this the early use of punctuation as a graphic device to help
would make legal language more intelligible to con- people read texts aloud. Most legal documents are
sumers, saving much and money, and
time, anxiety, purely written records, so there would have been little

would also greatly simplify the job of lawyers them- need punctuate them. When these documents came
to
tobe compositors - doubtless influenced by
printed,
selves. Those who defend the complexity of legal
the inconsistency of manuscript punctuation -
language argue that its characteristics are the developed the practice of printing texts without any
product of centuries of effort to devise an unam- punctuation at all. Gradually, the tradition grew that
biguous, reliable, and authoritative means of regu- punctuation had no part to play in legal writing.

lating human society and resolving conflict. In their

386 X LANGUAGE IN THE WORLD

 . . . .

SPOKEN LEGAL LANGUAGE

The language of the courtroom is familiar to mil- The characteristics of legal language
lions only through the media of films and tele-
(if • Common words with un stood meanings (terms of malice
vision), but only recently has there been any common meanings: art'): nominal sum
action = law suit appeal defendant reasonable care
systematic research into the complex rules of
avoid = cancel bail felony undue interference
linguistic behaviour that participants are expected hand = signature contributory injunction
to follow. Most lay people fail to grasp the extent presents = this legal docu- negligence libel • The use, conversely, of
of these rules until they have had the experience ment words and phrases to ex-
of acting as a witness. After the event, the almost said = mentioned before • Less precise terms and press precise meaning:
specialty = sealed contract idioms, in standard use in irrevocable
universal reaction is frustration - of not being
daily legal discussion inperpetuity
allowed to say what they wanted in the way they (sometimes referred to as nothing contained herein
• Old and Middle English
wanted. Once people are in court, they must follow words no longer in general legal argot'): This convention motivates
its procedures, and use its language; if they do not, usage: alleged the use of long lists of near
issue of law synonyms in documents,
they may be held 'in contempt'. aforesaid thenceforth
forthwith thereby objection as in these phrases from a
There are several everyday functions of language order to show cause standard form, in which a
hereafter theretofore
that witnesses are not allowed to use. They must heretofore whereby strike from the record person is released:
not report what other people have said ('hearsay'), said (adjec- witnesseth superior court

evaluate other people or events ('opinion'), give tive) without prejudice from any and all manner of
action or actions, cause
their listeners extra context (i.e. they must simply • Formal or ceremonial
• Latin words and and causes of action, suits,
'respond to the question'), or show such emotions words and constructions in debts, dues, sums of
phrases, including some
as humour. Similarly, the legal experts are subject that have become part of written documents and in money, accounts, reckon-
to linguistic constraints, such as how to introduce the language as a whole spoken courtroom lan- ings, bonds, bills, special-
(e.g. affidavit, alias, alibi): guage: ties,covenants, contracts,
evidence or cross-examine witnesses. Books have Signed, sealed, and deli-
corpus delicti per stirpes controversies, agreements,
been written on court 'tactics' - how to manipulate vered promises, trespasses,
ejusdem gen- quasi
witnesses, impress judges, and influence juries. eris res gestae Whereas ... (in contracts) damages, judgments,
At a trial, language counts for everything. In ex post facto retraxit You may approach the executions, claims, and de-
in personam suijuris bench mands whatsoever, in law
terms of structural analysis (p. 79, §20), a trial
lex loci actus vis major Comes now the plaintiff or equity, which against
is little more than a giant narrative, with a begin- Your Honour
nolle prose- him have had, now have,
I

ning (the opening statements), middle (the presen- qui May please the court
it or which my heirs, execu-
tation of evidence), and end (the closing arguments Hear ye, hear ye, hear ye tors, or administrators,

and verdict). However, unlike most stories, this one Ido solemnly swear. . hereafter can, shall, or may
• Words derived from
The the whole truth,
truth, have, for or by reason of
is told by many people, including two 'official' French (many now in
and nothing but the truth any matter, cause, or thing
story-tellers (counsel for the defence and for the general use, e.g. appeal,
whatsoever, from the be-
assault, counsel, crime,
prosecution), and exists in at least two conflicting • The conscious use of ginning of the world to the
plaintiff, verdict):
versions. Resolving the conflict depends totally on vague words and phrases day of the date of these
demurrer fee simple
to permit a degree of flexi- presents . .

the linguistic skills of all concerned. easement lien

bility in interpretation:
estoppel tort
adequate cause Fairly comprehensive, one
as soon as possible imagines!
• Technical terms with fair division (After D. Mellinkoff, 1963,
precise and well-under- improper Ch.2.)

Effective courtroom strategies

Lawyers Witnesses
1 Vary the way in which you ask questions. 1 Vary the way in which you give answers.
2. Give your own witnesses a chance to speak at length; restrict the 2. Give long answers wherever possible; make the opposition lawyer stop
opportunity of witnesses under cross-examination to short, direct answers you frequently during cross-examination, to give the impression of reluctance
to the specific questions asked. to have your full story placed before the jury.
3. Convey a sense of organization in your interviews of witnesses and your 3. Try to confuse the organization that the opposition lawyer has planned
remarks to the jury. for the cross-examination.
4. Adopt different styles of questioning with different kinds of witnesses, e.g. 4. different styles of answering questions asked by different
Adopt
women, the elderly, children, expert witnesses. questioners (e.g. deference to the judge, no rehearsed answers while under
direct examination, no hostility to the opposition lawyer).
5. Remain poker-faced throughout; do not reveal surprise even when an 5. Do not show surprise even when questions are unexpected, save
answer is totally unexpected; save dramatic reactions for special occasions. dramatic reactions for special moments.
6. Rhythm and pace are important; do not bore the jury with slowness; use 6. Use rhythm and pace to advantage. Upset the opposition lawyer's pace
silence strategically. with variations in response timing (e.g. asking Would you please repeat the
question? after an especially long or complex question).
7. Repetition can be useful for emphasis but should be used with care
it 7. React to a cross-examiner's repetition of material, e.g. by saying Why

so as not to bore the jury. do you keep asking me the same question ?
8. Avoid interrupting a witness, especially when being given a responsive 8. Interrupt the opposition lawyer by volunteering answers, as soon as you
answer; gives the impression you want to hide some of the facts.
it can see the drift. This gives the impression that you are cooperative and
serves to confuse the lawyer's style.
9. Use objections sparingly; they not only call attention to the material being 9. Blurt out relevant facts and opinions on cross-examination, even though
objected to, but also convey an impression of attempting to conceal the opposition lawyer may attempt to limit your answer. These attempts will

information. give the impression that the lawyer is trying to conceal some of your
evidence.
(After W. M. O'Barr, 1982, Table 7.1 .)

63 LANGUAGE FOR SPECIAL PURPOSES 387

 .

Language and the press

The world of modern newspaper and magazine
publishing presents a wider range of linguistically
If the weather's bad. .

distinctive varieties than any other domain of lan- BLAME THE COMPUTER IN FUTURE
guage study. Within the pages of a daily paper there HAT big black cloud for joined the ranks of the Meteorological
will be juxtaposed such diverse categories as news T.ever hanging over the Office, Bracknell, Berkshire,
reports, editorial comments, imaginative articles,
heads of Britain's weathermen has yesterday.
reviews, letters, captions, headlines, sub-headings,
announcements, television programme descrip- vanished. He can scan half a million weather
tions, of sports results, cartoon dialogues,
lists Those "Yah, ha-ha-got-it-wrong- reports from all over the world —and
competitions, crossword clues, and many kinds of
again" remarks from the weather- come up with the answers in one and
advertising. Sunday papers in some countries are
even more diversified, providing a variety of sup- conscious public won't trouble them a half hours: Blow, blaze, or below

plements on different topics. too much any more. freezing.

With such a range of content, there is no likeli-
For should that "continuing dry" He requires an operating staff of
hood of finding a single style of writing used
forecast develop into a depressing three and another 50 processers to feed
throughout a paper, nor of finding linguistic char-
acteristics that are shared by all papers. Although downpour they can blame the new in weather statistics. But he can cope
each paper has its distinctive visual 'house style' member of the staff: Mr. Comet. with 1 million calculations a second.
(p. 13) and follows a set of general norms laid down Mr. Comet —a £500,000 computer. And he is hardly ever wrong.
by its editorial staff, the idea that there is a homo-
(Daily Express 3 November, 1965)
geneous method of writing used by all journalists
('journalese') seems to have little foundation. 'Jour-
nalese', like 'jargon' and several other labels, is for
many people no more than a loaded word, identify-
Weather Forecasting
ing a style of newspaper writing they dislike.
There are of course certain superficial similarities by Numbers
between newspaper styles arising out of the funda- From Our Science Correspondent
mental constraints of the medium. Information has Revolutions can begin quietly. One aid was given him As well as preparing
.

to be compressed into a limited space, usually in began yesterday at the Meteorological his own forecast chart of pressure
columns, and without loss of legibility. Interest has Office headquarters at Bracknell. For distribution— always the first step in
to be focused, captured, and maintained through the first time in routine procedure an forecasting — he received a second
the use of large type, dramatic headlines, frequent electronic computer contributed to the chart drawn from the computer-made
sub-headings, short paragraphs, and succinct sen- forecast chart published today on this For the issued chart he
calculations.
tences. The occurrence of photographs, the recency page. To look at there is nothing could make use of either or both as
special about the chart. The change for he pleased.
of the information reported, and the need to main-
the forecaster was only that an extra
tain human interest will in variousways influence (The Times, 3 November, 1965)
the choice of vocabulary and grammar. For exam- Style wars The contrast in newspaper styles is best seen when different papers deal
ple, most sentences will be narrative statements with the same story. The opening lines of a 1 965 news item, taken from two London
(rather than questions or exclamations); and the papers - Daily Express (top) and The Times (bottom) - demonstrate the difficulty of
use of the past tense will dominate (except in head- arriving at satisfactory generalizations about newspaper language'.

lines and captions). But within general constraints

of this kind, stylistic preferences vary enormously.
In just a few instances, features of style have Headlinese
developed that are idiosyncratic to the genre of Complaint on eggs upheld (Quoted in G. W. Turner, lines by making a distinc-
newspaper/magazine writing and are frequently Man finds girl in car 1972.) tive use of word play, such
Most headlines differ from However, despite the as these items taken from
used, thus giving credence to the notion of 'journa-
everyday language by syntactic restrictions, there recent issues of the Guar-
lese'.Well-known examples from English are the omitting many of the less an opportunity for
is still dian (London):
altered order of subject and verb (e.g. . . . com- important words in a sen- variation, and in fact the Plans for chess fight un-
mented Dr Brown), and the use of long lists of tence, to produce an ellipti- style chosen for headlines checked
descriptive adjectives (e.g. Tall, blue-eyed, 32-year- cal, telegrammatic' con- and sub-headings often (Arrangements for a chess
struction.They also display provides one of the most match are to go ahead)
old publisher John Brown said .). The distinctive
. .
a very restricted range of distinctive features of a Getting a true bill of fare
grammar of headlines provides a further illust- sentence structures. In re- newspaper. At one ex- (Foreigner who can't afford
ration. But, on the whole, there are few linguistic cent decades, for example, treme we find such plain, his fare home tries to be
features that are restricted to the world of journalis- the English press has unemotional wordings as deported by deliberately
tic writing.
made considerable use of Christmas unemployment not paying a restaurant bill)
on (in the sense of 'about') total at record level. At the Give us this way our daily
as part of 'headlinese': other, we find such drama- bread
Bishops disagree on di- tic (and, out of context, un- (Article on breadmaking)
vorce items as Crash,
intelligible) How to compose yourself
Protest on rail cuts bang, wallop! or Oh yes (A young composer's atti-
New move on libraries she is! Several papers also tude to life)
Concern on smoking add interest to their head-

388 X LANGUAGE IN THE Vi'ORLD

 : —

ctP^ZZryX^J^T&rTjjjvQf j— *. -__z ._
— <£^

u mSAX, tOW (1700 B.C.E.) cKF<5rfi NfWS OF THE PAST VOL. I, NO. 3

SODOM AND GOMORRAH WIPED OUT

IN WORST DISASTER SINCE FLOOD
Burning Sulphur* Violent Blasts, HYKSOS TAKE Mysterious Blaze-Quake
Flames Shooting Up to the Sky... EGYPTIAN Sweeps Valley of Siddim
Bodom and Gomorrah, Many of the houses that cities must all have been News
M well at a ooodly por- lay In the path of the killed or trapped within a CAPITAL (Chronicles Service)

of the Valley of Sid-

flames but had not yet few minutes of the out- HEBRON, 11 Nisan. — Nature's four basic elements — earth, fire,
tion
Mm, art visible from a been touched by them had
already
break of the disaster.
Meanwhile, the nearby
(Foreign

ZOAN. — Memphis,
Sews Service)
wind, and water — combined today to bring terror and death to the
heinht fust east 0/ Hebron collapsed. Indi- capi- twin-cities of Sodom and Gomorrah, in the worst catastrophe the world
(glonei Mamref). A frnr cating that a violent cities of Admah and Ze- tal of northern Egypt and
earthquake had preceded volm were rapidly being has seen since the Great Flood.
early risers were near this the serious
the fire. That also helped evacuated, last obstacle The disaster, a deadly combination of storm, earthquake, fire
spot when the disaster their Inhabit-
struck, and it i* from one account for the fact ants fleeing toward the to the Hyksos' conquest of and flood, struck with a suddenness that provided no opportunity for
that that no one was fleeing range of mountains which the entire country, surren-
of these ei/ev>itnesses
two escape. All the inhabitants of the two cities are believed to have been
tee received the following The Inhabitants of the was serving as our lookout dered to the Invader yes-
point. killed.
account. terday, after a siege which The earth tremors be-
was difficult to make
LOT SAVED! lasted 7 months. gan at sunrise, mounting
I
rise
was watching the sun
behind the hills In HEBRON. IX Nlsan <CNS).
Lot of Sodom and
out what exactly was hap-
pening Inside the cltle«*
because of the heavy black
To all intents and pur-
Advancing Waters In force with such start-
ling swiftness that a few
tileeast when It happen- Judge poses, the capture of Mem- minutes later there was
ed.
the
At first there was Just
sound —
a low, omi-
hN
was
danghterj, are safe. It smoke billowing upward to- phis marks the end of
reliably reported ear- ward the darkened skv
Egypt's 20-year resistance
Threaten Zoar not a house In either
city that was not wreck-
nous rumble which, in a ly this evening. They are The sun had long since t gainst the Hyksos. (Chronicles News Service) or damaged. Whatever
gioment. turned Into a the only persons known to disappeared. Earth and sky Thebes, capital of the sou- From Zoar, near the alive or whole
city are Lot and his dauf
deafening roar. have come oat of the holo- seemed to have become thern zone. Is not expect- southern end of the Sid- ters. first of the refug aulckly
Seeking the source of caust of Sodom an ] Go- one. and there were mo- ed to make more than a comes word —
ments when it appeared dim Valley, and the only ones f
the sound T looked up to morrah alive. token stand. that the population Is flee- Sodom and Gomorrah
the heavens But at that TT»e story of their mira- that the flames were not
ing the city In panic, fol- to reach Zoar this
very moment T became culous escape Is told by a shooting up to the skv but Population Starving lowing the realization that ing.
aware that something was Zoar man who fled to He- that, on the contrary, it the waves of the Salt Sea
The destruction, during Most of the fugitive
transpiring In the vallev bron when his city was was the heavens that were
below. T looked
«»w
— and pouring fire down upon the the early weeks of the
f-nrth. =ifre. of the nvmv water-
are
cirv.
bearing down on the seeking shelter f
nljrht in eaves in fhe
this Is w*~t l
ways in the vicinity upon During- most of the day tains to the w<
Ahln which Mr rr» phis dopendM the people remained calm.
predion wn- 'or Its Wftter supply too;r ^3. except for n few tre-
to-' heightened not only by the ^ heavy toll
among the mors early this morning,
were ranldly being de-
voured by a fiercely burn-
frequent flashes of lightn- flty's
Inhabitants. Tho3? the city (formerly called PRIESTS
CO,
ing that streaked down In-
ing blaze that seemed to
who are not already dead Hela) had been spared In
to the valley and moment-
have sprung up from no- arily lit
are in a more or less ad-
up the ghastly vanced state of dying.
the general disaster thit
had overtaken the vallry.
ON OlSAf
where and was moving In scene, but also by the ap- (Chronicles Nt
a westerly or northwest- The starved and plague- Zoar was considered safe.
pearance of gTeat balls of The priests
erly direction. Portion of Flooded Area that ridden population of the A number of refugees from
fire plunged earth- the stricken cities to tha Salem, and Je
ward, exploded with a re- long-besieged must j

(The spot where T was city

standing soon became a threatened by the waves of sounding boom, and then have presented a horrible north came to Zoar In the called together
common vantage oolnt for the Salt Sea late this after- shot up again in the form sight to the conquerors, but course of the day. to seek Ing. In their nj
a haven behind its wails tltes. for erne,
hundreds of Hebronltes noon. It is doubtful whether 'hev
of belching flames.
Towards evening, how- concerning th
Judging from the pung- were very much disturbed
i

who had been awakened At sunrise this morning,

by the terrific noise and
were hurrvlng to the hill
top. Individually and In
groups, they arrived and
stared, wide-eyed,
horror below. )
Jfirirho RflDorts
watchmen on the northern ent odour which now be- by It, as they went ..bout ever, there came a change morrah dlsast /
for the worse. Details of 1 1
wall of Zoar noticed a gan tt> reach
us on the
their systematic work of
group of people approach* mountain top. I would say pillage and destruction. The men at the lookout are lacking,
lng the gates of the city that these posts on the wall suddenly licved that V
In a half-run. As the fig-
fire-balls were By order of King SaJItu saw something that caused paratlons
at the globs of sulphur .... of the Hyksos, an image the blood to freeze In their flees In
ures came nearer, the Further proof that an of the god Sutekh was set come as a
watchmen recognized Ab- earthquake had taken place up today veins:
in the Temple of The waters of the Salt these disci
raham's nephew. Lot. The was
provided for us shortly Ra. to take the place of the Sea were advancing across Announc
judge Is a well-known fig- before
ure In these parts. With our
noon, when —
to Egyptian deity as top god the valley, along Its entire the pries'/
i

renewed amazement of the land. (The Egyptian width, and were rapidly among a'j
Salt Sea Receding:
him were his two beautiful —
we noticed that a large equivalent of the god Su-
daughters and two uniden- (Continued on P. bearing down on the city! villages
2. Col. 6) tekh is Seth.) Among those fleeing the upon ev<
(Chronicles News Service) tified men.
sacrifice
JERICHO. 11 Nlsan — Visit Unexplained mlly t
I

"Ine waters of the Salt Sea Asked who the strangers pease
are slowly and steadily re-
treating southward, threat-
were, Lot replied that only
he andhis daughters
Scene of Disaster - Before and Aftei gods.
AbraTT
ening to leave the port city sc-jght entrance to the city, Hebron session*
of Beth Hoglah 'high and and without any fur-
so. as the representative ***T'——-^_
dry". ther hesitation, the guards his god. aroused the ire of back^TT^*—
The movement of the sea opened the gates of the
Agere
the local populace by refus- reached suchuTTT—^j>lace»
^^^
began early this afternoon. city to their honoured and lng to attend the discus- as Jericho. Shechem, Beth-
It was supposed at first unexpected guests. 8l0DS -
Shemesh, and Gezer.
that the strange phenome- When Lot declined to
non was merely a "freak state the reason for his
low tide". But this hopeful sudden appearance at Zoar
supposition had gradually at this unearthly hour, it Abraham's Conduct Arouses
to be discarded, as It be- was supposed that there
came increasingly evident had been another dispute
that this was no movement between him and the men Suspicion Among Hebronites
of the tides but, —
ra- of Sodom.
By a Staff Writer
been quicker to reach the
ther, a violent and funda- Speculation on the mat-
mental change In the coun- ter, however, was soon In- Among those who watch- spot than the younger men
try's terrain: terrupted, as the great ho- ed the progress of the ca- who hurried there as soon
The sea is receding, and locaust broke over Sodom tastrophe from the heights as they heard the *lrst noi-
the eye can actually fol- and Gomorrah. around Hebron was Abra- ses, at sunrls?. Abraham
low its progress as it make* ham the son of Terach. mi'st. therefore, have ar-
its way southward, leaving Some of the other obser- rived at hi? vantage point
behind it dry land. vers of the tragedy are some time before sunrise
(At the moment, the wa- LARSA FALLS quick to link the Hebrew's —
and before anything had
ters of the Jordan are name with the disaster that begun to happen In the
struck the Siddun Valley valley below.
spreading over the vacated
area; but. once the river
TO HAMMURABI with such unprecedented (2) Three strangers vl-

63 LANGUAGE FOR SPECIAL PURPOSES 389

 .

psychological and linguistic, are essential: they

The language of advertising
The aim of advertising is to draw attention to a
product or service in order to sell it. Whether we
are shopping, reading the paper, travelling to work,
watching television, or simply lazing around, we
cannot avoid seeing advertisements — probably, if
we bothered to count, several hundred every day.
They come in an extraordinary range of forms and
contexts. The largest and most noticeable group
belong to commercial consumer advertising; but
there are also such categories as trade advertising
(from manufacturers to retailers), retail advertising
(from shops to customers), prestige advertising
combine to produce a single 'brand image' of a
product. However, little objective evidence is avail-
able to show how (or whether) ads succeed in their
aims. A great deal of market research is carried
out by firms and advertising agencies (e.g. asking
people whether they can recall the content of an
ad), but the link between language and sales
remains unclear.
Analyses of advertising style by linguists and pro-
fessional copywriters have drawn attention to
several important features of this variety. Most
obviously, the language is generally laudator)',
positive, unreserved, and emphasizing the unique-

(e.g. by government departments), classified adver- ness of a product {There's nothing like X . .). The

tising (want ads, house sales, etc.), and direct mail- vocabulary' tends to be vivid and concrete. Figura-
ing. The activities involve posters, signs, notices, tive expressions are common {eating sunshine (cer-

showcards, samples, circulars, catalogues, labels, eals), smiling colour (hair shampoo)). Rhythm,

wrapping paper, price tags, tickets, footballers' rhyme, and other phonetic effects are noticeable
shirts, and many other devices. The ears can be {Wot a lot I got, Milk has gotta lotta bottle). There
Poster pillars These adver-
assailed as well as the eyes, with slogans, jingles, may be deviant spellings, especially in the brand tisements for theatres, res-
loudspeaker messages, and the range
street cries, names {Rice Krispies, p. 204). And considerable taurants, political parties,
of auditory effects heard in radio and television use is made of inexplicit grammatical construc- and other local issues, are a
advertising. tions, which lend an air of vagueness — and thus common sight in several Eur-

In most cases, it is the visual content and design safety - to the claims for the product: X gets clothes opean countries.

of an ad that makes the initial impact and causes cleaner (than what?), X costs less (than what?),
us to take note of it. But in order to get people Many people say ... (who?), X treats aches and An ad from a successful
modern campaign, where
to identify the product, remember its name (or at pains (all?).
the language counted for
least make them feel that it is familiar), and per- The field of advertising is a controversial one, everything. The original slo-

suade them that worth buying, ads rely almost

it is
as people dispute the ethics and effects of 'hard' gan, launched by the lager
selling tactics, fraudulent claims, commercial spon- firm, Heineken, read:
totally on the use of language. Both elements,
'Heineken refreshes the
soring in sport, the intrusiveness of advertisements,
parts other beers cannot
A 19th-century theatre poster. The densely
late 18th- early and their effect on children. Its language therefore reach'. This slogan was so
packed language of this advertisement was typical of the needs careful investigation and monitoring. But it successful that it became
period. The long messages in small print provide a sharp
is not an easy field to make generalizations about. possible for the firm, within a
contrast with the short messages in large print which are
Its boundaries blur with other forms of persuasive very short time, to assume
typical today.
everyone knew it, and to
language, such as speeches, sermons, and public
introduce a series of linguis-
announcements. And within the genre, there is so tic jokes based on the word
much variation in subject matter that it is impos- part, e.g. Heineken re-

sible to maintain a single attitude that will en- freshes the parrots other
beers cannot reach', and the
compass everything. Whatever our view about
1 illustration below. This is an
m'r-""*' . ml. HUT advertisements for cigarettes, washing powders, or
sns^SSL-c-— >' ,,0 -' r
cough remedies, it is unlikely to be the same as
unusual use of language,
and an uncommon advertis-

•^rWest Indian. the view we hold about advertisements dealing ing technique. Anyone lack-

with the dangers of smoking, the sale of houses, ing a knowledge of the origi-
nal version would find very it
or the needs of the Third World.
difficult to see the point of

such sentences!

Heineken refreshes the pirates other beers cannot reach.

390 •
X LANGUAGE IX THE WORLD
 .

Advertising
Fair advertising compounds
An illustration produced by and premises, and thus ments for jobs which have agree, point out that, in one Novel compound words,
the Equal Opportunities bears directly on advertis- in the past been done way or another, the ad used as adjectives, are pro-
Commission of Britain, ing language. mainly by men or women must make clear that the bably the most noticeable
showing the force of the In an advertisement only (e.g. mechanic, typist) vacancy is open to both feature of advertising lan-
Sex Discrimination Act of checklist issued by the could not be understood to men and women. guage. Such compounds do
1975. This act states that it Commission, five guide- indicate a preference for • Pictures can give a occasionally become part of
is unlawful to treat anyone, lines are recommended: one sex. the everyday language, such
biased impression too. If
solely because of his or her • If the ad contains words they are used, ensure that as top-quality, economy-
sex, less favourably than • Watch out for words like like he, she, him, her, make men and women are size.
anyone opposite sex.
of the Salesman, Storeman, wo- sure that they are used as shown fairly, in both day-in-day-out protection
Apart from its relevance in man. If these are used, alternatives, e.g. he or she numbers and prominence. satin-soft skin
employment, education, make sure the ad also or him/her, and are consis- Otherwise a bold dis- craftsman-made furniture
and training, the act applies clearly offers the job to both tent throughout the adverti- claimer should be placed top-quality bulbs
to the public provision of sexes. sement. as close to the illustration economy-size packet
goods, facilities, services, • Make sure that advertise- • If your client does not as possible. feather-light flakes
creamy-mild soap
chocolate-flavoured cereal
relief-giving liquid

WAITER/ longer-lasting
coffee-pot-fresh
shave

WAITRESS all-purpose fertilizer

ready-to-eat cereal
REQUIRED
E xpenenoe no! necessary as (raining
up-to-the-minute styling
be g.ven r"~ -anywhere refrigerator
Apply in writing to
INTERNATIONAL RUBBER CO erG.N. Leech, 1966.)
SPORTS ASOCIAL CLUB
Stanley Road. Bonon, Longfield

Top twenties
In a study of the vocabulary
TELEVISION ADVERTISING used in television advertis-
This form of advertising shares many of the linguis- ing, the 20 most common ad-

tic features of consumer advertising in general, but jectives and verbs, in order
of frequency, were:
there are certain important differences. Less use
Adjectives Verbs
is made of written language, partly because of the
1 new 1 make
limited size of screen, but also because there is no
2 good/ better/ best 2 get
time for the viewer to read lengthy material. On 3 free 3 give
the other hand, the medium makes available the 4 fresh 4 have
infinite possibilities of the spoken language, with 5 delicious 5 see
6 full 6 buy
its reliance on voice qualities, spoken dialogues,
7 sure 7 come
adventure dramas, and many other kinds of inter- 8 clean 8 go
action. 9 wonderful 9 know
The simultaneous
use of sound and vision can 10 special 10 keep
11 crisp 11 look
present problems, however. Most advertisements
12 fine 12 need
use speech to make their main linguistic claims,
13 big 13 love
and use writing to reinforce what is said, or to 14 great 14 use
add any disclaimers (e.g. Battery not included, 15 real 15 feel

While supplies last). But it is easy to ignore the 16 easy 16 like

17 bright 17 choose
'small print' on the screen. It is not there for long,
18 extra 18 take
and the viewer may turn away at the crucial 19 safe 19 start
moment. Also, the fact that television advertising 20 rich 20 taste
takes place in real time can lead to difficulties of Good and new were over
twice as popular as any other
comprehension and evaluation. In printed advertis-
adjective. The verbs seem
ing, there is always time to reread the material,
unremarkable, until we recall
and thus to analyse what the ad actually says. On the special contexts in which
television, this opportunity is lacking, and there they are used in this variety,

is thus little chance of evaluating the nature of the eg- '

Made by Smith's.
The barrage of advertising in London's Piccadilly Circus.
daily claims that are made. Some analysts have argued
A walk through the centre of any city places us in contact
Get XXX today.
that these issues pose a particular problem when X gives you everything.
with thousands of advertisements, in the form of posters,
physical models, and neon signs. But it is unlikely that we
considering the kind of advertising that is aimed When you have an X . .

'see' (that is, consciously register) more than a tiny fraction. at children (M. L. Geis, 1982). (After G.N. Leech, 1966.)

63 LANGUAGE FOR SPECIAL PURPOSES- 391

 .

silence are of course important; but radio is par

The language of broadcasting
excellence the speaker's medium. Nowhere else
Editing
Much of the professional

Broadcasting, as a national medium, has existed does the human voice receive such undivided atten- 'finish' of many radio and
television programmes is
only since the 1920s; but its popularity and power tion. And as a consequence, great care is needed achieved by tape editing.
have been so great that it has already given lan- both to maximize its effects (especially through a Even spontaneous inter-
guage several new varieties. People now take for lively use of prosody, §29) and to avoid idio- views and talks can be
granted such styles as newsreading, weather report- syncrasies (which radio tends to exaggerate). cleaned up' before being
ing, programme announcing, disc-jockey patter, Above all, broadcasters have to pay special atten- broadcast by removing major
non-fluencies (such as too
and sports commentary, and they can easily forget tion to the problem of how much listeners can hear
many ers and urns' - a tech-
that these styles are only about two generations and take in at a time. There is no opportunity for nique sometimes called 'de-
old. The medium has also greatly increased popular immediate playback if something is misunder- umming').
awareness of linguistic diversity. An evening's lis- stood. As far as possible, therefore, broadcasting In films, dubbing is another
much-used technique -
tening or viewing provides an encounter with many language has to be clearly organized, and make
modifying or adding to the
regional accents, social dialects, and occupational use of sentences that are relatively short and sound track. When a change
uses of language (Part n). Only the seasoned tra- uncomplicated. of language is involved, this
vellerwould have encountered such a linguistic For the linguist, radio has uniquely interesting is one of the most complex

range a century ago. features. It is person-to-person communication linguistic operations imagin-

able. Accents have to be
There is, accordingly, no such thing as a single that is mouth-to-ear, but not face-to-face, and
matched to convey similar
homogeneous 'language' of broadcasting. In aim- where direct feedback is not possible. The totally social effects, and words
ing to inform, educate, and entertain, the medium auditory world of disembodied sound can involve have to be found that look
reflects all aspects of contemporary society and the emotions and imagination of the listener in like the originals. One writer
clearly illustrates the extent
incorporates most of its language. The result is a ways no parallel. Its simultaneous recep-
that have
of the problem:
range of linguistic variety that exceeds even the tion by millions promotes the language it uses as
Yes can easilybecome /a,
heterogeneity of the press (p. 388): discussions, a standard (e.g. 'BBC English') and gives it an un- especially when the speaker
news reports, soap operas, situation comedies, equalled status and authority within a community is a slack-lipped American,
games, popular science, cartoons, plays, children's (§61). The question of the kind of language profes- but it is hard to turn it into

igen or kylla or naam or ne

programmes, and much more - including, of sional broadcasters should use is therefore a con-
or evet. A cowboy saying yah
course, a considerable proportion of recent cinema troversial one, and in several countries the relative or yeah might just about be
output. merits of standard vs regional and formal vs infor- saying ouais, if not oui, but
mal usage continue to be debated. Sir John Gielgud affirming

The uniqueness of radio with precise actor's diction in

close-up is excruciating to
Because we can see speakers and context, the lan- dub. (Anthony Burgess,
guage used in television programmes plays a less 1980, p. 302.)
prominent role than it does on the radio. Here,
speech is everything. Sound effects, music, and 1ST HOTEL CORRIDOR SICHT STIDIO
Screenplays 1} (o.io) ,
} .

REYNOLDS IS RAPPING AT KINGS DOOR. OBVIOUSLY NOT

Broadcasting language has FOR THE FIRST TIME HE WAITS. LISTENS AT THE DOOR.
its written side too, in the AND THEN. DECIDING THAT KING MUST BE ASLEEP, HE
form of the scripts that pro- TAKES A PLASTIC CARD. OR AN IMPLEMENT, FROM HIS
PERSON. AND UNLOCKS THE DOOR
vide the basis of the pro-

% 4»« L^ grammes. These display

several interesting linguistic

rV : BB c
features of layout and termi-
nology, as can be seen from
from a script by
this extract
Dennis Spooner for a TV ser-
i«. INT.KINGS HOTI-.L ROOM NIGHT STIDIO (» «)

DOOR BEHIND HIM HE LOOKS AROUND THE ROOM

ANOTHER ANGLE
REYNOLDS MOVES ACROSS THE LIVING ROOM. PART OF
i«.

REYNOLDS MOVES INTO KING'S ROOM. AM) CLOSES THE

if
ies (Jason King). This is one
of several possible formats in
THE SUITE. AND OPENS A DOOR THAT LEADS INTO THE
BEDROOM
current use, but several of its
OS R1YSOLDS
conventions are standard.
AS HE LOOKS INTO THE (UNSEEN) BEDROOM. AND NOTES
Each shot is numbered and THAT KING IS NOT THERE HE CLOSES THE DOOR. AND
headed: Ext ('Exterior') vs TURNS AWAY
Int. (Interior), the scene lo- ASOTHER ASCLE
cation, Night vs Day, the REYNOLDS CROSSES TO THE DESK. LOOKING FOR A CLUE
type of set, and the time it AS TO WHERE KING HAS GONE AT THE DESK. WHICH IS BY
A BBC ne wsreader of the 1 950s, dressed in e isuit
THE WINDOW. HE HEARS THE MURMUR. OF VOICES BELOW.
should last. The change of
for the oc( :asion, spoke in a correspondingly fo rmal REYNOLDS TURNS TO THE WINDOW WE ARE ON HIM AS
shot is indicated by cut to (vs
style. But the old order changeth . Fashions 3f radio
. . i HE LOOKS OUT.
language (and, indeed, of clothing) have altere d, just
dissolve). Camera positions
and movements are identi-
as in othe areas of language use. In Britain, th ese
r
fied usingsuch phrases as
changes r lave been particularly noticeable in r scent 15. EXT LOSDOS HOTEL SICHT LOCATION, (oof) 'S
another angle, on .., and
years, wh ere there has been a strong trend tov /ards
POv'Cpoint of view', i.e. the REYSOLDSPOV
the introdi jction of regional and informal speec h.
scene is shot as it would WE SEE KING BELOW. MOVING AWAY. AND GIVING A
However, the new styles have not gone unnoti led by SALUTORY WAVE TO THE COMMISSIONAIRE, WE HEAR.
be seen by the character).
those whc see themselves as defenders of trai jitional
) IN MUFFLED TONES.
Directions are in capitals.
linguistic \ values. The BBC continues to receive i a large KING
(After M. Hulke, 1982,
postbag fr om people complaining about what t hey see A walk rounJ the Park might help mc sleep
p. 205.) AND KING TURNS. AND MOVES AWAY.
as a declii ie in linguistic standards (§1 ).

392 •
X LANGUAGE IN THE WORLD
The power of language General
semantics
A recurrent theme of the present section has been
Alfred Korzybskl (1879- bal labels: different senses Terrorism, religious intoler-
the concern and controversy that can arise when
1950) A Polish-U.S. scien- can be distinguished by ance, bigotry.
people encounter the powerful influence of lan- tist and philosopher whose numbering (e.g. fascism^ A large Florida rattlesnake
guage in special settings. Between professionals, of system of linguistic philoso- and fascism 2 ) or dating in summer.
course, there is no problem: whether the subject phy, known as general se- (e.g. H/f/e/-1930 vs H/f/e/-1939 ),

mantics, attracted consi- to distinguish exactly which Chase found 1 5 distinct

matter is medicine, science, or baseball, the ability
derable popular interest in aspect of a notion is being concepts in the answers,
to use a specialized variety of language is a necess-
the 1 930s and 1 940s (nota- referred to. he comments: 'Multiply the
ary part of professional competence. The difficul- bly through his book Several attempts to ap- sample by ten million and
ties arise only when others come into contact with Science and Sanity ( 1 933) ply these ideas followed, in picture if you can the ag-

it, by accident or design, and find themselves threat-

and the popularizations
that ensued). It still has a
ened by its lack of familiarity or clarity — as happens
certain following, especially
so often in such fields as science, medicine, religion, inthe U.S., though its em-
and the law. Proposed solutions are complex, and phasis on word meaning
range from large-scale recommendations for (rather than on sentences
or contexts) has not made
reform to proposals that accept the linguistic com-
the approach appeal to
plexity, and introduce children to these varieties
modern semanticists (§17).
while at school. General semantics looks
such fields as psychother-
apy and language teach-
ing. One of Korzybski's
popularizers, Stuart Chase,
illustrated the kind of practi-
cal level of analysis in-
volved. In The Tyranny of
Words (1 938), he investi-
gated the way nearly 100
people used the word Fas-
gregate mental chaos. Yet
this is the word which is so-
berly treated as a definite
thing by newspapers, auth-
ors, orators, statesmen,
talkers, the world over.' He
concludes that only a pro-
gramme of semantic train-
ing can help people to con-
trol their reliance on words

In the case of the mass media, the issues are some- the way people
critically at cism: everyone he asked at the expense of realities.

what different. Here the chief anxiety relates to use words without carefully disliked it (it was a 'snarl' 'What the semantic discip-
considering what they word), but there was little line does is to blow ghosts
the use of language to convey the truth. Whether
mean. It recommends the agreement as to what it out of the picture and
we are faced with a newspaper editorial, a radio analysis of meaning as a meant, as can be seen create a new picture as
news report, a film documentary, or a piece of tele- way of promoting mutual from the following selection close to reality as one can
vision advertising, we are confronted with the understanding, both be- of comments: get . . . The probability of

results of language selection: someone has made

tween individuals and A dictator suppressing all better judgements is

across generations. Words opposition. greatly improved, for [a

a decision about what shall be communicated and are seen as deceptively person] is now swayed
One-party government.
what withheld. Inevitably, then, questions arise stable entities, which ob- Obtaining one's desires by more by happenings in the
about the reasoning used, and the form of its scure the variety of mean- sacrifice of human lives. outside world than by re-
linguistic expression. Suspicion about motives is ings that people give to Hitler and Mussolini. verberations within his
them. People continue to War. Concentration camps. skull.' (S. Chase, 1938,
universal: 'Don't believe everything you read/
use the same word even Empiricism, forced control, pp. 129-30, 141.)
hear.' though there may have quackery.
These issues vary in seriousness, depending on been a major change in the Same thing as commu-
the subject matter, and the kind of society in which reality to which once re-
it nism.
ferred. One proposed solu- Exaggerated nationalism.
they are raised. There is an extensive everyday ter-
tion is the indexing of ver- Lawlessness.
minology that illustrates the many ways in which
the abuse of linguistic power shows itself. At one
extreme we are faced with such 'mild' notions as
salesmanship, exaggeration, and sensationalism; at
the other we find a wide range of strongly pejora- Speakers' Corner, Hyde
tive labels, such as bias, prejudice, progaganda, Park A place universally
recognized as a symbol of
misinformation, censorship, indoctrination, brain-
free speech, and used as
washing, and psychological (which usually means such by dozens of amateur
linguistic) warfare. These words are used in all orators each week.
kinds of social situations where people are in con-
flict - most commonly when the conflicts are 'orga-
nized', as in politics, and trade union
religion,
negotiations. However, it impossible to agree
is

about their meaning, as they often function only

as emotional 'snarl words'.
Ironically, there is a far less extensive vocabulary
available to express the various kinds of freedom
and openness of expression that people aspire to.
In a democratic society, there are of course many
kinds of activities that represent freedom in action,
such as public enquiries, opinion polls, and press
conferences. But the concept of openness in public
debate and dissemination seems not to have moti-
vated a corresponding supply of 'purr words' for
everyday use.

63 LANGUAGE FOR SPECIAL PURPOSES • 393

PART XI
Language and communication
In this final Part, the aim is to place the study of
language broader intellectual perspective. This
in a
is done, first of all, by showing how language
relates to other modes of communication, within
the more general heading of semiotics. The main
'design features' of human language are reviewed,
and compared with the properties displayed by ani-
mal communication. There seems to be little in
common: the creativity and structural complexity
of language cannot be found in the natural commu-
nicative behaviour of other species (though the
extent to which language can be taught to chimpan-
zees requires our separate consideration). There is
a clear contrast, also, with the various domains
of nonverbal expression. We look at such areas
as facial expression, bodily gesture, and tactile
communication, as well as at some of the language-
based codes and surrogates that are in use around
the world. By noting the similarities and differences
between language and these other areas, we can
more easily identify the subject's boundaries
though the search for a clear definition of 'lan-
guage' seems no nearer as a result.
A plenary session on syntax and semantics at the XIII
national Congress of Linguists, held in Tokyo in 1 982.
The second and final section outlines the devel-
opment and current scope of linguistics, the science
of language study. The opening perspective is his-
torical, dealing with the emergence of systematic
ideas about the nature of language among the scho-
lars of Ancient Greece, Rome, and India, and trac-
ing the persistent interest in language through the
middle ages and the Renaissance to modern times.
The most rapid period of development has been
the 20th century, where we find unprecedented
progress in linguistic scholarship. A brief account
is given of the history of ideas from the pioneering
statements of Ferdinand de Saussure to the present
day, in particular recognizing the influential role
played by Noam Chomsky in giving direction to
the modern subject. A separate section deals with
the nature of linguistic data,
and with contempor-
ary methods of handling it computationally. Part
xi then concludes with a summary of the main
branches of linguistics, and of the various points
— of contact the subject has with other academic
fields.
Inter-
64 Language and other communication systems

A widely recognized problem with the term 'lan- used by most modern linguists, is to identify the
guage' is the great range ofits application. This various properties that are thought to be its essen-
word has prompted innumerable definitions. Some tial defining characteristics. The aim is to determine
focus on the general concept of 'language', some what 'counts' as a human language, as opposed
on the more specific notion of 'a language'. Some to some other system of communication. Two main
draw attention to the formal features of phonology kinds of enquiry have been used. One focusses
(or graphology), grammar, and semantics (Parts upon identifying the universal structural properties
m-vi). Some emphasize the range of functions that of language, and this is discussed in Part ill
language performs (Parts i, n). Some stress the dif- (§§13-15). The other is to contrast language with
ferences between language and other forms of non-human forms of communication and with
human, animal, or machine communication (see other forms of human communication.
below). Some point to the similarities. At one
extreme, there are definitions that are highly techni- DESIGN FEATURES OF
cal in character; at the other, there are extremely COMMUNICATION
general statements, reflecting the way in which the The most widely acknowledged comparative
notion has been applied figuratively to all forms approach has been that proposed by the American
of human behaviour, such as the 'language' of linguist Charles Hockett (1916- ), who used a

music, cookery, or the cinema.
Most textbooks in the subject avoid the problem,
preferring to characterize the notion of language
rather than define it. They recognize that the ques-
tion of identifying an individual language has no
single, simple answer, because formal and social
criteria are often in conflict (§47). Similarly, they
note the correspondingly complex problems that
arise when attempting to construct a definition of
language in general that makes a precise and com-
prehensive statement about formal and functional
universal properties. The set of definitions given
below exemplifies the way different writers
attempted to tackle the problem, and illustrates
some of the difficulties involved. There seems little
to be gained by trying to summarize the content
of the present volume in a single sentence
it is the banal observation that 'language'
this encyclopedia is about!
A more useful approach to language, and one
zoological mode of enquiry to identify the main
points of connection between language and other
systems of communication, especially those found
in animals. His set of 13 design features of commu-
nication using spoken language were as follows:
• Auditory-vocal channel Sound is used be-
tween mouth and ear, as opposed to a visual,
tactile, or other means (pp. 401-3).
• Broadcast transmission and directional recep-
tion A signal can be heard by any auditory sys-
tem within earshot, and the source can be
located using the ears' direction-finding ability
have 142).
(p.
• Rapid fading Auditory signals are transitory,
and do not await the hearer's convenience
(unlike animal tracks, or writing, §31).
- unless • Interchangeably Speakers of a language can
is what reproduce any linguistic message they can
understand (unlike the differing courtship beha-
viour of males and females in several species).

Language definitions
'Language is a purely hu- Language is 'the institution communicating ideas or pression and the use of as the particular interpre-
man and non-instinctive whereby humans communi- feelings by the use of con- words in human inter- tation that the biologist puts
method of communicating cate and interact with each ventionalized signs, course . . . significant com- upon it and the unravelling

ideas, emotions and desires other by means of habitually sounds, gestures, or marks munication. of its more detailed implica-
by means of voluntarily pro- used oral-auditory arbitrary having understood mean- 4. a special manner or use tions within some currently
duced symbols.' (E. Sapir, symbols'. (R. A. Hall, 1964.) ings. of expression. accepted theoretical frame-
1921.) 2c. an artificially con- (
Webster's Third New Inter- work that nourish the biolo-
'A language a system of
is A dictionary definition structed primarily formal national Dictionary, 1 961 .) gist's day-to-day specula-
arbitrary vocal symbols by 1. the words, their pronun- system of signs and sym- tions and research. So it is

means of which the ciation, and the methods of bols (as symbolic logic) And a comment for the linguist in relation to
members of a society inter- combining them used and including rules for the 'The question "What is lan- the question "What is lan-
'

act terms of their total cul-

in understood by a consider- formation of admissible ex- comparable with
guage?" is guage?"
ture.' (G.Trager, 1949.) able community and estab- pressions and for their - and, some would say, (J.Lyons, 1981, p. 1.)
A language is 'a set (finite or lished by long usage. transformation. hardly less profound than -
infinite) of sentences, each 2a. audible, articulate, 2d. the means by which "What is life?", the presup-
finite in length and con- meaningful sound as pro- animals communicate or positions of which circum-
structed out of a finite set of duced by the action of the are thought to communicate scribe and unify the bio-
elements'. (A. N. Chomsky, vocal organs. with each other. logical sciences ... it is not
1957.) 2b. a systematic means of 3. the faculty of verbal ex- so much the question itself

396 •
XI LANGUAGE AND COMMUNICATION
 5

Total feedback Speakers hear and can reflect

upon everything that they say (unlike the visual
displays often used in animal courtship, which
are not visible to the displayer).
Specialization The sound waves of speech have
no function other than to signal meaning (unlike
the audible panting of dogs, which has a biologi-
cal purpose).
Semanticity The elements of the signal convey
meaning through their stable association with
real-world situations (unlike dog panting, which
does not 'mean' a dog is hot; it is 'part of' being
hot).
Arbitrariness There is no dependence of the
element of the signal on the nature of the reality
to which it refers (unlike the speed of bee 'danc- remote inspace or time from the situation of The 'language' of bees One
ing',which directly reflects the distance of the the speaker (unlike most animal cries, which ref- of the most closely investi-

nectar from the hive). gated forms of animal com-

lect immediate environmental stimuli).
munication is the 'dance'
Discreteness Speech uses a small set of sound Productivity There is an infinite capacity to performed by a honey bee
elements that clearly contrast with each other express and understand meaning, by using old when it returns to the hive,
(unlike growling, and other emotional noises, sentence elements to produce new sentences which conveys precise infor-
where there are continuous scales of variation (unlike the limited, fixed set of calls used by ani- mation about the source and
in strength). amount of food it has dis-
mals).
covered. Several kinds of
Displacement It is possible to talk about events Traditional transmission Language is trans- movement pattern have
mitted from one generation to the next primarily been observed. In the 'round
by a process of teaching and learning (unlike dance' (above, left) used
the bee's ability to communicate the source of when the food source is
close to the hive, the bee
nectar, which is passed on genetically).
moves in circles alternately
The applicability of the 1 3 design features to six systems of Duality of patterning The sounds of language to and right. In the 'tail-
left
communication (after C. F. Hockett, 1 960, pp. 1 0-1 1 ). The have no intrinsic meaning, but combine in differ- wagging dance' (above,
music column refers only to western music since the time of right), used when the source
ent ways to form elements (such as words) that
Bach. A question mark indicates that it is unclear or unknown is further away, the bee
whether a system has a particular feature. A blank space do convey meaning (unlike animal calls, which
moves in a straight line while
indicates that a feature cannot be determined because other cannot be analysed into two such levels of struc- wagging her abdomen from
information is lacking. ture). side to side, then returns to
her starting point. The
Western straight line points in the dir-
ection of the food, the liveli-
Stickleback meadowlark Instrumental
ness dance indicates
of the
Bee dancing courtship song Gibbon calls Language music
how rich is, and
a source it

the tempo of the dance pro-

The vocal-auditory auditory, vides information about its
no no yes yes yes
channel not vocal distance. For example, in
one study, an experimental
Broadcast transmis- feeding dish 330 metres from
sion and the hive was indicated by 1
directional reception yes yes yes yes yes yes complete runs through the
pattern in 30 seconds, where-
Rapid fading ? ? yes yes, repeated yes yes as when the dish was moved
to 700 metres distance, only
Interchangeability limited no ? yes yes ? 1 1 runs were carried out in
? that time. No other animal
Total feedback no yes yes yes yes
communication system
Specialization ? in part yes? yes yes yes seems able to provide such
a quantity of precise informa-
Semanticity yes no in part? yes yes no (in general) tion- except human lan-
guage. (After K. von Frisch,
Arbitrariness no if semantic, 1962.)
yes yes yes

Discreteness no ? ? yes yes in part

Displacement yes, always ? no yes, often

Productivity yes no ? no yes yes

Traditional transmis-
sion probably not no? ? ? yes yes

Duality of patterning no ? no yes

64 LANGUAGE AND OTHER COMMUNICATION SYSTEMS •

397
CHIMP COMMUNICATION A quite different way of proceeding was intro- Washoe's words
The formal and functional complexity of language duced chimpanzee called
in the case of a 5-year-old Washoe's typical vocabulary
is such a distinctive human trait that many scholars Sarah, in a research programme that began in 1954 can be seen from the signs
she used in a study of her
think the designation homo loquens ('speaking (D. &c A. J. Premack, 1983). She (and, later, several
responses to 500 questions.
man') to be a better way of identifying the species others) was taught a form of written language -
The signs were grouped into
than any other single criterion that has been sug- to arrange and respond to vertical sequences of 13 general types:
gested (such as tool using) (p. 291). This is not plastic tokens on a magnetic board. Each token
Proper names (her com-
to disregard the complex patterns that have been represented a word, e.g. small blue triangle =
panions)
observed in the natural communicative systems of apple, small pink square = banana. In due course, Don, DrG, Greg, Roger,
birds, insects, apes, and other animals (the subject the trainer was able to teach Sarah to respond cor- Linn, MrsG, Susan, Washoe
matter of the field of zoosemiotics). But no animal rectly to several basic semantic sequences (e.g. 'give
Pronouns
system remotely compares with the level of sophis- Mary apple'), including a number of more abstract me, we, you
tication found in human language. The evolution- notions, such as 'same/different' and 'if/ then' (e.g.
Common nouns
ary gap is very wide. Only the recent experiments ?apple different banana).
baby dirty nut
in teaching language to chimpanzees have sug- Chimp language research attracted considerable bath drink pants
gested that this gap may be somewhat narrower media publicity in its early years, with reporters bed flower pencil
than has traditionally been assumed. focussing on the implications of the work. What berry food purse
bird fruit ride
Early experiments to teach chimpanzees to com- would chimps say if they could use language? What
blanket gun shoe
municate with their voices failed because of the would they think of the human race? Would they book hammer smoke
insufficiencies of the animals' vocal organs (p. 290). claim civil rights? Such speculations were wholly brush hat spoon
However, when attempts were made to communi- premature, given the limited findings of the bug ice swallow
car key sweet
cate with them using the hands, by teaching a selec- research to date. These findings are in any case
cereal leaf tree
tion of signs from American Sign Language (ASL, controversial, range of reactions
receiving a
chair listen water
see Part vi), dramatic progress was claimed. The extending from total support to total antipathy. cheese lollipop window
first subject was a female chimpanzee named A variety of interpretations seems possible. It is clothes look wiper
Washoe, whose training began in 1966 when she evident that chimps can learn to imitate signs, com- comb man woman
cow meat
was less than a year old. It took her just over four bine them into sequences, and use them in different
years to acquire 132 ASL signs, many of which contexts, but the explanation of this behaviour is Possessives
bore striking similarities to the general word mean- less clear. Many scholars believe that the chimps' mine, yours

ings observed in child language acquisition (Part behaviour can be explained as a sophisticated imi-
Traits
vn). She also began to put signs together to express tation ability rather than as evidence for some form funny, good, hungry, stupid
a small set of meaning relations, which resembled of linguistic processing, and they argue the need
some of the early sentences of young children, such for fuller accounts to be provided of chimp beha- Colours
black, white, green red
as want berry, time drink, there shoe (B. T. 8c R. viour, and of the training methods used, in order
A. Gardner, 1975). to evaluate the claims being made about learning.
Temporal
Since then, several other chimpanzees (and also More systematic data have begun to be collected, time
gorillas) have acquired a vocabulary of signs, and but it will be some time before these questions can
alternative teaching procedures have been tried. be resolved. Negative
enough, no
can't,
For example, in the case of the chimps Moja and
Pili, sign language teaching began soon after birth,
Imperative
and training was carried out by native signers. Both gimme, help
chimps began to sign when they were about 3
months old, and had over a dozen signs by the Appetitive
please, want
age of 6 months - a marked contrast with Washoe,
who had only 2 signs after 6 months of training. Quantitative
hurry, more
NOUNS VERBS CONCEPTS ADJECTIVES

X CONDITIONALS (COLOURS)

^
Verbs

1
^^^T chocolate
is

red
bite, catch, cry, go, hug,
open, peekaboo, smile, tickle

Sarah A
M Locatives
out, up there

H
different in, .

give
Typical sequences
V
no- not
yellow
Me Washoe
You me out
Food fruit
Time drink
take Susan bite Good me
Mary banana
there
name

t
of
brown
»^ Sarah's symbols (left)
apncot
pail
Symbols used in communi-
colour of
'Peony nose touch' Peony, one of the second generation' cating with Sarah and the
of chimps trained in the Premack study, carries out this
instruction, which her trainer has placed on the magnetic
board. (D. Premack & A. J. Premack, 1 983, p. 29.)
Mm
/////////

dish rasin
A
wash
AA7 if-then
green
other chimps (colour is not
shown). (From D. Premack &
A. J. Premack, 1983, p. 21.)

398 • XI LANGUAGE AND COMMUNICATION

 language. Further writing-based codes, such as
Semiotics
semaphore and morse, would also be included here. :-.-.i:;:i:i
Language can also be studied as part of a much Non-linguistic forms of visual communication
wider domain of enquiry: semiology, or semiotics. include the systems of facial expression and bodily
This field investigates the structure of all possible gesture, which are the subject matter of kinesics
sign systems, and the role these play in the way (p. 402).
we create and perceive patterns (or 'meanings') in
sociocultural behaviour. The subject is all-inclu- TACTILE
dealing with patterned human com-
sive, therefore, Tactile communication has very limited linguistic
munication in all its modes (sound, sight, touch, function, apart from its use in deaf— blind commu-
smell, and taste) and in all contexts (e.g. dance, nication and in various secret codes based on
film, politics, eating, clothing). The subject matter spoken or written language (p. 58). Its main uses
of the present book would form but a small section are non-linguistic, in the form of the various ways
of any proposed encyclopedia of semiotics. in which bodily contact and physical distance
between people can signal contrasts of meaning
AUDITORY-VOCAL — the subject matter of proxemics (p. 401).
The diagram below shows the relationship between The communicative use of the visual and tactile
language, as identified in Parts in—vi, and other modes is often referred to as 'nonverbal communi-
aspects of human communication. The structured cation', especially in academic discussion. In every-
use of the auditory— vocal mode, or channel (p. day terms, it is the area of 'body language'.
400), results in the primary manifestation of lan-
guage: speech. But non-linguistic uses of the vocal OLFACTORY AND GUSTATORY
tract are also possible: physiological reflexes, such There seems to be little active role for the olfactory
as coughing and snoring; musical effects, such as and gustatory modes in human communication (a
whistling; and the communication of identity, in marked contrast with the important use of these
the form of voice quality (§6). The suprasegmental senses for communicative purposes in the animal
aspects of vocal expression (§29) are usually kingdom). However, they do play an important
included within the study of language, though it part in our reception of information about the out-
is difficult to draw a clear-cut boundary line side world (e.g. in smelling and tasting food). The
between some of these effects (those placed under communicative use of body odour seems to have
the heading of 'paralanguage', such as giggling and a mainly sexual role in human
society; but there
Other modes? This is the
whispering) and those that clearly fall outside lan- are several anecdotes of use in other domains.
its pictographic message trans-
guage. One linguist even claimed to be able to tell when mitted into space by the Are-
his informants (p. 410) were under strain (and per- cibo radio telescope in Puer-
to Rico in 1 974. The signal
VISUAL haps therefore were being less reliable) by the dif-
was aimed at the cluster of
The visual mode is used for a variety of purposes, ferent body odour they exuded! 300,000 stars, known as
some linguistic, some not. The primary way in M13, in the Hercules constel-
which visual effects have a linguistic use is in the lation.

various deaf sign languages (Part vi). In addition, The message consists of a
series of radio pulses which
there is the historically derivative use of the visual
can be arranged into a picto-
mode that resulted in the development of written gram. It includes data on the
chemical basis of life on
earth, the human form, and
SEMIOTICS the solar system. It assumes,
of course, that the communi-
Language in relation to other
cative system of the receiv-
aspects of semiotics.
ing species is capable of re-
sponding to the same
semiotic contrasts as are dis-
played in the pictogram
(shape, length, etc.). If the
entity receiving the signal

Gustatory happens to have a communi-

cative system based on, say,
heat, the astronomers will

have wasted their time!

The Hercules constellation
is24,000 light years away -
which means that, if any one
or thing is there to receive it,

Voice Sign NX'nting kinesics Deaf- Secret Proxemics and chooses to reply, the
Speech Physiological Musica
vocal reflexes effects qualities languages i
Blind codes response should arrive in
anguage about 50,000 years' time.
Codes

'Body language'
(non-verbal communication

64 LANGUAGE AND OTHER COMMUNICATION SYSTEMS .

399
AUDITORY-VOCAL EFFECTS can convey precise distinctions. With very few
The main systems of communication using the exceptions, each 'syllable' of whistle corresponds
auditory-vocal channel have been described else- to a syllable of speech. Ambiguity is uncommon,
where in this volume (Part iv). However, from time because the topic of the conversation is usually
to time linguists have reported types of auditory something evident in the situation of the speakers.
communication that fall outside the normal use of However, it is important for both speakers to use
the vocal apparatus - notably, the whistled speech the same musical key, otherwise confusion may
of several rural populations. This is found in some arise.

Central and South American tribes, as well as in Whistled dialogues tend to contain a small
the occasionalEuropean community (e.g. in Tur- number of exchanges, and the utterances are short.
key and the Canary Islands, based on Turkish and They are most commonly heard when people are
Spanish respectively). at a distance from each other (e.g. when working
the land), but they can also be found in a variety
Whistled speech of informal settings. Although women are able to
Eusebio Martinez was observed one day standing in understand whistled speech, it is normally used
front of his hut, whistling to a man a considerable only by and between males.
distance away. The man was passing on the trail below,
going to market to sell a load of corn leaves which he
was carrying. The man answered Eusebio's whistle with
a whistle. The interchange was repeated several times
with different whistles. Finally the man turned around,
retraced his steps a short way and came up the footpath
to Eusebio's hut. Without saying a word he dumped his
load on the ground. Eusebio looked the load over, went
into his hut, returned with some money, and paid the
man his price. The man turned and left. Not a word
had been spoken. They had talked, bargained over the Drum signalling
price, and come to an agreement satisfactory to both
In several parts of the world with lips varying in thick- tionsand additions. The
parties - using only whistles as a medium of - notably Africa, the Ameri- ness, thus allowing several Jabo rarely use these
communication. (G. M. Cowan, 1948, p. 280.) cas, and the Pacific - different tones to be pro- drums for communicating
drums, gongs, horns, and duced. Two straight sticks with other villages (unlike
This conversation took place between Mazateco other musical instruments are used for beating, and the drum signalling found in
speakers, of a tribe that lives in and
members have been used to simulate further tonal variations can many other parts of Africa).
around the State of Oaxaca, Mexico. The whistled selected features of be made by altering the The words and syllables

conversations closely correspond to patterns of speech (primarily, tones way these sticks hit the of Jabo are
tonal (§29):
and rhythms). In Africa, drum. Other types of drum there are four basic tones,
spoken language, as has been shown by having the
drums are the usual instru- are also used for different which are often linked by
whistlers translate their tunes into speech. It is thus ments involved, and quite purposes (such as danc- glides, and these interact
quite unlike the unstructured whistling patterns elaborate systems of com- ing)- with aspects of the vowel
used as attention signals (e.g. 'wolf-whistling') in munication have developed. The drummer, an official and consonant system.
Euro-American culture. For example, in the follow- One system, used of the town's law-enforcing There is also considerable
among the Jabo tribe of authority, controls the way variation in the length of
ing sequence of whistled utterances (where the
Eastern Liberia, makes use meetings take place, using these tonal contrasts,
tones are classified from 1 (high) to 4 (low), and of a wooden signal drum' special signals to do such which accounts for several
glides between tones are marked by a dash), quite (actually, more like a bell, things as call for order, of the drum patterns used.
specific meanings are signalled, as the following as it has no skin covering) summon people, and end Some examples of these
- a hollowed tree trunk, of- the meeting. These signals signals, with a transcription
transcription of Mazateco shows:
ten over 2 metres in length. consist mainly of fixed for- in Jabo, are given below.
1,1,3,3,2,4 hme 1
cV st
3
kt
3
-Pai -ve2 4
This has a longitudinal slit mulae, with a few varia- (From G.Herzog, 1945.)
'What did you bring there?'
Pa na 4 hme -ni
1,4,1,1
'It is

1,3,3,4,3hnd
x

a load of corn.'
1
ti
3
-?m? koai
l l

4
-">ni
3
"7
4
T*)H *
ba
2
IJiSj
22 l 2
J*

2
J*

2
M 1
t <
ni di le ba po le kpe'le
'Well where arevou going with it?'
Greetings!' Come ye quick! Put ye your effort there!'
3,2,4,2,3,4 te
3
na nko 4 ti2 -vhP koa 4
2

'I am taking it to Tenango.'

3,3,3,3,2,3,2-4,3 V-fr -
3
ka 3 te2 na 3 -m 2 ^ni 3
'Are you going to sell it then?'
W it i J
itr-j
t
,">•.« J t J * \* i* *H
3 3 2 2~3 3
2,3,3,2,2-3 tf-vhi ka te na cla wle'
2 2
'Gbj nl
4
ba
2
te* •Zk'le
2
ba
2
bj
2
do
3
do
'I am going to sell it.'
Soldiers all! Stop ye the noise. Speak ye one by one!' (Played in the mens assembly
1,1,3,2,4,4,2,3,1-3,4 ho 1
th\
l
PaP-^ni2
4 4 2 3 l -3 4 when the discussion threatens to get out of hand.)
">i -ta te na -nai -vi
'How much
4-3,4,3,3,3,2,4 ka
'I will take
vou take then? Sell it to me here.'
4
will
3
tqg k'oa 3 nka hnko ka
S2.50 a box.'
3 4- 3
2
la
4

(G. M. Cowan, 1948, pp. 284-5.)

ii
•Du
2
i
2
i
bloV
, y ft
2
ka no
*
2 -'
r
t M T T^T
"Gwt'ne
'Gws
1
,

a
3
J>

ml 2 'Du
2
i
2
bloV
Jfch,

(o 1

The whistled tunes are based on the patterns of To collect fines - hunger is raging - we are going to collect fines (indeed)' (Played
before the assembly sends out a group to collect fines imposed by the court.)
tone and rhythm used in the spoken language, and

400 •
XI LANGUAGE AND COMMUNICATION
TACTILE EFFECTS
The communicative use of touching behaviour, The amplified
proxemics, has in recent years attracted a great deal
hand-shake
of research by psychologists, sociologists, and
In a culture where hand-
anthropologists. A very wide range of activities is shaking is a normal form-
involved, as is suggested by this small selection of ality, extra warmth can be
terms expressing bodily contact: expressed only by extra ac-
tivity, such as increased
embrace lay on (hands) punch firmness, longer duration,
guide link (arms) shake (hands) and more vigorous vertical
hold nudge slap movements. The second
hand may also be brought
kick pat spank into play, as shown in the
kiss pinch tickle diagrams, which illustrate
The communicative value of tactile activities is increasing warmth: (a)
hand clasping, (b) arm
usually fairly clear within a culture, as they com-
clasping, shoulder
(c)
prise some of the most primitive kinds of social clasping, and (d) shoulder
interaction (several of the activities are found embracing. (From D. Mor-
between animals). They express such 'meanings' ris, 1977, p. 93.)

as affection, aggression (both real and pretend),

sexual attraction, greeting and leave taking, con-
gratulation, gratitude, and the signalling of atten- and misinterpretations. Latin
easily lead to conflict Distance zones
tion. They operate within a complex system of Americans, for example, prefer to stand much An American study suggests
social constraints: some of the acts tend to be found closer to each other than North Europeans, so that that there may be four proxi-
only in private (notably, sexual touching); some when the former and the latter converse, there may mity zones when people
are specialized in function (e.g. the tactile activities be a problem. The present author recalls one such interact:
• Intimate Less than
carried on by doctors, dentists, hairdressers, or conflict during a conversation with a student from
45 cm, used for intimate rela-
tailors); and some are restricted to certain ceremo- Brazil, who came and stood before him at some tionships.
nies (e.g. weddings, graduation, healing). Everyone 45 cm distance - a normal interaction distance for • Personal Between 45 cm
has a subjective impression about how these activi- her, but too close for him. He instinctively
much and 1 .3 metres, for reason-

ties take place, and what they mean. But there are ably close relationships.
moved back to the distance he found most comfor- • Social consultative Be-
many differences in behaviour between individuals table - nearer 1 metre. However, as he did so, the tween 3 and 4 metres, for
and groups, and it is not easy to make accurate student moved forward, unconsciously maintain- more impersonal relation-
generalizations about society as a whole. ing her own norm. He retreated further, not wish- ships.
It is difficult to study tactile activity in an objec- • Public Above 4 metres,
ing to be so close to the student. After both had
for public figures and public
tive way: a basic problem is how to obtain clear circled the desk several times, he capitulated, and
occasions.
recordings in which the participants are unaware asked her to sit down! (E.T.Hall, 1959.)
of the observer (especially if the behaviour is being
filmed). There are thus few detailed accounts of Tadoma communication The rules of Indian caste

the range of communicative tactile acts in a society, (p. 38) illustrate the point
Tadoma a method of tactile speech communication that
is
even more precisely. Ac-
and of the factors governing their use. It is evident, has evolved between people who are both deaf and blind.
cording to tradition in one
however, that some societies are much more toler- Speech is perceived by placing a hand against the face of part of India, members of
ant of touching than others, so much so that a the speaker and monitoring the articulatory movements
each caste may not ap-
involved. Usually, the thumb is used to sense the movements
distinction has been proposed between 'contact' proach each other within the
of the lips, and the fingers fan out over the side of the face following distances:
and 'non-contact' societies — those that favour and neck. Devised in Norway in the 1 890s, it get its name
touching (such as Arabs and Latin Americans), and from its first use in the U.S. with two deaf-blind children, Tad Brahmins - Nayars: 2 metres
those that avoid it (such as North Europeans and Chapman and Oma Simpson. (R. Vivian, 1 966.) Nayars - Iravans: 8 metres
Several other tactile methods of communication are used Iravans - Cherumans:
Indians). In one study of couples sitting together
with the handicapped, such as braille (p. 280). It is also 1 metres
in cafes, it was found that in Puerto Rico the people possible to 'translate' such codes as morse and finger spelling Cherumans - Nayadis:
touched each other on average 180 times an hour; (p. 225) into tactile form. 20 metres
in Paris it was 110 times an hour; whereas in Lon-
don there was no touching at all (S. M. Jourard,
1963).
The distance people stand from each other, and
p MC|H|^\<f^^] The rules, which are still fol-
lowed in some areas, work in
an additive way: thus, a
Nayadi may not come closer
to a Brahmin than 40 metres
the way they hold their bodies when interacting, (M.Argyle, 1975).
are other important facets of proxemic behaviour.
There are norms of proximity and orientation
within a culture that communicate information
about the social relationship between the partici-
pants. A common research procedure is to observe
1
the point at which people are made to feel uncom-
fortable when
others invade their 'body space', by
moving too close to them (e.g. in a queue, outside
a cinema, on a beach). Any cultural variations can

64 LANGUAGE AND OTHER COMMUNICATION SYSTEMS •

401
VISUAL EFFECTS Eyebrow flashing
The of non-verbal visual communication,
field When people greet each other at a distance, wishing to show that they are ready to make
kinesics, can be broken down into several compo- social contact, they raise their eyebrows with a rapid movement, keeping them raised for
about one-sixth of a second. The behaviour has been noted in many parts of the world, and
nents: facial expression, eye contact, gesture, and
is considered universal (though some cultures suppress it, e.g. the Japanese, who consider
body posture. Each component performs a variety it indecent). We are not usually aware that we use this signal, but it evokes a strong response

of functions. Movements of the face and body can in a greeting situation, and is often reciprocated. To receive an eyebrow flash from someone

give clues to a person's personality and emotional we do not know is uncomfortable, embarrassing, or even threatening. (After Eibl-Eibesfeldt, I.

state. The face, in particular, signals a wide range 1972.) Below: an eyebrow flash made by a Samoan (left) and a Waika Indian (right).

of emotions, such as fear, happiness, sadness,

anger, surprise, interest, and disgust, many of the
expressions varying in meaning from culture to cul-
ture. In addition, the face and body send signals
about the way a social interaction is proceeding:
patterns of eye contact show who is talking to
whom; feedback to the
facial expression provides
speaker, expressing such meanings as puzzlement
or disbelief; and body posture conveys a person's
attitude towards the interaction (e.g. relaxation,
interest, boredom). Several kinds of social context
are associated with specific facial or body beha-
viours (e.g. waving while taking leave). Ritual or Come here?
official occasions are often primarily marked by Beckoning can be carried

a hj
such factors as kneeling, standing, bowing, or out with the palm of the
hand facing up or down.
blessing.
People used to the former
Visual effects interact very specifically with could interpret the latter to
speech. Gestures and head movements tend to coin- mean Go away! The chart
cide with points of emphasis. Hand movements in shows the preferred pat-
tern in countries between
particular can be used to add visual meaning to
Britain and North Africa.
what has been said ('drawing pictures in the air').
(After D. Morris etal.,
Patterns of gaze distinguish the participants in a 1978.)
conversation: a listener looks at a speaker nearly
twice as often as the speaker looks at the listener.
Being humble
They also assist in marking the structure of a con- Body transcription
Points of similarity as well as
versation (§20): for example, speakers tend to look Some of the symbols, or kinegraphs, which have been
difference can be seen in the
up towards the ends of their utterances, thus giving used order to transcribe the various movements of
in
expression of an attitude
their listeners a cue that an opportunity to speak face and body. Different sets of symbols have been
among various cultures. In
devised for different areas of the body: head, face,
is approaching. one early study, the commu-
trunk, shoulder arm wrist, hand fingers,
Several visual effects may well be universal, but nication of humility was
hip/ leg/ankle, foot activity, and neck. The symbols
found to make use of such
the focus of interest in recent years has been on below are from the set for facial activities. (From R.
body postures as the follow-
the cultural differences that can be observed in face L. Birdwhistell, 1952.)
ing:
and body movements. Some societies use many ges- -cr>- Blank-faced 4> $> Slitted eyes
tures and facial expressions (e.g. Italian); others • Join hands over head and
Single raised e e Eyes upward bow (China).
use very few (e.g. Japanese). Moreover, a visual
brow " indicates
(
• Extend or lower arms
effect may seem to be shared between societies, but brow raised) Shifty eyes (Europe).
in fact convey very different meaning. Thus, in • Stretch arms towards per-
France, using a finger to pull down the eyelid means Lowered brow '"<» <»' Glare son and strike them together
that the speaker is aware of something going on, (Congo).
Medial brow 3> Tongue in cheek • Crouch (Fiji, Tahiti).
whereas in Italy the same gesture means that the contraction • Crawl and shuffle forward;
listener must become aware. Cultural variations in Pout walk on all fours (Dahomey).
visual effects are among the first things a foreigner Medial brow
— • Bend body downward
nods Clenched teeth
notices, but it can be very difficult working out (Samoa).
• Permit someone to place
what they mean, and even more difficult deciding
Raised brows *B* Toothy smile a foot on one's head (Fun-
whether one is permitted to use them.
dan, Tonga).
Wide eyed Square smile • Prostrate oneself, face
down (Polynesia).
Wink Open mouth • Bow, extend right arm,
then move it down, up to
Sidewise look S<0>L Slow lick — lips
head, and down again (Tur-
key, Persia).
% ° Focus on auditor <»<2> Quick lick — lips • Throw oneself on the
back, roll from side to side,
Stare oo Moistening lips and slap outside of the thighs
(Batokas).
<§>(§) Rolled eyes CP Lip biting (After M. H. Krout, 1942.)

402 XI LANGUAGE AND COMMUNICATION

Tick-tack talk Sign 'language' sports players or officials can signal the state
One of the most intriguing Many gestural systems have evolved to facilitate of play, or an intention to act in a certain way.
sights at dog tracks and communication in particular situations. They are entertainment a group of performers can
racecourses in Britain is the
often referred to as 'sign languages', but few have coordinate their activities, such as acrobats,
system of tick-tack signing
used to circulate information
developed any degree of structural complexity or musicians.
about the way bets are being communicative range, and it is therefore important theatres/ cinemas ushers can signal the
placed. A signer acts as an to distinguish them from language proper'
'sign number and location of seats.
agent for a group of book- - the natural signing behaviour of the deaf (Part casinos officials can report on the state of
makers who have bought his
vi). Several might properly be described as 'res- play, or indicate problems that might affect the
'twist card', on which the
dogs or horses are given dif- tricted languages' (p. 56). participants in a game.
ferent numbers to those on In many parts of the world, such as India, Thai- sales I auctions auctioneers can convey the
the official race card. The land, and Japan, pantomime and dance have come type and amount of selling and buying.
same set of tick-tack signs is
to use complex systems of symbolic hand gestures aviation marshalling ground staff can send
used by all signers, identify-
in association with facial expressions and body information about the position of an aircraft,
ing the amount of a bet, a
horse or dog number, and movements. The events of a story, its deeper mean- the state of its engines, and its desired position.
the number of a race; but ing, and the emotional states of the characters may radio / television direction producers and
only those who have an indi- all be conveyed in this way. For example, in the directors can signal to performers the amount
vidual signer's twist card will
Bharata Natya-sastra ('principles of dramatic art'), of time available, instructions about level of
be able to interpret what a
number refers to. the 6th-century bc manual of Hindu dance, there loudness or speed of speaking, and information
are over 4,000 picture patterns for the hands [mud- about faultsand corrections.
Number Signs ras). diving divers cancommunicate depth, direc-
1 right hand on top of
tion, time, and the nature of any difficulties they
hat • Religious or quasi-religious groups and secret
2 right hand on nose
have encountered.
societies often develop ritual signing systems so
3 right hand under truck driving drivers can exchange courtesy
that members can recognize and communicate
chin signals, give information about the state of the
4 hand sweeps a with each other. Such signs are used in Freema-
right road, or show they are in trouble.
curve sonry, practised by some 6 million people mainly
heavy equipment drivers people controlling
5 right hand on in the USA and Britain, and in many of the secret cranes, hoists, and other equipment signal the
shoulder
societies of the Far East, such as the Hung direction and extent of movement.
6 sign 5 then 1
Society.
7 sign 5 then 2 fire service firemen can send directions about
8 sign 5 then 3 • Several monastic orders developed signing sys-
the supply of water, water pressures, and the
9 sign 5 then 4 tems of some complexity, especially if their
use of equipment.
10 clap hands members were vowed to silence, as in the case
£5 right hand held up, bookmaking bookies send signals about the
of the Trappist monks, a development of the
palm outwards, number of a race or horse, and its price (see left).
fingers spread medieval Cistercian order.
noisy conditions environmental noise may
£10 both arms held up • Simple signing systems are found in a wide range
with fingers spread
make verbal communication impossible (e.g. in
of professions:
£50 clenched fists held cotton mills) and a signing system may result.
together
£100 left hand held up
with fingers spread
£500 hands outline a cir-

cle
Eurhythmy
£1,000 hands play imagin- The bodily representation of the sound a, expressing
ary piano (grand the meaning of astonishment and wonder, as
piano' = 'a grand' recommended in eurhythmy (R. Steiner, 1 931 p. 40).
,

= £1,000) This approach, developed by the founder of

anthroposophy, Rudolf Steiner (1861-1925), aimed to
Some signs for odds promote a close harmony between the sounds of
Evens arms held in front, speech and patterns body movement. Eurhythmy
of

moving up and down was seen as visible speech', with thebody reflecting
11/10 hands together, form- in its physical shape the forms of sounds as they are

ing a pyramid articulated. The different sounds are interpreted

6/4 one right finger in the symbolically (§30), e.g. u is seen as the expression
left ear-hole of something which chills or stiffens, and this is shown
in the body by a pressing together of the arms and

legs. According to Steiner, The entire universe is

expressed when the whole alphabet is repeated from
beginning to end.'

No bet. I don'l want i! Nine to four agarnsl

'!
wanl lo pay !o lose
I /

64 LANGUAGE AND OTHER COMMUNICATION SYSTEMS •

403
 65 Linguistics

Language has been an object of fascination and supported the former viewpoint. He saw the reality
a subject of serious enquiry for over 2,000 years. of a name to lie in its formal properties or shape,
Often, the observations have been subjective and its relationship to the real world being secondary

anecdotal, as people reflected on such topics as the and indirect: 'no name exists by nature, but only
nature of meaning, ideals of correctness, and the by becoming a symbol.'
origins of language (§§1, 49). But from the earliest These first ideas developed into two schools of
periods, there has also been an objective approach, philosophical thought, which have since been
with scholars investigating aspects of grammar, labelled conventionalist and naturalistic. Modern
vocabulary, and pronunciation in a detailed and linguists have pointed out that, in their extreme
organized way. At the end of the 18th century, forms, neither view is valid (p. 101). However, var-
the subject attracted an increasing number of spe- ious modified and intermediate positions were also
cialists (§50), so much so that it rapidly became argued at the time, much of the debate inspiring
possible to see the emergence of a new field of scien- a profound interest in the Greek language.
Plato (c. 427-c. 347 bc)
tific research with language analysis as its focus. Another theoretical question was discussed at
This approach, first known as philology, dealt this time: whether regularity (analogy) or irregular-
exclusively with the historical development of lan- ity (anomaly) was a better explanation for the
guage. In the present century, the subject has linguistic facts of Greek. In the former view, lan-
•flRLSTl
broadened to include the whole range of subject guage was seen to be essentially regular, displaying
matter represented in this book, and it is now symmetries in its rules, paradigms, and meanings.
generally called linguistics (or linguistic science). In the latter, attention was focussed on the many
Linguistics today is a widely practised academic exceptions to these rules, such as the existence of
discipline, with several domains of application (p. irregular verbs or the lack of correspondence
408). between gender and sex (p. 93). Modern linguistics
does not oppose the two principles in this way:
Early history languages are analysed with reference to both rules
and exceptions, the aim being to understand the
A religious or philosophical awareness of language relationship between the two rather than to deny
can be found in many early civilizations (p. 384). the importance of either one. The historical signifi-
In particular, several of the important issues of lan- cance of the debate is the stimulus it provided for
guage analysis were addressed by the grammarians detailed studies of Greek and Latin grammar.
and philosophers of Ancient Greece, Rome, and In the 3rd century bc, the Stoics established more
India. formally the basic grammatical notions that have
since, via Latin, become traditional in western
THE GREEKS thought. They grouped words into parts of speech, Aristotle (384-322 bc)
The debate is found in
earliest surviving linguistic organized their variant forms into paradigms, and
the pages of Plato (c. 427-347 bc). Cratylus is a devised names for them (e.g. the cases of the noun).
dialogue about the origins of language and the Dionysius Thrax (c. 100 bc) wrote the first formal
nature of meaning — first between Socrates and grammar of Greek - a work that became a standard
Hermogenes, then between Socrates and Cratylus. for over 1,000 years.
Hermogenes holds the view that language origi- The focus throughout the period was entirely
nated as a product of convention, so that the rela- on the written language. The word grammar
tionship between words and things is arbitrary: 'for (Greek: grammatike) in fact originally meant 'the
nothing has its name by nature, but only by usage art of writing'. Some attention was paid to basic
and custom'. Cratylus holds the opposite position, notions concerning the articulation of speech, and
that language came into being naturally, and there- accent marks were added to writing as a guide to
fore an intrinsic relationship exists between words pronunciation. But the main interests were in the
and things: 'there is a correctness of name existing fields of grammar and etymology, rather than pho-
by nature for everything: a name is not simply that netics. A doctrine of correctness and stylistic excel-
which a number of people jointly agree to call a lence emerged: linguistic standards were set by
thing.' The debate is continued at length, but no comparison with the language of the ancient
firm conclusion is reached. writers (e.g. Homer). And as spoken Greek (the
The latter position is more fully presented, with koine) increasingly diverged from the literary stan-
divine origin being invoked in support: 'a power dard, we also find the first arguments about the
greater than that of man assigned the first names undesirable nature of linguistic change (§1): the
to things, so that they must of necessity be in a language had to be preserved from corruption.
correct state.' By contrast, Aristotle (384-322 bc)
in his essay De interpretation ('On interpretation')

404 XI LANGUAGE AND COMMUNICATION

THE ROMANS
Roman Greek precedents
writers largely followed
and introduced a speculative approach to language.
On the whole, in their descriptive work on Latin,
they used Greek categories and terminology with
little change. However, the most influential work
of the Roman period proved to be an exception
to this trend: the codification of Latin grammar
by Marcus Terentius Varro (116-27 bc) under the
headings of etymology, morphology, and syntax.
De lingua latina ('On the Latin language') consisted
of 26 books, though less than a quarter of these
survive. Varro's work takes into account several
differences between Latin and Greek (e.g. the ab-
sence of the definite article in the former). He also
held the view (which is remarkably modern) that
language is first and foremost a social phenomenon
with a communicative purpose; only secondarily
is it a tool for logical and philosophical enquiry.
Especially towards the end of the millennium,
several authors wrote major works in the fields
of grammar and rhetoric (§12) — notably, Cicero
(106-43 bc) on style, and Quintilian (1st century
ad) on usage and public speaking. Julius Caesar
wrote on grammatical regularity — it is said, while
crossing the Alps on a military campaign. Aelius
Donatus (4th century ad) wrote a Latin grammar
(Ars maior) that was used right into the middle
ages, its popularity evidenced by the fact that it
was the first to be printed in wooden type, and
had a shorter edition for children (the Ars minor).
In the 6th century, Priscian's Institutions gramma-
ticae ('Grammatical categories') was another
influential work that continued to be used during
the middle ages: it contains 18 books, and remains
the most complete grammar of the age that we
have.
i s v
~>^L'
The main result of the Roman period was a
model of grammatical description that was handed
down through many writers in Europe, and that
ultimately became the basis of language teaching
in the middle ages and the Renaissance. In due
course, this model became the 'traditional'
approach to grammar, which continues to exercise
its influence on the teaching of English and other
modern languages (§§1, 62).
THE INDIANS
During the above period, techniques of minute des-
criptive analysis were being devised by Indian
linguists, which could have been of great influence
had these descriptions reached the western world
(something that did not take place until the 19th
century). The motivation for the Indian work was
quite different from the speculative matters that
attracted Greek and Roman thinkers (though they
did not ignore those topics). The Hindu priests
were aware that their language had diverged from
that of their oldest sacred texts, the Vedas (p. 384),
in both pronunciation and grammar. An important
part of their belief was that certain religious cere-
monies, to be successful, needed to reproduce accu-
rately the original form of these texts. Change was
not corruption, as in Greece, but profanation.
Several ancillary disciplines (Veddnga, 'limbs of the
Vedas'), including phonetics, etymology, grammar,
and metrics, grew up to overcome this problem.
Their solution was to establish the facts of the
old language clearly and systematically and thus
to produce an authoritative text. The earliest evi-
dence we have of this feat is the work carried out
by the grammarian Panini, sometime between the
5th and 7th centuries bc, in the form of a set of
4,000 aphoristic statements known as sutras
('threads'). The AstadhyayT ('Eight books'), dealing
mainly with rules of word formation, are composed
in such a condensed style that they have required
extensive commentary, and a major descriptive tra-
dition has since been established. The work is
remarkable for its detailed phonetic descriptions:
for example, places of articulation are clearly des-
cribed, the concept of voicing is introduced, and
the influence of sounds on each other in connected
speech is recognized (the notion otsandhi). Several
concepts of modern linguistics derive from this tra-
dition.
Cicero (Roman Imperial copy of a late contemporary portrait)
65 LINGUISTICS .
405
THE MIDDLE AGES language study. Missionary work produced a large
Very little is known about the development of quantity of linguistic material, especially from the
linguistic ideas in Europe during the 'Dark Ages', Far East. The Chinese linguistic traditions were dis-
though it is evident that Latin, as the language of covered. Arabic and Hebrew studies progressed,
education, provided a continuity of tradition the latter especially in relation to the Bible. In the
between and medieval periods. Medieval
classical 16th century, several grammars of exotic languages
learning was founded on seven 'arts', of which came to be written (e.g. Quechua in 1560). There
three — grammar, dialectic, and rhetoric — formed was a more systematic study of European lan-
one division, known as the trivium. Grammar guages, especially of the Romance family. The first
(mainly using Priscian and Donatus) was seen as grammars of Italian and Spanish date from the 15th
the foundation for the whole of learning. A tradi- century. Major dictionary projects were launched
tion of 'speculative' grammars developed in the in many languages. Academies came into being (p.
The 'first

13th and which grammatical

14th centuries, in 4). The availability of printing led to the rapid dis-
grammarian'
notions were reinterpreted within the framework semination of ideas and materials. The Prose Edda is a 1 3th-
century textbook on poetic
of scholastic philosophy. The authors (the Modis- As we approach modern times, fresh philosophi-
style and construction, writ-
tae')looked to philosophy for the ultimate explana- cal issues emerged. The 18th century is character- ten by the Icelandic chief,
tion of the rules ofgrammar. A famous quotation ized by the arguments between 'rationalists' and Snorri Sturluson. Appended
from the period states that it is not the grammarian 'empiricists' over the role of innate ideas in the to the manuscript are four
but 'the philosopher [who] discovers grammar' development of thought and language. Such ideas treatises on grammar, writ-
ten in the mid-1 2th century,
{philosophus grammaticam invenit). The differ- provided the basis of certainty in knowledge, the first of which has at-
ences between languages were thought to be super- according to Cartesian philosophy, but their exis- tracted special attention be-
ficial, hiding the existence of a universal grammar tence was denied by philosophers (such as Locke, cause of the originality of its

(§14). Hume, and Berkeley) for whom knowledge derived thought. The authorship of
this 'First Grammatical Trea-
The middle ages also saw the development of from the way the mind operated upon external
tise' is unknown, but the
western lexicography (§18) and progress in the sense impressions. The issue was to resurface in writer has come to be known
field of translation, as Christian missionary activity- the 20th century (p. 409). as the first grammarian'. The
increased. In the East, Byzantine writers continued Several other important trends have been noted picture shows part of the

during the 17th and 18th centuries: the breakdown opening leaf of the manu-
to expound the ideas of the Greek authors. There
script.
was a strong tradition of Arabic language work of Latin as a universal medium of communication,
This early exercise in
related to the Qur'an (which was not to be trans- and replacement by modern languages (§59);
its spelling reform (p. 215) sum-
lated, p. 384). From around the 8th century, several the many proposals for universal languages, marizes and illustrates the

major grammars and dictionaries were produced, shorthand systems, and secret codes (§§33, 58); principles needed to improve
the use of the Latin alphabet
as well as descriptive works on Arabic pronuncia- the beginnings of a systematic approach to phone-
for writing Old Icelandic. It is
tion. For a long time, these remained unknown in tics(§27); the development of 'general' grammars, the only work of this period
Western Europe. Opportunities for contact with based on universal principles, such as the 17th- to draw attention to the pro-

the Greek, Arabic, and Hebrew linguistic traditions century grammar of Port Royal (§14); and the blems involved in applying
Latin letters to a vernacular
only came later, as a result of the Crusades. major elaborations of traditional grammar in
language. It contains several
schools (§1). Then, as the 19th century
acute phonetic observations,
THE RENAISSANCE approached, the first statement about the historical and, in its emphasis on find-
The rediscovery of the Classical world that came relationship between Sanskrit, Greek, and Latin ing symbols to express

was made, ushering in the science of comparative sound contrasts, anticipates

with the 'revival of learning', as well as the disco-
the basis of 20th-century
veries of the New World, transformed the field of philology (§50).
phonological theory (§28).

A translation of part of the

opening page (E. Haugen,
1972):

Ihave written an alphabet for

*w* $*?.. e ]pm mm or mmimfo$mr fnktfjio arafarffcu. .***
us Icelanders also, in order
that it might become easier
9**c loo Cm feu mn abafcr \fa |juft i tmvnw Hf aj Rn 'k-umooX to write and read ... have I

used all the Latin letters that

seemed to fit our language

pxtvm ftawi> k^j 01 x cnafamr io«u ^cmi rr» (jnfKlajnro fbjmm

atfa
well and could be
nounced, as well as some
rightly pro-

fj^
W
* * ljm? ™ «** %u xtxm m nj klfo, mji'mij eifftonw
«ir4ro ftapx nc
'

,
i

|£p rrR|r f1Tlu f^.- tmrbioMinsTa £u1)ufrt£a

iri „ y
W other letters that seemed
needful to me, while those
were taken out that did not

-nm#> snuS fT> f A Srurm ™*ti fh*a jw v& fomr/fc* v*m a* hi at su}»
suit the sounds of our lan-
guage. Some of the conso-
nants of the Latin alphabet
were rejected, and some
new ones added. No vowels
ip na l^t^ti-T-m ram* to iinfciiit, M |wir^ wte»T bak*** were rejected, but a good
many were added, since our
language has the greatest
number of vowel sounds.

406 XI LANGUAGE AND COMMUNICATION

Twentieth-century linguistics Saussurean principles
The growth of modern linguistics, from the end Some of Saussure's most is being played, it is pos- signified', or 'concept').
of the 18th century to the present day, has in large central ideas were ex- sible toassess the state of This relationship of signi-
part already been summarized in earlier sections pressed in the form of pairs the game by studying the fied to signifier Saussure
of concepts: position of the pieces on calls a linguistic sign. The
of this volume. The majority of the concepts used
C the board. sign is the basic unit of
in the discussion of language history, acquisition,
communication within a
structure, substance, and use stem from this pers- Langage vs langue vs community: langue is seen
pective (reflecting the background of the author). parole as a 'system of signs'.
The many senses of the
However, there remain several loose ends of a his- word 'language' prompted Syntagmatic
torical and theoretical nature that need to be drawn Saussure to introduce a
together in this final part of the book. three-fold set of terms, the She + can +
last two of which were cen-
EUROPE AND AMERICA D
tral to

is
his thinking.
the faculty of speech pre-
Langage

Two main approaches to language study, one Euro- sent in all normal human
pean, one American, unite to form the modern sub- Diachrony vs synchrony beings due to heredity - He will run
ject of linguistics. The first arises out of the aims He sharply distinguished our ability to talk to each
historical ('diachronic') and other. This faculty is com-
I may sit
and methods of 19th-century comparative philo-
non-historical ('synchro- posed of two aspects: lan- You might see
logy (§50), with its focus on written records, and
approaches to lan-
nic') gue (the language system)
its interest in historical analysis and interpretation. guage study. The former and parole (the act of etc. etc. etc.

The beginning of the 20th century saw a sharp sees language as a contin-
change of emphasis, with the study of the principles ually changing medium; the
latter sees it as a living
governing the structure of living languages being
whole, existing as a 'state'
introduced by the Genevan linguist, Ferdinand de at a particular moment in
Saussure (1857—1913). Saussure's early work was time. In his diagram, AB re-
in philology, but he is mainly remembered for his presents a synchronic 'axis
theoretical ideas, as summarized in the Cours de
- a lan-
of simultaneities'
guage state at some point
linguistique generate ('Course in general linguis-
in time; CD is a diachronic
tics'), which is widely held to be the foundation 'axis of successions' - the
of the modern subject. This book was in fact pub- historical path the lan-
lished posthumously in 1916, and consists of a guage has travelled.
In this view, it is always
reconstruction by two of Saussure's students of his
necessary to carry out
lecture notes and other materials. some degree of synchronic
The second approach arose from the interests work before making a
and preoccupations of American anthropologists, diachronic study: before we
can say how a language
who were concerned to establish good descriptions
has changed from state X
of the American Indian languages and cultures to state Y, we need to know
before they disappeared. Here, there were no writ- something about X and Y.
ten records to rely on, hence historical analysis was Correspondingly, a syn-
ruled out. Also, these languages presented very dif- chronic analysis can be
made without referring to
ferent kinds of structure from those encountered
history. Saussure illus-
in the European tradition. The approach was there- trates this point using an
fore to provide a careful account of the speech pat- analogy with a game of
terns of the living languages. A pioneer in this field chess: if we walk into a
room while a chess game
was Franz Boas (1858-1942), who published the
first volume of the Handbook of American Indian
speaking). The former is
the totality of a language,
which we could in theory
discover by examining the
memories of all the lan-
guage users: 'the sum of
word-images stored in the
minds of individuals'.
Parole is the actual, con-
crete act of speaking on
the part of a person -a
dynamic, social activity in
particular time and place.
Signifiantvssignifie
Saussure recognized two
sides to the study of mean-
ing, but emphasized that
the relationship between
the two is arbitrary (p. 404).
His labels for the two sides
are signifiant ('the thing
that signifies', or 'sound im-
age') and signify ('the thing
Syntagmatic and associ-
ative (or paradigmatic)
A sentence is a sequence
of signs, each sign contri-
buting something to the
meaning of the whole.
When the signs are seen
as a linear sequence, the
relationship between them
is called syntagmatic, as in
a She + can + go. When a
sign that' is present is seen
as contrasting with other
signs in the language, the
relationship is called asso-
ciative (in later studies,
paradigmatic), as in Shevs
He, can vs will, go vs run
in the above sentence.
These two dimensions of
structure can be applied to
phonology, vocabulary, or
any other aspect of lan-
guage. The result is a con-
ception of language as a
vast network of interrelated
structures and mutually de-
fining entities - a linguistic
system.

Languages in 1911. Ten years later, another

anthropologically oriented book appeared: Lan-
guage by Edward Sapir (1884—1939). These works
proved to be a formative influence on the early
development of linguistics in America. The new
direction is forcefully stated by Boas (p. 60): 'we
must insist that a command of the language is an
indispensable means of obtaining accurate and
thorough knowledge, because much information
can be gained by listening to conversations of the
natives and by taking part in their daily life, which,
to the observer who has no command of the lan-
guage, will remain entirely inaccessible'.

65 LINGUISTICS •
407
LATER DEVELOPMENTS Schools of thought
Both European and American approaches devel-
Many different approaches to linguistics emerged in
oped rapidly. In Europe, Saussure's ideas were the middle decades of this century, some of which
taken up by several groups of scholars (especially have attracted a great deal of support. The
in Switzerland, Czechoslovakia, France, and Den- distinguishing feature of five of these approaches is

mark), and schools of thought emerged based on outlined below. (For corresponding developments
within the field of generative linguistics, which has
Saussurean principles (notably, the Linguistic Cir-
been dominant since the 1 960s, see p. 409.)
cle of Prague, which was founded in 1926). The
Functional sentence perspective
field of phonology (§28) was the first to develop,
An approach used by the Prague School of linguists
with later progress coming in such areas as gram- to analyse utterances in terms of their information
mar and style. Saussure's influence continues to be content, and still widely used in Czechoslovakia and
strong today, with his notion of a language 'system' other East European countries. The semantic
becoming the foundation of much work in semio- contribution of each major element in a sentence is
Roman Jakobson (1896-
rated with respect to the 'dynamic' role it plays in
tics and structuralism (pp. 79, 399). 1982) Jakobson was one of
communication. the founders of the Prague
In America, the development of detailed proce-
Dependency grammar School of linguistics and a
dures for the study of spoken language also led major contributor to many
A type of formal grammar developed in the 1 950s,
to progress in phonetics and phonology, and espe- notably by the French linguist, Lucien Tesniere (1893- fields of study in Slavonic
cial attention was paid to the distinctive morpho- 1954). It explains grammatical relationships by setting languages and general
logy and syntax (§16) of the American Indian up 'dependencies' (or 'valencies') between the linguistics. In 1941 he moved

elements of a construction. to America, where he held

languages. The first major statement synthesizing
professorial posts at Colum-
the theorv and practice of linguistic analysis was Tagmemics bia and Harvard between
Language by Leonard Bloomfield (1887-1949), A theory developed since the 1 950s by the American
1946 and 1967.
which appeared in 1933. This book dominated
linguist, K. L Pike (191 2-), which focusses particularly
on the need to relate linguistic 'forms' and 'functions'.
linguistic thinking for over 20 years, and stimulated A central the contrast between the 'emic'
notion is

many descriptive studies of grammar and phono- units, which are functionally contrastive in a language

logy. In due course, the Bloomfieldian approach (such as phoneme and morpheme), and the 'etic' units
that give them physical shape (cf. phonetics, §28).
came to be called 'structuralist', because of the var-
ious kinds of technique it employed to identify and Stratif icational grammar
A theory devised by the American linguist S. M. Lamb
classify features of sentence structure (in particular,
(1 929-) in the 1 960s that views language as a system
the analysis of sentences into their constituent
of related layers ('strata') of structure.
parts, p. 96) It also represented a behaviourist view
.

Systemic linguistics
of linguistics, notably in its approach to the study
A theory developed since the 1 960s by the British
of meaning (p. 101). However, its appeal dimi- linguist, M. A. K. Halliday (1 925-) in which grammar is

nished in the 1950s, when there was a sharp reac- seen as a network of 'systems' of interrelated contrasts;
tion against the limitations of structural linguistic particular attention is paid to the semantic and
pragmatic aspects of analysis (§§1 7, 21 ) and also to
methods, especially in the area of grammar (p. 96).
the way intonation is used in the expression of
This extract from an obituary of Bloomfield, meaning
Louis Hjelmslev (1899-
(§29).
written by Bernard Bloch in the journal Language 1 965) The leading theoreti-

in 1949 (p. 93), summarizes this scholar's achieve-

cian of the Copenhagen
John Rupert Firth (1890-1960) (below, left) J. R. Firth, Pro- School of linguistics, Hjelms-
ment: fessor of General Linguistics in London from
the University of lev propounded a formal ap-
1 944 to 1 956, was a key figure in the development of British proach to language study in
There can be no doubt that Bloomfield's greatest
linguistics. A central notion in his approach is that the patterns the 1 930s known as glosse-
contribution to the study of language was to make a
of language that appear at a particular level of description matics.
science of it. Others before him had worked
(§13) cannot be explained using a single analytic system.
scientifically in linguistics; but no one had so »«
Different systems may need to be set up at different places, rui< ip\ i"
uncompromisingly rejected all prescientific methods, or inorder to handle the range of contrasts involved (an
had been so consistently careful, in writing about approach known as polysystemicism).
language, to use terms that would imply no tacit reliance
on factors beyond the range of observation ... It was
Bloomfield who taught us the necessity of speaking
Leonard Bloomfield (below, right)
0U*
about language in the style that every scientist uses when
he speaks about the object of his research: impersonally,
precisely, and in terms that assume no more than actual
$%
observation discloses to him.
k
Bloomfield's opposition to unscientific impression-
ism
the
in language studies is neatly summarized by
wry comment he made on one occasion: if
you want to compare two languages, it helps to
know one of them!'
\
Daniel Jones (1881-1967)
The leading British phoneti-
cian in the first half of this
century, Jones was Profes-
sor of Phonetics at the
University of London from
1921 to 1949.

408 •
XI LANGUAGE AND COMMUNICATION
CHOMSKY to make a contribution to our understanding of
In 1957, Avram Noam Chomsky, Professor of the nature of the human mind.
Linguistics at the Massachusetts Institute of Tech- The essence ofthe approach is summarized by
nology (1928— ), published Syntactic Structures, Chomsky 1986 book {Knowledge of Lan-
in a

which proved to be a turning point in 20th-century guage, p. xxvi) as providing an answer to the ques-
linguistics. In this and subsequent publications, he tion 'How comes it that human beings, whose
developed the conception of a generative grammar contacts with the world are brief and personal and

(p. 97), which departed radically from the structur-

limited, are nevertheless able to know as much as

alism and behaviourism of the previous decades. they do know?' By studying the human language
Earlier analyses of sentences were shown to be faculty, it should be possible to show how a person

inadequate in various respects, mainly because they
failed to take into account the difference between
'surface' and 'deep' levels of grammatical structure.
At a surface level, such sentences as John is eager
to please and John is easy to please can be analysed
in an identical way; but from the point of view
of their underlying meaning, the two sentences
diverge: in the first, John wants to please someone
else; in the second, someone else is involved in
pleasing John. A major aim of generative grammar
was to provide a means of analysing sentences that
constructs a knowledge system out of everyday
experience, and thus move some way towards solv-
ing this problem.
Noam Chomsky
A major feature of Chomsky's approach was the
technical apparatus he devised to make the notion
of competence explicit - the system of rules and
symbols that provides a formal representation of
the underlying syntactic, semantic, and phonologi-
cal structure of sentences (aspects of thisapparatus
have been referred to in earlier sections, pp. 97,
1 07, 1 62) A primary notion - the transformational
.

took account of this underlying level of structure. rule (p. 97) — led to the approach being commonly

To achieve this aim, Chomsky drew a fundamen-
tal distinction (similar to Saussure's langue and
parole) between a person's knowledge of the rules
of a language and the actual use of that language
in real situations. The first he referred to as compe-
tence; the second as performance. Linguistics, he
referred to as transformational grammar (or TG).
Since the 1950s, much of linguistics has been taken
up with proposals to develop the form of generative
grammars, and the original theory has been refor-
mulated several times. During the same period,
also, there have been several proposals for alterna-

argued, should be concerned with the study of com- tive models of grammatical analysis to those

petence, and not restrict itself to performance — expounded by Chomsky and his associates, some
something that was characteristic of previous of which have attracted considerable support. As
linguistic studies in their reliance on samples (or a consequence, linguistic theory, the core of scienti-
fic language study, is now a lively and controversial
'corpora') of speech (e.g. in the form of a collection
field.
of tape recordings). Such samples were inadequate
because they could provide only a tiny fraction of
the sentences it is possible to say in a language;
they also contained many non-fluencies, changes Some theoretical consequences of generative theory
of plan, and other errors of performance. Speakers Since the 1960s, several tivegrammar. Further an ab-
for alternatives to
fresh theoretical ap- levels of phrase structure stract,formal approach to
use their competence to go far beyond the limi-
proaches to grammatical are recognized, and dis- grammar. This particular
tations of any corpus, by being able to create and
analysis have emerged, tinguished using different theory adopts a pragmatic
recognize novel sentences, and to identify perfor- most of which can be seen numbers of bar symbols. view of language as social
mance errors. The description of the rules govern- as a development of interaction (§21), and sets
ing the structure of this competence was thus the Chomsky's proposals - or Montague grammar up 'functional' units ofa
as a reaction against them. This approach derives from pragmatic and syntactic
more important goal. the work of the American kind within sentence struc-
Chomsky's proposals were intended to discover Case grammar logician Richard Montague ture.
The American linguist,
the mental realities underlying the way people use
Charles Fillmore (1 929-),
(1930-70), and is based on
language: competence is seen as an aspect of our the study of logical lan- Realistic grammar
devised a theory which fo-
guages. A close correspon- Grammatical analyses
general psychological capacity. Linguistics was cusses on the semantic dence is set up between should be 'psychologically
thus envisaged as a mentalistic discipline - a view roles (or 'cases') played by
the categories of syntax according to this ap-
real',
elements of sentence
that contrasted with the behavioural bias of pre- and semantics. proach. Formal patterns
structure.
vious 20th-century work in the subject, and con- should be related to the
Relational grammar
Generalized phrase psychological factors that
nected with the aims of several earlier linguists structure grammar
This approach views gram- underlie linguistic be-
(such as the Port-Royal grammarians, p. 84). It This theory does not recog- haviour, such as compre-
matical relations (e.g. 'sub-
was also argued that linguistics should not simply ject', 'object') as central, nize the role of transforma- hension and memory.
tions in a generative gram-
limit itself to the description of competence. In the rather than the formal cate-
mar. Instead, it focusses on Network grammar
long term, there was a still more powerful target: gories (e.g. 'noun phrase',
This kind of grammar has
developing the phrase
verb phrase') of earlier
to provide a grammar capable of evaluating the structure dimension to developed out of research
generative theory (§1 6).
adequacy of different accounts of competence, and grammatical analysis into artificial intelligence,

of going beyond the study of individual languages X-bar(X) theory
The theory provides an al-
to the nature of human language as a whole (by
ternative account of phrase
discovering 'linguistic universals', (§§14, 38). In structure within a genera-
this way, it was hoped, linguistics would be able
(p. 96). aiming to simulate the way
in which people understand
Functional grammar the sentences of a lan-
Several approaches look guage.

65 LINGUISTICS • 409
 anticipated. Alternatively, attempts can be made A cautionary tale
Obtaining linguistic data to make the speaker forget about the recording,
The informant arrived and
Many procedures are available for obtaining data such as by keeping the tape recorder out of sight, we started our work. 'How do
about a language. They range from a carefully or using radio microphones. A useful technique is you say I run in your lan-
planned intensive field investigation in a foreign to introduce a topic that quickly involves the guage?' The Indian was
quiet for a while. First he
country to casual introspection about one's mother speaker, and stimulates a natural language style
looked down; then he looked
tongue carried out in an armchair at home. (e.g. asking older informants to talk about how
out. Suddenly his face lit up
times have changed in their locality). as if struck by a sudden flash
Informants An audio tape recording does not solve all the of inspiration. He spoke very

linguist's problems, however. Speech is often rapidly. If had been able to

someone has
I

In all cases, to act as a source of lan-

transcribe what he said, it
guage data — an informant. Informants are (ideally) unclear or ambiguous. Where possible, therefore,
would have spread across
native speakers of a language who provide utter- the recording has to be supplemented by the the page several times. I

ances for analysis and other kinds of information observer's notes about the non-verbal behaviour gulped and bravely started to
about the language (e.g. translations, comments of the participants, and about the context in write; but after a few syll-

general. A facial expression, for example, can dra- ables, I was already hope-
about correctness, or judgments on usage). Often, lessly bogged down.
when studying their mother tongue, linguists act matically alter the meaning of what is said (p. 402). 'How did you say that?' With
as their own
informants, judging the ambiguity, Video recordings avoid these problems to a large his repetition added two I

acceptability, or other properties of utterances extent, but even they have limitations (the camera more syllables, then bogged
can be highly intrusive, and cannot be everywhere), down again. When asked
against their own intuitions. The convenience of
I

for the third repetition, the in-

this approach makes it widely used, and it is consi- and transcriptions always benefit from any ad-
formant began to waver and
dered a primary datum in the generative approach ditional commentary provided by an observer (p. finally tochange his story,
to linguistics (p. 409). But a linguist's personal 231). and so had to give up en-
I

tirely. To my self-justifying
judgments are often uncertain, or disagree with the
and half self-accusing 'But
judgments of other linguists, at which point Elicitation
that surely doesn't all mean
recourse is needed to more objective methods of Linguists also makegreat use of structured ses- just / run', he said, 'Why of
enquiry, using non-linguists as informants. The lat- sions, in which they systematically ask their infor- course not. It means was I

ter procedure unavoidable when working on for-

is mants for utterances that describe certain actions,
sitting here with you; then I

looked out of the door and

eign languages, or in such mother-tongue fields as objects, or behaviours. With a bilingual informant,
saw a deer, so quickly I

child speech (§38) or language variation (§§8-12). or through the use of an interpreter, it is possible grabbed my spear and now
Many factors must be considered when selecting to use translation techniques ('How do you say Iam running after it.' Then,
informants - whether one is working with single table in your language?', 'What does gua mean?'). almost philosophically, he
speakers (a common situation when languages added to himself, Only a fool
A large number of points can be covered in a short would run for nothing.'
have not been described before), two people inter- time, using interview worksheets and question- (J. A. Loewen, 1 964, p. 1 89.)
acting, small groups, or large-scale samples. Age, naires. Often, the researcher wishes to obtain infor-
sex, social background, and other aspects of iden- mation about just a single variable, in which case
tity are important, as these factors are known to a restricted set of questions may
be used: a particu-
influence the kind of language used (Part n). The lar feature of pronunciation, for example, can be
A reverse lexicon An ex-
topic of the conversation and the characteristics elicited by asking the informant to say a restricted tract from the Brown Univer-
of the social setting (e.g. the level of formality) are set of words. There are also several indirect meth- sity Corpus listing words in

also highly relevant, as are the personal qualities ods of elicitation, such as asking informants to fill reverse alphabetical order.
of the informants (e.g. their fluency and consis- the blanks in a substitution frame (e.g. / see — REDEMPTION
tency). For larger studies, scrupulous attention has a car), or feeding them the wrong stimulus for cor- EXEMPTION
to be paid to the sampling theory employed. And rection ('Is it possible to say / no can see}'). GUMPTION
in all cases decisions have to be made about the RESUMPTION
best investigative techniques to use.
Corpora PRESUMPTION
CONSUMPTION
A representative sample of language, compiled for
ASSUMPTION
Recording the purpose of linguistic analysis, is known as a
OPTION
Today, data from an informant are often tape corpus. Acorpus enables the linguist to make ADOPTION
recorded. This enables the linguist's claims about objective statements about frequency of usage, and SORPTION
the language to be checked, and provides a way it provides accessible data for the use of different ABSORPTION
of making those claims more accurate ('difficult' researchers. Its range and size are variable. Some ERUPTION
pieces of speech can be listened to repeatedly). But corpora attempt to cover the language as a whole, INTERRUPTION
obtaining naturalistic, good-quality data is never taking extracts from many kinds of text; others CORRUPTION
easy. People talkabnormally when they know they are extremely selective, providing a collection of DISRUPTION
are being recorded, and sound quality can be poor. material that deals only with a particular linguistic
DESERTION
A variety of tape-recording procedures have thus feature. The size of a corpus depends on practical
INSERTION
been devised to minimize the effects of the
ASSERTION
factors, such as the time available to collect, pro-
EXERTION
'observer's paradox' (how to observe the behaviour cess, and store the data: it can take up to several
ABORTION
of people when they are not being observed). Some hours to provide an accurate transcription of a few PORTION
recordings are made without the speakers being minutes of speech (p. 23 1 Sometimes a small sam-
) .
PROPORTION-
aware of the fact — a procedure that obtains very ple of data will be enough to decide a linguistic APPORTION
natural data, though ethical objections must be hypothesis; by contrast, corpora in major research CONTORTION

410 •
XI LANGUAGE AND COMMUNICATION
 Computer corpora
A 'standard' corpus is
large collection of data
available for use by many
researchers. In English
linguistics, there are now
three standard computer
corpora, all in machine-
readable form, and thus,
principle, available any-
where in the world.
The London-Lund Cor-
pus of Spoken English
This corpus of educated
spoken British English con-
sists of the spoken material
a collected as part of the Sur- lish This corpus is drawn British English This is the on corpus studies in Eng-
vey of English Usage (see from U.S. printed sources British equivalent of the lish: the International Com-
below). The data consist of published in 1 961 It com-
. Brown corpus. It was com- puter Archive of Modern
87 texts of 5,000 words prises 500 samples of about piled by researchers in the English (ICAME), based at
each. It was transferred to 2,000 words each repre- Universities of Lancaster Bergen University. Its aims
computer tape in the 1 970s senting 1 5 main varieties of and Oslo, and prepared for are to compile an archive of
at the Survey of Spoken the language. It is available computer analysis at the English-language material
in English, University of Lund, viacomputer tape, printout, Norwegian Computing available for computer pro-
and is also partly available and microfiche. Apart from Centre for the Humanities in cessing, and to collect and
in printed form. In addition the running text, there are Bergen. Facilities are avail- distribute information on
to the running text, a lexical lexical concordances, word able similar to those pro- research that uses this
concordance has been frequency lists, and a re- vided by the Brown corpus. material.
compiled. verse alphabetical list. ICAMEThereisnowa
The Brown University The Lancaster- clearing centre for storing
Corpus of American Eng- Oslo/ Bergen Corpus of and distributing information

projects can total millions of running words. An prehension, and to carry out a variety of analytical The Survey of
important principle is that all corpora, whatever procedures. English Usage
their size, are inevitably limited in their coverage,
This survey, begun in Lon-
and always need to be supplemented by data der- Reconstruction don in 1 960 by the British
ived from the intuitions of native speakers of the The limiting case of linguistic study, one might linguist Randolph Quirk

language, through either introspection or experi- imagine, is when no data are available at all - as (1 920-), aims to describe

the grammatical repertoire of

mentation. in the case of the historical study of language where
adult educated native
written records are lacking. But it is possible to speakers of British English.
Experiments break through even this apparent barrier, by using The corpus comprises 200
Experimental techniques are widely used in linguis- the 'reconstruction' techniques of comparative phi- texts of spoken or written
material, classified as follows
tics, especially in those fields that have been lology (§50). The forms of Proto-Indo-European
(figures refer to the number
influenced by the methods of sciences where experi- and other reconstructed languages may be totally of texts of each type):
mentation is routine. Phonetics (§24) is the subject hypothetical in status, but they have nonetheless
Origin in writing (100)
most involved in this approach, but experimental become a major field of linguistic enquiry.
Printed (46)
testing is also common in several other areas, such Learned arts (6)
as child language acquisition (§38) and language Learned sciences (7)
pathology (§46). In grammar and semantics, exper- Instructional (6)

imental studies usually take the form of controlled Press: general news (4)
Press: specific reporting (4)
methods for eliciting judgments about sentences or Administrative/official (4)
the elements they contain. Informants can be asked Legal and statutory (3)
to identify errors, to rate the acceptability of sen- Persuasive writing (5)
tences, to make judgments of perception or com- Prose fiction (7)

Non-printed (36)
Continuous writing: imagina-
tive (5), informative (6)
Social letters: intimate (6),
Tagging a text equal (3), distant (4)
Many of the operations that pus with a label that indi- suprasegmental features VA + O main verb, base Non-social letters: equal
a computer can perform on cates its word class (§16). of pronunciation, §29): form (4), distant (4)
a corpus are linguistically This enables the user to dis- CD that used as Personal journals (4)
trivial, though they save an tinguish between such su- subordinator (See further, J. Svartvik & R.
As spoken (18)
enormous amount of time perficially identical items as NP proper noun Quirk, 1980, from which the
Drama (4)
(e.g. listing of words in fre- oear (animal) and bear (ac- RA personal classification of Survey of
Formal scripted oration (3)
quency of use or alphabeti- the many different
tion), or pronoun, subject EnglishUsage texts (right)
Broadcast news (3)
cal order). More interesting syntactic functions of that. RB personal has been taken.)
Talks: informative (4),
is the possibility of automati- Larger constructions (such pronoun, object
imaginative (2)
cally analysing the structure as different kinds of clause, VA + D main verb, past
Stories (2)
of the corpus, from a gram- p. 95) can also be tagged, to tense
matical, semantic, or phono- the retrieval of
facilitate VA + G main verb, -ing Origin in speech (100)
logical point of view (§26). grammatical information. form Monologue (24)
This is the aim of several Two tagged sentences VA + N main verb, past Prepared (but unscripted)
current research pro- from the London-Lund cor- participle oration (6)
grammes. pus are given (from J. Svart- VB + 5 was form of to be Spontaneous oration (10)
A first step is to provide vikera/., 1982, p. 57). Ab- *VH + contracted form Spontaneous commentary:
an automatic means of tag- breviations are as follows of have, present sport (4), non-sport (4)
ging' each word in the cor- (other symbols refer to tense
Dialogue (76)
Surreptitious: intimate (24),
0101000563 B KRA> llknew<VA -[>> th, it<CD> he<RA> was<VB+5> distant (10)
c v onunql<vA+G> Non-surreptitious: intimate
0101000564 B I' ve<RA*VH+0>
heard- :VA+N> Stan <NP>I
'Carter<NP> (20), distant (6)

m-"ention< VA + 0> b h imi RB> Telephone: intimate (10),

distant (6)

65 LINGUISTICS -411
 .

The domain of linguistics Interdisciplinary fields

The development of linguistics, the science of lan- Anthropological Unguis- the analysis of disorders of computer science, and sta-
tics spoken, written, or signed tistics (§15).
guage, has been particularly marked in recent
The study of language var- language (§46). Neurolinguistics
decades. There has been an increased popular inter- iation and use in relation to Computational linguis- The study of the neuro-
est in the role of language in relation to human the cultural patterns and logical basis of language
tics
beliefs and behaviour (§§1, 63), and an accompa- beliefs of the human race, The study language
of development and use in
nying awareness of the need for a separate acad- as investigated using the using the techniques and human beings, especially of
theories and methods of
concepts of computer the brain's control over the
emic discipline to deal adequately with the range
anthropology (§§2-5). science, especially with processes of speech and
and complexity of linguistic phenomena. The understanding (§45).
reference to the problems
university teaching of linguistics emerged during Applied linguistics posed by the fields of ma- Philosophical linguistics
the 1960s, and since then, several branches of The application of linguistic
chine translation (p. 350), The study of the role of lan-
linguistic enquiry have been established.
theories, methods, and information retrieval, and guage in the elucidation of
findings to the elucidation
The subject has now developed a clear identity of language problems that
artificial intelligence (§26). philosophical concepts,
— notwithstanding occasional uncertainty over its Educational linguistics and of the philosophical
have arisen in other do-
The application of linguistic status of linguistic theories,
name and coverage. Linguistics is the usual desig- mains. The term is especi-
methods, and observations
ally used with reference to
theories and methods to
nation, with linguistic science often used as a para- the study of the teaching (§§5,17).
the field of foreign lan-
phrase. The of phonetics (§27) is sometimes
field and learning of a language Psycholinguistics
guage learning and teach-
considered to be a separate discipline, because of ing (§62), but it applies
(especially a first language) The study of the relation-
its emphasis on the 'pre-linguistic' aspects of speech in schools and other edu- ship between linguistic be-
equally to several other
cational settings haviour and the psycho-
analysis; but it is more commonly included within fields, such as stylistics
(§§44,62). logical processes (e.g.
the coverage of linguistics, as many see it as an (§12), lexicography (§18),
translation (§57), and lan- Ethnolinguistics memory, attention) thought
indispensable foundation for language research. to underlie (§§7, 38).
guage planning (§61), as The study of language in it

Also, the label for the person who practises general well as to the clinical and relation to ethnic types and Sociollngulstics
linguistics has caused some concern: 'linguistician' educational fields below. behaviour, especially with The study of the interaction
is sometimes used, but not popular among
it is reference to the way social between language and the
Biological linguistics interaction proceeds (§9). and functioning of
structure
students of the subject, who normally refer to them-
The study of the biological society (§§10-12, 60-3).
selves as 'linguists'. There is thus occasional ambi- Geographical linguistics
conditions for language de-
The study of the regional Statistical linguistics
guity with the general use of the term 'linguist' velopment and use in hu-
distribution of languages The study of the statistical
meaning 'fluent in many languages'. man beings, with reference
and dialects, seen in rela- or quantitative properties of
Different dimensions of the subject can be dis- both to the history of lan-
tion to geographical factors language (§15).
guage in the human race
tinguished, depending on the focus and interests
and to child development
in the environment (§8).
Theolinguistics
of the Diachronic (or historical) and
linguist. (§§45,49). Mathematical linguistics The study of the language
synchronic linguistics have developed as a result The study of the mathema- used by biblical scholars,
of the distinction introduced by Saussure (p. 407): Clinical linguistics tical properties of lan- theologians, and others in-
The application of linguistic guage, using concepts volved in the theory and prac-
the former is the study of language change (§54);
theories and methods to from such fields as algebra, tice of religious belief (§63).
the latter the study of language states, regardless
of their history. When linguists try to establish
general principles for the study of all languages,
Envoi
they are said to be practising theoretical (or Linguistics has provided would include the consider- same time, by restricting
general) linguistics. When they concentrate on the conceptual framework ation of a large number of the book to a single
establishing the facts of a particular language sys- within which this encyclo- authors, including Ludwig perspective, it has been
tem, they practise descriptive linguistics. And when pedia has been written. Lit- Wittgenstein, A. J. Ayer, possible to give more nu-
tlereference has therefore and Gilbert Ryle - a major merous, detailed, and sys-
the focus is on the similarities and differences
been made to the other enterprise in its own right. tematic illustrations of the
between languages, the subject is often referred to academic traditions of lan- Similarly, many eminent use and structure of lan-
as comparative (or typological) linguistics (§§14, guage study mentioned on psychologists, such as guage than would other-
50). this page. It has occasion- A. R. Luria, L. S. Vygotsky, wise have been the case.
ally been possible to ack- and Jean Piaget, have writ- It has enabled the empha-
Linguistics shares with other sciences a concern
nowledge them; but there ten at length on language, sis to be more on the di-
to be objective, systematic, consistent, and explicit
has been no attempt to and there is a long tradition verse patterns of language
in its account of language. Like other sciences, it give a systematic account, of experimental psycho- structure and the variety of
aims to collect data, test hypotheses, devise models, as they operate within a logical enquiry into linguis- language functions, and
and construct theories. Its subject matter, however, quite different intellectual tic behaviour, which would less on the many ap-
perspective, and use radi- have to be carefully con- proaches and methods that
is unique: at one extreme, it overlaps with such
cally different procedures sidered. Issues of similar have been devised to ana-
'hard' sciences as physics and anatomy; at the of study. standing would have to be lyse these matters, and the
other, it involves such traditional 'arts' subjects as An integrated account of faced in the language various controversies
philosophy and literary criticism. The field of the history of ideas in lan- historiography of other which accompany them. A
linguistics includes both science and the humani-
guage scholarship is fields, such as anthro- comprehensive history of
beyond the scope of the pology, sociology, and linguistic thought, paying
ties, and offers a breadth of coverage that, for many
present volume. In philoso- mathematics. proper attention to the con-
aspiring students of the subject, is the primary phy alone, for example, The absence of syste- tribution of all these aca-
source of its appeal. there is a major academic matic reference to these demic issues and traditions
tradition,focussing on the other major traditions is is long overdue. But that
study of ordinary language, thus the main limitation of will have to be left for some
which to treat responsibly the present volume. At the other encyclopedia . .

412 •
XI LANGUAGE AND COMMUNICATION
Appendices
I Glossary

This glossary contains a brief definition of all the specialized language • Synonymous terms are given in bold type, preceded by the word
terms used in the text of this encyclopedia, along with some of the 'also'.

associated linguistic terminology likely to be encountered by the general

• Most entries lack exemplification, as this can be found within the
reader. The glossary excludes four types of term: body of the encyclopedia; in a few cases, where the main text does
not provide sufficient illustration, examples are given in parentheses,
(i)words in everyday use that do not raise any particular problem without the use of 'e.g.'.
of meaning (such as the names of punctuation marks); (ii) names of • At the end of each entry, there is a page reference to a section
different theories and approaches as in linguistics and language teach-
of the encyclopedia where related subject matter may be found.
ing); (hi) the very detailed terminology of grammatical description
and particular schools of thought; and (iv) background terms from Abbreviations used
related disciplines, such as anatomy, acoustics, or medicine. Names
of languages, language families, dialects, and scripts are given in Appen- acou acoustics Lat. Latin
dix vi. A selection of more specialized dictionaries of linguistic terms anat anatomy ling general linguistics
is given at the end of Appendix iv. app applied linguistics neuro neuroltngutstics
din clinical phonet phonetics
E. English phonol phonology
Glossary conventions phys physiology
esp. especially
• The alphabetical arrangement of the glossary is letter by letter. Fr. French poet poetics
• Each head-word is followed in parentheses by an abbreviated indica- gen general application prag pragmatics
tion of the main sub-field to which it belongs e.g. sem = semantics). Ger. German psycho psycholinguistics
The abbreviations used are given below. gram grammar rhet rhetoric
• Within entries, words or phrases that are themselves defined else- graph graphetics / graphology sem semantics
where in the glossary are preceded by + Superscript numerals are . hist historical linguistics semiot semiotics
used when it is important to distinguish a particular sense within It. Italian socio sociolinguistics
cross-references (e.g. grammar 1
). J- Japanese styl stylistics

abessive gram An + inflection 1

that typically expresses the meaning of 'with- adjacency pair socio) A single sequence of stimulus—utterance +
out'. 92 response— utterance by two different speakers, e.g. question + answer. 118
ablative [gram) An --inflection
1
that typically expresses such meanings as adjective (gram) A type of word identifying an attribute of a -t-noun a red
'by/with/from'. 92 chair,, in many languages showing + degree contrasts. 91
ablaut hist A -t- vowel change that gives a word a new grammatical function adjunct [gram) A less important or omissible element in a grammatical con-
drink —* drank); also, gradation. 297 struction (She ran quickly). 95
abstract see concrete adnominal [gram) Any element in a -i-noun phrase that is a -(-modification 1

accent 1 phonet Features of pronunciation that signal regional or social of the noun. 95
identity; cf. + dialect. 24 2 {phonol} A type of emphasis given to a spoken adverb (gram) A word whose main function is to specify the kind of action
word or syllable. 164 3 [graph, A mark above a letter, showing its pronun- expressed by a -(-verb (He spoke angrily); other functions include acting
ciation. 194 as +intensifier (very big) and as a -(-sentence connector (Moreover, they
acceptable ling) Said of any usage that + native speakers feel is possible in laughed .91
a language. 410 adverbial (gram) Said of -(-words, -(-phrases, or -(-clauses that function as
accidence [gram) Changes in the -t-form 2 of words signalling different gram- -(-adverbs. 95
matical functions [walking/ walked . . .); cf. + morphology. 90 aerometry phonet The measurement of air flow during speech. 139
accommodation socio Adjustments that people make to their speech, affective sem) Said of the emotional or attitudinal meaning of an utterance.
influenced by the speech of those they are talking to. 51 103
accusative gram An -inflection 1
that typically identifies the +object of a affirmative (gram) A -(-sentence or +verb that has no marker of -(-negation
+verb; 92
also, objective. He's running). 95
The branch A meaningful form that is attached to another form, to make
acoustic phonetics phonet
cal properties of speech sounds. 132
of -(-phonetics that srudies the physi- affix [gram)
a more complex -(-word tua- + kind —
ness cf. —infix, +prefix, ->-suffix.
;

acquired elm Said of any linguistic disorder that results from injury or dis- 90
ease; cf. -i-developmental. 271 affixing language ling, A language that uses -(-affixes as its main way of
acquisition see language acquisition expressing grammatical relanonships. 293
acrolect In -creole studies, the most prestigious -*- variety of a lan-
socio affricate phonet Said of a -(-consonant in which a complete -(-closure of
guage, seen in contrast with other varieties. 24 the -t-vocal tract is ([pf] Ger. pfennig). 157
gradually released
acronvm gen A word made up out of the initial letters of a phrase (laser). agent(ive) (sem) A form expressing who or what is responsible for
linguistic
90 an action T7ie man laughed, farmer 'one who farms'). 93
|

acrostic gen A poem or other text in which certain letters in each line make agglutinative agglutinating language ling A type of language in which
a word. 64 -t-words consist of lengthy strings of forms. 293
active 1 gen Said of language that a person actually uses - as opposed to agnosia din Loss of ability to interpret sensory information: auditory agno-
language that is known but not used passive knowledge;. 374 2 'phonet) sia, affecting speech sounds. 271

Said of an + articulator that moves towards an immobile, passive, articula- agrammatism din A language disorder that produces speech of a typically
tor,. 130 -t-telegrammatic quality man see ball). 271
active voice see voice agraphia see dysgraphia
acuity phonet The ability to detect and discriminate sound. 145 agreement see concord
adessive gram An -t- inflection
1
that typically expresses the meaning of 'on' air-stream mechanism phonet) An arrangement of parts of the -i-vocal tract
a place. 92 that acts as a source of energy' for speech sound production. 124

414 APPENDICES
alaryngeal (din) Said of speech without the +larynx. 276 apico- (phonet) Said of a sound using the tip (or -f-apex) of the tongue, e.g.

alexia see dyslexia 'apico-dental'. 155

alienable (gram) Applied to relationships where a possessed item is seen as apocope (hist) The omission of a final -(-syllable, sound, or letter in a word.
having a temporary or non-essential dependence on a possessor (the man's 328
car); cf. inalienable, where the dependence is permanent or necessary (the apostrophe (rhet ) A + figurative expression in which an idea, inanimate object,
man's brain). 93 or absent person is addressed as if present.70
allative (gram) An -(-inflection that typically expresses the meaning of 'to'
1
appellative (sem) A personal name used as an everyday word (a sandwich).
a place. 92 112
alliteration (poet) A sequence of words beginning with the same sound, espe- applied linguistics (ling) The application of the theories, methods, or findings
cially as used in poetry. 74 of -(-linguistics to the solution of practical problems. 412
alio- (ling) A variation in the +form 2 of a linguistic unit that does not alter apposition (gram) A series of + nouns or +noun phrases with the same mean-
its basic identity, e.g. allophones (variants of a -(-phoneme), allomorphs (var- ing and grammatical status (Mr Jones, the baker). 95
iants of a -(-morpheme), allographs (variants of a +grapheme). 90, 160, appropriate (ling) Said of any use of language considered to be compatible
194 with a given social situation; cf. -(-correctness. 2
allograph see allo- approximant (phonet) A -(-consonant in which the organs of -(-articulation
allomorph see allo- approach each other, but without -(-closure or audible friction ([1], [j]); also,
allonym (gen) A name an author assumes that belongs to someone else; cf. frictionless continuant. 157
-(-pseudonym. 112 approximative system see interlanguage
allophone see allo- apraxia (din) Loss of ability to carry out voluntary muscular movements
alphabet (gen) A writing system in which a set of symbols ('letters') represents for the production of speech; also, dyspraxia. 271
the +phonemes of a language; cf. -(-dual alphabet. 202 aprosody see dysprosody
alphabetism (gram) A word made of initial letters, each being pronounced aptitude (app) A person's natural ability to learn a language; evaluated using
(VIP). 90 an aptitude test; also, prognostic test. 371
2
alternation (ling) The relationship between the different -(-forms of a linguis- aptronym (gen) A name that fits a person's nature or occupation (Mr Clever,
tic unit, usually symbolized by 90 ~ (cat ~ cats). Mr Smith). 112
alveolar (phonet) Said of a consonant in which the tongue makes contact
-I- arbitrariness (ling) The absence of any physical correspondence between
with the bony prominence behind the upper teeth ([t], [n]). 155 linguistic signals and the entities to which they refer; cf. -(-iconic. 397
ambilingual (gen) Someone who can speak two languages with equal facility; archaism (gen) An old word or phrase no longer in general spoken or written
also, balanced bilingual. 362 use. 330
amelioration (hist) A change of meaning in which a word loses an originally area (ling) A geographical region identified on the basis of its linguistic charac-
unpleasant reference; cf. -(-deterioration. 330 teristics. 33
amplitude (acou) The intensity of a sound. 134 areal linguistics see geographical linguistics
anacoluthon (gram, rhet) An unexpected break in a -(-sentence (John might argot (gen) Special vocabulary used by a secretive social group, e.g. gypsies.
— Are you listening?). 52 58
anacusis (din) Total deafness. 266 article (gram) A word that specifies whether a +noun is -(-definite or indefinite
anagram (gen) A word or phrase formed by changing the order of letters (the/ a). 91
in another word or phrase. 65 articulation (phonet) The physiological movements involved in modifying a
analects (gen) A selection of passages taken from an author. 66 flow of air to produce speech sounds. 130
analogy (ling) A change that affects a language when -(-regular forms begin articulator (phonet) A + vocal organ involved in the production of a speech
to influence less regular forms. 234, 330 sound. 130
analytic 1 (gram) see isolating 2 (sem) Said of sentences expressing a -(-tauto- articulatory phonetics (phonet) The branch of -(-phonetics that studies the
logy (Bachelors are unmarried); contrasts with synthetic. 107 way speech sounds are produced by the -t-vocal organs. 124
ananym (gen) A name that has been written backwards. 112 artificial language (gen) 1 An invented language used to facilitate international
anap(a)est (poet) A unit of + metre consisting of two light beats followed communication; also, auxiliary language. 352 2 An invented language used
by a heavy beat. 74 in computer programming, e.g. basic. 351

anaphora (gram) A feature of grammatical structure referring back to some- artificial larynx (din) A portable device that provides a source of vibration
thing already expressed; the -(-pronoun in When Mary saw John, she waved for speech, for people who have no +larynx. 276
is 'anaphoric'; cf. -t-cataphora, +exophoric. 119 artificial speech (phonet) The output of a -(-speech synthesizer. 149
anarthria see dysarthria ascender (graph) A part of a letter that extends above the height of the letter
animate (gram) Said of words (esp. +nouns) that refer to living things, and x. 190
not to objects or concepts (inanimates). 91 aspect (gram) The duration or type of temporal activity denoted by a +verb,
4
anomia (din) A -(-language disorder in which the primary symptom is diffi- e.g. completion or non-completion of an action; cf. -I- perfective. 93
culty in remembering the names of things. 271 aspiration (phonet) Audible breath that may accompany the -(-articulation
antecedent (gram) A part of a +sentence to which some other part grammati- of a sound (E. pen [p h en]). 161
cally refers (This is the cat that chased the rat). 119 assimilation (phonol) The influence exercised by one sound upon the + articu-
anthropological linguistics (ling) The study of (esp. non-western) languages lation of another, so that the sounds become more alike. 164
or cultural patterns and beliefs. 412
in relation to social associative meaning (sem) The sense associations that are not part of a word's
anthroponomastics (sem) The study of personal names. 112 basic meaning (birthday—* presents, party, etc.). 103
anthropophonics (phonet) The study of the human potential for vocal sound. assonance (poet) The repeated use of -l-vowels to achieve a special effect.
18 74
anticipatory see regressive asterisked form 1 (ling) A usage that is not + acceptable or not -(-grammatical
2
antonym (sem) A word that is opposite in meaning to another word (good- (*do have gone). 88 2 (hist) A form in linguistic history for which there
/bad, single/ married). 105 is no written evidence (Indo-European *penk w e 'five'). 292

aorist (gram) A form of the + verb in some -(-inflecting languages, esp. referring asyndeton (rhet) The omission of -(-conjunctions to achieve an economical
to an action without any particular completion, duration, or repetition. 93 form of expression (They ran with haste, with fear). 91
aperiodic see periodic atelic see telic
apex (phonet) The tongue. 131
tip of the attested (ling) Said of linguistic forms where there is evidence of present or
4
aphasia (din) A +language disorder resulting from brain damage, which past usage. 292
affects a person's ability to produce or understand -(-grammatical and attribute 1 (phonet) An identifiable feature of sound sensation, e.g. +pitch,
-fsemantic structure; also, dysphasia. 270 -(-loudness. 144 2 (sem) A defining property of the meaning of a word
aphasiology (din) The study of -faphasia. 270 (round is an attribute of ball). 107
aphesis The loss of an +unstressed -fvowel from
(hist) the beginning of a attributive (gram) Said of -(-adjectives or other forms that are + modifiers
word ('mongst). 328 of a +noun within the +noun phrase (the big table); contrasts with predica-
aphonia see dysphonia tive uses (The table isbig). 95

aphorism (gen) A succinct statement expressing a general truth (More haste, audiogram (din) A graph used to record a person's ability to hear +pure
less speed). 53
tones. 266

1 GLOSSARY- 415
audiolingual (app) Said of a language-teaching method based on the use of breaking see voice mutation
drills and dialogues for speaking and listening; also, aural-oral. 374 breath group {phonet) A stretch of utterance produced within a single breath
audiology (din) The study of hearing and hearing disorders, esp. their diagno- expiration. 124
sis, assessment, and treatment. 266 breathy (phonet) A +voice quality that involves the use of audible breath.
audiometer {din) An electronic instrument that measures the sensitivity of 128
hearing. 266 broad (phonet) Said of a + transcription of speech that shows only the major
auditory agnosia see agnosia -(-phonetic contrasts; cf. -(-narrow, -(-phonemic transcription. 158

auditory discrimination {pbonet) The process of distinguishing between (esp. Broca's area (neuro) An area of the brain that controls the expression of
speech] sounds. 145 spoken language; cf. -(-Wernicke's area. 260
auditory phonetics (phonet) A branch of +phonetics that studies the way buccal (phonet) Applied to sounds made in or near the + cavity of the cheek.
people perceive sound. 142 127.
aural-oral see audiolingual
automatic translation see machine translation
autonomous speech see idioglossia cacography (gen) Bad handwriting or spelling. 274
autosegmental (phonol) An approach to +phonology that includes the study cacology (gen) Unacceptable pronunciation or use of language. 2
of features of sound that extend beyond individual -(-segments. 161 cacophony (gen) Unpleasant, harsh sounds, esp. of speech. 2
auxiliary language 1 (sodo) A language adopted by different speech communi- caesura (poet) A break in the -(-rhythm of a line of poetry. 74
ties for purposes of communication. 352 2 (gen) see artificial language calligraphy (gen) The an of beautiful handwriting. 188
auxiliary verb (gram) A -(-verb used along with a -i-lexical verb to make gram- caique (hist) A + borrowed item in which the pans are translated separately
matical distinctions (She is going/ might go). 91 into the new language (E. superman from Ger. Ubermensch); also, loan trans-
lation. 330
baby talk (gen) 1 A simplified speech style used by adults to children. 235 2 cant (gen) The special speech of a group with low social standing, e.g. gypsies,
An immature form of speech used by children. 244 thieves. 58
back (phonet) Said of sounds made in the back part of the mouth ([h]) or cardinal number (gram) The basic form of a numeral (one, etc.); cf. -(-ordinal.

with the back pan of the tongue ([k], [o]). 131 99

back-formation (hist) A process of -1-word formation where a new word is cardinal vowels (phonet) A set of reference points, based on auditory and
formed bv removing an imagined -t-affix from another word (editor—* edit). aniculatory criteria, used to identify +vowels. 154
330 caretaker speech (psycho) The speech of adults when they talk to children;
back slang (gen) A secret language in which words are said backwards. 59 also, motherese. 236
balanced bilingual see ambilingual case (gram) In an -Hnflecting language, the form of a +noun, -(-adjective,
basal readers (app) The first textbooks used in a graded reading programme. or -Hpronoun, showing its grammatical relationship to other words. 93
251 catachresis see malapropism
base (ling) A component of a + transformational grammar, in which the basic catalect (gen) Any part of an author's literary work seen as separate from
sentence patterns of a language are -(-generated. 97 the rest. 66
basilect (sodo) In -fcreole studies, a language +variety furthest away from cataphora (gram) A feature of grammatical structure that refers forward to
the one that carries most prestige (the +acrolect). 24 another unit; (in John said this:, the +pronoun is 'cataphoric'); cf. -(-anaphora
behaviourism (gen) The study of observable and measurable behaviour (here, and -Hexophoric. 119
of the linguistic stimuli and responses made by participants in speech situ- catenation (ling) The linking together of a series of linguistic forms, e.g. sounds
ations). 408 or words. 95
bel (acou) Unit for the measurement of acoustic intensity; cf. -(-decibel. 134 catenative (gram) A -(-lexical verb that governs another lexical verb (try to
bidialectal (sodo) Applied to someone who is proficient in the use of two run). 91
+ dialects. 24 causative (gram) A linguistic element that expresses the notion of 'cause' (the
bidialectism (socio) An
educational policy that recommends the teaching of causative verb kill = 'cause to die'). 93
a non-standard +dialect along with a -(-standard one. 26 cavity (phonet) An anatomically defined chamber in the + vocal tract, e.g.
bilabial (phonet) Said of a -(-consonant made with both lips ([b], [m]). 155 oral, nasal. 124
bilingual (gen) Said of an individual or a community that regularly uses two central see centre
languages; cf. -(-ambilingual. 360 centre (phonet) The top part of the tongue, between -(-front and -t-back;
binary (ling) Said of any linguistic analysis that sets up an opposition between involved in central sounds. 131
two alternatives. 79 centum language (hist) An Indo-European language that kept the sound [k]
binary feature (phonol) Any -(-phonetic variable that enables sounds to be in such words as centum ('hundred'); cf. -t-satem language. 328
channel (gen) A medium selected for communication (e.g. speech, writing).
1
classified into two mutually-exclusive possibilities, e.g. -(-voice ('voiced' vs
'voiceless'). 162 48
binaural (phonet) Using both ears. 142 character (graph) A graphic sign used in a writing system, esp. one that is
biolinguistics (ling) The study of the biological preconditions for language not part of an -(-alphabet. 200
development and use in human beings, both as individuals and as a race; chereme (ling) The smallest xmtrastive unit in a +sign language. 221
also, biological linguistics. 412 cherology (ling) The study of +sign language. 221
bisyllable (phonet) A word with two + syllables. 164 chest pulse (phonet) A contraction of the chest muscles that forces air into
blade (phonet) The pan of the tongue between the +apex and the -l-centre; the +vocal tract. 164
also, lamina. 131 chiasmus (rhet) A balanced pattern in which the main elements are reversed.
blend (gram) The result of two elements fusing to form a new word or con- 70
struction (breakfast + lunch = brunch); cf. -(-coinage. 90 chirography (graph) The study of handwriting forms and styles. 186
block (din) In -(-stuttering, an obstruction experienced by the speaker that chrestomathy (gen) An anthology of passages usually used for learning a lan-
prevents the production of speech. 278 guage. 374
body language (semiot) Communication using body movement and appear- chroneme (phonol) An abstract unit that accounts for differences in the
3
ance, as opposed to speaking, writing, or -(-sign 399 . + consonants. 408
-(-duration of speech sounds, e.g. long vs short
body size (graph) The size of a piece of type. 190 chronogram (gen) A which letters that are also Roman
phrase or sentence in
borrow (hist) To introduce a word (or some other linguistic feature) from numerals (e.g. C, X) combine to form a date. 64
one language or + dialect into another; vocabulary borrowings are usually chunking (psycho) Dividing an utterance into parts, e.g. to make it easier
known as loan words. 330 to remember. 171
bound form (gram) A -(-morpheme that cannot occur on its own as a -(-word cipher (gen) A secret -(-code in which letters are transposed or substituted.
1

(E. de-, -tiori). 90 58

boustrophedon (graph) Writing in which lines run in alternate directions. circumlocution (gen) The use of more words than is necessary to express
185 a meaning. 2
brachygraphy (graph) Shorthand writing. 206 class see word class
bracketing (ling) A way of showing the internal structure of a string of ele- classifier (gram) A -(-morpheme which indicates that a word belongs to a
ments ((The girl) (ate) (a cake)). 97 particular -(-semantic class, e.g. animates, large objects. 91

416 •
APPENDICES
clause {gram) A structural unit smaller than the + sentence but larger than comparative linguistics (ling) A branch of -(-linguistics that relates the charac-
-fphrases or +words;
+dependent, +main clause. 95
cf. teristics of different languages or -(-varieties. 84
clavicular breathing (din) A way of breathing, in which inhalation comes comparative method (hist) A technique that compares forms taken from + cog-
from using the neck muscles to raise the collar bones. 125 nate languages to see if they are historically related. 292
cleft palate (din) A congenital fissure in the middle of the +palate, often comparative philology (hist) The study of the historical relationship between
found along with a split in the upper lip (cleft lip, also 'hare lip') and teeth languages. 292
ridge. 277 compensation (phonet) An alternative -(-articulation that counteracts the
cleft sentence (gram) A + clause has been split
sentence in which a single effect of some abnormality in the + vocal organs. 18
1
into two its own +verb (It was Mary who arrived). 95
sections, each with competence (ling) Unconscious knowledge of the system of + grammatical
cliche (gen) An expression which has become so overused that it no longer -(-rules in a language; cf. -(-communicative competence, -(-performance. 409
1

conveys much meaning, and is criticized (a fate worse than death). 2 complement (gram) A -(-clause element that completes what is said about
click (phonet) A sound produced using the +velaric + air-stream mechanism some other element, such as the -(-subject (That book looks nice). 95
(E. [j] 'tut'). 126 complementarity (sem) A type of oppositeness of meaning; two words are
clinical linguistics (ling) The application of linguistics to the analysis of dis- complementaries if to assert one denies the other (single /married). 105
orders of spoken, written, or -l-sign language. 412 complementary distribution (phonol) A property of sounds that cannot appear
cupping (gram) A process of +word formation in which a new word is pro- 1
in the same -(-phonetic -(-environment (E. [t* ] and [t]). 161
1

duced by shortening (examination —> exam); also, reduction. 90 complex sentence (gram) A +sentence consisting of more than one -(-clause
clitic (gram) A form that resembles a -(-word but that cannot stand on its (esp. including a +dependent clause). 95
own as a normal utterance because it is structurally dependent on a neighbour- complex tone (acou) A sound wave consisting of two or more -(-pure tones.
ing word (Fr. je). 91 133
close (phonet) Said of a + vowel made with the tongue in the highest position component 1 (ling) The major sections of a -(-generative grammar. 82 2 (sem)
possible without causing audible friction (e.g. [i], [u]); vowels a degree lower A basic feature of word meaning (girl = human, female, etc.). 107
3
are half/ mid-close; cf. -fopen . 153 componential analysis (sem) The analysis of vocabulary into a finite set of
2
closed 1 Said of any +word class whose membership is limited to
(gram) basic elements (-(-components ). 107
a small number of items, e.g. +pronouns, -(-conjunctions; cf. -t-open 1 91 2 . compound 1 (ling) Said of a linguistic unit composed of elements that can
(phonol) Said of a +syllable ending in a -Hconsonant; cf. +open 2 164 . function separately elsewhere, e.g. a compound + word/ -(-sentence. 90 2
closure (phonet) A contact made between + vocal organs in order to produce (socio) Said of +bilinguals who are thought to have a single meaning system
2
a speech sound. 157 underlying their use of words in both languages; cf. -(-coordinate 362 .

cloze procedure (app) A

technique used in the teaching and testing of reading, comprehension (gen) The ability to understand and interpret language; cf.
in which readers guess words omitted at intervals from a text. 377 -(-production. 261
cluster (phonol) A series of adjacent + consonants occurring at the beginning compressed speech (phonet) Speech that has been acoustically altered so that
or end of a -(-syllable (stray, books). 164 it uses a smaller range of -(-frequencies than normal. 138

cluttering (din) A -(-speech disorder in which utterances are produced in an computational linguistics (ling) The application of the concepts and techni-
excessively rapid and unrhythmical way. 278 ques of computer science to the analysis of language. 412
coalescence (hist) The fusing of originally distinct linguistic units. 328 computer language see language 2
coarticulation (phonet) An -(-articulation involving the simultaneous or over- concatenation see catenation
lapping use of more than one point in the +vocal tract ([pk], [bd]). 156 concord (gram) A + grammatical 1 relationship in which the +form 2 of one
cochlea (anat) The part of the inner ear that contains the organ of hearing. element requires the corresponding form of another (She eats). 95
143 concordance (gen) An ordered list of the words used in a particular text or
code 1 (gen) Any system of signals used for sending messages, often in secret -(-corpus. 411
form. 58 2 (sodo) A language, or -I- variety of language. 48 concrete (gram) Said of -Hnouns that refer to physical entities (book, train);
1
code switching (socio) Changing from the use of one language or +variety contrasts with abstract. 91 2 (phonol) Said of analyses that emphasize the
to another; also, language mixing. 363 +phonetic reality of speech sounds; contrasts with abstract. 163
codify (app) To
provide a systematic account of a language (esp. its -(-gram- conditional (gram) 1 Said of a + clause that expresses a hypothesis or condition
1
mar and vocabulary). 364 (Kit rains, you'll get wet). 95 2 Said of a -(-verb form that expresses hypothe-
cognate (hist) A language or linguistic form that is historically derived from tical meaning (Fr. 'conditional tense' je marcherais 'I would walk'). 93
the same source as another, e.g. Spanish and French are 'cognate languages', conditioning (ling) The influence of linguistic -(-context on a +form 2 (E. 1

both deriving from Latin. 292 a—> an when followed by a +vowel). 164
cognitive meaning see denotation conductive (din) Said of a hearing loss where sound fails to reach the -(-coch-
coherence (ling) The underlying logical connectedness of a use of language. lea. 266
119 conjugation (gram) The set of + verbs that occur in the same forms in an
cohesion (ling) The -(-formal linkage between the elements of a -t-discourse
1
-(-inflecting language. 293
or +text (the -Hpronoun is 'cohesive' in The man left. He.. .). 119 conjunction (gram) A word that connects words or other constructions (cat
coinage (gen) The creation of a new word out of existing elements (postper- and dog). 91
son);d. -I- blend. 90 connective/ connector (gram) An item whose function is to link linguistic units,
collective noun (gram) A -(-noun that denotes a group of entities (army, gov- e.g. -l-conjunctions, certain +adverbs (however). 91
ernment). 91 connotation (sem) The personal associations aroused by words; cf. -(-deno-
collocation (sem) The habitual co-occurrence (or mutual selection) of + lexical tation. 103
items. 105 consonance (poet) The repetition of sounds in the same position in a sequence
coloratura (gen) A soprano singer with a high vocal range. 18 of words. 74
comitative (gram) An -(-inflection that typicallv expresses the meaning 'with'.
1
consonant (phonol) A speech sound that functions at the -(-margins of -(-syll-
92 ables, produced when the + vocal tract is either blocked or so restricted
command (gen) A type of + sentence in which someone is told to do (or that there is audible friction ([k], [s], etc.); cf. -(-vowel, + semi-vowel. 155
not do) something. 121 constative (ling) An utterance that is a descriptive statement, analysable into
comment Part of a -(-sentence that says something further about the
(ling) truth values (The table is red); cf. -(-performative. 121
sentence -(-topic (The car was in the garage); also, new information. 94 constituent (gram) A linguistic unit that is a component of a larger construc-
comment clause (gram) A -(-clause that adds a parenthetic remark to another tion. 96
clause (The answer, you see, is complicated). 52 constituent analysis (gram) A process of analysing a construction into its ma jor
common noun (gram) A +noun that refers to a class of objects or concepts components (immediate component being further ana-
constituents), each
(chair, beauty); cf. -Hproper noun. 91 lysed until a set of irreducible elements is left (ultimate constituents). 96
communicative approach (app) An approach to language teaching that constriction (phonet) A narrowing in the + vocal tract, in order to produce
focusses on language -(-functions 2 and -(-communicative competence, and a speech sound. 157
not on -(-grammatical structure. 374 1
contact (socio) Said of languages or -(-dialects in close geographical or social
communicative competence (ling) A person's awareness of the + rules 1 govern- proximity, which thus influence each other. 33
ing the -(-appropriate use of language in social situations. 48 content word (gram) A type of word that has an independent, 'dictionary'
comparative see degree meaning (chair, run); cf. -(-function word. 91

1 GLOSSARY 417 •
context (ling) 1 environment of an element. 82 2 The non-
The linguistic (cryptograms). 58
linguistic situation inwhich language is used. 100 cryptograms see cryptanalysis, cryptography
continuant (phonet) A speech sound made with an incomplete -(-closure of cryptography (gen) The preparation of secret messages (cryptograms), using
1
the -(-vocal tract. 157 -l-codes and -(-ciphers. 58
1
continuous see progressive cryptology (gen) The study of -I- cryptography and -(-cryptanalysis. 58
contoid {phonet) A -(-consonant defined solely in -(-phonetic terms. 153 cryptophasia see idioglossia
contour (phonol) 1 A distinctive sequence of -t-prosodic features (esp. cued speech [din) A method of -(-speech-reading in which manual cues help
1
-i-tones ). 169 2 Said of a -l-tone language that uses -(-gliding tones. 172 to distinguish sounds. 225
1 {gram) A shortened linguistic -(-form attached to an adjacent
2
contraction cuneiform (graph) An ancient writing system that used wedge-shaped char-
form [I'm), or a + fusion of forms (Fr. de le —* du). 164 2 {poet) The acters. 198
-(-elision of -(-syllables to keep a line's -(-metre regular. 74 cursive (gen) A form of handwriting in which separate characters in a sequence
contradictory see complementarity have been joined. 186
contrary see antonym cycle (acou) A single complete vibration, forming part of a -(-sound wave.
contrast {ling) Any -(-formal
1
difference that serves to distinguish meanings 133
in a language; contrastive differences are also known as distinctive, functional,
significant. 160
contrastive see contrast
contrastive analysis {app) The identification of structural differences between dactyl (poet) A unit of rhythm in poetic -(-metre, consisting of one heavy-
languages, seen as points of potential learning difficulty. 372 beat followed by two light beats. 74
contrastive stress {phonol} Extra emphasis given to a word, in order to draw dactylology (din) Signing in which each letter of the alphabet is given its
attention to its meaning {John bought a red car). 169 own sign; also, finger spelling. 225
convention gen) The tacit agreement of speakers to use the same -(-rules 1
dative (gram) An -(-inflection
1
that typically expresses an -(-indirect object
in order to communicate. 404 relationship (Lat. Dedi epistolam puellae i gave the letter to the girl'). 93
conventionalism [sem) The view that there is a relationship of + arbitrariness daughter language see parent language
between words and things; also, nominalism; cf. + naturalism. 404 decibel phonet) A unit for measuring the relative -(-intensity of sounds, esp.
convergence {socio) A process of linguistic change in which + dialects or + ac- in the assessment of hearing loss. 134
1
cents become more like each other; contrasts with divergence. 51 decipher (gen) To work out the meaning of a message in -(-code 1 (esp. in
conversational implicature (prag) An implication deduced from an utterance, -(-cipher). 58
using the + cooperative principles that govern the efficiency of conversations declarative (gram) A grammatical construction used in expressing a -(-state-
(A bus! = 'We must run'). 117 ment (The dog barked). 121
conversational maxims prag, General principles thought to underlie the effi- declension (gram) A set of -(-nouns, -(-adjectives, or -(-pronouns that show
1
cient use of language, e.g. speakers should be relevant and clear. 117 the same -(-inflections (decline). 93
conversation analysis ling A method of studying the structure of conversa- decline see declension
tions using the techniques of -t-ethnomethodology. 116 decode (gen) 1 To use the brain to interpret an incoming linguistic signal.

converseness {sem) A type of oppositeness of meaning, such that one word 262 2 To convert a secret message into intelligible language. 58
presupposes the other {buy/sell). 105 deconstruction (styl) An approach to literary theory that aims to show the
conversion [gram) A type of -(-word formation in which an item changes contradiction in -(-structuralist principles of textual analysis. 79
its -(-word class without the addition of an -(-affix {smell = verb/noun). decreolization 'socio) Change in a -(-Creole that makes it more like the -(-stan-
90 dard language of an area. 336
cooperative principle {prag) A
agreement between speakers to follow
tacit deep grammar structure (ling) An underlying level of grammatical organiza-
the same when communicating. 117
set of -(-conventions ('maxims') tion that specifies how sentences should be interpreted; cf + surface grammar-
.

coordinate 1 {gram) Said of -(-clauses displaying -(-coordination. 95 2 {socio) /structure. 98

Said of +bilinguals who are thought to have different meanings for the corres- defective (gram) Applied to words that do not follow all the rules of the
1
ponding words in their two languages. 362 class to which they belong (E. +auxiliary verbs, which lack the usual verb
coordination {gram) The linking of linguistic units that have the same gram- -(-inflections ). 91
1
2 (graph) A writing system consisting only of -(-consonant
matical status, e.g. two -(-noun phrases the cat and the dog,. 95 symbols. 202
coordinator gram) A -(-conjunction used in -(-coordination 'and, but). 95 defining see restrictive
coprolalia [din] Uncontrolled use of obscene language. 264 defining vocabulary (app) A core set of words used to define other words.
copula (gram) A +verb whose main role is to link other elements of the Ill
-(-clause (It is ready). 95 definite [gram, sem) Said of a specific, identifiable entity or class of entities
coreference 'sem) The use of elements that can be interpreted only by referring (thecar); contrasts with indefinite a car:. 91
to another element in a text. 119 degree (gram) A contrast of comparison in -(-adverbs or -(-adjectives: usually
coronal {phonet) Said of sounds where the -(-blade of the tongue is raised identified as positive big,, comparative bigger), and superlative (biggest).
to the hard -(-palate. 155 92
corpus {ling) A collection of language data brought together for linguistic deixis [ling] Features of language that refer directly to the personal, temporal,
analysis. 411 or locational characteristics of the situation deictic forms) [you, now, here).
correctness gen) An absolute standard of language use deriving from the 106
rules of institutions (e.g. language academies) or respected publications (e.g. deletion (ling) Omitting an element of sentence structure (that in / said he
dictionaries); cf. -(-appropriate. 2 was ready). 97
correlative {gram) Said of constructions using a pair of connecting words demonstrative (gram) Applied to forms whose function is to distinguish one
(either/ or). 95 item from other members of the same class this/ that). 99
countability see countable denasal 1 phonet Said of sounds whose -(-nasality has been reduced or
countable (gram) Said of + nouns denoting separable entities, as shown by removed. 130 2 (din) Said of a -Hvoice qualitv with poor nasal -(-resonance.
their use with such forms as a (dog, chair); count(able) nouns contrast with 276
uncountable non-count -(-mass) nouns. 91 denotation (sem) The objective ('dictionary') relationship between a word
creaky phonet) A -(-voice quality produced by very slow vibration of the and the reality to which it refers; also, cognitive referential meaning; cf.
-^ vocal folds. 128 -(-connotation. 100
creativity (ling) A characteristic of language that enables speakers to produce dental (phonet) Said of a -(-consonant made by the +apex and rims of the
and understand an indefinitely large number of sentences. 397 tongue against the teeth. 155
Creole (socio) A -(-pidgin that has become the mother tongue of a speech dependent 'gram, Said of any element whose -(-form 2 or -(-function 1 is deter-
community (through a process of creolization). 336 mined by another part of the sentence (in the red car, the -Harticleand -^-adjec-
critical period psycho) A period of time in child development during which tive depend on the -(-noun); also, subordinate. 95
language is thought to be most easily learned. 263 derivation 1 (gram) A major process of -(-word formation, esp. using -(-affixes
cross-sectional gen) Said of studies that sample the language of a group of to produce new words act —» action 90 2 (gram) The 1
cf. -(-inflection ; .

individuals at a single point in rime; cf. +longitudinal. 229 set of analytical steps required to -(-generate a sentence. 97 3 (hist) The
cryptanalysis (gen The process of + deciphering or + decoding secret messages
) origins or historical development of a language or form. 292

418 •
APPENDICES
descender {graph) A part of a letter that extends below the depth of the letter dissimilation (phonol) The influence sound segments have on each other, so
x. 190 that they become less alike. 328
description (ling) An objective and systematic account of the patterns and dissonance (gen) The use of sounds to convey unpleasant effects. 74
use of a language or +variety; cf. +prescription. 2 distinctive (phonol) Said of a feature capable of making a difference of mean-
deterioration [hist) A change of meaning in which a word acquires a negative ing between otherwise identical forms, e.g. -l-vocal fold vibration; cf. -(-con-

evaluation; also, pejoration; cf. + amelioration. 330 trast. 160

determinative (graph) A
+ logogram that indicates its + semantic
part of a distribution (ling) The total set of linguistic + environments 1
in which an item
content; also, radical; +phonetic. 199
cf. can occur. 161
determiner (gram) An item that co-occurs with a -f-noun to express such mean- disyllable (phonol) A word of two -(-syllables. 164
ings as number or quantity (the, some, each). 96 ditransitive (gram) Said of + verbs that take two +objects (give, show). 95
determinism see linguistic relativity divergence see convergence
developmental (din) Said of any linguistic disorder that arises out of an abnor- dominant language (socio) 1 The most important language in a + multilingual
mal process of development in the child, e.g. 'developmental +aphasia'; cf. speech community. 360 2 The language a + bilingual knows best. 362
+acquired. 2~1 dorsal (phonet) Said of sounds made with the -(-back ('dorsum') of the tongue
developmental (psycho)linguistics (ling) The study of the acquisition of lan- ([k],[g]).131
guage in children. 226 doublet (gen) A type of word game in which a series of single-letter substitu-
deviance (ling) Failure to conform to the -(-rules of the language. 88 1
tions links pairs of words. 65
devoiced (phonet) Said of a sound in which the normal amount of -l-vocal downdrift (phonol) A gradual lowering of -(-tones throughout an utterance
1
fold vibration (-l-voice has been reduced. 163
) in a +tone language. 172
diachronic see historical, synchronic drift (hist) A gradual series of related changes in the historical development
diacritic (graph) A mark added to a symbol to alter its value, e.g. an + accent3 . of a language. 328
154 dual (gram) A -(-grammatical 1 contrast of -fnumber in some languages, refer-
diadochokinesis (din) The ability to carry out rapid repetitive movements ring to 'two of. 92
of the -l-vocal organs. 271 dual alphabet (graph) The use of capital and small letters in a single system.
diagnostic test (app) A test to show what a language learner knows and does 186
not know. 377 dualism (sem) A theory that postulates a direct, two-way relationship between
diagramming see parsing linguistic forms and the entities to which they refer. 100
dialect (ling) A language -( variety in which the use of grammar and vocabulary duality of structure (ling) The structural organization of language into two
identifies the regional or social background of the user; cf. + accent 24 1
. abstract -I- levels meaningful units (e.g. words) and meaningless segments
1
:

dialect continuum (socio) A chain of dialects whose end-points are not (sounds, 397letters).
mutually intelligible. 25 duration (phonet) The length of time involved in the + articulation of a sound
dialectology (ling) The study of (esp. regional) dialects; also, dialect geogra- or -(-syllable. 169
phy. 26 dynamic 1 (gram) Type of -fverb that expresses activities and changes of
The study of
dialinguistics (ling) the range of -(-dialects and languages in a state,
1
allowing such forms as the -l-progressive (He's running); cf. -t-stative.
speech community. 26 93 2 (socio) Said of language analyses that take account of temporal change.
dichotic listening (psycho) A technique for determining which half of the brain 328
is primarily involved in processing auditory effects. 259 dyne (acou) A unit of measurement for sound pressure. 134
diction (gen) The effective choice of words, esp. the vocabulary used by a dysarthria (din) A motor speech disorder that leaves someone unable to articu-
poet or other writer. 73 late speech sounds; in severe form, also, anarthria. 271
diglossia (socio) The use of two + varieties of a language throughout a -(-speech dysfluency (din) The loss of ability to control the smooth flow of -(-speech
community, each with a distinct set of social functions. 43 production, resulting in hesitancy, poor -(-rhythm, -(-stuttering, etc. 278
digraph (graph) 1 A -(-graphic unit in which two symbols have combined dysgraphia (din) A + language 4 disorder that primarily affects the ability to
to function as one (encyclopaedia). 365 2 Any sequence of two letters pro- write; also, agraphia. 272
nounced as a single sound (ship, wood). 365 dyslalia (din) A disorder of -(-articulation that has no clear phvsical cause.
dimeter (poet) A line of verse containing two units of rhythm ( + foot). 277
74 dyslexia (din) A + language 4 disorder that affects the ability to read; also,
diminutive (gram) An -(-affix with the general meaning of 'little' (It. -ino). alexia, word blindness. 272
90 dysnomia see anomia
diphthong (phonet) A +vowel in which there is a perceptible change in quality dysphasia see aphasia
during a +syllable (time, road); cf. + monophthong, -(-triphthong. 154 dysphemism (rhet) A use of language that emphasizes unpleasantness (a horr-
diplomatics (graph) The study of legal and administrative documents. ible dirty day); cf. -(-euphemism. 61
187 dysphonia (din) The loss of ability to use the -l-vocal folds to produce normal
An utterance intended to get other people to do (or not do) form, aphonia. 276
1
directive (prag) -(-voice ; in severe
something (Sit down); also, command. 121 dyspraxia see apraxia
direct method (app) A method of language teaching that emphasizes speaking dysprosody (din) The loss of ability to produce speech with a normal + intona-
2 1
in the + target language and avoids the conscious learning of -(-grammar . tion. 276
374 dysrhythmia (din) The loss of ability to produce normal +rhythm in -fspeech
direct object (gram) A
element immediately affected by the action
-(-clause production. 278
of the H-verb (She hit him); contrasts with a less directly affected (indirect)
object (I gave John a letter). 95
direct speech (gen) The actual utterance spoken by a person; cf. -(-indirect ear training (phonet) A technique in + phonetics to train the ability to identify
(or reported) speech. 77 speech sounds. 158
discontinuous (gram) The splitting of a grammatical construction by the
inser- echolalia (din) The automatic repetition of all or part of what someone has
tion of another unit (switch the light on). 95 said. 271
1
discourse (ling) A continuous stretch of (esp. spoken) language larger than economy (ling) The use of the smallest possible number of -Hrules and sym-
a -(-sentence. 116 bols in carrying out a linguistic analysis. 163
discourse analysis (ling) The study of patterns of linguistic organization in educational linguistics (ling) The application of -(-linguistics to language teach-
+discourses. 116 ing and learning in schools and other educational settings. 248
discovery procedure (ling) A set of techniques automatically applicable to egocentric speech (psycho) Speech that does not take account of the needs
2
a sample of language to produce a correct -(-grammatical analysis. 408 of the listener. 235
discrete (ling) Said of linguistic elements that have clearly defined boundaries. egressive (phonet) Said of sounds produced using an outwards-moving -(-air-

397 stream mechanism. 125

disjunction (sem) alternative or contrastive relationship between elements
An ejective (phonet) A +consonant produced using the -t-glottalic -(-air-stream
we're early or the bus is late). 107
in a sentence (Either mechanism. 126
displacement (semiot) The ability of language to refer to contexts removed elaborated code (socio) A relatively formal, educated language use involving
from the speaker's immediate situation (/ was angry yesterday). 397 a wide range of linguistic structures; cf. -(-restricted code. 40

I GLOSSARY •
419
 1 »

dative (gram) An +inflection

1
that typically expresses the meaning 'out of less of its antecedents or subsequent history. 407

a place. 92 • ethnography of communication (socio) The study of language in relation to

electroaerometer [phortet) An instrument that records air flow during speech. the social and cultural variables that influence human interaction. 48
139 ethnolinguistics (ling) The study of language in relation to ethnic groups and
electrokymograph (pbonei) An instrument that records the changes in the behaviour. 412
air flow from mouth and nose during speech. 139 ethnomethodology (socio) The detailed study of the techniques used during
electrolaryngograph phonet, An instrument that records -t-vocal fold vib- linguistic interaction. 116
ration. 141 etic(phonet) The analysis of the physical patterns of speech without reference
electromyograph (phonet) An instrument that records muscular contractions to their function within the language; cf. +emic. 408
during speech. 139 etymological fallacy (hist) The view that an earlier (or the oldest) meaning
electropalatograph (phonet) An instrument that makes a continuous record of a word is 330
the correct one.
1
of the contacts between tongue and +palate during speech. 140 etymology The study of the origins and history of the -t-form and
(hist)

elicit (ling) To obtain utterances or linguistic judgments from + informants. meaning of words. 330
410 etymon (hist) The -(-form 1 from which a later form derives (Lat. mater —
elision (phonol, poet) The omission of sounds in connected speech (bacon Fr. mere). 330
V eggs). 164 euphemism (gen) The use of a vague or indirect expression in place of one
ellipsis (gram, rhet) The omission of pan of a sentence (e.g. for economy, that is unpleasant or offensive (pass away for die). 61
emphasis), where the missing element is understood from the -t-context 1 (A: euphony (gen) A pleasing sequence of sounds. 74
Where's the book? B: On the table). 94 exclamation (gram) An emotional expression marked by strong +intonation
elocution gen) The art of speech training to produce effective public speaking. in speech or by an exclamation point in writing (Good grief'.); cf. -(-command,
70 + question, -(-statement. 121
embedding (gram) Inserting one +grammatical unit within another (The man
1
exclusive (gram) Said of a first- + person +pronoun (we) that does not include
who left was my uncle). 95 the person being addressed; cf. + inclusive. 92
emic (phonol) An to speech analysis that sets up a system of abstract
approach excrescent (ling) Said of a sound added to a word to make the pronunciation
+contrastive units, esp. +phonemes; cf. +etic. 408 easier. 328
emotive meaning (sem) The emotional content of a use of language. 10 exegesis (gen) An interpretation of a text, esp. of a biblical kind. 385
empty word (gram) A meaningless word that expresses a grammatical rela- existential (gram) A sentence emphasizing the idea of existence (There is a
tionship (It's today he goes); also, prop word; cf. -l-content word. 91 book on the table). 95
encipher (gen) To write a message using a -(-cipher. 58 exocentric see endocentric
enclitic (gram) An + unstressed form attached to a preceding word (cannot). exophoric (gram) Said of a linguistic unit that refers directly to the -t-extral-
91 inguistic situation (there, him); +endophoric. 119 cf.

encode (gen, To give linguistic shape to a meaning, as part of communication. expansion 1 (gram) The process of adding new elements to a construction,
262 2 To convert a message from one system of signals into another (esp. without its basic structure being affected. 95 2 (psycho) An adult response
2
for secrecy); cf. -(-decode . 58 to a child which adds grammatical elements that the child has omitted. 23
1
endocentric (gram) Said of a construction where there is a -(-grammatical experimental phonetics (phonet) The use of instrumentation and experimental
-t-head (the tall men); cf. -t-exocentnc, which lacks a head (People left early). techniques to investigate the properties of speech sounds; also, instrumental
95 phonetics. 138
endophoric (gram) Said of the relationships of +cohesion that help to define expletive (gen) An exclamatory word or phrase, usually obscene or profane.
the structure of a -t-text; cf. -t-exophonc. 119 61
enjamb(e)ment (poet) The running on of a sentence between two couplets expression {ling) 1 Any string of elements treated as a unit for analysis, e.g.
of verse without pause. 74 a -(-sentence, -t-idiom. 95 2 All aspects of linguistic -t-form 1 (as opposed
environment 1 (ling) The parts of an -(-utterance or -(-text that are adjacent to meaning). 82
to an item of language. 163 2 (socio) The social or cultural situation in expressive 1 (gen) Said of a use of language that displays or affects a person's
which a particular use of language takes place. 48 emotions. 10 2 (din) Said of disorders of language + production, e.g.
epenthesis (phonol) The insertion of an extra (epenthetic) sound in the middle 'expressive aphasia'. 265
of a word. 328 extension 1 (sem) The class of entities to which a word is correctly applied,
epicene (gram) A -t-noun that can refer to either sex without changing its e.g. the extension of flower is rose, daffodil, etc.; cf. -Hntension. 107 2
form (teacher). 47 (hist) Widening the meaning of a word. 330
epiglottis (anat) A structure that closes over the -(-larynx during swallowing. extralinguistic (ling) Said of anything (other than language) to which language
124 can relate. 82
epigram (gen) A short, witty statement, in verse or prose. 53 extraposition (gram) Moving an element to a position at one end of a -(-sen-
epigraph (gen) 1 An inscription on stone, buildings, coins, etc. 187 2 A tence (Working here is nice —» It's nice working here). 95
phrase or quotation above a section in a book or on the title page. 53 extrinsic (anat) Said of sets of muscles that control the gross movements of
epigraphy (gen) The study of inscriptions, esp. their interpretation in ancient certain -t-vocal organs, e.g. tongue, -(-larynx. 131
times. 187 eye dialect (gen) A way of spelling words that suggests a regional or social
epithet (gen) Any item that characterizes a -t-noun and is regularly associated way of talking (Thankee kotndly, zur). 180
with it (Ethelred the Unready). 105 eye rhyme (poet) A pair of words that seem to rhyme from the spelling, but
eponym (gen) The name of a person after whom something, e.g. a place, have different pronunciations (come/ home). 74
a book title, is named (Washington, Hamlet). 112

equative (gram) Applied to a -(-clause which relates two elements that are false friends (app) Words in different languages that resemble each other in
identical in meaning (Mr Jones is a butcher). 95 but express dissimilar meanings (Fr. demander = 'request', not
1
-(-form ,

ergative (gram) Applied to a construction in some languages where the 'demand'); also, faux amis, false cognates. 347
--object of a +transitive verb and the -(-subject of an -(-intransitive one are false vocal folds see ventricular folds
in the same +case. 93 family (hist) A set of languages that derive from a common ancestor -(-parent) (

error 1 (neuro) An
inaccuracy in the spontaneous use of language attributable language, and are represented as a family tree. 292
to a malfunctioning of the neuromuscular commands from the brain. 262 Any typical, noticeable, or -t-contrastive property of a -l-level 1
2 feature (ling)
(app) A language learner's systematic use of a linguistic item that does not of language. 82
2
conform to the -l-rules of the -i-target language, because knowledge of these
1
feedback 1 (prag) The ongoing reaction speakers receive from their listeners,
rules is incomplete; contrasts with unsystematic, -(-performance faults (mis- which helps them to evaluate the efficiency of their communication. 118 2
takes). 372 (phonet) The information speakers obtain by monitoring their own speech
error analysis (app) The systematic interpretation of the unacceptable forms activity. 262
used by someone learning a language. 372 felicity conditions (prag) The criteria that must be satisfied if a + speech act
esophageal see oesophageal is to achieve its purpose. 121

essive (gram) An -i-inflection that typically expresses the meaning 'at' a place.
1
feminine see gender
92 festination (din) Abnormal increase of speed while speaking. 278
etat de langue (ling) The 'state of a language' seen at a particular time, regard- field see semantic field

420 • APPENDICES
fieldwork (ling) The principles and procedures of obtaining linguistic data causing any change of meaning. 161
from + informants, esp. in their home environment. 410 frequency (acou) The number of -(-sound waves per second produced by a
figurative (gen) Said of an expressive use of language when words are used source of vibration. 133
in a non-literal way to suggest illuminating comparisons and resemblances fricative (phonet) Said of a + consonant made when two + vocal organs come
(figures of speech). 70 so close together that the air moving between them produces audible friction
filled pause (ling) A vocal hesitation (erm). 172 ([{], [z]); also, spirant. 157
filter (acou) A device used to separate the +frequency components of a frictionlesscontinuant see approximant
+ sound wave. 133 front (phonet) Said of sounds made in the front part of the mouth or by
filtered speech (phonet) Speech passed through + filters to alter its acoustic the front part + blade) of the tongue ([i], [t]); cf. +back, -(-centre. 131
(

characteristics. 133 fronting 1 (phonol) +Articulation of a sound further forward in the mouth
finger spelling see dactylology than is normal. 155 2 (gram) Moving a -(-constituent from the middle or
finite (gram) A form of a +verb that can occur on its own in a +main clause end of a +sentence to the front (Smith his name was). 95
and permits variations in -Htense +number, and +mood (They ran, He
1
, full verb see lexical verb
is running); contrasts with non-finite. 93 full word see content word
finite-state grammar (ling) A simple kind of -(-generative device that is able function (ling) 1 The relationship between a linguistic form and the other
to process only a very limited range of sentences. 97 elements of the system in which it is used, e.g. a +noun as -(-subject or
first language (gen) The language first acquired as a child (mother tongue, +object of a -(-clause. 95 2 The role language plays in communication
native language), or preferred in a + multilingual situation. 368 (e.g. to express ideas, attitudes) or in particular social situations (e.g. religious,
first person see person legal). 10
'fis' phenomenon (psycho) A child's refusal to accept an adult imitation of functional 1 (ling) Said of linguistic approaches that treat the notion of -(-func-
what it has just said. 240 tion as central, esp. to show -(-grammar communicating meaning in social
fixation (graph) A period of relative stability between rapid eye movements. interaction.409 2 (din) Said of a linguistic disorder that has no apparent
208 physical cause. 264 3 (ling) see contrastive
flap (phonet) A + consonant produced by a single rapid contact between two functional change 1 (hist) An alteration in the role of a linguistic feature
organs of articulation, e.g. the tongue tip movement [r] in very. 157 in historical development. 328 2 (gram) The use of a +word in different
flexion see inflection grammatical roles (round the corner, a round table). 92
fluency (gen) Smooth, rapid, effortless use of language; cf. +dysfluency. 278 functional load/ yield (phonol) The use a language makes of a + contrast in
flyting (poet) An exchange of curses or personal abuse in verse form. 60 -(-phonology (E./p/~/b/ has a 'high' functional load, often distinguishing
focal area (socio) A region where + dialect forms are relatively homogeneous pairs of words). 161
and tend to influence the forms used in adjoining areas. 28 function word (gram) A word whose role is largely or wholly to express
focus (gram) An element in a sentence to which the speaker wishes to draw a grammatical relationship (to, a); also form/grammatical/structural word.
special attention (It was John who wrote to me). 107 91
folk etymology (hist) Altering an unfamiliar word to make it more familiar functor see function word
(asparagus -» sparrow-grass); also, popular etymology. 330 fundamental frequency (acou) The lowest- + frequency component in a com-
foot (phonol, poet) A basic unit of -(-rhythm, esp. used in describing poetic plex -(-sound wave, of particular importance in determining a sound's -(-pitch.
-(-metre. 74 133
foregrounding (poet) Any departure from a linguistic or socially accepted fusion (ling) The merging of distinct linguistic elements (Lat. -us -t-noun ending
norm, esp. in literary language. 71 simultaneously signals -(-number, +gender, and +case). 293
foreign language (app) A
non-native language, esp. one that has no official fusional language see inflecting language
status in a country; cf. + second
language. 342 futhork (graph) The runic alphabet; also futhark. 203
forensic linguistics (ling) The use of linguistic techniques to investigate crimes
in which language data constitute part of the evidence. 69
form 1 (ling) The outward appearance or structure of language, as opposed geminate (phonol) A sequence of identical adjacent sounds in one + morpheme
to its function, meaning, or social use (formal vs -(-notional ). 82 2 (gram) 1
(It. notte 'night'). 160
gender (gram) A way of grouping words into different + formal
1
The variations in which a linguistic unit can appear (the 'forms' walk, walks, classes, using
etc.). 91 such labels as masculine, feminine, neuter, animate. 93
1
formal 1 see form 2 see formality genealogical see genetic classification
formalist Said of an approach that studies the structural (-(-formal 1 )
(styl) generalization (psycho) A process in -(-language acquisition in which a first
basis for literary effects in great detail. 78 use of a linguistic feature is extended to a class of items (+plural -s on
formality (socio) A scale of language use, relating to situations that are socially E. regular +nouns); cf. +overgeneralization. 234
careful or correct (formal) or otherwise (informal). 40 generate see generative
2
formal universal (ling) An obligatory feature of -(-grammar2 construction;
1 1
generative (ling) Said of a -(-grammar that uses a set of -(-formal -(-rules
1
cf. + substantive universal. 85 to define the membership of (generate) the infinite set of -(-grammatical
formant (acou) A concentration of acoustic energy, esp. distinctive in -(-vow- + sentences in a language. 97
1
els and -(-voiced sounds. 135 generic (gram) A word or sentence that refers to a class of entities (the Chinese,
formative (gram) An irreducible grammatical element that enters into the con- the rich). 91
struction of larger linguistic units. 90 genetic classification (hist) The grouping of languages into -(-families based
form class (gram) A set of items that display similar or identical grammatical on their historical relationships; also genealogical classification. 293
features. 91 genitive (gram) An -(-inflection
1
that expresses such meanings as possession
formulaic (ling) Said of a sentence that does not permit the usual range of and origin (girl's bag, man's story); also applied to related structures (the
grammatical variation (Many happy returns); also, fossilized or stereotyped cover of the book). 93
sentences, or routines. 52 genre (gen) An identifiable category of artistic (here, literary) composition,
form word see function word e.g. the novel. 52
fortis (phonet) Said of -(-consonants made with relatively strong muscular geographical linguistics (ling) The study of languages and -t-dialects in terms
effort and breath force ([f], [p]); cf. -t-lenis. 157 of their regional distribution; also, area! linguistics. 33
fossilized (ling) Said of any construction that lacks -(-productivity, e.g. gerund (gram) A +noun derived from a -(-verb (a 'verbal noun'), esp. as found
+idioms (spick and span), -(-formulaic utterances (So be it!). 52 in Latin grammar, or in grammars based on Latin (amandum 'loving'). 91
frame (gram) A specific -(-structural 2 -(-context within which a class of items gerundive (gram) An -(-adjective derived from a +verb (a 'verbal adjective'),
can be used. 95 esp. asfound in Latin grammar, or in grammars based on Latin (amandus
free form (gram) A minimal grammatical unit that can be used as a -Hword 'lovable').91
without additional elements; also known as a free -(-morpheme; cf. -(-bound ghost form (hist) A word originating in an error during the copying, analysing,
form. 90 or learning of a language, which does not exist in the original language.
free translation (gen) A -(-translation expressing the meaning rather than the 187
1
-l-form of the -(-source language; contrasts with literal (word-for-word) given see topic
translation. 344 glide (phonet) 1 A transitional sound made as the + vocal organs move tow-
free variation (phonol) The substitution of one sound for another without ards (on-glide) or away from (off-glide) an -(-articulation ([j] in puny [pJu:ni]).

I GLOSSARY •
421
 137 2 A + vowel 1
where there is an audible change of quality- (-(-diphthong, word or phrase; occurs as consonant harmony and vowel harmony. 161
-(-triphthong). 154 3 A -ftone 1 involving a change of +pitch level. 172 head 'gram) The main element in a -(-phrase on which other elements depend,
3
gliding see glide and which controls the function of the phrase as a whole (All the new books
global aphasia din) Aphasia involving a severe disorder of +production
-t- from the library are on the table 95 .

and +comprehension. 271 headword (app) The item that occurs at the beginning of a -(-dictionary entry.
glossal 'anat) Pertaining to the tongue. 131 108
glossary gen, An alphabetical list of the terms used in a special field. Ill hertz (acou) The unit for measuring sound vibration; once known as 'cycles
glossogenetics (ling) The studv of the origins and development of language. per second'. 133
290 heterographs 'gen) Words that have the same meaning or pronunciation, but
glossograph (phonet) An instrument that records the movements of the tongue differ in spelling (bear\s bare). 106
during speech. 140 heteronyms (gen) Words that differ in meaning but are identical in either
glossolalia [gen) Speaking in tongues, as practised bv certain religious groups. pronunciation or spelling (threw vs through). 106
11 heterophemy (gen) An unintentional error in spoken or written language.
glottal (phonet) Said of sounds made in the + larynx resulting from the + clo- 164
sure or narrowing of the -fglottis, e.g. +whisper, +creaky. 155 heterotopy (gen) A misplaced sound during (esp. fast) speech. 278
glottalic (phonet) Said of the + air-stream mechanism that uses the + glottis hexameter (poet) A line of verse containing six units of rhythm
as the source of vibration for -fejective and -t-implosive sounds. 126 (+toot). 74
glottalization (phonet) An + articulation involving a simultaneous + glottal hiatus 1 (phonet) The use of adjacent 1
-(-vowels in different -(-syllables. 164 2
-f-constriction. 126 (gen) A break in a -(-sentence that leaves it incomplete. 52
glottal stop (phonet) The audible release of a -(-closure at the -(-glottis; (bottle hierarchy {ling) A classification of linguistic units into a series of successively
as [bo'l]). 128 subordinate +levels 3 esp. an analysis of -(-sentences into -(-clauses, -(-phrases,
,

glottis (anat) The aperture between the -t-vocal folds. 128 -(-words, and -l-morphemes. 82, 95
glottochronology \hist) An approach to language history- in which a statistical hieroglyphic (graph) A writing system using mainly pictorial symbols; esp.
technique (-t-lexicostatistics) is used to quantify how far languages have applied to Egyptian. 199
diverged from a common source. 331 high phonet, phonol) 1 Said of -(-vowels (and sometimes -(-consonants) made
1

glottograph (phonet) An instrument that monitors the extent of + glottal open- by raising the tongue towards the roof of the mouth ([i], [k]). 153 2 Said
ing, using a light source. 141 of -(-tones 1 that use a relatively high level of -(-pitch range. 172 3 (socio)
goal (sent) The entity affected by the action of a -(-verb (The man kicked Said of the more prestigious -(-variety in -t-diglossia. 43
the ball); also, patient, recipient. 95 hiragana see kana
govern see government The study of development of language and lan-
historical linguistics (ling)
government (gram) A type of grammatical linkage in which one word requires guages over time; also, diachronic linguistics or (with different emphasis)
a specific -f-form 2 of another (Lat. ad 'governs' the -(-accusative form: ad comparative philology. 407
Roman 'to Rome'). 95 hold (phonet) To maintain a single position of the -(-vocal organs for a period
gradable gram, sem) Said of a word (esp. an -(-adjective) that can be compared of time. 157
or intensified (big—* very big, bigger). 105 holograph 'gen) A document that is entirely written in the handwriting of
gradation see ablaut its author. 187
gradience gram A pattern of gradually increasing irregularity at the bound- holophrase (psycho) A grammatically unstructured utterance, usually consist-
ary of a + word class. 92 ing of a single word, typical of the earliest stage of language learning in
grammar (ling) 1 The study of -(-sentence structure, esp. with reference to children (dada, allgone). 242
-(-syntax and -(-morphology, often presented as a textbook or manual. 88 2 homographs 'gen) Words with the same spelling but different meanings (wind
A systematic account of the -(-rules governing language in general, or specific = 'air' vs wind = 'turn'). 106
languages, including -(-semantics, -(-phonology, and often -(-pragmatics. 82 homonyms (gen) Words with the same -(-form
1
but different meanings (ear
grammatical (gram) 1 Pertaining to -(-grammar 88 2 Said of constructions 1
.
= 'corn' vs ear = 'body 106
part').
that conform to the -(-rules 1 of a -(-grammar2 those that do not are ungram-
; homophones (gen) Words with the same pronunciation but different meanings
matical. 88 (rode/rowed). 106
grammatical word see function word homorganic (phonet) Said of sounds made at the same place of -(-articulation
graph graph, The smallest + discrete segment in a stretch of writing or print ([p],[b],[m]). 157
(t,T,r,etc). 194 honorific (socio) A use of language (esp. of grammar 1
) to express levels of
grapheme (graph) The smallest -t-contrasrive unit in the writing system of politeness or respect. 99
a language (t, e, „ ?). 194 hybrid gram) A -i-wordcomposed of elements from different languages tele-
graphemics 'graph) The study of -(-graphemes. 194 vision, from Greek and 90
Latin).
graphetics (graph) The study of the visual properties of written or printed hydronymy (gen) The study of the names of rivers, lakes, etc. 1 14
language. 185 hyperacusis (din) An extremely acute sense of hearing. 142
graphic see graph hyperbole (gen, rhet) Emphatic exaggeration (There were millions of people
graphology (graph) 1 The (study of the) writing system of a language. 194 2 in the cinema). 70
The analysis of handwriting to discover the writer's character. 189 hypercorrection (socio) A linguistic -(-form that goes beyond the norm of 1

groove (phonet) A type of -(-fricative -(-consonant produced when the tongue a -(-target
2
+ variety,
because of the speaker's desire to be correct; also, hyper-
is slightly hollowed along its central line ([s],
[J], etc.). 157 urbanism, overcorrection. 2
group see phrase hypercreolization socio) The development of a kind of +creole that is a reac-
tion away from the -(-standard language. 336
habitual (gram) Said of a form (esp. a -(-verb or -(-adverb) expressing repetition hypernasality (din) Excessive -(-nasal -(-resonance in speech. 276
of an action (often). 93 hyper-urbanism see hypercorrection
half-close -open see close, open hyp o acusis (din) An impairment of auditory acuity. 266
hapax legomenon [styl) A word that occurs only once in a text, author, or hypocoristic (gen) A pet name Bill, honey). 1 12
language. 67 hyponasalify (din) Lack of normal -t- nasal -(-resonance in speech. 276
haplography (graph) An omission made in a sequence of identical letters hyponymy (sem) The relationship between specific and general words, where
(occurrence —* ocurrence). 213 the former is included in the latter (cat is a hyponym of animal). 105
haplology (phonol) The omission of sounds in a sequence of similar + articula- hypostatize (gen) To speak of an abstract quality as if it were human. 70
tions (probably—* /probli/). 328 hypotaxis (gram) The linking of a -(-dependent (hypotactic) clause to another
hard contact (din) A very -(-tense2 -(-articulation heard in -(-stuttering when part of the sentence using -(-conjunctions (The boy left when the bell rang);
the speaker attempts a difficult word. 278 cf. -(-parataxis. 95
hard palate see palate hysterical (din)
1
Said of disorders of -(-voice or hearing that are psychological
hare lip see cleft palate in origin. 276
harmonic acou, A -(-frequency of sound vibration that is a multiple of the
-(-fundamental frequency; also, overtone. 133 iamb (poet) A unit of -(-metre consisting of an unstressed -(-syllable followed
harmony (phonol) Similarity of -(-articulation between sounds in the same by a -(-stressed syllable ('To be/ or not/ to be/'). 74

422 APPENDICES
iconic {sent) Said of signals whose physical form corresponds to features of infinitive (gram) A -(-non-finite form of the -t-verb, which in many languages
the entities to which they refer (as in +onomatopoeia, e.g. cuckoo). 220 acts as the basic form donner 'to give'). 93
(E. run, Fr.
ictus (poet) The + stressed + syllable in a unit of + metre. 74 infix (gram) An + affix added within a +root'. 90
idealization {ling) The ignoring of certain kinds of variability in linguistic inflecting/inflected/inflectional language (ling) A language in which -(-words
data, in order to reach general conclusions. 409 1
express grammatical relationships by using -(-inflections ; also, synthetic/fu-
ideation (psycho) The cognitive process of forming ideas and relationships sional language. 293
of meaning, prior to their formulation in language. 262 inflection/inflexion (gram) An + affix that signals a grammatical relationship,
ideational function (ling) The use of language to refer to the people, events, e.g. -t-case, -Htense
1
(girl's, walked). 90 2 (phonet) Change in voice +pitch
etc. in the world; cf. + interpersonal, -(-textual functions. 10 during speech. 169
ideogram (graph) A symbol used in a writing system to represent a whole informal see formality
word or concept; also, ideograph. 198 informant (ling) Someone who acts as a source of data for linguistic analysis.
ideograph see ideogram 410
idioglossia (gen) An invented form of speech whose meaning is known only information (ling) The way a message content is structured, e.g. into given
to the inventor, e.g. the language sometimes used by twins; also, autonomous and new. 120
speech, cryptophasia. 247 ingressive (phonet) Said of all sounds produced with an inwards-moving air
idiolect (ling) The linguistic system of an individual speaker. 24 stream. 125
idiom (sent) A sequence of words that is a unit of meaning (kick the bucket inhalation see inspiration
= 'die'). 105 initiator ((phonet) The -I- vocal organs that are the source of air movement,
illative (gram) An -(-inflection
1
that typically expresses the meaning of 'into' e.g. lungs. 124
92
a place. innateness hypothesis (psycho) The view that a child is born with a biological
ill formed (gram) Said of any +ungrammatical 1 sentence; cf. +well formed. predisposition to learn language, involving a knowledge of its -(-universal
88 structural principles; also, nativism. 234
illocutionary act (prag) A + speech act involving a + performative verb (bap- inner ear (anat) The part of the ear containing the +cochlea. 143
promise, request, etc.); cf. +locutionary/+perlocutionary act. 121
tize, inspiration (phys) The act of drawing air into the lungs; also, inhalation.
imagery 1 (gen) The use of -(-metaphor, -(-simile, and other +figurative lan- 124
guage, esp. in a literary context. 70 2 (psycho) Language that produces institutional linguistics (ling) The study of the problems involved in -(-language
clear or vivid mental pictures. 103 planning. 364
imitation (psycho) The copying of linguistic behaviour, esp. while learning instructive (gram) An -(-inflection
1
that typically expresses the meaning 'by'.
a language; cf. -(-comprehension, -(-production. 234 92
immediate constituent see constituent instrumental (gram) An + inflection 1
that typically expresses the meaning 'by
immersion (app) Said of a -(-bilingual programme in which + monolingual means of. 292
children attend a school where another language is the medium of instruction. instrumental phonetics see experimental phonetics
367 intensifier (gram) A word or phrase that adds force or emphasis (very good,
imperative (gram) A grammatical +mood expressing a -(-command (Look!). awfully pretty). 91
93 intension (sem) The set of defining properties that determines how a term
imperfect (gram) A -t-tense form expressing such meanings as past duration
1
is to be used (table—* 'legs', 'flat surface', etc.). 107
and continuity (Lat. amabam i was loving/used to love'). 93 intensity (acou) The power transmitted along a -(-sound wave. 134
imperfective see perfective interchangeability (semiot) The ability of a signalling system to be mutually
impersonal (gram) Said of construction or -(-verbs with an unspecified -I- agent transmitted and received by members of the same species. 396
(It's raining). 95 interdental (phonet) A -(-consonant made by the -l-apex of the tongue between
implicational universal (ling) A type of -(-universal statement that takes the the teeth ([G], [8]). 155
form 'If a language has X, then it also has Y.' 85 interference see transfer
implicature see conversational implicature interjection (gram) A class of -l-words with -(-emotive meaning, which do
implosive (phonet) A -(-consonant made using the + glottalic +air-stream not form grammatical relationships with other classes (Gosh!, Yuk!). 91
mechanism with inwards-flowing air ([6], [cf]). 126 interlanguage (app) The language system used at an intermediate stage of
inalienable see alienable foreign language learning. 372
inanimate see animate intermediate vowel (phonet) A -l-vowel that falls between two adjacent -(-car-
1

incapsulating language see polysynthetic language dinal vowels. 154

inceptive (gram) Said of a -(-verb form that specifies the beginning of an action internal evidence (hist) Linguistic features in a text that indicate when the
('be about to'), e.g. Lat. -escere; also, inchoative. 92 work was written. 187
inchoative see inceptive internalrhyme (poet) The rhyming of words within lines of verse. 74
inclusive (gram) Said of a first- + person -(-pronoun that refers to both the interpersonal function (ling) The use of language to establish and maintain
speaker and someone else, as when we means 'me and you'; cf. -(-exclusive. social relations; cf. -(-ideational function, -(-textual function. 10
92 interpret (gen) To make an oral -(-translation
1
. 349
incompatibility (sem) A feature of mutually-defining items where the choice interrogative (gram) A type of + sentence or -(-verb form used in the expression
of one excludes the use of the others (The ink is red/ blue). 105 of -(-questions (Who is hef, Are they there?); cf. -(-declarative. 95
incorporating language see polysynthetic language interrogative word (gram) A word used at the beginning of a -(-clause to
indefinite see definite mark it as a -l-question (
Who is here?). 95
indefinite vowel see schwa intervocalic (phonet) A -(-consonant used between two + vowels 1
(/p/in
independent clause see main clause apart). 164
indexical 1 (ling ) Said of features of speech or writing (esp. -(-voice quality) intonation (phonol) The -Hcontrastive use of -(-pitch in speech. 169
that reveal the personal characteristics of the user, e.g. age, sex. 171 2 intonation contour see tone unit
(sem) see deixis intransitive (gram) Said of a -(-verb or -(-sentence that cannot take a -(-direct
indicative (gram) A grammatical + mood that expresses objective statements object (she's going); cf. -(-transitive. 95
(My car is new). 93 intraoral pressure (phonet) The build-up of air inside the mouth needed to
indirect object see direct object produce certain speech sounds, e.g. +plosives. 124
indirect question (gram) A -l-question as expressed in -(-indirect speech (He intrinsic (anat) Said of sets of muscles that control the fine movements of
asked if she was in). 77 certain +vocal organs, e.g. tongue, +larynx. 131
indirect speech (gram) A construction in which the speaker's words are made intrusion (phonet) The use of sounds in connected speech that do not appear
-(-subordinate to a -(-verb of 'saying' (She replied that she had); also, reported when words or +syllables are heard in isolation, e.g. 'intrusive r' between
the
1
speech; cf. -(-direct speech. 77 -(-vowels (as in law(r) and order). 164
indirect speech act (prag) An utterance whose linguistic -(-form
1
does not intuition (gen) A person's instinctive knowledge of language, which
directly reflect its communicative purpose (using It's cold in here to mean decides whether sentences are acceptable and how they can be interrelated
-I-

'Close the window'). 121 410

inessive (gram) An -(-inflection
1
that typically expresses location or position invariable word (gram) A word that does not undergo any change in structure
within a place. 92 (under, but); cf. + variable word. 91

I GLOSSARY •
423
 .

inversion (gram) A reversed sequence of elements (He is going — » /s he going?) result of brain damage. 270
243 language loyalty (socio) The personal attachment to a language that leads
irony (gen) Language that expresses a meaning other than that literally con- to its continued use in a country where other languages are + dominant.
veyed by the words (That's marvellous, said of poor work). 70 367
language maintenance (socio) The continued use of and support for a language
1
irregular (gen) Said of a lingustic -(-form that is an exception to a pattern
404 in a -(-bilingual or -I- multilingual community. 360
1
stated in a rule .

isochrony/isochronism (phonet) A rhythmic pattern in which -(-stressed -(-syll- language pathologist/pathology see speech pathologist/ pathology
ables fall at roughly regular intervals throughout an utterance. 169 language planning (socio) Official intentions and policies affecting language
isogloss (socio) A line on a map showing the boundary of an area in which use in a country. 364
a linguistic feature is used; the lines mark such features as vocabulary (isolex), language shift (socio) A permanent change in a person's choice of language
-(-morphology (isomorph), -(-phonology (isophone), -(-semantics (isoseme), for everyday purposes (esp. as a result of immigrant movement). 360
or socio-cultural use isopleth 28 .
language therapist see speech pathologist
isolating language (ling) A language in which -(-words are -(-invariable and langue (ling) The language system shared by a -(-speech community; cf. +par-
grammatical relations are shown mainly by -(-word order, e.g. Chinese; also, ole. 407
analytic/ root language. 293 laryngeal (phonet) A speech sound made in the -(-larynx. 128
iterative (gram) A 4-form 1 that expresses the repetition of an action (fre- laryngectomee (din) Someone who has had a + laryngectomy. 276
quently), esp. as part of the -(-aspect system. 93 laryngectomy (din) The surgical removal of some or all of the -(-larynx. 276
laryngology (din) The study of the anatomy, physiology, and diseases of the
jargon 1 (gen) The technical language of a special field. 56 2 (gen) The +larynx. 128
obscure use of specialized language. 379 3 (psycho) Unintelligible utterance laryngopharynx (anat) The lower part of the +pharynx, between -(-larynx
with good -(-intonation, used by young children when learning to talk. 237 4 and -(-oropharynx. 130
(din) Unintelligible speech in some -(-language disorders. 271 laryngoscope (din) A device inserted into the mouth to enable the -(-larynx
juncture (phonot) -(-Phonetic boundary features that demarcate units of gram- to be seen. 129
mar, e.g. certain features of -(-pitch, -(-duration, pause. 164 larynx (anat) The part of the -(-trachea containing the +vocal folds. 128
lateral (phonet) Said of a -(-consonant in which air escapes around one or
2
kana (graph) Either of the two Japanese -I- syllabic writing systems, hiragana both sides of a -(-closure made in the mouth, as in the various kinds of
and katakana. 195 /sound. 157
katakana see kana lateralization/laterality (neuro) The primary involvement of one hemisphere
kernel (gram) A basic type of -Hsentence structure, as used in early -(-generative of the brain in the exercise of a bodily function, e.g. language. 258
Applied to any -(-grammar that is based on the terms and
1
grammar. 97 latinate (gram)
kin(a)esthesis (phys) Awareness of the movements and positions of the + vocal categories used in classical Latin grammar. 2
organs during speech; also, kin(a)esthetic feedback. 124 law (hist) A
statement of the predictable relationships (esp. in the use of
kineme (semiot) The smallest -t-contrastive unit of body expression. 402 sounds) between different languages or states of a language. 328
kinesics (semiot)The systematic use of facial expression and bodily gestures- lax see tension
/movements to communicate meaning. 399 leading (graph) The spacing between lines of type. 190
koine (socio) The spoken language of a locality that has become a -(-standard lect (socio) A collection of linguistic phenomena that has a functional identity
language. 43 within a speech community, e.g. a regional or social +variety. 24
kymograph (phonet) An early device for recording information about + vocal length see duration
organ movements. 138 lenis (phonet) Said of -(-consonants made with a relatively weak degree of
muscular effort and breath force ([b], [v]); cf. +fortis. 157
labial (phonet) The active use of one or both lips in the -(-articulation of lenition (phonet) A relaxation of muscular effort during -(-articulation. 157
asound([f],[u]). 155 lento (phonet) Said of speech produced slowly or with careful +articulation.
labialization (phonet) + Rounding the lips while making a speech sound. 156 169
labio-dental (phonet) Said of a + consonant in which one lip actively contacts lesion (din) An abnormal change in body tissue due to injury or disease.
the teeth ([f], [v]). 155 259
labio- velar (phonet) A speech sound made at the + velum with simultaneous level 1 (ling) A major dimension of the structural organization of language,
lip -(-rounding ([w], [u]). 155 capable of independent study, e.g. -(-phonology, -(-syntax. 82 2 (gram) A
laminal (phonet) Said of a -(-consonant made with the H-blade (or lamina) kind of representation recognized within the -(-derivation 2 of a sentence,
of the tongue in contact with the upper lip, teeth, or -(-alveolar ridge ([s], e.g. +deep vs -(-surface grammar. 409 3 (gram) One of a series of structural
[t]). 157 layers within a -(-sentence (+clause, -(-phrase, +word, etc.); also, rank. 95 4
langage (ling) The human faculty of speech. 407 (phonol) A degree of -(-pitch height or -(-loudness during speech. 170 5
language (gen) 1 The systematic, conventional use of sounds, signs, or written (socio) A mode of expression felt to suit a type of social situation (formal,
symbols in a human society for communication and self-expression. 396 2 intimate, etc.). 40
A specially devised system of symbols for programming and interacting with lexeme (sem) The
smallest +contrastive unit in a -(-semantic system (run,
computers. 396 3 The means animals use to communicate. 397 4 (din) cat, switch on); also, lexical item. 104
The symbolic aspects of language 1 excluding -(-phonetics (and often -(-phono-
, lexical diffusion (socio) The gradual spread of a linguistic change through
logy). 265 a language. 332
language acquisition 1 (psycho) The process of learning a -(-first language lexical item see lexeme
in children. 226 2 (app) The analogous process of gaining a -(-foreign or lexical field see semantic field
+ second language; 368 lexical verb (gram) A +verb expressing an action, event, or state; also full-
language-acquisition device (psycho) The innate capacity that enables children main verb; cf. -(-auxiliary verb. 91
to learn their mother tongue; often, LAD. 234 lexical word see content word
language attitudes (socio) The feelings people have about their own language lexicography (gen) The art and science of dictionary-making (by lexicogra-
or the language(s) of others. 1 phers). 108
language centre/center (neuro) A brain area controlling -(-production or lexicology (sem) The study of the history and present state of a language's
-(-comprehension. 260 vocabulary. 108
language change (hist) Change within a language over a period of time; cf. lexicon (sem) The vocabulary of a language, esp. in dictionary form; also,
+ language shift. 328 lexis. 108
language contact (socio) A situation of prolonged association between the lexicostatistics (hist) A method for comparing the rates of change in sets of
speakers of different languages. 360 words in hypothetically related languages; cf. -t-glottochronology. 331
language disorder (din) A serious abnormality in the system underlying the lexis see lexicon
use of language. 264 liaison (phonol) The pronunciation of a -(-consonant at the end of a word
language laboratory (app) A classroom that uses tape-recorder booths to when the next word begins with a + vowel (Fr. C'est un ... 'It is a ...');
enable students to listen and respond to foreign utterances. 377 cf. -(-linking. 164
language loss 1 (socio) The gradual loss of ability to use a language, e.g. ligature (graph) A character in which two or more letters have been joined
in immigrant situations. 360 2 (din) The sudden loss of language as a together (ee, a). 194

424 • APPENDICES
 . ;

linear (graph) Said of + scripts using simply drawn characters instead of pictor- question 'how?' (quickly). 91
ial writing. 183 manual alphabet see dactylology
lingua franca (gen) A medium of communication for people who speak differ- manualism (din) The teaching of +sign 3 to the deaf, to the exclusion of speech
ent + first languages. 357 cf. +oralism. 267
lingual/linguo- (phonet) Said of any sound made with the tongue. 131 margins (phonet) Sound -(-segments that form the boundaries of a -(-syllable
linguist 1 (gen) Someone who is proficient in several languages. 412 2 (ling) ([k],[p]inc«p). 164
A student or practitioner of the subject of -(-linguistics; also linguistician. marking markedness (ling) The presence/absence of a particular +contras-
412 tive feature in a language or languages. 85
linguistic 1 (gen) Pertaining to + language 1
. 2 (ling) Pertaining to -(-linguis- masculine see gender
tics. 412 mass (gram) Said of + nouns that typically express general concepts and lack
linguistic atlas (ling) A set of maps showing the geographical distribution an indefinite -(-article or -l-plural (information); cf. -(-countable. 91
of linguistic items; also, dialect atlas. 30 matched guise (socio) Recording two languages or + dialects by the same
linguistic change see language change speaker, in order to elicit listener -(-language attitudes. 23
linguistic geography see geographical linguistics mathematical linguistics (ling) The study of the mathematical properties of
linguistician see linguist 2 language, esp. using statistical or algebraic concepts. 412
linguistic relativity/determinism (ling ) The hypothesis that a language's struc- matronymic (sew) A name derived from that of a person's mother (Marjori-
ture governs the way in which its speakers view the world. 15 son); also, metronymic. 112
linguistics (ling) The science of language. 404 maxims see conversational maxims
linguistic science(s) see linguistics measure see metre
linking (phonol) A sound introduced between two +syllables or -(-words, mechanical translation see automatic translation
for ease of pronunciation (E. 'linking /r/' of car and ..); cf. +liaison. 164 medium (gen) A dimension of message transmission, esp. speech, writing,
linking verb see copula + sign 3 123
.

lipogram (gen) A text from which a specific letter has been omitted through- mel (acou) A unit of measurement for -(-pitch. 144
out. 65 mentalistic (ling) Said of the study of language through introspection rather
lipreading see speech reading than through the description of behaviour; cf. -(-behaviourism. 409
liquid (phonet) [1]- or [r]-type -(-consonants. 166 merger (hist) The coming together of linguistic units that were originally dis-
lisp (din) An abnormal -(-articulation of a -(-sibilant -(-consonant, esp. [s]. tinguishable. 328
277 mesolect (socio) In +creole studies, a -(-varietv between +acrolect and +basi-
literal (gen) The usual meaning of a word or phrase; cf. -(-figurative. 70 lect. 336
literal translation see free translation metalanguage (ling) A language used for talking about language. 248
loan translation see caique metanalysis (hist) A word deriving from a word-boundary error (E. a naddre
loan word borrow see -> an adder). 328
localization (neuro) The control of a specific kind of behaviour, e.g. speech, metaphor (gen) A -(-figurative expression in which one notion is described
by a specific area of the brain. 260 in terms usually associated with another (launch an idea). 70
locative (gram) A form that expresses location (at the corner). 93 metathesis (ling) Alteration in a normal sequence of elements, esp. sounds
locutionary act (prag) The -(-speech act of making a meaningful utterance; (aks tor ask). 32S
cf.+illocutionary act. 121 metonymy A -(-semantic change where an attribute is used for the whole
(hist)
logocentrism (styl) A language- or word-centred view of literature or other (crown =
king). 70
behaviour. 79 metre/meter (poet) A rhythmical verse pattern; also, measure. 74
logogram (graph) A symbol that represents a -Hword (as in Chinese); also, metrics (poet ) The study of metrical structure. 74
logograph. 200 metronymic seematronymic
logograph see logogram microlinguistic (ling ) Said of highly detailed studies of language data. 404
A word puzzle using -I- anagrams. 65 Said of a -(-vowel 1 -(-articulated between -(-high and -How
1 1
logogriph (gen) mid (phonet)
logop(a)edics see speech pathology tongue positions ([e], [a]); cf. -(-close, +open 3 154 .

logop(a)edist see speech pathologist middle ear (anat) Part of the ear between the ear drum and the -(-inner ear.
logorrhoea (gen, din) Excessive, uncontrolled, incoherent speech. 271 142
long (phonol) Said of a -(-phoneme that -(-contrasts because of its greater minim (graph) A single downstroke of the pen. 187
-(-duration (the -(-vowel of beat compared with bit). 153
1
minimal pair (phonol) Words that differ in meaning when only one sound
longitudinal (gen) Said of studies that follow -(-language acquisition over a is changed. 160

period of time; cf. + cross-sectional. 229 minuscule (graph) A form of writing consisting of small letters; cf. -Hmajus-
look-and-say (app) A method of teaching reading that focusses on the recogni- cule. 186
tion of whole words; also, whole word; cf. + phonics. 251 miscue (app) An error made by someone learning to read; studied by miscue
loudness (phonet) The auditory sensation that primarily relates to a sound's analysis. 250
intensity; also, volume. 44 mismatch (psycho) A child's H-semantically inappropriate use of a word,
low (phonet, phonol) 1 Said of -(-vowels made with the tongue in the bottom
1
where there is no apparent basis for the error. 244
area of the mouth ([a], [a]). 153 2 Said of -(-tones that use a relatively 1
mistake see error2
low level of -(-pitch range. 172 3 (socio) Said of the less prestigious -(-variety mixing see code-switching
in +diglossia. 43 modal (gram) A +verb that signals contrasts in speaker attitude (+mood),
e.g. may, can. 93
machine translation (gen) The use of a computer to carry out the task of modality 1 (semiot) A + medium of communication. 396 2 (gram) The sys-
automatic translation. 350
-(-translation; also, tem of + modal expression. 93
macrolinguistics (ling) A broad conception of linguistic enquiry, including mode (semiot) A +medium of communication. 396
psychological, cultural, etc. factors. 404 modelling (app) Providing language examples for a learner to follow. 368
main clause (gram) A -(-clause that does not depend on any other part of modification 1 (gram) The structural dependence of one element (a modifier)
a +sentence (Tie man arrived after the bus left); also, independent clause; upon another. 95 2 (phonet) Movement that affects the air flow in the
cf. -(-subordination. 95 +vocal tract. 130 3 (hist) Any of several kinds of -l-formal 1 change in a
maintenance see language maintenance word (man —» men). 328
main verb see lexical verb modifier see modification
majuscule (graph) A form of writing consisting of capital letters; cf. -(-minus- monaural (phonet) Using one ear; cf. -(-binaural. 142
cule.186 monitoring (app) Critical self-listening. 372
malapropism (gen) An inappropriate word, used because of its similarity in monogenesis (hist) The view that all languages come from an original lan-
sound to the intended word (a paradigm of virtue). 77 guage; cf. -Hpolygenesis. 291
malformation (gen) An unacceptable -l-word formation due to a wrong +ana- monoglot see monolingual
logy (gooses for geese). 330 monolingual (gen) Said of a person/community with only one language; also,
manner 1 (phonet) The specific process of + articulation used in a sound's unilingual; cf. + bilingual, + multilingual. 360
+ produaion (+plosive, etc.). 157 2 (gram) An + adverbial answering the monologue (gen) Speech by a single person. 48

1 GLOSSARY • 425
monometer (poet) A line of verse containing a single unit of rhythm + foot). (
only (brillig). 90
74 non-count see countable
monomorphemic {gram) Said of a +word consisting of a single +morpheme. non-defining see restrictive
90 non-finite see finite
monophthong (phonet) A +voweP with no detectable change in quality dur- non-restrictive see restrictive
ing a -(-syllable [car). 154 non-standard see standard
monosyllabic {phonot) Said of a +word consisting of a single -(-syllable. 86 non-verbal (semiot) Said of communication that does not use words, e.g. ges-

mood (gram) Attitudes of fact, wish, possibility, conveyed by a

etc. -t-verb tural. 399
(a +modal) or -Hclause, e.g. + indicative, -(-subjunctive. 93 normative see prescription
mora (phonol) A minimal unit of rhythmical time equivalent to a short + syll- notation see transcription
able. 74 notional 1 (gram) Said of a grammar whose terms rely on +extralinguistic
1
morph (gram) The physical form of a -(-morpheme. 90 notions, e.g. action, duration, time; cf. -(-formal . 91 2 (app) Said of a
morpheme (gram) The smallest -(-contrastive unit of -(-grammar (man, de-, syllabus based on an analysis of sentence meanings and functions; cf. -(-com-
-tion. -s, etc); cf. + bound form, -(-free form. 90 municative approach. 374
morphemics (gram) The study of -(-morphemes. 90 noun (gram) A -t-word class with a naming function, typically showing con-
morphology (gram) The study of -i-word structure, esp. in terms of -(-mor- trasts of -fcountabiliry and -(-number, and capable of acting as -(-subject
phemes. 90 or -l-object of a +clause. 91
morphophonemics see morphophonology noun phrase (gram) A -(-phrase with a +noun as -l-head (the tall man in
morphophonology (gram) The study of the relations between -I- morphology a hat). 95
and H-phonology. 90 nuclear see nucleus
morphosyntactic (gram) Said of a category whose definition involves both nucleus (phonol) The -(-syllable in a -Hone group that carries maximum
+ morphology and -(-syntax, e.g. + number. 90 -(-pitch prominence (nuclear tone, tonic) (She went to London). 170
morphotactics (gram) The arrangement of -(-morphemes in a linear sequence. number (gram) The grammatical category that expresses such contrasts as
90 '+ singular/ +plural/+ dual' (cat/cats, he is /they are). 93
motherese see caretaker speech
mother tongue see first language object (gram) A -(-clause element that expresses the result of an action (cf.

motor phonetics see articulation -(-direct/ -(-indirect object). 95

move (prag) A unit of speech in a -(-discourse. 116 objective see accusative
multilingual (gen) Said of a person/community with several languages; cf. object language (ling) A language that is the object of analysis (using a -(-meta-
-(-monolingual. 360 language). 82
mutation (gram, hist) A sound change in a word due to the influence of adjac- oblique (gram) Said of anv +case form of a -(-word except the -(-nominative.
ent -(-morphemes or -l-words (Welsh pen 'head' —> fy mhen 'my head'). 90 93
mutism (din) Involuntary inability to speak. 264 obsolescent (gen) Said of a word or sense no longer used. 330
mytheme (styl) The smallest contrastive unit of structure found in mythical obstruent (phonet) Sounds made with a constriction (-(-plosives, -(-fricatives,
narratives. 79 +affncates). 157
obviative (gram) A fourth- -(-person form used in some languages, typically
narrow (phonet) Said of a -(-transcription that shows many -(-phonetic details; contrasting with the third person to mean 'someone/ something else'. 92
cf. +broad. 158 occlusion (phonet) The length of the -(-closure during the -(-articulation of
nasal see nasality a -(-stop +consonant. 157
nasality (phonet) Sound made with the soft -(-palate lowered, thus allowing oesophageal/esophageal (phonet) Said of sounds or -l-voice produced 1
in the
air to resonate in the nose (nasals), e.g. [m], [n], or nasalized sounds, e.g. upper part of the oesophagus, esp. after -(-laryngectomy. 276
[a]. 130 off-glide, on-glide see glide
nasopharynx (anat) The part of the -(-pharynx adjoining the nasal -(-cavity. onomasiology (sem) The study of sets of associated concepts in relation to
130 their linguistic forms. 100
native language see first language onomastics (sem) The srudv of the +etvmology and use of -(-proper names.
native speaker (gen) A person whose language is a -I- first language or 'mother 112
tongue'. 368 onomatology see onomastics
nativism see innateness hypothesis onomatopoeia (sem, poet) Words that imitate the sounds of the world (splash,
naturalism (sem) The view that there is a close, 'natural' connection between murmur). 174
words and things; cf. -(-conventionalism. 404 ontogeny (ling) Growth and decay (here, of language) in the individual; cf.
natural language (gen) A language with +native speakers; cf. -(-auxiliary lan- -t-phylogeny. 228
guage 2 -(-language 2 352
, . open 1 (gram) Said of a -(-word class with unlimited membership (+noun,
negation (gram) A process expressing the denial or contradiction of some -(-adjective, -(-adverb, -(-verb); cf. -l-closed
1
. 91 2 (phonol) Said of a -(-syll-
2
or all of the meaning of a sentence; negative forms (negators) include not, able that ends in a -(-vowel 1
; cf. -(-closed . 164 3 (phonet) Said of -(-vowels 1

un-, etc.; cf. + affirmative. 243 made with the tongue in the lowest possible position ([a], [a]); -(-vowels
negative, negator see negation a degree higher are half-/ mid-open. 153
neologism (gen) A new or invented word or expression (linguistified). 73 opposition (phonol) A linguistically important contrast between sounds. 160
neurolinguistics (ling) The study of brain structure and function in relation optative (gram) A +mood of the -(-verb in some languages expressing desire
to language use, acquisition, and disorder. 261 or wish. 93
neuter see gender oracy (app) Ability in speaking and listening. 248
neutralization (phonol) The loss of a -(-contrast between two -(-phonemes oral (phonet) Said of sounds made in the mouth (as opposed to the nose,
in a particular -(-environment
1
(/t/ vs /d/ is 'neutralized' in stop). 161 + nasal). 152
neutral vowel see schwa oralism [din) The teaching of speech to the deaf, to the exclusion of +sign 3 ;

new see comment cf. -t-manualism. 267

node see nodule ordinal (gram) A class of numerals (first, etc.); cf. -(-cardinal number. 99
nodule (din) A small localized swelling ('node'), esp. on the +vocal folds. oropharynx (anat) The part of the -fpharvnx adjacent to the oral cavity.
276 130
noise (acou) A complex -(-sound wave with irregular vibrations. 137 orthoepy (gen) The study of correct pronunciation, esp. as practised in the
nomenclature (gen) A system of terms used in a specialized field. 380 17th/T8th centuries. 329
nominal (gram) A -t-noun or noun-like item. 91 orthography (gen) The study of the use of letters and the rules of spelling
nominalism see conventionalism in a language. 194
nominalization (gram) Forming a +noun from some other -(-word class (red- orthophonist see speech pathologist
ness, my answering .). 91 . . oscillograph (acou) An instrument that provides a graphic representation of
nominative (gram) An -(-inflection 1 that typically identifies the -(-subject of -(-sound waves (an oscillogram). 138
a -(-verb (Ger. Der Mann seht den Mann 'The man sees the man'). 93 oscilloscope (acou) An instrument for the visual displav of -(-sound waves.
nonce formation (ling) An invented or accidental linguistic form, used once 138

426 APPENDICES
 1

ossicles (anat)The bones of the +middle ear. 143 pejoration see deterioration
otology (din) The study of diseases of the ear. 266 pejorative (gen) Said of a linguistic form that expresses a disparaging meaning
oto(rhino)Iaryngology (din) The study or diseases of the ear, nose, and throat. (goodish). 330
266 pentameter/pentametre (poet) A line of verse containing five units of rhythm
overcorrection see hypercorrection ( + foot). 74
overextension see overgeneralization perfect (gram) A -(-tense
1
form typically referring to a past action that has
overgeneralization (psycho) A learner's extension of a word meaning or gram- present relevance (1 have asked); cf. -(-pluperfect. 93
matical -t-rule beyond its normal use (men —» mens); also, overextension.
1
perfective (gram) A +verb +aspect typically stressing the completion of an
244 action; contrasts with imperfective. 93
overtone see harmonic performance (ling) The language actually used by people in speaking or writ-
oxymoron (rhet) A + figurative combination of incongruous or contradictory ing; cf. -(-competence. 409
words. 70 performative (prag) An -(-utterance or +verb that performs an action (pro-
oxytone (gen) A word with heavy -(-stress on the final -(-syllable (represent). mise, baptise). 121
169 periodic (acou) Said of a -(-waveform that involves a repeated pattern of vib-
ration; contrasts with aperiodic (random) vibration. 133
1
paedography (graph) A writing system devised to help children to read. 1 94 periphrasis 1 (gram) The use of separate -l-words instead of -(-inflections
to express a -(-grammatical relationship (periphrastic) (more happy for hap-
1
palaeography (graph) The study of ancient writings and inscriptions. 187
palatal (phonet) Said of sounds made in the area of the hard -(-palate ([q],
pier). 922 see circumlocution
[J])- 155 perlocutionary act (prag) A -fspeech act that achieves a particular effect on
palatalization (phonet) An + articulation in which the tongue moves towards a listener (frightens, persuades); cf. -iTocutionary act. 121
the hard -(-palate while another sound is being made. 156 perseveration (din) Involuntary continued use of a linguistic form. 271
palate (anat) The arched bony structure that forms the roof of the mouth; person (gram) A grammatical form (esp. a +pronoun or +verb) referring
divided into the hard palate and soft palate (velum). 124 directly to the speaker ('first person'), addressee ('second person'), or others

palato-alveolar (phonet) Said of a -(-consonant made between the -(-alveolar involved in an interaction (esp. 'third person'). 93
ridge and the hard -l-palate ([/]). 155 personal pronoun see person
palatography (phonet) The instrumental study of tongue contact with the personification (poet, rhet) The + figurative attribution of human qualities
-l-palate, displayed as a palatogram. 140 to non-human notions. 70
palilalia (din) Involuntary repetition of words or phrases. 270 petroglyph (gen) An ancient stone inscription; also, petrogram. 196
palilology (rhet) Word repetition for emphasis. 70 petrogram see petroglyph
palindrome (gen) Words or expressions that read the same backwards or pharyngeal (phonet) Said of sounds made in the -(-pharynx ([h], [?]). 155
forwards. 65 pharyngealization (phonet) Narrowing of the -(-pharynx while another speech
pangram (gen) A sentence that contains every letter of the alphabet. 65 sound is being made. 156
paracusis (din) Any hearing abnormality. 266 pharynx (anat) The part of the throat above the -(-larynx. 130
paradigm (gram) The set of -(-inflectional -(-forms of a word (Lat. amo/ama-
1 1
phatic (ling) Said of language used to establish atmosphere or maintain social
s/amat. .). 90
.
contact. 10
paradigmatic (ling) The relationship of -l-substitution between a linguistic unit philology see comparative philology
and other units at a particular place in a + structure 407 2
.
philosophical linguistics (ling) The study of language in relation to philosophi-

paradox (gen) An apparent contradiction that contains a truth. 70 cal concepts. 412
paragram (gen) A play on words by altering a letter, esp. in humour. 63 phon (acou) Unit of measurement for the -(-loudness level of a sound. 144
paralanguage (ling) Features of speech or +body language considered to be phon: aksthenia (din) An abnormally weak -(-voice quality 1 276 .

marginal to language; studied by paralinguistics. 169 phon(a)esthetics (phonet) The study of the aesthetic or symbolic properties
paralinguistics see paralanguage of sound. 174
parallelism (styl) The use of paired sounds, words, or constructions. 60 phonation (phonet) The production of -(-voice 1 through the use of the +vocal
paraphasia (din) An involuntary error in the production of words or phrases. folds. 128

270 phone (phonet) The smallest perceptible + discrete + segment of speech sound.
paraphrase (gen) An alternative version of a sentence that does not change 152
its meaning. 107 phoneme (phonol) The smallest -(-contrastive unit in the sound system of
pararhyme (poet) The repetition of the same initial and final consonants in a language. 160
different words (tail/ tall). 74 phonemics (phonol) The analysis of -(-phonemes. 160
parataxis (gram) Constructions joined without the use of -l-conjunctions (/ phonemic transcription (phonol) A -(-transcription of the -(-phonemes in an
had tea, eggs. .); cf. +hypotaxis. 95
.
utterance. 160
parent language (hist) A language from which other languages descend, e.g. phonetic 1 (phonet) Pertaining to phonetics. 152. 2 (graph) Pan of a
Latin is the parent of daughter languages French, Spanish, etc., which are -(-logogram that indicates its pronunciation; cf. -I- determinative. 200
thus sister languages to each other. 292 phonetic alphabet see phonetic transcription
parole (ling) The concrete utterances of a speaker; cf + langue. 407 .
phonetician (phonet) A -l-phonetics specialist. 152
paronomasia (gen) A play on words, or pun. 63 phonetics (phonet) The science of speech sounds, esp. of their production,
2
paronym (hist) A word that comes from the same -l-root as another (wise/ wis- transmission, and reception (' + acoustic/ +articulatory/ + auditory phone-
dom). 90 tics'). 152
paroxytone (phonol) A word with heavy -(-stress on the penultimate syllable phonetic spelling (gen) A spelling system that represents speech sounds in
(telegraphic). 169 a one-to-one way. 213
parsing (gram) Analysing and labelling the grammatical elements of a -fsen- phonetic transcription (phonet) A + transcription of all distinguishable phones
tence; also, diagramming. 249 in an utterance, using special symbols (a phonetic alphabet). 158
1
participle (gram) A word derived from a -(-verb and used as an -(-adjective phoniatrics (din) Study of pathologies affecting -Hvoice quality and pronun-
(a smiling face). 91 ciation. 264
particle (gram) An -(-invariable word with a -(-grammatical function (fogo, 1
phonics (app) A method of teaching reading that trains recognition of the
not). 91 sound values of individual letters; cf. + look-and-say 25 .

partitive (gram) A form that refers to a part or quantity (some, piece, ounce). phonogram (graph) A symbol representing a speech sound; cf. -(-logogram.
92 199
part of speech see word class phonography (graph) A writing system that represents individual speech
pasigraphy (gen) The use of a system of symbols understood between lan- sounds. 197
guages (1, 2, +,£). 200 phonologist (phonol) A -(-phonology specialist. 160
passive see active, voice
2 phonology (phonol) The study of the sound systems of languages. 160
patient see goal phonostylistics (poet) The study of the expressive use of sound, esp. in poetry.

patois (gen) A provincial -(-dialect. 24 74

patronymic (gen) A name derived from that of a person's father (Peterson). phonotactics (phonol) The specific sequences of sounds that occur in a lan-
112 guage. 160

I GLOSSARY -427
phrasal verb (grant) A +verb consisting of a lexical element and -t-particle(s) preliterate (hist) Said of a language before a writing system has developed.
(get up).91 196
phrase (gram) A group of words smaller than a + clause, forming a +gramma- premodification (gram) Items that occur within a -(-phrase before the +head
1
tical unit (in a box). 95 (the funny clown). 95
phrase marker (gram) A structural representation of a sentence in +generative preposition (gram) Items that -(-govern and typically precede +nouns, -(-pro-
grammar, usually in the form of a +tree diagram. 96 nouns, and certain other forms (in the box, to me, by running). 91
phrase-structure grammar (gram) A -(-generative grammar that provides an presby(a)cusis (din) Gradual loss of hearing acuity as a result of old age.
analysis of sentences into -(-constituent elements. 96 266
phylogeny (hist) Historical development (here, of language) in communities prescription (gen) An authoritarian (prescriptive or normative) statement
or in the human race as a whole; cf. -(-ontogeny. 328 about the correctness of a particular use of language; cf. -(-description. 2

physiological phonetics see articulatory phonetics prescriptive see prescription

pictogram/ pictograph (graph) A symbol used in picture writing. 197 presupposition (sem) The information that a speaker assumes to be already
pidgin (socio) A language with a reduced range of structure and use, with known; cf. + focus. 120
no -(-native speakers. 334 preterite (gram)
1
-(-simple past +tense form of a -(-verb (/ saw). 93
The
pidginize (socio) To
develop into a -(-pidgin. 334 prevocalic (phonet)
1
Preceding a -(-vowel 164 .

pitch (phonet) The auditory sensation of the height of a sound. 133 principal parts (gram) The -(-forms 2 of a -(-verb required to determine its
place of articulation (phonet) The anatomical point in the -l-vocal tract where -(-conjugation (Lat. amo/amare/amavi/amatum). 91
a speech sound is produced (+labial, -(-dental, etc.). 155 proclitic (gram) An unstressed word that depends on and is pronounced with
pleonasm (gen) The unnecessary use of words (in this present day and age). 2 a following word (an), cf. -(-enclitic. 91
plethysmograph (phonet) An instrument that records changes in air volume production (ling) The active use of language; cf. -(-comprehension. 261
during speech. 125 productivity (ling) The creative capacity of language users to produce and
plosive (phonet) Said of a -(-consonant made by the sudden release of a com- understand an indefinitely large number of sentences. 397
plete -(-closure in the -(-vocal tract ([p], [k]). 157 proficiency test (app) A test that measures how much of a language someone
pluperfect (gram) A
-t-verb form that typically expresses completion of an knows. 377
action before a specific past time (/ had jumped); also, past perfect. 93 pro-form (gram) An item that substitutes for another item or construction
plural (gram) A -(-word form typically expressing 'more than one' in -(-number (so does John). 119
(boys, them). 93 prognostic test see aptitude test
plurilingualism see multilingualism progressive 1 (gram) A +verb form that typically expresses duration or incom-
plurisegmental see suprasegmental pleteness (He is running); also, continuous; cf. -(-simple. 93 2 (phonol)
pneumograph (phonet) An instrument that measures chest movements during Said of an -(-assimilation when one sound causes a change in the following
breathing. 139 sound ([J] -» [tf] in did she). 164
pneumotachograph (phonet) An instrument that measures air flow from nose prolongation (din) The abnormal or controlled lengthening of a sound in
and mouth. 139 + stuttering. 278
poetics (poet) The linguistic analysis of poetry (and sometimes of other crea- prominence (phonet) The degree to which an element stands out from others
tive language use). 73 in -(-environment 1 169
its .

point size (graph) A system for measuring the size of pieces of type. 190 pronominal (gram) An item that functions as a -(-pronoun. 91
polarity (gram) The system of contrast between -(-affirmative and -(-negative pronoun (gram) An item that can substitute for a +noun or +noun phrase
in a language. 93 (he, who, himself). 91
polyalphabetic (gen) Said of a -(-cipher that makes use of many letter transfor- proper name/noun (gram) A +noun that labels a unique place, person, ani-
mations. 58 mal, etc. and lacks the grammatical forms of a -(-common noun. 1 12
polygenesis (hist) The view that languages come from several original sources; proposition (sem) A unit of meaning in -(-statement form that is asserted to
cf. +monogenesis. 291 be true or false (The cat is 107
asleep).
polyglot/ polylingual see multilingual prop word see empty word
polysemia/polysemy (sem) Several meanings of a word (plain = 'dull/ob- proscriptive (ling) Said of -(-prescriptive + rules 2 that forbid a usage, e.g. criti-

vious/ .. .'). 106 cism of very unique. 2

polysemic/polysemous (sem) Showing +polysemy. 106 prosodic features see prosody 1
polysyllabic (phonet) Having more than one -(-syllable. 87 prosody 1 (phonol) The linguistic use of -(-pitch, -(-loudness, -(-tempo, and
polysynthetic (ling) Said of a language that uses long -(-word forms with a + rhythm. 169 2 (poet) The study of versification. 74
complex -(-morphology; also, incorporating, incapsulating. 293 prosthesis (phonol) The insertion of an extra sound at the beginning of
polysystemic (ling) Said of an analysis that sets up different linguistic systems a word. 328
2
at different places in -(-structure 408 . proto-language 1 (hist) A hypothetical ancestor language or form ('Proto-
popular etymology see folk etymology Indo-European'). 292 2 (psycho) A stage before the emergence of a recog-
portmanteau (gram) A +morph that can be analysed into more than one nized linguistic form (proto-word). 237
-(-morpheme (Fr. au = a le). 90 proto-word see proto-language
positive see affirmative, degree, polarity proverb (gen) A short, pithy, rhythmical saying expressing a general belief.
possessive (gram) A linguistic form that indicates possession (my, mine, 53
John's). 93 proxemics (semiot) The study of the communicative function of body distance,
postal veolar (phonet) Said of a -(-consonant made at the rear of the -(-alveolar posture, etc. 399
ridge. 155 pseudepigraphy (gen) The false ascription of an author's name to a written
postcreole continuum (socio) A related series of + varieties that develops when work. 187
-l-creole speakers are taught in the -(-standard language. 336 pseudolinguistic (gen) Said of vocal behaviour with a superficial resemblance
postmodification (gram) Items that occur within a +phrase after the -l-head to language. 11
(the man in a suit). 95 pseudonym (gen) A fictitious name, esp. of an author. 112
postposition (gram) A -(-particle that follows the +noun it -(-governs (Jap. psittacism (gen) Meaningless repetitive ('parrot-like') speech. 270
X kara Y made 'from X to Y'); cf. -(-preposition. 92 psycholinguistics (psycho) The study of language in relation to psychological
post-structuralism (styl) A reaction to the + structuralist analysis of literary processes. 412
texts. 79 pulmonic (phonet) Pertaining to the lungs. 125
postvocalic (phonet) Following a + vowel 1
. 164 pure tone (acou) A -(-sound wave of a single -(-frequency; cf. -(-complex tone.
pragmatics (prag) The study of the factors influencing a person's choice of 132
language 1 120 . pure vowel (phonet) A -(-vowel 1
that does not change in quality during a
predicate (gram) The + clause element that gives information about the -(-sub- -(-syllable; cf. -(-diphthong. 154
ject (He saw a dog). 94 purism (gen) The view that a language needs to preserve traditional standards
predicative see attributive of correctness and be protected from foreign influence. 2
prefix (gram) An -(-affix added initially to a -(-root
1
(unhappy). 90
prelinguistic (psycho) Said of child utterance before the emergence of lan- qualifier (gram) A word or phrase that limits the meaning of another element
guage. 228 (red car). 95

428 •
APPENDICES
quality {phonet) The characteristic + resonance, or + timbre, of a sound. 133 a reduced range of structures; cf. + elaborated code. 40

quantifier (sent) An item expressing amount {all, some, each). 91 restricted language (socio) A
highly reduced linguistic system found in nar-
quantitative linguistics see mathematical linguistics rowly defined settings, e.g. heraldry, weather reporting. 56
quantity {phonol) The relative +duration of +contrastive sounds and -(-syll- restrictive (gram) Said of a -(-modifier that is an essential part of the identity
ables. 169 of another element (my brother who's abroad); also, defining; contrasts with
question {gram) A sentence that asks for information or a response. 121 non-restrictive or non-defining, where the modification is not essential (my
brother, who's abroad). 95
radical see determinative retracted (phonet) Said of the backwards movement of an -(-articulator, e.g.
3
rank see level the -Hapex of the tongue. 155
readability formula {app) A measure of the ease with which a written text retroflex (phonet) Said of sounds made when the + apex of the tongue is

can be read. 252 curled back in the direction of the hard -(-palate ([t], [r]). 155
Y'
realization {phonol) The physical expression of an abstract linguistic unit. rewrite rule (gram) A -l-rule 1
in -(-generative grammar of the form 'X—»
82 (= 'replace X by Y'). 97
rebus {gen) A combination of letters, pictures, and pictograms to make words rheme The new information conveyed in a sentence; cf. +theme. 120
(ling)
and sentences. 65 rhetoric (rhet) The study of effective speaking and writing. 70
received pronunciation {phonol) The regionally neutral, prestige accent of rhetorical question (gram) A -(-question to which no answer is expected. 121
British English. 39 rhinolalia/ rhinophonia (din) +Nasal resonance. 276
receptive aphasia {din) A disorder of language* + comprehension caused by rhotacism (din) A defective use of [r]. 277
brain damage; cf. +expressive 2 270 . rhotic area (socio) A -(-dialect area in which /r/ is pronounced following
recipient see goal a + vowel (car). 28
reciprocal 1 {gram) An item that expresses the meaning of mutual relationship rhoticization (phonet) The -(-articulation of -(-vowels with r-colouring. 153 1

{each other). 91 2 {phonol) A type of -(-assimilation in which sounds rhyme (poet) A correspondence of ^-syllables, esp. at the ends of poetic lines.
influence each other. 164 74
reconstruction {hist) The -(-comparative linguistic analysis of extant texts to rhythm (phonol) The perceived regularity of prominent units in speech. 169
work out an earlier, non-extant state of a language. 292 roll see trill

recursive {gram) 1
Said of a -(-grammatical +rule' that is capable of repeated romanization (graph) The use of the Latin alphabet to transcribe non-Latin
application. 97 writing systems. 313
reduction {gram) The lack of one or more of the normal + constituents
1 root (gram) The base form of a word, from which other words derive (mean-
1
in a construction {gone to town); cf. -(-ellipsis. 95 2 {phonol) A -Hvowel 1
ingfulness);cf. +stem. 90 2 (hist) The earliest form of a word. 330 3
that becomes + central when a word is -(-unstressed ([a] —» [a] as in he can (phonet) The furthest-back part of the tongue. 131 4 see isolating lan-
—» he c'n go). 164 3 {hist) A narrowing of meaning. 330 4 see clipping, guage
redundant {ling) Said of a feature that is unnecessary for the identification rounded see rounding
or maintenance of a linguistic -(-contrast. 146 rounding (phonet) The visual appearance of the lips, permitting + contrasts
2
reduplication {gram) 1 A +form involving a repeated element (Lat. curro of rounded ([u]) and unrounded ([i]). 152
'run' — cucurri 'ran'). 175 2 A type of -(-compound word using repeated see formulaic
1
routine
elements {helter-skelter). 90 rule (gram) 1 A generalization about linguistic structure. 97 2 A -(-prescrip-

reference {sem) The relationship between linguistic forms and entities in the tive recommendation about correct usage. 3
world (referents). 102
referent see reference
referential see denotation, reference saccades (psycho) Rapid eye movements used when searching for an object.
reflexive {gram) A construction or -t-verb in which -(-subject and + object 208
relate to the same entity {She washed herself). 93 salience (phonet, psycho) The perceptual prominence of a sound. 145
reflexiveness {semiot) The capability of language to 'talk about' itself; cf. sandhi (gram) A sound change affecting a word used in a specific grammatical
-(-metalanguage. 397 1
-l-context (do-* don't). 405
regional dialect see dialect satem language (hist) An Indo-European language that replaced [k] by [s]

register 1 {phonet) A physiologically determined range of the human -(-voice 1

,
in such words as centum ('hundred'); cf. -t-centum language. 328
e.g. falsetto. 18 2 {socio) A socially defined -(-variety of language, e.g. scienti- scansion (poet) The analysis of -(-metre. 74
fic, 52 3 {phonol) Said of a +tone language that does not use
legal, etc. scheme (rhet) A -(-figurative effect, e.g. -(-rhyme, that changes the structure
-(-gliding tones. 172 of language without affecting its meaning; cf. + trope. 70
regression {psycho) A backward eye movement while reading a line of print. schwa/ shwa (phonet) An + unstressed -(-vowel 1 [a] made in the centre of the
208 mouth, heard at the end of such words as after and the. 153
regressive {phonol) Said of an -(-assimilation when one sound causes a change script (graph) Any system of written signs. 194
in the preceding sound ([t] —» [p] in hot pig); also, anticipatory. 164 secondary articulation (phonet) The lesser point of + stricture
in a sound
1
regular {ling) Said of a linguistic form that conforms to the -(-rules of the involving two points of 156
-(-articulation, e.g. lip -(-rounding.
language. 404 second language (app) A non-native language that has an official role in a
related (hist) Said of languages or forms that share a common origin. 292 country. 368
relative clause see relative pronoun second person see person
relative pronoun (gram) The item that introduces a -(-dependent -(-clause (rela- segment (phonet) A -(-discrete unit that can be identified in the stream of
tive clause) in a -l-noun phrase, referring back to the +noun (the car which speech. 161
was sold . . .). 95 segmental phonology (phonol) The analysis of speech into -(-phones or + pho-
relativity see linguistic relativity nemes; cf. -(-suprasegmental phonology. 160
release (phonet) + Vocal organ movement away from a point of + articulation, segmentation (phonet, gram) The process of analysing speech into + segments.
esp. in H-plosives. 157 96, 160
relexification (socio) A process in the development of -(-pidgins in which origi- selection(al) features see collocation
nal Portuguese vocabulary replaced by native language words. 337
is semantic see semantics
relic area (socio) A -(-dialect area that preserves linguistic features from an semantic component (sem) An element of a word's meaning (girl — » young,
earlier period. 28 female, human). 107
repair (prag) The correction of a misunderstanding or error made during a semantic differential (psycho) A technique for measuring the emotional asso-
conversation. 116 ciations of words. 103
repertoire (socio) The range of languages or + varieties that a speaker has semantic feature see semantic component
available. 48 semantic field (sem) An area of meaning identified by a set of mutually defining
reported speech see indirect speech words (colour, furniture, etc.). 104
resonance (phonet) Air vibrations in the -t-vocal tract that are set in motion semantic relations (sem) The -(-sense relations that exist between words, e.g.
by -l-phonation. 130 +hyponymy. 105
respiration (phys) The act of breathing. 124 semantics (sem) The study of linguistic meaning; also, semasiology, semato-
restricted code (socio) An informal + variety of language thought to display logy, semology. 100

I GLOSSARY •
429
semasiology/ sematology see semantics sound system (phonol) The network of + phonetic + contrasts comprising a
semi-consonant see semi-vowel language's -(-phonology. 165
semiology see semiotics sound wave (acou) A wave-like air disturbance from a vibrating body, which
semiotics (semiot) The study of the properties of signs and signalling systems, transmits sound. 132
esp. as found in all forms of human communication; also, semiology, signifies. source language (ling ) A language from which a word or text is taken. 344
399 spectrograph (phonet) An
instrument that gives a visual representation of
semi-vowel (phonet) A sound that displays certain properties of both -(-conso- the acoustic features of speech sounds, in the form of a spectrogram; also,
nants and H-vowels 1 ([1], [j]); also, semi-consonant. 153 sonagraph/sonagram. 136
semology see semantics spectrum (acou) The range of -I- frequencies that comprise a -l-sound wave.
sense relations (sent) The meaning relations between words, as identified by 135
the use of -(-synonyms, -(-antonyms, etc.; cf. + reference. 102 speculative grammar (ling) A type of grammatical treatise written in the mid-
sensorineural (din) Said of hearing loss due to damage to the -(-inner ear. dle ages. 406
266 speech 1 (gen) The oral medium of transmission for language (spoken lan-
sentence (gram) The largest structural unit that displays stateable -(-grammati- guage). 123 2 (din) The -(-phonetic -(-level 1 of communication (where dis-
1 2 4
cal relationships, not -l-dependent on any other -(-structure 94 . order can occur); cf. -I- language 265 .

sequencing 1 (psycho) Psychological processing of a series of linguistic ele- speech act (ling) An + utterance defined in terms of the intentions of the
ments. 275 2 (app) The order in which a graded series of items is given speaker and the effect it has on the listener, e.g. a -(-directive. 121
to a learner. 374 3 (prag) The rule-governed succession of utterances in speech community (socio) A group of people, identified regionally or socially,
a -(-discourse. 120 who share at least one language or -(-variety. 48
shwa see schwa speech defect (din) A regular, involuntarv deviation from the norms of speech.
sibilant (phonet) A -(-fricative made with a groove-like -(-stricture in the front 264
part of the tongue, to produce a hissing sound ([s], [J]). 157 speech disorder (din) A serious abnormality in the system underlying the
sight vocabulary (app) Words that can be recognized as wholes by someone use of spoken language. 264
learning to read. 250 speech event (prag) A specific act or exchange of speech (greeting, sermon,
sigmausm 1 (din) Abnormal pronunciation of [s], esp. as a -t-lisp. 277 2 conversation, etc.). 48
(poet) The repetitive use of [s] for effect. 74 speech impairment see speech defect
sign 1 (semiot) A feature of language or behaviour that conveys meaning, speech pathologist (din) A person trained to diagnose, assess, and treat
esp. as used conventionally in a system; also, symbol. 407 2 (graph) A -(-speech disorders; also, language pathologist/ therapist, logop(a)edist, orth-
mark used as an element in a writing system; also, symbol. 194 3 (ling) ophonist, speech therapist. 264
Deaf +sign language. 220 speech pathology (din) The study of all forms of involuntary, abnormal
signifiant (sem) That which signifies; contrasts with signifie, that which is linguistic behaviour; also, language pathology, logop(a)edics, speech therapy.
signified. 407 264
significant see contrastive speech perception (psycho) The reception and recognition of speech by the
signification (sem) The relationship between signs and the things or concepts brain. 145
to which they refer. 100 speech processing (psycho) The stages involved in the perception and produc-
signifies see semiotics tion of speech. 262
signifie see signifiant speech production (psycho) The planning and execution of acts of speaking.
sign language (ling) A system of manual communication, esp. one used by 262
the deaf. 220 speech reading (gen) A method of interpreting a speaker who cannot be heard
simile (rhet) A -(-figurative expression that makes an explicit comparison (as by following the movements of the mouth; also, lip reading. 225
tall as a tower). 70 speech reception see speech recognition
speech recognition (psycho) The initial stage of the -(-decoding process in
1
simple (gram) Said of a -(-tense 1 form that has no -(-auxiliary verb ('simple
1
present' He runs, etc.);
cf. -(-progressive
1
93 2 Said of a -(-sentence contain-
. -(-speech perception. 149
ing one -(-clause; cf. -(-complex sentence, -(-compound 1 95 . speech srience(s) (ling) The study of all factors involved in -(-speech production
sine wave (acou) A simple -(-waveform that produces a +pure tone. 132 and reception. 123
singular (gram) A form that typically expresses 'one of in -(-number (dog, speech stretcher (phonet) A device that presents a slowed but undistorted
It is). 93 recording of speech. 138
sister language see parent language speech surrogate (ling) A communication system that replaces the use of
situation (ling) The +extralinguistic setting in which a use of language takes speech (as in drum- or whistle-languages). 400
place. 48 speech synthesizer (phonet) A device that simulates the -(-speech-production
slang (gen) 1 Informal, -(-non-standard vocabulary. 53 2 The -(-jargon 1
of process. 146
a special group. 56 speech therapist/therapy see speech pathologist/pathology
slot (gram) A place in a construction where a class of items can be inserted spelling pronunciation (gen) The pronunciation of a word based on its spelling
(the -car). 95 (says as /seiz/). 180
social dialect see dialect spelling reform (gen) A movement to make spelling more regular in its relation
sociolect (socio) A social -!- dialect. 38 to speech. 215
sociolinguistics (socio) The study of the relationship between language and spirant see fricative
society. 412 spirometer (phys) An instrument for measuring the air capacity of the lungs.
soft palate see palate 125
solecism (gen) A minor deviation from what is considered to be linguistically split infinitive (gram) The insertion of a word between to and the -(-infinitive

correct.2 form of the -fverb in English (to boldly go). 2

sonagram/ sonagraph (phonet) see spectrograph splitting (hist) One -(-phoneme becoming two as a result of -l-sound change.
sonant (phonet) A -(-voiced sound. 128 328
sone (acou) Unit of measurement of -(-loudness. 144 spondee (poet) A unit of -(-rhythm in poetic -(-metre, consisting of two
sonorant (phonet) A -(-voiced sound made with a relatively free passage of -(-stressed + syllables. 74
air ([a], [1], [n]). 157 spoonerism (gen) The transposition of sounds between words, which gives
sonority (phonet) The relative prominence or 'carrying power' of a sound. a new meaning (queer old dean for dear old queen). 262
164 spread (phonet) Said of sounds made with lips stretched sideways ([i]). 152
sound change (hist) A change in the sound system of a language, over a period stammering see stuttering
of time. 328 standard (socio) A prestige -(-variety, used as an institutionalized norm in
sound law (hist) A regular, predictable series of -(-sound changes. 328 a community; forms or conforming to
varieties not this norm are non-stan-
sound pressure level (acou) The level of a sound as measured in -(-decibels. dard or (pejoratively) sub-standard. 24
134 standardization (socio) Making a +form 2 or -(-usage conform to the -(-stan-
sound shift (hist) A series of related -I- sound changes. 328 dard language. 364
sound symbolism (phonet) A direct association between the sounds of lan- starred form see asterisked form
guage and the properties of the external world. 174 state see stative

430 APPENDICES
 statement {gram) A sentence that asserts or reports information (The dog (go/ goes/ going/ gone/ went). 90
saw the cat). 121 suprasegmental (phonol) A vocal effect extending over more than one -(-seg-
static see stative ment, e.g. -(-pitch; also, plurisegmental. 169
statisticallinguistics (ling) The study of the statistical properties of lan- surface grammar/ structure (ling) A -(-syntactic representation of a -(-sentence
guage^). 86 that comes closest to how the sentence is actually pronounced. 98
stative (gram) Said of + verbs that express states of affairs rather than actions switching see code switching
(know, seem); also, static/state verbs; cf. +dynamic 1 93 . syllabary (graph) A writing system in which the symbols represent +syllables.
steganography (gen) The use of techniques to conceal the existence of a mes- 201
sage. 58 syllabic 1 (phonol) Said of a -(-consonant that can be used alone as a + syllable
stem (gram) The element in a word to which + affixes are attached; cf. 4-root 1 . (/l/ in bottle). 164 2 (graph) Said of a writing system in which the symbols
90 represent -(-syllables. 201
stenography (graph) Shorthand writing. 206 syllabification (phonol) The division of a -(-word into -(-syllables. 164
stereotyped see formulaic syllable (phonol) An element of speech that acts as a unit of + rhythm, consist-
stop (phonet) A
+consonant made by a complete +closure in the +vocal ing of a +vowel, +syllabic ,or vowel/ -(-consonant combination. 164
1

tract ([p], [6]). 157 syllable-timed (phonol) Said of languages in which the -l-syllables occur at
stratification (ling) A model of language as a system of related layers, or regular time intervals; cf. + isochrony. 169
2
strata. 83 symbol see sign
1 '

stress (phonet) The degree of force with which a +syllable is uttered; syllables syn(a)esthesia (sem) A direct association between -(-form 1 and meaning (si-

may be stressed or unstressed in various degrees (heavy, weak, etc.). 169 in slimy, slug, etc.). 174
stressed see stress synchronic (ling) Said of an approach that studies language at a theoretical
stress-timing see isochrony 'point' in time; contrasts with diachronic. 407
stricture (phonet) An + articulation in which the air stream is restricted to syncope (hist ) The loss of sounds or letters from the middle of a word (bo' sun).
some degree. 157 328
string (ling)A linear sequence of linguistic elements. 95 syncretism (hist) The merging of + forms 2 originally distinguished by -(-inflec-
strong form (phonol) A -(-stressed -l-word form. 169 tion
1
. 328
strong verb (gram) A +verb that changes its -l-root +vowel when 1
changing syndeton (gram) The use of -l-conjunctions to link constructions. 95
its + tense 1
(sing/ sang). 91 synonym (sem) A word that has the same meaning (in a particular -l-context 1
)

structural see structure as another word (a nice range/selection of flowers). 105

structuralism (ling) An approach that analyses language (or any human insti- syntactic (gram) Pertaining to -(-syntax. 94
tution or behaviour) into a set of + structures 1 79 .
syntactics see syntax
structural(ist) linguistics (ling)The study of a language's system of -(-formal 1 syntagm(a) (gram) A string of elements forming a unit in -(-syntax. 94
patterning (esp. in -(-grammar and -(-phonology), rather than of the meaning syntagmatic (ling) Said of the linear relationship between elements in a word
the patterns convey. 408 or construction. 407
structural semantics (sem) The studv of the +sense relations between words. syntax (gram) 1 The study of + word combinations; also, syntactics; cf. -(-mor-

105 phology. 2 The study of -(-sentence structure (including word structure).

structural word see function word 94
structure (ling) A
system of interrelated elements, which derive their (struc-
1 synthesis see speech synthesis
2
tural) meaning from the relations that hold between them. 96 2 A sequential synthetic 1 see inflecting language 2 see analytic
pattern of linguistic elements, at some analytical + level 2- 3
; cf . + deep/surface systematic phonology (phonol) An approach that represents the speaker's
structure. 98 knowledge of the -(-phonological relations between words (telegraph>/ telegra-
stuttering (din) A disorder of speech -I- fluency marked by hesitancy, -(-blocks, phy,etc.). 160

sound repetitions, etc.; also stammering. 278 systemic (ling) Said of an approach that analyses language into systems of
stylistics (ling) The study of systematic variation in language use (style) charac- -(-contrasts, and studies their functional use in social communication. 407
teristic of individuals or groups; also, stylolinguistics. 66
stylolinguistics see stylistics T (socio) Said of a linguistic form (esp. a -(-pronoun) used to express social
stylometrics see stylostatistics closeness or familiarity; cf. + V. 45

stylostatistics (ling) The quantification of -(-stylistic patterns; also, stylo- taboo (gen) Said of a linguistic form whose use is avoided in a society. 61
metrics. 67 tachistoscope (psycho) A device used in reading research that gives a very
subject (gram) The -(-clause -(-constituent about which something is stated brief exposure to a visual image, e.g. a letter. 208
(in the -(-predicate) (
The books are on the table). 94 tachygraphy (graph) Shorthand writing. 206
subjective see nominative tactics (ling) The systematic arrangements of linguistic units in linear sequence.
subjunctive (gram) A grammatical -l-mood used in some -(-dependent -(-clauses 82
to express doubt, tentativeness, etc. (Were he here ...); cf. -(-imperative, tag (gram) An element attached to the end of an utterance, esp. a tag question
-(-indicative. 93 {...isn't it?). 171
subordinate see dependent tagmeme (gram) A grammatical unit that relates an item's -(-form 1 and -(-syn-
1
subordination (gram) The dependence of one grammatical unit upon another, tactic -(-function the central notion in tagmemic analysis. 408
;

as in subordinate clauses (They left after the show ended). 95 tambre, tamber see timbre
subordinator (gram) A -(-conjunction used in -(-subordination (since, because). tap (phonet) A + consonant made by a single rapid tongue contact against
95 the roof of the mouth (as sometimes heard in the /t/ of writer). 166
sub-standard see standard target 1 (phonet) The theoretical position adopted by the +vocal organs dur-
substantive (gram) A +noun or noun-like item. 91 ing the + articulation of a sound. 137 2 (app) The language or -(-variety
substantive universal (ling) Basic elements that a -(-grammar 2 requires to ana- that is the goal of an activity, e.g. into which a -(-translation is being made.
lyse language data. 85 344
substitution (ling) The replacement of one element by another at a specific tautology (gen) An unnecessary repetition of a word or idea. 386
2
place in a -(-structure 119 . taxonomic (ling) Said of a linguistic approach that is mainly concerned with
2
substitution frame (gram) A specific + structure in which a + substitution classification. 408
takes place (a — cat). 95 technography (graph) A writing system devised for a specialized field. 194
substrate/ substratum (hist, socio) A -l-variety that has influenced the structure teknonymic (sem) A parent's name that derives from that of a child. 1 12
or use of a more dominant variety or language (the superstratum) in a commu- telegrammatic/telegraphic (psycho) Said of speech that omits + function
nity. 333 words and +dependent -(-content words (Man kick ball). 243
suffix (gram) An -(-affix that follows a +stem. 90 telescoped word see blend
superfix (phonol) A vocal effect that extends over more than one + segment, telestich (gen) An -(-acrostic based on the last letters of words or lines. 64
e.g. -(-stress. 169 teletex(t) (gen) The transmission of -(-graphic data from a central source to
superlative see degree a television screen. 193
superstratum see substratum telic (gram) Said of a -l-verb when the activity has a clear terminal point
suppletion (gram) The use of an unrelated form to complete a +paradigm (kick); contrasts with atelic verbs (play). 93

I GLOSSARY- 431
tempo Relative rate of speech. 169
(phortol) trill(phonet) A -(-consonant made by the rapid tapping of one +vocal organ
tense (gram) A change in the -(-form 2 of a +verb to mark the time at which
1 against another (trilled /r/); also, roll. 157
an action takes place (past, present, etc.). 93 2 see tension trimeter (poet) A line of verse containing three units of rhythm
tension (phonet) The muscular force used in making a sound, analysed as (+foot). 74
strong (tense), weak (lax), etc. 157 triphthong (phonet) A -l-vowel
1
containing three distinct qualities (tow-
tetrameter tetrametre (poet) A line of verse containing four units of rhythm er/taoa/)- 154
(+foot). 74 trisyllable (phonol) A word containing three +syllables. 164
text (ling) A stretch of spoken or written language with a definable communi- trivium (gen) The medieval study of grammar, rhetoric, and logic. 406
cative function (news report, poem, road sign, etc.). 116 trochee (poet) A unit of + rhythm in poetic -(-metre, consisting of a -(-stressed
1
textlinguistics (ling) The study of the linguistic ^-structure of +texts. 116 followed by an unstressed -(-syllable. 74
textual function (ling) The use of language to identify +texts. 119 trope (rhet) A -(-figurative effect, e.g. + metaphor, that changes the meaning
thematization (ling) Moving an element to the front of a sentence, to act of language; cf. +scheme. 70
as the + theme (Smith his name is); also, topicalization. 120 turn (prag) A single contribution of a speaker to a conversation (a conversatio-
theme (ling) The element at the beginning of a sentence that expresses what nal turn). 118
is being talked about (The cat w as in the garden); cf. -t-rheme. 120 typography (graph) The study of the graphic features of the printed page.
theography (styl) The study of the language people use to talk about God. 190
51 typological linguistics (ling) The study of the structural similarities among
thesaurus (gen) A book of words grouped on the basis of their meaning. languages, regardless of their history. 84
104
third person see person ultimate constituent see constituent
timbre (phonet) A sound's tonal quality, or 'colour', which differentiates umlaut (hist) A -(-sound change in which a + vowel 1
is influenced by the vowel
sounds of the same +pitch, +loudness, and +durarion. 133 in the following (*gosi—* geese). 328
-(-syllable
tip see apex uncial (graph) A form of writing consisting of large, rounded letters. 186
tmesis (rhet) The insertion of a word or phrase within another (absobloom- uncountable see countable
inglutely). 70 underextension (psycho) The use of a word to refer to only part of its normal
tone 1 (phonol) The distinctive +pitch level of a +syllable. 169 2 see pure meaning, e.g. a child's use otshoe to mean only 'own shoe'. 244
tone underlying structure see deep grammar/ structure
tone group/unit (phonol) A distinctive sequence (or +contour') of -Hones 1
ungrammatical see grammatical
in an utterance. 169 unihngual see monolingual
tone language (ling) A language in which word meanings or +grammatical 1
universal (ling) A property found in the analysis of all languages; cf. -t-formal-
-(-contrasts are conveyed by variations in +tone. 172 /+ substantive universal. 84
toneme (phonol) A +contrastive -t-tone 1 172 . universal grammar (ling) A -l-grammar2 specifying the possible form a lan-
tonetics (phonet) The study of the +phoneric properties of -(-tones 1 170 . guage's grammar can take. 84
tone unit see tone group um vocalic (gen) A written composition that uses only one -(-vowel 2 65 .

tonic see nucleus unmarked see marked

tonicity (phonol) The placement of -(-nuclear syllables in an utterance. 171 unproductive (ling) Said of a linguistic feature that is no longer used in the
topic (ling) The subject about which something is said (The pen is red); also, creation of new forms (the -th of length, width, etc.). 90
given information; cf. -(-comment. 94 unrounded see rounding
topicalization see thematization unstressed see stress
toponomasiology, toponomastics, toponomatology see toponymy unvoiced see voiceless
toponymy (gen) The study of place names. 112 urban dialectology (socio) The study of the speech patterns used within a
trachea (anat) The passage between lungs and -(-larynx. 124 modern community. 32
city
trade language (socio) A -l-pidgin used to facilitate communication while trad- usage (gen) The speech and writing habits of a community, esp. when there
ing. 334 is a choice between alternative forms (divided usage). 2

traditional (gram) Said of the attitudes and analyses found in language studies utterance (ling) A physically identifiable stretch of speech lacking any gramma-
that antedate -(-linguistic science. 3 tical definition; cf. -(-sentence. 94
transcription (phonet) A method of writing speech sounds in a systematic uvula (anat) The small lobe hanging from the bottom of the soft -(-palate.
and consistent way, from a particular point of view (+phonetic/ -(-phonemic 130
transcription, -(-narrow/ -(-broad); also, notation, script. 158 uvular (phonet) Said of a -(-consonant made by the -(-back of the tongue
transfer (app) The influence of a foreign learner's -(-mother tongue upon the against the +uvula ([r]). 155
2
-(-target language; positive transfer facilitates learning, whereas negative
transfer (interference) hinders it. 372 X (socio) +form 2 (esp. a -(-pronoun) used to express polite-
Said of a linguistic
transform(ation) (ling) A
formal 1 linguistic operation (a transformational ness or distance; +T. 45 cf.

rule) that shows a correspondence between two structures, e.g. active and valency (gram) The number and type of bonds that -(-syntactic elements may
passive -(-voice 2 sentences. 97 form with each other. 408
transformational grammar (ling) A -(-grammar2 that uses -(-transformations. variable rule (socio) A -(-rule
1
that specifies the +extralinguistic conditions
409 governing the use of a linguistic feature (or variable). 32, 332
transition 1 The way adjacent sounds are linked (+glide, +liaison,
(phonol) variable word (gram) A -(-word that expresses + grammatical relationships 1

etc.). 164 2 (acou) An acoustic change reflecting the movement of the by changing its +form 2 (walk/walks/ walking); cf. -(-invariable word. 91
-(-vocal organs towards or away from a +consonant (esp. -(-plosive) -(-articu- variant (ling) A linguistic -(-form 2 that is one of a set of alternatives in a
lation. 137 3 (socio) Said of a geographical region (a transition area) where given -(-context 1 (E. plural /s/, /z/, /iz/). 90
there is no clear boundary between adjacent + dialects. 28 variety (socio) A situationally distinctive system of linguistic expression (legal,

transitive (gram) Said of a +verb taking a -(-direct object (She saw a dog); formal, etc.). 48
cf. -(-intransitive. 93 velar (phonet) Said of -(-consonants made by the -(-back of the tongue against
translation (gen) 1 Conversion from one language into another. 2 Conver- the soft -(-palate, or velum ([k]). 155
sion of written texts from one language into another; cf. -(-interpret. 344 velaric (phonet) Said of sounds, e.g. -(-clicks, when the air has been set in
translative (gram) An -(-inflection that typically expresses the meaning of
1
motion by a 126
-(-closure at the soft +palate.
change from one place to another. 93 velarization (phonet) An + articulation in which the tongue moves towards
transliteration (gen) Conversion of one writing system into another. 346 the soft -(-palate while another sound is being made. 156
tree diagram (gram) A diagram used in -(-generative grammar to show the velopharyngeal (anat) Said of the area between the soft +palate and the back
1
hierarchical -(-structure of a -(-sentence. 96 wall of the -(-pharynx, which separates oral and nasal -(-cavities. 130
tremor (din) Involuntary shaking of the voice. 19 velum see palate
trial (gram) A grammatical contrast of -(-number in some languages, referring ventricular folds (anat) Bands of tissue that lie above the + vocal folds. 128
to 'three of. 92 verb (gram) A +word class displaying such contrasts as -(-tense +aspect, 1
,

2
trigraph (graph) Three written symbols representing one speech sound -(-voice , +mood, and typically used to express an action, event, or state
(manoeuvre). 213 (run, know, want). 91

432 APPENDICES
verbal group see verb phrase which can function as the centre of a -(-syllable ([e], [i]). 153 2 (graph)
verb phrase (gram) 1 A group of words that have the same grammatical func- The analogous sign in a writing system. 202
tion as a single +verb (has been running); also, verbal group. 95 2 In
+ generative grammar, the whole of a sentence apart from the first +noun wave see sound wave
phrase. 96 waveform (acou) A graph of the movement of air particles in a -hsound wave.
verbless (gram) A construction that omits a +verb (Although angry, he .). . . 132
95 wavelength (acou) The distance travelled by a -(-sound wave during a single
vernacular (socio) The indigenous language or -(-dialect of a community. 35 + cycle of vibration. 133
viewdata (gen) The interactive transmission of data between a central source weak form (phonol) The -(-unstressed form of a -t-word in connected speech
and a local television set. 193 (of—* cup of tea). 164
[a] in
vocal abuse (din) Overuse of the voice, resulting in a + voice disorder. 276 weak verb (gram) A +verb that forms its past -(-tense by adding an -(-inflec- 1

vocal folds (phonet) Two muscular folds in the + larynx that vibrate as a tion (walk -» walked); cf. -(-strong verb. 90
1

source of sound; also known as vocal cords/lips/bands. 128

1
well formed (ling) Said of a sentence that can be -(-generated by the -(-rules
2
vocalic (phonet) Pertaining to a +vowel'. 153 of a -(-grammar cf. — ill formed. 88
;
I—

vocalization (phonet) Anv sound or utterance produced by the +vocal organs. Wernicke's area (anat) An area of the brain that controls language -(-compre-
124 hension; cf. +Broca's area. 260
vocal nodules see nodules whisper (phonet) Speech produced without -(-vocal fold vibration. 128
vocal organs (phonet) The parts of the body involved in the production of whistled speech (ling) A form of communication in which whistling substitutes
speech sounds. 124 for the -(-tones of normal speech. 400
1

vocal tract (phonet) The whole of the air passage above the + larynx. 124 whole word see look-and-say
vocative (gram) A form (esp. a -t-noun) used to address a person, animal, widening (phonet) Enlarging the -(-pharynx to produce a different -(-vowel
1

etc. (Excuse me, sir); in some languages identified by an -finflection 1 93

. quality. 153
vocoid (phonet) A speech sound lacking +closure or audible friction; includes word (gram) The smallest unit of -(-grammar that can stand alone as a com-
1
-(-vowels and vowel-like sounds ([1], [j]). 153 plete utterance, separated by spaces in written language and potentially by
voice 1 (phonet) The auditory +vocal fold vibration (voiced sounds,
result of pauses in speech. 91
[b], [z], [e]); cf. 2 (gram) A grammatical system
+voiceless, +devoiced. 128 word blindness see dyslexia
varying the relationship between -(-subject and -(-object of the +verb, esp. word class (gram) A set of words that display the same -(-formal 1
properties,
contrasting active and passive voices (The cat saw the dog vs The dog was esp. their -(-inflections
1
and -(-distribution (-(-verb, +noun, etc.); also known
seen by the cat). 93 as part of speech. 91
voiced see voice 1
word ending (gram) An -(-inflection
1
used at the end of a word (boys, walking).
voice disorder (din) An involuntary, abnormal -l-voice quality that interferes
1
90
with communication; cf. +dysphonia. 276 word-finding problem (din) Inability to retrieve a desired word, symptomatic
voiceless (phonet) Said of sounds made without -l-vocal fold vibration ([{], of -(-aphasia. 271
[p]); also, unvoiced. 152 word formation (gram) The process of creating words out of sequences of
voice mutation (phonet) The development of an adult -l-voice quality 1 after -(-morphemes. 90
puberty; also, breaking. 19 word order (gram) The sequential arrangement of -(-words in a language.
voice onset time (phonet) The point when -(-vocal fold vibration starts relative 98
to the release of a -(-closure. 137
voiceprint (phonet) A + spectrographs display of the acoustic structure of x height (graph) The height of the small letter x. 190
a person's voice. 20
voice quality (phonet) 1 The permanent, background, person-identifying fea- zero (ling) An abstract unit used in an analysis that has no physical realization
ture of speech. 129 2 A specific tone of voice. 169 in speech. 90
volume see loudness zoosemiotics (semiot) The study of the properties of animal communication.
vowel (phonet, phonol) 1 A sound made without + closure or audible friction, 398

I GLOSSARY -433
II Special symbols and abbreviations

Non-phonetic symbols and abbreviations

All non-phonetic symbols and abbreviations

used in this encyclopedia are listed in alpha-
betical order below. Page references are given
when different sources cited use the same sym-
bol with different meanings.

a adjective (p. 108) f. formant

first POV point of view
a article (p. 86) h second formant PSG phrase structure grammar
A adverbial FL foreign language pu purpose of
adj adjective FLL foreign-language learning qu quantity of
adv adverb FLT foreign-language teaching r spoken response
ae affected entity FN" first name R non-linguistic response (p. 101)
AL artificial language fr from R response utterance (p. 118)
ALPAC Automanc Language Processing Fr French RA personal pronoun, subject
Advisory Committee Fx fundamental frequency RB personal pronoun, object
ALPS Automated Language Processing G -ing form of verb RLP Roman Lipi Parishad
Systems I initiating utterance [p. 118 RP received pronunciation
Ameslan American Sign Language I intrinsic muscles 'p. 131 s speech stimulus (p. 101)
Anat anatomy usage I CAME International Computer Archive of s substantive (p. 356)
AS Anglo-Saxon Modern English S subject (p. 95)
ASL American Sign Language incl inclusive pronoun S non-linguistic stimulus (p. 101)
at attribute of LNT interior sg singular
Aux auxiliary LPA International Phonetic Alphabet S'g signation
BASIC British American Scientific H high variety SL second language
International Commercial Hz hertz SP specification of
BBC British BroadcastingCorporation IC immediate constituent SPL sound pressure level
BEY Black English Vernacular ita initial teaching alphabet st state of
C complement p. 95 L low vanetv p. 43 su substance of
C sound velocity (p. 133) L Latin (p. 108) T tu (familiar) pronoun (p. 45)
C consonant p. 152 Lx laryngeal waveform T teacher (p. 248)
CALL computer-assisted language LI first language tab tabula
learning L2 second language TG transformational grammar
CB citizen band L3 third language TLN title with last name
CD subordinate that LAD language-acquisition device UPSID University of California, Los
CHILDES Child Language Data Exchange LARSP Language Assessment, Angeles, Phonological Segment
System Remediation, and Screening Inventory Database
CO containment of Procedure V verb
cps cycles per second LL Late Latin V vos (polite) pronoun (p. 45)
CSL Chinese Sign Language 1/m litres per minute V verb (p. 95)
CULT Chinese University Language LN last name V vowel p. 152)
Translator lo location of VA main verb
D past tense m masculine VB verb be
DAF delayed auditory feedback MLU mean length of utterance VH verb have
dB decibel msec milliseconds VHF very high frequency
DET determiner MT machine translation VOT voice onset time
dez designator n noun VP verb phrase
E extrinsic muscles N noun (p. 96) wph words per hour
-ed past tense form N nasal feature (p. 163) X invented category
EEG electroencephalography N past participle (p. 411) \ wavelength
EMG electromyography Non-SLIP Non-speech Language Initiation ? usage of doubtful
-en past participle form Program acceptability grammaticality
EURALEX European Association tor NP noun phrase (p. 96) *
unacceptable or ungrammatical
Lexicography NT proper noun (p. 411) usage
EURODICAUTOM O object enclose optional grammatical
European Automatic Dictionary OED Oxford English Dictionary elements (p. 97)
excl exclusive pronoun P pronoun enclose linguistic variables (p. 32)
EXT exterior P pupil(p. 248) + semantic component applicable
f feminine P phrase (p. 96) - semantic component inapplicable
F feedback (p. 118) Pathol pathology usage base form of verb
F frequency 'p. 133) PGSS Paget-Gorman Sign System 5 was form of be
F„ fundamental frequency Pi plural

434. APPENDICES
Phonetic symbols

This list comprises all the phonetic symbols

illustrated in this book, with the addition of
several variant forms. Diacritic and other conventions
Example
of use
open front unrounded vowel o mid-open back rounded vowel o uncertain segment ©
front unrounded vowel between P voiceless bilabial plosive pulmonic ingressive s

mid-open and open PPP voiceless bilabial trill voiceless B

central unrounded vowel between p voiceless lingualabial plosive voiced s
'
mid-open and open q voiceless uvular plosive ejective P'
h h
a open back unrounded vowel r voiced alveolar trill aspirated P
=
D open back rounded vowel r voiced alveolar tap/flap unaspirated P
b voiced bilabial plosive j voiced postalveolar fricative breathy-voiced/murmured b
b- voiced bilabial fricative (esp. US) voiced retroflex tap/flap n dental t
C
6 bilabial implosive -t voiced retroflex approximant a bidental h
bbb voiced bilingual trill r voiced alveolar fricative trill
;
or .
retroflex ft
B voiced lingualabial plosive R voiced uvular trill/ tap/flap u reverse labiodental tj

C voiceless palatal plosive B voiced uvular fricative labialized t

"
C voiceless palato-alveolar affricate S voiceless alveolar fricative y
°r palatalized f tj t
1

(esp. US) S voiceless palato-alveolar fricative or 1

? voiceless palatal fricative (esp. US) . or ~ velarized/pharyngealized 1 t

c voiceless alveolo-palatal fricative voiceless retroflex fricative ~ laryngealized b

d voiced alveolar plosive t voiceless alveolar plosive F
nasal fricative m F

4 voiced alveolar fricative (esp. US) t voiceless retroflex plosive tense articulation f

4 voiced retroflex plosive tf voiceless palato-alveolar affricate lax articulation m

d alveolar implosive J dental click
«'
very short articulation m
voiced palato-alveolar affricate u close back rounded vowel »«« reiterated articulation P-P-P
n n
D voiced alveolar tap « close centralrounded vowel non-audible release b
e mid-close front unrounded vowel Ul close back unrounded vowel () enclose a mouthed articula- (0
a or 3 central unrounded vowel u or O or u centralized back rounded vowel be- tion
a^ory r-coloured central vowel tween mid-close and close syllabic n
e central rounded vowel v voiced labio-dental fricative »or« simultaneous kp
f voiceless labio-dental fricative u voiced labio-dental approximant or .
raised e e

g voiced velar plosive mid-open back unrounded vowel or lowered e or e

g voiced velar fricative (esp. US) voiced labio-velar approximant + advanced u+ u
velar implosive voiceless labio-velar fricative - retracted i i-
ff
G voiced uvular plosive voiceless velar fricative centralized e
h voiceless glottal fricative close front rounded vowel ~ nasalized a
J
h voiceless pharyngeal fricative voiced palatal fricative/approxi- or" r-coloured o J o"
fi voiced glottal fricative mant (esp. US) long i:

e simultaneous fi and x centralized rounded vowel

front half-long i

i close front unrounded vowel between mid-close and close " non-syllabic u
i close central unrounded vowel z voiced alveolar fricative or w rounding f P
i or i central front unrounded vowel be- z voiced palato-alveolar fricative 3 more rounded 3>

tween mid-close and close (esp. US) C less rounded y<

'ma
1

voiced palatal fricative/approxi- voiced retroflex fricative primary stress

mant voiced alveolo-palatal fricative 1
secondary stress ,ma
J or j voiced palatal plosive p voiced bilabial fricative high level tone "ma
J voiced palato-alveolar affricate V voiced velar fricative . low tone
level .ma
'

(esp. US) e mid-open front unrounded vowel high rising tone 'ma
k voiceless velar plosive e voiceless dental fricative , low rising tone ,ma
*

1 voiced lateral approximant a voiced dental fricative high falling tone 'ma
t voiceless lateral fricative (esp. US) voiced palatal lateral v
low falling tone ,ma
"
4 voiceless lateral fricative voiceless bilabial fricative rising-falling tone "ma
fe
voiced lateral fricative x voiceless uvular fricative falling-rising tone "ma
I voiced retroflex lateral ; voiceless palato-alveolar fricative brief pause
J voiced alveolar lateral flap l palatalized voiceless palato-alveo- short pause
L voiced lingualabial lateral lar fricative -" long pause
m voiced bilabial nasal 3 voiced palato-alveolar fricative extra-long pause
rrj voiced labio-dental nasal 3 palatalized voiced palato-alveolar / or or |
1

M voiced lingualabial nasal fricative or# tone-unit boundary

n voiced alveolar nasal M voiced labio-palatal approximant t step-up in pitch
n voiced palatal nasal (esp. US) ui voiced velar approximant °or| onset of pitch movement
n. voiced retroflex nasal voiced pharyngeal fricative / boundary of foot
n voiced palatal nasal ?or? glottal plosive l
high pitch level
n voiced velar nasal O bilabial click
2
mid-high pitch level
3
N voiced uvular nasal I postalveolar click mid-low pitch level
4
O mid-close back rounded vowel b lateral click low pitch level
mid-close front rounded vowel mid-close back unrounded vowel [] enclose phonetic units (phones or dis-
ce mid-open front rounded vowel tinctive features)
a. open front rounded vowel // enclose phonological units (phonemes)

II SPECIAL SYMBOLS AND ABBREVIATIONS. 435

Ill Table of the world's languages

Nearly 1,000 living languages are listed in alphabetical order in the spoken. It is impracticable to list minority usage, which is often scat-
folowing pages, along with information about the language family tered over a wide area. Similarly, it is not possible to list all the alterna-
they belong to, where they are mainly spoken, and approximately how tive names given to some languages (see §47). For further details, and
many speakers there are. in all cases of omission from the following table, reference should
Languages with fewer than about 10,000 speakers have been omit- be made to C. F. &
F. M. Voegelin, Classification and Index of the
ted, except where they have particular linguistic significance (e.g. the World's Languages (1977).
sole survivor of a family). Only the main pidgins and Creoles have
Deen muuueu {bcc luriner, p. jj oj. rui numDcrs are appruxn lauuiib.
Abbreviations
bearing in mind the problems o speaker-counting discussed in §48.
:

Estimates var) enormously in certain parts of the world; and in cases

•
AR Autonomous Region
where there is disagreement both the lowest and the highest estimates ASSR Autonomous Soviet Socialist Republic
are given. Wh :re a language has widespread use as a second language, C Central
the estimated number of speakers is given in paren thesis. E East
Figures are given in thousands: 10 = 10,000 etc. Where nillions I./Is. Islands
of speakers are involved, the abbreviation m is used: 3m = 3 million, L Lake
etc. N North
The level of classification used in specifying language families reflects PR People's Republic
the discussion in the bodv of th e text (§§51-3). It should b e noted R River
that in some cases (mainly in Africa and the Ame ricas) a very general Rep. Republic
classification i s used - far more abstract than in other cases (such S South
as Indo-European). The question of levels is reviewed on p. 292 SSR Soviet Socialist Republic
The countries or areas listed are those where il language is mainly W West

Name Where spoken Language family How many 000) Name Where spoken Language family How many (000)

Abe Ivory Coast Niger-Congo 20-25 Arapesh New Guinea Indo-Pacific 18

Abelam New Guinea Indo-Pacific 30 Araucanian Chile, Argentina Penutian 200-350
Abkhaz USSR (Abkhaz ASSR) Caucasian 80-100 (Araukan)
Achinese N Sumatra Austronesian lm-2m (Mapuche)
(Atjehnese) Ari Ethiopia Afro-Asiatic 32
Acholi Uganda, Sudan Nilo-Saharan 500-70 Armenian USSR (esp. Armenian Indo-European 4m-5m
Adhola Uganda, Kenva Nilo-Saharan 153 SSR), Turkey, Iran, (Armenian)
Adyghe Adygei) USSR (Adygei AR) Caucasian 100 Syria, Lebanon
Adyukru Ivory Coast Niger-Congo 20-30 Arumanian see Romanian
Afar Danakil Ethiopia, Djibouti Afro-Asiatic 500-700 Ashanti see Akan
Afrikaans South Africa, Indo-European 4m-5m Asmat New Guinea Indo-Pacific 34
Namibia Germanic) Aso Burma Sino-Tibetan 95
Afusare Nigeria Niger-Congo 30 Assamese India (Assam), Bhutan Indo-European 7m-llm
Agar Sudan, Ethiopia Nilo-Saharan 16 Indo-Aryan)
Aguacatec Guatemala Penutian 8-10 Asu EC Africa Niger-Congo 100
Ainu Japan (Hokkaido), Isolate 16 Atjehnese see Achinese
Sakhalin &c Kuril Is., Aushi EC Africa Niger-Congo 29
USSR Atayal Taiwan Austronesian 36
Akan (Twi) Ghana, Ivory Coast Niger-Congo 5m-8m Avamba see Ambo
(Fante) Avar USSR (Daghestan, Caucasian 300-400
(Ashanti) Azerbaijan SSR)
Akoli see Acholi Aymara Bolivia, Peru Andean-Equatorial 600- lm
Akye Ivory Coast Niger-Congo 50-60 Azande see Zande
Alaba see Kambata Azerbaijani (Azeri) USSR (Azerbaijan Altaic (Turkic) 7m- 12m
Albanian Albania, and Indo-European 3 m—4jm SSR), Iran, Afghanistan
surrounding countries (Albanian) Aztec see Nahua(tl)
Altai USSR (Altai region) Altaic Turkic) 50-60 Baakpe NWC Africa Niger-Congo 16
Alur Uganda, Zaire, Sudan Nilo-Saharan 200-600 Bachama Nigeria Afro-Asiatic 12
Amba NEC Africa Niger-Congo 48 Badaga S India (Tamil Nadu) Dravidian 67-85
Ambo (Ovambo) Angola, Namibia Niger-Congo 225-700 Bade Nigeria Afro-Asiatic 32
(Avamba) Bafang Cameroon Niger-Congo 68
Amhanc Ethiopia Afro-Asiatic 8m-13m Bafou Cameroon Niger-Congo 41
Amhannyai Bagirmi Chad Nilo-Saharan 30-40
Amuesha Peru Andean-Equatorial 4-9 Bagobo Philippine Is. Austronesian 21
Amuzgo Mexico Oto-Manguean 8-12 Bahasa Indonesian see Indonesian
Andoni Nigeria Niger-Congo 50 Bahnar S Vietnam Austro-Asiatic 85
Angami Burma Sino-Tibetan 35-40 (Mon-Khmer)
Angas Nigeria Afro-Asiatic 55-280 Bai see Pai
Ankwe Nigeria Afro-Asiatic 14 Bakue Liberia, Ivory Niger-Congo 16
Anuak Sudan Nilo-Saharan 56 Coast
Anyi Ghana, Ivory Coast Niger-Congo 450 Balanta Guinea-Bissau, Senegal Niger-Congo 167
Ao' Burma Sino-Tibetan 15-30 Balantak S Sulawesi (Celebes) Austronesian 125
Arabic North Africa, Middle Afro-Asiatic 120m-150m(175m) Bali NEC Africa Niger-Congo 38
East, Arabian Peninsul a Balinese Bali Austronesian 2m-3m
Aramaic (Modern Iran, Iraq, USSR, Afro-Asiatic 50-100 Balkar see Karachay
Middle East Baluchi (Balochi Pakistan, Afghanistan, Indo-European Hm-3m

436 • APPENDICES
Name Where spoken Language family How many (000) Name Where spoken Language family How many (000)

Iran, Arabian Peninsul a (Iranian) Budu NEC Africa Niger-Congo 83

Bambara Mali, Senegal, Burkina Niger-Congo lm-2m Buduma Chad Afro-Asiatic 45
Faso, Ivory Coast Buginese S Sulawesi (Celebes) Austronesian 2}m—4m
Bamileke Cameroon Niger-Congo lm-2m Bugombe NEC Africa Niger-Congo 12
Bamoun Cameroon Niger-Congo 75 Buja NWC Africa Niger-Congo 100
Banda WC Africa Niger-Congo 350 Bukidnon Philippine Is. Austronesian 41
Banen NWC Africa Niger-Congo 27 Bulgarian Bulgaria and surround - Indo-European 8m-8im
Bangante Cameroon Niger-Congo 67 ing areas (Balto-Slavic)
Bangba WC Africa Niger-Congo 700 Bnli Ghana, Burkina Faso Niger-Congo 63
Banggai S Sulawesi (Celebes) Austronesian 200 Bulu NWC Africa Niger-Congo 170
Banna Ethiopia Afro-Asiatic 23 Bunak Timor Is. Indo-Pacific 50
Banoni Solomon Is. Austronesian 12* Bungku C Sulawesi (Celebes) Austronesian 180
Banyun Guinea-Bissau, Senegal Niger-Congo 18 Bunun Taiwan Austronesian 21
Barambu Sudan, Congo Niger-Congo 46 Buol N Sulawesi, Indonesia Austronesian 150
Bare'e C Sulawesi (Celebes) Austronesian 325-350 Burji Ethiopia Afro-Asiatic 15
Bargu Benin, Togo, Nigeria Niger-Congo 240 Burmese Burma Sino-Tibetan 20m-27m
Ban Sudan, Uganda, Zaire Nilo-Saharan 250-500 Bum Nigeria, Cameroon Afro-Asiatic 100-500
Basa NWC Africa Niger-Congo 150-170 Burun Sudan Nilo-Saharan 18
Basari Guinea, Senegal, Niger-Congo 12 Buningi Tanzania Afro-Asiatic 11
Gambia Burushaski India (NW Kashmir), Isolate 20-27
Basherawa Nigeria Niger-Congo 20 Pakistan
Bashkir Bashkir ASSR Altaic (Turkic) 821-lim Buryat USSR (Buryat ASSR) Altaic (Mongolian) 290-300
Basque SW France, NW Isolate 500-700 Busa Benin, Nigeria Niger-Congo 30
Spain, USA Bushman see San
Batak see Toba Busoong NWC Africa Niger-Congo 30
Bats Georgian SSR (Zemo- Caucasian 2*-3 Bute Cameroon Niger-Congo 16
Alvani village) Butung SE Sulawesi (Celebes) Austronesian 200
Batta Cameroon Afro-Asiatic 23 Buyi see Puyi
Beach-la-Mar see Bislama Bviri WC Africa Niger-Congo 16
Bebele NWC Africa Niger-Congo 24 Bwaka WCAfrica Niger-Congo 17
Beja Sudan, Ethiopia Afro-Asiatic 500-lm Byelorussian see Belorussian
Belorussian Belorussian SSR, Indo-European 9m-10m Caga NEC Africa Niger-Congo 237
(Byelorussian) Poland (Balto-Slavic) Cakchiquel Guatemala Penutian 225-500
Bemba EC Africa Niger-Congo 170-2im Cambodian see Khmer
Bena EC Africa Niger-Congo 158 Campa Peru Andean-Equatorial 25-33
Bengali Bangladesh, India Indo-European 80m-150m Carib Brazil, Suriname, Ge-Pano-Carib 5
(W Bengal, Tripura) (Indo-lranian) Guyana, French
Berba Benin Niger-Congo 44 Guinea
Berber see Kabyle Riff, Shluh, Tamashek
, Catalan NE Spain, France Indo-European 5m-7m
Berta Ethiopia, Sudan Nilo-Saharan 28 (Rousillon), Andorra (Romance)
Bete Ivory Coast Niger-Congo 250-lJm Cayapo NW Brazil Ge-Pano-Carib 10
Bhili India (Rajasthan, Indo-European 2m-3m Cebuano see Sebuano
Madhya Pradesh, (Indo-lranian) Cham Vietnam, Kampuchea Austronesian 150
Gujarat, Maharashtra Chamba Nigeria, Cameroon, Niger-Congo 500-lm
Bhojpuri India (Uttar Pradesh, Indo-European 8m—23m Chad
Bihar) (Indo-lranian) Chamorro Guam Austronesian 51
Bhotia see Dzongkha Chatino Mexico Oto-Manguean 18-20
Bhutani see Dzon gkha Chattis-garhi India (Madhya Indo-European 3m-6m
Bhutia (Murmi) India (Sikkim), Nepal Sino-Tibetan 30-40 Pradesh, Bihar) (Indo-lranian)
Biafada Guinea-Bissau Niger-Congo 12 Chechen Georgian and Caucasian 500-800
'Bidyo Chad Afro-Asiatic 14 Kazakh SSR
Bihari India, Nepal Indo-European 40m-65m Cheremis see Man
Biisa EC Africa Niger-Congo 42 Cherokee USA (N Carolina, Macro-Siouan 20-50
Bikol Philippine Is. Austronesian 2m-3im Oklahoma)
Bilaan Philippine Is. Austronesian 51 Chewa see Nyanja
Bilin Ethiopia Afro-Asiatic 32 Cheyenne USA (Montana, Macro-Algonquian 3^»
Bini (Edo) Nigeria Niger-Congo 300-2im Oklahoma)
Binji WC Africa Niger-Congo 64 Chimbu New Guinea Indo-Pacific 119
Bira NEC Africa Niger-Congo 32 Chinantec Mexico Oto-Manguean 25-30
Birifo(r) Ghana, Burkina Faso Niger-Congo 90 Chinese China, Taiwan, Sino-Tibetan 1,000m
Bisa Ghana, Burkina Faso Niger-Congo 127 SE Asia, USA
Bisaya see Sebuano Chingpaw Burma, S China Sino-Tibetan 170-200
Bislama (Beach-la - Vanuatu, Solomon Is. English-based 100 Chipewyan USA (Alaska) Na-Dene 4-7
Mar) pidgin/creole NW Canada
Bitare Nigeria Niger-Congo 50 Choctaw USA (Oklahoma) Macro-Algonquian 10-13
Blackfoot Canada (Alberta, Macro-Algonquian 5-6 Chokwe see Ciokwe
Saskatchewan), USA Choi Mexico Penutian 30-33
(Montana) Chontal Mexico Penutian 15-20
Bobo Mali, Burkina Faso Niger-Congo 390 Chord Guatemala, Honduras Penutian 33
Bodo NE India (Brahma- Sino-Tibetan 200-250 Charau S Vietnam Austro-Asiatic 15
putra Valley) (Mon-Khmer)
Boiken New Guinea Indo-Pacific 17 Chuave New Guinea Indo-Pacific 21
Bolewa Nigeria Afro-Asiatic 50-200 Chuj Guatemala Penutian 13-14
Bomougoun- Cameroon Niger-Congo 25 Chukchi USSR (Chukchi Palaeosibehan 11-15
Bamenjou Peninsula) (but see p. 306)
Bondei EC Africa Niger-Congo 28 Chung-chia see Puyi
Bontok Philippine Is. Austronesian 21-32 Chuvash USSR (Chuvash region ) Altaic (?Turkic) l|m-2m
Brahui Pakistan (Baluchistan, Dravidian 300-400 Ciga NEC Africa Niger-Congo 272
Sind) Cinga NWC Africa Niger-Congo 13
Breton France (Brittany) Indo-European 600-lm Ciokwe (Chokwe) SWC Africa Niger-Congo 600-lJm
(Celtic) Circassian see Kabardian
Bra N and S Vietnam, Austro-Asiatic 40 Cocama see Kokama
Laos (Mon-Khmer) Coorg see Kodagu

III TABLE OF THE WORLD'S LANGUAGES -

437
 m

Name Where spoken Language family How many (000) Name Where spoken Language family How many (000)

Cree Canada, USA Macro-Algonquian 35-62 Luxemburg, Monaco,

(Montana) and many parts of Afr -

Croatian see Serbo-Croat(ian) ca, Oceania, and the

Cua S Vietnam Austro-Asiatic 10-15 Americas; widespread
Mon-Khmer) 2nd language use
Cuicatec Mexico Oto-Manguean 10-12 Frisian Netherlands, W Indo-European 250-350
Cuna Panama, Columbia Macro-Chibchan 20-21 Germany (Germanic)
Czech W Czechoslovakia Indo-European (Balto 9m-10m Friulian N Italy Indo-European 350-400
Slavic) (Romance)
Daba Cameroon Afro-Asiatic 13 Fulani (Ful) W Africa Niger-Congo 10m-15m
Dagara Ghana, Burkina Faso Niger-Congo 75 Fuliro NEC Africa Niger-Congo 56
Dagari Ghana, Burkina Faso Niger-Congo 200 Fur Sudan, Chad Nilo-Saharan 200-400
Dagbani Ghana, Togo Niger-Congo 409 Ga Ghana, Togo, Benin Niger-Congo 250-400
Daghur (Dagur) NW Manchuria Altaic (Mongolian) 50-100 Gadba EC India Dravidian 8-9
Daju Chad Nilo-Saharan 60 Gaddang Philippine Is. Austronesian 17-28
Daka Nigeria, Cameroon Niger-Congo 53 Gaelic see Irish Gaelic, Scottish Gaelic
Chad Gagauz USSR (Moldavian Altaic (Turkic) 150
Dakota Sioux) USA (N and S Dakota, Macro-Siouan 10-20 SSR), Bulgaria,
Montanaj, Canada Romania
(Manitoba) Gahuku New Guinea lndo-Pacific 23
Dan Liberia, Ivory Coast Niger-Congo 100 Galla see Oromo
Danakil see Afar Gambai Chad, Central Nilo-Saharan 235
Dangaleat Chad Afro-Asiatic 16 African Rep.
Danish Denmark, W Germany Indo-European 5m-5Jm Gan Burkina Faso Niger-Congo 51
USA Germanic Ganda (Luganda) NEC Africa Niger-Congo lim— 3im
Dargwa (Khjurkili) C Daghestan SSR) Caucasian 160-230 Garhwali India (Uttar Pradesh) Indo-European 800-Hm
Dasenech Ethiopia Afro-Asiatic 18 (Indo-lranian)
Derasa Ethiopia Afro-Asiatic 250 Garo NE India (Assam) Sino-Tibetan 300-350
Digo NEC Africa Niger-Congo 32 Gayo N Sumatra Austronesian 120
Dinka Sudan, Ethiopia Nilo-Saharan 500-2m Gbari Nigeria Niger-Congo 155
Diola Guinea-Bissau, Niger-Congo 200 Gbaya WC Africa Niger-Congo 300
Gambia Georgian Georgian SSR, and a Caucasian 3m—4m
Dogon Mali, Burkina Faso Niger-Congo 150 (Gruzinian) few surrounding areas
Duala NWC Africa Niger-Congo 500-lm German E 8c W Germany, Indo-European 95m-100m
Dukawa Nigeria Niger-Congo 19 Austria, Switzerland, (Germanic) (100m)
Duna New Guinea lndo-Pacific 14 other pans of E Eur-
Duru Cameroon Niger-Congo 32 ope, USA, S Africa,
Dutch see Netherlands Latin America
Dye Togo Niger-Congo 17 Gidder Cameroon, Chad Nilo-Saharan 38
Dyimini see Tagbana-Dyimini Gilbertese Kiribati and surround- Austronesian 44-60
Dyula Ivory Coast, Burkina Niger-Congo Um (Kiribati) ing areas
Faso, Ghana Gilyak EUSSR(R. Amur, Palaeosiberian (but see 3—4
Dzindza NEC Africa Niger-Congo 67 Sakhalin I.) p. 306)
Dzongkha Bhutan Tibeto-Burman 800 Gimi New Guinea lndo-Pacific 17
(Bhutani, Bhotii )
Gipsy see Romany
Edo see Bini Gisiga Cameroon Afro-Asiatic 18
Efik Nigeria Niger-Congo 2m—4 m Goajiro Colombia, Venezuela Andean-Equatorial 40-45
Ekoi Cameroon, Niger-Congo 90 Gogo EC Africa Niger-Congo 270
Nigeria Gola Liberia, Sierra Leone Niger-Congo 150
Erabo NEC Africa Niger-Congo 200 Gondi India Madhva Dravidian lim-2m
Enga New Guinea lndo-Pacific 119-130 Pradesh)
Engenni Nigeria Niger-Congo 12 Gorontalo N Sulawesi, Indonesia Austronesian 900
English USA, Canada, UK, Aus-Indo-European 300m-350m (700m- Grebo Liberia, Ivory Coast Niger-Congo 50
tralia, New Zealand, (Germanic) 1,400m) Greek Greece, Cyprus, Indo-European lOm-lOim
pans of Asia and Turkey, and (Greek)
Africa, and widespreac surrounding areas
2nd language use Greenlandic see In jit
Epera Columbia, Panama Macro-Chibchan 20 Grishun see Romansch
Epie Nigeria Niger-Congo 26 Gruzinian see Georgian
Eskimo see Inuit Guahibo Colombia, Venezuela Andean-Equatorial 15-45
Estonian Estonian SSR Uralic (Finno-Ugric) lm Guan(g) Ghana Niger-Congo 53
Eton NWC Africa Niger-Congo 112 Guarani see Tupi
Even see Lamut Guaymi Panama Macro-Chibchan 10
Evenki Siberia, NE China Altaic(Manchu- 15-30 Gujarati India (Gujarat, Indo-European 20m-35m
Tungus) Maharashtra, and (Indo-lranian)
Ewe Ghana, Togo, Benin Niger-Congo 2m-3m surrounding areas),
Ewondo NWC Africa Niger-Congo 93 Pakistan
Faeroese (Faroese) Faeroe Is. Indo-European 35-44 Gumuz Ethiopia, Sudan Nilo-Saharan 53
(Germanic) Garage Ethiopia Afro-Asiatic 500-lm
Fall Cameroon Niger-Congo 50 Gurkhali see Nepa li

Fang NWC Africa Niger-Congo 200-2m Gonna Togo, Burkina Faso Niger-Congo 200- lm
Fante see Akan Gusii (Kisii) NEC Africa Niger-Congo 255-1
Faroese see Faeroese Gutob India (Andhra Pradesh Austro-Asiatic 30-60
Farsi see Persian Orissa (Munda)
Fiba EC Africa Niger-Congo 78 Gwandara Nigeria, Niger Afro-Asiatic 15
Fijian Fi|i Is. Austronesian 200-300 Gwere NEC Africa Niger-Congo 162
Finnish Finland, Sweden, Uralic (Finno-Ugric) 4im-5m Gypsv see Romany
USSR Ha NEC Africa Niger-Congo 286
Flemish see Netherlandic Had(iy)ya Ethiopia Afro-Asiatic 100-700
Fon Benin Niger-Congo lm—2 m Haitian Creole) Haiti Romance (French 4im-5m
Frafra Ghana Niger-Congo 138 Haka Burma Sino-Tibetan 85
French France, Canada, Bel- Indo-European 60m-70m Haling S Vietnam Austro-Asiatic 10
gium, Switzerland, 'Romance (220m) (Mon-Khmer)

438 • APPENDICES
Name Where spoken Language family How many (000) Name Where spoken Language family How many (000)

Hangaza NEC Africa Niger-Congo 54 Kaje Nigeria Niger-Congo 25

Hanunoo Philippine Is. Austronesian 6 Kako NWC Africa Niger-Congo 37
Hausa Nigeria, Niger, and Afro-Asiatic 15 m— 25 m Kaliko Sudan, Zaire Nilo-Saharan 18
adjacent areas (30m-40m) Kalinga Philippine Is. Austronesian 41
Havu NEC Africa Niger-Congo 50 Kalmyk (Kalmuk) USSR (Kalmyk region) Altaic (Mongolian) 135-150
Havunese Sawu Is., Raidjua Is. Austronesian 40 Kam SE China Kam-Sui (?Tai) 700-800
Hebrew- Israel, USA, Europe Afro-Asiatic 3m—4m Kamano New Guinea Indo-Pacific 34
Hebe EC Africa Niger-Congo 192 Kamba (Kikamba) NEC Africa Niger-Congo 600-2m
Herero SWC Africa Niger-Congo 25 Kambari Nigeria Niger-Congo 67
Higi Cameroon Afro-Asiatic 15 Kambata (Alaba) Ethiopia Afro-Asiatic 300
Hindi N and C India, Africa, Indo-European 130m-200m Kanakum Nigeria Afro-Asiatic 11
Fiji, Surinam, Guyanas (lndo-lranian) (300m-700m) Kanarese see Kannada
Hlota India (Assam) Sino-Tibetan 18-22 Kanjobal Guatemala Penutian 27-40
Hmong see Miao Kankanay Philippine Is. Austronesian 91
Hopi NE Arizona Aztec-Tanoan 5 Kannada India (Mysore) Dravidian 18m-25m
Hre S Vietnam Austro-Asiaric 80 (Kanarese)
(Mon-Khmer) Kanuri Nigeria, Niger Nilo-Saharan 2m—4m
Huastec Mexico Penutian 40-60 Kaonde WC Africa Niger-Congo 38
Huave Mexico Penutian 7 Karaboro Burkina Faso, Niger-Congo 25
Huela Ivorv Coast Niger-Congo 50 Ivorv* Coast
Huku NEC Africa Niger-Congo 13 Karachay (Balkar) USSR (NW Caucasus) Altaic (Turkic) 120-150
Hub New Guinea Indo-Pacific 54 Karakalpak Karakalpak ASSR Altaic (Turkic) 228-300
Hunde NEC Africa Niger-Congo 34 Afghanistan
Hungarian Hungary, Romania Uralic (Finno-Ugric) 13im-14m Karamojong Uganda, Kenya Nilo-Saharan 100
(Magyar) and nearbv areas Karelian Karelskaya ASSR Uralic (Finno-Ugric) 100-120
Hungu WC Africa Niger-Congo 66 Kari Chad, Cameroon, Niger-Congo 40
Ibanag Philippine Is. Austronesian 150-300 C African Rep.
Ibo see Igbo Kasem Ghana, Burkina Faso Niger-Congo 74
Icelandic Iceland, USA Indo-European 230-250 Kashmiri India and Pakistan Indo-European (Indo- 2m-3m
(Germanic) (Kashmir) Iranian)
Idoma Nigeria Niger-Congo 118 Kasimbar N Sulawesi (Celebes) Austronesian 100
Ifugao Philippine Is. Austronesian 70-80 Katab Nigeria Niger-Congo 23
Igala Nigeria, Benin Niger-Congo 100 Karu S Vietnam, Laos Austro-Asiatic 20-30
Igbirra Nigeria Niger-Congo 150 (Mon-Khmer)
Igbo (Ibo) Nigeria Niger-Congo 6m— 13m Kazakh (Kazak) Kazakh SSR, China, Altaic (Turkic) 5m-7m
Ijo (Ijaw) Nigeria Niger-Congo lm-ljm Mongolian PR,
Da EC Africa Niger-Congo 130 Afghanistan
Ilocano (Iloko) Philippine Is. Austronesian 3m-6m Kebu Togo Niger-Congo 17
Ilonggo Philippine Is. Austronesian 2m-3m Keigana New Guinea Indo-Pacific 27
Indonesian Indonesia Austronesian 30m Kekchi(Quekchi) Guatemala Penutian 250-400
Ingush Kazakh SSR Caucasian 100-150 Kela NWC Africa Niger-Congo 80-100
Inuit (Inupiac, Greenland, Eskimo- Aleut 65-90 Kele NWC Africa Niger-Congo 26
Inukritut, Canada, USA Kenga Chad Nilo-Saharan 20-25
Greenlandic, (Alaska) Kera Cameroon Niger-Congo 15
Eskimo) Kerebe NEC Africa Niger-Congo 31
Iraqw Tanzania Afro-Asiatic 111 Kerintji W Sumatra Austronesian 170
Iraya Philippine Is. Austronesian 6-8 Ket USSR (Yenisey R.) Paleosiberian (but see 1

Irish Gaelic Ireland (esp. west) Indo-European 30-100 p. 306)

(Celtic) Kewa New Guinea Indo-Pacific 47
Ishan Nigeria Niger-Congo 90 Khakas Khakas ASSR Altaic (Turkic) 56
Island Carib Honduras, Belize, Andean-Equatorial 30 Khalka (Mongol) Mongolian PR, China Altaic (Mongolian) 700
Guatemala (Inner Mongolian (2Jm-4m)
Isoko see Urhobo Region)
Italian Italy, Switzerland, Indo-European 56m—60m Khandeshi India (Maharashtra Indo-European 500-lm
San Marino, Vatican (Romance) (60m) and surrounding (lndo-lranian)
City, Sardinia, Yugo- areas)
slavia, parts of N and S Khanty (Osryak) Khanti-Mansi region, Uralic (Finno-Ugric) 14J-15
America, N Africa USSR
Ixil Guatemala Penutian 20-25 Kharia India (Bihar and Austro-Asiatic 110-180
Iyala Nigeria Niger-Congo 23 surrounding areas) (Munda)
Jacaltec Guatemala Penutian 12 Khasi India (Assam) Austro-Asiatic 200-500
Japanese Japan, Brazil, USA Isolate (?Altaic) 120m Khasonke Mali Niger-Congo 53-71
Jarai Vietnam, Kampuchea Austronesian 150 Khjurkili see Dargwa
Javanese Java, Malaysia Austronesian 45m— 65m Khmer (Cambo- Kampuchea, Vietnam, Austro-Asiatic 5m-8m
Jeh S Vietnam Austro-Asiatic 10 dian Thailand (Mon-Khmer)
(Mon-Khmer) Khmu' Laos, Thailand Austro-Asiatic 100
Jita NEC Africa Niger-Congo 97 (Mon-Khmer)
Jivaro Ecuador, Peru Andean-Equatorial 20 Khorcin China Altaic 900
Jongor Chad Afro-Asiatic 14-16 Kibet Sudan, Chad Nilo-Saharan 16-22
Juang India (Orissa) Austro-Asiatic 13-16 Kikamba see Kamba
(Munda) Kikuvu NEC Africa Niger-Congo lm^m
Judaeo-German see Yiddish Kilba Nigeria Afro-Asiatic 23
Jui see Puyi Kinga EC Africa Niger-Congo 57
Jukun Nigeria,, Cameroon Niger-Congo 32-37 Kirghiz (Kirgiz) Kirghiz SSR, and Altaic (Turkic) lim-2m
Kaba Chad, C African Rep. Nilo-Saharan 29 surrounding areas)
Kabardian USSR (Kabardino- Caucasian 280-350 Kiribati see Gilbertese
(Circassian) Balkar and Karachay- Kin Liberia, Sierra Leone, Niger-Congo 250
Cherkess regions) Guinea
Kabre Togo, Benin Niger-Congo 157 Kisii see Gusii
Kabyle Algeria Afro-Asiatic lm-2m Kodagu (Coorg) SW India (Mysore) Dravidian 80
Kadara Nigeria Niger-Congo 18 Ko'ho S Vietnam Austro-Asiatic 100
Kafa Ethiopia, Kenya Afro-Asiatic 170 (Mon-Khmer)
Kagulu EC Africa Niger-Congo 60

III TABLE OF THE WORLD'S LANGUAGES- 439

 m

Name Where spoken Language family How many (000) Name Where spoken Language family How many (000)

Kokama (Cocama) Peru, Colombia, Brazil Andean-Equatorial 10 Latvian Latvian SSR Indo-European ljm-2m
Kolami India (Bombay, Dravidian 50 (Balto-Slavic)
Andhra Pradesh) Lefana Togo Niger-Congo 15
Komi (Zyryan) Komi ASSR and Uralic (Finno-Ugric) 375-400 Lega NEC Africa Niger-Congo 33
surrounding area Lele NWC Africa Niger-Congo 26
Konda India (Orissa) Dravidian 13 Lendu Uganda, Zaire Nilo-Saharan 110-250
Konde see Makonde Lenje EC Africa Niger-Congo 33
Kongo WC Africa Niger-Congo 5m-7m Letzebuergesch see Luxemburgish
Konkani India (Goa, and Indo-European lim-2m Lezghian Daghestan, Georgian Caucasian 165-350
surrounding areas) (Indo-Iranian) (Lezgi, Kuri) and Azerbaijan SSR
Konkomba Ghana, Togo Niger-Congo 130 U China (Hainan I.) Kadai(?Tai) lm-lim
Kono Liberia, Mali, Sierra Niger-Congo 112 Ligbi Ivory Coast Niger-Congo 50
Leone Liko NEC Africa Niger-Congo 26
Konso Ethiopia Afro-Asiatic 60 Limba Guinea, Sierra Leone Niger-Congo 174
Konyagi Guinea Niger-Congo 85 Lingala see Losengo
Konzo NEC Africa Niger-Congo 162 Lisu China (Yunnan, Sino-Tibetan 250
Korean N and S Korea, Japan, Isolate (?Altaic) 50m-60m Sichuan)
China, USSR (60m) Lithuanian Lithuanian SSR Indo-European 2im-3m
Koria NEC Africa Niger-Congo 94 (Balto-Slavic)
Koryak NEUSSR Palaeosiberian (but see 6-8 Lobi(ri) Burkina Faso, Niger-Congo 211
p. 306) Ivory Coast
Kosali India (Madhya and Indo-European (Indo- 20m Logo Sudan, Zaire Nilo-Saharan 54-60
Urtar Pradesh) Iranian) Loinang C Sulawesi (Celebes) Austronesian 100
Kota S India Kotagin Dravidian lm Loko Sierra Leone, Guinea Niger-Congo 76
Kota NWC Africa Niger-Congo 28 Loma Liberia, Guinea Niger-Congo 260
Kotoko Chad Afro-Asiatic 30-50 Lombo NWC Africa Niger-Congo 11
Koya India (Andhra Dravidian 160-170 Lore C Sulawesi (Celebes) Austronesian 100-140
Pradesh: Losengo (Lingala) NWC Africa Niger-Congo lm-ljm
Kpa NWC Africa Niger-Congo 15 Lotuho Sudan Nilo-Saharan 66
Kpelle Liberia, Guinea Niger-Congo 250-500 Lozi SWC Africa Niger-Congo 70
Kposo Togo Niger-Congo 17 Luba -Lulua WC Africa Niger-Congo 3im-5m
Krahn Liberia, Ivory Coast Niger-Congo 100 Lubu E Sumatra Austronesian lm
Krio Sierra Leone English-based Creole 250 Lucazi SWC Africa Niger-Congo 60
Kru-Krawi (Kru) Liberia, Ivory Coast Niger-Congo 100-125 Luganda see Ganda
Kua see Makua Lugbara Uganda, Zaire Nilo-Saharan 350
Kui India (Orissa) Dravidian 500 Luhya NEC Africa Niger-Congo 650-3
Kukele Nigeria Niger-Congo 20 Lumbu NWC Africa Niger-Congo 12
Kulango Ivory Coast Niger-Congo 47 Lunda WC Africa Niger-Congo 82
Kulele Ivory Coast Niger-Congo 15 Lundu NWC Africa Niger-Congo 24
Kumam Uganda Nilo-Saharan 96 Luo Kenya, Tanzania Nilo-Saharan 800-2Jm
Kumauni India (Uttar Pradesh) Indo-European lm Lusatian see Sorbian
(Indo-Iranian) Lushai Burma Sino-Tibetan 70
Kumyk Daghestan ASSR Altaic Turkic) 186 Luwu' S Sulawesi (Celebes) Austronesian 500
Kunama Ethiopia Nilo-Saharan 41 Luwunda WC Africa Niger-Congo 20
Kunda EC Africa Niger-Congo 73 Luxemburgish Luxemburg Indo-European 300-400
Kurdish Turkey, Iraq, Indo-European (Indo- 5m-10m (Letzebuergesch) (Germanic)
Iran, Syria, USSR Iranian) Lwalu WC Africa Niger-Congo 21
Kuri see Lezghian Lwena SWC Africa Niger-Congo 90
Kurku India (Madhya Austro-Asiatic 175-190 Lwo Uganda Nilo-Saharan 20
Pradesh) (Munda) Lyele Burkina Faso Niger-Congo 61
Kurukh(i) NE India Dravidian lm-Um Ma(a)sai Tanzania, Kenya Nilo-Saharan 190-400
(Oraon(i)) Mabiha SE Africa Niger-Congo 70
Kurumba Burkina Faso Niger-Congo 86 Mabuyag Torres Strait Is. Australian 2-4
Kusal(e) Ghana, Burkina Faso Niger-Congo 122 Macedonian Yugoslavia, Bulgaria, Indo-European lm-Um
Kusu NWC Africa Niger-Congo 26 Greece, Albania (Balto-Slavic)
Kuru EC Africa Niger-Congo 15 Machiguenga Peru Andean-Equatorial 6-10
Kuyu see Kikuyu Mada Nigeria Niger-Congo 24
Kwaa Liberia Niger-Congo 25 Madi Uganda, Sudan Nilo-Saharan 114
Kwadi Angola Khoisan 15 Madurese Madura, NE Java Austronesian 5m-10m
Kwangwa SWC Africa Niger-Congo 27 Magahi India (Bihar and Indo-European 3m-10m
Kweni Ivory Coast Niger-Congo 210 surrounding areas) (Indo-Iranian)
Kyama Ivory Coast Niger-Congo 12 Magyar see Hungarian
La din Italy (S Tyrol) Indo-European 12-20 Maithili India (Bihar), Nepal Indo-European 6m
(Romance) (Indo-Iranian)
Lahnda India, Pakistan Indo-European 15m-20m Maji Ethiopia Afro-Asiatic 18
(Punjab) (Indo-Iranian) Maka NWC Africa Niger-Congo 52
Lak(k) (Lakh) C Daghestan SSR Caucasian 64-80 Makonde (Konde) SE Africa Niger-Congo 281-lm
Laid SE Sulawesi (Celebes) Austronesian 125 Makua (Kua) SE Africa Niger-Congo 3m-6m
Lala EC Africa Niger-Congo 64 Malagasy Madagasar Austronesian 9m
Lalia NWC Africa Niger-Congo 30 Malay Malaysia, Singapore, Austronesian 10m
Lamba Benin Niger-Congo (Gur) 29 Brunei, Thailand, (100m-160m)
Lamba EC Africa Niger-Congo (Bantu) 70-80 Sumatra, Borneo, Java
Lamnso Cameroon Niger-Congo 22 and surrounding areas
Lampung S Sumatra Austronesian 200 Malayalam India (Kerala, Dravidian 17m-25m
Lamut (Even) E Siberia Altaic(Manchu- 7-12 Laccadive Is.)

Tungus) Maldivian Maldive Islands Indo-European 100-155

Langi NEC Africa Niger-Congo 95 (Indo-Iranian)
Lango Uganda Nilo-Saharan 534-900 Male Ethiopia Afro-Asiatic 12
Languedoc see Occitan Malila EC Africa Niger-Congo 18
Lao (Laotian) Thailand, Laos Tai 9m-10m Malinka (Malinke) W Africa Niger-Congo Hm-3m
Lapp(ish) (Lappic Norway, Sweden, Uralic (Finno-Ugric) 30-35 Maltese Malta Afro-Asiatic 300-350
Finland, USSR Malto India (NE Bihar) Dravidian 90
Malvi India (Rajasthan and Indo-European lm

440 APPENDICES
Name Where spoken Language family How many (000) Name Where spoken Language family How many (000)

surrounding areas (Indo-Iranian) Mongol see Khalka

Mambila Cameroon, Nigeria Niger-Congo 16 Mongo-Nkun-du NWC Africa Niger-Congo 200-250
Mambwe EC Africa Niger-Congo 15 Monguor China (Gansu, Altaic (Mongolian) 50-60
Mam(e) Guatemala, Mexico Penutian 285-350 Qinghai)
Mamuu-Efe Uganda, Zaire Nilo-Saharan 60 Moni New Guinea Indo-Pacific 15-20
Man see Yao Monjambo WC Africa Niger-Congo 13
Manchu China (Xinjiang, Altaic (Manchu- 15-20 Mordvin(ian) Mordvin ASSR, and Uralic (Finno-Ugric) lm
Heilongjiang) Tungus) surrounding area
Mandara Cameroon, Nigeria Afro-Asiatic 30-40 More see Mossi
Mangbetu Uganda, Zaire Nilo-Saharan 100-500 Mom Sudan Nilo-Saharan 23
Mangisa NWC Africa Niger-Congo 14 Mossi (More) Ghana, Burkina Faso Niger-Congo 2m—6m
Manipuri see Meithei Move Panama Macro-Chibchan 15
Manjaku Senegal, Portuguese Niger-Congo 84 Mpoto EC Africa Niger-Congo 58
Guinea Mubi Chad Afro-Asiatic 23
Mano Liberia, Guinea Niger-Congo 45-150 Mumuye Cameroon, Nigeria Niger-Congo 103
Mansi (Vogul) USSR (Khanti-Mansi Uralic (Finno-Ugric) 4-4i Mundang Chad, Cameroon, Niger-Congo 45
region) C African Rep.
Maori New Zealand Austronesian 100 Mundari India (Madyha Austro-Asiatic lim-ljm
Mapuche see Araucanian Pradesh, Orissa, Bihar) (Munda)
Marant Sudan, Chad Nilo-Saharan 42 Mu'o'ng N Vietnam Austro-Asiatic 390
Marathi India (Maharashtra, Indo-European (Indo- 30m-50m (Mon-Khmer)
and surrounding areas ) Iranian) Murle Ethiopia Nilo-Saharan 40
Margi Nigeria Afro-Asiatic 151