Вы находитесь на странице: 1из 364



This volume follows on from The Lesser-Known Varieties of English

(Cambridge University Press, 2010) by documenting a further range
of varieties that have been overlooked and understudied. It explores
varieties spoken by small groups of people in remote regions as diverse
as Malta, the Netherlands Antilles, Brazil, the Cook Islands and Palau.
The varieties explored are as much a part of the big picture as major
varieties and it is the intention of this collection to spark further
interest in the sociolinguistic documentation of minority Englishes in
a postcolonial world. Language endangerment is a very real factor for
the vast majority of lesser-known varieties of English, and this book
aims to highlight that documentation and archiving are key initial
steps in revitalization and reclamation efforts.
This book will be of interest to historians of English, and scholars
in dialectology, language birth and death, language contact, typology,
and variation and change.
jeffrey p. williams is Professor of Ethnology and Linguistics at
Texas Tech University. He previously taught at the University of
Sydney and Cleveland State University. Most recently he edited
The Aesthetics of Grammar: Sound and Meaning in the Languages of
Mainland Southeast Asia (Cambridge University Press, 2013).
edgar w. schneider is Professor and Chair of English Linguistics
at the University of Regensburg. He has published and lectured on all
continents on topics in the dialectology, sociolinguistics, history, and
semantics of English and its varieties. He edited the scholarly journal
English World-Wide for many years and has written and edited about
twenty books, including Handbook of Varieties of English (2004/2008),
Postcolonial English (Cambridge University Press, 2007) and English
around the World (Cambridge University Press, 2011).
peter trudgill is Adjunct Professor of Sociolinguistics at the University of Agder. He has carried out research on dialects in English,
Norwegian, Greek, Albanian and Spanish and has written and edited
more than thirty books on sociolinguistics and dialectology including
Sociolinguistic Variation and Change (2002), A Glossary of Sociolinguistics (2003) and New-Dialect Formation (2004).
daniel schreier is Associate Professor of English Linguistics at the
University of Zurich. He has taught and lectured in New Zealand,
Germany and the USA. His previous publications include Isolation
and Language Change (2003), Consonant Change in English Worldwide
(2005) and St Helenian English (2008).

studies in english language

General editor
Merja Kyto (Uppsala University)
Editorial Board
Bas Aarts (University College London)
John Algeo (University of Georgia)
Susan Fitzmaurice (University of Sheffield)
Christian Mair (University of Freiburg)
Charles F. Meyer (University of Massachusetts)
The aim of this series is to provide a framework for original studies of English,
both present-day and past. All books are based securely on empirical research,
and represent theoretical and descriptive contributions to our knowledge of
national and international varieties of English, both written and spoken. The
series covers a broad range of topics and approaches, including syntax,
phonology, grammar, vocabulary, discourse, pragmatics, and sociolinguistics, and
is aimed at an international readership.
Already published in this series:
Daniel Schreier, Peter Trudgill, Edgar W. Schneider and Jeffrey P. Williams
(eds.): The Lesser-Known Varieties of English: An Introduction
Irma Taavitsainen and Paivi Pahta (eds.): Medical Writing in Early Modern English
Colette Moore: Quoting Speech in Early English
David Denison, Ricardo Bermudez-Otero, Chris McCully and Emma Moore
(eds.): Analysing Older English
Jim Feist: Premodifiers in English: Their Structure and Significance
Steven Jones, M. Lynne Murphy, Carita Paradis and Caroline Willners:
Antonyms in English: Construals, Constructions and Canonicity
Christiane Meierkord: Interactions across Englishes: Linguistic Choices in Local and
International Contact Situations
Haruko Momma: From Philology to English Studies: Language and Culture in the
Nineteenth Century
Raymond Hickey (ed.): Standards of English: Codified Varieties around the World
Benedikt Szmrecsanyi: Grammatical Variation in British English Dialects: A Study
in Corpus-Based Dialectometry
Daniel Schreier and Marianne Hundt (eds.): English as a Contact Language
Bas Aarts, Joanne Close, Geoffrey Leech and Sean Wallis (eds.): The Verb Phrase
in English: Investigating Recent Language Change with Corpora

Martin Hilpert: Constructional Change in English: Developments in Allomorphy,

Word Formation, and Syntax
Jakob R. E. Leimgruber: Singapore English: Structure, Variation and Usage
Christoph Ruhlemann: Narrative in English Conversation
Dagmar Deuber: English in the Caribbean: Variation, Style and Standards in
Jamaica and Trinidad
Jock Onn Wong: English in Singapore: A Cultural Analysis
Eva Berlage: Noun Phrase Complexity in English
Nicole Dehe: Parentheticals in Spoken English: The Syntax Prosody Relation
Anita Auer, Daniel Schreier and Richard J. Watts: Letter Writing and Language
Marianne Hundt: Late Modern English Syntax
Irma Taavitsainen, Merja Kyto, Claudia Claridge and Jeremy Smith:
Developments in English: Expanding Electronic Evidence
Arne Lohmann: English Co-ordinate Constructions: A Processing Perspective on
Constituent Order
John Flowerdew and Richard W. Forest: Signalling Nouns in English: A
Corpus-Based Discourse Approach
Jeffrey P. Williams, Edgar W. Schneider, Peter Trudgill and Daniel Schreier:
Further Studies in the Lesser-Known Varieties of English
Earlier titles not listed are also available

edited by
Texas Tech University

University of Regensburg

University of Agder

University of Zurich

University Printing House, Cambridge cb2 8bs, United Kingdom

Cambridge University Press is part of the University of Cambridge.
It furthers the Universitys mission by disseminating knowledge in the pursuit of
education, learning and research at the highest international levels of excellence.
Information on this title: www.cambridge.org/9781107021204

C Cambridge University Press 2015

This publication is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2015
A catalogue record for this publication is available from the British Library
isbn 978-1-107-02120-4 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
urls for external or third-party internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.


List of maps
List of tables
List of contributors

page ix

1 Introduction

Jeffrey P. Williams, Edgar W. Schneider, Peter Trudgill and

Daniel Schreier

part i europe
2 Maltese English


Manfred Krug

3 Gibraltar English


David Levey

4 Irish Traveller English


Maria Rieder

part ii the americas

5 American Indian English


Elizabeth L. Coggshall

6 Bequia English


James A. Walker and Miriam Meyerhoff

7 Saban English


Jeffrey P. Williams and Caroline Myrick

8 St. Eustatius English


Michael Aceto



9 The English of Gustavia, St. Barthelemy


Ken Decker

10 Anglo-Paraguayan English


Danae M. Perez-Inofuentes

11 Gullah West: Texas Afro-Seminole Creole


Ian Hancock

part iii asia and the pacific

12 Palmerston Island English


Rachel Hendery

13 Pasifika Englishes in New Zealand


Donna Starks, Andy Gibson and Allan Bell

14 Palauan English


David Britain and Kazuko Matsumoto





Maltas three inhabited islands

Malta in its wider geographical context
Island of Bequia


page 10


2.1 Question: How well do you speak . . . ?

page 16
8.1 Results of 1974 Census in St. Eustatius according to place
of birth
8.2 Pronouns in Statian English
9.1 Phonetic realizations of the lax vowels of Gustavia English
9.2 Phonetic realizations of the tense vowels and diphthongs of
Gustavia English
10.1 Anglo-Paraguayan English vowel system
11.1 Articles in Afro-Seminole Creole
11.2 Subject pronouns in Afro-Seminole Creole
11.3 Object pronouns in Afro-Seminole Creole
11.4 Possessive pronouns in Afro-Seminole Creole
11.5 Demonstratives in Afro-Seminole Creole
11.6 Non-derived adverbs in Afro-Seminole Creole
12.1 Minimal consonant inventory for Palmerston Island English
12.2 Pronominal forms in Palmerston Island English
12.3 Present-tense verb paradigm for Palmerston Island English
12.4 Standard English and Palmerston Island English verbal and
adjectival predicates
13.1 Consonants of Pasifika English
13.2 Pasifika vowels as compared to General NZE
14.1 Palauan English vowels


michael aceto is Professor of Linguistics in the Department of English

at East Carolina University. His published work on Caribbean Englishes
has mostly made use of primary data gathered in the field in Panama,
Barbuda, St Eustatius and Dominica. His future work aspires to bring
the discipline of linguistics in contact with the millennia of works and
thought by Buddhist scholars.
david britain is Professor of Modern English Linguistics at the University of Bern. His research interests embrace language variation and
change, varieties of English, dialect contact and attrition, and the
dialectologyhuman geography interface, with particular interest in
applying insights from social geographys Mobilities paradigm to social
dialectology. He is editor of Language in the British Isles (Cambridge
University Press, 2007), co-editor of Social Dialectology (with Jenny
Cheshire, 2003) and co-author of Linguistics: An Introduction, 2nd
edition (with Andrew Radford, Martin Atkinson, Harald Clahsen and
Andrew Spencer, 2009). He is also currently an associate editor of the
Journal of Sociolinguistics and is on the editorial board of the Journal of
Linguistic Geography and English World-Wide.
elizabeth l. coggshall is a doctoral candidate in the linguistics department at New York University; she has a Masters from North Carolina
State University. Her research has focused on two large questions: how
does language use help create ethnic identity and place identity, and
how do these two identities intersect linguistically? To this end, she has
worked extensively on the Lumbee and Eastern Cherokee of North Carolina, as well as with White, Black, Latino and Filipino communities
of New York City and Jersey City, New Jersey. Her past research used
prosodic rhythm in American Indian English to better understand the
relationship between ethnic change and language change, and her current research concentrates on vowel quality to understand place identity


List of contributors
in a relatively small urban community (Jersey City) in the shadow of a
much larger city (New York City).

ken decker is currently the International Coordinator for Language

Assessment with SIL International. He has worked for over twentyfive years in the field of Applied Linguistics. He has a Masters degree in
Sociolinguistics from the University of Texas at Arlington and is working
on a PhD from the University of Groningen. He has published on the
languages of Pakistan, Creole languages, Belize Kriol, language vitality,
orthography development and language assessment.
andy gibson of the Institute of Culture, Discourse and Communication
at Auckland University of Technology has research interests including
the sociophonetics of singing in popular music, and the portrayal of
Pasifika varieties of New Zealand English in performative contexts. He
is Editorial Associate of the Journal of Sociolinguistics and co-edited a
theme issue of the journal entitled The Sociolinguistics of Performance.
ian hancock is the Representative to the UN (ECO-SOC/NGO Category II) and to UNICEF for the Romani people, and was appointed by
President Bill Clinton to represent Roma on the US Holocaust Memorial Council in 1997. In the same year he was awarded the prestigious
Rafto Foundation Prize for Human Rights (Norway) and was recipient
of the Gamaliel Chair in Peace and Justice for 1998. He publishes and
lectures widely on Romani civil and human rights, and on the fate of
the Romani victims of the Holocaust. His publications include We Are
the Romani People, A Handbook of Vlax Romani, The Pariah Syndrome:
An Account of Gypsy Slavery and Persecution and International English
Usage. He is currently writing three new books: one on the construction
of identity, one on the linguistic and historical origins of the Romani
people, and one a grammar of the Maskogo Creole language spoken in
south Texas.
rachel hendery is the Senior Lecturer in Digital Humanities at the
University of Western Sydney. Her main research interest is the intersection of typology and historical linguistics, with a recent focus on
contact languages. Since 2009 she has been working on Palmerston
Island English in the Cook Islands, and has also carried out fieldwork in
Indonesia and East Timor. Previously she was involved in the AustKin
project: a large interdisciplinary research project on change in Australian
Aboriginal kinship terminology and systems.

List of contributors


danae m. perez-inofuentes graduated from the University of Zurich

with a Masters degree in Spanish and English Linguistics and Anthropology, and is now working on her PhD on language contact and shift in
Paraguay. She is currently employed as a research and teaching assistant
in the English Department. Her research interests include colonization
and language contact with a regional focus on the Americas.
manfred krug is Chair of English and Historical Linguistics at the
University of Bamberg, after previous appointments in Mannheim and
Freiburg, where he completed his PhD and postdoctoral research project.
His research centres around synchronic variation and diachronic change,
typically including a corpus-based perspective. He holds an MA in
Applied Linguistics from Exeter University and was a visiting professor
at Portland State University. He has written and co-edited two books
on English modal verbs and edited, with Julia Schluter, the handbook
Research Methods in Language Variation and Change (Cambridge University Press, 2013). Apart from Maltese English, his recent publications
focus on the Great Vowel Shift, auxiliary verbs, grammaticalization as
well as methodological aspects of designing and conducting interviews
and questionnaires.
davd levey is Associate Professor at the University of Cadiz and lectures in
Phonetics, Phonology and Sociolinguistics. His research interests include
accent variation, language contact and change, speech perception and
language testing and evaluation. As well as publishing books and articles
on various aspects of language acquisition, he has written extensively on
the language situation in Gibraltar and is the author of Language Change
and Variation in Gibraltar (2008).
kazuko matsumoto is Associate Professor of Sociolinguistics at the University of Tokyo. Working primarily within sociolinguistics, her research
interests include language variation and change, dialect contact and
new dialect formation, language death, language contact, bilingualism,
borrowing, heritage language maintenance, language ideologies, and
pragmatic-discourse markers. She is currently engaged in collaborative
research projects investigating Japanese dialect contact and obsolescence
both in the Republic of Palau in Micronesia and in Mexico, as well as
Palauan English as a newly emerging postcolonial variety. She is currently
on the editorial board of the Japanese Journal of Language in Society.
miriam meyerhoff is Professor of Linguistics at Victoria University of
Wellington. Her research has often focused on the role that language


List of contributors

variation plays in informing our understanding of the linguistic dynamics of language change in Creole speech communities and the social
indexicalities that language variation serves in them. She has done fieldwork in the Caribbean and the Pacific, and is currently working on
language variation in northern Vanuatu.
caroline myrick is currently a PhD student and teaching assistant
at North Carolina State University. She received a BA in Linguistics
and Communication Studies from the University of North Carolina at
Greensboro, and an MA in English (Linguistics) from North Carolina
State University. Her research interests include language variation and
change, island studies, and dialect awareness and education. She has
carried out sociolinguistic field research on the island of Saba, which
served as the basis for her MA thesis and various publications.
maria rieder is currently a PhD student at Trinity College Dublin, where
she is carrying out a sociolinguistic and ethnographic study of the Irish
Traveller language and culture. She has completed a BA in Linguistics
and Spanish and the First State Examination for Teachers in English
and History at the University of Regensburg. Her research interests
include the fields of contact linguistics and language change combined
with and explored through the approaches of linguistic anthropology,
ethnography of communication and folk linguistics.
edgar w. schneider is Full Professor of English Linguistics at the University of Regensburg, after previous appointments in Bamberg, Georgia
and Berlin. He has written and edited several books (including American Earlier Black English, 1989; Introduction to Quantitative Analysis of
Linguistic Survey Data, 1996; Focus on the USA, 1996; Englishes around
the World, 1997; Degrees of Restructuring in Creole Languages, 2000;
Handbook of Varieties of English, 2004; Postcolonial English, Cambridge
University Press, 2007; English around the World, Cambridge University
Press, 2011) and published widely on the dialectology, sociolinguistics,
history, semantics and varieties of English. For many years he was editor
of English World-Wide and its associated book series Varieties of English
around the World. He is President-Elect of the International Society for
the Linguistics of English (ISLE).
daniel schreier has taught in Switzerland, New Zealand, Germany
and the USA, and is Professor of English Linguistics at the University of
Zurich. He is the author of Isolation and Language Change (2003), Consonant Change in English Worldwide (2005), St Helenian English: Origins,

List of contributors


Evolution and Variation (2008) and English as a Contact Language (with

Marianne Hundt, Cambridge University Press, 2013). He has published
in the Journal of Sociolinguistics, Language Variation and Change, American Speech, English Language and Linguistics and is co-editor of English
donna starks is a Senior Lecturer in Applied Linguistics in the Faculty
of Education at La Trobe University, Australia. Her research into the
Pasifika languages has centred on the Niue language and its speakers.
Her recent interests have focused on creating a framework for a better
understanding of the relationship between language and identity within
language education and applied linguistics in Language Education and
Applied Linguistics: Bridging the Two Fields (2014).
peter trudgill is a theoretical dialectologist who is Professor of Sociolinguistics at Agder University; Emeritus Professor of English Linguistics at
Fribourg University; and Honorary Professor of Sociolinguistics at the
University of East Anglia. His most recent monograph is Sociolinguistic
Typology (2011). A collection of his essays Investigations in Sociohistorical
Linguistics appeared in 2010 and was published by Cambridge University
james a. walker is Associate Professor of Linguistics in the Department
of Languages, Literatures and Linguistics at York University (Toronto).
He has conducted research on variation and change in Sango (Central
African Republic), Swedish and Brazilian Portuguese, although most
of his research has involved varieties of English (African American,
Canadian and Caribbean). He is the author of Variation in Linguistic
Systems (2010), co-author of Bequia Talk (with Miriam Meyerhoff, 2013)
and editor of Aspect in Grammatical Variation (2010).
jeffrey p. williams is Professor of Ethnology and Linguistics at Texas
Tech University. He has conducted fieldwork throughout the West Indies
since 1982, and also in Papua New Guinea and Australia, and with
Montagnard refugees in the southern United States. He is the co-editor
of Contact Englishes of the Eastern Caribbean (with Michael Aceto, 2003),
The Lesser-Known Varieties of English (with Daniel Schreier, Peter Trudgill and Edgar Schneider, Cambridge University Press, 2010) and editor
of The Aesthetics of Grammar (Cambridge University Press, 2013).
allan bell is Professor of Language & Communication, and Director
of the Institute of Culture, Discourse & Communication, at Auckland


List of contributors

University of Technology, New Zealand. He has made pioneering contributions on media language and discourse, the theory of style (Audience Design) and New Zealand English. His research interests include
multilingualism in New Zealand, performance language, language and
identity, and social and linguistic aspects of the internet. He has led
major research projects on New Zealand English, language style, Pasifika languages, television violence, and the World Internet Project New
Zealand. He has published many papers in leading journals and edited
collections, as well as six books. His 2013 Guidebook to Sociolinguistics is
a comprehensive, research-based textbook road map to the field. He is
co-founder and Editor of the Journal of Sociolinguistics.

c h a p ter 1

Jeffrey P. Williams, Edgar W. Schneider, Peter Trudgill
and Daniel Schreier

With the appearance of the seminal piece on lesser-known varieties of

English by Trudgill in 2002, a research trajectory was charted that gathered together scholars involved in the documentation of overlooked and
understudied varieties of English, many of which were spoken by very
small groups of people in remote and isolated locales. This assorted group
of lessers gained prominence in the literature for their value in providing
insights into larger questions in linguistics and sociolinguistics, culminating in the first instance in an edited collection of documentary descriptions
and analyses (Schreier, Trudgill, Schneider and Williams 2010). We continue to expand our treatment of lesser-known varieties of English (LKVEs)
in this second volume of further documentary descriptions.
As we stated at the outset of the first introductory volume to the documentation and study of LKVEs, one fundamental problem has to do with
how to evaluate and demarcate the status of the term lesser-known. To aid
the reader, we once again reproduce the set of characteristics we outlined
in the previous volume here in order to set the template for the individual
contributions that follow.
Lesser-known Englishes:
1. are spoken as first languages and not as ESL or EFL varieties, often in
environments where bi- or multilingualism is restricted;
2. are identified as distinct varieties by their respective speech communities and other groups in their social environment;
3. are associated with stable communities or regions;
4. are typically spoken by minorities; they are usually delimitated (not
necessarily isolated but socially or regionally distinct) to small communities which are embedded into a larger (regional) population
5. were, many of them, originally transmitted by settler communities or
adopted by newly formed social communities that emerged early in
the colonial era, so that they substantially derive from British inputs;

williams, schneider, trudgill and schreier

6. were formed by processes of dialect and/or language contact (which

makes it impossible to ascribe them genetic status, e.g. creoles or koines,
see below);
7. frequently take the function as identity carriers by their respective
8. are very often endangered.
The last point deserves more discussion here than we were able to provide
in the first volume. The notion of endangered varieties of a seemingly
voraciously dominating language such as English may seem insincere to
some. Wolfram (2008) has written passionately about the Ocracoke Brogue
as being an endangered language that challenges the established canon of
linguistic endangerment:
As it turns out, our classification of the Ocracoke Brogue as an endangered
language variety has challenged the established canon of endangerment
in linguistics. After several invitations to speak at language endangerment
conferences early in our studies where we presented the case for labeling
the Ocracoke Brogue as an endangered language variety, we have now
been excluded from conferences and workshops on this topic, reflecting the
marginalisation of English dialects in terms of the language endangerment
canon. In fact, after one of my presentations at a national conference on
language endangerment, a colleague congratulated me on the presentation
only to follow up with the comment, Do you think anyone takes you
seriously when you argue that isolated dialects of American English should
be considered as endangered? I would like to challenge the exclusion of
dialects from the endangerment canon on several bases. Indeed, it seems
like the endangerment canon is based on some questionable assumptions
about the nature of language variation. (Wolfram 2008: 9)

Attitudes such as those that Wolfram is challenging here are pervasive in

the field. And while we might seem to be swimming against the tsunami,
we are in strong agreement with Wolframs statements; not just on ethical grounds, but also on the grounds of first-hand experience of working
with languages that are unquestionably, within the endangerment canon,
endangered. Williams has worked with endangered American Indian languages (Comanche, Cheyenne, Arapaho, Navajo) and has observed parallel
although not identical sociolinguistic conditions between these communities and those in isolated island communities in the anglophone West
Indies. Endangerment as a sociolinguistic condition cannot be determined
by only global considerations but must be evaluated according to local conditions and a nuanced sociohistorical understanding of language varieties.


As the first volumes production led us to envision a second volume, which

now appears, the production of this second volume leads us to consider
a more contentious project that chronicles the endangerment of English
varieties in a third volume.
In an attempt to broaden the scope of the lesser-known typology, we
have included more varieties that have a deeper chronology extending
back to the seventeenth century in some cases. We have also explored
the boundaries of the upper extent of genesis of new varieties, as the
contribution on Pasifika Englishes demonstrates. Expanding our domain
of documentation in this volume, we were both challenged and engaged
by the paucity of coverage of lesser-known varieties on the global stage.
Over half of the contributions to this volume are from the New World, not
surprisingly since English has a long and dominant colonization there. As
we stated in the first volume, our list of varieties there was not exclusive or
exhaustive. The expansive list we develop here is also equally unbalanced
in terms of thoroughness of documentation.
In developing and realizing this second collection on LKVEs, it was
essential to continue with the same frameworks of presentation and typological organization. The contributors were selected and solicited based
on their expertise in the relevant varieties, and were provided with a
basic framework for documentary presentation. We provided the same
basic prompts for (i) sociohistorical origins, (ii) sociodemographic data,
(iii) structural features (phonetics/phonology/morphology/syntax/lexicon)
and (iv) assessment of the future of the variety.
Finally, as far as grouping and classification are concerned, the fact that
the LKVEs display considerable heterogeneity in their social and contact
histories made it difficult (in fact, nearly impossible) to assign them to
separate categories on typological grounds.
Again, following suit from the first volume, we organized the contributors chapters geographically instead of opting for a new scheme based
on some sort of sociolinguistic-typological classification. We now briefly
introduce the contributions from Europe, the Americas and the Caribbean,
and the South Pacific.

In this volume we extend our focus from the British Isles, as in the first
volume, to the broader geopolitical landscape of insular Europe. The three
varieties that are described by Krug, Rieder and Levey all developed and
are primarily spoken in island communities.

williams, schneider, trudgill and schreier

Malta, in spite of its proximity to Italy, is part of the British Commonwealth and a member of the EU. Manfred Krug explains its multilingual history, under the rule of Greeks, Romans, Phoenicians, Byzantines,
Arabs and Normans, which is reflected in the linguistic stratigraphy1 of the
English spoken there. In the early nineteenth century, Malta became part
of the British Empire and has remained an anglophone outpost ever since.
Maltese English, as opposed to most other lesser-known varieties, is not
In his chapter, David Levey describes another insular European outpost
of English on Gibraltar: an overseas British territory located off the southernmost tip of Spain. In spite of Englishs status as the only official language
in Gibraltar, Yanito the local variety that contains elements of Andulusian
Spanish and British English is generally preferred in vernacular contexts.
However, as Levey points out through his description, the use of English
in such contexts is on the rise with younger speakers in particular.
Maria Rieder provides a foundational description of Irish Traveller
English. This formerly undocumented variety differs from other kinds of
English through its incorporation of Shelta (also known as Gammon or
Cant) into Irish English, which evidences archaisms and dialect mixture.
As Rieder points out, the combination of Shelta lexicon with Archaic Irish
English grammar creates an unintelligible code that promotes in-group
cohesion and solidarity.

The Americas and the Caribbean

As is well known, the transplantation of English to the Americas has
produced a wide range of sociolinguistic outcomes. Beyond the catastrophic
consequences for the indigenous languages of North America, there is also
the emergence of new and stable varieties through the mechanisms of
language/dialect contact that is emblematic of the Caribbean region.
In her chapter on American Indian English (AIE), Elizabeth Coggshall
provides an overview of the socially and linguistically related varieties
of English spoken by indigenous populations of the United States and
Canada. No one would argue about the lesser-known status of AIE, and

The use of the term stratigraphy is intentional in this description. The intent is to draw a parallel
with the concept of stratigraphy in geosciences that dates back to the mid seventeenth century.
Biostratigraphy is likely to be the best analogue for linguistic stratigraphy since it accounts for the
formation and extinction of species. In this case, the layers of linguistic influence in the language
can provide clues to periods of influence and contact, and cultural realms of contact, as well as other
aspects of sociolinguistic history.


Coggshall addresses the challenge of providing a unified description of a

set of disparate varieties that share some features that set them off from
other anglophone varieties.
The English language has a lengthy and complex history in its West
Indian setting, with its incipient transportation to the region in the early
seventeenth century. Williams (2012) estimates over sixty varieties of
English spoken throughout the region, but the vast majority of those are
lesser known. This volume enhances our knowledge of the lesser-known
West Indian varieties of English with chapters on Bequia in the Vincentian
Grenadines, Saba (one of the former Windward Netherlands Antilles),
St Barthelemy and St Eustatius (also one of the former Windward
Netherlands Antilles).
In spite of its relatively small size, the island of Bequia exhibits a great
deal of linguistic variation within its English-origin varieties. James A.
Walker and Miriam Meyerhoff discuss that variation within the context
of providing an overview of Bequia English. The variety of linguistic
inputs to the overall sociolinguistic landscape of the island including
whalers from the northeastern US region, former indentured servants from
eastern Barbados, creole English speakers from other islands in the eastern
Caribbean is not atypical for the insular speech communities of the
West Indies.
Saba, an island of less than thirteen square kilometers in area, also exhibits
significant village-level variation, like Bequia. Jeffrey P. Williams and
Caroline Myricks description of Saban English focuses on the varieties
spoken primarily in the villages of Windwardside and Hells Gate. Saban
English predates Bequia English by almost a century, with Saba being
colonized by anglophones in the mid seventeenth century. Internal isolation
has characterized Saban social interaction over the centuries of European
settlement and colonization, resulting in distinctive village dialects.
St Eustatius, or Statia, was one of the most important entrepots for
African slaves during the middle to later seventeenth century. Its cosmopolitan character, based on a diversity of merchants, sets the island
apart from most others in the West Indies during the same period. Michael
Acetos contribution on St Eustatius is important because of the islands
prominent place in the regions history, as well as its distinctiveness visa`-vis other regional varieties. Aceto explains this divergence in terms of
socio-economic focus: St Eustatius was a commercial centre and not an
agricultural centre.2

A similar explanation holds for St Martin English.

williams, schneider, trudgill and schreier

Ken Decker outlines the sociolinguistic history and grammatical features of Gustavia English a late eighteenth-century arrival. Gustavia
English is an endangered variety of English spoken by a very small population in the town of Gustavia on the Francophone island of St Barthelemy
that has been overshadowed by the surrounding sociolinguistic complexity
of the French-origin varieties of the island.
While Paraguay is well known to sociolinguists because of its societal
bilingualism involving Spanish and Guarani (or Guaran Paraguayo), it is
very much lesser known in terms of the place and history of English in the
country. Danae M. Perez-Inofuentes describes the English of Paraguay
that was brought in through immigration from the British Isles and more
importantly Australia, during the nineteenth century.
Ian Hancock discusses and describes a lesser-known variety of English
that shares its origins with the well-known Sea Islands Creole, often referred
to as Gullah or Geechee, that is spoken by a declining elderly population of
fewer than three hundred in south Texas, central Oklahoma and northern
Mexico. This English-based creole took form on a former reservation,
leading to spatial and cultural isolation for the speakers and fostering a
social situation where the language has retained more original features
than Sea Islands Creole its closely related kin.

The South Pacific

The South Pacific was an area of linguistic and sociolinguistic diversity
and complexity prior to the transplantation of the English language into
the region. While responses to English have produced a wide range of
sociolinguistic outcomes, many of the South Pacific varieties cannot be
classified as lesser-known varieties of English.
Rachel Hendery, however, describes an extremely isolated and lesserknown community on Palmerston Island a tiny atoll in the Cook Islands.
Unlike the situation with many insular locations of English, Palmerston
was uninhabited at the time of the first anglophone settlement. After
140 years of near-total isolation and with a total population of fewer than
seventy-five individuals, there are no salient linguistic differences between
the three social groups of the island.
In their chapter, Pasifika Englishes in New Zealand, Donna Starks,
Andy Gibson and Allan Bell discuss varieties of English spoken by
Polynesian peoples in New Zealand. As the authors point out, Pasifika
English is one of the youngest lesser-known varieties in the region,
having been formed around sixty years ago. As with many third- and


fourth-generation New Zealand communities, the process of language

shift from their Polynesian languages to English is evident and progressing.
David Britain and Kazuko Matsumoto describe Palauan English and
in doing so provide the first documentation of a Micronesian variety of
English. Although Palauan English has arisen under different sociocultural
conditions than have other Postcolonial Englishes (Schneider 2007), it still
exhibits many of the same features.
In conclusion, our second volume documenting further studies in lesserknown varieties of English enhances the body of published work on the
wide-ranging variation in English as a global language. There are a number
of further issues and topics that we would have liked to be able to cover in
this volume, and we intend it, as was our intention with the first, to spark
further interest in fieldwork and documentation of minority Englishes
in a postcolonial world. Endangerment is a very real factor for the vast
majority of LKVEs, with documentation and archiving as key initial steps
in revitalization and reclamation efforts. We will leave those discussions
and topics for another volume.

Schneider, Edgar W. 2007. Postcolonial English: Varieties around the World.
Cambridge University Press.
Schreier, Daniel, Peter Trudgill, Edgar W. Schneider and Jeffrey P. Williams, eds.
2010. The Lesser-Known Varieties of English: An Introduction. Cambridge
University Press.
Trudgill, Peter. 2002. The history of the lesser-known varieties of English. In
Alan Watts and Peter Trudgill, eds., Alternative Histories of English. London:
Routledge, 2744.
Williams, Jeffrey P. 2012. English varieties in the Caribbean. In Raymond Hickey,
ed., Areal Features of the Anglophone World. Berlin: Mouton de Gruyter,
Wolfram, Walt. 2008. When islands lose dialects: the case of the Ocracoke Brogue.
Shima: The International Journal of Research into Island Cultures 2(1): 113.

part i






5 km

Map 2.1 Maltas three inhabited islands






Map 2.2 Malta in its wider geographical context

ch a p ter 2

Maltese English
Manfred Krug


Due to migration, at least 100,000 speakers of English with a Maltese background (including approximately 50,000 speakers of Maltese, a language
historically derived from Arabic; see Fabri 2010; Stolz 2011) live outside
the Maltese islands, notably in Australia, Canada, the UK and the USA.
The designation Maltese English (or, for short, MaltE) in the linguistic
literature, however, applies only to the varieties spoken in the Republic of
Malta, i.e. on the archipelago located in the Mediterranean about 100 km
south of Sicily and about 300 km north of Libya and east of Tunisia. Its two
biggest inhabited islands are Malta (population of c. 380,000) and Gozo
(population of c. 30,000); the third inhabited Maltese island, Comino, has
a permanent population of less than ten (Census of Population and Housing
2005; preliminary report of Census of Population and Housing 2011). (See
Maps 2.1 and 2.2.)
It should be noted at the outset that, depending on the nature and
intensity of language contact, the English spoken by individual people
in Malta may be indistinguishable from a variety spoken and written in,
say, England or Australia. For obvious reasons, these are not the varieties
I will describe in the present chapter. Instead, I will focus on acrolectal
Maltese English,1 which is, of course, a cover term conflating a number
of language-internal and -external factors. The term is used here to refer
to an idealized language variety spoken by university-educated speakers,
who speak at least some English at home and at work, who have spent no
more than short periods of their lives in English-speaking countries outside
Malta and whose parents were both born in Malta.

In line with some earlier publications (e.g. Platt and Weber 1980: 467; Kachru and McArthur 1992:
506) and an increasing number of recent publications on World Englishes (e.g. Baskaran 2008: 611;
Bautista and Gonzalez 2006: 133, 137; Bolton 2006: 293; Mahboob 2008: 252; Mesthrie 2008b: 256;
2008c: 307; Tayao 2008: 294; Biewer 2012; Hickey 2012a: 392), I will use the term acrolect (as well as
mesolect and basilect) without presupposing a creole history or a postcreole continuum of the variety.



manfred krug

Acrolectal MaltE thus correlates with higher socioeconomic strata,

although this is more true of older than of younger speakers because Malta
has in recent times encouraged tertiary-level education by grants and has a
high proportion of young people from all social strata attending university.
When I use MaltE in this chapter without qualifications such as acrolectal
or basilectal, I will be referring to the English spoken by educated Maltese
speakers, who have typically received or are still receiving a tertiary-level
education, but with no assumptions regarding the languages spoken at
home or at the workplace. It is also this larger group of speakers whose
language forms the bulk of the material that is compiled for components
of the International Corpus of English (ICE). In addition to the pertinent
literature, this chapter is based on data from the Maltese component of ICE
under compilation at the University of Bamberg (hereafter ICE-Malta; see
Hilbert and Krug 2010 for details) and data from a questionnaire for lexical
and morphosyntactic variation in English (see Krug, Hilbert and Fabri, in
press, for detail).
As is often the case in places with a colonial history involving British
rule, the varieties of English that are spoken in Malta represent in actual
fact a continuum between an acrolectal variety (a near-RP pronunciation
with a grammar and lexicon that is very similar to standard BrE) on the
one hand, and basilectal varieties on the other. The latter are characterized
by typical EFL learner features and more structural parallels with Maltese,
i.e. contact features, plus extensive code-switching (cf. the continuum
described in Vella 1994, Bonnici 2010 and such notions as mixed Maltese
As will be seen, MaltE exhibits not only interspeaker but also
intraspeaker, including stylistic, variation. I will concentrate here on the
more formal English spoken and written at the workplace rather than that
used at home or in private, informal emails (or other digitally transmitted informal messages), but point to important stylistic differences where
necessary. In essence, acrolectal MaltE is thus taken to be a standardizing edulectal variety that is oriented towards the traditional exonormative
British standard, but to trained linguists at least noticeably different
from it.

Sociolinguistic history and current status of Maltese English as

a lesser-known variety

Maltese English shares many of the typical characteristics of a lesser-known

variety listed by Schreier, Trudgill, Schneider and Williams (2010: 4): it is

Maltese English


far from exhaustively described (although there has occurred a surge in

scientific interest recently); it is lesser known outside the country; it is
associated with a stable region and was (in fact, still is being) formed by
language and dialect contact. And while the notion of a specific, independent variety of Maltese English, replacing or developing in addition to
the exonormative British standard would have seemed almost inconceivable to the majority of the Maltese population until quite recently, MaltE is
increasingly conceived as a distinct variety by its own speakers. Like many
other lesser-known varieties of English, MaltE is thus an identity carrier
for the speech community, sometimes consciously used, but often below
the level of awareness.2 MaltE is also felt to be distinct by speakers who
come into contact with the variety, including speakers from other varieties
of English. That this perception is justified on linguistic grounds will be
seen in Sections 4 to 7 below.
There are other features of typical lesser-known varieties that MaltE
shares not at all or only very arguably (figures from Census of Population
and Housing 2005):
r Being spoken by over 300,000 people (88 per cent of the population

aged 10 and older speak at least some English), MaltE is not an endangered variety of English.
For the vast majority of the Maltese population, MaltE is a second
language. It is Maltese that is the or a first language for around
93 per cent of the population.
Unlike in many regions with lesser-known varieties of English, bilingualism is therefore the norm rather than the exception in Malta.
Unless the term MaltE is understood in a narrow sense (i.e. restricted
to just those c. 9 per cent of the population who use English as the or
a main language in the home), then MaltE is not a minority language
within the relevant regional confines.
British colonial rule and substantial British inputs are clearly in evidence, but MaltE was not transmitted by a settler community nor did
it develop early in the colonial era in a newly formed community.

The next section will offer more figures and explanations for the claims
laid out above.

See below on overt style markers of MaltE. It should be kept in mind, however, that the identitycarrying function of language is probably characteristic of every speech community and indeed of
every form of human verbal interaction and not restricted to lesser-known varieties of English.


manfred krug

History, language policies, education system

Malta and Gozo were first settled as early as about 5000 bc, probably by
farmers from Sicily. The modern Republic of Maltas population density is
the highest of all countries in the European Union and one of the highest
on the planet. The island of Malta has an area of 246 square kilometres. Gozo (known as Gawdex in Maltese) is less densely populated with
approximately a quarter of the bigger islands area (67 square kilometres)
but less than 10 per cent of the Maltese population.
Due to its strategic importance in the Mediterranean, Malta has a long
history of varying ruling powers, including Greeks, Phoenicians, Romans
and Byzantines. The islands were under Arab control from 870 to 1090/91,
when the Normans conquered Malta. The monastic order of the Knights of
St John (also known as the Knights of Malta) ruled on the archipelago from
the sixteenth century until 1798, when Napoleon conquered Malta, but the
French were evicted only two years later by the Maltese, with British and
Italian support. In 1814 Malta officially became part of the British Empire,
i.e. a British colony. While remaining part of the Commonwealth, Malta
became independent in 1964 and a republic in 1974. In 1979, the last
British troops left the country. Since 2004, Malta has been a member of
the European Union (EU); it adopted the euro in 2008.
Even though a small number of British citizens have stayed in Malta
and a greater number have retired there, most British people who formerly
worked for the administration and armed forces left the country after
independence. The current administration is almost exclusively of Maltese
descent. And the vast majority of the current educated elite learned English
as a second language, often initially by reading rather than speaking the
language, and only rarely through intense language contact with native
speakers of British dialects (though this will often have occurred at later
stages of their professional careers during stays abroad). Such facts explain
why a number of especially non-acrolectal MaltE features are triggered by
English spelling (e.g. the vowel pronunciations in words like secondary, the
/l/ in palm, /n/ clusters in ringing; for qualifications see below). There is,
then, no unbroken line from British colonial settlers to the current elites
or other linguistically significant social groups in Malta. Maltese English is
therefore a rather young variety of English and not a settler variety.
Italian gained prominence during the rule of the Knights (15301798)
and used to be a prestige language in the middle and upper classes until
well into the twentieth century. Italian was the official language in Malta
until 1934, i.e. for over a hundred years under British rule, when it was
replaced by two co-official languages: English and Maltese. Both languages

Maltese English


still enjoy official status, but Maltese Malti (see Fabri 2010 for a synopsis
of the current situation and the history of the language) is considered
the national language. Maltese is the only Semitic language written in
the Latin alphabet. While its grammar and core lexicon are solidly Arabic,
Maltese has a high proportion of Italian and Sicilian as well as an increasing
number of English loanwords.3 Until the 1930s English and Italian were the
more prestigious languages, but Maltese has since steadily gained in status
and recently seen a further boost: in 2002, i.e. two years before Maltas
accession to the EU, Maltese became an official language of the European
There is no grammar or dictionary of Maltese English, only a number
of bilingual dictionaries of Maltese and standard (British) English. Maltese
English is therefore not overtly codified. Given the relatively short period
since English became one of the official languages, this may be no surprise.
With the notable exception of Mazzon (1992), most early studies of MaltE
focused on code-switching and the phonology of the variety. More recently,
empirical studies of morphosyntax (Hilbert and Krug 2012; Krug et al. in
press) and lexical items (Krug and Rosen 2012) have followed. Effects of
globalization as well as different or changing stylistic and intergenerational
preferences have not been thoroughly studied for MaltE except for a few
phenomena (notably quotatives and rhoticity; see Bonnici 2010).
As indicated in Section 2, Maltese is the native language for about
93 per cent of the population. According to census data from 2005,
nearly 90 per cent of the population aged 10 and older claim competence in English, although degrees of proficiency vary considerably: while
65 per cent of the over 10-year-olds claim to speak English well, 12 per cent
do not speak it at all. Detailed statistics for the main languages in Malta
are given in Table 2.1. While code-switching is common across the Maltese
islands and across all social strata, only about 9 per cent use English as a
main language at home and about 1 per cent of the population claim that
English is their only L1 (census data from 2005; Sciriha and Vassallo 2006:
26). L1-speakers of English are found primarily in the higher socioeconomic
strata; in addition, higher usage rates of English vary regionally along the
typical urbanrural cline. Particularly high rates have been reported for
areas where tourism, administration or high education levels are concentrated. These include the capital of Valletta, the nearby conurbation of

Tas-Sliema and San Giljan;

and towns like Is-Swieqi, L-Ibragg , Attard
and Melliea.

Italian and Sicilian together account for about 50 per cent of the Maltese lexicon; estimates for the
proportion of English loanwords range from about 5 to 20 per cent (see Brincat 2005).


manfred krug

Table 2.1 Question: How well do you speak . . . ? (Census of Population

and Housing 2005, data for population aged 10 years and over)

A little
Not at all















Not only bilingualism, then, but also tri- and multilingualism are
widespread in Malta, an aspect that is also well documented in the literature
(e.g. Camilleri 1991; Sciriha 2001; Sciriha and Vassallo 2006). As Table 2.1
shows, 57 per cent of the Maltese population claim (at least some) competence in Italian, 21 per cent in French, 6 per cent in German and 4 per cent
in Arabic. Both in terms of speaker numbers and in terms of average competence, these languages trail well behind Maltese and English. This multilingual situation is partly due to historical language contact (see above),
in particular as far as English and Italian are concerned. Other important
factors are trade and tourism (which includes a significant English language
teaching branch) in a country whose other official language has a small
number of native speakers by international standards. Hence, Maltese
has no privileged status in teaching syllabi outside Malta and the fact
that English is sufficient for communicative purposes on the archipelago
is a counterincentive for potential learners of Maltese as a foreign
Obviously, Maltese language and education policies play a role as well:
Schooling in Malta is mandatory until the age of 16. There is no official
policy on the classroom use of languages, but the National Minimum
Curriculum from 1999 issued by the Ministry of Education emphasizes
the importance of English and Maltese as official languages and states
further that pupils in secondary schools are expected to learn a third or
fourth language. Both Maltese and English are used from school entry, but
Maltese is naturally more prominent in primary schools, while English is
more prominent in secondary schools and dominant at tertiary level (in
class). There is a tendency for state schools to use less English than Catholic
(church) and private (independent) schools (see Fabri 2010 for detail). In
particular, English is the dominant language of reading and writing. The
language of instruction depends to a great extent on the language of the

Maltese English


textbooks, most of which are in (British) English. For the same reasons,
the subjects Maltese and History are largely taught in Maltese. In spoken
interaction, code-switching is widespread among both pupils and teachers,
although a change in teacher education has shifted the balance somewhat
towards Maltese: until the 1970s, teachers were trained by British religious
orders, but more recently teachers have been trained by bilingual Maltese
native-speaker scholars at the University of Malta.


Compared to other linguistic levels, it is the MaltE phonology that is

probably most independent of exonormative standards. In this section I
use standard Southern British English pronunciation (commonly known
as Received Pronunciation or RP) as a reference point, which historically
was and for many Maltese speakers still is the exonormative standard.
This is in line with previous studies on MaltE phonetics and phonology,
from which much of what follows is adopted, though often adapted (cf. in
particular Vella 1994: 5786; Calleja 1987; Mazzon 1992: 1269; Camilleri
1991: 1089; Beer 2011; Bonnici 2010; Borg and Azzopardi-Alexander 1997:
4.1 Consonants
4.1.1 Clear and dark /l/
RP has complementary allophonic variation for the phoneme /l/. Wordinitially before a vowel (as in lead or like), RP has clear (or light) /l/. In
prepausal and preconsonantal positions (as in bell or world), we find the
velarized allophone [], known as dark /l/ in RP. Intervocalic /l/ (as in belly)
and /l/ before yod (a possible realization in lure) are somewhere in between,
but in RP (unlike General American, for short: GenAm) tending towards
clear /l/ (see Johnson and Britain 2007; Wrench and Scobbie 2003). The
vast majority of MaltE speakers have only clear /l/ (e.g. Vella 1994: 778)
and this applies also to the acrolectal end of the continuum. Vocalized
allophones of dark /l/, which are common in BrE in coda-position
(e.g. bottle, hill ), are rarely heard in MaltE.
4.1.2 Syllable-coda devoicing
The devoicing of word-internal, syllable-final lenis obstruents before fortis
consonants (as in lobster, bagpipe or drugstore) is common in varieties of


manfred krug

English around the world, although this assimilation process mostly goes
unnoticed for speakers and hearers alike.
Word-final devoicing of consonants (as in feed ) is also common in normproviding inner-circle Englishes, including RP and GenAm, especially
before fortis and in prepausal position. While existent in other varieties,
syllable-coda devoicing is more widespread and conspicuous across all environments among many speakers of MaltE, though not at the acrolectal end
of the MaltE continuum. Where it does occur, minimal pairs like feedfeet,
mugmuck, fleesfleece, figsfix, Hobbeshops become homophones with a
truly voiceless fortis (rather than just devoiced) final consonant or consonant cluster. Since Maltese exhibits word-final devoicing consistently,
language contact is the most likely immediate source. It should also be
noted that word-final and syllable-coda devoicing is frequent crosslinguistically and that MaltE shares this feature also with many ESL- and
EFL-varieties, so that learner strategies and typological factors in all likelihood play additional roles in what prima facie appears to be a MaltE
4.1.3 Inter-sonorant and word-final voicing
The reverse assimilation process, i.e. voicing between two sounds that rank
high on the sonority scale, can be observed too, notably with intervocalic /s/
in MaltE. While the same process is probably responsible for the variation
found in RP and GenAm in words like intrinsic or forensic, lexical effects
seem to play a role in MaltE too, as intervocalic voicing is more consistent
in some words than in others: in the spoken ICE-Malta data, for instance,
/z/ in basically is almost canonical across all social strata and styles of MaltE
and can thus be said to be a hallmark, if not a norm, of the variety; /z/ is
also common, though less consistent, in acrolectal MaltE pronunciations of
basic, basis and usage. As for non-phonetic factors, the Maltese and Italian
equivalents of basically and basis/base i.e. Malt. bazikament, bazi and It.
basicamente, base also have the voiced sibilant /z/, just like MaltE, so that
language contact plays a role, too.
In addition, this particular lexical effect seems to be connected to the
previous point, the devoicing of word-final consonants and consonant
clusters: based is frequently pronounced /bezd/ in MaltE, and this is
in all likelihood the lexical effect of base apart also triggered by
hypercorrection, because MaltE speakers here avoid precisely the final
cluster /st/ that is often produced word-finally due to devoicing, for example
in such words as raised, prized, gazed.

Maltese English


4.1.4 Spelling <ng> as /, , n/

Like some northern British English varieties, Maltese English features the
cluster // in words like ringing (potentially twice) or running, though
not to the same extent in all contexts. In the ICE Malta corpus data,
some interesting differences in distribution at the word boundary can be
observed across all styles:
(i) Coda-cluster // is most frequently articulated if it precedes a vowelinitial word, and in such cases the final // is often resyllabified as an
onset of the next syllable. Thus, running on /rn()n/ and bring
it /br()t/ become /rnn/ and /brt/, respectively.
(ii) In prepausal contexts, the cluster // is common, too.
(iii) Before consonants (other than // or /k/, which I had to exclude
from the analysis), the cluster // is rarely realized, so that in this
context MaltE conforms to RP in typically featuring the velar nasal
(iv) Word-final alveolar nasal -/n/ instead of the velar RP realization -//
is infrequent in MaltE.
Lexical and grammatical effects play a role too, however:
(v) In ICE Malta, the high-frequency item going features less often the
cluster // in prevocalic contexts than running, for example.
(vi) An exception to (iv) above is the going to future, in which /t/
varies with /nt/ and even /n/ (see below for the quality and variation of vowels). MaltE thus shows reductive frequency
effects, which are typical of grammaticalization generally and therefore common in varieties of English worldwide (see Krug 2000: ch. 5;
4.1.5 Rhoticity
MaltE is generally considered a rhotic variety (Mazzon 1992: 127; Vella
1994: 76, Beer 2011) although non-prevocalic /r/ is commonly not sounded,
especially at the acrolectal end of the continuum. Bonnici (2010: ch. 6), for
instance, finds that postvocalic /r/ in the speech of L1 English-dominant and
Maltese-English bilinguals favours the null realization in the vast majority
of cases (around 80 per cent of the time). Exceptions in her data are the
contractions youre, theyre, were, which are more often than not /r/-ful,
probably due to functional reasons as the loss or vocalization of /r/ in
these would lead to the loss of an entire phoneme and potentially to


manfred krug

homophones (cf. your, their, there, which in Bonnicis data are usually
Beer (2011) finds MaltE to be essentially rhotic if a different spectrum of
the society is analysed. With around 80 per cent of realized non-prevocalic
/r/, he obtains in fact almost exactly the inverse result of Bonnici (2010).
Nevertheless, the two studies are compatible (see Bonnici 2010: 205).
The main reason is that Beers informants are overwhelmingly Maltesedominant L1 speakers. Furthermore, in his subsample analysed for selfreported home language, those speakers who report using mostly English
and only some Maltese at home show dramatically lower rhoticity rates
(of about 50 per cent) and a virtually categorical use of approximants, and
thus near-absence of the more consonantal taps and trills.
As regards allophonic variation, MaltE /r/ has four important allophones,
three of which are found in Maltese (on which see Borg and AzzopardiAlexander 1997; Stolz 2011): poly-vibrant trill [r], alveolar tap [] and a
retroflex approximant similar to AmE []. The fourth MaltE allophone
is the postalveolar frictionless approximant [] known from RP. All /r/
allophones occur essentially in free variation, but the following tendencies
hold according to Beer (2011: ch. 4):
(i) At about 75 per cent overall, approximants clearly outnumber the
remaining /r/ allophones in all positions.
(ii) With about 20 per cent of all /r/ allophones, taps are neither infrequent nor particularly frequent. Their share is higher in intervocalic
positions, both word-internally (as in very, Europe, sorry) and across
word boundaries (e.g. number of ).
(iii) At around 5 per cent of all /r/ tokens, trills are rare, regardless of the
phonetic environment, speech style and social characteristics of the
(iv) If /r/ is realized in non-prevocalic positions (as in hard, start, yesterday
or prepausal singer), then the proportion of approximants is even
higher than on average and, as a concomitant, that of taps and trills
(v) Men use taps and trills more frequently than women, who in turn
prefer approximants disproportionately.
In view of related sociolinguistic research (e.g. Labov 1990), it appears that
the approximant is currently becoming, or has already become, the prestige
norm in MaltE.

Maltese English


4.1.6 TH-stopping and other substitutes

Although Maltese lacks the dental fricatives // and //, acrolectal speakers
of MaltE usually produce them as in RP, if sometimes slightly retracted, i.e.
with an alveo-dental place of articulation. In this latter case, the fricatives
are perceived as or, if further retracted, become alveolar [s, z] (see Vella
1994: 76). Fully fledged TH-stopping occurs occasionally in the acrolect,
in particular prepausally and in preconsonantal position, where the stop
may be unreleased (e.g. with men, with people; personal observation). For
basilectal varieties, stops [t] and [d] are common realizations, and the
voiceless variant is often strongly aspirated. The voice contrast found in
RP is typically preserved in all MaltE varieties for stops as well as fricative
realizations of // and //, except of course in cases of final devoicing.


The vowel inventory of Maltese provides the repository from which the
MaltE vowels are recruited for most speakers, except for those that, due to
extensive training or language contact with an inner-circle variety, approximate an exonormative standard (usually RP). Maltese has a large inventory
of diphthongs as well as long and short monophthongs covering almost the
entire vowel space; MaltE vocalic realizations are therefore as a rule fairly
close to their RP counterparts. A notable exception is schwa, which thus is
the first vowel in the detailed discussion that follows.

Schwa and vowels followed by non-prevocalic /r/

comma, support, secondary
Maltese has no schwa. Nor does Italian, and this fact seems noteworthy even
though it is a much less important contact language for MaltE. However,
since Italian was widely used among the higher social strata of the Maltese
society before and even after the spread of English, the Italian phoneme
inventory too can be expected to have had an effect on the development of
the MaltE acrolect. While at the acrolectal end of the MaltE continuum all
full vowels tend to have realizations very close to RP, it is a lesser degree of
vowel reduction in unstressed syllables that is characteristic of the variety
and common to all styles and strata in Malta.
What is a short schwa in RP may range from a schwa to a short, full
vowel in MaltE. Full vowels are categorical in the basilect but also common
in the acrolect. The exact quality in each word depends on the spelling of
the relevant vowel and on typical soundspelling correspondences found
elsewhere in RP and MaltE. The reduced syllables in comma and support


manfred krug

thus vary in MaltE between [] and [], the second vowel in secondary
between [] and []. When orthography alone triggers the pronunciation,
a word like support can also feature [] in the unstressed syllable.
Words like secondary normally preserve four syllables, and the penultimate vowel varies between schwa and the full vowels [] (thus rhyming
with MaltE berry) or [], if orthography holds sway over perceived vowel
quality (cf. similarly the unstressed syllables in comma or about). As in the
second syllables of RP wanted or ordinary, schwas with spelling <e> or
<i> tend to vary between [] and []. In MaltE, [] also occurs commonly
for other spellings before consonants that are often syllabic in RP, i.e. /l/,
/n/ or /m/, as in bottle, bottom or button (see Vella 1994: 74).
bird, nurse, father
There are rhotic and non-rhotic speakers of the acrolect (and a vast majority
that is variably rhotic; see above on rhoticity in MaltE). This results in a
number of different pronunciations of words like nurse and bird. Both
acrolectal and basilectal MaltE rhotic speakers who have no central vowel
tend to use [r]. Non-rhotic acrolect speakers often have [] (long schwa),
as in RP. Rhotic acrolect speakers vary between an r-coloured schwa (long
or short) and a short schwa with any allophone of /r/, which is indicated
in the following phonetic transcriptions as italicized [r]. Due to the lack
of long schwa, some acrolectal speakers produce homophones for words
like were and where (or per and pair etc.). Such non-rhotic speakers tend
to vary between [w] and [w], and rhotic speakers between [wr] and
[w:r]. For the RP short schwa in unstressed syllables ending in <er>, as
in father, the same vowel qualities are found in MaltE as for the RP long
schwa, but the duration is always short.
start, north, force
There is no perceptible difference in MaltE for the vowels of north and
force. In words like start and force, rhotic MaltE has both short and
long [] and [], respectively, either of which is followed by any MaltE
allophone of /r/. Non-rhotic MaltE accents feature the vowels given in the
relevant sections for bath and thought.
near, square, cure
In the sets near, square, cure, rhotic MaltE has typically [()r], [()r]
and [()r], respectively. Notice, though, that the length of the non-central
vowel varies considerably and that the MaltE schwa is optional or may be
lowered to []. Furthermore, the MaltE vowels in near and cure have

Maltese English


somewhat tenser starting points than RP. The vowel qualities of non-rhotic
and rhotic MaltE are similar for all three environments.
Triphthongs of RP
The RP triphthongs commonly found in tower, tire are typically [r]
and [r], respectively, in MaltE. Non-rhotic accents typically have []
and []. Diphthongization to [] and monophthongization to []
(or similar vowel qualities), known as smoothing from BrE (Wells 1982:
2389), are very rare in MaltE.

Other vowel contrasts between MaltE and RP

The RP trap vowel // has no equivalent in Maltese. In MaltE [] varies
with []. Both can be short or long, typically mirroring the low-level (i.e.
sub-phonemic) lengthening found in RP before voiced consonants as in
bad, sad or lag. Naturally, lengthening applies less frequently when speakers
display final devoicing, but even here pronunciations with long vowels like
[bt] are common.
kit, foot, fleece, goose
The two lax high vowels // and // tend to be more peripheral in MaltE
than their RP equivalents. The same is true for their tense and longer
counterparts /i/ and /u/, which are typically even tenser in MaltE and
less noticeably diphthongal in open syllables and before voiced consonants
than in RP (e.g. see, shoe, seed, shoes).
The face vowel /e/ is [i] in MaltE and thus similar to RP. As in many
other varieties of English, including RP (see Wells 1982: 240), monophthongal realizations [()] and [e()] are also commonly heard for this phoneme
(see Stolz 2011: 243 on the variation between open and close allophones of
what is usually labelled the Maltese phoneme // in the literature).
The RP pronunciation [] for goat is similar to the typical MaltE
pronunciation []. As in many other varieties of English, including RP
(see Wells 1982: 240), monophthongs [o()] and [()] are heard for this
phoneme too (see Stolz 2011: 243 on the variation between open and close
realizations of the Maltese phoneme //).


manfred krug

bath, palm
The Maltese phoneme system has only a central /a/, both long and short,
with the quality typically being []. All varieties of MaltE use both long []
and short [] for the bath and palm vowels more frequently than the more
back RP quality []. The <l> in palm or calm is not consistently silent
across varieties of MaltE, with mesolectal and basilectal varieties favouring
/l/-ful pronunciations. If /l/ is pronounced in MaltE words like palm and
calm, then the vowel tends to be short.
4.4 Minor vocalic differences between MaltE and RP
thought, price, choice, mouth
As in RP, the thought vowel is pronounced [] in MaltE and sometimes
slightly raised. The raising diphthongs found in MaltE for price, mouth
and choice, i.e. [] [] and [] respectively, are in most environments
indistinguishable from their respective RP pronunciations [a], [a] and
happy, kit
Like in RP, the happY vowel is usually pronounced [i] or [i] in MaltE.
Since /i/ is tenser in MaltE than in RP, the rather central allophone [i] found
in conservative RP is rarely used in Malta except by English retirees.
Like in RP, the dress vowel /e/ in MaltE is [ e] or [ ] and thus auditorily
indistinguishable in the two varieties, even if the MaltE realization is
generally somewhat more centralized.
The lot vowel in MaltE is typically [] and thus somewhat higher than the
RP realization []. In addition, the MaltE realization tends to be marginally
more central than its RP counterpart.
The quality of the RP strut vowel // is [] in MaltE and thus almost
identical to RP, if perhaps marginally more central. It can merge with the
vowels of trap, palm, start and, less commonly, bath (see above).

Maltese English


Suprasegmental aspects

4.5.1 Fewer weak forms and contractions, less cluster reduction

The phonological integrity of individual words is more frequently preserved
and, in all varieties of MaltE, vowels in unstressed syllables are generally less
reduced than in RP. As a consequence, MaltE speakers unless they were
exposed to extensive language contact with, e.g. BrE or AmE use fewer
weak forms (of e.g. is, was, have, not) and contractions (like weve, theyre,
didnt, wanna). In a similar vein, Vella (1994: 76) observes a tendency for
glottal stop insertion at word boundaries for MaltE, which in sequences
like heis or goeast makes the liaison (frequently found in RP) by /j/ or
/w/ unlikely.
In the spoken ICE Malta data, contraction ratios are then also consistently and significantly lower than for British English in the LondonLund
Corpus and Bank of English Corpus (on which see Krug 1998). In that
study, a group of items that can immediately precede have and clitic ve
(I, you, we, they, who, there, where, how, here) was investigated while controlling for a number of phonological and syntactic factors. It transpired
that there was a diachronic trend towards higher contraction ratios for BrE
from the 1960s to the 1990s. Another twenty years later, the share of clitic
forms in acrolectal MaltE is on average about half the share found for
educated BrE from the 1960s and 1970s, which shows that MaltE is very
conservative with regard to the use of contractions. The same trend can be
observed for written data from MaltE journalistic prose (on which see Krug
et al. in press). It seems uncontroversial that the low spoken contraction
ratios as well as adherence to an older written exonormative norm are the
fundamental factors underlying the situation found in written MaltE.
4.5.2 Rhythm
Maltese English has a number of characteristics that have an effect on the
perceived rhythm: fuller vowels relative to RP; // or // before /l/, /m/ or
/n/ instead of syllabic consonants (in words like bottle, bottom; see Section
4.2.1 above); greater phonological integrity of individual words; less syllable
compression and thus syllables of more equal length. In particular at the
basilectal end of the continuum, where these features are most in evidence,
MaltE sounds rather syllable-timed (a tendency noted already in Calleja
1987 and Mazzon 1992) and thus close to an Italian EFL variety in terms of
suprasegemental phonology. An ancillary role in this phenomenon is played
by the creation of open syllables through full release and concomitant schwa
epenthesis, which Vella (1994: 767) finds after closed syllables ending in


manfred krug

stops or affricates, as for instance in stop, red, but, huge giving [stph ],
[rd], [bth ], [hju:] a trait that figures similarly in Italian EFL
varieties, which commonly feature open syllables in such contexts, too, e.g.
[stpe] or [bat].
4.5.3 Stress
Calleja (1987: 7185) and Vella (1994: 7985) find postponed stress in
MaltE for words that in RP have stress on the antepenultimate or an earlier
syllable, as in examples (1) to (3). On a related note, single-stress words
and compounds with an early primary stress in RP such as (4) and (5)
often have late stress in MaltE or receive two full stresses, in which case
a secondary stress is typically promoted to a primary one. Compare the
following examples (from Calleja, Vella, Bonnici and ICE Malta):

RP cri.ti.ci.sm vs MaltE cri.ti. ci.sm

(2) RP e.xer.cise vs MaltE e.xer. cise

(3) RP cen.ti.me.tre vs MaltE cen.ti. me.tre
(4) RP dish.wa.sher vs MaltE dish. wa.sher
(5) RP trai.ning part.ners vs MaltE trai.ning part.ners
It is also striking that MaltE words ending in -ism (like tourism, fascism,
plagiarism, socialism, communism) fairly consistently receive stress on the
penultimate syllable. With some exceptions, English words ending in -ism
have equivalents in Maltese ending in -izmu, which in turn are loans from
Italian ending in -ismo. Both languages have their stress in these words on
the penultimate syllable. Furthermore, Maltese has regular stress on heavy
final syllables (i.e. syllables with a long vowel or diphthong, and syllables
with a short vowel followed by consonant clusters or geminates; Fabri 2010:
800). Hence, late stress in words like criticism in MaltE can be borrowed
directly from Italian or, even more likely, via Maltese.
More generally, transfer of the Maltese pattern heavy late syllables receive
stress seems to be applied to MaltE commonly, e.g. in the above MaltE
examples (1) to (3): except in cases of vowel epenthesis, criticism features
a consonant cluster in the final syllable, exercise a diphthong; and -metre
a long vowel in the penultimate. Not all words are affected by the MaltE
tendency to postpone stress, however. Vella (1994: 80) notes exceptions
like messenger and characteristically, which are stressed as in RP. Apparently,
therefore, stress shifting in MaltE depends on more than just which syllable

Maltese English


is stressed in RP, Maltese or the Maltese equivalent of an English word, and

it seems certain that lexical effects and effects related to secondary stress
and degree of reduction in RP play important roles, too (see Vella 1994:
4.5.4 Intonation
Calleja (1987: 112) finds that changes in pitch patterns and the occurrence
of tonic stresses are much more frequent in MaltE than RP. This is confirmed in essence and refined by Vella (1994: ch. 5), who provides a detailed
investigation of intonation in MaltE interrogatives and demonstrates
Maltese influence on MaltE intonation. Notably, Vella finds sentence-final
high rise patterns and post-nuclear stressed syllables with a rather high
pitch across (almost) all investigated structures polar questions, interrogatives guised as statements and interrogatives with primary verbs and
auxiliaries (examples below from Vella 1994: 23640). This is noteworthy,
because in RP questions often show a fall or risefall at the end.

Is that /all?


You under- /stand?


Can we com- /pare?


Is there an abandoned cot- /tage?

(10) Do you have the old /mill?

In MaltE, final rises also occur more frequently in declarative sentences
and imperatives than in RP. For instance in the Bamberg Questionnaire for
Lexical and Morphosyntactic Variation in English (see Krug et al. in press,
for the full questionnaire; Krug and Sell 2013 for methodological detail),
the sentences given below, read out by a female acrolectal bilingual speaker,
also have a final rise on but (here meaning though or however; see next
section for a discussion of syntactic aspects) in (11); and a final rise on the
last syllable in each of the examples (12) to (16). Crucially, such post-nuclear
rising intonation patterns are no exceptions; they are paralleled by many
other examples in the questionnaire recordings and also occur in MaltE
spontaneous interaction.

I like this painting, I prefer the other one, /but.

(12) My sister and me got along very well when we were youn- /ger.
(13) This car is more fast than the one I drove yester- /day.


manfred krug

(14) French I do not use a /lot.


American English does not spell like British Eng- /lish.

(16) Dont stay walking on the /grass!


Prominent morphosyntactic features of acrolectal MaltE include want

constructions with a subjective pronoun in the dependent clause, sentencefinal but, definite article omission and marked uses of the progressive.

Special want constructions

In addition to the standard English want construction (Do) You want

me to get us some ice-cream? speakers of Maltese English produce want
constructions with an overt subject in the subjective case and a finite verb
in both the matrix and complement clause. They occur overwhelmingly in
questions and offers (see Krug et al. in press for details):
(17) Do you want I stand over here?4
(18) Do you want I get us some ice-cream?
(19) You want I buy a drink for you?
Maltese has volitional constructions which are construed exactly like (17) to
(19) above (see Borg and Azzopardi-Alexander 1997: 32; Krug et al. in press).
Language contact is therefore clearly the source of the MaltE structures. In
the questionnaire data, usage ratings drop considerably from the spoken
to the written mode, though. This suggests that we are dealing with a style
marker or an overt marker of informal spoken MaltE. In addition, want
interrogatives without do support such as (19) consistently produce higher
usage ratings than those with do support like (17) and (18). Such data lend
empirical support not only to claims made in the literature about the use of
MaltE want constructions but also to the widely held belief that questions
guised in statement syntax (i.e. lacking do support, e.g. Mazzon 1992: 141)
enjoy a high acceptance rate and are common in if not typical of MaltE.
Informal syntactic constructions, like questions without do support, are of
course an ideal context for a nonstandard MaltE construction to emerge,

This sentence was produced spontaneously by a Maltese university lecturer.

Maltese English


whose prototypical frame is an informal setting. As to the question of

whether operator-free interrogatives are indeed typical of MaltE, it should
be kept in mind that this phenomenon is the second-most common among
235 features in vernacular varieties of English worldwide (Kortmann and
Lunkenheimer 2011; eWAVE feature 229; attested in 91 per cent of the
documented varieties).

Sentence-final but

Another classic in treatments of MaltE grammar is sentence-final adversative but (see Mazzon 1992: 142). Here, too, a parallel structure exists in
Maltese, but also in Italian (see Krug and Rosen 2012: 132; Bonnici et al.
2012; Krug et al. in press, for details), so there is a complex contact scenario
underlying the existence of Maltese English sentences like:
(20) I like this painting, I prefer the other one, but.
There may be more general factors than language contact at work as well.
The feature is even more widespread than eWAVE data suggest (feature
211, attested in 23 per cent of the documented varieties). It is, for instance,
also found in Irish and Northern Irish English (Walshe 2009: 1234). And
the Bamberg questionnaire data show high usage ratings for sentence-final
but even in spoken educated AmE not in the written mode, though,
which points to normative pressures. Probably cognitive factors like the
realization of a contrast only after starting the adversative second sentence
and the desire on the part of the speaker to either express or stress the
contrast help to explain the late position of but. In addition, many contrast
markers in English and other languages figure predominantly in sentencefinal position (e.g. though); yet others can figure sentence-initially, -medially
and -finally (like however). Analogy with a semantic equivalent that has a
different syntactic distribution in either English or a contact language, and
thus general cognitive or L2-strategies, are consequently further possible
triggers for sentence-final but in MaltE.
As in the case of want constructions above, usage ratings drop considerably from spoken to written MaltE, which again suggests that we are
dealing with an overt style or identity marker of the variety. In addition,
the questionnaire data reveal an interesting sociolinguistic pattern: while
male Maltese subjects show only minimal differences between the two registers in their usage ratings of sentence-final but, the stylistic gap for female
Maltese student informants is considerable.


manfred krug

Definite article omission

This section focuses on definite article omission, even though the definite
article in MaltE may also occur where it does not normally figure in normproviding inner-circle varieties, e.g. with so called quasi-locatives such
as college, church, school, hospital or university (see Quirk et al. 1985: 277
for details and differences between BrE and AmE). Witness the following
acrolectal spoken example from ICE Malta:
(21) So, NAME, when we look at Standard Arabic this is what we would
learn at the school.
Definite article omission in MaltE is common with language adjectives in
attributive function as in example (22), governmental bodies like cabinet,
government as in examples (24)(25), and posts (i.e. when not used as a
title followed by a proper name), e.g. rector, receptionist, director, dean, as
in example (23):
(22) and it was quite good you know and it was for the translated version
so the Albanian student translate uh answered the questionnaire in
Albanian language.
(23) So when Rector went off to China last year to sign the relative
agreements . . .
Omission is more likely when the person or institution is clear from the
context, in other words when the referent is inherently definite or unique
(see Krug et al. in press for the relevant familiarity hierarchies). In such
contexts, institutions and posts become similar to proper names, which
are ungrammatical with the definite article in the major norm-providing
varieties of English. Uniqueness, for instance, is in evidence for rector or
university in Malta because there is only one university in the country and
only the head of this institution (and not that of a college, for instance) is
referred to as (the) Rector. Following this principle, cabinet and government
occur without the article in MaltE newspaper language only when they refer
to the Maltese government or cabinet. Compare the following examples
from Krug et al. (in press):
(24) These are precisely the sort of proposals that Cabinet should be
discussing, if we are to continue providing a decent service to future
generations. [ed_M_059]
(25) He said the public should know why government was going back
on its decision for offshore wind technology, by now considering

Maltese English


land-based farms. People have the right to ask whether governments

decision is based on technical, or political criteria, and why is keeping
government from publishing these studies.
In a 100,000-word press corpus of Maltese English, cabinet occurs overwhelmingly without the definite article and is in these cases consistently
capitalized, which indicates similarity to a proper noun and hence uniqueness. A stylistic difference is observable with government, which typically
lacks the definite article in editorials, but usually figures with the definite
article in press reportage (for quantitative detail see Krug et al. in press).
The following factors interact in definite article deletion in Maltese
(i) Nouns that rank high for definiteness, familiarity and referentiality
(like rector or dean of a given university or faculty; or government and
cabinet when the country is understood) omit definite articles more
readily than nouns that rank low on these scales. It seems likely that
these hierarchies play a role in definite article omission in varieties
of English around the world.
(ii) The emergence of article deletion is facilitated if not triggered by
the existence of variation between zero and obligatory definite article
in standard inner-circle varieties, whose regulation may be partially
free or complex, as in the case of titles, institutions (as opposed to
abstract uses) and languages. Compare, e.g.: Ill go to university vs
Ill visit the university today; Minister X /Rector Gerrardi announced vs
the minister/rector announced; for English vs for the English language;
an irrational belief in government as a form of magic fairy vs (the)
Government has made it clear.
In such contexts, acrolectal speakers employ the article more frequently
in closely monitored speech than in spontaneous situations. This hints at
a stylistic preference rather than at a grammaticalization process nearing
completion anytime soon. Finally, it is noteworthy that inner-circle varieties
commonly exhibit definite article omission with language adjectives, too:
questionnaire data show relatively high usage ratings for spoken informal
English for the following sentence not only in Malta, but also in the US
and the UK:
(26) Mary has a very good knowledge of Spanish language.
Each of the above varieties rate example (26) on average as Could be said
by many people and in the US (unlike in MaltE and BrE), usage ratings do


manfred krug

not even drop significantly from informal spoken to semi-formal written

English. Rather than exhibit nonstandard syntax in the case of definite
article omission, therefore, MaltE seems to be taking part in a global trend,
being similar to AmE and BrE in informal spoken English while siding
with BrE in differentiating stylistically.
The following examples (all taken from acrolectal MaltE speech, all from
radio interviews with academics) can serve to illustrate the variation found
in standard Englishes around the world. A zero article would be normal
in (27) and (28), less likely in (29) and (30) and out for most in formal
English in (31):
(27) Now the ideal, the idea is to introduce and or enhance the teaching
of Chinese language and culture.
(28) In simple English it is the headquarters of this idea of fostering
Chinese language and culture.
(29) Malta is different from the UK and from Greece, this doesnt really
take away also that even Maltese culture is becoming very quickly
(30) yet that boys talents and fluency in Italian language cannot be
assessed at the moment in the island simply because were bound
to the pen and the paper
(31) Okay so uhm when we look at Arabic language you have the script
thats different
What the above examples have in common is that the adjectives make
language inherently definite, which renders the definite article redundant.
Norm-providing inner-circle varieties are variable in most of the above
examples and omit the article obligatorily only when the adjective is nominalized (cf. when we look at Arabic; fluency in Italian).
A related case of variation is the inherent definiteness of universities.
Compare the variation found in most inner-circle varieties and illustrated
in (32) and (33):
(32) I taught at the University of Cambridge. vs I taught at Cambridge
(33) I taught at the University College of London. vs I taught at UCL.
Like the norm-providing inner-circle varieties, MaltE has grammaticalized
the zero article for acronyms like UCL or UoM (short for University of

Maltese English


Malta), but it can also do without the definite article when the full form is
given, as in (34):
(34) NAME is the head of the Department of Communication Therapy
and Deputy Dean of the Faculty of Health Sciences at University of
Malta, our latest faculty, our newest faculty.
Ordinal numbers are equally unambiguous and logically, therefore, they
need no additional definite article. MaltE more often than BrE omits it in
such contexts, e.g.:
(35) The first year you know they are all new to the system yes finding
their system, you dont really want to overload them. But from second
year, especially third year, and fourth year I think it is very important
to . . .
Finally, seasons of the year when used in a generic sense that is without
reference to a specific year occur in free variation with and without
the article in standard BrE and AmE (see Quirk et al. 1985: 2789), i.e.
in (the) summer/winter/autumn/spring. Current BrE (as represented in the
British National Corpus, hereafter: BNC) tends to use the definite article
in the majority of cases, even though no definite summer or winter is being
referred to.5 MaltE, by contrast, uses a higher proportion of examples
without the definite article, e.g.:
(36) And the reason for that is uhm because if you buy a cheap house
which is rated F, that means its uh uh it has thin walls, it has single
glass window panes its, its roof is very slim, its too hot in summer
and too cold in winter, you will end up spending a lot of money
on your electricity and energy bills throughout the lifetime of that
In summary, it is not contact with Maltese or other contact dialects that
motivate definite article omission in MaltE but general cognitive and grammatical factors. These are universal and thus help to explain why some of the
phenomena described in this section enjoy a more global spread. Maltese
English (like other varieties too, I suspect) seems to use the definite article
most consistently only when it is semantically or grammatically required:
it frequently omits the definite article when this would be redundant (as in

In the spoken BNC (c. 10m words), uses of seasons with the definite article outnumber those without
the article by a margin of about 5:1 when they follow the preposition in. The ratios range from 4:1
(for winter) via 5:1 (for autumn) to 7:1 (for spring and summer).


manfred krug

the case of unambiguous posts, institutions, languages and ordinal numbers; or because the referent has been established in the discourse). Vice
versa, MaltE avoids the definite article commonly where the NP is not
definite but generic (as with summer/winter).
5.4 Progressives
In press language, MaltE usage of the progressive conforms closely to the
exonormative British standard, with the noteworthy difference that it combines more frequently with modal verbs and modal constructions (Hilbert
and Krug 2012). As in BrE, progressives have a higher text frequency in
spoken MaltE than in journalistic prose. Spoken MaltE also features progressives with a limited number of stative verbs, notably have and be,
which then often indicate dynamic meaning (with have) and temporary
behaviour or transitory state (with be), e.g.:
(37) dance can be a wider, uh, can have a wider influence and use than
its currently having.
(38) Muscat may well score crucial political points by accusing Gonzi of
misleading the public. But the sad truth of the matter is both sides
are being deceptive here: . . .
Often, however, as in (39) and (40) for example, semantic changes to
temporariness or dynamic meaning which would license progressive be
and have constructions in standard BrE or AmE (e.g. he is being funny; he
is having a bath) are hard to detect in MaltE:
(39) so thats why I think were having the the older population that were
having apart from uh science which nowadays getting more into act
which the older people are being uh more healthier
(40) but we have so much jargon here and there that the winding roads
that we are having its quite expensive yeah
As in most major varieties of English, the progressive in MaltE can have
future time reference, e.g.:
(41) Were leaving Saturday.
The progressive in MaltE can also occur with habitual meaning, which
would normally trigger simple (present or past) tense in standard BrE:
(42) so every month were just gathering this all this this data

Maltese English


On a related note, MaltE has an additional aspectual marker, stay + V-ing

(Bonnici 2010, example taken from ibid.), which appears to overlap with
iterative meaning:
(43) To go abroad, we dont want to stay changing currency and losing
money off the currency.
Language contact plays an important role here because Maltese has a similar
progressive construction. Most Maltese verbs form their progressive with
the imperfective form following either qieged (the present participle of
qagad he stayed, he was located) or qed (a shortened form of qieged;
Borg and Azzopardi-Alexander 1997: 230).
Finally, innovative uses of the progressive for emphasis with verbs of
personal (dis-)preference like (44) receive much lower usage ratings by
Maltese than by British or American students:
(44) Im (totally/so) wanting/liking/hating this!
This indicates that current MaltE is conservative with regard to this feature
because an older grammatical constraint operates more rigorously than
in BrE and AmE: no progressives for stative verbs. This is a recurrent
pattern for MaltE, since the same is true for other colloquial features that
have recently become common in many varieties of English, e.g. quotative
be like (on which see Bonnici 2010) and theres existentials with plural
complements (see Krug 2007). Young acrolect speakers of MaltE seem to
be adopting them, however.

Further features of especially mesolectal and basilectal MaltE

As can be expected, mesolectal and basilectal MaltE exhibit a great number

of features that are either clear cases of language contact and interference
from Maltese, typical EFL and ESL features (i.e. in line with typological
markedness constraints and/or due to rule rearrangement and simplification) or a combination of these. Such features can be obtained from the
usage rankings and comments on MaltE for eWAVE given by Bonnici,
Hilbert and Krug (2012); appendix A in Bonnici (2010); or Mazzon (1992).
The exact patterning of these features according to language-internal, stylistic and social criteria, however, still requires further research:
r absence or underuse of morphological third-person marking in the

simple present tense (e.g. he go);


manfred krug

r simple present for continuative or experiential perfect, as in I live here

since 1999;

r both nonstandard regularized (like leaved for left) and irregular past


tense and participial forms (like mound for minded);

invariant dont for all persons in the present tense;
differential use of prepositions (over- and underuse of certain prepositions, absence of prepositions where standard inner-circle varieties of
English have one; Philipp 2008);
much figures with plural nouns in non-acrolectal MaltE, in particular
with plurals lacking an {-s} morpheme like people; language contact is
relevant here as afna is Maltese for both much and many in standard
different or unsystematic division of labour between present perfect
and simple past;
would in irrealis conditional clauses not indicating politeness, habitual
or volitional aspects, e.g. I would go if he would come along.
inconsistent use of which and who with animate and inanimate referents;
relative pronoun that used in non-defining contexts;
absence of future markers and object pronoun drop, e.g. Will you have
fresh basil next week? On Monday I have;
use of would for future marking instead of will (Fabri 2008);
regularized plural forms, e.g. houses /hasz/ (as frequently found in
AmE), sheeps;
double comparatives like more easier, more better, more healthier.

Yet other features occasionally mentioned in the literature on MaltE require

more detailed quantitative studies, including comparisons with other standard spoken Englishes, because they seem good candidates for standard
colloquial features of English around the world. Examples include some of
the above, but in particular the following:
r absence of subjects recoverable from context and/or of auxiliaries or

copulas (see Krug 2000: 17880; Bonnici 2010: 292), e.g. Wanna dance?
You sure? You want ice-cream? Gotta go home now.
r adjectives and adverbs have identical forms, e.g. He did it good/quick/real
r parataxis of two main clauses (with optional coordinator and), i.e.
absence of if in conditional contexts, e.g. You go there, (and) hell go
r resumptive pronouns, subject doubling and topicalization, e.g. This
soup, I love it; This is the car which I saw it yesterday;

Maltese English


r loosening of sequence of tense rule (Kmetova and Sciriha 1992);

r preverbal negator never, e.g. I never did it.

6 Discourse markers
Discourse markers have not received the attention they deserve in the literature on MaltE, even though their use, certainly in combination, seems
quite unique. What follows are findings based on the Bamberg questionnaire and corpus data.
6.1 Discourse markers no and eh
Like sentence-final but and special uses of the progressive (see Sections 5.2
and 5.4 above), discourse markers no and eh are not distinctive of acrolectal
MaltE, but characteristic of it. The invariant tag eh is clearly above the
level of awareness among educated speakers of MaltE, as unusually strong
reactions during the elicitation test (laughter, or Yes, we do that a lot!)
showed when Maltese informants were exposed to the recording of the
following test sentence:
(45) Ive got no chance, eh?
A notable differentiation applies to the phonetic realization of the tag:
while MaltE // is commonly perceived as acrolectal or even conceited,
monophthongal realizations are typical of meso- and basilectal MaltE (Ray
Fabri, p.c.). Furthermore, extreme differences in the usage ratings for spoken and written MaltE confirm the salience and stylistic markedness of the
eh tag. It is also striking that the text frequency of eh is more than twice
as high in the preliminary version of the ICE Malta subcorpus than in the
spoken BNC. Apart from the fact that the tag exists in many varieties of
English, including all major national varieties, the usage in MaltE is propagated by contact with Maltese (and maybe Italian), where similar tags
exist, notably as // and //. Below are some naturally occurring examples
of both eh and no.
(46) I think here in Malta we do have a lot of very good, very creative
researchers; perhaps even due to the limitations of our size eh. Sometimes w- we we we learn how to work you know and I think . . .
(47) We have the plate [pause]; and the sculptor [pause]; uh craft uh
great craftsman eh This is the madonna here. Its really something
wonderful eh. Alright yes this is . . .


manfred krug

(48) you know Francia buildings no? And there was uh Lady Francia. She
used to go and all . . .
Unlike invariant eh, no tags seem to operate essentially below the level
of consciousness, as is evidenced by almost identical questionnaire usage
ratings for spoken and written English for the following sentence:
(49) Your sister is older than you, no?
On a more global level, high text frequencies of no tags might have the
potential to become a widespread marker of EnglishRomance language
contact, most obviously so for areas where English is in contact with Spanish
or Italian. The facts that sentence negators are common tags typologically
and that the invariant tag no exists in most codified varieties of English,
too, should facilitate its further spread and may be responsible for the lack
of stylistic markedness of no tags (according to the Bamberg questionnaire
data, at least) in Maltese, Gibraltarian as well as Puerto Rican English.

Greeting alright?

In Maltese English, alright? /rat/ or /:rat/, both with a rising and

falling intonation are more commonly heard as a greeting than in most
inner-circle varieties of English. As in other varieties of English, MaltE
alright? is relatively informal and roughly equivalent to British or American
English Hello! or How are you? and shortened versions thereof, like Hiya?
It may well be that the greeting Alright? is even more typical of Maltese
than of Maltese English (Ray Fabri, p.c.), but the exact ratio is difficult
to determine due to frequent code switching and the common usage of
English-origin discourse markers in Maltese (e.g. alright, okay, bye, thanks).

Multiple yes

Triple or even quadruple assertion by yes, often with other assertive expressions in the co-text is exceptionally frequent in MaltE. In our transitional
ICE component of spoken MaltE, the discourse frequency of three or more
consecutive yes is ten times higher than in the spoken BNC. Language contact may play a role: Italian commonly features the triple assertion si, si,
si. Maltese has iva, iva, yes, yes, typically reduced to /i:vi:va/ (Ray Fabri,
p.c.), so that the number of syllables is identical to English triple yes and
Italian triple si. The phenomenon occurs utterance-internally as well as

Maltese English


(50) I think so, I think the weather makes the people be in one way or
another, I think its true, yes, yes, yes, for sure, definitely.
(51) a: So we have to await that publication?
b: Yes yes yes, Im uh Im very very very interested in that part.
6.4 Discourse marker anyways
While most regional varieties of BrE unlike AmE predominantly use
anyway, the speakers recorded for ICE Malta also prefer anyway but show
a higher discourse frequency of anyways than BrE in the spoken BNC.
Such corpus data receive strong support from the Bamberg questionnaire
data: educated speakers from England clearly prefer anyway, while AmE
has free variation. MaltE (like Channel Island and Australian English) leans
towards BrE, i.e. shows a preference for anyway but less unequivocally so
than speakers in the UK. There is thus some evidence that MaltE is part
of a globalizing trend towards an increasing use of the discourse marker

Even I

Rather unobtrusive features of MaltE develop where an identical structure exists in norm-providing inner-circle varieties but the MaltE meaning
differs in one of the following ways: (i) pragmatic enrichment by implicatures; (ii) generalization by semantic bleaching; (iii) specification or subtle
semantic shift due to a lexical gap or under-/overdifferentiation in a contact language. Even I, for instance, in MaltE often just means me too
or I also (if a verb follows); it thus lacks the emphatic or counter-toexpectation meaning aspects of, say, a British or American lexicon entry
for even.6 Maltese lacks a lexical item synonymous with English even (Ray
Fabri, p.c.), but it has two lexemes for also/too: Semitic ukoll and the
Italian borrowing anki. And since such brief and common phraseological
units like anchio may have been borrowed into Maltese in toto during the
long era of MalteseItalian contact, the MaltE usage of even I might be
a calque from Italian that made its way into MaltE via Maltese anki jien
(also I). Such subtle semantic differences would be interesting topics for
further research. But even in the absence of more detailed studies, the fact

I am grateful to Sarah Grech for bringing this example and the typical Maltese interpretation to
my attention. She also pointed out to me that even I is identical structurally and semantically with
Italian anchio me too.


manfred krug

that the discourse frequency of even in the spoken ICE Malta material is
three times higher than in the spoken BNC strongly suggests that MaltE
indeed employs even in the more general meaning too.
6.6 First of all
The discourse-structuring device first of all enjoys a higher text frequency in
ICE Malta than in the BNC, although the function appears to be the same
in both varieties. It often figures as a discourse marker and can occur turninitially or following another turn-taking signal. Compare the following
starts of speaker turns from the MaltE ICE material:
(52) Well first of all perhaps we could start off by saying a word about the
institute for which you are now working.
(53) Thats right. First of all the theme is dyslexia drama and self-esteem.
(54) First of all however I would like to put much more in, in the context,
I mean, how how did it emerge . . .

Lexicon: British, American or globalizing?

The findings reported here stem from a lexicon questionnaire that was
conducted among acrolectal as well as meso- and basilectal speakers. The
informants gave judgements regarding their personal usage of lexical and
orthographic alternatives that are preferred (or formerly were preferred) in
British or American usage, e.g. centre vs center, autumn vs fall, lorry vs truck.
For clarity of exposition, I will simplify in what follows and label the more
(or traditionally more) British variants merely as British; the alternative
merely as American (see Krug and Sell 2013 for methodological detail).
As can be expected given the countrys history, MaltE is overall much
closer to British than to American English as far as the lexicon is concerned.
About half of the sixty-eight lexical items investigated in the questionnaire
score close to categorically British usage. This is to say that the overwhelming majority of Maltese informants reported using only the British variants.
It seems that, in particular, household and everyday vocabulary items tend
to remain relatively stable and British due to their high entrenchment and
predominant usage in domains that are not subjects of formal teaching.
Examples include the British items nappies, pushchair, dummy, postman,
dustbin, tap, fish fingers and football, which are strongly preferred over their
American equivalents diapers, stroller, pacifier, mailman, trash can, faucet,
fish sticks and soccer.

Maltese English


There is, however, an interesting tendency for younger speakers in Malta

to opt less frequently for the British variants, a fact which points to Americanization or perhaps, if such tendencies should be found in more varieties,
to globalization. On the basis of apparent-time data gleaned from the questionnaires (age bands: (i)  25; (ii) 2650; (iii)  51), one can confidently
confirm Americanization for a number of MaltE lexical items. They fall
into three major groups: the first consists of items that already exhibit considerable Americanization across the entire Maltese population, but the
tendency is nevertheless still in progress because younger speakers choose
the American variants even more frequently than older speakers. To this
group belong lorry, ill, forwards and sport, which in current Maltese English
have already been marginalized by their more American alternatives truck,
sick, forward and sports.
The second group of Americanizing items are close to neutral across
the Maltese English speech community at large. This group contains
the items parcel [package] and to let [for rent]. Overall these are almost
free variants, but there is a clear trend for younger speakers to use the
American variants given in brackets more frequently. For the last group
of Americanizing items, it seems at this point more appropriate to speak
merely of weakening British exonormative influence, rather than of
Americanization proper, because the items still have solidly British mean
values overall. Relevant examples are the above-mentioned discourse
marker Anyway . . . [Anyways . . . ] as well as autumn [fall], biscuit [cookie],
boot [trunk], dummy [pacifier], laund(e)rette [laundromat] and (potato)
crisps [(potato) chips]. Interestingly, shopping trolley [shopping cart] as well
as chemists [drugstore] belong to the group of items that show a tendency
towards a less regular British usage in MaltE, which in all probability is
connected to the near-categorical use of shopping cart and drugstore in
e-commerce. These two trends of MaltE are therefore not unlikely to be
or to become globalizing developments.
In conclusion, in the MaltE lexicon, Americanizing tendencies are often
in evidence and they are for the most part globalizing tendencies (see Krug
and Rosen 2012 for further discussion). There are some exceptions to this
general trend, though: an ever strengthening preference for the more British
variant in younger MaltE speakers is currently observable for aluminium
[aluminum], anticlockwise [counterclockwise], backwards [backward],7 centre

The apparently incompatible tendencies towards backwards and forward in MaltE may be due to
the fact that nominal uses of forward known from the domain of sports (e.g. centre forward) affect
peoples ratings.


manfred krug

[center], holiday [vacation], jacket potato [baked potato], globalisation, liberalisation, organisation [-ization], lift [elevator]. It is apparently the effect of
formal teaching that encourages the use of traditional British variants for
such salient and easily learnable items.

8 Some generalizations: MaltE as economical, crosslinguistically

unmarked or globalized
Maltese speakers of English have regularly two distinct linguistic repertoires
at their disposal. And over 50 per cent of the population and an even
higher share of acrolect speakers command a third repertoire, Italian,
which is not only spoken by many but also the source of about half of
the Maltese lexicon. Since there is also extensive code-switching and codemixing, Maltese speakers can combine these inventories and select from
them often below the level of linguistic awareness individual features
in the formation of their own idiolect of MaltE. Abstracting away from
idiolectal variation (and idealizing somewhat, to be sure), the analyst will
notice that MaltE is in various respects more economical than standard
international English and standard inner-circle varieties of English, and
also more in line with unmarked crosslinguistic preferences. This is especially obvious in the domain of phonology, where features tend to occur
automatically, i.e. phonetically driven, and thus at a subconscious level,
r final obstruent devoicing
r intervocalic voicing of /s/ to /z/, e.g. in basically (cf. /t/-flapping in other

national standards)

r creating more optimal syllables by resyllabification of // in prevocalic

// clusters, thus creating onset in sequences like something else

r creating onsets (speaker economy) and thus at the same time keeping

neighbouring vowels more distinct (hearer economy) through intervocalic glottal stops (as in Malta is)
r creating stronger onsets and thus at the same time keeping neighbouring
vowels more distinct by choosing intervocalic taps (as in there is), rather
than approximant /r/, which dominates elsewhere in rhotic accents.
None of the above features occurs in 100 per cent of the relevant contexts
in any native speaker of MaltE, but their frequency in all types of MaltE is
clearly higher than in, e.g., RP or GenAm.
In the domains of morphology and syntax, stigmatization of nonstandard features is more prominent and thus phenomena driven by economy

Maltese English


and markedness occur mainly in mesolectal and basilectal MaltE. Various examples of regularization (e.g. plural forms like sheeps) and paradigm
smoothing are given in Section 5.5. In the acrolect, economy-driven patterns tend to occur only in contexts where there is variation in inner-circle
standard Englishes in the form of genuinely free variation or a complex
distribution. In such cases even acrolectal MaltE occasionally regularizes
the distribution. This comes in different shapes:
(a) MaltE makes the pattern or structure more economical (i.e. shorter
or less complex);
(b) MaltE makes the pattern more consistent internally;
(c) MaltE makes a new distribution accord with cross-linguistic, i.e.
typological, tendencies and preferences.
Some aspects of the above points are reminiscent of koineization and
trends in AmE when compared to BrE, and may thus be characteristics of
transplanted varieties more generally that have lost their close contacts with
(or become independent of ) their former colonizers and thus the stabilizing
influence of BrE (see Rohdenburg and Schluter 2009). A straightforward
case is the aforementioned omission of definite articles, which is not a
contact feature from Maltese but frequently due to the redundancy of
doubly marking definiteness or even uniqueness (e.g. omission of the article
in Talk to rector since there is only one rector on the island). Another is
the omission of the definite article with language adjectives (e.g. Italian
language) and seasons (e.g. summer) when used in a generic sense.
It therefore seems that, when deviating qualitatively or quantitatively
from standard reference grammars of English, the emerging variety of
Maltese English employs rule simplification more liberally and cognitive
and economical factors more consistently than the more rigidly codified
inner-circle varieties (see Trudgill 2004: chs. 9 and 11 on regularization and
paradigm simplification). However, more consistently is not meant to imply
categorically as the exonormative standard and MaltE systems coexist.
Needless to say, principles of pattern and rule economization often apply
at the same time by way of universal L2-learning strategies, and thus most
often for non-acrolectal speakers of MaltE.
A related case of pattern smoothing can be observed in the MaltE lexicon,
where sports has supplanted the older BrE variant sport. Crucially, MaltE is
not the only variety adopting this strategy: in Channel Island English, too,
sports is the second-most Americanized item in the Bamberg questionnaire
(see Krug and Rosen 2012). Since also Gibraltar English and, to a lesser
extent, even questionnaire data from Wales, England and Australia display a


manfred krug

similar tendency, one may tentatively propose a globalizing trend towards

sports, though at differential speeds in different varieties. And while in
this case globalization is synonymous with Americanization, one should
probably not call it that, because maths (the solidly stable British variant)
shows few signs of giving way to American math (which is equally stable
in the US) in any of the varieties mentioned. Instead, what we seem to
be witnessing around the world in these two cases is a cognitively driven
regularization process by a semantic-structural analogy, where sports fits in
with other school and university subjects (like physics, maths, linguistics)
and features a form with -s that governs a singular verb.


9.1 Convergence, divergence and globalization in

different linguistic domains
Maltese English is not overtly codified and, from a birds-eye perspective, it
still displays a clear orientation towards British English, most obviously so
in its lexicon and morphosyntax. As early as 1998, Trudgill discusses convergence and divergence scenarios in the domains of phonology, grammar
and lexicon. He points out that while the situation for grammar is difficult
to determine, the homogenisation in the direction of North American
usage at the lexical level contrasts with divergence between British and
American English and between varieties of English around the world in
general for the level of phonology (1998: 302). The data reported here
essentially confirm this view. It was seen that differences between acrolectal
MaltE and standard Southern British English are more obvious for phonological features and lexical items than for morphosyntax, and that some
lexical items show signs of Americanization or globalization (e.g. towards
truck, sick, sports, package, for rent). In acrolectal MaltE grammar, especially
in formal written genres, the exonormative BrE model is still very much
adhered to, although differences in usage can be observed, for instance in
the domains of progressives, verb complementation and determiners.

Stylistic variation, identity markers and issues related to prestige

As for stylistic and genre-specific variation, the data presented here suggest
that while the BrE exonormative standard is still the model for formal text
types of acrolectal MaltE, attitudes are currently changing for the spoken
mode, in particular at the more informal end. Substantially higher usage

Maltese English


ratings for informal spoken than for semi-formal written MaltE, on the one
hand, and significant differences between educated MaltE and BrE only for
informal spoken language, on the other, show that morphosyntactic and
discourse features can indeed carry the potential to act as salient identity
and style markers in cases where language contact (with Maltese plus
maybe Italian) plays an important role. This was found to apply in MaltE
to special want constructions, sentence-final but and the invariant tag
As in many anglophone countries and regions, the notion of a specific,
independent variety, enjoying a similar prestige as the exonormative British
standard, would have seemed quite extraordinary to the majority of the
Maltese population until quite recently. And while a codified endonormative standard remains extraordinary for formal writing, it is in informal
spoken language that covert prestige is already being transmitted via MaltE
identity markers. Such a less-than extraordinary linguistic situation with
overt and covert prestige forms can reconcile what to many lay people
must seem a contradiction in terms: two co-existing, diverging prestige
What we encounter in Malta is in fact even more complex: competing
systems as in all varieties of English including a conservatively (though,
even by linguists, often poorly) understood, non-flexible standard English,
which is usually taught in a prescriptive manner in schools, and a whole
gamut of systems with varying discourse frequencies of MaltE features.
On a related note, it is an interesting finding that in the Bamberg
questionnaire data stylistic differentiation in MaltE is consistently more
pronounced for women than for men. Such data allow for two interpretations: one is that the male students who participated in the questionnaire
are less aware of stylistic nuances. The second is that these men are aware
of the stylistic differences but do not feel the need or desire to opt for
overtly prestigious forms in the written medium and instead almost indiscriminately choose forms with covert prestige that have greater potential
to serve as MaltE identity markers.
MaltE is typically classified linguistically as a Phase III variety in Schneiders (2007) model, i.e. as undergoing nativization. It was seen in this
chapter, however, that there exist contact-induced stylistic differences in
MaltE which suggest that the emergence of a local norm as an identity carrier is well under way. Malta is therefore not only sociopolitically further
advanced than Phase III but, at least in some respects, also linguistically
approaching Phase IV (endonormative stabilization), even though other
hallmarks of that stage (like codification and literary creativity in the new


manfred krug

variety) are not, or only very arguably, in evidence (see Thusat et al. 2009;
Bonnici et al. 2012; Hilbert and Krug 2012).
There has been some discussion as to whether MaltE is a first- or secondlanguage variety of English. The fact that in Kortmann and Lunkenheimer
(2012) MaltE consistently patterns with L2-varieties, suggests that it is
despite the undisputed existence of L1-speakers indeed most appropriately
labelled an indigenized L2. And as in many ESL contexts, there are two
important forces continuously at work in (the) Maltese society: display
of local identity and display of formal education involving prescriptivist
overtones. In formal situations such as professional contexts within Malta as
well as in international contexts, educated speakers of MaltE who are aware
of specific MaltE features will generally opt for a conservative, inner-circle
feature. The more relaxed and colloquial the speech situation, however,
and the higher the number of Maltese participants in a conversation, the
more likely it becomes for educated MaltE speakers to use the features
discussed in this chapter. It seems likely that the same general rule applies
for most L2 varieties of English.

The future of Maltese English

As has been shown, most features of MaltE are grounded in a complex

interaction of the following factors: language contact (typically from Maltese, sometimes reinforced by Italian); general learner strategies (e.g. rule
simplification, generalization) and crosslinguistic, typological preferences.
The future of MaltE is consequently difficult to predict but will be the
outcome of precisely the interaction of the above factors with exonormative pressures, globalizing forces and further endonormative developments.
Let us nevertheless dare to venture briefly into the future of more specific
linguistic domains and features of the variety.
More intense language contact with informal English among the younger
generation is likely to further propagate the usage of colloquial constructions with global reach in informal genres, for instance, quotative be like
and theres existentials followed by plural complements. For the acrolect
and formal genres, adherence to an international standard grammar and
British lexicon are likely to remain (except for globalizing items like sports
and package, which are spreading at the expense of sport and parcel ). It
seems improbable, however, that RP or any other national accent will
become a unanimously adhered to model for MaltE in the foreseeable
future. While part of the educational elite in Malta certainly regard RP as
their model, there also exists some antagonism towards RP when spoken

Maltese English


by the indigenous population in Malta due to this accents socio-economic

associations. Occasional stop release of syllable-final <ng> as //, variable
rhoticity, predominance of clear /l/ and fuller unstressed vowels than in RP
are therefore likely to stay part of educated MaltE for some time to come.
What complicates matters is that what prima facie may look like an RPinduced language change (e.g. decreasing rhoticity among younger speakers
of higher socioeconomic strata, as found by Bonnici 2010) can equally plausibly be interpreted as an independent, economy-driven change. The exact
impact of RP on the current and future development of MaltE is therefore
far from clear.
To conclude on a more definitive note, there can be no doubt about the
vitality of a variety which displays such systematic stylistic and registerspecific differences if it is spoken by over 300,000 people, even when
the vast majority of them acquire their most important identity marker,
Maltese, first, and Maltese English only as a second language.

Baskaran, Loga. 2008. Malaysian English: morphology and syntax. In Mesthrie,
ed., 61023.
Bautista, Maria Lourdes S. and Andrew B. Gonzalez. 2006. Southeast Asian
Englishes. In Kachru et al., eds., 13044.
Beer, Max. 2011. Rhoticity in Maltese English. Unpublished Masters thesis, University of Bamberg.
Biewer, Carolin. 2012. Acrolectal English in the South Pacific: emerging standards?
In Hickey, ed., 33653.
Bolton, Kingsley. 2006. Varieties of World Englishes. In Kachru et al., eds., 289
Bonnici, Lisa. 2010. Variation in Maltese English: the interplay of the local and the
global in an emerging postcolonial variety. Unpublished PhD dissertation,
University of California, Davis. Available at http://linguistics.ucdavis.edu/
pics-and-pdfs/DissertationBonnici.pdf; accessed 12 Dec. 2012.
Bonnici, Lisa, Michaela Hilbert and Manfred Krug. 2012. Maltese English. In
Bernd Kortmann and Kerstin Lunkenheimer, eds., The Mouton World Atlas
of Variation in English. Berlin and New York: Mouton de Gruyter, 65368.
Borg, Albert and Marie Azzopardi-Alexander. 1997. Maltese. London and New
York: Routledge.
Brincat, Joseph M. 2005. Maltese: an unusual formula. MED Magazine (The
monthly webzine of the Macmillan English Dictionaries) 27. Available
at http://macmillandictionaries.com/MED-Magazine/February2005/27-LIMaltese-print.htm; accessed 12 Dec. 2012.
Calleja, Marisa. 1987. A study of stress and rhythm as used by Maltese speakers of
English. Unpublished B.Ed. thesis, University of Malta.


manfred krug

Camilleri, Antoinette. 1991. Crosslinguistic influence in a bilingual classroom: the

example of Maltese and English. Edinburgh Working Papers in Linguistics 2:
Census of Population and Housing 2005. Valletta: National Statistics Office, Malta (2
vols., published 2007, available at www.nso.gov. mt; accessed 12 Dec. 2012).
Census of Population and Housing 2011: Preliminary Report. Valletta: National Statistics Office, Malta (published 2012, available at www.nso.gov. mt; accessed
12 Dec. 2012).
Fabri, Raymond. 1995. The tense and aspect system of Maltese. In Rolf Thieroff,
ed., Tense Systems in European Languages II (Linguistische Arbeiten, 338).
Tubingen: Niemeyer, 32743.
2008. Maltese-English language contact. Paper given at Bamberg, July.
2010. Maltese. In Christian Delcourt and Piet van Sterkenburg, eds., The Languages of the New EU Member States. Revue Belge de Philologie et dHistoire
88(3): 791816.
Hickey, Raymond. 2012a. Glossary. In Raymond Hickey, ed., Standards of English:
Codified Varieties around the World. Cambridge University Press, 392406.
ed. 2012b. Standards of English: Codified Varieties around the World. Cambridge
University Press
Hilbert, Michaela and Manfred Krug. 2010. The compilation of ICE Malta: state
of the art and challenges along the way. ICAME Journal 34: 5463.
2012. Progressives in Maltese English: a comparison with spoken and written
text types of British and American English. In Ulrike Gut and Marianne
Hundt, eds., Mapping Unity in Diversity (Varieties of English around the
World). Amsterdam: Benjamins, 10336.
Johnson, Wyn and David Britain. 2007. L-vocalisation as a natural phenomenon:
explorations in sociophonology. Language Sciences 29 (special issue on Issues
in English Phonology, ed. Philip Carr and Patrick Honeybone): 294315.
Kachru, Braj B. and Tom McArthur. 1992. Indian English. In McArthur, ed.,
Kachru, Braj B., Yamuna Kachru and Cecil L. Nelson, eds. 2006. The Handbook
of World Englishes. Oxford: Blackwell.
Kmetova, Tanya and Lydia Sciriha. 1992. Why bother about those tenses?
Sapostavitelno ezikoznanie 17(3): 4557.
Kortmann, Bernd and Kerstin Lunkenheimer, eds. 2011. The Electronic World
Atlas of Varieties of English [eWAVE]. Leipzig: Max Planck Institute for
Evolutionary Anthropology. www.ewave-atlas.org.
Kortmann, Bernd and Kerstin Lunkenheimer, eds. 2012. The Mouton World Atlas
of Variation in English. Berlin and New York: De Gruyter Mouton.
Krug, Manfred. 1998. String frequency: a cognitive motivating factor in coalescence, language processing and linguistic change. Journal of English Linguistics
26(4): 286320.
2000. Emerging English Modals: A Corpus-Based Study of Grammaticalization.
Berlin and New York: Mouton de Gruyter.

Maltese English


2007. Modern methodologies and changing standards in English linguistics. In

Mara Losada Friend, Pilar Ron Vaz, Sonia Hernandez Santano and Jorge
Casanova, eds., Proceedings of the 30th International AEDEAN Conference.
Huelva: Servicio de Publicaciones de la Universidad de Huelva. [CD-Rom]
2011. Auxiliaries and grammaticalization. In Bernd Heine and Heiko Narrog,
eds., Handbook of Grammaticalization. Oxford University Press, 54758.
Krug, Manfred and Anna Rosen. 2012. Standards of English in Malta and the
Channel Islands. In Raymond Hickey, ed., Standards of English: Codified
Varieties around the World. Cambridge University Press, 11738.
Krug, Manfred and Katrin Sell. 2013. Designing and conducting interviews and
questionnaires. In Manfred Krug and Julia Schluter, eds., Research Methods
in Language Variation and Change. Cambridge University Press.
Krug, Manfred, Michaela Hilbert and Ray Fabri. In press. Maltese English morphosyntax: corpus-based and questionnaire-based studies. Il-Lingwa Taghna
(special issue: Towards a Description of Maltese English, ed. Alexandra Vella
and Ray Fabri).
Labov, William. 1990. The intersection of sex and social class in the course of
linguistic change. Language Variation and Change 2: 20554.
McArthur, Tom, ed. 1992. The Oxford Companion to the English Language. Oxford
University Press.
Mahboob, Ahmar. 2008. Pakistani English: morphology and syntax. In Mesthrie,
ed., 24458.
Mazzon, Gabriella. 1992. Linglese di Malta. Naples: Liguori Editore.
Mesthrie, Rajend, ed. 2008a. Varieties of English, vol. 4: Africa, South and South
East Asia. Berlin and New York: Mouton de Gruyter.
2008b. Introduction: varieties of English in Africa, South and South East Asia.
In Mesthrie, ed., 2331.
2008c. Synopsis: the phonology of English in Africa and South and Southeast
Asia. In Mesthrie, ed., 30719.
National Minimum Curriculum. 1999. Issued by the Maltese Ministry of Education. Available at www.curriculum.gov.mt/docs/nmc english.pdf ; accessed
12 Dec. 2012.
Philipp, Isabell. 2008. Maltese English: the variable use of prepositions. Unpublished Masters thesis, University of Bamberg.
Platt, John and Heidi Weber. 1980. English in Singapore and Malaysia: Status,
Features. Functions. Kuala Lumpur: Oxford University Press.
Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, Jan Svartvik. 1985. A
Comprehensive Grammar of the English Language. London: Longman.
Rohdenburg, Gunter and Julia Schluter. 2009. New departures. In Gunter
Rohdenburg and Julia Schluter, eds., One Language, Two Grammars? Differences between British and American English. Cambridge University Press,
Schneider, Edgar W. 2007. Postcolonial English: Varieties around the World.
Cambridge University Press.


manfred krug

Schreier, Daniel, Peter Trudgill, Edgar W. Schneider and Jeffrey P. Williams. 2010.
Introduction. In Daniel Schreier, Peter Trudgill, Edgar W. Schneider and
Jeffrey P. Williams, eds., The Lesser-Known Varieties of English. Cambridge
University Press, 114.
Sciriha, Lydia. 2001. Trilingualism in Malta: social and educational perspectives.
International Journal of Bilingual Education and Bilingualism 4(1): 2337.
Sciriha, Lydia and Mario Vassallo. 2006. Living Languages in Malta. Malta: Print
IT Printing Services.
Stolz, Thomas. 2011. Maltese. In Bernd Kortmann and Johan van der Auwera, eds.,
The Languages and Linguistics of Europe. Berlin and New York: De Gruyter
Mouton, 24155.
Tayao, Ma. Lourdes G. 2008. Philippine English: phonology. In Mesthrie, ed.,
Thusat, Joshua, Emily Anderson, Shante Davis, Mike Ferris, Amber Javed, Angela
Laughlin, Christopher McFarland, Raknakwan Sangsiri, Judith Sinclair, Victoria Vastalo, Win Whelan and Jessica Wrubel. 2009. Maltese English and
the nativization phase of the dynamic model. English Today 25: 2532.
Trudgill, Peter. 1998. World Englishes: convergence or divergence? In Hans
Lindquist, Staffan Klintborg, Magnus Levin and Maria Estling, eds., The
Major Varieties of English: Papers from MAVEN 97. Vaxjo: Acta Wexionensia,
2004. Dialects. 2nd edition. New York: Routledge.
Vella, Alexandra. 1994. Prosodic structure and intonation in Maltese and its influence on Maltese English. Unpublished PhD thesis, University of Edinburgh.
Walshe, Shane. 2009. Irish English as Represented in Film (University of Bamberg
Studies in English Linguistics, 53). Frankfurt am Main: Lang.
Wells, J. C. 1982. Accents of English, vol 1. Cambridge University Press.
Wrench, Alan A. and James M. Scobbie. 2003. Categorising vocalisation of English
/l/ using EPG, EMA and ultrasound. Proceedings of the 6th International
Seminar on Speech Production, 31419.

ch a p ter 3

Gibraltar English
David Levey


Gibraltar is a British overseas territory situated on the southern tip of the

Iberian Peninsula at the western entrance to the Mediterranean with the
northern coast of Africa lying just 30km across the Strait of Gibraltar. It
measures just 6.8 km2 and is one of the most densely populated areas in
Europe (4.9 people per km2 ). It has very few natural resources and its
economy is based mainly on tourism, shipping and financial services. Most
of the population lives in the small town area in the shadow of the famous
Rock, or el Penon as it is known in Spanish, which soars magnificently
to a height of some 426 metres. This Rock together with Jebel Musa, the
mountain on the Moroccan coast opposite, are said to have formed the
mythological Pillars of Hercules which, legend has it, were forced apart to
form the African and European continents.
The British have been in Gibraltar for more than three centuries although
the question of sovereignty remains contentious, and Spain has sought
its return ever since it was ceded to Britain by the terms of the Treaty
of Utrecht in 1713. Gibraltar and its people, for their part, have always
rejected any suggestion that they are Spanish and have resisted all attempts
by Spain to reclaim the Rock, vehemently defending their right to selfdetermination. This has been shown in the two referendums which have
been held on the matter. In 1967, 99.6 per cent voted against the proposal
of Spanish sovereignty and thirty-five years later, in 2002, 98.5 per cent of
Gibraltarians voted against the option of joint sovereignty. Gibraltarians
have traditionally felt strongly British and manifestations of this sentiment
are clearly visible on the Rock in the form of Union Jacks, pictures of
the Queen and pro-British slogans. In recent years, however, particularly
when the local population has felt let down or abandoned by the UK
government, a stronger sense of Gibraltarian nationalism has emerged.
Although Gibraltar remains under the jurisdiction of the United Kingdom,


david levey

which is still responsible for matters such as defence and foreign affairs, the
new 2006 Constitution gave the local government increased autonomy in
running its own affairs.
Of the 29,752 people who live in Gibraltar, 24,288 of these are Gibraltarians, 3,042 are UK British and 2,422 are classified as Others.1 The
Gibraltarian population is predominantly Roman Catholic (88 per cent)
with Church of England and other Christians accounting for 6.6 per cent
and a further 2 per cent of the population are Jewish. Hindus, Muslims,
Atheists/Agnostics and Other make up the remaining 3.5 per cent.2
English is the only official language in Gibraltar, yet largely due to
geographical proximity and historical as well as family ties, Spanish and/or
the local variant Yanito (also spelled Llanito) still arguably remain the
most common forms of expression in the home domain and in informal
contexts. The situation, however, appears to be changing and English use
is increasing, particularly amongst younger speakers (see Levey 2008a: 95
8). While the Spanish spoken in Gibraltar has certain distinctive features
unique to the Rock, its dialectal form and accent, in general terms, are not
dissimilar to, and are sometimes indistinguishable from, those used in the
neighbouring Andalusian towns.3
Yanito is tricky to define and classify since it often implies different
things to different people. For some, it is simply Gibraltarian Spanish, a
variant of Andalusian Spanish but with some locally specific lexical items
incorporated. These are often words borrowed or adapted, not only from
English, but also from the languages of the immigrant communities who
have settled on the Rock. For others, however, Yanito refers fundamentally
to the local tendency to code-switch between Spanish and English.4 Some
would therefore argue that there are three distinct languages spoken in
Gibraltar: English, Spanish and Yanito (Ballantine 2000: 11819), albeit
with considerable overlap.
Most Gibraltarians can converse, to varying degrees, in two or more languages, sometimes independently and sometimes simultaneously. Although
many would consider themselves multilingual, this does not mean that

Figures were kindly supplied by the Statistics Office, Government of Gibraltar and are correct as of
31 December 2013. The last full census was carried out in 2001.
Figures are taken from the Census of Gibraltar (2001).
According to Lipski (1986: 41719) and Ballantine (2000: 119), Gibraltarian Spanish shares certain
phonetic features with the Canary Islands and South America.
As well as being the name given to the local vernacular, Yanito is also the demonym, used both
locally and in neighbouring Spain, to describe someone from Gibraltar. It probably derives from
Gianni (the diminutive of the name Giovanni), and harks back to the eighteenth and nineteenth
centuries when Italians, particularly Genoese, abounded in Gibraltar.

Gibraltar English


they speak each language with equal ease and proficiency. Language choice
and preference often depend on the situation and domain and are conditioned by factors such as age, social class, education and ethnic background (Levey 2008a: 958). Older speakers, for example, who have not
gone on to further education, may prefer to speak in Spanish or Yanito,
and may find difficulties maintaining an extended complex conversation
in English. Younger generations, on the other hand, having been educated
in English, tend to consider their vocabulary is more extensive in English,
and may feel less comfortable speaking formal Spanish in non-colloquial
It is not unusual to find different languages spoken within the same
nuclear family. While parents may choose to speak to their children in
English for the benefit of their education, they may speak to each other in
Yanito and to their parents in Spanish. Although English is seen as the prestige language and encouraged, this does not mean that other languages are
rejected. Gibraltar has always been a multicultural speech community and
language is seen fundamentally as a means of communication. Although
some may view language choice as an act of identity or a declaration of
allegiance, for the most part, Gibraltarians do not consider it paradoxical or a contradiction to feel staunchly British yet choose to speak in

2 Sociohistorical background and influences

on the variety
Gibraltar and the surrounding area is one of the oldest inhabited areas
in Europe with evidence of Phoenician settlements in the vicinity dating
back to about 950 BC. Various tribes subsequently settled including the
Carthaginians, Goths and Visigoths. The period of Visigothic rule ended
in 711 when an Arab and Berber force under Tariq-ibn-Zeyad invaded
the strategically important Rock. It became known as Jebel-al-Tarik (the
Mountain of Tariq) in honour of the conquering leader, and it is from
this that the toponym Gibraltar is derived. Except for a 25-year period
(130933), Gibraltar was to remain in Arab hands until 1462 when it was
finally reconquered. But it was not until 1501 that Gibraltar finally came
to form part of the Spanish Crown, thus putting an end to more than
750 years of Muslim rule. The Arabs and Jews who had once inhabited
the Rock were expelled and, subsequently, the Rock was repopulated by
Spanish Christians.


david levey

The period of Spanish rule was to last 200 years until the turn of the
eighteenth century when, during the struggle for power that surrounded
the War of Spanish Succession (170114), the Anglo-Dutch allied naval
fleet opportunistically seized the Rock on 4 August 1704 in the name of
Archduke Charles of Austria, one of the contenders to the Spanish throne.
Hostilities were finally brought to an end in 1713 with the signing of the
Treaty of Utrecht. As part of this agreement, Gibraltar was ceded to
the British Crown by Philip V of Spain in perpetuum. This treaty remains
the legal justification for British sovereignty in Gibraltar.5
After Utrecht, most of the Spanish left and the national and ethnic
balance of Gibraltar was to change considerably as immigrants, particularly
from Genoa, began to arrive, seeing the trading potential that Gibraltars
unique location and status offered. In 1721, according to one surviving
account, there were 310 citizens able to bear arms of whom 169 were
Genoese, 96 were Spanish and 45 were British (Howes 1991: 2). This,
of course, was not the whole population but is, nevertheless, a useful
indication of the ethnic distribution in Gibraltar at the time. There is,
however, one conspicuous omission from this account: the Jews, who were
undoubtedly an important and influential presence in the early years of
the British period. The reason why they are not officially mentioned is
that by the terms of the Treaty of Utrecht, neither Jews nor Arabs were
allowed to reside in Gibraltar.6 When the first full census appeared in
1753, however, their presence was reflected. Of the 1,816 persons officially
living in Gibraltar, 597 were Genoese, 574 were Jews, 434 were British,
185 were Spanish and 25 were Portuguese. Thus, initially at least, the
Genoese together with the Sephardic Jews formed the backbone of the
new population, influencing its early cultural and linguistic development.
The British, who were mostly military personnel, by and large, kept to
Italian and Spanish were the most spoken languages amongst the civilian
population. Indeed, up to 1830, when Gibraltar became a British Crown
Colony, important proclamations, decrees and announcements were given

Article X of the Treaty of Utrecht (1713) states: The Catholic King does hereby, for himself, his heirs
and successors, yield to the Crown of Great Britain the full and entire propriety of the town and
castle of Gibraltar, together with the port, fortifications, and forts thereunto belonging; and he gives
up the said propriety to be held and enjoyed absolutely with all manner of right forever, without any
exception or impediment whatsoever (translation from the original Latin from Valle Galvez and
Gonzalez Garca 2004: 461).
The fifth paragraph of Article X states: Her Britannic Majesty, at the request of the Catholic King,
consents and agrees that no permission shall be given, under any condition, to neither Jew nor Moor
to reside or dwell in the said town of Gibraltar.

Gibraltar English


in Spanish and Italian as well as in English. Several languages and dialects,

mostly of Romance origin, would have co-existed on the streets and docks
of Gibraltar. Many of the Jews would have spoken Ladino (Judeo-Espanol),
a language derived from Old Castilian, which was widely spoken by the
exiled Sephardic Jews since their expulsion from Spain at the end of the
fifteenth century. It has been suggested that a type of Mediterranean lingua
franca or Romance-based pidgin (Kramer 1986: 53) was also spoken during
the eighteenth century. However, given the limited evidence, it is difficult
to know for sure whether this did indeed exist as such and, if it did, how
uniform or widespread it was. Speculation seems to partially stem from
a brief and tantalizing allusion made by an eighteenth-century Spanish
historian. In his Histora de Gibraltar published in 1782, Ignacio Lopez
de Ayala mentions the existence of an international vernacular spoken in
Gibraltar which was apparently understood by all:
i tanto estos (los Genoveses) como los Judios hablan bien o mal el Castellano
e Ingles, i un dialecto o jerga comun a todas las naciones, sin excluir las
Africanas. (Lopez de Ayala 1782: 374)
[both the Genoese as well as the Jews speak Castilian and English well or
badly as well as a dialect or jargon common to all nations including Africans]

In the first half of the eighteenth century the population was to fluctuate
considerably as deadly epidemics ravaged Europe. The demographics were
also to change as people from different nations sought refuge from the wars
and upheavals that marked the period. A new wave of Genoese immigrants,
both Christian and Jewish, headed for the Iberian peninsula after the
Ligurian Republic, as Napoleon Bonaparte renamed it, was annexed by
France in 1805. There was also immigration from France, Portugal, Spain
and also Minorca, which, like Gibraltar, had been ceded to Britain by the
terms of the Treaty of Utrecht. Italian-speaking Maltese immigrants, who
were to play an important role in Gibraltars sociolinguistic and cultural
development, began to arrive after the Mediterranean archipelago became
part of the British Empire in 1814.
While other European ports and cities saw their commercial activities
badly affected by events in Europe, Gibraltar was able to prosper and hang
up a business as usual sign. As the population increased, overcrowding
became a major problem and it was not uncommon for six or seven family
members to live in one room. In these cramped and usually unsanitary
conditions, the contagious diseases such as influenza, cholera and typhoid
which appeared in waves in Europe in the first half of the nineteenth
century spread particularly rapidly, devastating the local population. Yet


david levey

there was no shortage of new immigrants prepared to take their place and
by 1871 official census figures reveal the population had grown to 18,695.
As a result, the local authorities felt it necessary to curb immigration by
introducing the Aliens Order in Council in 1873 which permitted free
access to the Rock only to the British and required Aliens to obtain special
permits. During the twentieth century, immigration continued, mostly
from other British colonies such as Malta and later from India. Economic
migrants from Morocco also began arriving in the 1970s and 1980s to fill
the labour vacuum left after the closing of the border with Spain, adding
to the racial and cultural fusion.7
Tentative attempts by the authorities to encourage the use of English had
limited success. English was the official language, the language of education
and administration, yet Spanish and its variants were the everyday means
of communication. This of course is not to say that nobody spoke English.
The UK serviceman and their families naturally spoke English along with
other UK residents in Gibraltar. The more affluent classes, who could
afford to, sent their children to private English schools. Many Jews and
Indians, whose home language was often not Spanish, placed considerable
importance on multilingualism. However, taking the population as a whole,
it seems fair to say that levels of English proficiency remained low until
well into the second half of the twentieth century.
The Second World War was to be an important turning point in Gibraltars social and linguistic development. The Rock was of vital strategic
importance to Britain and largely for its own safety the whole civil population of Gibraltar was evacuated for the duration of the war. The majority
eventually went to the UK. For many it was the first time they had left
their homes and were put in a situation where they were obliged to speak
English. When they were eventually repatriated, they returned to Gibraltar
with a stronger sense of national pride and a better level of English. The
first seeds had been sown but there was no revolutionary change in the
language habits. Many, or perhaps most, had started to acquire English
as a second language. On returning to their homes, despite the best of

Whereas in 1981, during the Spanish blockade, there were 2,140 Moroccans listed as usually resident,
the 2001 census figures showed that numbers had dropped by 55 per cent to just 961. Although this
considerable fall in numbers may partially be explained by the inherent difficulties of enumerating
the Moroccan community, this does not detract from the fact that there is a significant decrease
in the numbers of Moroccans living and working in Gibraltar as employment figures clearly show
(Census of Gibraltar 2001: xvi). After the reopening of the border and increased freedom of movement
for EC members, it has become increasingly difficult for Moroccan workers to obtain work permits
and work.

Gibraltar English


intentions in some cases, it inevitably proved more comfortable and comforting to return to familiar language forms.
Although English language competence increased gradually after the
war, the policies adopted by Spain under General Francos dictatorship
arguably did more, directly and indirectly, to improve the levels of English
in Gibraltar than any local or UK initiatives could. In 1969, in a misguided
attempt to starve the local population out, the frontier or verja, as it is
known, was closed overnight, thereby cutting Gibraltar off from Spain
and the rest of Europe. The decision had drastic and sometimes tragic
consequences for those living on both sides of the border. The blockade
was to last thirteen years and Gibraltarians, particularly older ones, find it
hard to forgive and forget this singular action which divided families and
changed lives. Besides the political consequences, the closing of the frontier
had considerable effects on Gibraltars linguistic development.
Resentment and hostility towards Spain intensified and for some language choice became a declaration of allegiance and provided a motivation
for learning and speaking English. At the same time, the need or excuse
to speak Spanish diminished. With the frontier closed, regular social contact with Spanish friends and family ceased as did commercial relations
with Spain. Spanish labourers, domestic workers and childminders, who
had always provided an important language input in the home and the
workplace, and have been cited as one of the chief reasons for language
maintenance, could no longer cross the border. As contact with Spain and
its language decreased, so language contact with the UK increased. Coinciding with the 1970s tourist boom, there were now more regular flight
connections between Gibraltar and the UK. With trips and holidays to
Spain no longer possible, those who could afford to, flew to Britain. During this period, young Gibraltarians, helped by government grants, began
to further their studies at UK universities in greater numbers. Therefore,
when the border finally reopened on 15 December 1982 after thirteen
years of isolation from its Spanish neighbours, a new speech community
emerged with a stronger national identity and with greater English language
Many years have passed since then and, on the surface at least, it would
appear that things have returned to relative normality. Cross-border relations have resumed, but some would argue that the close relationships that
once existed have been lost and time is still needed to rebuild bridges.
English language competence has notably increased and Gibraltar English
is more widely used than ever before, but this does not mean that Spanish
has been displaced.


david levey

Features of Gibraltar English

Considerable variation exists within the community and although Gibraltar English is gaining a degree of stability, there are notable generational


Previous studies (West 1956; Ballantine 1983; Enriles 1992) note that Gibraltarians often fail to distinguish between kit and fleece vowels, thus chip
and cheap may be perceived and produced identically.8 fleece tends to
be shortened and realized as [i] or [ i] and kit may be produced with a
certain degree of tensing, especially before /l/. While kitfleece merger
still features in the speech of some Gibraltarians, particularly older ones,
younger speakers today generally distinguish the two vowels by quality (see
Cal Varela 2001; Levey 2008a).
Enriles (1992: 30); Errico (1997: 141) and Kellermann (2001: 362) noted the
absence of a long close back vowel. More than twenty years ago, Enriles
(1992: 26) wrote that foot and goose tend to merge and were both
pronounced [u + ]. Although short realizations are still evident, a longer
goose vowel is becoming more common and complete footgoose
merger was found in less than 10 per cent of pre-adolescents and adolescents
Levey (2008a: 1045).
The thought vowel in Gibraltar tends to be noticeably short and may
merge with lot, with both being realized as []
and therefore cot/caught
or wok/walk may not be differentiated. While merger is the norm amongst
speakers over the age of 45, the two vowels are now distinguished by younger
Gibraltarians with lot being lower than thought.
trap/strut and start
trap and strut are not distinguished in traditional Gibraltar English
(Ballantine 1983: 50; Enriles 1992: 245), and thus the differences between

This is also suggested by spellings in local Yanito dictionaries (e.g. scrin screen, tipa teapot, pisup
pea soup).

Gibraltar English


cat/cut or match/much are often not perceived. Recent studies, however,

reveal that more than 80 per cent of children and adolescents differentiated
between the two vowels (Levey 2008: 110) with trap increasingly being
produced as a more open and front vowel, tending towards [].
Although some speakers pronounce bath or start with a back vowel,
a short centralized short vowel is favoured, often overlapping with strut.
Amongst the less anglophone speakers, triple trap/strut and start
merger may occur with all three vowels being realized as [a].
In traditional Gibraltar English, the pronunciation of this vowel may be
conditioned by orthography and accompanied by postvocalic /r/. Thus
shirt and bird would be pronounced [ir], nurse and turn [ur], earth and
earn [er] and word or work [or]. In modern Gibraltar English, which is
non-rhotic, nurse is a realized as a notably short front vowel [] although
[] is making inroads.
The use of weak vowels is not common in Gibraltar. letter tends to
be produced as [a] by older speakers and as [] by younger speakers. The
schwa is also beginning to appear.
cure near square
cure [u
a], square [e  ea] and near [i  ia] tend to be produced
as broad opening diphthongs although shorter centring glides towards the
schwa [] are increasingly favoured by younger speakers. Monophthongization remains uncommon.


/p, t, k/
Initial plosives are generally released with aspiration with voice-onset time
(VOT) values similar to those found in standard British English (Cal Varela
2001: 37). Final /p, t, k/ tend to be audibly released and, in the case of /t/,
may be accompanied with some affrication.
/b/ /v/
Labiodental /v/ has no phonemic status in Spanish where the letters <b>
and <v> are both produced bilabially. In Gibraltar English /b/ /v/ merger
occurred very occasionally among older speakers. In a small number of


david levey

cases, the use of [] substituted /v/, particularly after nasals (e.g. involved
// /t/
// /t/ merger was once a typical feature of Gibraltar English and older
Gibraltarians may not distinguish between minimal pairs (e.g. shoes/choose;
shes/cheese; wash/watch), realizing both as fricatives. This, however, is considerably less common amongst the new generation.9
Initial s + consonant
Hispanophones speaking English often insert an epenthetic vowel before
words beginning with /s/ + consonant. This is also a feature of traditional
Gibraltar English (e.g. start [estat]; strong [estron]), although it is not
widespread in modern Gibraltar English.
Gibraltar English is non-rhotic. Approximant [] and flap [] cohabit in
Gibraltar. The use of Spanish coloured trill [r], which Kellermann (2001:
398) notes in a few informants born in the 1940s and 1950s, has now
practically disappeared.
Gibraltar English /l/ is notably clear in all positions although there are signs
of darkening amongst younger informants (see Levey 2008a: 1589).
Although an aspirated [h] realization is the norm in Gibraltar English, Hdropping is occasionally present in intervocalic environments (e.g. behind
[baind], behave [beiv]) and in initial <hu> lexical sets (e.g. huge [judz];
human [juman].
TH-stopping and TH-fronting
Voiced TH-stopping is present in Gibraltar, particularly in common words
(e.g. this, that, the) although // is the norm among younger speakers.
Recently, incidences of TH-fronting have also been noted in the speech of
young Gibraltarians (see Levey 2008b). This is an interesting development

The fricative [] has no phonemic status in Standard Castilian, existing only as an allophonic variant
of [t]. It is a dialectal feature of the Spanish of Cadiz and the surrounding area including Gibraltar
where muchacha (girl), for example, would be pronounced as [muaa] rather than the standard
Castilian [mutata].

Gibraltar English


as // /f/ and // /v/ mergers are unusual in Spain in both adults and
children, and this would therefore suggest that young Gibraltarians are
acquiring a more native English phonetic repertoire.
T-glottalling, was found to be present sporadically in the speech of young
Gibraltarians, but rarely occurred in intervocalic positions. No cases were
recorded for speakers older than 45. As was the case with TH-fronting, it
is significant that the glottal stop is not a feature of Spanish and indeed
is quite difficult for hispanophones to produce. While T-glottalling in
the UK has traditionally been associated with less prestigious lower-class
accents, although this perception is undoubtedly changing, in Gibraltar, it
is not a social marker and is essentially a fast speech phenomenon reflecting
English language fluency.


Little has been written on prosody in Gibraltar and it is an area, which

is worthy of further study. Here I can merely offer a few brief impressions. Gibraltar English is fundamentally different from most UK regional
varieties in that it has a syllable-timed rhythm rather than a stresstimed one and weak forms are rarely used. This, to some ears, makes
it sound non-native. It has distinctive stress and intonation patterns.
Compound nouns such as seashell, car ferry or dockyard, for example, are
stressed on the second syllable and accompanied by a characteristic rising
It is often claimed that Gibraltar English is simply English coloured by
Spanish. Findings suggest that this is a fallacy. When a Gibraltarian speaks
English it is usually evident that he or she is not from the UK. But, to
the trained ear, it is also clear that his or her accent is markedly different
from that of an Andalusian speaking English. The Gibraltarians tone and
rhythm reveal different primary and secondary influences, perhaps as the
direct or indirect result of a Genoese, Sephardic or Maltese ancestry.

4 Lexicon
Gibraltarians are renowned for their code switching. On entering Gibraltar
one is soon struck by they way that locals maintain conversations in more
than one language. As an example of the way this works I cite a short
exchange I overheard in a shop on my last visit to Gibraltar:


david levey

speaker 1: Guapo, excuse me tienes cambio for 10 pounds please?

speaker 2: Sorry, no tengo, my dear. Business is slow today. Estamos en
[Excuse me handsome, have you got change for 10 pounds please?]
[Sorry my dear I havent. Business is slow today. Theres a recession
Codeswitching may occur intersententially, with one person speaking in
Spanish, say, and the other responding in English, or intrasententially
with English and Spanish being used concurrently within the same phrase.
Moyer (1993: 247) identifies four code-switching patterns in Gibraltar:
(i) alternate use of two languages by different participants in verbal
exchange; (ii) combination of different syntactic constituents within the
sentence; (iii) insertion of individual lexical items; (iv) insertion of ritualized expressions with culture-specific content. Attempts to determine why
one language or another should be used at a given moment proved unsuccessful, leading her to conclude, the language of lexical (or constituent)
insertion is truly random (1993: 251).
There is also a characteristic Yanito vocabulary, which has given rise
to two or three locally published dictionaries as well as various web pages,
wikis and blogs that have compiled popular words, expressions and sayings
from Gibraltar.10 While some of these may be widely used within the
Gibraltarian speech community as a whole, others are anecdotal, invented
or exaggerated for comic effect. Some may have existed once but now have
largely or completely fallen into disuse. There are other entries which,
although existing in Gibraltar, are not really exclusive to it and may be
heard further afield.11
It is not my place here to analyse or gauge if, when and to what extent
these words are used today and by whom. We should, however, be aware
of the dangers of falling into the trap of popular or false stereotyping.
In the same way that some might be led to believe that Cockneys today
speak in rhyming slang constantly and consistently, so many of the words


In addition to my own research and observation, four main sources were consulted when compiling
this section: Cavillas (1990) Diccionario yanito, Vallejos (2001) The Yanito Dictionary and Montero
Sanchezs (2010) El habla del campo de Gibraltar and Kramers (1986) English and Spanish in
The presence of Anglicisms and English transfer in the Spanish of the towns and villages in the
surrounding area known as el Campo de Gibraltar has been well documented (Garca Martn 1996,
1997; Montero Sanchez 2010). These are particularly notable in La Linea de la Concepcion where,
for example, there is a well-known street and car park near the border called and spelt Focona which
derives from Four Corners.

Gibraltar English


often attributed to the speech of Gibraltarians are arguably not nearly as

widespread as claimed, or at least not now.
Many of the entries in Yanito word lists are English borrowings which
originated in past times when little English was spoken. Words such as
beki sangui bacon sandwich, trafilai traffic lights or siticonsi City Council are really phonetic representations of Andalusian Spanish transfer.12 If
these realizations were heard over the border in La Linea (Spain) in an ESL
class they would probably be treated as English pronunciation errors and
corrected. This raises the hundred-dollar question whether these should be
considered bona fide lexical items or not. To take the debate one stage further, in terms of their pronunciation, given that English is the only official
language in Gibraltar and Gibraltar is British, should they be accepted as
legitimate British regional variants?
The idea of the Gibraltarian speaking Spanish with a thick Andalusian
accent throwing in a word of badly pronounced English has a quaint
and sometimes comical appeal. While this stereotype might have been the
case in previous generations and may still exist in some older speakers,
times have changed and continue to change. As levels of education and
contact with the English language increase, there is a new generation of
Gibraltarians who are competent in English and speak with a less marked
pronunciation. Spanish transfer is less evident and if, for example, they do
say rolipo, it is not because they cant pronounce lollypop. If they choose
to pronounce it in the Yanito way it may be for comic effect (Gibraltarians
have a keen sense of humour and are not adverse to self-parody!), or, in
some cases, it may be a case of accommodating to the speech of their
Much of the food imported to cater for British tastes from the late
nineteenth century to the second half of the twentieth century was new to
Gibraltar. The autochthonous population pronounced these new products
as heard or read. This gave rise to: arishu Irish Stew; combi or carne con
b corned beef; greivi gravy; Oso Oxo stock cube; Quekaro porridge
(from Quaker Oats); chinchiba ginger beer; saltipina salted peanuts;
bequipagua baking powder; capotn cup of tea; liqueribar liquorice bar;
chinga chewing gum.
As is to be expected, there are numerous words related to work, particularly in the docks and construction. The fact that Gibraltarians have always
worked side by side with Spanish workers may also partially explain why

Andalusian Spanish tends to elide certain intervocalic and final consonants, thus the city of Cadiz
is commonly pronounced (and popularly written) Cai.


david levey

the pronunciation of English words was not anglicised. Some of the most
common words included: doquia dockyard; forme foreman; cren crane;
cimen cement; esprin spring; guasha (tap) washer; gerda girder; winchi
winch; manpagua manpower; iunio or working iunio trade union; penshi
pension. At home, there was juva vaccum cleaner (from Hoover), hacer
londri do the laundry.
While Spanish remained the principal home language, English was the
enforced language of education. Not only was this reflected in the language
of the classroom (e.g. tishe/tisha teacher; cho chalk; sepli say please)
but also extended into the playground. The game of meblis marbles, for
example, was popular amongst past generations of Gibraltarian children
and had its own language to convey the dos and donts of flicking or
shooting (e.g. follinacle fold in knuckle; fondinga fold in finger). It is
interesting to note that the game, along with its corresponding English
lexicon, spread well outside the confines of Gibraltar, as attested to by
numerous Spanish colleagues who grew up in the 1960s and 1970s in the
provinces of Cadiz and Malaga. Needless to say, in the modern computer
age, the game and its vocabulary mean little to the younger generations.
English gerunds are commonly used after Spanish verb forms, giving rise
to: tomamos un trinqui(ng) Lets have a drink, tomar a guashi(ng) have a
wash, hacer nitin a champa to knit a jumper. Amongst younger speakers
today, on both sides of the border, this tendency continues: voy shopping
Im going shopping, tengo training hoy Ive got training today. It is
also common to create Spanish-sounding words from English words: afordarse to afford; chitera cheating; dampista dumptruck driver; pipando
piping hot; plomero plumber. Perhaps the most well-known example of
what some would term Spanglish is the verb afolinarse which means to
line up and comes from the British military term to fall in. There is
another slightly different variation of the hispanification process whereby,
due to influence from English, existing Spanish words are given a different or second meaning, which they do not usually have. Examples of
these false friends include: aplicacion (job) application (Sp = solicitud );
apologa apology (Sp = disculpa); soportar support (Sp = apoyar); vacancia
vacancy (Sp = vacante); documentario documentary (Sp = documental );
orden order (Sp = pedido). There are other Yanito words such as estacion
de polica police station or teatro de operaciones, operating theatre which
are literal translations from English but make no sense in standard Spanish.
While most of Gibraltars lexicon is English or Spanish in origin, there
are several items, which have come about through contact with those
immigrant communities who settled on the Rock. While the Maltese and

Gibraltar English


Italian influence is notable in several local names and some of the typical
Gibraltarian dishes such as rosto, panissa and calentita, the lexical legacy,
which was undoubtedly once more evident, is now limited to a handful
of words which are increasingly disappearing from use. In his Diccionario
yanito, Cavilla includes the following examples: buchero a din or racket
(It. = buscherio); tana hiding place (It. = tana); chufo tuft of hair (It. =
ciuffo). The Jewish community has also contributed various words. Some
unsubstantiated sources have put the figure as high as 500, but my own
research seems to suggest that this figure is greatly exaggerated.13 Amongst
those words of Hebrew or Haketia origins which are recognized by the wider
Gibraltarian community are: bizim balls, guts; ha ham important person;
haiznear observe carefully; las nogas synagogue; haremos who what can
we do? (expression of resignation); echar el who to curse someone. Arabic
has provided words such as jara pigsty; flus money; zup penis; bicef
enough; and chuni nice might come from German (schon). There are
other items whose origins are less clear such as aliquindoi, meaning keep
a lookout, which Vallejo (2001: 25) suggests might come from Calo, a
Gypsy language spoken in Spain, or possibly from the French un clin doeil
a twinkling of the eye.14
Although it might have once been considered a sign of lexical deficiency
or semi-lingualism, Yanito is seen by many as a badge of identity which is
worn with evident pride. Although not prestigious, it is not overtly stigmatized either and Gibraltarians generally look on it with a certain fondness
and see it is a linguistic expression of their unique cultural heritage (Kellermann 2001: 1345; Fernandez Martn 2003: 1901; Levey 2006: 725). The
new generation of Gibraltarians is increasingly competent in two languages
and has a degree of language confidence that many of their parents and
grandparents lacked. If they use Spanglish it does not necessarily imply
language weakness and Yanitadas, as they are known locally, are often
expressions of multilingual word play. Thus, when Gibraltarians say es un
cachofinger its a joke, they are consciously playing with different languages
and dialects.15 There is also a recent trend to add the English suffix -tion
to Spanish words for deliberately absurd effect. This may take place when


I would like to thank all those who took part in my survey. I am particularly grateful to Levi Attius
and Albert Borrell for their very helpful comments and clarifications.
The use of aliquindoi is not confined to Gibraltar. Estar al liquindoi or al liqui are recorded in Pedro
Payans (1991) popular lexicon of Cadiz. The expression is also used in Malaga and can be found in
Antonio del Pozos (2005) and Juan Cepas (2010) respective lexicons. Pozo suggests that it comes
from the English look and do it.
The correct Spanish word is cachondeo, but as Andalusians tend to omit intervocalic <d>, the final
element is hypercorrected to dedo and then translated as finger for comic effect.


david levey

the speaker cannot call to mind the English word or when there is no easy
equivalent (e.g. armondigations meat balls [Sp = albondigas]; asergation de
Spanish torti chard pie [Sp = torta de acelgas]).

5 Conclusion
At present, it is difficult to talk about a homogeneous and uniform Gibraltarian speech community. Education, social class, ethnic background and
particularly the age of the speaker condition language choice and proficiency (cf. Levey 2008a). That children speak differently from parents
and grandparents in any speech community is of course usual, but social
and political events as well as educational policies have combined and
contributed to widening the generation gap in the case of Gibraltar. The
closing of the frontier from 1969 to 1982 was to leave its mark on the speech
community and served as a catalyst for language change.
Although most Gibraltarians can communicate, to varying degrees, in
two or more languages, it would be wrong to assume that everyone in
Gibraltar is multilingual. There have always been an important number
of monolingual UK residents in Gibraltar and monolingual Spanish day
workers have been crossing over the border for centuries. English and
Spanish have always cohabited in Gibraltar although the communitys true
bilingual potential has never been exploited. English is the only official
language on the Rock and given the contentious question of sovereignty,
it seems inconceivable that Spanish could ever be given any official status.
English has gained ground and today it is difficult to find young Gibraltarians who are not reasonably fluent. However the question is whether
or not this has taken place at the expense of Spanish. Although colloquial
Spanish variants are widely spoken in homes and on the streets, concerns
have been voiced in certain sectors that the levels of formal and written
Spanish have declined in recent years. Further research is necessary to gauge
whether this is indeed the case. Spanish is not obligatory in schools but students can choose it as a second language or alternatively French, Russian,
Portuguese or Italian. Perhaps not surprisingly, of all the GCSE and A level
subjects available, Spanish is one of the most popular.16 English as a second
language is of course not an option in Gibraltar, which follows the UK

Official government statistics reveal that in 2013, 405 Gibraltarian schoolchildren took Spanish
GCSE with a pass rate of 88%. Of these passes 68% obtained A or A grades. All but one of the
167 students who took Spanish A level passed, with 18% obtaining A or A (Abstract of Statistics
2013: 414).

Gibraltar English


National Curriculum, but it is interesting to note that English Literature

is the third most popular A level subject after Spanish and Psychology.
It is important to point out that at present, Gibraltar does not have its
own university and so practically all nationals who want to further their
studies and obtain a university degree have to do so in the UK.17 Despite
proximity, there are no Gibraltarians currently studying in Spain, except
those sent there on Erasmus exchange programmes from UK universities.
The reason for this partially lies in the incompatibility of the respective
education systems. Despite the Bologna Agreement, recognizing UK qualifications is not automatic in Spain and can be a lengthy and complicated
process.18 This inevitably means that it is difficult for Gibraltarians to attain
a high academic knowledge of Spanish. While their Spanish language and
vocabulary are sufficient for dealing with everyday situations, they may be
lacking in more formal or complex contexts.
In an interesting recent development, the Spanish government has now
established an Instituto Cervantes in Gibraltar.19 This initiative was met with
mistrust by certain sectors of the local population who see it as a Spanish
Trojan Horse and have questioned why the Instituto Cervantes, a public
institution which promotes Spanish language and culture throughout the
world, would want to set up shop in a place where Spanish is apparently
already widely spoken. The Instituto finally opened its doors in spring 2011
and according to the Spanish news agency Europa Press (5 May 2011) it had
150 enrolments divided into 20 groups before courses had even started.
Gibraltar enjoys a lovehate relationship with its Spanish neighbours.
Periodically there is cross-border tension for one of a number of reasons:
disputes over fishing rights, accusations of encroachment into territorial
waters, protests over planned royal visits, fears over bilateral talks between
UK and Spanish governments in which the future of Gibraltar may be
discussed. Gibraltarians will complain about the exasperatingly long queues
at the Spanish border posts, which they see as deliberate provocation, while
the Spanish authorities will justify the need to control contraband and
illegal immigration. The press on both sides of the border add fuel to the
fire and the question of sovereignty will inevitably be raised once again.
But then things die down and normality returns. Both neighbours have
learnt to live with the situation and with each other. As long as the sensitive
issues are avoided they get on fine and have much in common. National

In 2013, there were 247 student enrolments at UK universities (Abstract of Statistics 2013: 46).
At present the Spanish Ministry of Education has a huge backlog of unresolved applications from
European nationals, many of whom have been waiting for up to two years for a resolution.
This was one of the five resolutions of the trilateral Cordoba Agreement of 18 September 2006.


david levey

sentiment does not preclude Gibraltarians from chatting in Spanish and

enjoying Andalusian food. Although some will follow the BBC on cable
television, others will watch popular Spanish programmes and read Spanish
magazines and newspapers. There is no contradiction or incongruence, as
far as they are concerned, in feeling Gibraltarian and British while enjoying
their neighbours language, customs and culture. This has always happened
and there is no reason to suspect that it will change substantially in the
near future. The Gibraltarian view is clearly summed up by Luis Montiel,
the former Minister for Employment.
The privilege of the Gibraltarian is to live two cultures, two worlds: the
Anglo-Saxon culture and the Spanish culture. We like the good things of
both countries. So we live two cultures and enjoy the best of each. We reject
the worst of one and the worst of the other. But we choose what we want.
Thats the privilege of the Gibraltarian.20

Abstract of Statistics 2013. Government of Gibraltar.
Ballantine, S. 1983. A study of the effects of English-medium education on initially monoglot Spanish-speaking Gibraltarian children. MA dissertation,
University of Valencia.
2000. English and Spanish in Gibraltar: development and characteristics of two
languages in a bilingual community. Gibraltar Heritage Journal 7: 11524.
Cal Varela, M. 2001. Algunos aspectos sociolingusticos del ingles gibraltareno: analisis
cuantitativo de tres variables en el nivel fonico. Santiago de Compostela: Universidad de Santiago de Compostela servicio de publicaciones.
Cavilla, M. 1990. Diccionario yanito, 2nd edition. Gibraltar: MedSUN.
Census of Gibraltar 2001. Government of Gibraltar.
Cepas Gonzalez, Juan. 2010. Vocabulario popular malagueno. Malaga: Arguval.
Enriles, J. M. 1992. The vowel system of Gibraltarian English. MA dissertation,
University College London.
Errico, E. 1997. Gibraltar: a hybrid of language and culture. PhD dissertation,
University of Bologna.
Fernandez Martn, C. 2003. An Approach to Language Attitudes in Gibraltar.
Madrid: Umi-ProQuest information on Learning.
Garca Martn, J. M. 1996. Materiales para el estudio del espanol de Gibraltar:
aproximacion sociolingustica al lexico espanol de los estudiantes de ensenanza
secundaria. Cadiz: Servicio de Publicaciones de la Universidad de Cadiz.
1997. El espanol en Gibraltar: panorama general. Demofilo. Revista de cultura
tradicional de Andaluca 22: 14154.

The quote is from an interview with Luis Montiel in Spanish, which is published in Oda-Angel
(1998: 45). My English translation originally appeared in Levey (2011: 76).

Gibraltar English


Howes, H. W. 1991. The Gibraltarian: The Origin and Development of the People
of Gibraltar, 3rd edition. Gibraltar: MedSUN.
Kellermann, A. 2001. A New New English: Language, Politics and Identity in Gibraltar. Norderstedt: Books on Demand.
Kramer, J. 1986. English and Spanish in Gibraltar. Hamburg: Helmut Buske Verlag.
Levey, D. 2006. Yanito. In K. Brown, ed., Encyclopedia of Language and Linguistics,
2nd edition, vol. 13. Oxford: Elsevier, 7245.
2008a. Language Change and Variation in Gibraltar. Amsterdam and Philadelphia: John Benjamins.
2008b. The changing face of Gibraltarian English: TH-fronting on the Rock. In
Ewa Waniek-Klimczak, ed., Issues in Accents of English Language Change and
Variation in Gibraltar. Newcastle: Cambridge Scholars Publishing, 91101.
2011. National identity and allegiance in Gibraltar. In C. McGlynn, A. Mycock
and J. W. McAuley, eds., Britishness, Identity and Citizenship: The View from
Abroad. Bern: Peter Lang, 7393.
Lipski, J. M. 1986. Sobre el bilionguismo anglo-hispanico en Gibraltar. Neuphilologische Mitteilungen 87(3): 41427.
Lopez de Ayala, I. 1782. Historia de Gibraltar. Madrid: Antonio de Sancha.
Montero Sanchez, Sebastian. 2010. El habla del Campo de Gibraltar. Cadiz:
Moyer, M. 1993. Analysis of code-switching in Gibraltar. PhD dissertation, University of Barcelona.

F. 1998. Gibraltar: la herencia oblicua: aproximacion sociologica al
contencioso. Cadiz: Servicio de Publicaciones de la Diputacion de Cadiz.
Payan Sotomayor, Pedro. 1991. El habla de Cadiz. Cadiz: Quorum.
Pozo, Antonio del. 2005. El habla de Malaga. Malaga: Miramar.
Valle Galvez, A. del and Gonzalez Garca, I., eds. 2004. Gibraltar, 300 anos. Cadiz:
Servicio de Publicaciones Universidad de Cadiz.
Vallejo, T. 2001. The Yanito Dictionary. Gibraltar: Panorama Publishing.
Wells, J. C. 1982. Accents of English, 3 volumes. Cambridge University Press.
West, M. 1956. Bilingualism in Gibraltar. Overseas Education 27: 14853.

ch a p ter 4

Irish Traveller English

Maria Rieder


Ireland is often considered to be a homogeneous bilingual country, and

before the influx of foreigners in the boom years of the Celtic Tiger the
Irish themselves may have regarded their country as a monocultural society.
However, even though still widely disregarded in their cultural distinctiveness, the Irish Travellers have stood out as a separate community in Irish
society for several centuries. The Irish Travellers are a native Irish community with a nomadic background, who naturally share a lot of history
with settled Irish people, but culturally, religiously and linguistically they
have preserved their own identity. According to researchers and the Irish
Travellers themselves, there is no or very little genetic connection with
other European nomadic or Gypsy groups, even though they may share
with them many traditions and values, such as the preference of selfemployment, birth, marriage and burial customs, and values concerning
morality, taboos and purity (Freese 1980: 5363).
The Irish Travellers are recognized as an ethnic group in Northern
Ireland and in the UK. In the Republic of Ireland their legal status has
been widely discussed but remains insecure.1 According to the Census of
2011, there are 29,573 Irish Travellers living and travelling in the Republic
of Ireland, which accounts for 0.6 per cent of the overall population in
Ireland.2 There are also Irish Travellers living in Britain, Australia and the
Linguistically, Irish Travellers differ from the settled community in a
twofold way. Firstly, their in-group code Shelta (also known as Gammon
or Cant), a distinctive communicative tool used in specific, Travellerrelated contexts, provides the possibility to have private conversations in
situations where settled people are present, such as trade and business

www.itmtrav.ie/keyissues/myview/20 (8 July 2012).


www.cso.ie/en/ (23 July 2012).

Irish Traveller English


situations, contexts where warnings need to be exchanged or when talking about taboo topics. Morphosyntactically, Shelta is a mixture of Irish
English grammar and the Travellers own lexicon, a majority of which is
derived from Irish Gaelic and disguised in various ways by means of transposition (deliberate switching around of consonants, insertion and deletion
hAodha 2002
of syllables, etc.; see Hickey 2007b: 382 for an overview, O
for a detailed description of the Shelta lexicon), while a smaller amount is
of unknown, though possibly very old origin. The combination of Shelta
lexicon with Irish English grammar allows Travellers to speak privately
without raising suspicion (for more information on Shelta: Binchy 1994,
1995, 2002; Browne 2002; Cauley 2006; Grant 1994; Hancock 1973, 1984,
hAodha 2002).
1986; Macalister 1937; N Shuinear 2002; O
Besides Shelta, also the Irish Travellers variety of English distinguishes Travellers from settled speakers and general Irish English. However,
Traveller English has not yet been researched as a variety of its own and
therefore the term Traveller English is not yet commonly established. The
linguistic analysis of the variety in this contribution will be based on a
modest corpus of 40,000 words (Rieder, unpublished data). The corpus
stems from a two-year ethnographic project carried out among the Irish
Traveller community in the West of Ireland and consists of seven audiorecorded, semi-structured group interviews of about 40 minutes each. Both
men and women were interviewed; the age group ranged from 18 to 65 and
the participants came from varied socioeconomic backgrounds.
In what follows, the phonological, morphosyntactic and lexical characteristics of Traveller English will be analysed in depth, after a brief account
of the Travellers sociolinguistic history.

2 Sociolinguistic history and current status of the variety


Outline of demographic and cultural history

In exactly what period of Irish history the Travellers emerged as a distinct

cultural group is difficult to determine due to a lack of historical records
typical of nomadic and oral cultures. A DNA study carried out in 2011 with
samples taken from 40 Travellers confirmed that Travellers have a shared
heritage with settled people but that they separated at some point between
1000 and 2000 years ago (Irish Examiner, 31 May 2011).3

www.irishexaminer.com/ireland/dna-study-travellers-a-distinct-ethnicity-156324.html (26 July



maria rieder

This very vague time frame gives room for a lot of debate and speculation.
For instance, MacNeill (1937: 82) and Gmelch and Gmelch (1976: 227)
suggested that the Travellers may be the descendants of a Celtic tribe dating
back to pre-Christian times, which stayed nomadic and was, in the course
of several centuries, joined by peasants, beggars, farmers, seasonal workers
and monastic scholars, all of whom lost their permanent accommodation
due to eviction, unemployment or other misfortunes. Underscoring that
suggestion, Gmelch and Gmelch (1976: 227) reproduce a historical record
from the fifth century ad, which testifies the existence of itinerant groups
in Ireland at that time. Many centuries later, in 1175, the word tinkler and
tynker appeared for the first time in written records as a surname and
occupational name. As Irish Travellers are still often referred to as tinkers
(a term coming from the sound of a hammer hitting metal, which points
to occupations as tinsmiths and jewellers), this led some researchers to
believe in a historical connection of todays Travellers with the tinklers in
the twelfth-century records.
However, several historical facts need to be clarified in order to establish such connection. The above-mentioned early records do not reveal
whether the featured itinerants or tinklers were a clearly distinct social
group different from the rest of the Irish population. The Irish society
was generally highly nomadic until well into modern times, and clearly
distinguishable categories, such as nomadism vs sedentariness, Traveller
vs settled, did not exist with the same connotations in the twelfth century
as today (see Bhreathnach 2007: 32). Still today, the Irish Traveller community are very heterogeneous in terms of occupations, nomadic traditions
etc., and cannot be reduced to any one occupation. Also, we cannot be sure
of any cultural continuity of features attributed to Irish Travellers, such as
family and marriage patterns, gender roles, religious beliefs, occupations
and value system. Therefore, we do not know whether todays Travellers
with all their cultural characteristics have a connection with one or several
different itinerant groups of the past, and that the tinklers mentioned
in these first records are the ancestors of present-day Irish Travellers, even
though the above-mentioned DNA study may suggest that.
The first direct hints of tinkers as culturally comparable to our presentday Travellers can be found in the records of The Commission on the
Condition of the Poorer Classes of the year 1834 (see Gmelch, Langan and
Gmelch 1975: 10). Also, Shelta as a common code must have been well
established before the time of the Great Famine of 1845 to 1848, because
Travellers who emigrated to America during that time have held on to
Shelta as an in-group marker until today. Therefore, there are also linguistic

Irish Traveller English


reasons that support the hypothesis of dating the Travellers as a firmly

established cultural group to the first half of the nineteenth century. We
generally get a clearer picture of the Irish society in nineteenth-century
records, as property ownership started to become important for the rural
middle class and a polarity between nomadism and sedentariness was
emerging (see Bhreathnach 2007: 34). Even though the many famines and
other cases of personal disadvantage forced many people onto the roads,
the Travellers could now be distinguished from those for whom travelling
was only a temporary necessity and who otherwise adhered to a sedentary
lifestyle (see Helleiner 2000: 35).
In the twentieth century an increasing number of written records and
memories of older members of the community give us a better idea of
their culture and lifestyle. Traditional occupations were manifold: until
the 1970s many made a living as tinsmiths or repairers of metal items and
china, helped out in farms and horse stables, worked as fortune tellers, and
sold handmade goods, cattle and horses in markets, fairs or from house to
Changes in the Irish economy in the second half of the twentieth century
resulted in a change in occupational orientation and in general lifestyle.
The introduction of plastic led to a gradual replacement of handmade
tin ware. Technological and industrial progress meant that work on farms
became scarce. The Travellers therefore began to move into towns and
cities, where large temporary settlements at town entrances began to be
seen as a public problem by the settled community (Helleiner 2000: 135ff.).
Several actions by the Irish government and local authorities, such as the
placement of boulders at roadsides, poorly serviced campsites and laws
concerning illegal camping were aimed at settling Travellers in standard
accommodation. Today, the majority of Travellers live in permanent social
housing estates or Traveller halting sites outside bigger towns.
Nomadism was and still is a substantial pillar of Traveller culture and
ideology and the endeavour to settle them is slowly affecting the other
cultural characteristics. The lack of flexibility of accommodation makes
leading a self-sufficient life extremely difficult. Today, some Travellers still
deal in horses or collect and sell scrap, car parts and other metals. Few
have gone into standard employment, and 74.5 per cent of Travellers are
currently unemployed.4
Also, nomadism used to be deeply connected to the importance of family
life, another core value of Traveller culture. Family weddings, funerals,

CSO 2006.


maria rieder

christenings and other celebrations kept the contact with the extended
family alive, and were also the setting where partners were found and new
families planned (McDonagh 1994: 89). Travelling therefore contributed
to the reinforcement of family ties. Families are traditionally quite large
and the age profile is very young with 41 per cent of the community under
14 years of age. Life expectancy, in turn, is considerably lower than in the
settled community.5
Connected to social aspects, nomadism also had a cultural function.
By regularly meeting friends or family, traditions, folk wisdoms, language, values and beliefs could be shared and kept alive (McDonagh
1994: 89). Most values and the moral code centre around the Roman
Catholic faith, in which a number of the Travellers own religious practices have been integrated. Some of these are older Catholic practices,
for example novenas, praying for special intentions such as illnesses and
relatives, faith in and visiting spiritual healers, old Irish superstitions and
omens of good and bad luck. Traveller womens faith is very strong and
openly expressed by wearing religious jewellery and carrying religious items,
such as prayer books, prayer cards, saints images, holy water, oils and
ointments. Most Traveller men usually display their faith less strongly
than women. They participate in the sequence of sacraments but usually
attend mass only on special occasions. Family meetings at events such
as funerals, christenings and weddings usually draw a large number of
Travellers from all over the country and keep these values and beliefs
In conclusion, impeding the freedom of movement of Travellers is slowly
wiping out traditional core values. Nevertheless, the Traveller community is
trying to keep their cultural identity alive while at the same time expanding
their networks towards the settled community.

Linguistic history and current status of the variety

Traveller English is a local variety in the sense that it is spoken in Ireland,

and by Travellers who have emigrated from Ireland. In Ireland, Traveller
English is an overarching social or cultural variety that, in contrast to settled
Irish English which has much dialectal variation, is a more cohesive entity
regardless of geographical location of the speakers (N Shuinear 1994: 58). A
reason for this may be that Travellers, who in the past never stayed in one
place for longer than a few months, did not pick up any one variable from

CSO 2006.

Irish Traveller English


a certain region, but are instead reproducing a peculiar dialect that exhibits
mixed dialectal characteristics.
Phonological, morphosyntactic, lexical and pragmatic features of Traveller English differ slightly from mainstream general Irish English of today
in two respects. Firstly, Traveller English has retained Irish English features
that would have been widespread decades ago among settled Irish people
and can be called archaic Irish English features, which are becoming recessive in general Irish English. These features will be referred to as Archaic
Irish English in the following sections (for a detailed list of Irish English
feature see Bliss 1979; Kallen 1994; Wells 1982), in contrast to the General
Irish English spoken by settled people today, as collected in the ICEIreland Project (Kallen and Kirk, 2008), which takes into consideration
the many different regional dialects of settled Irish English in the Republic
of Ireland. Irish English features have been described as resulting from
the influence of an Irish substrate, as well as from historical and dialectal
features of English settlers, that were retained due to a long period of language shift from Irish to English, and lastly from other, primarily internal,
principles of historical change and variation (Kallen 1997: 3ff.). Looking at
the historical and modern situation of Irish Travellers, a secluded life, limited contact with the wider Irish society and learning English mainly from
speakers based in rural areas may have preserved and reinforced Archaic
Irish English features.
Secondly, Traveller English is experienced as being different beyond
these Archaic Irish English features. To out-group speakers, Travellers are
instantly recognizable by their language and often difficult to understand
even for native Irish people for reasons that will be pointed out below.
This led to suggestions of another substratum that could underlie Traveller
English. N Shuinear, for example, suggested a Gammon underlay (1994:
58), i.e. a possible but lost Cant grammar, which still influences Traveller
Baoill 1994: 157). The
English besides Archaic Irish English features (O
distinctive elements lie primarily in the phonology and prosody, particularly
the intonation of Traveller English, but are also found on all the other levels
of language. The theory of a Gammon substratum, however, has not yet
been proved historically.
Travellers themselves are well aware of their distinctive variety of English
and have described it as a flat accent in contrast to settled Irish English
(Rieder, unpublished data), which again refers to phonology primarily. In
terms of attitudes, Travellers usually defend their variety strongly against
outsiders who might feel General Irish English to be superior to Traveller
English, and claim that they would always refuse to adapt their speech


maria rieder

in situations of contact with settled Irish people. In real-life situations,

however, one can perceive a degree of convergence towards General Irish
English, with a gradual appropriation of standard features also in in-group
situations (OSullivan 2008: 55).


The influence of strong Archaic Irish English features, as well as other,

unknown developments, is most strongly felt in Traveller English phonology. Particularly the greater tendency to mid-centralize kit, trap, strut
and unstressed vowels towards //, the close mid or mid-central onsets of
many diphthongs, and rhoticity, pre-R breaking and pre-Schwa laxing processes surrounding near, square, cure centring diphthongs distinguish
Traveller English from settled Irish English. Apart from these observations, however, single vowel realizations do not differ much from the Irish
English still heard in rural areas, and it may rather be for prosodic reasons,
i.e. intonation, rhythm etc. that Traveller English is perceived to be different. In what follows, peculiarities of Traveller English will be described
in detail, especially in comparison to features of the General Irish English
of the settled community (as described in Kallen 1994; Wells 1982; Hickey

Short vowels
kit //
A first, slight difference between Traveller English and General Irish English
can be found in kit words, which often have a more rounded and central
// in Traveller English than in RP or General Irish English, and can
approximate [] also in many stressed syllables. In contrast, other words
such as his, big or bit are lengthened to [i].
A peculiarity of Traveller English that may not be found in General Irish
English is the approximation of certain kit words towards strut in an
[]/[] realization: wrist [wrst], mirror [mr].
dress /e/

While kit words often become centralized in a [] or at least a [] sound,

words that would be in the dress category in RP, especially vowels followed
by a nasal such as den, Ben, then, but also in get, settled etc., often have a
raised [ e] or kit []. This is also a very common feature of General Irish

Irish Traveller English


trap //
The trap group displays a great deal of variation in Traveller English.
trap vowels are often more raised to [ ] and can approximate the openmid vowel [] in syllable-final position. Also the words many, any, which
most settled Irish realize as [] instead of the RP [] and which is seen as
a striking Irishism (Wells 1982: 423), are mostly pronounced with a more
raised [ ] by Irish Travellers.
lot //
As in General Irish English, LOT words frequently have the unrounded
variant [] in Traveller English, especially before nasal consonants (e.g.
long). Some words in this group are raised as far as to [], e.g. clock. In all
other instances it is usually []: got.
strut / /
Realizations of the strut vowels can be similar to the General Irish English
mid centralized back somewhat rounded vowel [ ] (Wells 1982: 422), an
intermediate between [] and [], as in bus [b s), summer [s m]. The
realization of [ ] can be found in some lexemes in Traveller English, but is
less pronounced than in General Irish English. Most realizations of strut
would either have [], which for some words may be influenced by the
spelling, e.g. come, done, other, but also pub, but etc. are pronounced with
an []. Some words have [ ] or [], e.g. in husband, run, and many others
are pronounced with a foot [], e.g. spuds.
As in General Irish English, some words that would have an onset strut
vowel in General English can be realized with a kit [i] in Traveller English,
e.g. onion [n].
foot //
Some words in the strut group have not even been lowered to // or / /,
but have a foot // vowel. This indicates that the footstrut Split has
not entirely taken place and words such as spuds, cut, bucket have an //,
resulting in some homophones with foot words, e.g. look and luck, which
can still be found in vernacular forms of General Irish English.
Many foot words have retained the historical /u/, which is also still
present in General Irish English, a retention of the Middle English /o/
which underwent the raising but not the shortening. Therefore, many
words of the foot group can be included in the mood group, such as


maria rieder

book, cook, crook (see also Wells 1982: 423). This feature is also present in
General Irish English, though it is becoming recessive.
A peculiarity of Traveller English absent in General Irish English is
that words with an onset <u> are often aspirated, e.g. us [hs], under
[h nd].
Weak vowels
Similar to General Irish English, Traveller English uses schwa extensively,
especially in unstressed word-final syllables, where /i/ and // are often
merged, e.g. happy [hp]. Words ending in -er either have an r-coloured
mid-central vowel: e.g. letter [lt] or, more commonly, the schwa
absorbed, e.g. better [btr]. These features are also found in General Irish
English, though especially the absorbed schwa would be more common
in Traveller English.
Words ending in -ow also commonly have a [], and can even be raised
to an []: follow [fl], which can be lengthened: window [wndi]. The
raising to [] or [i] is peculiar to Traveller English and may not be found
in General Irish English.
The -ing suffix is mostly reduced to /-n/ or /-n/ in Traveller English,
while schwa absorption (Wells 1982: 434) is common for words ending in
dentals: putting [ptn], sitting [stn]. Likewise the endings of morning
and evening are usually reduced to /n/. OSullivan (2008: 34ff.) studied
the reduction of the -ing suffix by the example of doing and going in a
comparison of the Limerick Corpus of Irish English and her own corpora of
Traveller English and found that this feature occurred in almost 100 per cent
of all -ing forms used by Irish Travellers, in contrast to about 7 per cent by
General Irish English speakers.
To in all meanings is usually weak in Traveller English and has either
a schwa, [t], or an unstressed front open vowel [ta]. Other, normally
stressed words such as I, what, when, and occasionally verbs like went
followed by a stressed preposition, are often used in their weak forms
with a schwa. Also the weak form of my [mi] is very commonly used.
Traveller English is very similar here to General Irish English, but again
there may be quantitative differences in regard to the articulation of weak
vowels and further quantitative research would be required for more precise
Unstressed prefixes of multisyllabic verbs are often not audible: remember
[memb], I decided [ai saidd]. This may occasionally occur in General
Irish English in connected speech, but is used very noticeably and consistently in Traveller English.

Irish Traveller English


3.1.2 Diphthongs
choice /a/
The realization of choice diphthongs is typically shifted to price /a/ as
in boy [ba], noise [nas], annoyed [anaid] and is more advanced/fronted
than the General Irish English /b/.
price /a/
This diphthong is unremarkable in Traveller English. While General Irish
English tends to neutralize the opposition /ai/ and /i/ by a low central
onset: Irish [r]  [r]  [ r]  [r], Traveller English usually shifts both diphthongs into the price direction with a slightly more
advanced/fronted onset: Traveller English Irish [ari], boy [ba].
mouth /o/
mouth diphthongs have close/mid-back onsets: /o/  // as in Traveller
English bouncer [bons]  [bns], in contrast to a low central onset
in General Irish English.
3.1.3 Centring diphthongs
The RP vowels //, // and // are absent in Traveller English as they are
in General Irish English, but in contrast to traditional or rural General Irish
English, Traveller English has only light rhoticity [], and often inserts a
schwa sound between the vowel and the following /r/: beer: General Irish
English [bir] Traveller English [bi] (pre-R breaking, Wells 1982:
213f.). At the same time, the subsequent process of pre-schwa laxing is
carried through, by which formerly long vowels [i, e, o, u] are shortened
to [, , , ]: beer: [b ]. Words in the near, square and cure groups
are pronounced in this way, with an unrounded front starting point
moving towards a mid-central position: [], [] and [] respectively.
start /a/
start words tend to be realized as // plus pre-R schwa: mark [mak] 
[mk] and may have a somewhat shorter vowel than General Irish English
for some members of the community. Travellers therefore seem not to have
appropriated the pre-fricative lengthening that was completed around the
end of the seventeenth century for RP (Wells 1982: 203ff.) and which
General Irish English seems to perform to a greater degree than Traveller


maria rieder

nurse //
The nurse merger is not completely carried through for Traveller English
and therefore displays a great variation. nurse words that were pronounced
as // or // before the merger was completed by the seventeenth century
(Wells 1982: 196), usually have an // realization, e.g. heard [hd], bird
[bd], in Traveller English, and may have a lengthened vowel as in Germany
[dmn]. Words that have an <i> are pronounced as []: girl [gl].
Words spelled with <o> tend towards an // sound: word [wd], and
those spelled with a <u> usually have a centralised // or // vowel, e.g.
curb [k b], turnip [t np].
north //
Like many other dialects north is merged with force in Traveller
English. Both vowel groups have a shorter vowel than General Irish English:
north [n].

Long monophthongs
fleece /i/ vs /e/
Pairs such as meet and meat, which were merged in the so-called fleece
merger by 1700, can still be distinct in General Irish English and
consistently so in Traveller English. Therefore meet and meat are not
homophonous in Traveller English: meat [met]  [m et]  [met], and
likewise eat [et], seat [set], tea [te]. Wells (1982: 195) explains in regard
to this phenomenon that rival pronunciations of the <ea> group were
current until well into the eighteenth century. One can argue that the Irish
Travellers as a very isolated and rural group would have preserved this
feature more than settled speakers, for whom it is becoming recessive and
restricted to rural areas (Wells 1982: 196).
face /e/
General Irish English face has not or only variably undergone the longmid diphthonging completed around 1800 (Wells 1982: 211), hence also
Travellers use mostly an /e/ vowel for the face group, though more
consistently than settled Irish people would, e.g. today [tde], name [nem].
Words ending in the /e/ diphthong typically approximate fleece /i/, as
in they [d i:], or // in say [s].
bath/palm //
The vowels /a/, //, // may not be distinct in Traveller English, and generally the bath/palm vowel can be slightly more raised in Traveller English

Irish Traveller English


than in General Irish English and RP depending on the environment.

Therefore, words such as calm, balm would have a vowel approximating
the RP vowel /a/, whereas man, Ann, tend towards //. Father, which in
General Irish English is often pronounced as [f], usually has a slightly
raised short []. Likewise, the vowels in dance, advance and similar words
are shortened and would therefore fall into the trap category.
thought //
The thought vowel is unremarkable in Traveller English, though it may
be a slightly shorter [] than in RP.
goat /o/
Travellers have preserved the traditional use of the monophthong /o/ for
//, which is, similar to the face group, a sign of the absence of long-mid
diphthonging. This is also a feature of General Irish English, though it
would be more consistent in Traveller English. A similarly recessive feature
in General Irish English, but widespread in Traveller English, is that some
of the goat words have a second variant with a mouth [a]: old [old] /
[ald], bold [bold] / [bald], which has a jocular and non-literal meaning
(Wells 1982: 427). [ald] has a sentimental connotation when talking about
times long gone by or affectionately about other people. Other words have
as their only realization an approximation towards an /a/ diphthong, e.g.
cold [kald], told [tald], shoulder [ald]. This feature is recessive in
General Irish English of today (OSullivan 2008: 48) but very present in
Traveller English.
mood /u/
The mood vowel is a very close, back long vowel in Traveller English.
As mentioned above, some RP foot words have a long /u:/ in Traveller
English: cook [kuk], book [buk]. This feature is becoming recessive in
General Irish English.
3.1.5 Consonants
As for vowel realizations, many consonantal features of General Irish
English vernacular that are already or are becoming recessive can still
be found extensively in Traveller English. A detailed comparative research
and analysis will be required for a clear picture as to the quantitative difference of usage between General Irish English and Traveller English of
the features summarized below. Those characteristics that may distinguish
Traveller English from General Irish English and may have different origins


maria rieder

are clearly marked and listed at the end of each subsection (examples from
Rieder, unpublished data): Alveolar and dental stops
Features common in both General Irish English and Traveller English (see
also e.g. Kallen 2005; Wells 1982), though more noticeable in Traveller
r Fortition of dental fricatives /, / to slightly aspirated dental plosives:

think [t k], that [d t ], therefore almost complete neutralization of

oppositions between []/[t] and []/[d]. The contrast may be maintained depending on factors such as word position, position of the
segment, and phonetics of the following segment by slight degrees of
aspiration and a more dental vs more alveolar articulation in order
to distinguish between minimal pairs such as thank [t k] and tank
r Sometimes further dentalization of plosives to aspirated alveolar plosives
or tapping: water [wt ]  [w];
r Lenition to a slit fricative (Wells 1982: 429) in word-final, postvocalic
position: hit [h t];
r Further lenition of /t/ to /h/ in some intervocalic positions: what it was
[hhwz], later [leh], I bought one [abhwon];
Peculiar to Traveller English:
r A peculiarity of Traveller English that may not be found in General

Irish English is the dentalization of plosives between two vowels: city

[st i].

Alveolar fricatives

r Alveolar fricatives /s, z/ tend to become palato-alveolar in word-final

position especially before approximants: it is yeah [hij], god bless you

[gdblj]. This feature is also common in General Irish English;
r Traveller English extends this rule to the dental fricative //, which tends
to become palato-alveolar in word-final position before approximants:
with you [wt j]. This feature is less frequent in General Irish English,
but can occur in connected speech.

The liquids

r Rhoticity or the presence of postvocalic and word-final //;

r Palatal rather than velar /l/.

Irish Traveller English


Both of these features are characteristic for General Irish English as well
as Traveller English. Other consonants
Common in both General Irish English and Traveller English, though
more noticeable in Traveller English:
r Historical retention of the aspirated glide cluster /hw/ or // for words

spelled <wh>;

r Schwa epenthesis in clusters consisting of a liquid and a nasal: film

[flm], harm [harm];

r Strong aspiration of word final /p, k/.

Peculiar to Traveller English:

r The nasal /m/ in word-final positions is often moved from its bilabial

to an alveolar position: from [frn]. This feature also occurs in General

Irish English, but usually only in connected speech. Traveller English
uses it consistently in sentence-final position as well.
3.1.6 Phonological processes
Common in both General Irish English and Traveller English, though
more frequent in Traveller English:

Yod coalescence in stressed syllables: did you [dd], tune [tun];

Yod dropping in unstressed syllables: education [edken];
Articulation of -ing forms as [n] or merely [n];
/h/ dropping in her [], him [m], humour [jum];
D epenthesis between an /r/ and a following /n/: different [difrdnd],
burn [brdn].
r /k/ dropping in /kt/ clusters: picture [pt];
Features peculiar to Traveller English:
r Metathesis of /sk/ clusters: ask [ks], which has disappeared in General

Irish English, but is still very common in Traveller English;

r Reduction of syllables in multisyllabic words: automatically [to

mkli]. This feature may not be found in General Irish English;

r Epenthetic /h/ in certain clusters: conversation [knvsen]. This

feature is absent in General Irish English.


maria rieder

3.1.7 Prosody
Most accountable for perceived differences between Traveller English and
General Irish English are probably Traveller English prosodic aspects such
as stress and intonation. Very limited study has been carried out in this field,
however, and therefore the following points are only tentative descriptions.
A combination of the above named features, especially the tendency of
kit, trap and strut vowels and some diphthongs towards a more midcentral realisation and the elimination of some unstressed syllables create
a singular rhythm, which can sometimes impede comprehension between
General Irish English and Traveller English speakers.
In terms of word stress, Traveller English varies a lot, but the main
stress in polysyllabic words often falls on the second-last syllable: washing
machine, she recognised me. Difference between stressed and unstressed
syllables may also be less marked than in General Irish English, with
secondary stresses in many polysyllabic words, which give Traveller English
a very strong rhythm: she recogNISED me.
The most recognizable feature of Traveller English is, however, intonation and pitch on the sentence level, which usually starts out with a very
high starting point falling to a lower level. Towards the end of a phrase
pitch rises again. The last stressed syllable in any phrase is lengthened and
marked by a slight fall of intonation:
(H) If I GET a qualifiCAT\ion now ( = )
(L) that I get a  GOOD STEAdy HOU\se;
(L) THEN Id have a  qualifCATion for /ME\
Questions follow the same pattern but go up again slightly in cases where
there is a fall on an unstressed syllable.
For a more concise picture of Traveller English prosody further quantitative and qualitative research is required, particularly the exact measuring
and computing of Traveller English speech rate, lexical tone and rhythm,
detailed intonation transcription and analysis of voice quality in comparison with General Irish English are highly desirable, as prosody is such an
important point of contrast between Traveller English and General Irish
3.1.8 Some conclusions about Traveller English phonology
The picture that emerges from the phonological analysis of Traveller
English reveals two main patterns. Firstly, Traveller English displays Irish
English characteristics that are recessive in General Irish English and would
be called Archaic Irish English. In terms of vowel realizations, Traveller

Irish Traveller English


English only partially seems to have adopted processes such as the Great
Vowel Shift processes, mergers and splits that would be associated with
modern English. The monophthongal quality of vowels in face and goat,
the low starting point of the diphthong in choice and the central starting point of the diphthong in mouth are, though not unique in the
English-speaking world, the most distinguishing features used consistently
in Traveller English, while mostly abandoned by General Irish English
speakers. Also the consonants are marked by Archaic Irish English characteristics. The segregated lifestyle of Travellers as well as learning English
mostly from rural people might be the reasons why Travellers have been
slower in adapting to the new standard and retained many Middle English
Besides aspects related to General Irish English, Traveller English also features characteristics that are not found in settled General Irish English, such
as cases of epenthesis, metathesis, reduction of syllables and the extending of General Irish English dentalization or alveolarization constraints.
However, the main aspects differentiating Traveller English from General
Irish English are found in prosody. The many weak vowels, reduced words,
and unpronounced unstressed syllables, as well as a singular intonation
pattern give Traveller English a distinctive and unique rhythm and sound


Also in terms of morphosyntax, Traveller English displays strong vernacular and Archaic Irish English features, which, according to Fordes (2005)
corpus-linguistic, lexico-grammatical analysis of modern Irish English
speakers, are slowly being abandoned by the settled Irish population. Irish
Travellers in turn, have held on to most features outlined in Forde and this
section will therefore align itself to his taxonomy, while also pointing out
some distinguishing characteristics of Traveller English.

The noun phrase Plural formation

Traveller English speakers usually avoid redundant plural marking of quantity nouns following a numeral:
(1) might be two or three time a year they go
(2) its about 25 mile back


maria rieder

While this redundancy rule is also still present in General Irish English
vernacular, Traveller English speakers often extend this constraint to nonredundant cases:

if you asked anythin else in year now gone by

The nonstandard use of quantifiers is also a feature that is distinctive in

Traveller English: lots of is often used with an indefinite article:

twas a lots o things there

Also absent in General Irish English is that many and much are frequently
used interchangeably and can be followed by a plural noun:

thatll tell ya how much crowds that was there Definite article

The nonstandard use of the definite article is still a distinguishing feature in
General Irish English (Forde 2005: 26) as well as in Traveller English. In four
of all categories of nonstandard definite article Traveller English deviates
mostly from General Irish English, for which these features are becoming
recessive. Differences between General Irish English and Traveller English
are therefore of a quantitative nature:
(a) Non-count concrete nouns are often found with a definite article:

theyd sit you at the table and give you the tea and the dinner

Forde, in his analysis of General Irish English regarding this feature found
that 5 out of 15 instances of non-count concrete nouns included a definite
article (Forde 2005: 27). Traveller English has this feature slightly more
frequently: 9 out of 15 cases of tea and dinner are preceded by the definite
(b) Same and both are usually expressed by the numerals one and

the two of them.


Cant and Gammon is all the one.

The word same is exclusively used in the sense of equal:


every woman is the same.

(c) The definite article is often used in a possessive sense:

(10) I just wonderin never see you an the husband

Irish Traveller English


(d) Very typical of Traveller English as well as of General Irish English

vernacular is the occasional use of a definite article with county names:

he loved the Clare

(12) that is the County Galway Pronouns
The pronoun systems are fairly standard. Subject pronouns follow general
patterns of most standard varieties: I, you, he/she/it for singular, we for
the first-person plural, the second-person plural pronoun is distinguished
from the singular by ye, and they is the third-person plural pronoun. Object
pronouns are unremarkable.
Reflexive pronouns can differ slightly from standard varieties as even
the plural pronouns are usually composed of a possessive adjective plus the
singular of self: meself, yourself, hisself, herself, ourself, yerself, theirself.
(13) the Travellers used that among theirself
The bold forms are absent in General Irish English and cannot be said to
be Archaic Irish English either. The first-person reflexive pronoun meself
could be seen as the weak form of the possessive adjective my, which is
commonly used on its own as well:
(14) I remember me poor father and mother now Adjective comparison
A feature that gives the impression of hypercorrection is the comparative
form of adjectives: Several instances in the data show a tendency towards
-er suffixation of adjectives that are already in their irregular comparative
form, e.g. lesser, worser. Sometimes already inflected forms are preceded
by a periphrastic more in a double comparative, which functions as an

its supposed to be more deeper

(16) so twas a simple life, and twas .. more happier

(17) its getting lesser as it goes
Both of these observations do occur in other dialects of English as well.
In comparison with settled Irish English speakers they occur much more
frequently in Traveller English than in General Irish English.


maria rieder Prepositions
Prepositions in Traveller English can often go unnoticed by the listener or
are indeed missing. Very similar prepositions, such as in and on can sound
identical due to centralisation of the weak vowels in both of these prepositions. In other cases conjoined syllables result in a missing preposition:
(18) I was goin up to my niece, she was livin the other side Oranmore.
(19) But then you could go other parts in the country.
Two other nonstandard usages of prepositions that are now very rare
in General Irish English can be found in the preposition on, which is
commonly replaced with of when referring to days of the week:
(20) we brought her of a Tuesday
(21) I dont eat meat of a Friday
and in the frequent intensification of in with inside:
(22) inside in the place

The verb group Irregular verb forms

Some Traveller English verb forms were found to display the same characteristics as irregular comparatives, where a regular suffix is added to an
existing irregular form. Examples are hurted, seened, growned and borned,
all of which are regularly used preterite forms. This phenomenon does not
occur in General Irish English.
Other verb features of Traveller English can also be observed in General
Irish English, even though less frequently (Forde 2005: 35). For example,
the reduced number of irregular forms: seed or seen, and done are commonly
used as a preterite:
(23) her daughters never seed her mother
(24) I seen the photos
(25) its a person that done somethin for the poor and the sick
The forms broke, lighten, wrote and went are used for preterite and as past
participles alike:
(26) it got broke
(27) all the candles lighten

Irish Traveller English


(28) weve already wrote our names on them now

Regular verbs have often lost the preterite or past participle suffix:
(29) later years then it start comin to feen, didnt it?
A further form of deletion is evident in some cases of existential sentences,
where the copular verb can be deleted:
(30) Im not able to go up to Winnies, that my niece Subjectverb agreement
Nonstandard subjectverb concord is one of the most striking and typical
features of Traveller English. As with many other features, some of the types
of nonstandard concord listed below can, though becoming recessive, still
be found in General Irish English and are also present in other dialects
of English. However, again Traveller English speakers perform them to a
greater quantitative degree than settled Irish English speakers would.
Most instances of nonstandard subjectverb agreement in Traveller
English fall under the Northern Subject Rule, a system of verbal concord widespread in northern English varieties, which states that the use
of the present-tense verb -s suffix can be extended to the first and second
persons singular and plural, except when the subject is a personal pronoun
that immediately precedes the verb. The Northern Subject Rule of verbal
concord therefore relies not only on features of person and number, but
also on the syntactic position and morphological features of the subject.
Examples of instances according to this rule in Traveller English are:
(31) different countries has different languages (Subject  Pronoun)
(32) they just thinks that they are just like everybody else (Subject =
Pronoun, but not adjacent)
When searching the corpus for present-tense verbs immediately following
the subjects my brothers, the people and Travellers, in 12 out of 27 cases the
verb had an -s suffix.
Apart from the Northern Subject Rule Traveller English deviates from
standard subjectverb concord by developments common in many other
English dialects, such as:
(a) the extensive use of the -s suffix as a marker of historic present when
introducing reported speech:
(33) I says some poor mouths waitn for it


maria rieder

(b) the equally widely used -s suffix as a marker of habitual and generic
present tense:
(34) thats where they does the Irish
(35) because they knows it
(c) the levelling of the contrast between was and were on was, as found
in many other varieties of English:
(36) my brothers was born in Offaly
(37) we was talkin in front of him
(d) the frequent deletion of the verbal -s suffix after third-person subjects,
which is also found in English dialects worldwide:
(38) he dont always use that language
(39) coffee dont make any difference to me Habitual aspect
From the three different traditional General Irish English ways of marking
habitual aspect (inflected do, inflected be, inflected do plus non-finite be;
Kallen 1994: 180) the corpus revealed one instance of the third one:
(40) I walk three mile every morning and I swear I do be dead after it
The do be habitual form can also be negated, which is very rare in General
Irish English but quite frequent in Traveller English:
(41) we dont be travelling now anymore Perfect aspect
The area of tense and aspect in General Irish English has been described
as being one of the most influenced by Gaelic (Forde 2005: 40). Irish does
not have a perfect tense, the expression of which is therefore substituted
by other means e.g. by loan-translations from Irish (Trudgill and Hannah
2002: 103). Structures such as the after perfect for recently completed
events, as in what I was after tellin ya, the extended-now perfect, as in he
is dead for many years now, the resultative/accomplishment perfect with
a split perfect, as in she has a good bit picked up now from him, and the
indefinite-anterior perfect, as in we never went there in years, are all features
of Traveller English that are still present in General Irish English, with a
possibly higher frequency in Traveller English.

Irish Traveller English


3.2.3 Complex sentences

Several more features that are associated with General Irish English vernacular are equally present in Traveller English:
r topicalization for reasons of contrast and reassertion: Its Travellers that

youre hearing;

r the construction for to + infinitive to express purpose: if they were doin

churnin the butter, you had to put your hands to the churn, for to put
luck on it;
r multiple negation: you cannot say nothn in Cant to the guards;
r retention of question-inversion and frequent lack of subordinator in
indirect questions: we can go to the library and see do they have any
books on that.
Two other features became evident in the Traveller English corpus which
are not found in General Irish English or Archaic Irish English:
r nonstandard negation: that mornin the pain not allowed me; everything

is not fitted well;

r the corpus also revealed three examples of questions without the stan-

dard subjectverb inversion: why she didnt come back?

3.2.4 Some conclusions
Morphosyntactic features of Traveller English draw a picture similar to
Traveller English phonology: on the one hand, many General Irish English
features that have been abandoned by settled speakers have been retained
by Travellers. On the other hand, Traveller English exhibits characteristics
that are not found in General Irish English and could be seen as extensions
of existing General Irish English constraints. Among these are cases of
plural formation, reflexive pronouns, subject-verb concord, variable word
order, nonstandard negation and most of all a phenomenon that Bliss
(1979: 284) described as regularity resulting from analogical reformation
of irregular forms, apparent in the formation of some irregular verb forms
and adjective comparatives.

Lexical and pragmatic features

Many Traveller English lexical features are found in nonstandard English

around the world, and are common features of General Irish English
vernacular, though they are becoming rarer for settled speakers. Examples
are the use of childer for children and the substitution of teach with learn,


maria rieder

as in she wants to learn us that. The corpus revealed 5 cases of standard use
of teach, but in 7 cases learn was used in the sense of teach.
Traveller English also makes wide use of nonstandard vocabulary and
General Irish English slang. Frequent terms would be baba child, baby,
yoke thing, object, spuds potatoes, young one child/person younger than
the speaker, often used in a superior way, my fellow my husband, my
small/young fellow my child, holy show a scene, spectacle, muppet fool.
Settled people are usually called buffers among Travellers, and the Cant
word pavee is used to refer to themselves. Other Cant words, such as lush
drink, lurk see, stall stop, beoir woman and feen man, have been
adopted into General Irish English slang and are frequently used by both
Some features can be considered exclusive to Traveller English or rarely
occurring in General Irish English. For example, mispronunciations of relatively modern words are frequent: ulcer is normally a homophone of ulster;
kilos is frequently pronounced as [kilgs] and traditional as [dinl],
which could be seen as a metathetic variant. Another commonly used
metathetic feature is found in the word ask, which is pronounced as [aks].
Also, several words are used with a slightly different meaning in Traveller English. Little, for example, is often used as a term of endearment,
meaning sweet, nice, cute. Especially when talking about people who
have had to experience misfortune or hardship, little is used to express
In the same way, the lexeme old-fashioned (pronounced as [aldfnd]
can have two semantic connotations in Traveller English: in a negative
sense it can describe badly educated, spoiled and rough children; used in
a positive way old-fashioned describes a clever, assertive and self-confident
The phrase god bless you (6 hits in the corpus) is frequently used to
express ones approval of someones (new-born) child:
(42) ann: lovely girl, oh god bless her
mary: bless her, int she?
Religious expressions in general are very frequent in Traveller English.
While in General Irish English god and oh my god are the most commonly
used religious expressions (OSullivan 2008: 44), Traveller English speakers
prefer oh Lord or God bless you for the above-mentioned meaning. Other
expressions frequently heard are the Lord have mercy on him/her (21 hits);
I swear to God (8 hits); with the help of God and our Blessed Mother (2
hits), God forgive me (1 hit). Traveller English is also rich in religion-related

Irish Traveller English


metaphors and colourful expressions such as he/she is a soul of a person or

he/she has a heart of gold.
Similar to Clancys (2011) findings in a comparative study of settled
speakers vs Travellers hedging, the corpus showed very little use of hedges,
such as like, actually, I think, etc. Clancy ascribes this to macro-social factors linked to socioeconomic and educational differences between the two
communities: the strength and primacy of their family network provides
Travellers with an assuredness of their position, which reduces the need
for Traveller family members to use hedges (Clancy 2011: 383). In contrast,
the settled community are characterised by a more individualistic ideology
and higher social mobility, which results in frequent family-external communicative situations where a higher amount of hedges is natural (Clancy
2011: 384f.).
A closer look at the type of hedges used by Travellers reveals an overwhelming use of you know in contrast to very few instances of I think.
Similar to you know hedges, other reassuring strategies such as directly
addressing their interlocutor or frequent expressions of solidarity point to
a preference of hedges that address the positive rather than negative face of
the interlocutor. Clancy (2011: 385) argues that this tendency is rooted in a
strong sense of community and serves to reinforce group bonds. However,
the absence of hedges that have a more assertive connotation in circles
of higher socioeconomic status may have a direct influence on [the Travellers] continuing marginalisation in modern-day Ireland (Clancy 2011:

4 Conclusion
Phonological, morphosyntactic, lexical and pragmatic features of Traveller
English have been shown to combine to create a variety of English that is
rich in unique cultural characteristics, while at the same time it displays a
great many Archaic Irish English features that are slowly being left behind
by the settled Irish population. In its distinctiveness from and similarities to General Irish English, Traveller English reflects the positioning of
Irish Travellers with regard to the settled Irish population. Despite their
indigenously Irish origin, a secluded way of life separate from mainstream
society as well as strong family ties and distinctive cultural aspects have
characterized Irish Travellers for centuries, and perpetuated their variety of
It needs to be mentioned though that Traveller English is not a homogeneous variety, and the degree to which the vernacular is spoken depends very


maria rieder

much on the level of education, accommodation and nature of networks of

the individual speaker. Until recently, the women in the community used
to be confined to their homes looking after a big family. This is now slowly
changing with women starting to look beyond the community boundaries
for work or free-time activities, which therefore opens up and loosens their
network ties. This development may have an impact on the use of vernacular norms and may bring about language change. OSullivans (2008: 55)
study of communicative shifts in Travellers casual speech revealed a certain degree of accommodation towards General Irish English with regard to
several pragmatic and morphosyntactic features, such as subjectverb concord, nonstandard negation, use of learn and teach, etc. However, Traveller
English phonological characteristics seem to be among the most resistant
to change. Pronunciation and intonation may also serve as a way to differentiate themselves favorably from the out-group in order to maintain a
positive social identity (OSullivan 2008: 14). After all, the Irish Travellers
are and perceive themselves as a separate cultural group. Years of denigration have led to a lot of opposition as well as the acquisition of a certain
pride, which may be symbolically expressed by linguistic separation and
the strong identification with their own variety of English.

Bakker, Peter and Maarten Mous, eds. 1994. Mixed Languages: 15 Case Studies in
Language Intertwining. Amsterdam: IFOTT/Foris.
Bhreathnach, Aoife. 2007. Confusing origins and histories: the case of Irish Travellers. Irish Journal of Anthropology 10(1): 305.
Binchy, Alice. 1994. Travellers language: a sociolinguistic perspective. In McCann
et al., eds., 13454.
1995. Travellers language. In A Heritage Ahead: Cultural Action and Travellers.
Dublin: Pavee Point Publications, 8793.
Baoill, eds., 1116.
2002. Travellers use of Shelta. In Kirk and O
Bliss, Alan. 1979. Spoken English in Ireland, 16001740. Dublin: Dolmen Press.
Britain, David, ed. 2007. Language in the British Isles. Cambridge University Press.
Browne, Marian. 2002. The syntactic structure of present-day Cant. In Kirk and
Baoill, eds., 6578.
Cauley, William. 2006. Canting with Cauley: A Glossary of Travellers Cant/
Gammon. Dublin: A. & A. Farmar.
Clancy, Brian. 2011. Complementary perspectives on hedging behaviour in family
discourse. International Journal of Corpus Linguistics 16(3): 37190.
Forde, Kieran. 2005. A corpus-based lexico-grammatical analysis of the Limerick
Corpus of Irish English (L-CIrish English): some nominal and verbal features.
Unpublished MA thesis, University of Limerick.

Irish Traveller English


Freese, Christoph. 1980. Zur Geschichte und Gegenwart der Zigeuner und Landfahrer in Deutschland: Versuch einer subkulturellen Erklarung. University of
Gmelch, Sharon and George Gmelch. 1976. The emergence of an ethnic group:
the Irish Tinkers. Anthropological Quarterly 49(4): 22538.
Gmelch, Sharon, Pat Langan and George Gmelch. 1975. Tinkers and Travellers.
Dublin: OBrien Press.
Grant, Anthony P. 1994. Shelta: the secret language of Irish Travellers viewed as a
mixed language. In Bakker and Mous, eds., 12350.
Hancock, Ian F. 1973. Shelta, a problem of classification. Journal of the Gypsy Lore
Society 52: 806.
1984. Shelta and Polari. In Trudgill, ed., 384403.
1986. The cryptolectal speech of the American roads: Traveler Cant and American Angloromani. American Speech 61(3): 20620.
Harper, Jared. 1973. Irish Traveler Cant in its social setting. Southern Folklore
Quarterly 37: 10114.
Helleiner, Jane Leslie. 2000. Irish Travellers: Racism and the Politics of Culture.
University of Toronto Press.
Hickey, Raymond. 2007a. Southern Irish English. In Britain, ed., 13551.
2007b. Irish English: History and Present-Day Forms. Cambridge University
Kallen, Jeffrey L. 1994. English in Ireland. In R. W. Burchfield, ed., The Cambridge
History of the English Language, vol. 5. Cambridge University Press, 14896.
1997. Focus on Ireland. Amsterdam and Philadelphia: John Benjamins.
2005. Internal and external factors in phonological convergence: the case of
English /t/ lenition. In Peter Auer, Frans Hinskens and Paul Kerswill,
eds., Dialect Change: Convergence and Divergence in European Languages.
Cambridge University Press, 5180.
Kallen, J. L. and J. M. Kirk. 2008. IC-Ireland: A Users Guide. Belfast: Clo Ollscoil
na Banrona.
Baoill, eds. 2002. Travellers and Their Language.
Kirk, John M. and Donall O
Belfast: Clo Ollscoil na Banrona.
Macalister, Stewart R. A. 1937. The Secret Languages of Ireland: With Special Reference to the Origin and Nature of the Shelta Language. Cambridge University
MacNeill, Eoin. 1937. Phases of Irish History. Dublin: M. H. Gill & Sons.
Sochain and Joseph Ruane, eds. 1994. Irish Travellers:
McCann, May, Seamas O
Culture and Ethnicity. Queens University of Belfast.
McDonagh, Michael. 1994. Nomadism in Irish Travellers identity. In McCann
et al., eds., 95109.
N Shuinear, Sinead. 1994. Irish Travellers, ethnicity and the origin question. In
McCann et al., eds., 5477.
Baoill, eds., 2141.
2002. The curious case of Shelta. In Kirk and O


maria rieder

hAodha, Mcheal. 2002. Travellers language: some Irish language perspectives.

Baoill, eds., 4763.
In Kirk and O
Baoill, Donall P. 1994. Travellers Cant language or register? In McCann
et al., eds., 15567.
OSullivan, Joan. 2008. Talkin Different: Linguistic Diversity and the Irish Traveller Minority. Newcastle upon Tyne: Cambridge Scholars.
Trudgill, Peter, ed. 1984. Language in the British Isles. Cambridge University Press.
Trudgill, Peter and Jean Hannah. 2002. International English: A Guide to the
Varieties of Standard English. London: Arnold.
Wells, J. C. 1982. Accents of English, vol. 2. Cambridge University Press.

part ii

The Americas

c h a p ter 5

American Indian English

Elizabeth L. Coggshall


American Indian English (hereafter AIE) is best described as a constellation of several varieties of English rather than a single variety. Because
the indigenous peoples of North America are not an undifferentiated,
monolithic group, it only follows that the English they speak is also not a
singular variety. But there is a relationship among various varieties of AIE,
both linguistically and socially, and it is this relationship that I focus on
in this chapter. AIE is spoken throughout the United States and Canada
primarily by people of indigenous heritage, and the speakers may number
as many as four million.1 However, the exact delineation of who is and
who isnt American Indian2 is not merely a social matter or based on a box
checked on the census, but rather, it is a long-fought political issue. As
Louise Erdrich (Chippewa) writes in her novel The Round House:
You cant tell if a person is an Indian from a set of fingerprints. You cant
tell from a name . . . You cant tell from a picture . . . From the governments
point of view, the only way you can tell an Indian is an Indian is to look at
that persons history. There must be ancestors from way back who signed
some document or were recorded as Indians by the US government, someone
identified as a member of a tribe. And then after that you have to look at that
persons blood quantum, how much Indian blood theyve got that belongs
to one tribe . . . On the other hand, Indians know other Indians without the

Based on the 2010 US Census (Humes, Jones, and Ramirez 2011) and the 2006 Canadian Census
(www.statcan.gc.ca/pub/89645-x/2010001/count-pop-denombrement-eng.htm); of the 4 million,
about 685,000 also identified as Hispanic or Latino on the US Census, meaning that they are
probably not in the population in question because they most likely originate from Central or South
America or the Caribbean.
A note on terms: The correct term to use when referring to the indigenous peoples above the Rio
Grande is a source of contention, usually between Native American and American Indian. For
this chapter, I use the two terms interchangeably. I use First Nations or aboriginal peoples when
discussing the indigenous people of Canada (the term First Nations does not include the Inuit or
Metis), and the name of the particular tribe when appropriate.



elizabeth l. coggshall
need for a federal pedigree, and this knowledge . . . has nothing to do with
the government. (2012: 2930)

For the purposes of this chapter, I leave the question of authenticity in

abeyance. But the tension described by Erdrich is an important part of AIE
use, as discussed in the conclusion.
The entertainment industry has painted a picture of the English spoken
by Native Americans that shares little resemblance with AIE. Barbra Meek
(2006) termed these representations as Hollywood Injun English and
includes such stereotypes as How! as a greeting, heap as an intensifier,
as well as subtler features such as lack of contractions and a slow pace that
suggests a lack of English fluency.3 Hollywood Injun English is not the
topic of this chapter, but it is important to acknowledge its place in the
popular consciousness and differentiate between this fiction and the facts
of AIE.
The academic study of language in Native American communities has a
long tradition. However, the study of AIE does not, hence its inclusion in
a book on under-studied varieties of English. Rather, the focus of linguistic
scholarship has been on the documentation and analysis of indigenous
languages (e.g. Boas 1911; Sapir 1933; Whorf 1941; Mithun 1999), the vast
majority of which have either become extinct or are now facing extinction
(Hinton 1999; Whalen and Simons 2012). This focus leads to the erasure
of American Indians who do not speak a language other than English, as
well as the English they do speak. Nonetheless, sociolinguists have long
called for more research on AIE, even if these calls have not been heeded.
Roger Shuy wrote in 1964: At the heart of the communication problem
of the American Indian is this question: in what way, if any, is the English
of the American Indian different from that of non-Indians of the same
relative social status and geographical environment? (52) This chapter is a
summation of some of the answers to Shuys question, as well as a further
call to action for more research.
Most varieties of AIE share some features that separate AIE from other
varieties of English. While many features, such as TH-stopping and copula
deletion, are common among nonstandard varieties of English, features
such as a small pitch range, insertions of glottal stops, and the variable
loss of the distinction between masculine and feminine in the third-person

Leechman and Hall (1955) suggest, based on historical documents, that features such as these were
actually once common and are indications of an earlier American Indian Pidgin English. See also
Miller (1967).

American Indian English


pronouns (he/she, him/her, his/hers) are rarely found in other varieties of

In what follows, I first demonstrate the status of AIE as a lesser-known
variety of English, as well as how AIE came about both as a variety and as
a code of indigenous identity. The challenge of my chapter is to provide
a unified description of these disparate varieties. To this end, I bring
together the various strands of research on AIE, showing commonalities
among different varieties of AIE as well as differences. Finally, I show the
need for greater research on AIE and general avenues of possible future

Sociolinguistic history and current status of

American Indian English

To understand AIE in its present form, it is important to understand the

history of colonization (and post-colonization) in the US and Canada.
The history of American Indians begins long before English was spoken in
the Americas. Before 1492 and the onset of major European colonization,
it is estimated that almost 300 languages were spoken from over fifty
language families, though this estimation is most likely low due to the
extinction of languages before they could be documented (Mithun 1999:
1). The reduction of linguistic diversity began almost immediately due to
the destruction of entire tribes through the introduction of disease and as a
consequence of warfare. Thus, many languages were lost due to the death
of all of their speakers, while others were lost when the remnants of tribes
joined together to create new communities (Maynor Lowery 2010). In at
least one case, the Lumbee of North Carolina, English became the language
of the newly formed tribe (Dannenberg 2002; Wolfram and Dannenberg
1999). It is estimated that there are only 209 indigenous languages extant
in the United States and Canada, all of which are endangered to a greater
or lesser extent (Whalen and Simons 2012: 1618). Twenty-two languages
have become extinct just since 1950 (Whalen and Simons 2012: 160).
For the most part, English has replaced these languages in indigenous
communities. There are exceptions, of course, like the Yaqui of Arizona who
use Spanish (e.g. Dozier 1956; Trujillo 1997) or the Metis of Saskatchewan,
Manitoba, and North Dakota who speak Michif, a mixed language of
French and Cree (e.g. Bakker and Papen 1997). Many other tribes adopted
a creole language, though few are still in use. Those creoles include Chinook
Jargon spoken from southern Alaska to Northern California (e.g. Silverstein
1972; Mithun 1999: 5879); Delaware Pidgin in New Jersey, New York, and


elizabeth l. coggshall

Pennsylvania (e.g. Goddard 1997; Mithun 1999: 5902); and Mobilian

Jargon from Florida to Texas to Illinois (e.g. Drechsel 1983; Mithun 1999:
As stated above, not all of the heritage languages are extinct, and these
extant languages are spoken by multilingual speakers (though there are a
very few monolingual speakers of heritage languages among the elderly
in some few communities) in communities across the US and Canada,
and there are still children whose first language is not English but the
heritage language (as has been my experience with the Navajo in Arizona
and the Eastern Cherokee in North Carolina), though, like monolingual speakers, they are not numerous. Speakers of indigenous languages
in the US and Canada number around 230,000 (Whalen and Simons
In some communities, AIE is the only Indian-related language tradition that speakers have access to, since it is the only one that is left; in
cases like this, fluency in AIE is an important social skill, its features having acquired heightened social significance (Leap 1993: 3, emphasis in the
original). But not all speakers of AIE are mono-dialectal; many speak a
more standard variety of English as well, and the use of one variety or the
other is often a political decision (Leap 1993: 4).
The reasons behind this language shift are many, and differ from community to community, thus a brief summation of the history of contact
between Native Americans and Europeans (specifically the English4 and
the English-speaking descendants) north of the Rio Grande is in order.
Prior to contact, there were between two and eighteen million people in
what is now Canada and the United States (Calloway 2004: 13). Then,
Christopher Columbus discovered the Americas in 1492 (the English did
not set up their first permanent settlement, Jamestown, until 1607). Disease, enslavement, and the destruction of communities (and the languages
they spoke) started more or less immediately upon landing. Diseases were
especially pernicious in that epidemics would often decimate populations
well in advance of European settlement, both giving the impression that
the land was virgin and free for the taking, and destroying languages
(and other cultural components) before they could be documented. Wars
between the indigenous peoples and the colonizers were also common well
into the nineteenth century. Disease, while not in epidemic proportions,
still plagues indigenous communities.

Besides the English, the Spanish, French, Russians, and Dutch also had a presence in the areas under

American Indian English


After 1776, Indian affairs became the burden of the newly formed US
government, which assigned this task to the War Department, though it
moved to the Departments of the Interior in 1849. Today, the department in
charge of Indian policy is the Bureau of Indian Affairs (BIA). The wars and
epidemics continued, and the US government signed several treaties with
different tribes. American Indians were not US citizens but instead existed
in a no-mans land between sovereign nations and regular citizens; they
finally became full citizens in 1924. These treaties became important later
to determine who counts as American Indian for government purposes, as
explained above.
In 1830, Congress passed the Indian Removal Act, allowing for the forced
relocation of the eastern tribes to Indian Territory in the west. Most
notable of these removals was the Trail of Tears in 1838, where approximately
4,000 Cherokees died on a forced march to present-day Oklahoma from the
Carolinas and Georgia. The reservation system that is still in existence in the
United States started in the 1850s. The reservation system was always seen
as suboptimal by the federal government, for whom the ultimate goal was
the civilization (i.e. assimilation) of Native Americans. In keeping with
this goal, in 1877, Congress appropriated the first funds to create schools
for Native American children (Calloway 2004: 345). Like the children of
European immigrants, Indian children were expected to jettison their old
ways and language and become English-speaking Americans (Calloway
2004: 344). Boarding schools were typically placed far from the home
reservations of the children, who were (often forcibly) removed from their
families. In 1887, an English-only policy was instituted (Fear 1980: 14), and
speaking a heritage language was grounds for (often corporal) punishment.
Most boarding schools separated children from the same tribe as a strategy
to force childrens use of English. This language education was part of
a larger plan of transformation to remake them as individual citizens,
not tribal members (Calloway 2004: 347). The boarding schools had an
impact on Native American language far outside their walls: graduates of
such institutions feared teaching their children anything other than English
lest these children would have to endure the same humiliation and torture
that their parents had endured.
The federal government did much more in the interest of civilizing
the savages, though many of the government officials in charge of Indian
policy were convinced they were doing what was best for a beleaguered
people. In 1887, in order to show Native Americans in the US the importance of private ownership of land, the Allotment Act was passed, which
divided many reservations into parcels of land that the American Indian


elizabeth l. coggshall

owners could retain or, in many cases, sell to non-Indians. In 1934, the
Indian Reorganization Act was passed that set up self-governing bodies on
many reservations. Soon after, however, Indian policy in the US turned
toward the termination of reservations and the special status of American
Indians. To this end, the Indian Claims Commission was set up in 1946 to
buy back land from tribes, and in 1956, a system of relocation was set up
to move Native Americans from rural reservations to urban centers. There
is an ironic consequence of the policies discussed thus far: the goal of the
termination and relocation policies, along with the English-only boarding
schools, had been to end the Indian problem through assimilation. But
instead of individuals ties to Indian identity being weakened, a panethnic
identity was created by these pressures, and the population of people claiming American Indian heritage grew rather than shrank (Lopez and Espiritu
1990; Nagel 1996; Nagel and Snipp 1993). In 1968, the Indian Civil Rights
Act was passed, and in 1975, Richard Nixon declared that the federal government would now have a policy of self-determination, allowing tribes to
do what they think is best for themselves, with certain restrictions based on
the fact that they are not sovereign nations but part of the United States.
The Canadian governments history with aboriginal peoples followed
a similar trajectory (see Dickason 2006), with most likely similar results
as far as their use of English is concerned. Waves of epidemics and wars
swept through the First Nations upon contact (and for hundreds of years),
decimating much of the indigenous population. Much of the initial contact
was made by French fur traders in search of beaver pelts, rather than by
English-speaking settlers moving west in search of land as was the case
in the US. A policy of assimilation gained momentum in the nineteenth
century, mostly based around what is called the Indian Act, passed in 1876.
Reserves were created for First Nations to live on, and then they were broken
apart for individual allotments. The Indian Act created Indian residential
schools to speed the acquisition of civilization, and used force to push
an English-only environment. Policies of denying religious and linguistic
freedom, as well as rights to use of land for hunting were maintained
into the twentieth century; aboriginal peoples were not given the right to
vote until 1956. In the 1969, the Minister of Indian Affairs suggested the
abolition of the Indian Act and thus the special status held by First Nations,
but the relevant legislation has never been passed. Since then, First Nations
have been continuing the struggle for civil rights and representation in the
The history of contact in the United States and Canada, as well as the
sorts of policies the governments of the two countries enacted, had a large

American Indian English


impact on language use in Native American communities. Most obvious is

the extinction, or near extinction, of the diverse languages spoken before
1492. The adoption of English may seem a straightforward reaction to
the events discussed above, but not all American Indians have become
monolingual English speakers, and those who are English speakers do not
necessarily speak a standard version of American or Canadian English.
The assimilation that these policies were meant to accomplish has not
been successful, and, in the last few decades, has been replaced by promulticultural policies and attitudes in the general public. This lack of
assimilation allowed for the creation of separate varieties of English, ones
that often replaced heritage languages as a locus for speakers to express
American Indian identity.
AIE fits most of the criteria of a lesser-known variety (LKVE) of English
as put forth by Schreier et al. (2010: 4) in their introduction. AIE is most
definitely a variety that is lesser known, at least to the outside world
(Schreier et al. 2010: 4) since much of present-day American Indian culture
is not understood or known to non-Indians; further, AIE has not received
much attention in the literature (Schreier et al. 2010: 4), especially in the
last two decades, happening to coincide with the publication of American
Indian English by William Leap in 1993. Before that date, in the 1970s and
early 1980s, the vast majority of the work was done by a small number of
scholars: Leap, H. Guillermo Bartelt, and Susan Penfield Jasper. Of these
three scholars, only Bartelt continues to work on AIE.
AIE fits many of the more specific criteria of an LKVE as well. AIE is
the first language of most of its speakers, and sometimes the only code
they know. AIE is considered its own variety, separate from those spoken
by non-Indians, the variety spoken in school, and other varieties spoken
by AIE speakers (Leap 1993). AIE is spoken by people in stable communities, many of these communities being several hundred if not thousands
of years old. AIE is spoken by minorities, American Indians and First
Nations being some of the most marginalized people, both socially and
geographically, in the US and Canada. They are among the poorest people, they are ethnically separate, and for generations they have mostly lived
in reservations or reserves far outside population centers. They have also
resisted outsiders attempts to assimilate them into the mainstream, such
as allotment, termination, and relocation. AIE was not really transmitted by settler communities or adopted by newly-formed social groups
(Schreier et al. 2010: 4) in that many varieties are not directly derived from
British English during colonial times but have more of a connection with
American and Canadian English. Though there is some controversy over


elizabeth l. coggshall

the exact genesis of AIE, no one would deny that language contact had a
hand in its formation. AIE is also an important identity marker for many
of its speakers. It is the only Native American-related code available to
them because of the lack of an indigenous language in the community
either through the eradication or slow depletion of the language. AIE is
also a way for American Indians to identify each other, and some even
claim they can tell what tribe someone is from based on their English.
AIE is not, however, endangered. I argue that as long as American Indians exist, there will be AIE. I do so for two reasons. First, the isolation
described above, especially the geographic isolation, promotes the existence
of a separate code for American Indians. Second, since the rise of the Red
Power movement in the 1960s/70s, Native American identity has earned
cachet, and so any identity markers associated with it will most likely be
While AIE is a lesser-known variety of English, some studies of AIE
in the US (but none, to my knowledge, in Canada) have been done, but
much of the work that was done was based on data gathered in the 1960s
and 1970s. The goal of much of this work was to ascertain the provenance
of these features, be it interference from the indigenous language (e.g.
Wolfram 1980), influence from other nonstandard varieties (e.g. Dillard
1972), fossilized features of second language learning (e.g. Leap 1974), or
remnants from a past English-lexifier creole or pidgin (e.g. Craig 1991).
Further, most of the research was funded by institutions, like the National
Institute of Education in the case of Wolfram, Christian, Leap, and Potter
(1979), that were primarily concerned with the educational application of
linguistic research.

3 Features of the variety

Because AIE is not one variety of English but instead several varieties united
by the ethnic character of its speakers, many features are unique to specific
varieties. The focus of the present study is not features such as these but
rather the large set of features that are shared among at least some varieties
of AIE. Scholars have noted that there are several commonalities that run
through most, if not all, varieties of AIE (Wolfram 1980; Leap 1982; Craig
1991). Wolfram (1980: 387) suggests that the common features, like those
discussed below, may unite American Indian English in opposition to
other non-mainstream varieties of English. The fact that many features
are shared among varieties raises questions about the origins of these features. The causes of these similarities have been considered, though no

American Indian English


definitive conclusion has been reached. Some believe that there was once a
creole spoken by American Indians, possibly learned from escaped slaves,
and that these features are the result of decreolization (Craig 1991; Dillard
1972). Others believe that a common source of English boarding and residential schools explains these shared features (Harvey 1974; Spicer 1967;
cf. Malancon and Malancon 1977). Many of these features are also common
in the English spoken by second language learners and thus may be the
result of fossilization of such features. Finally, some of these features fall
into what Chambers (2004) refers to as vernacular universals, where these
features are common throughout most nonstandard varieties of English, a
point made about AIE by Leap (1977c). Of the features discussed below,
Chambers theorizes that consonant cluster reduction, subjectverb nonconcord, negative concord, and null copula are vernacular universals (2004:
129). However, Leap warns, Even if two Indian English varieties appear
to share certain surface-level features in common, we cannot automatically
assume . . . that similar constructions are governed by similar underlying
causes (1982: 3).
This is not to say that all of the features discussed below are shared by
non-Indians. As Leap sums it up: in some cases, their linguistic details are
quite similar to those found in the English of their non-Indian neighbors,
coworkers, and classmates. More commonly, Indian English shows extensive influence from the speakers ancestral (or native) language tradition(s)
or from other language sources and differs accordingly from non-Indian
notions of standard grammar and appropriate speech (Leap 1993: 1).
Varieties of AIE vary on how much they resemble non-Indian varieties;
some are quite similar to those spoken by non-Indian neighbors, others
heavily influenced by non-English language traditions.
I start by looking at phonological patterns in AIE, then moving onto
morphosyntax, perhaps the best-studied aspect of these varieties. I then
briefly discuss some of the issues of the lexicon, and finish up with a
discussion of some pragmatic features of AIE.


The phonology of AIE has not been studied as extensively as the syntax. Many of the unique/interesting features of the phonology are related
to ancestral language traditions. However, recent research suggests that
some features, namely the extensive use of glottal stops (Rowicka 2005)
and a syllable-timed prosodic rhythm (Coggshall 2008), appear to have
spread from some varieties of AIE to other varieties. These findings lead to


elizabeth l. coggshall

some interesting questions that should be addressed in future research, as

described in the conclusion below.
3.1.1 Consonants
A noteworthy aspect of consonants in AIE involves differences between
these varieties and non-Indian varieties of English in consonant inventory.
Below, two such examples are discussed that occur in many varieties of AIE:
TH-stopping, where the interdental fricatives are lost from the inventory,
and glottal stops, an addition of a sound. There are more such examples
of differences in consonant inventory. For instance, a loss of consonant
distinctions can be found in Navajo English, Pima English, and Tsimshian
English, with the interdental fricatives /f, v/ replaced by the voiced bilabial stop [b] (Cook and Sharp 1966: 24; Nelson- Barber 1982: 125; Mulder
1982: 100). Further, the set of alveolar and palato-alveolar fricatives has
undergone changes in several varieties. In Tsimshian English and Pima
English, /z, , / are realized as [s] (Cook and Sharp 1966: 24; NelsonBarber 1982: 124), and Kotzebue English lacks a distinction between these
four phones, using them interchangeably (Vandergriff 1982: 1389). Some
varieties have consonants that are not found in non-Indian varieties of
English. For instance, Quinault English has a labialized voiceless velar fricative [xw ] in words beginning with wh- (Rowicka 2005: 307), and Kotzebue
English has a voiced velar fricative [] that replaces /g/ (Vandergriff 1982:
The other process involving consonants that is common to many varieties of AIE is consonant cluster reduction (CCR). CCR occurs usually,
but not always, at the end of words. The reduction refers to the deletion of
whole segments, like han for hand or des for desk. CCR is noted in almost
every description of varieties of AIE. CCR has been documented in the
Quinault (Rowicka 2005), Isletan (Leap 1977a), Mohave (Penfield 1977),
Hopi (Penfield 1977), Navajo (Penfield 1977), Cheyenne (Alford 1974), San
Juan (Wolfram et al. 1979; Wolfram 1980), Laguna (Wolfram et al. 1979;
Wolfram 1980), Lakota (Flanigan 1984, 1985), Lumbee (Torbert 2001), and
Brandywine (Gilbert 1986) varieties of AIE, as well as many non-Indian
varieties of English. TH-stopping
TH-stopping, that is, a stop where a standard variety would have one of
the interdental fricatives /, /, is one of the most commonly cited features
of varieties of AIE. The stops in question are dental or, less commonly,
alveolar. For instance, in Tsmshian English the is pronounced [d] with

American Indian English


the voiced alveolar stop, and northland is pronounced [nortlnd] with the
voiceless alveolar stop (Mulder 1982: 100). The fricatives are not always
replaced by stops; in Western Apache English and Navajo English, while
// is always the dental stop [d ] and word-initial // becomes [d ], word-final
and word-medial // become [f] (Bartelt 1986: 692). TH-stopping has also
been documented in the Brandywine (Gilbert 1986), Kotzebue (Vandergriff
1982), Alabama-Coushatta (Hoffer 1982), Hopi (Penfield 1977), and Pima
(Miller 1977) varieties of AIE. Glottal stops
Many varieties of AIE display extensive use of glottal stops, [], particularly
where oral stops are found in other varieties of English. For some varieties,
glottal stops only replace the voiceless stops /p, t, k/, e.g. Quinault English
(Rowicka 2005), while others allow glottal stops to replace voiced stops
/b, d, g/ as well, e.g. Navajo English (Penfield 1977). Also, depending on the
variety, these replacements can occur both word-medially and word-finally
or only word-finally. In Quinault English, utterances such as Wha[]s
u[]? for whats up and po[]latch for potlatch (Rowicka 2005: 31617)
are common, and, likewise in Navajo English, [bi] for big and [ga] for
god (Penfield 1977: 31). Some varieties, e.g. Pima English (Nelson-Barber
1982), also insert glottal stops after a consonant, such as [pIg] for pig
or [kek] for cake. Bartelt (1986) hypothesizes that the use of [] adds
to the choppy property of AIE. This kind of use of [] has also been
documented in the Mohave (Penfield 1977), Hopi (Penfield 1977), San Juan
(Stout 1977), Pima (Miller 1977), and Cheyenne (Alford 1974) varieties of
3.1.2 Vowels
Unlike the patterns noted with consonants, there are few commonalities between different varieties of AIE in regard to vowels. While AIE
vowels are definitely different from those in other varieties of English,
they also vary from variety to variety. In some varieties, such as Eastern
Cherokee English (Anderson 1999, Coggshall 2006) and Lumbee English
(Schilling-Estes 2004; Coggshall 2006), the vowels used by AIE speakers
and speakers of non-Indian English are very similar and show little influence from the substrate language. For instance, Eastern Cherokee English
has an extremely fronted goose vowel in line with the regional standard,
Appalachian English, even though the Eastern Cherokee language has a
fully backed goose vowel (Coggshall 2006: 59). Other varieties of AIE
show more influence from the substrate. For instance, Quinault English


elizabeth l. coggshall

(Rowicka 2005: 3089) and Navajo English (Cook and Sharp 1966: 223;
Bartelt 1986: 692) both lack glides on face and goat vowels due to a similar lack in Quinault and Navajo. Other varieties such as Navajo English
(Cook and Sharp 1966: 223) and Isletan English (Leap 1993: 46) lose
distinctions between tense and lax vowels.
3.1.3 Prosody
The suprasegmental phonology of AIE is one of the most marked aspects
of the varieties, as well as one of the least studied. Leap hypothesizes
that suprasegmental features contribute substantially to contrasts with
standard English and to contrasts that distinguish Indian English codes
from different tribal communities (Leap 1993: 50). These differences lead
to an impression that AIE speakers talk in more subdued tones, show little
expression or emotion in their voices, speak in a monotone, or speak in
sing song voice (Leap 1993: 52). Such impressions can be linked at least
in part to the prosodic features explored below: a smaller pitch range, high
rising terminal, and syllable timing.
Penfield (1977), in her work with speakers of Mohave English, Navajo
English, and Hopi English, found that they spoke with few pitch changes
and thus display a smaller pitch range than speakers of other varieties
of English. They remained level even when asking questions; in nonIndian varieties, questions usually have a rise in pitch at the end. In contrast to the varieties described by Penfield, other varieties have a high
rising terminal on declarative sentences. That is, there is a higher pitch
at the end of the sentence than in the rest of the sentence. This feature
has been attested in the Alabama-Coushatta (Hoffer 1982), Brandywine
(Gilbert 1986), and Tsimshian (Mulder 1982) varieties of AIE. Another
salient aspect of AIE prosody is rhythm, i.e. the relative length of adjacent syllables. Most non-Indian varieties of English have a stress-timed
rhythm such that stressed syllables are lengthened and unstressed ones
reduced, resulting in a difference in syllable length between the two types.
The resulting pattern has been likened to Morse code; it is found in
Germanic languages generally. In contrast, many varieties of AIE have
a syllable-timed rhythm, similar to that which occurs in most Romance
languages. That is, all syllables are of similar length, creating a rhythm
that is compared to a metronome. Syllable timing has been attested
to in Kotzebue (Vandergriff 1982), Brandywine (Gilbert 1986), Lumbee
(Coggshall 2008), and Eastern Cherokee (Coggshall 2008) varieties of

American Indian English




The sentence and word structure of several varieties of AIE (namely Isletan, Lakota, Lumbee, Mohave, Laguna, and San Juan) have been studied
extensively, and several of these structures are discussed below. I start with
variation in copula usage, and then move to other verbs, specifically to
features dealing with tense and aspect, as well as agreement. A short discussion of variation in nouns and pronouns follows, along with negative
concord. Variation among pronoun usage as well as deletion of pronouns
is explored after that, and I finish with a look at nonstandard word order
in AIE.
3.2.1 The copula
For some varieties of AIE, the copula is optional, as shown in (1) and (2).
For the sake of this discussion, and following Labov (1969), all instances of
to be, whether as copula or auxiliary, are discussed here.
(1) Papago (Tohono Oodham): But they __ going fishing. (Bayles and
Harris 1982: 6)
(2) Lakota English: This __ my grandpa. (Flanigan 1984: 85)
This feature has also been attested in Haliwa-Saponi (Hazen 2002), Lumbee (Dannenberg 1999), Tsimshian (Mulder 1982), Tlingit-Haida (Chessin
and Aurbach 1982), Hoopa (Chessin and Aurbach 1982), Isletan (Leap
1993), Mohave (Penfield Jasper 1980), Tohono Oodham (Bayles and Harris 1982), and Ute (Leap 1993) varieties of AIE. Other auxiliaries, especially
do, are sometimes deleted as well (Vandergriff 1982; Leap 1993).
When present, the copula can be uninflected, as in (3) and (4). While
uninflected be is often used as a habitual marker much like it is in African
American English, Leap (1974, 1977c) argues this is not always the case, as
demonstrated in (5).
(3) Eastern Cherokee English: I dont believe a woman be out there,
floppin around using ball sticks like that. (Coggshall 2005)
(4) Lakota English: They be goin home (Flanigan 1984: 85)
(5) Isletan English: I be home soon. (Leap 1977c: 83)
Uninflected be has also been documented in San Juan (Wolfram 1984),
Lumbee (Dannenberg and Wolfram 1998), Yakima (Chessin and Aurbach
1982), and Laguna (Wolfram 1984) varieties of AIE.


elizabeth l. coggshall

3.2.2 Verbs
The verbal systems of varieties of AIE are the locus of the most extensive
differences between AIE and non-Indian varieties of English. In particular, the aspectual and agreement systems are markedly different from
those found in non-Indian varieties; moreover, AIE is characterized by the
deletion of many inflectional morphemes. Tense and aspect
There are several contributing factors to the differences between AIE and
non-Indian varieties of English, the two most important being the deletion
of inflectional morphemes in varieties of AIE and influence from the
aspectual systems of indigenous languages. A common view on the latter
influence is that many of the ancestral languages of AIE speakers rely more
on aspect than on tense, which is the opposite of standard varieties of
English; thus, the substrate aspectual system leaves its imprint on AIE
in the form of larger, more extensive aspect systems. A few examples of
nonstandard inflection are illustrated below.
Moreover, no two AIE varieties have the same tense/aspect system. Leap
(1993: 637) has many examples of these differences. For instance, Isletan
English makes distinctions between delimited, distributive, and continuous verbs. This can be contrasted to Cheyenne English, which makes a
distinction between manifest and nonmanifest actions (Alford 1974: 6),
or Kotzebue English and its system set up around the completedness of
actions rather than how an event unfolds (Vandergriff 1982: 1308).
The ways in which varieties use the -ing suffix constitute a common
difference between varieties of AIE and non-Indian varieties. Sometimes
the suffix is used on verbs that are not in the progressive tense, as in (6).
Bartelt (1986) hypothesizes that in these cases -ing is marking non-punctual
(6) Navajo: I live by the beliefs that coming from both the Navajo culture
and Christianity. (Bartelt 1985: 501)
In other cases that are progressive, the -ing suffix is dropped, as in (7)
and (8).
(7) Lakota English: Our childrens are start, you know, really mixing it up.
(Flanigan 1985: 222)
(8) Yakima English: Youre going to be the one bring the money home.
(Chessin and Aurbach 1982: 180)

American Indian English


The most studied feature of tense in AIE is unmarked tense, where verbs
that are meant to describe something that happened in the past lack any
indication of past tense. Some of the instances of absence of overt tense
marking may be due to CCR, where the past-tense suffix of weak verbs is
deleted due to this phonological constraint, as shown in (9) and (10):

Mohave English: One guy got mash_ bad. (Penfield Jasper 1980: 82)

(10) Tohono Oodham English: His shirt got unzipper. (Bayles and Harris
1982: 6)
However, there are other examples where CCR cannot explain the lack
of tense, as in (11)(13).

San Juan English: Remember the time they fight for Unge. (Wolfram
et al. 1979: 49)

(12) Eastern Cherokee English: Youd go to the home they all speak in
Cherokee. (Coggshall 2005)
(13) Lakota English: He begin to look for her. (Flanigan 1985: 225)
Unmarked tense has also been documented in the Brandywine (Gilbert
1986), Laguna (Wolfram 1980, 1984, 1986; Wolfram et al. 1979), Western
Apache (Bartelt 1985, 1986), Isletan (Leap 1993), Quinault (Rowicka 2005),
Nisqually (Chessin and Aurbach 1982), Colville (Chessin and Aurbach
1982), and Navajo (Cook and Sharp 1966) varieties of AIE. Subjectverb nonconcord
In many varieties of AIE, verbs do not always display the agreement pattern
that shows up in standard English. This occurs in sentences with both plural
and singular subjects. Some of this lack of agreement may be due to CCR,
which would cause the inflectional -s ending on the verb to be lost due to
consonant deletion, as in (14), though, of course, it is impossible to tell if
CCR or another process is at work.
(14) Isletan English: This traditional Indian ritual that take place in June.
(Leap 1977d: 123)
Other examples, however, are clearly not the result of a phonological
process. In particular, forms of be, in both present and past tense, frequently
fail to display the agreement patterns of standard English, as shown in (15)
(16). Some varieties pattern along the lines of African American English,
as in Hoopa (17) and Lumbee (18) English, while others do not.


elizabeth l. coggshall
Hoopa English: Drugs is what is happening today. (Chessin and
Aurbach 1982: 179)

(16) Lumbee English: The barges was on the other side. (Wolfram and
Sellers 1999: 97)
(17) Isletan English: By this time, this one side that are fast have overlapped. (Leap 1977d: 123)
(18) Navajo English: I were looking for deer. (Cook and Sharp 1966: 27)
Another common verb subject to nonconcordance is to do, as in (19) and
(19) Lakota English: My brother, he do that every day. (Flanigan 1984:
(20) Navajo English: She dont know how to hold on to the horses. (Cook
and Sharp 1966: 27)
The subjectverb nonconcord discussed here is also found in the Quinault (Rowicka 2005), Laguna (Wolfram et al. 1979, Stout 1979), Eastern
Cherokee (Coggshall 2005), Mohave (Penfield Jasper 1980), and San Juan
(Wolfram et al. 1979) varieties of AIE, as well as many non-Indian varieties
of English.
3.2.3 Nouns
Inflectional endings on nouns, namely plural (21) and (22) and possessive
-s (23), are optional in varieties of AIE. The information contained in
these morphemes is often expressed instead through other means, such as
through overt expressions of number elsewhere in the sentence or by word
(21) Quinault English: My nephew they are dead people. (Rowicka
2005: 309)
(22) San Juan English: Three other place we went. (Wolfram et al. 1979:
(23) Navajo English: my sister husband, Jack father, my grandma house
(Cook and Sharp 1966: 25)
These features are also documented in the Mohave (Penfield Jasper 1980),
Laguna (Wolfram et al. 1979), Lakota (Flanigan 1984), and Brandywine
(Gilbert 1986) varieties of AIE.

American Indian English


The grammatical distinctions between count and mass nouns can be

lost, as shown in (24), where the count noun horse receives the modifier
much, which in most non-Indian varieties of English can only be attached
to mass nouns. (25) shows the opposite, where the mass noun pottery has
the plural morpheme typically reserved for count nouns.
(24) Lakota English: We ride much horses. (Flanigan 1985: 223)
(25) Mohave English: Theres a lot of potteries around there. (Penfield
Jasper 1980: 89)
This feature has also been attested in Koyukon English (Kwachka 1988).
3.2.4 Pronouns
Some varieties of AIE behave as pro-drop languages, where pronouns are
optional, as shown in (26) and (27).
(26) Hoopa English: Now when __ hear that some of my friends are
getting married, it a sad occasion. (Chessin and Aurbach 1982: 117)
(27) Lummi English: Something should be done to make __ possible.
(Chessin and Aurbach 1982: 117)
Pro-drop is also found in the Tlingit-Haida (Chessin and Aurbach 1982),
Swinomish (Chessin and Aurbach 1982), Tsimshian (Mulder 1982), Ute
(Leap 1993), and Mohave (Penfield Jasper 1980) varieties of AIE.
A particularly marked feature of many varieties of AIE is the variable
loss of gender distinction in the third person singular pronouns, he, she,
him, her, his, and hers. This leads to utterances such as (28) and (29) where
the pronoun does not have the same gender as its antecedent.
(28) Tohono Oodham English: The boys zipper got caught in her jacket.
(Bayles and Harris 1982: 6)
(29) Mohave English: My aunt plants corn in his own garden. (Penfield
Jasper 1980: 75)
This feature has also been documented in the Navajo (Cook and
Sharp 1966), Tsimshian (Mulder 1982), Lakota (Flanigan 1984, 1985), and
Cheyenne (Alford 1974) varieties of AIE.
3.2.5 Articles
In some varieties of AIE the articles a(n) and the are optional, as shown
in (30)(32). This feature may be a system of marking specific versus nonspecific or what is known to the speaker versus what is known to the


elizabeth l. coggshall

listener, as is found in creoles (John Singler p.c.), so an avenue of further

research is to determine if there is a pattern to article deletion or not.
(30) Nisqually English: Youre __ nice person. (Chessin and Aurbach
1982: 178)
(31) Navajo English: They found __ bone in __ dumpyard. (Cook and
Sharp 1966: 25)
(32) Brandywine English: And __ fellow looked around. (Gilbert 1986:
This feature is also documented in the Quinault (Rowicka 2005),
Mohave (Penfield Jasper 1980), Swinomish (Chessin and Aurbach 1982),
and Yakima (Chessin and Aurbach 1982) varieties of AIE.
3.2.6 Negation
A common feature of nonstandard varieties of English in general is the use
of multiple negative items in a single clause, what is referred to as negative
concord. Two examples of negative concord in AIE are shown in (33) and
(33) Eastern Cherokee English: My mother didnt make no pottery.
(Coggshall 2005)
(34) Laguna English: Then no police didnt catch us. (Stout 1979: 67)
Negative concord has also been found in the Mohave (Penfield Jasper
1980), San Juan (Wolfram et al. 1979), Lakota (Flanigan 1984, 1985), and
Isletan (Leap 1977b, 1974) varieties of AIE.
3.2.7 Prepositions
Prepositions are another source of difference between AIE and other varieties of English. Many varieties show nonstandard use of prepositions, as
in (35)(37).
(35) Tohono Oodham English: They were at fishing. (Bayles and Harris
1982: 17)
(36) Mohave English: He got fired of the church. (Penfield Jasper 1980:
(37) Cheyenne English: Lets ride on your car to Pizza Hut. (Alford 1974:

American Indian English


This kind of prepositional usage has also been documented in the Navajo
(Cook and Sharp 1966), Yakima (Chessin and Aurbach 1982), Brandywine
(Gilbert 1986), Tlingit-Haida (Chessin and Aurbach 1982), and Tsimshian
(Mulder 1982) varieties of AIE. Prepositions can also be optional in some
varieties of AIE, as shown in (38) and (39).
(38) Mohave English: He lives __ that second house. (Penfield Jasper
1980: 144)
(39) Lakota English: They live __ New York. (Flanigan 1984: 92)
Preposition deletion has also been documented in Nisqually (Chessin
and Aurbach 1982), and Brandywine (Gilbert 1986) English.
3.2.8 Word order
Word order in AIE has more options than many (though not all) nonIndian English varieties. Specifically, varieties of AIE use topicalization
and right-to-left syntactic constructions. Topicalization, the placement of
the focus or topic of a sentence at the beginning of the sentence, is more
common in AIE (as well as in other varieties of English, such as New York
City English (Feinstein 1980)), as demonstrated in (40) and (41).
(40) Apache English: That man, he went to town. (Liebe-Harkort 1983:
(41) Lakota English: All the neighbor boys, childrens, that he play with,
they all speak English. (Flanigan 1985: 227)
Topicalization has also been documented in the Mohave (Penfield Jasper
1980), Tohono Oodham (Bayles and Harris 1982), and Ute (Leap 1993)
varieties of AIE.
Right-to-left syntactic constructions, where the structure of the sentence
is left-branching rather than the usual right-branching of other varieties
of English, are less common in AIE but more marked than topicalization.
This construction can be seen in (42) and (43).
(42) Arapaho English: From the family is where we learn to be good.
(Leap 1993: 77)
(43) Mescalero Apache English: There are circle dance songs that we have.
(Leap 1993: 78)
This kind of sentence construction has also been documented in San
Juan, Lakota, Yavapai, and Tewa varieties of AIE (Leap 1993: 778).


elizabeth l. coggshall


The lexicon of AIE is understudied, with almost nothing written about it.
However, there are mentions of various sources of lexical items. Borrowings
from ancestral languages are obviously a major source for lexical innovation.
Kotzebue English has lexical items that come from the ancestral language,
loan translations, and the local dialect of English (Vandergriff 1982: 122).
Mulder (1982: 1067) details the kinds of words from Tsimshian that
are found in Tsimshian English; such borrowings are usually for terms
that English lacks. Innovative lexical items in Lumbee English have been
catalogued, including ellick for coffee, toten for a sign of impending death
or evil given by a spirit, juvember for slingshot, and brickhouse Indian for a
rich Lumbee (Brewer and Reising 1982, Wolfram et al. 2002: 63). Finally,
innit is a lexical item that spans many varieties of AIE. Innit, sometimes
spelled enit or ennit, is a tag question akin to yknow in other varieties of
English, but available for broader usage (Johansen 2007: 336). I have also
heard the variation is it spoken by Navajo youth to express surprise or
disbelief. Examples of enit can be found in the works of the novelist and
short story writer Sherman Alexie (Spokane/Coeur dAlene), as shown in
(44) Why dont you get in your BMW, thats what you drive, enit? (Alexie
2000: 50)
(45) Youre a fighter, enit?
I threw in the enit, a reservation colloquialism, because I wanted the
fighter to know that I had grown up on the rez, in the woods, with every
Indian in the world (Alexie 2012: 33, emphasis in original)
(46) So you must have eight or nine spirits going on inside you of you,
enit? (Alexie 2004: 183)
(47) Dont worry about the money, Thomas said. It dont make any
difference anyhow.
Probably not, enit? (Alexie 1994: 74)


Pragmatics in AIE has been the subject of intense study, especially by

those interested in improving American Indian childrens performance in
schools. This focus results from extensive intercultural miscommunication

American Indian English


between American Indians and non-Indians. Philips stresses this point,

stating Educators cannot assume that because Indian children . . . speak
English . . . that they have also assimilated all of the sociolinguistic rules
underlying interaction in classrooms and other non-Indian social situations
where English is spoken (1972: 392). Below are a few of the more salient
pragmatic features.
3.4.1 Silence
Compared to groups in contact with American Indians, speakers of AIE
are extremely quiet, to the point that may seem baffling or even rude.
Dumont, working in classrooms on the Cherokee and Sioux reservations,
described the mask of silence the students used while in the classroom
(1972: 346). Basso (1970) used the indigenous term for the extensive silence
his Western Apache informants used: To give up on words. He found
six particular instances where not talking was considered the correct thing
to do: when meeting strangers, when courting, when children came home
from boarding school, when getting cussed out, when being with someone in mourning, and being with someone undergoing a healing ceremony
(Basso 1970: 21724). He summarized this pattern thus: keeping silent in
Western Apache culture is associated with social situations in which participants perceive their relationships vis a vis one another to be ambiguous
and/or unpredictable (Basso 1970: 226).
3.4.2 Asking questions
Another point of differentiation between AIE and non-Indian varieties
is the avoidance of direct questions, which are considered by many AIE
speakers to be inappropriate, even rude (Leap 1993: 85). Apache English
speakers will answer, I dont know to any direct question, the inference
being that this is an inappropriate utterance and one should not have
asked, leading to understandable confusion across cultures (Liebe-Harkort
1983: 208). As a result of this constraint on directness in questions, requests
will often be framed in the form of a directive (Loan me . . . ) rather than a
question (Could you loan me . . . ?); in this speech community, a directive
is less rude than a direct question (Liebe-Harkort 1983: 208).
3.4.3 Humor
While not a locus for intercultural miscommunication, humor is a large
part of language use in American Indian communities. Basso, in his classic
study, Portraits of the Whiteman, shows humor in action in the Western
Apache community he studied (1979). However, that study was on Apache


elizabeth l. coggshall

speakers, and thus outside the purview of this chapter, but it does allow us
to see that humor pervades American Indian culture.
Vine Deloria, Jr. (Standing Rock Sioux), a prominent writer and activist,
has bemoaned the fact that this aspect of American Indians is not more
well known: It has always been a great disappointment to Indian people
that the humorous side of Indian life has not been mentioned by professed
experts on Indian Affairs (Deloria 1988: 146). In his book Custer Died for
Your Sins (its title itself being an example of this humor), he details many
examples of humor in AIE, usually satirical in nature, often lampooning
some of the worst things to ever happen to American Indians: Columbus,
Custer, the Bureau of Indian Affairs, missionaries, and white people in
general, as well as razzing members of other tribes (1988: 146167). In
Custer Died for Your Sins, Deloria gives a few examples of his favorite jokes,
in (48)(50).
(48) We also had a saying that in case of fire call the BIA and they would
handle it because they put a wet blanket on everything. (Deloria
1988: 1478)
(49) It is said that when Columbus landed, one Indian turned to the
another and said, Well, there goes the neighborhood. Another version has two Indians watching Columbus land and one saying to the
other, Maybe if we leave them alone they will go away. (Deloria
1988: 148)
(50) Custers Last Words occupy a revered place in Indian Humor. One
source states that as he was falling mortally wounded he cried, Take
no prisoners! Other versions, most of them off color, concentrate
on where those Indians are coming from. (Deloria 1988: 149)
Deloria hypothesizes that this emphasis on humor comes from the precontact method of social control where individuals who did not follow
cultural conventions were teased by other members of the tribe. This
kind of teasing was done in order to preserve the face of the individual:
since teasing is an indirect form of social control; such teasing was then
anticipated and making fun of oneself was an act of humility (Deloria 1988:


As long as American Indians exist as a separate social entity, AIE will

exist in some form. The reservation system in the US and the reserves in

American Indian English


Canada will also lead to the survival of AIE, since it is well attested that
isolation, either social or geographical, is a major factor in the creation
and maintenance of dialects (Labov and Harris 1986). And because it
will continue to exist, more research is needed on AIE. I suggest that
further work on AIE should take a particular path: new data on more
communities, especially in the growing urban communities, with a focus
beyond documentation to issues of identity work and changes in ethnicity.
Past research was mostly conducted in the 1970s, in the western part of the
United States, with an eye towards educational applications and questions
of genesis and documentation.
Because most of the research that has been presented in this study dates
back several decades, new research is needed to see how varieties of AIE have
changed over time, during decades of changing demographics and attitudes
towards Native Americans. Linguistic science has also progressed significantly since the 1970s, especially in field recording and acoustic phonetics.
Present-day work on AIE can use tools such as these to better understand
the workings of AIE, getting more precise data on the phonological features
and a larger corpus in which to look at morphosyntactic features.
Further, the majority of the work on AIE has focused on the western half
of the United States: e.g. Lakota, Navajo, Isletan, Mohave. This concentration is problematic because many Native American contact situations in
the eastern United States are significantly different from those in the west.
Eastern tribes were in contact with English speakers long before those in
the west. Further, eastern tribes were decimated early by disease and war;
they lack treaties signed between them and the United States government
and thus often cannot claim the special status of many western tribes;
they adopted English early on and lost their indigenous languages early,
too. These differences can perhaps lead to the reliance on and adoption of
cultural markers, such as linguistic features, to lay claim to an authentic
American Indian identity (Coggshall 2008). Further, little work has been
done on Canadian aboriginal peoples (Ball et al. 2006; Ball and Bernhardt
2008), who have yet a different history with regard to colonization and the
English language. Not only do these communities of AIE speakers have
different histories, the English speakers around them today can have an
effect on AIE. Language contact is a major force in language change and
new dialect formation, and studying contact situations that have been in
effect for various lengths of time enables a better understanding of the
linguistic and social facts of contact.
Another important area of concern is the large and growing population
of American Indians living off reservations and other rural enclaves, mainly


elizabeth l. coggshall

in urban centers throughout North America. In fact, in the US, more than
half of all people of indigenous descent live in cities. Urban Indian culture
differs greatly from that found on reservations and elsewhere (e.g. Lobo
and Peters 2001). So far, only one small study on urban Indians has been
published; Bartelt (1993) looked at a speech at a powwow in Los Angeles.
Perhaps the most important factor is that urban communities not only
include people from many tribes but also people of many other ethnicities.
This change in environment may lead to ethnic change. American Indian
ethnicity has undergone extensive reorganization since the 1970s (Nagel
1996), most strikingly in the advent of a pan-Indian identity for many
Native Americans, particularly in urban centers (Lopez and Espiritu 1990).
These changes in demographics and ethnicity suggest that there may
be repercussions in AIE. Recent studies (Coggshall 2008; Rowicka 2005)
suggest that the strict compartmentalization by tribe (as described above
in Section 3) may be deteriorating, and that certain features appear to be
spreading from one variety of AIE to another (see Leap 1982, 1993). This
change may be a result of panethnicity (Lopez and Espiritu 1990), where
the scope of identity expands to include a larger Indian identity on top
of a tribal-level identity. As the scope of identity changes, the language
used to express this identity may change as well, leading to a convergence
of different varieties of AIE. Leap (1993), on the other hand, has stated
categorically that there is no general AIE variety and that there will never
be one. Only further research can answer this question.

Alexie, Sherman. 1994. The Lone Ranger and Tonto Fist Fight in Heaven. New
York: Harper Perennial.
2000. The Toughest Indian in the World. New York: Grove Press.
2004. Ten Little Indians. New York: Grove Press.
2012. Blasphemy. New York: Grove Press.
Alford, D. 1974. The Cheyenne dialect of English and its educational implications.
Unpublished manuscript. Lame Deer, MT: Northern Cheyenne Title VII
Anderson, Bridget L. 1999. Source-language transfer and vowel accommodation in
the patterning of Cherokee English /ai/ and /oi/. American Speech 74: 33968.
Bakker, Peter and Robert Papen. 1997. Michif: a mixed language based on Cree and
French. In Sarah G. Thomason, ed., Contact Languages: A Wider Perspective.
Amsterdam and Philadelphia: John Benjamins, 295363.
Ball, Jessica and Barbara Bernhardt. 2008. First Nations English dialects in Canada:
implications for speech-language pathology. Clinical Linguistics & Phonetics
22(8): 57088.

American Indian English


Ball, Jessica, Barbara Bernhardt, and Jeff Deby. 2006. First Nations English
dialects. Unpublished monograph. University of Victoria and University
of British Columbia.
Bartelt, H. Guillermo. 1986. Language contact in Arizona: the case of Apachean
English. Anthropos 81: 46.
1993. Urban American Indian intertribal discourse. English World-Wide 14(1):
Basso, Keith. 1970. To give up on words: silence in Western Apache culture.
Southwest Journal of Anthropology 26(3): 21338.
1979. Portrait of the Whiteman. Cambridge University Press.
Bayles, Kathryn A. and Gail A. Harris. 1982. Evaluating speech-language skills in
Papago Indian children. Journal of American Indian Education 21(2): 1120.
Boas, Franz. 1911. Handbook of American Indian languages, vol. 1. Washington,
DC: Government Printing Office.
Brewer, Jeutonne P. and R. W. Reising. 1982. Tokens in the Pocosin: Lumbee
English in North Carolina. American Speech 57(2): 10820.
Calloway, Colin G. 2004. First Peoples: A Documentary Survey of American Indian
History. Boston: Bedford/St. Martins.
Chambers, J. K. 2004. Dynamic typology and vernacular universals. In Bernd
Kortman, ed., Dialectology Meets Typology: Dialect Grammar from a CrossLinguistic Perspective. Berlin: Walter de Gruyter, 12745.
Chessin, Linda and Emily Aurbach. 1982. Teaching composition to Northwest
American Indians. In H. Guillermo Bartelt, Susan Penfield Jasper, and Bates
Hoffer, eds., Essays in Native American English. San Antonio, TX: Trinity
University, 17386.
Coggshall, Elizabeth L. 2005. Cherokee English in the Appalachian context. Paper
presented at the Southeastern Conference on Linguistics 72, North Carolina
State University, April.
2006. Differential vowel accommodation among two Native American groups.
Unpublished Masters thesis, North Carolina State University.
2008. The prosodic rhythm of two varieties of native American English. Penn
Working Papers in Linguistics 14(2): 19.
Cook, Mary Jane and Margaret Amy Sharp. 1966. Problems of Navajo speakers in
learning English. Language Learning 16: 219.
Craig, Beth. 1991. American Indian English. English World-Wide 12(1): 2561.
Dannenberg, Clare. 1999. Grammatical and phonological manifestations of null
copula in a tri-ethnic contact situation. Journal of English Linguistics 27(4):
2002. Sociolinguistic Constructs of Ethnic Identity: The Syntactic Delineation of
an American Indian English. Raleigh, NC: Duke University Press for the
American Dialect Society.
Dannenberg, Clare and Walt Wolfram. 1998. Ethnic identity and grammatical
restructuring: be (s) in Lumbee English. American Speech 73(2): 13959.
Deloria, Vine. 1988. Custer Died for your Sins: An Indian Manifesto. Norman, OK:
University of Oklahoma Press.


elizabeth l. coggshall

Dickason, Olive Patricia and William Newbigging. 2006.A Concise History of

Canadas First Nations. New York: Oxford University Press.
Dillard, Joey Lee. 1972. Black English: Its History and Usage in the United States.
New York: Vintage Books.
Dozier, Edward. 1956. Two examples of linguistic acculturation: the Yaqui
of Sonora and Arizona and the Tewa of New Mexico. Language 32(1):
Drechsel, Emanuel J. 1983. Towards an ethnohistory of speaking: the case of
Mobilian Jargon, an American Indian pidgin of the Lower Mississippi Valley.
Ethnohistory 30(3): 16576.
Dumont, Robert V., Jr. 1972. Learning English and how to be silent: studies in
Sioux and Cherokee classrooms. In Courtney B. Cazden, Vera P. John, and
Dell Hymes, eds., Functions of Language in the Classroom. New York: Teachers
College Press, 34469.
Erdrich, Louise. 2012. The Round House. New York: Harper Collins.
Fear, Jacqueline. 1980. English versus the vernacular: the suppression of Indian
languages in reservation schools at the end of the nineteenth century. Revue
francaise detudes americaines 9: 1324.
Feinstein, Mark. 1980. Ethnicity and topicalization in New York City English.
International Journal of the Sociology of Language 26: 1524.
Flanigan, Beverly Olson. 1984. Bilingual education for Native Americans: the
argument from studies of variational English. In Jean Hanscome, R. Orem,
and P. B. Taylor, eds., On TESOL 83: The Question of Control: Selected
Papers from the Annual Convention of Teachers of English to Speakers of Other
Languages. Washington, DC: TESOL, 8193.
1985. American Indian English and error analysis: the case of Lakota English.
English World-Wide 6(2): 21736.
Gilbert, Glenn. 1986. The English of the Brandywine population: a triracial isolate in Southern Maryland. In Michael Montgomery and Guy Bailey, eds.,
Language Variety in the South: Perspectives in Black and White. Tuscaloosa,
AL: University of Alabama Press, 10210.
Goddard, Ives. 1997. Pidgin Delaware. In Sarah G. Thomason, ed., Contact Languages: A Wider Perspective. Amsterdam and Philadelphia: John Benjamins,
Harvey, Gina Cantoni. 1974. Dormitory English: implications for the classroom
teacher. In G. Bills, ed., Southwest Areal Linguistics. San Diego: Institute for
Cultural Pluralism, 28392.
Hazen, Kirk. 2002. Identity and language variation in a rural community. Language
78(2): 24057.
Hinton, Leanne. 1999. Trading tongues: loss of heritage languages in the United
States. English Today 15(4): 2130.
Hoffer, Bates L. 1982. The Alabama-Coushatta of East Texas: a case for multiplesource dialects. In H. Guillermo Bartelt, Susan Penfield Jasper, and Bates
Hoffer, eds., Essays in Native American English. San Antonio, TX: Trinity
University Press, 5566.

American Indian English


Humes, Karen R., Nicholas A. Jones, and Roberto R. Ramirez. 2011. Overview
of Race and Hispanic Origin: 2010. Washington, DC: United States Census
Johansen, Bruce. 2007. The Praeger Handbook on Contemporary Issues in Native
America: Linguistic, Ethnic, and Economic Revival, vol. 1. Westport, CT:
Greenwood Publishing Group.
Kwachka, Patricia. 1988. Oral and Written English of the Koyukon Athabaskan
Area: A Preliminary Analysis. APEL Research Report vol. 4. Nenana, AK:
Yukon-Koyukon School District.
Labov, William. 1969. Contraction, deletion, and inherent variability of the
English copula. Language 45(4): 71562.
Labov, William and Wendell Harris. 1986. De facto segregation of black and white
vernaculars. In David Sankoff, ed., Diversity and Diachrony. Amsterdam and
Philadelphia: John Benjamins, 124.
Leap, William L. 1974. On grammaticality in Native American English: the evidence from Isleta. International Journal of the Sociology of Language 2: 7990.
1977a. On consonant cluster simplification in Isletan English and elsewhere. In
William L. Leap, ed., Studies in Southwestern Indian English. San Antonio,
TX: Trinity University Press, 4553.
1977b. Two examples of Isletan English syntax. In William L. Leap, ed., Studies
in Southwestern Indian English. San Antonio, TX: Trinity University, 6578.
1977c. Uninflected be in Isletan English: a problem in accountability. In William
L. Leap, ed., Studies in Southwestern Indian English. San Antonio, TX: Trinity
University Press, 79100.
1977d. A note on subjectverb agreement in Isletan English. In William L.
Leap, ed., Studies in Southwestern Indian English. San Antonio, TX: Trinity
University Press, 12130.
1982. The study of American Indian English: status and direction of inquiry. In
H. Guillermo Bartelt, Susan Penfield Jasper, and Bates Hoffer, eds., Essays in
Native American English. San Antonio, TX: Trinity University Press, 122.
1993. American Indian English. Salt Lake City: University of Utah Press.
Leechman, Douglas and Robert A. Hall, Jr. 1955. American Indian Pidgin English:
attestations and grammatical peculiarities. American Speech 30(3): 16371.
Liebe-Harkort, Marie-Louise. 1983. A note on the English spoken by Apaches.
International Journal of American Linguistics 2: 2078.
Lobo, Susan and Kurt Peters. 2001. American Indians and the Urban Experience,
vol. 5. Lanham, MA: Altamira Press.
Lopez, David and Yen Espiritu. 1990. Panethnicity in the United States: a theoretical framework. Ethnic and Racial Studies 13(2): 198224.
Malancon, Richard and Mary Jo Malancon. 1977. Indian English at Haskell
Institute, 1915. In William L. Leap, ed., Studies in Southwestern Indian English.
San Antonio, TX: Trinity University Press, 14154.
Maynor Lowry, Malinda. 2010. Lumbee Indians in the Jim Crow South: Race,
Identity, and the Making of a Nation. Chapel Hill, NC: University of North
Carolina Press.


elizabeth l. coggshall

Meek, Barbara. 2006. And the Injun goes How!: representations of American Indian English in white public space. Language in Society 35(1): 93
Miller, Mary Rita. 1967. Attestations of American Indian Pidgin English in fiction
and nonfiction. American Speech 42(2): 1427.
1977. Children of the Salt River: First and Second Language Acquisition among
Pima Children. Bloomington: Indiana University Press.
Mithun, Marianne. 1999.The Languages of Native North America. Cambridge University Press.
Mulder, Jean. 1982. The Tsimshian English dialect: the result of language interference. In H. Guillermo Bartelt, Susan Penfield Jasper, and Bates Hoffer,
eds., Essays in Native American English. San Antonio, TX: Trinity University,
Nagel, Joane. 1996. American Indian Ethnic Renewal: Red Power and the Resurgence
of Identity and Culture. Oxford University Press.
Nagel, Joane and C. Matthew Snipp. 1993. Ethnic reorganization: American Indian
social, economic, political, and cultural strategies for survival. Ethnic and
Racial Studies 16(2): 20335.
Nelson-Barber, S. S. 1982. Phonologic variations of Pima English. In Robert
St. Clair and William Leap, eds., Language Renewal among American Indian
Tribes: Issues, Problems, and Prospects. Arlington, VA: National Clearing house
for Bilingual Education, 1223.
Penfield, Susan. 1977. Some examples of Southwestern Indian English compared.
In William L. Leap, ed., Studies in Southwestern Indian English. San Antonio,
TX: Trinity University, 2343.
1980. Selected grammatical characteristics of Mohave English. Unpublished
PhD dissertation, University of Arizona.
Philips, Susan U. 1972. Participant structures and communicative competence:
Warm Springs children in community and classroom. In Courtney B. Cazden, Vera P. John, and Dell Hymes, eds., Functions of Language in the Classroom. New York: Teachers College Press, 37094.
Rowicka, Grazyna J. 2005. American Indian English: the Quinault case. English
World-Wide 26(3): 30124.
Sapir, Edward. 1933. La realite psychologique des phonemes. Journal de Psychologie
Normale et Pathologique 30: 24765.
Schilling-Estes, Natalie. 2004. Constructing ethnicity in interaction. Journal of
Sociolinguistics 8(2): 16395.
Schreier, Daniel, Peter Trudgill, Edgar W. Schneider, and Jeffrey P. Williams,
eds. 2010. The Lesser-Known Varieties of English: An Introduction. Cambridge
University Press.
Shuy, Roger. 1964. The problem of American Indian English. In Roger W. Shuy,
ed., Social Dialects and Language Learning. Urbana, IL: National Council of
Teachers of English, 524.
Silverstein, Michael. 1972. Chinook jargon: language contact and the problem of
multi-level generative systems. Language 48(2): 378406.

American Indian English


Spicer, Edward Holland. 1967. Cycles of Conquest: The Impact of Spain, Mexico,
and the United States on Indians of the Southwest, 15331960. Tucson, AZ:
University of Arizona Press.
Stout, Steven. 1977. A comment on selection control in English expression at
Santa Ana. In William L. Leap, ed., Studies in Southwestern Indian English.
San Antonio, TX: Trinity University, 5563.
1979. Sociolinguistic aspects of English diversity among elementary-aged
students from Laguna Pueblo. Unpublished PhD dissertation, American
Torbert, Benjamin. 2001. Language history traced through consonant cluster
reduction: the case of Native American Lumbee English. American Speech
76(4): 36187.
Trujillo, Octaviana. 1997. A tribal approach to language and literacy development
in a trilingual setting. In J. Reyhner, ed., Teaching Indigenous Languages.
Flagstaff, AZ: Northern Arizona University, 1021.
Vandergriff, Jim. 1982. Kotzebue English: some notes on Inupiaq English. In
H. Guillermo Bartelt, Susan Penfield Jasper, and Bates Hoffer, eds., Essays in
Native American English. San Antonio, TX: Trinity University, 12156.
Whalen, Douglas H. and Gary Simons. 2012. Endangered language families.
Language 88(1): 15573.
Whorf, Benjamin Lee. 1941. The relation of habitual thought and behavior to
language. In Leslie Spier, A. Irving Hallowell, and Stanley S. Newman, eds.,
Language, Culture and Reality: Essays in Memory of Edward Sapir. Westport,
CT: Greenwood Press, 7593.
Wolfram, Walt. 1980. Dynamic dimensions of language influence: the case of
American Indian English. In Howard Giles, W. Peter Robinson, and Philip
M. Smith, eds., Language: Social Psychological Perspectives: Selected Papers from
the First International Conference on Social Psychology and Language. Oxford:
Pergamon Press, 37788.
1984. Unmarked tense in American Indian English. American Speech 59(1): 31
Wolfram, Walt, Donna Christian, Leap William L., and Lance Potter. 1979. Variability in the English of Two Indian Communities. Arlington, VA: Center for
Applied Linguistics.
Wolfram, Walt and Clare Dannenberg. 1999. Dialect identity in a tri-ethnic context: the case of Lumbee American Indian English. English World-Wide 20(2):
Wolfram, Walt, Clare Dannenberg, Stanley Knick, and Linda Oxendine. 2002.
Fine in the World: Lumbee Language in Time and Place. Raleigh, NC: North
Carolina State University.
Wolfram, Walt and Jason Sellers. 1999. Ethnolinguistic marking of past be in
Lumbee Vernacular English. Journal of English Linguistics 27(2): 94114.

c h a p ter 6

Bequia English
James A. Walker and Miriam Meyerhoff


Bequia is a small island located in St Vincent and the Grenadines in

the Eastern Caribbean whose sociolinguistic situation has been shaped by
two main geographic factors: its relative isolation and its small size (see
Map 6.1 ). The largest and northernmost of the Grenadines, Bequia lies
14 km (8 miles) south of St Vincent, separated by a deep channel. Until
the airport was built in 1992, the only way of reaching the island was by
ferry from St Vincent. Bequia is roughly hook shaped, approximately 11 km
(7 miles) from north to south and 8 km (5 miles) at its widest eastwest
point, and split between its leeward (west) and windward (east) coasts
by a low mountain ridge running roughly northsouth. Given the size
of the island and its relatively shallow timeline of settlement, the current
population of about 5,000 people features a surprising degree of diversity
in the varieties of English and English-based creole they speak.

Sociolinguistic history and current status of the variety

At the time of European arrival, Bequia was an uninhabited island, used

by the Caribs of St Vincent for fishing and hunting (Rochefort 1666). First
claimed by the French, under the Treaty of Paris (1763) and the Peace of
Versailles (1783), the island passed to the control of the British, who began
to develop large-scale sugar plantations and to import slave labour from
other Caribbean colonies and Africa. The collapse of the sugar industry
and the end of slavery in the early nineteenth century led to a shift to
cotton and arrowroot crops, which were ultimately unsuccessful too. At
the end of the nineteenth century, maritime work was more significant in

The French gave Bequia its current name, which is said to derive either from the Carib name becouya
island of clouds or from the French word bequille crutch (from the islands shape). The current
pronunciation [bkwe:] seems to be an English reading of the French spelling.


St Lucia



St Vincent

Industry Bay



Spring Bay
Port Elizabeth


Lower Bay



Friendship Bay


Paget Farm
Map 6.1 Island of Bequia

La Pompe


james a. walker and miriam meyerhoff

the economy and social life of the island (Adams 1996: 1089). Whaling
was established in 18756, and by 1890 there were six whaling stations on
Bequia. Although whaling is no longer the major industry, the International
Whaling Commission currently assigns Bequia a traditional peoples quota
of up to four whales a year. The arrival of electricity and telephones in 1969
70 and the building of an airport in 1992 reduced the islands isolation,
and its economy shifted to tourism and yachting.
The communities in Bequia where we have conducted research, while
geographically close to each other, are distinguished by different demographic histories. Although identification with a local village is an important part of social identity, different villages do not neatly correspond to
discrete ethnic or racial groups. The residents of Hamilton, a community situated at the northern end of Admiralty Bay, predominantly trace
their origins to the ex-slaves who worked the large plantation that once
dominated the area (Price 1988). Mount Pleasant, located in the hills east
of Port Elizabeth, was established by poor whites from Barbados who
were resettled throughout the Grenadines in 18589 (Price 1962; Sheppard
1977: 97). Although socially isolated in the past, Mount Pleasant now has
more contact with neighbouring communities. Southside refers to the contiguous settlements on the south shore (including Paget Farm, La Pompe
and Friendship) that developed out of fishing and whaling communities.
Because the Southside plantations were smaller than that of Hamilton, the
communities are believed to be more racially and ethnically mixed (Price

3 Features of the variety

Village is not only significant for social identity in Bequia but is also
cited when people discuss linguistic differences on the island. Each village
in Southside has a distinct character that sometimes emerges in speech:
anyone citing dialect differences on the island invariably mentions Paget
Farm, though Mount Pleasant is also mentioned. Hamilton speech is also
locally salient and appears to contain the most creole-like features, at least
among older speakers.
Local linguistic diversity is also reflected in a high degree of variation, not
only between villages but also even among and within individual speakers
from the same village (Meyerhoff and Walker 2007, 2012). This variation
consists in the presence or absence of specific features, the rates at which
features occur, and the conditioning of those features by elements of the
linguistic context. In the following sections, which are based on fieldwork

Bequia English


that we conducted in Bequia between 2003 and 2005 (see Meyerhoff and
Walker, 2012, 2013), we refer to features that occur generally across the
island, though we note the existence of variation and its distribution and
conditioning where relevant.


3.1.1 Short vowels

The front vowels kit and dress are generally short and lax ([] and []),
though they occasionally occur as tense vowels ([i] and [e]). The kit vowel
is sometimes lowered to something close to [] (e.g. miracle sounds like
[mrkl]). The trap vowel normally occurs as a low central [a], though
we have occasionally heard a more fronted []. This system is mirrored in
the back vowels to some extent. The foot and lot vowels are generally
short and lax ([] and []), though they are sometimes lowered to [] and
[a]/[], respectively. The strut vowel normally occurs as a central [], but
we have heard it lowered and backed to [] or [] (e.g. cup [kp]).
3.1.2 Long vowels
Long vowels tend to be monophthongal (i.e. without an off-glide), though
there appears to be variation between an off-glide and a full postvocalic
glide (e.g. [ej] rather than [ej ]) (Partridge 2009). The high vowels fleece
and goose tend to occur as [i:] and [u:], and the face and goat vowels
tend to occur as [e:] and [o:], though they occasionally have a schwa offglide (i.e. [e] and [o]). The bath vowel tends to occur as a long central
vowel [a:], though we have heard a more fronted []. The cloth, palm
and thought vowels tend to occur as a short central [a] (and they may be
distinguished from bath by length), but thought occasionally occurs as
a more back []. Partridges (2009) acoustic measurements show that [a]
and [] are realized very close together.
3.1.3 Diphthongs
The price and mouth diphthongs occur as [a] and [a], respectively,
though occasionally their onsets are raised to [] or backed and rounded to
[] (e.g. price [ps]). The choice diphthong usually occurs as [] but its
onset is often lowered, fronted and unrounded to [a] (e.g. choice [tas]).
3.1.4 Other vowel features
There is variability in the realization of vowels before /r/. The nurse vowel
tends to occur as a mid vowel, either central ([] or []) or backed [], with


james a. walker and miriam meyerhoff

lengthening of the vowel if /r/ is absent (e.g. nurse [n:s]). The near
vowel tends to occur as a long high [i:], though it is sometimes lowered
and merged with the mid square vowel [e:]. Some speakers seem to
distinguish here [h:] from hear [hi:], but there is no systematic distinction
between function words (like here) and content words (like hear). The
start vowel tends to occur as a central [a:] (though we have also heard a
more backed [:]) and the cure vowel tends to occur as [:]. We have not
noticed a difference in the realization of the north and force vowels,
although lord tends to occur with a more central variant [a:].
3.1.5 Prosody
The prosody of Bequia English is characteristically Caribbean, in that
there is a wide pitch range and unstressed vowels often receive their full
value. The latter pattern may be related to a rightward stress shift noted by
Wells (1982: 572) (emphatically or phrase-finally). There is a great deal of
variability in the realization of unstressed vowels: unstressed front vowels
tend to occur as [] or [i], while unstressed low vowels may occur as [a],
[] or []. Unstressed vowels with an underlying following /r/ may occur
with or without r-colouring.
3.1.6 Consonants
Bequia English has the full set of English stops: /p b t d k g/. Wordinitially, /t/ and /d/ have more dental articulations and /k/ and /g/ tend to
be palatalized (e.g. Coast Guard [kj o:s gj a:d]). Word-final /t/ is often fully
released and is sometimes deleted, especially in function words, such as
but [b] and about [ba]. Word-medially, /t/ generally occurs as an oral
stop (i.e. not flapped) or as a glottal stop, so that after may be pronounced
as [aft] or [af]. The fricatives are / f v s z ()/. Word-initial /h/
is variably present, normally omitted in function words such as here and
him. The interdental fricatives are most often realized as the alveolar/dental
stops [t] and [d] (e.g. think [tk], there [d]), though there appears to be
some variation according to style or topic (Ng 2009). The nasal consonants
are /m n/ and word-final //. After back vowels, word-final /n/ tends to be
velarized, so that Hamilton sounds like [hamlt]. The sonorant and glide
consonants are /w j l r/. The lateral /l/ is typically light rather than dark.
Generally, /r/ is realized as a retroflex [], though postvocalic rhoticity is
Consonant clusters involve a number of processes. There is widespread
stridentization of the first element in [str] clusters (so that industry is
pronounced [ndtri]), and palatalization in /tr/ clusters [t] (Partridge

Bequia English


2009). Final clusters may be devoiced (so that /dz/ in kids is realized as [ts])
and final stops in clusters are usually deleted, especially /t/ and /d/ (so respect
is realized as [rispk]). There is systematic pronunciation of ask as [ks],
and common metathesis of [sp] in some words (e.g. crisp pronounced as
[krps] and crispy as [krpsi]), though metathesis may be lexically restricted,
since mask undergoes final deletion [ma:s].


3.2.1 Plural formation

Plurality is indicated in several ways. Nouns may be marked with the
Standard English plural -s suffix and irregular forms (1) or with the regional
Caribbean strategy of postnominal (and) them (2). Unmarked (bare) nouns
also occur in contexts where the meaning is clearly plural (3).

Hear, the children these days, they live in a bed of roses. (Speaker 36)


a. When the stagaboys and them, which is the bugs them, coming
out to catch you, you want catch them quick. (Speaker 5)
b. Well, I hear old time people say turn you pocket an dem on the
wrong side. (Speaker 36)

(3) I still think the teacher used to try they best, eh? (Speaker 29)
3.2.2 Pronouns
Bequia English varies between the Standard English pronominal system,
distinguishing subjects (I, she, he, we, they) and objects (me, her, him,
us, them) with different forms for every pronoun except you, as well as
possessive pronouns (my, your, his, her, our, their), and a system in which
the same forms may be used for subjects, objects and possessives (4). With
reflexive pronouns speakers either combine the general pronoun with -self,
giving forms such as meself, weself, sheself, youself etc. (5), or use the Standard
English forms, myself, herself etc.
(4) a. He (done) ask me what me want, (but ?? I) tell him I want nothing,
(Speaker 6)
b. After me teach she, he come for L. G. [proper name]. (Speaker 6)
c. Not for we to live. (Speaker 306)
d. All of we descendants is from there. (Speaker 306)
e. Shoes for help they foot when the sun hot. (Speaker 12)
f. You put it over your shoulder and you get you corn and all thing
in it. (Speaker 26)


james a. walker and miriam meyerhoff

g. So who mother aint go have they young children and see um
something pick it up and go away with it? (Speaker 24)


a. Im a person who sit down by meself. (Speaker 6)

b. Everybody enjoy theyself. (Speaker 301)
c. When its youself. (Speaker 23)

There is also a gender-neutral third-singular pronoun e [i], which can be

used in all of the cases where Standard English uses it, as well as to refer to
human beings whose sex is known (6).
(6) a. E say it does vomit she. (Speaker 6)
She says it makes her vomit.
b. It na sound like e break. (fieldnotes)
It didnt sound like it [the glass] broke.
Pronouns may be emphasized using one (7), though we have only heard
this variant in first person singular.
(7) a. Q: So you alone live here?
A: Yeah. I mean [name] and dem doz come and go, but is me one
doz sleep and everything. (Speaker 10)
b. Is not me one, me and a cousin [go fishing together]. (Speaker 17)
c. Sometime me one me drop sleep. (Speaker 9)
Null subjects also occur in Bequia English, at higher rates than in Standard
English, but much less frequently than in typical null subject languages
such as Spanish.2 Subjects are most likely to be omitted in first person
singular or second person (8), though other topical subjects may also be
omitted (9).
(8) a. I said, you know, came from St Vincent. (Speaker 20)
I said, you know, [I] came from St Vincent.
b. Brad! Call you! Come here! (grandmother calling to grandchild
from a window, fieldnotes)
Brad! I am calling you! (from: me call you) Come here!
(9) Q: What would you say it [whale meat] taste like?
A: Well, taste like beef. (Speaker 20)

Based on a subsample of eighteen speakers from our corpus, balanced across Hamilton, Mount
Pleasant and Paget Farm, the rate of null subjects in affirmative declarative clauses is 5 per cent
[471/8820], while the rate in Standard English is estimated between 1 and 2 per cent in noncoordinated clauses (Meyerhoff 2000).


Bequia English

3.2.3 Tense and aspect

Tense and aspect are marked in Bequia English through an array of
morphosyntactic variants, though unmarked (bare) verbs are the most
common realization across all communities, for both present and past
reference (10).
(10) a. Even my boy child go out. (Speaker 14)
b. When we go round we play a ring song. (Speaker 1)
Present-tense verbs may be inflected with -s (11a) across all persons and
numbers, although very infrequently in all communities (Walker 2010).
The present progressive may also be used, occurring either with an overt
or null auxiliary (11bc). In addition, the verb may occur with will (12) or
doz (13) in contexts of habitual aspect, although both are very infrequent.

a. Of course, I pays my bonds. (Speaker 101)

b. We are not talking about the grave. (Speaker 23)
c. You going to dance, the same going on. (Speaker 1)

(12) These things will never show up on any test. (Speaker 27)
(13) But people doz hamper what your children they does.


In the past tense, verbs are variably marked with a number of pre- and
post-verbal morphemes. Weak verbs are generally unmarked for past tense
(14b), but inflection with the [t, d] suffix (14a) occurs at low rates across
all communities (though more frequently in Mount Pleasant). Strong and
semi-strong verbs are also variably marked for past tense, through stem
changes and/or final [t, d] (14c, 15). The irregular verbs be and have and
modal verbs are typically overtly marked for past tense (16).
(14) a. School stopped at age fifteen. (Speaker 313)
b. That was a night we always look forward to. (Speaker 20)
c. Yesterday I sit down here . . . (Speaker 13)

a. When I was sixteen, me brother send for me in Georgetown.

(Speaker 11)
b. In those days I built myself a boat, it was ten feet long. (Speaker 7)

(16) a. That was what she eating everyday, corn coocoo, and fish, bush
water tea. (Speaker 19)
b. But my we werent the fastest but we want to be the king.
(Speaker 13)
c. I had somebody so and I lose them. (Speaker 24)


james a. walker and miriam meyerhoff

d. Macintosh wouldnt need all that food for all of them. (Speaker 13)
e. I couldnt stay out late when I was a teenager. (Speaker 24)

Two creole-like preverbal markers are used in Bequia English: bin (17)
and done (18). When combined with stative verbs, bin seems to convey
simple past, as in (17a), but with dynamic verbs the interpretation may be
remote past (where bin alternates with did ) (17bc). Bin may also mark
irrealis mood (17d). Bin is largely restricted to Hamilton, and even there
occurs at a very low rate. The preverbal marker done indicates that an action
is completed and in some ways corresponds to the Standard English perfect
(which occurs very rarely in Bequia English) (18). We found occurrences
of done only in our Hamilton and Southside interviews, and at low rates.
(17) a.

Them not bin have no engine. (Speaker 1)

When I bin going to school . . . (Speaker 1)
The time I telling you, he bin there. (Speaker 6)
If rain na bin come, I bin going today by a lady name Miss F.
(Speaker 1)
If it hadnt rained, I would have gone today to visit a lady called
Miss F.

(18) a. I could deh here now; one minute you pass off, you done dead.
(Speaker 19)
I could be here now; the next minute you pass away and youve
b. They done call me already. (Speaker 314)
They had called me already.
c. I done accustom to home here.(Speaker 5)
Aspectual markers other than those we have already discussed may be
broadly divided into recurring events (habituals) and ongoing events (progressives). Habituals are marked with a broad range of pre- and postverbal
elements (19): deh, do(z), ah, V-ing, will/would and used to.
(19) a. All day you outside [when I was a kid], you deh running around.
(Speaker 102)
b. We doz close at ten. (fieldnotes)
c. It ah burn, did stop, but it used to burn me. (Speaker 17)
d. All them we na bin cussing. (Speaker 1)
e. We aint using no gun in the argument, you know, to settle it.
(Speaker 2)
f. Right now I dont think Ill walk alone in the dark. (Speaker 2)

Bequia English


g. Because if the fowls go and eat it, it would kill them. (Speaker 29)
h. We used to start Christmas as it were, say, from Nine Morning.
(Speaker 304)
Aspectual markers can combine with tense forms to yield more complex
distinctions: bin and V-ing together express past habitual (20), and did +
V may also express past habitual (21).
(20) When I bin going to school. (Speaker 1)
(21) Theres a lot of people did spoil [the] lobster industry. (Speaker 304).
3.2.4 Non-verbal predication
As in Standard English, predicates other than tensed verbs may take full or
contracted forms of be, but Bequia English also allows a zero variant, as do
other varieties of Caribbean English. In general, the linguistic contexts that
favour be-absence parallel those in other varieties of Caribbean English,
but we have found significant differences between different villages on
Bequia (Walker and Meyerhoff 2006).
Across the island, speakers rarely use be or any other copula before future
go(n)/gonna and going to or present participles (22)(23).
(22) a. Yeah, I think my boy gon done this year. (Speaker 5)
b. like if I- Is gon tell you something and like you know it already
(Speaker 101)
(23) a. They getting the clean clothes in a washing machine
(Speaker 102)
b. I say Im dancing the whole night (Speaker 12)
c. I deh looking, but I aint get through (Speaker 11)
Speakers from Hamilton and Paget Farm also usually omit be with a
following adjective (24), but speakers from Mount Pleasant are more likely
to retain be in this context.
(24) a. You think she old. (Speaker 101)
b. A little drink is good, keep you body good. (Speaker 1)
Before noun phrases (25), prepositional phrases (26) and locative adverbs
(27), speakers tend to retain be, though there is variation across communities and even among different speakers within each community.
(25) a. There are the poor folks who cant afford all of that so (Speaker 102)
b. Lo Bay a nice place. (Speaker 12)


james a. walker and miriam meyerhoff

(26) a. Yeah, we in the other one. (Speaker 102)

b. This pan deh in a mess. (fieldnotes)
(27) a. But they na there now. (Speaker 1)
b. She doesnt have her memory but shes there. (Speaker 102)
Invariant be also occurs (28), usually in hypothetical or counterfactual
(28) a. Me think you be [Nancys] son. (Speaker 1)
b. And then they be a burden to you. (Speaker 301)
Aside from be, the variant deh may occur if the predicate is locative (whether
an adverb or a prepositional phrase), as well as with present participles (29).
(29) a. I deh by the market working. (Speaker 6)
b. Else otherwise she deh there in America too. (Speaker 5)
c. And I deh eighty one now. (Speaker 6)
3.2.5 Existentials
There are three main types of existential construction in Bequia: the
dummy subject there followed by some form of the verb be, as in Standard
English (be-existentials); or the dummy subject it, either with some form
of have (have-existentials) (30ac), or with got (30d) (got-existentials).
(30) a.

It have some person like that. (Speaker 4)

It have a lime tree here . . . (Speaker 6)
It had a next one where you hold it (Speaker 2)
I know it got some doz call [it] Jack-o-lantern. (Speaker 12)

The overall frequency of each existential construction differs according

to village, with be-existentials preferred in Southside villages and haveexistentials preferred elsewhere. Hamilton has the highest occurrence of
got-existentials. Within each village, there is also some variation according
to the individual speaker.
Regardless of whether the postverbal subject NP is singular or plural,
the preferred form of be-existentials is singular agreement (there is/s/was)
and the preferred form of have-existentials is plural agreement (it have),
though again there is variation according to village and speaker.

Bequia English


3.2.6 Questions
In forming questions, speakers of Bequia English alternate between
English-like forms (31) and forms with neither do-support nor subject
auxiliary inversion (32).

a. What you think about him? (Speaker 20)

b. What changes you have seen in our island? (Speaker 306)

(32) What your parents really wanted you to do? Do you know?
(Speaker 306)
3.2.7 Negation
Negation in Bequia English takes several forms, the most frequent being
not/-nt, which is also found in Standard English, and aint (which may be
pronounced as [], [n], [nt]), found in many nonstandard varieties of
English and in English-based creoles. Other options are na and don.
(33) a. They en really have the oldest in our day [at school] (Speaker 306)
b. Oh God look he dead reach and nobody na know. (Speaker 6)
Oh God, look, hes fallen dead and nobody knew.
c. And when he is old, he is not depart. (Speaker 2)
d. Not even the moon. (Speaker 303)
e. If it hasnt have eggs, all well. (Speaker 304)
In addition, never and na can be used to express at no time; never and
not a as shown in the following examples.
(34) a. Our parents never let us go! (Speaker 306)
b. They used to sing, This vessel na builder . . . (Speaker 23)
In general, Bequia English observes negative concord, which is triggered
by the presence of negative elements elsewhere in the sentence as well as
verbal negation (33).
(35) a. Nobody couldnt beat me. (Speaker 5)
b. I aint into no jazz. (Speaker 2)
Some villages seem to favour one variant more than other villages. For
instance, na (33b) seems to be most common among Hamilton speakers
and less common among Southside and Mt Pleasant speakers (Walker and
Sidnell 2011).


james a. walker and miriam meyerhoff

3.2.8 Complementation
Bequia English uses a variety of strategies to introduce clausal complements, depending on whether the complement is subordinate (i.e. tensed),
infinitival or relative. As in other varieties of English, subordinate clauses
may be introduced with or without a complementizer (36). Verbs of saying
may also take an additional verb say to introduce indirect speech (37).
(36) a. You- you see that they running the sea-lion and them upon the
shore and eating them. (Speaker 25)
b. I aint think he dead yet. (Speaker 16)
(37) And when the girl come and tell me say me pass to go in Standard
Two, I tell them for tell these teacher them to write them down and
send them give me. (Speaker 1)
Subordinate infinitival clauses may be introduced with to, as in Standard
English, or a Caribbean variant f (from English for) (38), as well as
omitting to and f altogether (39).
(38) a. You want f know that, how that? (Speaker 11)
b. Put them f air out until they ready f call you f come back out
again. (Speaker 11)
c. You got know how to deal with the both types. (Speaker 2)
d. But you have f keep chuking it because it going stick.
(Speaker 17)
(39) You want turn off this a minute? (Speaker 2)
Relative clauses are formed with Standard English relative pronouns (that,
who, which) (40ab), or with nonstandard uses of which and that (40c),
as well as the zero relative pronoun (40d). There is also a general relative
pronoun weh [we] (40e) that seems to derive from Standard English where
or what.
(40) a. And the first man that carry me to sea, but he died. (Speaker 11)
b. But he was the one who never kept it up. (Speaker 7)
c. Until at last my- my aunt, which was living in Petit Martinique,
Aunt [name]? She came up. (Speaker 24)
d. E have some [people in Mt Pleasant] speak broken language
cf. There are some who speak . . . (Speaker 24)
e. Fion! Two roti weh you order. (fieldnotes)
Fiona! The two roti that you ordered [are done].

Bequia English



In addition to the common vocabulary inherited from English, Bequia

English contains many words that do not occur widely in the Englishspeaking world, although much of this vocabulary is found in other
Caribbean English varieties (or at least, in the eastern Caribbean). For example, boilin is a local word for a soup made from fish heads and dumplings,
which is used elsewhere in the Caribbean to refer to fish-head soup (without dumplings). Coastal trees in Bequia (and many other islands of the
Caribbean) are called sea grapes (coccoloba uvifera), because their fruit resembles bunches of grapes. There are a number of words of likely African origin,
such as dukuna corn pudding, which may be derived from Ga-Adangme
doko na sweeten (vb) mouth (Allsopp 2003: xxxiv, 207). In Bequia English,
when people are referring to the bad luck brought on by attention from
others, they talk of maljo(u), a term that occurs in varying forms throughout the Caribbean and perhaps derives from French mal dyeux bad eyes
(or Spanish mal de ojo or a Portuguese equivalent) (Allsopp 2003: 364).
Some words and phrases (like comess gossip and jumbie ghost) are
also found in other Caribbean English varieties, though some are more
preferred in Bequia or Bequians use different words to refer to the same
thing. For example, what is known as suck-teeth throughout the Caribbean
(the kissing noise made by intaking breath over closed teeth, a sign of
irritation or disapproval) is also called chups in Bequia English. Similarly,
we have recorded pickney children, which seems to be a more recent
adoption from elsewhere in the Caribbean. Finally, some of the words and
phrases that people use in Bequia English may be used in ways similar to
lesser-known dialects of English, such as to vex to annoy, bother and to
teef to steal (< thief ).

4 Conclusion
Although Bequia is a lesser-known island of the eastern Caribbean, it represents an interesting situation of dialect and language contact contained
within a relatively small geographic space. The diverse linguistic varieties
spoken in Bequia show contributions in phonology and morphosyntax
from nonstandard varieties of English (including British and Irish English,
as well as white Caribbean English) and from English-based creoles,
with additional lexical contributions from other languages such as French
and Portuguese. Yet, despite over 150 years of intensive contact, the
geographically proximate communities of Bequia have managed to retain


james a. walker and miriam meyerhoff

unique ways of speaking, in the presence or absence of some features, and

in the distribution and conditioning of others.
The speakers we sampled in our research represent the older generation in Bequia, who acquired their way of speaking before the island was
opened up to outside influence and greater mobility. The tourism industry
in Bequia continues to expand, with more and more outsiders from North
America and Europe building homes and resorts on the island. As a result
of increasing access to higher education and exposure to other, more mainstream linguistic models, there is some evidence that younger people are
losing some of the more distinctive features or have begun to use them in
ways that differ from that of their parents and grandparents (Daleszynska

Adams, Edgar. 1996. Linking the Golden Anchor with the Silver Chain: A Historical and Socio-economic Perspective on Shipping in St Vincent and the
Grenadines. Kingstown, St Vincent and the Grenadines: R&M Adams Book
Allsopp, Richard. 2003. Dictionary of Caribbean English Usage. Kingston, Jamaica:
University of the West Indies.
Daleszynska, Agata. 2012. Variation in past tense marking in Bequia creole:
apparent time change and dialect leveling. PhD dissertation, University of
Meyerhoff, Miriam. 2000. Constraints on Null Subjects in Bislama (Vanuatu):
Social and Linguistic Factors. Canberra: Pacific Linguistics Publications.
Meyerhoff, Miriam and James A. Walker. 2007. The persistence of variation
in individual grammars: copula absence in urban sojourners and their
stay-at-home peers, Bequia (St Vincent and the Grenadines). Journal of
Sociolinguistics 11: 34666.
2012. Grammatical variation in Bequia (St Vincent and the Grenadines). Journal
of Pidgin and Creole Languages.
2013. Bequia English. Westminster: Battlebridge Publications.
Ng, Zoe. 2009. A social dialect study of (th) stopping in Bequia English. MA
thesis, University of Edinburgh.
Partridge, Andrew. 2009. Mapping the vowel space in Bequia creole. MSc dissertation, University of Edinburgh.
Price, Edward. 1962. The Redlegs of Barbados. Journal of the Barbados Museum
and Historical Society 29: 4752.
Price, Neil. 1988. Behind the Planters Back: Lower-Class Responses to Marginality in
Bequia island, St Vincent. London: Macmillan.
Rochefort, Cesar de. 1666. A History of the Caribby Islands. Rendered into the
English by John Davy.

Bequia English


Sheppard, Jill. 1977. The Redlegs of Barbados: Their Origins and History. Millwood,
NY: KTO Press.
Van Herk, Gerard. 2000. The question question: auxiliary inversion in early
African American English. In Shana Poplack, ed., The English History of
African American English. Oxford and Malden, MA: Blackwell, 17597.
Walker, James A. 2010. Looking for agreement in the Eastern Caribbean: evidence
from Bequia. Paper presented at the annual meeting of the Society for Pidgin
and Creole Linguistics, Baltimore, MD.
Walker, James A. and Miriam Meyerhoff. 2006. Zero copula in the eastern
Caribbean: evidence from Bequia. American Speech 91: 14663.
Walker, James A. and Jack Sidnell. 2011. Inherent variability and coexistent systems:
negation on Bequia. In Lars Hinrichs and Joseph Farquharson, eds., Variation
in the Caribbean: From Creole Continua to Individual Agency. Amsterdam and
Philadelphia: John Benjamins, 3955.
Wells, J. C. 1982. Accents of English, vol. 3: Beyond the British Isles. Cambridge
University Press.

c h a p ter 7

Saban English
Jeffrey P. Williams and Caroline Myrick


Saba, like Bequia described by Meyerhoff and Walker in Chapter 6, is a

very small island in the Caribbean Sea whose sociolinguistic landscape has
been shaped by its size and isolation. Located in the Leeward chain of the
Lesser Antilles, Saba is approximately 45 kilometers south of St. Martin.
Formerly, Saba had been part of the Netherlands Antilles, a political unit
that was dissolved as an autonomous country within the Kingdom of the
Netherlands in 2010. Saba, along with Sint Eustatius and Bonaire, were
united as special municipalities of the kingdom.
With a total land area of 13 square kilometers, Saba is the smallest
inhabited island in the archipelago of the Leeward Antilles and one of the
smallest inhabited islands in the West Indies (see Map 7.1 ).2 The island
consists of the remnant of a volcanic cone and reaches a maximum elevation
of 887 meters at Mt. Scenery. Sabas physiography is one of steep peaks
and facing cliffs dissected by guts (ghat or ghaut from Hindi) that lead
the extensive rainwater to the sea. The rocky, almost entirely impenetrable
coastline and the lack of any natural harbor also discouraged ships from
setting anchor at the island.3 The completion of the Leo A. Chance Pier
in 1972 enabled larger vessels, as well as smaller tourist vessels, to visit the
island. Prior to that, ships had to anchor offshore and passengers and crew
would take a dinghy to shore. Sabas interaction with the outside world was
greatly enhanced by the construction of the Juancho E. Yrausquin airstrip


Fieldwork from which this chapter derives was conducted on several occasions over a period of just
over thirty years. Williams spent a month in the village of Windwardside in 1982 and revisited the
island in 2002 and 2004. Myrick spent a month doing fieldwork in Hells Gate, Windwardside, and
The Bottom in 2012 and 2014.
Petit Martinique a dependency of Grenada is the smallest inhabited island in the region, with
only 586 acres.
In a recent archaeological study, Espersen (2009) has shown that the now abandoned village of
Marys Point (also known as Palmetto Point) had a tide-dependent beach landing that facilitated
boat contact between the village and what were the Danish Antilles (US Virgin Islands).






Flat Point

Marys Point
Mtn 585 m
Sandy Cruz


Cove Bay
Spring Bay

Middle Island
Mt Scenery
(887 m)



Core Gut Bay


The Gap

The Level

The Bottom

Booby Hill

St Johns
Giles Quarter
Fort Bay
Great Level Bay
Map 7.1 Saba


800 m
0.5 mile


jeffrey p. williams and caroline myrick

in 1963 (Hartog 1975). Although the airstrip now serves as a commercial

airport, it is permanently closed to aircraft that are not pre-approved.4 At
the present, the Juancho E. Yrausquin Airport is the shortest commercial
runway in the world.
The islands rugged physiography has also hindered internal communication throughout the history of European occupation of the island. Prior
to the introduction of motor vehicles, all intra-island communication was
accomplished either on foot or by donkey. Stairs were hewn into the steep
cliffs to serve as trails between the isolated villages and the majority of
people, women in particular, did not venture out of their natal villages.
However, with the introduction of the first jeep in 1947 and the gradual
completion of the only paved vehicular road that connects all of the villages
on the island in 1963, intra-island communication became less arduous and
more common. However, as Hartog (1975: 22) points out, the inhabitants
of the four principal villages The Bottom, Windwardside, St. Johns, and
Hells Gate had very little contact with each other. He goes on to say that
this pattern of extremely limited interaction gave rise to distinctive styles
of English spoken in each village. Will Johnson, a Saban politician and
historian, provides the following case history.
Contact between villages was infrequent. One of my grandmothers, who
was born and raised in the village of Hells Gate, never visited The Bottom
until she was fifty years old. Incidentally, her husbands parents came from
that village. Although this could have been a case of not liking her in-laws,
as a boy I often heard old people saying that they had never had the time,
inclination, or desire to visit any other village. There are still a number of
people both young and old, even with all means of transportation today,
who have never been off Saba. (Johnson 1979: 50)5

In this chapter we provide a preliminary descriptive account of the variety of

English spoken on the island of Saba, which is located in the Leeward chain
of the Lesser Antilles. Saba, due to its settlement history and patterns of
social interaction between European- and African-descended populations
on the island during the colonial period, evidences little difference in
the grammars of whites and blacks.6 Our data are drawn primarily from

The pre-approved aircraft are limited to STOL (short take-off and landing) aircraft, which seat a
maximum of fifteen people.
On Williams first visit to the island in 1982, he spoke with older Saban women who had never left
their villages of birth.
Euro-Sabans are a relatively homogenous social group in opposition to Euro-descendants on other
anglophone islands where we find strong local differences between clear-skinned, red, and other
social classificatory groupings.

Saban English


white speakers in the historically white villages of Windwardside and Hells

Gate. Data from the village of The Bottom are from both black and white

Sociolinguistic history and current status of the varieties

Next to nothing has appeared in the literature on English dialectology,

sociolinguistics, or lesser-known studies that is focused on the varieties
of English spoken on Saba.7 However, Saba was the focus of research by
the late anthropologist Julia Crane, who published several works on the
ethnography of the island (Crane 1971, 1987). One of these works, Saba
Silhouettes, is a source of sociolinguistic information since it is an extensive
compendium of narratives by Sabans, written in a form that preserves the
grammatical features of the varieties.

Early European settlement and anglophonization

Saba was first settled by Europeans in the middle of the seventeenth century. While the island shows material evidence of pre-contact indigenous
occupation, the early accounts make no mention of an indigenous population present at the time Europeans commenced occupation and settlement.
The first permanent European settlements on Saba were established by Zeelanders who had come to the island from the nearby Dutch possession of
St. Eustatius in 1640 (Hartog 1975: 18). Passing reference to English settlers
are in evidence prior to the Zeelanders occupation in 1640, although, the
first official documentation of English-speaking settlers does not appear
until 1659 (Hartog 1975: 21). The 1659 document, a request for an Englishspeaking clergyman, gives the Saban population as being composed of 57
Dutch and 54 English, Irish, and Scots combined (Johnson 1979: 8). There
is no reference made to the means by which they arrived on the island,
or their specific provenance, but it appears that these anglophone settlers
arrived on the island without the direct knowledge or permission of the
Dutch colonial authorities on St. Eustatius.
Ethnohistorical evidence concerning the growth of the English-speaking
population on Saba and the dispersal of the indentured Europeans that
were emigrating from Barbados, St. Kitts, and Antigua during this time

Williams wrote a Masters thesis on white Saban English in 1984 and has provided information on
the Windwardside dialect embedded in various publications on the Euro-Caribbean varieties.


jeffrey p. williams and caroline myrick

strongly suggests that the anglophone population that arrived on Saba in

the time prior to 1659 had been transported to other islands in the West
Indies as indentured servants. The harsh conditions of indenturage, the
increasing dependence on African slave labor, as well as land shortages
forced many servants to seek refuge on the small, isolated islands of the
Caribbean Basin. In short, the English-speaking settlers had come to Saba
from various parts of the British Isles via St. Kitts, Barbados, and Antigua
during the exodus of indentured servants in the middle of the seventeenth
century. They had come to Saba to avoid the hardships and possible fatal
outcomes of indenturage. Due to its isolation and lack of colonial infrastructure, Saba was an ideal location for relocation; one where there was
little chance that the escapees would be caught and subject to harsh punishment and re-indenturage. The social circumstances that these escaped
servants found themselves in was identical to those that escaped slaves also
found themselves in during the height of forced labor transportation in the
Americas. These escaped servants, not unlike their African counterparts,
formed discrete, isolated communities and developed distinctive forms of
When the English took possession of the island in 1665, Henry Morgan
removed the seventy Dutch-speaking colonists to Sint Maarten.9 In the
archival accounts, a population record from 1699 indicates that almost all
of the population was from the British Isles (Johnson 1979: 15). While we
have not seen the original report, we can venture to speculate that the
account might not be completely accurate since many of the previously
Dutch surnames were anglicized early in the settlement history of the
In spite of several dispossessions by England and periods of English
rule, Saba has remained part of the Hollandish sociopolitical sphere since
the mid seventeenth century. Saba finally became a permanent Dutch
possession in 1816. As the linguistic evidence shows, Dutch rule has not
translated into Dutch linguistic or sociolinguistic influence on Saba. As part
of the 2010 Acts that changed the political status of Bonaire, St. Eustatius,
and Saba (Wet op de Openbare Lichamen Bonaire, Sint Eustatius en Saba),
the Kingdom of the Netherlands made the use of English by residents of
Saba and St. Eustatius for official business and governmental transactions
formally recognized.

For further discussion that draws parallels between escaped African slaves and escaped European
servants, see Beckles (1986).
Additionally, 102 African-descended slaves were transported to Jamaica at the same time.

Saban English


Een ieder kan de Nederlandse taal gebruiken in het verkeer met de in artikel
4b, eerste lid, bedoelde organen en personen.
2. Een ieder kan:
a. het Papiaments gebruiken in het verkeer met de organen van het openbaar
lichaam Bonaire;
b. het Engels gebruiken in het verkeer met de organen van het openbaar lichaam
Sint Eustatius of Saba.
[Anyone can use the Dutch language in communications with the bodies
referred to in Article 4b, first paragraph, and people.
2. Anyone:
a. can use Papiamentu with official public offices and persons representing
those offices in Bonaire;
b. can use English with official public offices and persons representing those
offices in Sint Eustatius and Saba.]


Sociolinguistic history of Saban varieties of English

English has dominated the sociolinguistic landscape of Saba throughout

its colonial and neo-colonial periods in spite of its political ties to the
Netherlands. A rugged terrain that was unsuited to the development of
sprawling, large-scale plantations as were developed in many parts of the
circum-Caribbean, contributed significantly to the islands isolated sociolinguistic character. Physical isolation, both internally and externally, was
one of several catalysts to social isolation, resulting in the islands diversity
of village-level dialects within a historically local population of around
1000 individuals.10
Saba has had five historically attested villages. Only four of these continue to the present day.11 While Saban identity has become a more highly
motivating "act of identity" for individuals in a broader regional and international context of social interactions, village identity was a key motivation for the genesis and consolidation of focused local accents and
dialects. Village identities on Saba as in other parts of the West Indies

The opening of a medical school in the early 1990s has changed the sociolinguistic profile of the
island and increased the resident population by approximately 30 percent.
The village of Palmetto (or Marys) Point was relocated between 1920 and 1934. Oral history provides
one set of explanations for the relocation while archaeological evidence provides another. Espersen
(2009) provides a detailed historical archaeology for the site, with a sophisticated nuancing of
oral histories and comparative information regarding other isolated, enclave communities in the
colonial Americas.


jeffrey p. williams and caroline myrick

were seemingly based on both ethnicity and color within a colonial social
Like I said, in each village tis different. Also, less than 1000 people . . . St.
Johns, those people they talkin more like Irishmen. In the Bottom, well
they have different. They have-the old white ones, those who are English,
they spoke London English, or tried to. [Windwardside, W//ma/1982]13

This local exegesis on ethnodialectology remains unchanged as the following transcription of a recording over thirty years later shows.
Well . . . The Bottom I mean talks a little different to the people over
here . . . Yeah. The St. Johns people, they talk a little different to The Bottom people too. Yeah, you can listen to you can hear it, I mean. Well the
Hells Gate and Windwardside, you know, its the same thing, you know?
Difference. [Hells Gate, W//ea/2012]

The varieties spoken in the historically Euro-descended villages of Windwardside, Hells Gate, St. Johns, and Marys Point belong to the EuroCaribbean Anglophone Linguistic Area (ECALA) as defined in Williams
(2012). The hallmark of these varieties is their koineized nature, exhibiting a structured integration and complex variation of forms drawn from
regional source input dialects. Euro-Saban varieties are no different in these
regards. Color has been a key social marker in Saban identity and it has
governed settlement and marriage as it has throughout the West Indies
throughout the colonial and neo-colonial periods.
It is evident that in the founder period of settlement of Saba, a great deal
of dialect and language contact took place. Plantations never developed on
Saba and the social economy remained at the founder stage of Mufwenes
development scheme. Minimally, the monolithic varieties of Dutch and
English were in contact as well as an unknown number of African languages

In his study of sojourners in the West Indies and the northeast coastal United States, Karras (1992)
discusses the pattern of ethnic segregation that was prevalent during the late colonial period.
Each recorded example provides the following information on the speaker: [village, ethnicity/sex/age
group/year recording was made]. The following abbreviations and conventions are used:

adolescent, approximately between the years of 14 and 20

child, approximately between the years of 0 and 9
elder, approximately beyond 66 years of age
middle age, approximately between 36 and 50 years of age
older age, approximately between 51 and 65 years of age
preadolescent, approximately between 10 and 13 years of age
white speaker; identifies as an individual of European descent
young adult, approximately between 21 and 35 years of age
male speaker
female speaker

Saban English


that were brought to the island with the slave population that came during
the latter part of the seventeenth century. The Dutch were not a social
factor for long in the sociolinguistic history of the island. However, the
English-speakers increased in number and importance.
The outcome of the dialect contact that occurred involved such factors
as ethnicity, kinship, locality, and demographic proportions. Each village,
being isolated developed its own sense of identity: a composite drawn from
those individuals who had come together to make up that community.
Euro-Sabans have continued to see themselves as descended from settlers from the British Isles, and not from the Dutch as the Netherlands
administration has attempted to portray them as.14 In point of fact, when
Saba became a permanent possession of the Netherlands in 1816, there was
formal stipulation that all official documents be translated into English
since there were no speakers of Dutch on the island (Hartog 1975: 22). This
act of identity has had linguistic consequences for the Euro-Sabans.
Although all Sabans are taught Dutch in school from the first grade on,
they do not use it among themselves in ordinary conversation because it is
associated with an identity that is distinctly non-Saban. The use of Dutch
on the island is restricted primarily to the classroom, to interaction with
native speakers of Dutch, and to some governmental and official events.
Since Dutch was the official language of the Netherlands Antilles until
the addition of Papiamentu and English in 1984, up until that time birth
records, passports, and the only available newspaper were in that language.
Even though Sabans have been very competent speakers of Dutch and
make use of opportunities to use the language in their interactions with
Netherlanders, the language itself would not have survived for the time that
it has were it not for the political system that mandated primary instruction
in Dutch.
With the development of a Saban national identity within the larger
framework of West Indian identity has come the genesis of a Saban
linguistic identity as well.15 This has brought about the development of
another focused variety of Saban English that can be identified, particularly by non-Saban West Indians as a discrete variety. This development has
been accompanied by Sabans increased interaction with other Caribbean

See Johnson 1979 for a full discussion of this point as well as recent discussions in The Saba Islander
a blog/newspaper published by Mr. Will Johnson of Saba.
Euro-Saban varieties of English might have influenced Bermudian English. After the outbreak of
the Boer War in South Africa, the English established prison camps on the island of Bermuda.
In 1902, over one hundred Sabans were employed at the camps. Many of those chose to remain
on Bermuda, and in the 1970s there was a population of nearly 200 Sabans residing on Bermuda
(Hartog 1975: 68).


jeffrey p. williams and caroline myrick

nationals, the recent opportunities for social and geographical mobility, and
the constantly increasing number of tourists who visit the island annually.
Sociolinguistic focusing that had taken place at the level of regional, i.e.
village, accents and dialects is in the process of becoming diffuse through
the process of the interaction of those factors. Many Sabans that we have
spoken with over the years have told us how the distinctive dialects were
being lost by the younger Sabans due to their extended periods of residence
off the island. In some regards, the dense local networks have been in a state
of decay, giving way to the creation of new, less dense networks that span
the geographical and social diversity of the island, and may even connect
individuals to another island such as Curacao, Aruba, or St. Eustatius, or
even another nation outside the Caribbean such as Canada or the United

3 Features of Euro-Saban English

The desctiption we provide here is broadly based on data collected across
village varieties, ethnic varieties, and expressing longitudinal variation.
These distinctions are rapidly disappearing as the island continues to

Segmental phonology16

Short vowels
Saban English shows lowering of the kit vowel ([] and []), with bit
being realized as low as [bt]. This pattern of lowering is identical to what
is evidenced in Bequia English (see Meyerhoff and Walker, this volume,
Chapter 6).
Saban English shows lowering of the dress vowel, with bet being realized
as low as [draes], or somewhere in between [] and [ae].
The trap/bath set is lowered and backed, occupying the lowest vowel
space in Saban English. In the speech of the older residents of Hells Gate,

The phonological description is based on actual acoustic analysis using PRAAT.

Saban English


the trap/bath vowel is merged with the lot vowel, which is fronted.17
Pass is realized as [ps].
Saban English demonstrates lowering of the foot vowel. In fact, the foot
vowel is realized lower than the boat vowel. There is evidence of additional
centralization of the foot vowel in the Hells Gate variety, causing most
lexical items with the foot vowel to be realized with the strut vowel.18
We find examples such as took being realized as [tk]. In the village of
Windwardside, the strut vowel is lowered to bought space, with tough
realized as [tf].
Acoustic analysis of Saban English shows absence of a central vowel. It is
noteworthy that Saban English speakers do tend to centralize unstressed
syllables as schwa, supporting Williams (2012: 147) position that "[n]onweakening of vowels in unstressed syllables is not a feature of the EuroCaribbean Anglophone Linguistic Area."
For those speakers who do not show a lot thought merger, the lot
vowel is realized as // For many speakers, this vowel shows raising and
backing, and is close to merged with the thought vowel (realized as
//). Saban Englishes differ from the General American lot thought
mergers by approximation, in which the thought vowel tends to lower
towards lot.

The face vowel is monophthongal; in older Hells Gate speakers, pre-nasal
fleece is realized as the pre-nasal face vowel, so that mean and main are
The near and square vowels are close/merged, so that fear and fair are

Vowels in pre-nasal position are not treated any differently in Euro-Saban English. In other words,
we do not find pre-nasal tensing of trap/bath vowel like we would see in North American English.
This pattern is reminiscent of the putputt merger we find in Scottish English.


jeffrey p. williams and caroline myrick

price is dipthongized; the nucleus is fronted to either [] or [], so that
price is realized as [prs] or [prs]. The nucleus of the price vowel
occurring before voiceless consonants (e.g. tight) is higher than when
before voiced consonants (e.g. tide).
The nucleus of the choice vowel is lowered, so that void is [vd], and
sometimes fronted, so that void is [vd].
goat [ou] is monophthongal; usually lengthened (duration), e.g. most as
[mo:st], or with an offglide of [], e.g. most as [most].
(1) whole road

[ho:l o:d]

(2) boat


In the mouth set, the nucleus is backed and raised to // so that house is
realized as /hs/.
3.1.3 Long monophthongs
The fleece vowel is realized as a long monopthong in Saban English, with
teach realized as [ti:t].
goose is monophthongal and very backed for older speakers. In younger
speakers, there is evidence that goose is moving forward.
In rhotic words with stressed nuclear [r], the nurse vowel merges with the
force vowel (Windwardside) or the strut vowel (Hells Gate). When
r-less, the vowel is realized as [], e.g. nurse as [ns].
The start vowel is realized as [r] and is homophonous with the north

Saban English


The force/north distinction/split (i.e. horse/hoarse distinction) has
been preserved on Saba, most strongly in the variety spoken in the village
of Hells Gate, but it appears to be undergoing merger by approximation.
Words with post-back vowel /r/ are typically r-ful in Saban English (see
/r/ discussion below); some lexical exceptions include the words more and
farm, which are typically r-less.

Consonants Initial /h/

Word initial h-dropping is common in Euro-Saban dialects. When initial
h is present it is realized as [j] as in the pronunciation of Hugo as "You-go."
This pattern is most common when there is a following consonant. Despirantization
The processes of despirantization, where interdental fricatives are realized as voicing equivalent plosives, is common in all varieties of Saban
English. Intervocalically and phrase finally, most are glottalized, e.g. math
as [ma] and birthday and [berdei]. In the section below we elaborate on /t/
glottalization. /t/ glottalization
As mentioned above, this process is common in all dialects of Euro-Saban
English. It is in evidence in intervocalic position as in water [wa]. Beyond
this, /t/ glottalization is also found in clusters, hospital as [haspl], and
also word finally, e.g. bet as [b:] and ate as [:]. v/w merger
Euro-Saban English exhibits some merging of [v] and [w] with realization
as [] in the speech of older speakers.
(3) seventh

[sn] /r/
Saban English is not classifiable as purely rhotic or non-rhotic due to a
two-way split for /r/. Post-vocalic /r/ is disfavored in unstressed syllables
and when following a front vowel, while /r/ is favored following a nuclear
/r/ or a back vowel (with the exception of lexical items more and farm).
Based on mixed-effects models run in Myrick (2014), Saban English thus


jeffrey p. williams and caroline myrick

shows the following hierarchy of /r/ favorability: unstressed syllable < postfront vowel < nuclear /r/  post- back vowel.
(4) BAELY any CARS

[Windwardside, W//ma/2012]

(5) the YEA I was BORN

[Windwardside, W//ma/2012]

(6) in the sand UNDA WATA

[Hells Gate, W//ma, 2012]

While /r/ vocalization can occur in word-final position when followed

by a consonant or a vowel, a following consonant favors r-lessness over a
following vowel.
Speakers with higher education (i.e. beyond the seventh grade) tend
to be more rhotic overall than speakers with lower education, supporting
Williams (2010) suggestion that level of education correlates negatively
with degree of rhoticity in speakers of Euro-Caribbean Englishes. Metathesis
Like other varieties of Euro-Caribbean English, Euro-Saban English evidences lexicalized metathesis in forms such as (7) through (9). The feature
is common across communities and ethnicities on Saba.
(7) ask


(8) sistern


(9) pattern

[ptrn] Nasal fronting

Alveolarization of velar nasals is variably realized in grammatical contexts
where the progressive aspect is marked by inflection on the verb in EuroSaban English. Hickey (1999: 45) has claimed that alveolarization is a
widespread areal feature in the anglophone world. Nasal backing
This feature is common in Euro-Caribbean dialects of English spoken
throughout the eastern Caribbean. As in these varieties, Euro-Saban dialects
have town typically pronounced as [ta], down as [da], and ground as
[gra]. Consonant cluster reduction
Consonant cluster reduction is a worldwide feature of nonstandard varieties
of English (e.g. [ms] for mast).


Saban English


3.2.1 Pluralization
Not unlike other Euro-Caribbean varieties (see Williams 2010), pluralization is variably marked in Euro-Saban English varieties. Plural nouns are
variably marked with the Standard English -s suffix as in (10) through (13)
(10) The bananas is put to come ripe.

[Windwardside, W//ch/1983]


I never lock no doors. [Windwardside, W//ma/1983]

While we to do the lightin of the candles. [Windwardside,


All them bones is broken.

[The Bottom, W//oa/2012]

Saban English does, however, use plural -s absence for count nouns
(e.g. mile, year, minute) that follow a quantifier.

sixteen year I spent there


half mile wide and six mile long

(16) she figure me ten cent

[Hells Gate, W//oa/2012]

[Hells Gate, W//oa/2012]

[The Bottom, W//oa/2012]

As noted in Williams (2010), Saban English does not make use of postpositional dem for pluralization. Undoubtedly, this is due to lack of development of a Creole-speaking population on the island.
3.2.2 Pronouns
The pronominal systems of Saban varieties of English make use of the
pronouns found in Standard English but with different case realizations
with some exceptions. Older speakers, in general, make use of subject
pronouns in all positions.

Them cost plenty money.

[Windwardside, W//ch/1983]


It was hard for he.


Took we down.


. . . leave she in charge of the church.

[Hells Gate, W//ma/2012]

[The Bottom, W//oa/2012]
[The Bottom,

In addition to the pronouns found in Standard English, Saban English

incorporates a second-person plural pronoun, aayu, stemming from an


jeffrey p. williams and caroline myrick

assimilated version of all you. It is used equally by all speakers (i.e. it

appears to be a stable feature), and typically appears in both interrogative
and declarative sentences, although only in subject form. Aayu can address
a specific pair or group of interlocutors, or refer to an abstract group (e.g.
all Americans).

Aayu has a Social Security thing [Hells Gate, W//oa/2012]

(22) Aayu know when you . . . boil custard? [Hells Gate,
(23) Aayu dont have this in America?

[The Bottom, B//ch/2012]

3.2.3 Possession
Object-form for possessive-form occurs with first-person singular possession, shown in examples (25) and (26), and subject-form for possessive-form
occurs with third-person singular possession, shown in example (24).

. . . had a problem with he hips


Hes me family anyhow


. . . with me heart

[Hells Gate, W//oa/2012]

[The Bottom, W//oa/2012]

[Hells Gate, W//oa/2012]


Tense and aspect

Tense Present and past tense

As in some Euro-Caribbean English varieties, Saban English varieties do
not always signal present or past tense, either through verbal inflection or
a separate preverbal marker. Example (27) shows an unmarked past-tense
(27) He never answer me.

[Windwardside, W//ma/1982]

In example (27), the semantics of the utterance refers to a specific instance

and the failure on the part of the individual to provide an answer to a question posed by the speaker. The use of the standard third-person inflectional
suffix would convey a different meaning in Euro-Saban English which
would be similar to Standard English whereby the person repeatedly
failed to answer the speaker.19 As discussed further along in our chapter,

It could be argued that what is absent in the phrase is the past-tense marker, rendered in Standard
English as He never answered me. The semantics of the phrase in Euro-Saban English is that the
individual did not answer a specific question and the time frame remains open for him to provide
an answer.


Saban English

third-person present indicative -s is reallocated as one grammatical resource

to signal habitual aspect. Aspect
Aspect is a diagnostic feature of the anglophone Caribbean and has three
exponent features in the region: habitual, progressive, and completive.
Euro-Saban English, not unlike other Euro-Caribbean varieties, evidences
multiple exponence in some categories. Habitual
Euro-Saban English has multiple exponence in the habitual category that
is most likely a product of earlier sociolinguistic patterns that have been
blended through village interactions over time.
Habitual be In Saban English, habituality can be denoted through the
use of be and the -s affix. Habitual be can occur before progressive verbs, as
in example (28), and before clauses beginning with with or without, as in
examples (29) and (30) respectively.
(28) Them bes helping their father with so much.
(29) I dont be with the hot water.

[Hells Gate, W//ma/2012]

(30) That how come I be without a walk.


It dont be nice.

[Hells Gate,

[The Bottom,

[The Bottom, B//oa/2012]

Verbal -s One predominant feature of Euro-Saban English is the use of

verbal -s in all persons to signal habitual aspect as the examples which
follow demonstrate.
(32) They puts on shoes.

[Windwardside, W//c/1982]

The Chinese drinks plenty tea.

[Windwardside, W//oa/1982]

(34) Sometimes they sings me a song.


He has to takes care of she.

(36) I gots to go down there.

(37) I dont knows she.

[Hells Gate, W//c/2012]

[Windwardside, W//ma/1982]

[Windwardside, W//ma/1982]

[Windwardside, W//ma/1982]

I knows how to do a tape recorder.

[Windwardside, W//c/1982]


jeffrey p. williams and caroline myrick

(39) When you goes to St. Johns?

[Hells Gate, W//oa/2012]

(40) Some gets married on the Fort Bay.

[Hells Gate, W//oa/2012]

The affix -s can be used in the present tense to mark habituality/nonstativeness and can be applied to both weak and strong verbs.

On the other islands we dos [du:z] it.

[Hells Gate, W//c/2012]

The verbal -s forms are conspicuously absent in contexts (example 42)

where its use in the historic present in narratives would be expected.
(42) She say I from St. Kitts.

[Windwardside, W//ma/1982] Progressive aspect

Especially in older speakers, this category is marked through "a-prefixing."
Longitudinally, this construction has been stable in Euro-Saban English
varieties over the last thirty years. The use of "a-prefixing" shows alignment
with similar uses in the rural American south: a schwa occurs before a
progressive verb whose first syllable contains primary stress, and whose
final // is fronted to [n], with vowel laxing.
(43) I was a-fishin yesterday.

[Hells Gate, W//oa/2012]

(44) You was a-buildin a house.

(45) Theys a-kissin you.

[Hells Gate, W//oa/2012]

[The Bottom, W//oa/2012]

(46) I was a-swimmin on the Cove Bay.

[Hells Gate, W//c/2012] Completive aspect

Completive aspect is marked with preverbal done as is common in many
other anglophone Caribbean varieties.
(47) You done ate what Is sent you?

[Windwardside, W//oa/1982]

3.2.5 Copula
The copula is typically not present in Euro-Saban English. Copula absence
may occur before the following environments, shown in the examples
below: progressive verb (48), adjective/adjective phrase (49), noun phrase
(50), preposition/prepositional phrase (51-52), and quotative like (53).
(48) I going with her

[Hells Gate, W//c/2012]

(49) You like an old woman!

[The Bottom, B//c/2012]


Saban English
(50) Who your family?

[Hells Gate, W//ma/2012]


He from Saba too

[The Bottom, W//oa/2012]


I know where you from


So I like, Mommy? Where we going?

[Windwardside, W//ma/2012]

(54) While we to do the lightin of the candles.


Is the onliest one here.


[Windwardside, W//ma/1982]

3.2.6 Questions
In older and more vernacular speech, questions are marked by intonational
patterns and not through subjectauxiliary inversion. Younger generations,
however, show a higher preference for subjectauxiliary inversion.
(56) What his name is

[Hells Gate, W//c/2012]

(57) You have been to see it

[Hells Gate, W//oa/2012]

3.2.6 Negation
Negation in Saban English has two main forms, the most common being
aint ([ ]  [n]  [nt]) that is found in many of the nonstandard and
lesser-known varieties of English as well as in English-based creoles. There
is frequent use of aint in the vernacular for singular and plural number as
well as being in place of didnt and havent. Standard English preverbal not
is also used in less vernacular contexts, where either social circumstances
or less dense network ties influence the shift.

There aint too many bad things

(59) That aint true

[Hells Gate, W//oa/2012]

[Hells Gate, W//c/2012]

(60) He aint send me nowhere.

(61) I aint say nothing
(62) She aint decide yet.

[The Bottom, W//oa/2012]

[Hells Gate, W//oa/2012]

[Windwardside, W//oa/2012]

3.2.7 Prepositions
Euro-Saban English varieties evidence prepositional usage that is not found
in Standard Englishes. Locative prepositions tend to be omitted before
macro locations, such cities, countries, other islands, as well as the villages


jeffrey p. williams and caroline myrick

on Saba. Examples (63) and (64) show examples of omissions before island
and state names.
(63) The other two was born Saba [Hells Gate, W//oa/2012]
(64) I went Sint Maartens for school [Windwardside,
When referring to micro locations, Saban English speakers typically employ
by or to for "at." These locations include homes, stores, restaurants, and
landmarks. The generalization of by to mean both "by" and "at" may be
a substrate effect from Dutch (as Dutch, like German, has only one word
to mean both "by" and "at"), supported by evidence of the same substrate
effect seen in the English of the Pennsylvania Germans (Wolfram, personal
communication, April 11, 2013) and Yiddish speakers (Benor 2012).
(65) The party was by my house.

[Windwardside, W//c/2012]

(66) there by Saba Treasures (the restaurant)

[The Bottom,
(67) the reception was kept by her parents home [Hells Gate,
(68) She teaches me to school.

[Windwardside, W//c/1982]

(69) (the house) right over to the end

[Hells Gate, W//oa/2012]

(70) Down here to the harbor

[The Bottom, W//oa/2012]


[Windwardside, W//c/1982]

She teaches me to school.

(72) We just shared it to the neighbors.

[Hells Gate, B//ma/2012]

3.2.8 Complementizers
Euro-Saban English also makes extensive use of for to complementizer
(73) They is ready for to come ripe.

[Windwardside, W//ma/1982]

3.3 Lexicon
Vocabulary of Saban English varieties exhibits similarities to other
Caribbean English varieties, both creole and non-creolized.
(74) drop
(75) smokey

to say

Saban English
(76) remind
(77) from



(78) molest

annoy or bother

(79) ground


(80) awoy/woy

a greeting

4 Conclusion
The varieties of Saban English are some of the more robust in the Caribbean
English sociolinguistic landscape, even in spite of the islands small size.
Relatively stable populations and favorable economic conditions have supported the maintenance of the local populations, although emigration for
education and employment has impacted males to a higher degee than
females. An emerging Saban identity has developed throughout the colonial period and contributed to a projected Saban sociolinguistic identity,
which has valued the vernacular.
Throughout its European historical period, isolation has played a key role
in the trajectories of Saban Englishes. Isolation, in the Caribbean context,
is not an absolute. Sabans we have spoken with described interactions with
other islanders, both on Saba and on other islands. The expanses of sea that
separate Saba also join it to other sociolinguistic communities. Large-scale
external impacts from tourism or immigration have not taken place on
Saba, as the island still does not have the capacity to receive larger aircraft,
with the worlds shortest commercial runway at just under 400 meters
in length. These social conditions have coalesced so that Saban English
varieties are not highly endangered, which also sets them apart from many
other lesser-known Englishes worldwide.

Beckles, Hilary. 1986. From land to sea: runaway Barbados slaves and servants,
16301700. In Gad J. Heuman, ed., Out of the House of Bondage: Runaways,
Resistance and Maroonage in Africa and the New World. London: Frank Cass,
Benor, Sarah. 2012. Becoming Frum: How Newcomers Learn the Language and
Culture of Orthodox Judaism. New Brunswick, NJ: Rutgers University Press.
Crane, Julia. 1971. Educated to Emigrate: The Social Organization of Saba. Amsterdam: Van Gorcum.


jeffrey p. williams and caroline myrick

1987. Saba Silhouettes: Life Stories from a Caribbean Island. New York: Vantage
Eliason, Eric A. 1997. The Fruit of Her Hands: Saba Lace History and Patterns.
Saba: Saba Foundation for the Arts.
Espersen, Ryan. 2009. From folklore to folk history: contextualizing settlement
at Palmetto Point, Saba, Dutch Caribbean. Masters thesis, University of
Hartog, Jan. 1975. History of Saba. Saba: The Saba Artisan Foundation.
Hickey, Raymond. 1999. Ireland as a linguistic area. In James P. Mallory, ed.,
Language in Ulster. Holywood, Ireland: Ulster Folk and Transport Museum,
Johnson, Will. 1979. Saban Lore: Tales from My Grandmothers Pipe, 3rd edition.
Published by the author.
Karras, Alan L. 1992. Sojourners in the Sun: Scottish Migrants in Jamaica and the
Chesapeake, 17401800. Ithaca: Cornell University Press.
Myrick, Caroline. 2014. Putting Saban English on the map: A descriptive analysis
of English language variation on Saba. English World-Wide 35(2): 16192.
Williams, Jeffrey P. 1984. White Saban English: a socio-historical study. Masters
thesis, University of Texas at Austin.
2010. Euro-Caribbean English varieties. In Daniel Schreier, Peter Trudgill, Edgar
Schneider, and Jeffrey P. Williams, eds., The Lesser-Known Varieties of English:
An Introduction. Cambridge University Press, 13657.
2012. English varieties in the Caribbean. In Raymond Hickey, ed. Areal Features
of the Anglophone World. Mouton: de Gruyter, 13360.

c h a p ter 8

St. Eustatius English

Michael Aceto


St. Eustatius is an English-speaking island in the Dutch Caribbean. This

dialect displays a handful of correspondences with other Englishes spoken
in geographically proximate areas, but what is most noteworthy about this
dialect is that so much of its grammar is significantly different from many
of those same nearby varieties. Historical, linguistic, and ethnographic
data are interwoven to make the case that Statian English sounds different
from most other Englishes of the Caribbean basin because the colonizing
and settlement patterns of the island differed from plantation societies
focusing on the production of cash crops. St. Eustatius was a commercial
center instead, offering an entrepot for goods (and, at times, slaves) for sale
to customers from the eastern rim of the Americas. In this importexport
context, English as a lingua franca of trade emerged with its own distinctive
St. Eustatius English is a dialect that has never been the focus of a single
piece of published research until Aceto (2006), even if it has been alluded
to in other works (e.g. Hancock 1987; Holm 19889: 4525). Williams
(1983: 95) notes that there are no published sources on the contemporary
Englishes of the Windward Netherlands Antilles (i.e. Saba, St. Martin,
and St. Eustatius), and this research characterization has not changed
much in the last nearly thirty years. Hancock (1987) presents fifty kernel
Statian phrases and sentences (displayed among thirty-three other English
varieties). Holm (19889: 4525) contains a short section on Statian. Parsons
(193343) includes St. Eustatius in her wide range of West Indian folklore
This chapter briefly examines the grammatical features of English on
St. Eustatius. The language data is based on several weeks of fieldwork
I would like thank the people of St. Eustatius who were generous in sharing their time and language
with me. Any errors or shortcomings are, as always, mine alone.



michael aceto

conducted on the island in the summer of 1998. The historical presentation

and linguistic data are more expansive in Aceto (2006). The present chapter
summarizes Aceto (2006) and then provides a discussion that explores the
role that Statian English played in the Americas and what dialect and creole
studies can learn from this English variety.

2 Background
St. Eustatius lies on the eastern edge of the Caribbean Sea approximately
40 miles to the south of St. Martin. Today, the residents of St. Eustatius
are, for the most part, people of African descent, though historically this
was not always the case (as is discussed below). There is some immigration
from Papiamentu-speaking islands such as Aruba and Curacao that are also
part of the wider Dutch Antilles as well as from geographically proximate
English-speaking areas such as Nevis and St. Kitts. Only one or two flights
a day in small propeller planes connect Statia with St. Martin, which is
one hub for air flights in the eastern Caribbean. Before this service was
established in the last fifty years, one could only travel to the island by
In 1933 Dutch became the official language of the public school system
in St. Eustatius; in 1976 English was made the language of public education. However, Dutch is again the language of education, though some
English language instruction unavoidably and undoubtedly occurs since,
for those who are locally born, Dutch is no ones native language. Statians
have changed their minds at the voting booth several times in the twentieth century about which language of instruction they would prefer general
classes to be held in, either English or Dutch. I was surprised by how
strongly locals felt that they should know some form of Dutch since (as
was explained to me) they are part of the Dutch Antilles. One consultant
even referred to the Netherlands as the mother country of Statia. Promising students who are more fluent in Dutch often go to the Netherlands
for study. Statian English is most folks native language, except for some
Papiamentu speakers, who also know at least some of the local vernacular,
and possibly several administrators and educators who originally arrived
from the Netherlands.

3 Statia in the colonial period

In the seventeenth century, St. Eustatius was sought after by various European colonial interests due to its central location and proximity to other

St. Eustatius English


islands in the eastern Caribbean. Williams (1983: 97) writes, St. Eustatius
was highly prized by the Dutch due to its proximity to St. Kitts and other
British possessions. This proximity to anglophone locations, as well as
robust trade with these same islands, encouraged the local emergence of
English as the vernacular of Statia. The island is relatively small and its
drought-ridden climate eventually made it largely unsuitable for use as a
significant plantation colony. Drought conditions continue to the present
French, Spanish and English colonists were already buying slaves at Statia
by 1675 (Hartog 1976: 49). Keur and Keur (1960: 39) state, The main traffic
was with St. Kitts, Barbados and St. Thomas. In 1679, one transport of
between 200 and 250 African slaves went directly to St. Eustatius (Postma
1990: 1956). Until this event, slaves were generally supplied from Curacao,
the center of the Dutch West India Company slave trade during this
period. Though there are not many demographic figures for the seventeenth
century, Postma (1990: 197) indicates that in 1665 Statia contained 330
Europeans, including children, and somewhere between 800 and 1000
In the eighteenth century, Statia briefly found its niche in the West
Indian economy as first a central slave-trading depot in 1720s, and later in
the 1770s, when it became known as the Emporium of the Caribbean
where all manner of material goods could be purchased and exported.
Postma (1990: 199) states, During the 1720s St. Eustatius briefly became
the cornerstone of the Dutch slave trade . . . even more than Surinam at
that time. Crane (1999: xv) writes, By the 1720s, although the West India
Company had four plantations on Statia, all were neglected or idle. The
production of agricultural goods via the plantation context was rarely the
economic focus of the island.
The local slave population declined to approximately 120 in 1705 and
then grew to 463 by 1709. Hartog (1976: 34) reports the islands population
in 1715 as 1,274 inhabitants, of which 750 were slaves. Postma (1990: 197)
argues that by 1734 the European population had grown to 519, while the
African population had increased to nearly 1,000. By this point in time,
depending on the sources consulted, the African and African-descended
population ranged from approximately one half to less than two-thirds of
the local population. These relative proportions hardly reflect the hugely
disproportionate demographic ratios associated with the language ecology
of most plantation societies (see Mufwene 2001).
Statia emerged as a local slave-trading center by about 1721, just as
Curacao was losing this distinction. During the 1720s, the Dutch West


michael aceto

India Company landed twenty-three ships in St. Eustatius holding approximately 11,000 African slaves who were subsequently sold and reshipped
to other colonies in the Americas, mainly francophone and anglophone
territories and islands (Postma 1990: 199). Attema (1976: 29) states that
during the 1720s, the Dutch transported 2,0003,000 slaves a year to
St. Eustatius (most were resold to other islands in the Caribbean), which
would make the sum of slaves transiting the island approximately 20,000
30,000 instead of 11,000 for that decade, as reported by Postma. Le Page
and DeCamp (1960: 58) also report, In the eighteenth century St. Eustatius
was distributing, particularly to the Leewards, 20003000 slaves annually.1
Hartog (1976) suggests that the Dutch trade with anglophone locations in
the eastern Caribbean such as Antigua and St. Kitts as well as other British
colonies in North America was central to the economy of Statia.2 Williams
(1983) states that St. Eustatius supplied St. Kitts with 1,000 slaves from
1721 to 1723. Only during the 1720s did the island of St. Eustatius, having
replaced Curacao as the Dutch slave depot in the Caribbean, briefly become
a more important market for slaves than Surinam. But after 1726 Suriname
became the dominant Dutch slave market in the West (Postma 1990:
The St. Eustatius slave trade reached its peak in 1726 and then seemed to
end abruptly by 1729 (Postma 1990: 200). From this brief peak in Statian
slave-trading, the island fell into a lull in general trade until the 1750s-70s
when it began to earn the names associated with great commerce such as
Money Mountain, Golden Rock, Diamond Rock, and Emporium of the
Caribbean. It is during this period that St. Eustatius became the commercial
center of the Caribbean (Keur and Keur 1960: 40). Ships originating from
the thirteen British North American colonies in what would eventually
become the United States of America used the facilities on Statia in order
to purchase goods and arms in fighting the subsequent American War
of Independence. Goslinga (1979: 83) reports that, in 1774, as many as
20 American ships at a time could be found in Statias harbor. Thus,
contact with English speakers was intense on St. Eustatius during the latter
half of the eighteenth century. Postma (1990: 223) writes, During the last
years of that decade [the 1770s] as many as 3000 ships anchored off St.
Eustatius annually, and as a neutral harbor the island played a significant
role in supplying American rebels with ammunition and other supplies.

Le Page and DeCamp (1960: 58) indicate that the British colony in Jamaica purchased a significant
number of their slaves in St. Eustatius.
The role of St. Kitts as a possible geographical dispersal point for creole features due to the islands
role in the settlement of the anglophone West Indies is explored in Baker and Bruyn (1998).

St. Eustatius English


Attema (1976: 31) writes, During the American War of Independence Sint
Eustatius specialized in the illegal delivery of weapons and gunpowder to
the rebellious English colonies . . . Trading relations between Sint Eustatius
and the North American rebels were so good that the new commander,
Johannes de Graaff, appointed in April 1776, opened the harbour to all
American shipping.
The number of ships visiting the island continued to grow significantly
in the 1770s, and the population of Statia grew to heights it would never
reach again. Hartog (1976: 43) states that the population of the island
in 1750 was 2,315 persons, of which 1,513 were slaves (approximately 65
percent). He puts this total figure in 1779 at 3,056 persons with slightly
more than half of them as slaves (1976: 4445).3 In 1790, Statia reached
the highest population in its history at 8,124 persons: 643 freed men of
color, 5,140 slaves (approximately 63 percent of the total population), and
2,341 Europeans (Hartog 1976: 52).4 Goslinga (1979: 85) estimates that the
number of whites passed 3,000 in 1791.
The free trade in slaves was forbidden in 1784. The French controlled
the island again from 1795 to 1801. The English took over again for one
year in 1801. The territory did not return to Dutch control until 1816. The
Netherlands abolished the slave trade in 1814 and the importation of slaves
from Africa to its islands in the Caribbean in 1821 (Attema 1976: 30).

4 Demographics
From the population peak of 8,124 persons (approximately 63 percent of
the population as slaves) in 1790, the number of Statias residents began to
dwindle. Once the economic base deteriorated at the end of the eighteenth
century, few immigrants moved to the island and most folks of European
descent moved away. After emancipation in 1863, many of the Africandescended males began to move about the Caribbean and the Americas
in general (often as sailors on whaling vessels; Crane [1999: xxiii]) in
search of work. This itinerant labor-related pattern continues to this day.
The following demographics corroborate this characterization (all figures
are from Hartog 1976: 102, 127, 134, except as indicated). In 1818, 2,668

Keur and Keur (1960: 40) assert that the population of St. Eustatius in 1780 was between 20,000
and 25,000, a figure that seems hard to reconcile in terms of housing and infrastructure for anyone
who has visited this small island; furthermore, such a large population on a relatively small island
would surely have been mentioned by Schaw in her journal (see Andrews and Andrews 1934).
For comparison, in the same year, nearby Dutch St. Martin contained 4,230 slaves and 1,290
Europeans and freed persons (Hartog 1976: 52).


michael aceto
Table 8.1 Results of 1974 Census in St. Eustatius
according to place of birth
St. Eustatius
St. Maarten (Dutch)
St. Martin (French)


Hong Kong 1
St. Thomas 2
Tortola: 1
St. Croix 2
Guyana 1
St. Vincent 3
Grenada 1
Dominica 2
St. Kitts 135
Nevis 31
Dominican Republic 5
St. Barths 1
Nassau, Bahamas 1


persons resided in Statia: 501 of European descent, 302 freed persons,

and 1,865 slaves. In 1863, 1,183 slaves were freed. By 1884, there were 1,600
persons on the island, and no more than 50 were Europeans or of European
descent. At this point, St. Eustatius reveals a population largely comprised
of African-descended peoples, and the total population figures begin a
decline that lasts for more than fifty years. In 1916, the total population was
1,431; in 1920, 1,315; in 1929, 965 (Goslinga 1979: 151); in 1935, 1,198; and
in 1948, 921, the lowest figure on record. This last demographic decrease
reflects the fact that many male residents had immigrated to Aruba and
Curacao for work.
In the last half of the twentieth century, the numbers begin to stabilize.
In 1960, the total was 1,014 and in 1974 there were 1,421 residents: 676 males
and 745 females. The 1974 census table arranged according to place of birth
(see Table 8.1; Hartog 1976: 136) reveals the considerable intra-Caribbean
migration that even small St. Eustatius has received. The pervasiveness
of intra-Caribbean migration since emancipation and its effect on language and culture have yet to be properly understood by researchers for
all areas of the Americas. Papiamentu speakers from Aruba, Bonaire, and
Curacao represent the largest non-English-speaking language group among
the immigrants in Statia. Speakers of English-derived varieties represent
the largest group of immigrants. The geographically proximate islands of
St. Kitts and Nevis represent the single largest source of immigration to
St. Eustatius. This observation is important in understanding the handful

St. Eustatius English


of creole-like structures that Statian English reveals (discussed below). The

population of Statia has stabilized at approximately 2,000 persons today.

Linguistic features of Statian English

The following data are all from my fieldwork except as indicated where
the data come from Hancock (1987) or Parsons (193343). I lived on St.
Eustatius for approximately two weeks in July 1998. I recorded most of
the following data (taken from approximately twelve hours of tapes) as
naturally occurring discourse in two locations in Oranjestad (or what
Parsons [193343: 376] labeled Orangetown): on a porch centrally located
in the community on one of the main roads and at a local bar in which
mostly but not exclusively men were drinking as well as playing cards
and dominoes. Data were corroborated by a limited set of interviews as
well. The approximately 1824 informants were mostly males in their 30s
50s.5 The data below are presented within a broad International Phonetic
Alphabet (IPA) transcription.
Unlike other anglophone Caribbean fieldwork locations where I have
gathered data, Statians do not typically use a discrete name for their language apart from the generic term English. For example, on the Panamanian island of Bastimentos (or Ol Bank as the locals call it), creole speakers
sometimes referred to their language as Guari-guari, and on Barbuda they
often called it simply dialect, a fairly common designation in the eastern
Caribbean. Both locations also called their language broken, raw, or
broad English among other pejorative labels. Recent fieldwork on the
island of Dominica reveals an English-derived creole that is called Kokoy
(Aceto 2010). The only term I could uncover on St. Eustatius was the term
English. I did not hear or record local folks using the term with any of
the aforementioned pejoratives either, but that certainly does not mean
that Statians have not internalized some of the same negative judgments
about their vernacular language that are so common in other anglophone
Caribbean locations. One informant explained that locals occasionally
identify their native language as bad or pidgin English.6 The situation
on St. Eustatius is not unique. Mufwene (2001: 85) points out that during

Some of the social contexts of card playing and dominoes for my audio recordings involved at least
four and as many as six informants being recorded at once.
However, another consultant insisted on two occasions that her English variety was no different
than mine even though, from a descriptive perspective, her variety of English exhibited nearly all of
the general features presented in this chapter.


michael aceto

his research on the Sea Coast Islands, one Gullah speaker told him, You
call [our vernacular] Gullah; we call it English.
There exist relatively little linguistic data on Statian English. The work
of Elsie Clews Parsons (193343: 37686) reveals some folk tales from Statia
which largely corroborate the data I gathered on the island. From both
my fieldwork and Parsons (193343) Statian English emerges as a dialect of
English, similar in regards to many features to African American Vernacular English (AAVE). In fact, one may notice that Statian English often
seems like a Caribbean version of AAVE. That is to say, there are few
features of Statian English that are so-called creole features though there
are at least three that may be considered so: unmarked verbs interpreted
as past, the preverbal future marker go, and the post-nominal pluralizer
dem. In the last ten or so years, it has been asserted that so-called creole
languages manifest structural characteristics different from non-creole languages (McWhorter 1998, 2000; Parkvall 2001),7 but many researchers have
remained unconvinced that these assertions/alleged diagnostic structural
features are exclusive to the group of languages researchers call creoles (Plag
2001). Rizzi concludes creoles do not look different from other natural
languages in any qualitative sense (1999: 466). Many creolists would probably agree with Mufwene (2000, 2001) that creolization is a social process
and not a structurally defined one. I will make reference below to Parsonss
ten pages of tales to corroborate my own data and/or to demonstrate that
she found features that I was unable to record in my naturally occurring
discourse or interviews.8
At least two language-external factors have greatly impacted the formation and emergence of Statian English and illustrate why this variety
may appropriately be considered as a dialect of English (if not all so-called
English-based or -derived creoles as well; see Mufwene 2008).9 First, there
seems to have been a significant segment of the local island population who
were first-language speakers of what was emerging in the eighteenth century
as North American English varieties (despite the fact that Statia was a Dutch


I have always found McWhorters (1998) analysis less than compelling because it does not discuss features that these languages display but rather features they lack; thus this sort of analysis is predisposed
to finding so-called creole languages comparatively deficient or lacking. Thus, subsequently designating them as simpler than other human languages seems only a step away from first presenting
them as lacking specific features.
It is important to remember that folk tales deliberately performed in story-telling mode may tend
to invoke features not normally associated with everyday vernacular language.
However, the term creole is still a useful one to designate the sociohistorical contexts in which many
of these languages emerged (Mufwene 2000).

St. Eustatius English


colony). Furthermore, a significant segment of the European-derived population was also comprised of second-language speakers of English (i.e.
Africans and their descendants were not the only persons acquiring a variety of English). Lastly, the demographic history of St. Eustatius reveals a
less disproportionate ratio between those of African and European descent
than was often found in other more typical creole-speaking plantation
areas such as Jamaica, Antigua, and even Suriname. That is, in Statia,
Africans and persons of African descent in subsequent generations had
a greater probability of hearing first- and second-language varieties spoken by Europeans, Africans, and slaves born locally than in many other
anglophone Caribbean locations. Africans and persons of African descent
only comprised as much as one half of the population (ranging to a high
of approximately 65 percent) at various times in the islands history. It is
often an overlooked fact in creole studies that many Englishes emerged
in the Americas, among peoples of both African and European descent,
that are structurally closer to dialects of the lexifier, e.g. varieties spoken
in the Bahamas (see Childs, Reaser, and Wolfram 2003), the Turks and
Caicos islands (see Cutler 2003), Anguilla (see Williams 2003), probably
the Cayman islands, and AAVE just to name a few of the likeliest candidates (see below and Aceto 2003 for a discussion of what he calls dialect
creole varieties).
Hancock (1987) provides one of the few printed sources of Statian linguistic data. Those data reveal a prevalent occurrence of the suppletive to
be form in Statian English: [wz] was, e.g. [tri hi frn waz d] Three
of his friends were there (283), [f yu wz stl di li:d] If you were still
the leader (321), [a js wz catn] I was merely chatting (322), which
was corroborated by my data and Parsons (193343: 376, 378, 3813, 386) as
well. Common constructions such as completive aspect done (e.g. [i dn
sat mai han]) I already played out my hand (in cards) and possessive constructions with successive noun phrases without inflectional morphology
(e.g. [de kaa] their car) are heard on the island, but these constructions
are present in a number of other Englishes as well (e.g. general Southern
American English and AAVE). Many other so-called creole features are
not found in this variety, as is discussed below. The following is a short
presentation of features found in Statian English based on my fieldwork,
Parsons (193343), and Hancock (1987).10 The following presentation does

Crane (1999) transcribed a series of autobiographical sketches for her informants. She then sent the
individual chapters back to the informants for their approval and corrections. Though it is possible
to glean some vernacular usages from this work, most of the stories are presented in a variety close
to general standards of written English.


michael aceto

not pretend to be exhaustive. For more a more complete data set, consult
Aceto (2006).

Phonological features

I selected the following phonological features of Statian English because

they offer a contrastive glimpse into the generally assumed picture of
Caribbean English phonology.
5.1.1 TH-stopping
The neutralization of the fricatives // and // as the stops /d/ and /t/,
e.g. /t/ thing and /fada/ father, is a common feature of many dialects of
Caribbean English as well as in regional, ethnic, and social dialects spoken
in North America and Great Britain (which often display reflexes different
from those in the Caribbean). This process creates new homonyms in the
specific dialects in question, e.g. thintin [tn], faithfate [fet], though
dough [do], breathebreed [brid]. So-called TH-stopping is the general
norm in the Caribbean, including in Statia, but the fricatives [] and [] are
also heard to a significant degree in naturally occurring speech in informal
contexts (i.e. playing poker or dominos, drinking in a bar). Parsons (1933
43) reveals a similar pattern going back approximately seventy years. For
example, within the same folk tale both [t] and [] are realized by the same
informant: three/tree (382), months/monts (383). Cutler (2003)
makes a similar observation about this feature in the English of Gran Turk
Island as does Williams (2003) about some varieties of English spoken in
(1) [ js wana ro m daun]
I just want to throw them down.
(2) [hau ya dun stupid ]
Why do you do stupid things?
5.1.2 /v//w/ merger
The contrast between /v/ and /w/ is often neutralized or merged in the eastern Caribbean. That is, many dialects of Caribbean English (e.g. Bahamian,
Bermudan, and Vincentian)11 may alternate [w], [] (the voiced bilabial

Of course, Bermuda is not in the Caribbean but is often tied geographically to North American
varieties of English. However, some of the linguistic features of Bermudan English pattern with
varieties spoken in the Caribbean. The Bahamas are also often considered part of the geographical
distribution of North American varieties of English but they too often pattern with Caribbean
varieties regarding many features.

St. Eustatius English


fricative), or [] (the voiced labiodental approximant) for words which

in metropolitan varieties begin with [v], e.g. village [wl]. This feature
may be related to component dialect varieties of English heard in the
Caribbean in the eighteenth century which contain this same alternation
(e.g. Cockney) or possibly to African languages that lacked the /v/ segment. Parsons (193343: 380, 382, 3846) robustly documents this same
5.1.3 Rhoticity
Statian English is primarily non-rhotic, although post-vocalic [r] is variably pronounced in some contexts by some speakers. Williams (2003)
documents a similar pattern for the Webster dialect of Anguilla.
(3) [mi gon er]
Im going there.
(4) [wan man gat wan ausan rntr]
One man has a thousand renters.
(5) [ ausan rntr i hav]
He has a thousand renters?
(6) [a don wan hir]
I dont want to hear it.
(7) [kaava ga mor skuln dn yu]
Kava has more schooling than you.

Grammatical features
Tense, mood, and aspect Past tense

What is perhaps most surprising about Statian English is that the data
I collected do not reveal even a single instance of a preverbal past-tense
marker. Parsons (193343: 37686) and Hancock (1987) also do not reveal
any such grammatical markers. Many other varieties of restructured English
in the western hemisphere, especially those languages often identified as
creoles by linguists, exhibit one or more of the following preverbal relative
past markers bin, mi(n), woz, or di(d) as grammatical strategies to indicate
relative past events. One interviewed informant rejected preverbal di(d)
and bin as possible local forms. The Statian English data below mostly
reveal verb phrases unmarked overtly for the past tense but also suppletive
forms (e.g. was as the past tense of is), strong verb forms indicated by


michael aceto

a change of vowel (e.g. brought), and even a few instances of past-tense

bound inflectional morphology (but always in my data, with the root verb
to want [i.e. wanted ]). The data gathered by Jeffrey Williams as displayed
in Hancock (1987) do not reveal a single instance of preverbal past-tense
marker for St. Eustatius nor for the geographically proximate Englishderived variety spoken in Saba. Unmarked verbal forms are exemplified by
the data:

[si bg tu km nau agn]

She asked to come now and again.


[a tl ar s ms]
I told her that she must.

Events overtly marked for past events include at least one verb form revealing bound inflectional morphology, several strong verbs that indicate the
past by a change in vowel as occur in other dialects of English spoken in
the Americas, and the suppletive form was, which is, of course, common
in many locations where varieties of English are spoken.
(10) [a wantd tu kl it an nks wan]
I wanted to kill it and the other one.
(11) [an luk wa si brat ya]
And look what she brought you.
(12) [no, a gan bak p di ailan]
No, I went back up the island.
(13) [i gan ardi]
He went/has gone already.
(14) [faama waz a nais man]
Farmer was a nice man.
Hancock (1987) does not reveal any strong verb forms or verbs marked
for the past with bound inflectional morphology, though the suppletive
form was occurs in several instances. However, Parsons (193343) reveals
instances of both verb types: dipped (376), started (377), tought (378),
met (379), tol (381), took up (381), shtruck (384), sent (385), trew
(385), and slept (385). Parsons (193343: 13343) also reveals many instances
of went (376, 378, 379, 382, 383, 384, 385, 386) as well as was (378, 382). It
is curious that gone as a past form of go does not occur in her data since
it is so common in the anglophone Americas. One might suspect that

St. Eustatius English


she refined the data; however, Parsons goes to some length to represent
the vernacular features of Statian English within the conventions of the
standard English spelling system. Its a puzzling absence. Progressive aspect
Present progressive aspectual constructions are overwhelmingly signaled
by the bound inflectional morpheme [-in] attached to verbs. This feature
is a common progressive form in the English-speaking world and is often
associated with so-called mesolectal varieties if one subscribes to the terms
and assumptions associated with creole continuum and the related concept
of decreolization. The perspective in this chapter is that, not withstanding
normal diachronic change that all human languages exhibit everywhere,
Statian English looks or sounds similar to the variety spoken during the
colonial period. In other words, I will not resort to so-called decreolization
and the purported creole continuum to try to explain the synchronic shape
of Statian English. For a detailed discussion of the limitations of the creole
continuum abstraction, consult Aceto (1999), (2002a), and (2003). I was
unable to locate any deep or so-called basilectal consultants and it seems
unlikely that Statian English ever exhibited many features associated with
this often-presumed historical ancestor in creole studies. In fact, Parsons
data, though similar to mine, appears more standard-like, and thus one
could argue that Statian has shifted for some speakers somewhat towards
features associated with nearby islands like St. Kitts and Nevis in the last
sixty plus years or so.
In my data, there are several instances of preverbal a (and even a single
instance of da [d]) to indicate this grammatical distinction as well, despite
the fact that several informants rejected preverbal a (and de) as possible local
progressive aspectual forms when interviewed. Hancock (1987) and Parsons
(193343) reveal only the verbal suffix -in to indicate progressive aspect.
Preverbal a is a form strongly associated with the eastern Caribbean and it
is likely that preverbal grammatical marker a was brought in with speakers
immigrating from one of the other islands in the Lesser (or even Greater)
Antilles. Small private boat traffic between Statia and other geographically
proximate islands of the eastern Caribbean (specifically St. Kitts, Nevis)
that regularly reveal this feature is common.12 It is important to remember
that immigrants from St. Kitts and Nevis represent the single largest group

Parsons (193343: 376) reveals that one of her informants for St. Eustatius, though locally born,
lived in St. Kitts for forty years; another informant was born in St. Kitts and lived there for some
unspecified amount of time before moving to Statia.


michael aceto

according to the 1974 census (see above). Past progressive constructions are
typically indicated by was + Verb-in.
(15) [ai gon nau]
Im going now.
(16) [a train tu sii]
Im trying to see.
(17) [a js pkn fait nau]"
Im just picking a fight now.
(18) [hu a bt]
Whos betting?
(19) [das wai am lukn ta tl ya wan taim]
Thats why Im looking to tell you once.
(20) [a tl ya, de du ratn ings tu ya, don de]
I told you, they are doing rotten things to you, arent they?
(21) [we de d luz ni ya wan]
What are they losing? Anything you want.
(22) [a woz waakin n di rod]
I was walking in the road. Completive aspect
Preverbal completive aspectual done is common in Statia, as it is in many
parts of the anglophone Americas, e.g. [i dn sat mai han] I already
wasted my hand [of cards]. Hancock (1987: 296) corroborates this form,
but, strange as it may seem given its common occurrence throughout the
Americas, it does not occur in Parsons (193343). Again, it is a puzzling
absence. Habituality
The use of preverbal doz to indicate habitual aspect is often considered diagnostic of anglophone eastern Caribbean varieties (see Holm 19889: 158
60). However, it appears to be absent from Statian English.13 The geographically proximate island of Saba, which shares a similar Dutch/anglophone

Habitual doz is generally thought to be absent from Jamaican Creole English, but I have documented
it in speakers from Bastimentos, Panama, whose ancestors were largely immigrants from Jamaica
(as well as from Providencia and San Andres) in the late nineteenth and early twentieth centuries.

St. Eustatius English


history, also appears to lack this feature. In Statia this aspectual distinction
is indicated by bare or unmarked present tense verbs (e.g. [si sii si brd]
she sees her brother (on weekends) (Hancock 1987: 288) as it is in many
varieties of English in the Americas. Additionally Statian English may add
adverbials such as aal taim all the time or aalweiz always to indicate
habitual, continuous, or repetitive actions. Past habitual actions are indicated by used to [yuustu], which is confirmed by Parsons (193343: 383).
However, as is common in many Englishes of the Americas, progressive
and habitual aspectual strategies often semantically overlap; consequently,
it seems that habituality in Statian English can also be signaled by aspectual
forms such as preverbal a and the suffix -in as the following data illustrate:
(23) [a tl ya, de du ratn ings tu ya, don de?]
I told you, they are doing rotten things to you, arent they?
(24) [yu hav siin r la taim]
Have you been seeing her for a long time? Futurity
Futurity is indicated by preverbal [go] (or one of its reflexes, e.g. [ga]) or
present forms in some instances. This feature is one of three or four features
associated with so-called creole languages that may distinguish this variety
from what are traditionally considered dialect varieties of English spoken in
the western hemisphere. Two other such features are the preverbal negator
no and the post-nominal pluralizer dem. Regarding most other grammatical
features Statian English resembles English dialects spoken in the Americas.
In the western Caribbean, gwain is a common future-tense marker but
was rejected by all interviewed informants in Statia even though goin, which
is diachronically the origin of gwain, is commonly heard on the island;
forms related to gonna are spoken as well (e.g. [gn]).14 No occurrences
of gwain were recorded in Statia. In at least two instances, (25) and (27),
futurity was indicated by the present-tense verb form and some reference
to time (e.g. two oclock, next time, etc.) instead of any overt grammatical
marker. Hancock (1987) presents goin (290) (see discussion of progressive
aspect above), gon [go] (301, 302), and go (304) as future markers, all of
which were recorded in my data as well. Parsons (193343) presents many
instances of goin as a future marker (e.g. 277, 380, 386), but no occurrences
of either go or some reflex of going to (e.g. gonna).

The potential future tense marker gwain, which is so common in the western Caribbean, was
rejected in other anglophone eastern Caribbean research locations such as Barbuda (see Aceto
2002b) and Dominica (Aceto 2010) as well.


michael aceto

(25) [ai km daun di rod tu oklak]

Ill come down the road at two oclock.
(26) [aim gn kc m . . . gn kc m]
Im going to catch him . . . going to catch him.
(27) [a sii ya neks taim smwier]
Ill see you next time somewhere.
(28) [a ga rez yu mo do]
Im going to raise you more though.
(29) [mo go aks ya]15
Am I going to ask you?
(30) [wi go fain yu]
Well find you.
(31) [a wan tl ya wa a ga gi ta yu]
I want to tell you what Im going to give to you.
(32) [ya go an drink]
Will you have a drink?/Are you going to drink?
(33) [yu no go pe]
Youre not going to pay?
(34) [a gn wk tumaro]
Im going to work tomorrow.
5.2.2 Copula forms
By far, the overwhelming number of copula forms are is for the present
and suppletive was for past contexts. However, the nominative copula a
occurred as well as instances of no overt copula form at all. The locative
copula de never occurred in my data nor did it occur in Hancock (1987) or
Parsons (193343). Consultants interviewed rejected de as a local form as
well. The copula was unrealized in attributive positions before an adjective,
which is quite common among many English dialects. However, it was
frequently absent before noun phrases, which is reminiscent of similar
patterns in AAVE.

It is not uncommon for pronominal forms preceding go to exhibit phonetic effects similar to vowel
harmony rules. That is, the vowel of the verb go seems to have affected the vowel of the pronoun
mi. I have documented a similar pattern in Barbuda (Aceto 2002b: 2334).

St. Eustatius English


(35) [aiz an di rod tu oklak]

Ill be on the road at two oclock.
(36) [hu a dm]
Who are they?
(37) [daiman a blak]
Diamond is the block.
(38) [hu dm]
Who are they?
(39) [bai di we, si lat fkn prablm]
By the way, shes a lot of fucking problems.
(40) [yu a hom]
Were you at home?
(41) [vribadi sk]
Everybody is sick.
(42) [wn move bl haus hir si w hir]
When Move built a house here, she was here.
(43) [wn a miit move n sent kroi, move waz a krIscn]
When I met Move in St. Croix, Move was a Christian.
(44) [buki a mi paatna]
Buki is my partner/friend.
(45) [i ste gud . . . ai js stei kul]
Im good; Ill just stay like this [I wont draw any more cards].
5.2.3 Unrealized subjects
Speakers occasionally start utterances without explicit subjects when the
subject can easily be inferred from the context or previously in the discourse. Hancock (1987) and Parsons (193343) offer no corroboration of
this feature, though it also occurs in other anglophone Caribbean locations
where I have carried out fieldwork. It is also documented for Turks Island
English (Cutler 2003: 62).
(46) [iz taim f ren]
Its time for rain.


michael aceto

(47) [he boi, ink wi bta st ya maut tude]

Hey, boy, I think we better shut your mouth today.
(48) [iz mo straa]
Its stronger.
5.2.4 Questions
Interrogative forms show the common vernacular pattern in which there
is no inversion of subjects and verbs or do support in questions lacking
modal verbs. However, Parsons (193343) reveals at least two instances of
subjectauxiliary verb inversion for interrogatives: Is it dis? (380) and
what is dat? (385). Hancock (1987) reveals no instances of subjectverb
inversion for questions.
(49) [hau yu ple]
How did you play?
(50) [a taak wa mi du]
Are you talking about what I do?
(51) [wat ya min]
What do you mean?
(52) [we i go]
Where did he go?
(53) [wa da gem]
What is the game?
(54) [we yu drink]
What do you drink?
(55) [hu gt kaa]
Whos got a car?
(56) [si gon skul]
Is she going to school?
(57) [si aks ya fa mi]
Did she ask you for me?
5.2.5 Negation
Verb phrases are negated by a range of typical strategies found among
vernacular varieties spoken in the anglophone Americas. Preposed before
the main verb of a predicate are the following options in Statian English:

St. Eustatius English


en (< aint), don (< dont), or no. Some auxiliary verbs have a postposed
clitic-like nasal -n, e.g. shouldn shouldnt, didn didnt. The modal cant
is indicated by [kaan]. These strategies are confirmed by Hancock (1987)
for en (287, 300, 302, 303) and don (299). He reveals no instances of the
common creole language negator no for Statian English and neither does
Parsons (193343). Parsons reveals only instances of dont and not as negators
as well as nasal clitics attached to the end of auxiliary verbs.
(58) [ en gat tu rid]
I dont have to read.
(59) [ no sen]
Im not saying.
(60) [a ein ga no fkn man]
I dont have a fucking thing, man.
(61) [mebi si don km]
Maybe she wont come.
(62) [ ddn no]
I didnt know.
(63) [ no cit]
I didnt cheat.
(64) [a dont got gi ya aal]
I dont have to give you everything.
(65) [a don no di tknk]
I dont know the technique.
(66) [de kmn . . . yu en km]
Theyre coming; youre not coming?
(67) [a kaan go . . . pe yo mni]
I cant go. Pay your money.
(68) [ en gat pe]
I dont have to pay.
5.2.6 Plurality
Statians have several options for indicating plurality in their language.
The common anglophone eastern Caribbean plural marker [an dm] (see
Aceto 2002b) is heard in St. Eustatius but it is less common than simple


michael aceto

postnominal [dm]. In at least one instance below the data reveal a common dialect construction with prenominal [dm] that indicates not only
possession but plurality as well. In instances of this nature, a (redundant)
postnominal plural marker is rarely if ever heard. Hancock (1987) only
reveals postnominal dem. Parsons (193343) does not indicate any of the
aforementioned strategies; she only indicates the plural explicitly with the
metropolitan English pattern of suffixation by a bound inflectional morpheme. Again, its puzzling since the strategies documented below are heard
so robustly on the island.
(69) [di poskaad dm]
The postcards.
(70) [wai ya taakn manwl dm]
Why are you talking to Manuel and his buddies?
(71) [hi rid . . . a no i rid m . . . i bai dm bk]
He reads. I know he reads them. He buys books.
(72) [de gat tbako an dm fram kyuba]
They have/get tobacco and other related things from Cuba.
5.2.7 Possession
In Statian English possession is indicated by the common vernacular pattern of juxtaposing the possessor noun phrase before the possessed noun
phrase, e.g. [de kaa] their car (Hancock 1987: 291) and [m dadi haus]
my fathers house (284). This pattern is corroborated many times in Parsons (193343), e.g. wife bosom, bull milk (383). I encountered no instances
of possession indicated by bound inflectional morphology in Statian
5.2.8 Infinitives
All infinitive verbs were indicated by to only. There were no instances
in my data of an infinitive marked by any reflex of for as is common
in many Englishes in the western hemisphere. Hancock (1987) and Parsons (193343) corroborate the previous characterization. One informant
rejected infinitivals marked by fi/fu as not local forms.
(73) [trai tu kc mi]
Try to catch me.
(74) [a haf tu ple daimn]
You have to play diamonds.

St. Eustatius English


Table 8.2 Pronouns in Statian English

1st person
2nd person
3rd person

a, ai, (subject), mi (mostly object,

possession), mai (poss.)
(h)i he/she/it (subject/possessive)
(h)m (object)
si she (subject/possessive)
(h)(r) (object/possessive)
t it

1st person
2nd person
3rd person

wi, s (object)
de (subject/possessive), dm (object)

(75) [nobadi wan tu ple for mor gem, man]

Nobody wants to play four more games.
(76) [mebi a want tu nir m]
Maybe you wanted to get close to/narrow the gap with him.
(77) [hau pipl dm mndz tu lv]
How do people manage to live?
5.2.9 Pronouns
Table 8.2 presents the pronominal forms which are heard in St. Eustatius.
All forms should be considered to have multiple functions as subject,
object, and possessive pronouns unless otherwise indicated.
In most ways, the pronominal system of Statian English is quite similar to metropolitan English with some phonological differences (more
below). In my Statian English data there were few instances of [mi] as
a first-person subject pronoun; Hancock (1987) and Parsons (193343)
record none.16 However, the first-person possessive pronoun was usually [mai] or [m], though a few instances of [mi] occurred as well.
Parsons also presents several instances of [mi] as a possessive pronoun
(e.g. 381). Statian English always distinguishes gender differences between
the third-person singular pronouns. Statian English also lacks the object

One informant insisted that mi as a first-person singular subject pronoun was not a local form.
Though I recorded several instances, it could possibly have been imported from one of the nearby
anglophone islands such as St. Kitts where it is heard more regularly.


michael aceto

pronoun om that is so common in much (though not all) of the eastern

Many eastern Caribbean varieties, as reported in Hancock (1987: 298),
lack the second person plural form [unu] or any of its reflexes that are
so common in anglophone western Caribbean varieties.18 Instead, eastern
varieties mostly reveal the common regional form [aayu] or [alyu] or some
reflex of those forms. However, in Statian English, the only second-person
plural pronoun attested in my data and the printed sources is [yu]. Accordingly, Statian English also does not reveal the common eastern Caribbean
first-person plural pronoun [aa(l)wi]; instead it exhibits only [wi]. Both of
these features resemble those found in metropolitan varieties of English.
The pronominal system of Statian English is similar to other restructured
varieties of English spoken in Anguilla (Williams 2003) and the Turks and
Caicos islands (Cutler 2003). These forms may be related to forms spoken
earlier in the history of St. Eustatius during its period of commerce in the
late eighteenth century rather than assuming them to be recent changes
due to purported decreolization.

Discussion and conclusions

For such a relatively small population, linguistic diversity and cultural

contact was often intense on colonial Statia. Hartog (1976: 23) states that
various European powers passed Statia back and forth among themselves
twenty-two times. Contact with English speakers began as early as 1663.
From 1665 until the Treaty of Breda in 1667, Statia remained under English
control. However, Dutch speakers ruled the island again in 1672 until 1673
(Attema 1976: 18). After the Dutch military left, the English retook the
island; however, it was returned (at least officially) again to the Dutch
via the Treaty of Westminster in 1674. The Dutch took physical control
of the island in 1679. The French occupied the island in 1689 until a
joint Dutch/English force drove them out about one year later (Attema
1976: 21). What language or lingua franca this joint military force spoke is
unspecified by Attema and other researchers; however, the linguistic history
of the island suggests it was most likely some variety of English spoken by
both native and non-native speakers alike.
By 1689, Attema (1976: 16) states, besides Dutch, there were also
English, French, Germans, Scots, Irish and Koerlanders living on the

Other analogues such as [am] and [m] are also heard in the eastern Caribbean.
One informant rejected unu as a local form. For him, unu is a form associated with Jamaican Creole
English. He added that he couldnt understand Jamaicans at all.

St. Eustatius English


island. Hartog (1976: 29) suggests that Statia was always multilingual from
it earliest colonialization, and that, because it was amidst other islands
in the Caribbean being concurrently colonized by the British, English
soon became the common language of trade. This is an important point
because unlike other Dutch-controlled islands in which English emerged
as the local language (e.g. Saba and St. Maarten), English emerged on Statia
due to its pragmatic link with commerce. Hartog explicitly states that the
Dutch customarily adopted the language of the colonized people, whereby
Dutch remained as a sort of ruling language for the upper-ten (1976: 29).
It seems reasonable to assert that some English variety was spoken on St.
Eustatius as a lingua franca among persons of both African and European
descent well before the nineteenth century. Keur and Keur (1960: 43) report
that the Dutch language was gradually replaced by English, and by 1780
St. Eustatius had adopted and [sic] English pattern of life. The churches
asked for bilingual preachers from the homeland. Continued relations with
the United States of America after 1780 kept the English language alive on
the islands [i.e. both St. Maarten and St. Eustatius] to the present day.
The St. Eustatius Gazette, an English-language newspaper, was established
in 1793. Crane (1999: xv) reports that during the colonial period, British
citizens from nearby St. Kitts . . . commuted for years across the narrow
channel and maintained huge warehouses.
English was replacing Dutch even in official records early in the islands
history. Hartog (1976) notes that at the beginning of the eighteenth century, the registrar for births entered family relations with English words,
e.g. godfather instead of peetom. Place names in St. Eustatius dating
back to the eighteenth century are mostly in English: Gallows Bay, Sugar
Hole, Negro-Path, Jenkins Bay. Other place names appear in historical
records in both English and Dutch, e.g. White Wall/Witte Hoek, Turtle Bay/Schildpadsbaai or Schildpaddenbaai. Hartog (1976: 29) writes, in
1816 there was, with the exception of the few European Netherlanders, no
one on St. Eustatius who could express himself properly in Dutch.19
It seems that from the beginning of European settlement, contact with
anglophone locations in the Americas and later the USA resulted in the
widespread use of some variety of English. Keur and Keur (1960: 42) report
that in 1781 attacking British forces met fifty armed merchantmen from
the USA. Jameson (1975: 90) notes that a British official in Amsterdam
made the following comments about the island in the latter half of the

Hartog (1976) reports that when a Dutch bishop visited the island in 1836, he could speak no English
and, since the Governor of the island could speak little Dutch, communication was difficult.


michael aceto

eighteenth century, St. Eustatius is the rendezvous of everything and

everybody meant to be clandestinely conveyed to America. It is easy to
get oneself carried thither, and military adventurers of all nations have
congregated at the island. This intense multilingual setting seems to have
contributed to the emergence of some English variety as the lingua franca
of the island (see Baker [2000: 48] for what he terms a medium for
interethnic communication), specifically when the recipients of many
of Statias goods were English-speaking sailors and merchants headed for
colonies in North America and later the USA.
Religious affiliations in the eighteenth century provide further indication
of the multilingual nature of Statian society. Hartog (1976: 115) lists four
general religious practices: Lutheran, English Presbyterian, Anglican, and
Judaism. At least two of these Christian groups explicitly received religious
instruction in some English variety. Tombstones in the Jewish cemetery
on Statia, as confirmed by this researcher, reveal inscriptions written in
both Portuguese and Hebrew, suggesting that these languages were most
likely used in Jewish homes and religious observances. Many of these Jews
may have also spoken Ladino, a language common among Sephardic Jews
whose ancestors were expelled from the Iberian Peninsula beginning in the
late fifteenth century (see Aceto 1997). There are no Jews on Statia today,
though archeological remains in the form of a cemetery, and the ruins of a
ritual bath and synagogue testify to their earlier previous presence.20
The multilingual and multiethnic diversity of St. Eustatius in the eighteenth century is indicated by Andrews and Andrews (1934: 136), who
present the diary of a Scottish visitor, Janet Schaw. In 1775 she wrote, But
never did I meet with such variety; here was a merchant vending his goods
in Dutch, another in French, a third in Spanish, etc. etc. They all wear
the habit of their country, and the diversity is really amusing. The first
that welcomed us ashore were a set of Jews. Hartog (1976: 40) states that
Turks, Greeks, and Levantines (i.e. presumably people from the modern
Middle Eastern area of Lebanon) were also present among the merchant


Most of the Jews on Statia were Sephardim whose ancestors derived from Spain and Portugal as
result of the expulsions in the late fifteenth century, which Schaws diary confirms (Andrews and
Andrews 1934: 1367). In 1722, the Jewish population stood at 21 persons in total: six adult males
with their wives and families. In 1781, at the end of the great commercial period, there were 350
Jews in total, including 101 adult males, their wives and families. Rodney deported approximately
one-third of the adult male Jewish population of Statia when he finished plundering the island. By
1790, there were only 157 Jews remaining on St. Eustatius. In 1818, only five Jews remained, and in
1846 the last remaining member of this once thriving community died.
However, ethnic, linguistic, and commercial diversity were not always welcome on the island. In
1739, Commander Faesch complained to the Amsterdam Chamber that both French and English

St. Eustatius English


There is no empirical or historical evidence that a restructured variety of

Dutch ever emerged in the Dutch Windward Antilles (as occurred nearby
on the island of St. Croix) but instead a dialect of English. Nor is there any
empirical evidence that a deeper Jamaican-like or Antiguan-like Englishderived variety was ever spoken on St. Eustatius. Why did an English dialect
emerge during the colonial period in the Dutch-controlled colony of St.
Eustatius (as well as on Saba, St. Martin)? In the Dutch Windward islands
it seems that European competition over the Caribbean necessitated that
they welcome colonists, settlers, and merchants from any location as long
as they aided in populating and settling the island. In the case of Statia,
the Dutch welcomed colonists from many ethnicities and nations who
spoke several different languages. English was one of several languages that
were spoken on the island as well as Dutch, French, Portuguese, Hebrew,
and presumably several African languages spoken among the first slaves.
Despite the fact that St. Eustatius was (and still is) a Dutch colony, the
largest speech community on the island was comprised of English speakers
(of both European and African descent). When considering the island
during its period of great commerce, English seems to have emerged as
an interethnic lingua franca in which commerce was facilitated by the
use of this language in a multilingual context. In other words, Statian
English might best be considered as a one of several commercial Englishes
of the eighteenth century. This English variety arose in an environment
where plantation slavery was not the norm, and instead an importexport
economy was the focus.
The few creole-like features heard in Statian English are perhaps
accurately and simply accounted for by assuming they are due to postemancipation immigration from islands whose English-derived varieties
exhibit these same features. Small boat traffic among the islands of the
Lesser Antilles continues today; it is informal, unofficial, and more intense
(and thus difficult to gauge since there are often no records) than generally acknowledged in Caribbean studies. Statians have historically and
contemporaneously made contact with St. Kitts and Nevis (and vice versa)
whenever they had access to maritime routes of transportation. This contact with St. Kitts is most likely to be responsible for the three or four
creole-like features examined in this chapter.
are bringing European goods to the market here, and foreigners even set up shops and trade here
(Attema 1976: 38). Goslinga (1979: 82) claims that the economic success of Statia and its dependence
on smuggling in the second half of the eighteenth century "brought many unsavory characters to
the island." He leaves unexplained what this assertion specifically means in terms of ethnicity,
culture, and language.


michael aceto

What seems unusual about the case of St. Eustatius is that for many nonEnglish-speaking Europeans as well as African slaves, an English dialect
emerged and was identified as a target language (I use the word target
here only to indicate that some variety of English was identified by residents
of the island as the most prestigious [at least in public life], economically
powerful, socially expedient, or common language spoken on the island). It
wasnt only slaves that were grappling with English varieties heard around
them but many Europeans on the island who natively spoke languages
other than English as well. That is, some of the Europeans spoke some
variety of English as a native language, but they do not seem to have
been the largest part of the population at least in the early colonial history
of the island since speakers of European languages on the islands came
from many different speech communities. However, two hundred years
later, by the time of emancipation, most of the islands population was
largely of African descent and mostly monolingual in the local variety of
English. I have no colonial artifacts that demonstrate conclusively that the
earliest Europeans and Africans spoke more or less the same variety (regular
language variation and change not withstanding), but the high degree of
second-language speakers of European descent suggests that the dynamics
of language creation, emergence, and shift was different on St. Eustatius
than on the typical Caribbean island in which the largest segment of the
European population was derived from the homeland of the colonial power
in question.
There are at least four features of Statian English generally associated with
so-called creole languages: postnominal plural marker dem, verbs unmarked
for past contexts, preverbal negator no, and preverbal future-tense marker
go. Of course, none of these features is heard exclusively among so-called
creole languages, and no individual feature is diagnostic of this group either.
I am excluding features such as the preverbal completive marker done since
it is found in a range of varieties that are uncontroversially considered
dialects of English. See Mufwene (2000, 2001, 2008), who argues that all
European-language-derived creoles may be considered as dialects of the
lexifier since creolization is a social, not a structural process.
One of the more popular ways for linguists to imagine creole varieties
without many so-called creole features is simply to insist (even in the
face of no artifacts) that varieties like Statian English must have once
looked more like, say, Jamaican or Antiguan, and that the variety in
question has subsequently decreolized since emancipation. This seems
too easy a way out of the challenge of trying to understand the different
social and linguistic circumstances (or language ecologies a` la Mufwene

St. Eustatius English


[2001, 2008]) under which Englishes emerged in specific locations in the

Americas. Language shift undoubtedly occurs in what might be labeled
creole-speaking as well as non-creole-speaking communities. Purported
decreolization in its most coherent version simply describes the role and
effect of literacy, familiarity with institutional varieties of English (local or
otherwise), and the effect that varying degrees of fluency in that standard
dialect have had on specific idiolects and community-based dialects in
Aceto (2010) describes Kokoy, an English creole or dialect spoken on
the island of Dominica. Older residents who mostly spoke Kokoy in their
youth insisted that Kokoy is still spoken today, even if the number of
speakers is declining. Many of these same key consultants reported that a
process of shift occurred fairly rapidly within one generation in the 1950s
70s in which young people started identifying the institutional English
variety used mostly in schools as the most prestigious target (see Garrett
2003 for the role of institutional varieties on community-based dialects
in the Caribbean). Kokoy speakers exhibit almost no mixing of Kokoy
and intermediate English features. When motivated, one switches between
these varieties as if one were switching from one grammatically unrelated
variety to another such as from the local French Creole to English. For
some speakers, Kokoy is simply not part of their repertoires; for some it is,
and these folks can switch from Kokoy to another language variety in the
same manner as traditional codeswitching. The concept of decreolization
has been undergoing a process of re-examination and deconstruction in
the last ten years or more (see Aceto 1999; Satyanath 2006) and certainly
data from Kokoy do not validate this abstraction as having any linguistic
reality in Dominica.
In any event, the term decreolization is largely meaningless. All living
languages change. Just because a dialect or so-called creole spoken by
people largely of African descent changes doesnt mean the language is
decreolizing. A speaker cannot undo the process of creolization (even
if that term could be coherently distinguished from regular cultural and
linguistic emergence, variation, and change and even that doesnt seem
possible); a speakers language will change since all grammars, all idiolects,
are fluid and in the flux of becoming. Why should change in the Caribbean
be labeled uniquely as decreolization? Just because its purportedly done
by people of African descent is not a sufficient motivation for such a term.
The term is problematic and should be carefully considered before use, if
used at all. Many of my students from eastern North Carolina can shift
between local rural community-based dialects and institutional varieties of


michael aceto

English, and some even make an effort to replace their vernacular with a
dialect variety containing more standard-like features.
Le Page (1998: 91) makes several important points: I came to realize
too late, unfortunately, to stop David DeCamp and others from taking
up the continuum model that it was a false representation. Le Page
also writes that DeCamp wrote him that he regretted the concept had
been taken up with such enthusiasm, since he [DeCamp] had never found
that it could provide an account of more than 30% of his Jamaican data
(1998: 92). That is, 70 percent of the data still could not be accounted for
with the continuum concept. If subsequent researchers have cherry-picked
their data and then sorted it along a continuum so it looks tidy and neat,
then perhaps that says more about the practitioners of creole studies in
terms of methodology than it does about the purported uniqueness of
the speech communities in question. On the other hand, if one wants
to read how scientists have regularly tampered with data, sometimes even
with successful results, then read John Wallers Fabulous Science (2006).
It seems it is a more common (albeit secret) practice than we were led to
believe by our mentors. Furthermore, creolists used to insist (in the 1970s,
1980s) that the alleged variation of the continuum was due to decreolization, which was perceived as a unilateral force affecting all Englishes of
the Caribbean. When this strong view of decreolization was questioned
and criticized with data that showed change was not unidirectional, then
creolists of the 1990s abandoned decreolization but still kept the continuum (even if decreolization was the explanatory factor) no matter how
flawed a concept it was. Now it is simply asserted in creole studies that
a continuum exists uniquely in the anglophone Caribbean (and why not
the francophone or hispanophone Caribbean as well?). Either all natural
human languages exhibit a continuum of variation or none do. Clearly all
living languages exhibit variation, but can this variation be arranged along
a continuum without forcing the data to fit the continuum model? This
insistence by creolists that something vaguely different or unique happened
in the anglophone Caribbean that was qualitatively different from the rest
of the language-speaking world (both past and present) is troubling.
What would prevent me from arranging data from English speakers
in eastern North Carolina along a continuum from vernacular to the
standard? Many long-term locals, even so-called white folks, regularly
reveal constructions like she mean with no explicit copula verb, or say
<with> with a word-final [f], [daet] with the word-initial stop (see above),
and verbal forms like my brothers out a-walkin. They also say carry for
take, cuss for curse, reach for arrive, wait on for wait. Sure, those

St. Eustatius English


same words are heard in the Caribbean but they are heard in other Englishspeaking regions beyond as well. I also have many students who are the first
ones in their families to go to college who manifest the aforementioned
forms as well as also more bookish forms associated with institutionalized
literacy or the standard. I could arrange the data on a continuum easily
from (left to right) vernacular forms to standard forms. Couldnt this been
done in any speech community in which a standard form associated with
education and literacy has been institutionalized?
It is also worth remembering that in the early twentieth century, Danish
physicist Niels Bohr defined the nature of light as neither particle nor wave
but both (just not at the same time). That is, Bohrs complementarity
principle states that photons of light (and electrons) could behave either
as waves or as particles, but it is impossible to observe both the wave
and particle aspects simultaneously. This example from another scientific
discipline wrestling with the dual nature of natural phenomena suggests
that the conclusions we draw from our data depend crucially on our
theoretical orientation and the questions we are trying to answer.
My research in the Caribbean reveals a situation that is not all that
different from any other location in which the local vernacular norms are
sufficiently different from the institutional norms of the lexically related
standard. However, decreolization, even if incoherently used by most
creolists, is only one type of externally motivated change and does not represent change in and of itself in creole-speaking communities (see Mufwene
[2000: 77] as well). However, it is my feeling that an over-reliance on this
purported phenomenon as an explanatory force has obscured the varied
details of language emergence or language ecology in specific anglophone
locations in the Americas.
Aceto (1999) pointed out that a reliance on the concepts associated with
the purported creole continuum and the concept of decreolization has
a tendency to view all changes in, at least, anglophone creole-speaking
Caribbean societies only in terms of whether the feature in question was
similar to or different from more standard varieties of English in terms of
form and function. The creole continuum terms acrolect, mesolect, and
basilect reveal little about linguistic or sociolinguistic processes invoked
by speakers involved in the creation, distribution, and maintenance of their
language, except for the assumption (which is undoubtedly true in some
specific well-defined cases) that some speakers are consciously or unconsciously shifting their language toward norms associated with metropolitan
or institutional English and that these same speakers have competency in
a range of lects, as undoubtedly do many speakers around the globe who


michael aceto

have some familiarity with literacy and institutional varieties of dialects

that are historically related to their community-based vernaculars. These
terms acrolect, mesolect, and basilect only measure whether a given
feature (or bundle of features) appears more like its metropolitan or standard version of the lexifier language or not. From this perspective, the
common assumption in the use of these terms is that so-called mesolectal varieties have simply decreolized under pressure from more standard
varieties of English and that basilectal creoles have not undergone this
same unidirectional change toward features associated with the lexifier. It
must be remembered that the variety of Jamaican found in Bailey (1966)
represented an abstract bundle of features (i.e. the bundle is abstract,
not any individual feature) associated with the so-called basilect, and
DeCamp (1971) considered this type of abstraction to be a necessary first
step in understanding variation in creole-speaking communities (the idea
of the so-called [post-creole] continuum originated with DeCamp 1971).
This abstraction suggests very strongly that, both from a diachronic and
synchronic perspective, there have always been few if any speakers of the
purported basilect revealing all of its associated features; in other words,
few individual so-called creole speakers can be said to have ever exclusively
displayed all the features of the basilect. The reification basilect is simply a compilation of all the features that are considered typologically the
furthest or most different from varieties of metropolitan English. Furthermore, any language, including so-called creole languages, can change in
ways left unexplored by the assumptions of the creole continuum, even
if the effects of internally induced change have been left largely unexplored by researchers studying creole-speaking societies (see Aceto 1999).
In addition, some types of change, whether they are externally or internally
motivated, may occur which do not resemble metropolitan varieties of
English. I am in complete agreement with Satyanath (2006), who writes,
decreolization . . . is a working hypothesis and not an empirically tested
claim (186).
The assumption in this chapter is that, not withstanding regular
diachronic change that all human languages exhibit everywhere, Statian
English sounds quite similar today to the colonial variety spoken 200300
years ago. In other words, it is not necessary to reify decreolization and
the purported creole continuum to try to explain the synchronic shape of
Statian English. For a detailed discussion of the limitations of the creole
continuum abstraction, consult Aceto (1999), (2002a), and (2003). I am
in agreement with Mufwene (2008), who makes the case that so-called
creole languages are just the latest (i.e. in the last 500 years) stage of

St. Eustatius English


the dispersal of a subgroup of the Indo-European family of languages

that began 10,00012,000 years ago. In other words, the English-derived,
French-derived, etc., varieties spoken in the Americas, Africa, and Asia
are dialects or daughters of their lexifiers in the same way French and
Spanish, for example, are daughters of Vulgar Latin spoken by the Romans
in the administration of their empire. However, every language has what
Mufwene calls its own specific language ecology (i.e. the history of who
came in contact with whom, the linguistic variants available, whatever social
factors are attached to those variants as well as the specific ethnographic
circumstances [2008: 54]). Therefore [l]anguages evolve in non-uniform
ways (Mufwene 2008: 14), without needing decreolization to account
for them. This chapter and Aceto (2006) describe the emergence of Statian

Aceto, Michael. 1997. Saramaccan Creole origins: Portuguese-derived lexical correspondences and the relexification hypothesis. In A. K. Spears and D. Winford, ed., The Structure and Status of Pidgins and Creoles. Amsterdam: John
Benjamins, 21939.
1999. Looking beyond decreolization as an explanatory model of language
change in Creole-speaking communities. Journal of Pidgin and Creole Languages 14: 93119.
2002a. Going back to the beginning: describing the (nearly) undocumented
anglophone creoles of the Caribbean. In G. G. Gilbert, ed., Pidgin and
Creole Linguistics in the 21st Century. New York: Peter Lang, 93120.
2002b. Barbudan Creole English: its history and some grammatical features.
English World-Wide 23: 22350.
2003. What are creole languages? An alternative approach to the anglophone
Atlantic world with special emphasis on Barbudan Creole English. In
M. Aceto and J. P. Williams, eds.,Contact Englishes of the Eastern Caribbean.
Amsterdam and Philadelphia: John Benjamins, 12140.
2006. Statian Creole English: an English-derived language emerges in the Dutch
Antilles. World Englishes 25: 41135.
2010. Dominican Kokoy. In Daniel Schreier, Peter Trudgill, Edgar W. Schneider, and Jeffrey P. Williams, eds., The Lesser-Known Varieties of English: An
Introduction. Cambridge University Press.
Aceto, Michael and Jeffrey P. Williams, eds. 2003. Contact Englishes of the Eastern
Caribbean. Amsterdam: John Benjamins.
Andrews, Evangeline Walker and Charles McLean Andrews. 1934. Journal of a
Lady of Quality; Being the Narrative of a Journey from Scotland to the West
Indies, North Carolina, and Portugal, in the Years 1774 to 1776. New Haven,
CT: Yale University Press.


michael aceto

Attema, Y. 1976. St. Eustatius: A Short History of the Island and Its Monuments.
Zutphen: De Walburg Pers.
Bailey, Beryl L. 1966. Jamaican Creole Syntax: A Transformational Approach.
Cambridge University Press.
Baker, Philip. 2000. Theories of creolization and the degree and nature of restructuring. In Ingrid Neumann-Holzschuh and Edgar W. Schneider, eds., Degrees
of Restructuring in Creole Languages. Amsterdam: John Benjamins, 4163.
Baker, Philip and Adrienne Bruyn, eds. 1998. St. Kitts and the Atlantic Creoles.
London: University of Westminster Press.
Bhatt, Parth and Ingo Plag, eds. 2006. The Structure of Creole Words: Segmental,
Syllabic and Morphological Aspects. Tubingen: Max Niemeyer.
Childs, Becky, Jeffrey Reaser, and Walt Wolfram. 2003. Defining ethnic varieties
in the Bahamas: phonological accommodation in black and white enclave
communities. In Aceto and Williams, eds., 128.
Crane, Julia G., ed. 1999. Statia Silhouettes. New York: Vantage Press.
Cutler, Cecilia. 2003. English in the Turks and Caicos Islands: a look at Grand
Turk. In Aceto and Williams, eds., 5180.
DeCamp, David. 1971. Toward a generative analysis of the post-creole continuum.
In Dell Hymes, ed., Pidginization and Creolization of Languages. Cambridge
University Press, 34970.
DeGraff, Michel, ed. 1999. Language Creation and Language Change: Creolization,
Diachrony, and Development. Cambridge, MA: MIT Press.
Garrett, Paul. 2003. An English Creole that isnt: on the sociohistorical origins
and linguistic classification of the vernacular English of St. Lucia. In Aceto
and Williams, eds., 155210.
Goslinga, Cornelis Ch. 1979. A Short History of the Netherlands Antilles and Surinam. The Hague: Martinus Nijhoff.
Hancock, Ian. 1987. A preliminary classification of the anglophone Atlantic creoles,
with syntactic data from thirty-three representative dialects. In G. G. Gilbert,
ed., Pidgin and Creole Languages. Honolulu: University of Hawaii, 264
Hartog, J. 1976. History of St. Eustatius. Aruba: De Wit.
Holm, John. 19889. Pidgins and Creoles, 2 volumes. Cambridge University Press.
Jameson, J. Franklin. 1975. St. Eustatius in the American Revolution. In Charles
Toth, ed., The American Revolution and the West Indies. Port Washington,
NY: Kennikat.
Keur, John Y. and Dorothy L. Keur. 1960. Windward Children: A Study in the
Human Ecology of the Three Dutch Windward Islands in the Caribbean. Assen:
Royal Vangorcum.
Le Page, R. B. 1998. Ivory Towers: The Memoirs of a Pidgin Fancier. A Personal
Memoir of Fifty Years in Universities around the World. Society for Caribbean
Le Page, R. B and David DeCamp. 1960. Jamaican Creole. London: Macmillan.
McWhorter, J. 1998. Identifying the creole prototype: vindicating a typological
class. Language 74: 788818.

St. Eustatius English


2000. Defining creole as a synchronic term. In Ingrid Neumann-Holzschuh

and Edgar W. Schneider, eds., Degrees of Restructuring in Creole Languages.
Amsterdam: John Benjamins, 85123.
Mufwene, Salikoko, S. 1996. The founder principle in creole genesis. Diachronica
13: 83134.
2000. Creolization is a social, not a structural, process. In Ingrid NeumannHolzschuh and Edgar W. Schneider, eds., Degrees of Restructuring in Creole
Languages. Amsterdam: John Benjamins, 6584.
2001. The Ecology of Language Evolution. Cambridge University Press.
2008. Language Evolution: Contact, Competition and Change. London and New
York: Continuum.
Parkvall, Mikael. 2001. Creolistics and the quest for creoleness: a reply to Claire
Lefebvre. Journal of Pidgin and Creole Languages 16: 14751.
Parsons, Elsie Clews. 193343. Folklore of the Antilles, French and English, 3 vols.
New York: American Folklore Society.
Plag, Ingo. 2001. The nature of derivational morphology in creoles and non-creoles.
Journal of Pidgin and Creole Languages 16: 15360.
Postma, Johannes Menne. 1990. The Dutch in the Atlantic Slave Trade 16001815.
Cambridge University Press.
Rizzi, Luigi. 1999. Broadening the empirical basis of universal grammar models:
acommentary. In Michel DeGraff, ed., Language Creation and Language
Change. Creolization, Diachrony, and Development. Cambridge, MA: MIT
Press, 45372.
Satyanath, Shobha. 2006. English in the New World: continuity and change, the
case of personal pronouns in Guyanese English. In Bhatt and Plag, eds.,
Waller, J. 2006. Fabulous Science: Fact and Fiction in the History of Scientific
Discovery. Oxford University Press.
Williams, Jeffrey. 1983. Dutch and English Creole on the Windward Netherlands
Antilles: an historical perspective. Amsterdam Creole Studies 5: 93103.
2003. The establishment and perpetuation of anglophone white enclave communities in the eastern Caribbean: the case of Island Harbor, Anguilla. In
Aceto and Williams, eds., 95119.

c h a p ter 9

The English of Gustavia, St. Barthelemy

Ken Decker


There is a variety of English spoken in the port city of Gustavia on the

French island of St. Barthelemy in the Leeward Islands in the northeastern
corner of the Caribbean. It has the smallest population of speakers of
English in the Caribbean and is probably the most endangered variety.1
The speakers of the language do not have a name for their speech other
than English. I call this variety Gustavia English (GE) due to its association
with the town of Gustavia rather than with the entire island. Most of the
speakers of GE are Afro-European while most of the rest of the population
of the island is white and French-speaking.
The name of the island is also spelled St. Bartholomew or St. Barths/
Bart(s). In 2007 the islands governance separated from Guadeloupe and
became the Overseas Collectivity of Saint Barthelemy. The 21 sq km
(8 sq. miles) island is within sight of St. Martin 19.5 km (12 miles) to
the northeast, and the islands of Saba, St. Eustatius (also called Statia),
and St. Kitts are about 48 km (30 miles) to the south and southwest. The
port town of Gustavia is the largest settlement on the island. The islands
population of 2,332 (CIA 2012) is primarily supported by tourism.
Most of the previous linguistic and sociolinguistic research on St. Barths
has focused on the varieties of French spoken on the rest of the island.
My previous article (Decker 2004a) is the most comprehensive description
of GE to date. That article focused primarily on identifying the historical
roots of GE, partly through an examination of some linguistic features. I
proposed that English came to St. Barths around 1785.
This chapter will build on the previous article and give further description of the phonology and morphology. I believe that GE represents an early
stage of contact between Africans and Europeans. The contact occurred in

The English of Webster Yard, Anguilla, described by Williams (2003), may be considered more
endangered, but the people are simply shifting to another close variety of English.


The English of Gustavia, St. Barthelemy


a context in which slaves and owners lived and worked closely; a context
which offered sufficient opportunity for the slaves to acquire English. There
is no evidence of a reduction and expansion process. Most nonstandard
features can be traced to regional varieties of English.

2 Sociolinguistic history and current status

Thorough sociolinguistic histories of St. Barths have been written by others
(Maher 1996), so I will only give a few significant dates and events here.
The first settlement on the island was in 1648 by a group of Frenchmen
from St. Kitts. Due to its small size and rugged interior the island was
never considered valuable for farming or other development. During the
first hundred years the settlers were attacked by the British numerous times,
as well as enduring a few attacks from the Spanish and Caribs. As a result
of these attacks there were several evacuations to St. Kitts.
In 1784 France sold St. Barths to Sweden. The Swedes made two very
significant policy decisions. First, they declared St. Barths to be a free port
open to trade with any nation. Shipping traffic rapidly expanded and by
1800 hundreds of ships were arriving each month. Secondly, protection was
offered to any immigrant who wanted freedom from debt. As a result of
these two rulings Gustavia grew from four landowners in 1785 to 5,000
by 1800 (Maher 1996: 382, 391).
The earliest report (Kohler and Runsten 2001: 9) of English on St. Barths
is from 1787 when English was said to be used as an official language, along
with French and Swedish. There never was a large population of Swedes
and the local government used English for all official communication. The
French speakers lived primarily in the rural areas outside Gustavia.
The pattern of the language development for GE was probably quite
similar to that of the neighboring Dutch islands. Aceto (2007) describes
the connections through shipping that St. Eustatius (Statia) had with the
American English colonies in the late eighteenth century. Even though
the island was administered most of the time by the Dutch, English was
the dominant language. When Saba became part of the Dutch Windward
Islands in 1816, a government document stated that all government documents and regulations shall be translated into the English language as there
is no one on this island [Saba] who understands Dutch (Aceto 2007). Jeffry
(1997) describes the early settlement of North Americans and Anguillans
on St. Maarten and the establishment of English as the language of both
the Dutch and French sides of the island. It appears that populations were
somewhat fluid and the fact that a Dutch, French, or Swedish government


ken decker

administered the island did not mean that English could not be the most
widely used language.
In 1781, the British destroyed much of the port of Statia and it never
recovered. It appears that some of the trade, particularly with the United
States, must have moved on to Gustavia. Furthermore, during the British
and French wars from 1793 to 1815 neutral St. Barths provided a place for
essential trade between the combatants.
It is important to understand the composition of the early population of
Gustavia as the foundation of GE. Shrimpton and Baker (1995: 90) relate
that the governor of Antigua complained that Englishmen, avoiding the
payment of taxes, were moving, with their slaves, to St. Barths. Shrimpton
and Baker (ibid.) also cite a report from the 1790s as saying, quite a
number of businessmen from the surrounding mainly English-speaking
islands, bringing with them their slaves, were moving to Gustavia. Maher
(1996: 382) says that, almost all new arrivals had English surnames. So,
there were many English speakers coming from neighboring islands, as well
as from England and America, and Africans.
The proportion of whites to slaves was usually fairly even on these islands.
The largest population of Afro-Caribbeans to be recorded on St. Barths
was 2,406 in 1812 (Maher 1996: 390). There are no other population figures
for that year, but total population of the island was no more than 6,000. A
third of those slaves may have been French Creole speakers working on
sugar plantations at the eastern end of the island.
Due to the proportional relationship of races and type of work, there
was more social contact than on plantations (Fleischmann 2005: 56). As
on Statia, the majority of slaves brought to Gustavia during this time were
used for unloading and loading ships. Maher (2010) notes that, Gradually,
Gustavias free coloreds came to outnumber the slaves. So it would seem
that the slaves had greater access to freedom also.
After a few decades of prosperity, trade began to decline, as did the
population. The emancipation of slaves came in 1846. As elsewhere in the
Caribbean, this caused further decline in the economy. In 1852 half of
Gustavia was destroyed in a fire, after which large numbers of the business
owners and laborers left the island. In 1877 the control of St. Barths returned
to France. After the French regained control of St. Barths the island sank
further into isolation and poverty.
For the next 100 years the population of St. Barths remained fairly small,
possibly no more than a few hundred people. The population was divided
between white speakers of an archaic French at the western end of the
island, white speakers of a French Creole at the eastern end of the island,

The English of Gustavia, St. Barthelemy


and Afro-Caribbeans and a few whites in and around Gustavia who spoke
English. To survive during this period many men worked on sailing vessels
in the region or worked for short periods of time on neighboring islands.
In 1958 David Rockefeller visited the island and saw the potential for
the development of tourism. He built several large hotels and the economy
slowly began to recover. However, it wasnt until the 1970s that the recovery
became evident. During the 1970s as new business was coming to the
island, and Gustavia was rebuilt, the language of Gustavia began to shift
from English to French. Today, there are only a couple hundred speakers
of GE, at most, remaining in Gustavia.
In my previous article (Decker 2004a) I proposed that the koineization
of the current variety of English on St. Barths probably began around 1785
to 1815. It is important to remember that the so-called English speakers
came predominantly from areas where nonstandard varieties of English
were spoken, either in the British Isles, the Caribbean, or North America.
While these nonstandard English2 (SE) speakers, with their slaves, may
have come from many neighboring islands, the social history indicates that
the environment of language development in Gustavia was quite different
from that of the plantations on islands such as St. Kitts or Antigua. If
many of the immigrants were fleeing debt, and had only a few slaves, it
can be assumed that they were not large land holders. Fleischmann (2005)
described the close social relationships between slaves and masters during
the mid seventeenth century. It seems that these immigrants to St. Barths,
at the end of the eighteenth century, may have still been living relatively
closely socially. The speech of the slaves would have been fairly similar to
that of their masters. As a result, the English of Gustavia is less divergent
from SE than the creole on neighboring islands. This supports Mufwenes
(2001) thesis that different social conditions have more relevance than a
unique creole developmental process.

3 Descriptive features
Gustavia English is worth studying and describing in as much as it is unique
and has features that distinguish it from other varieties of related speech.
In this section I will show that GE is a nonstandard, restructured variety of
English. There is no evidence of it having been a pidgin, nor having gone
through a creolization or decreolization process. There is possibly some

I use Cruttendens (2001) description of English as a model of SE. He includes comments on General
American (GA) and Received Pronunciation (RP).


ken decker
Table 9.1 Phonetic realizations of the lax
vowels of Gustavia English
Lexical set (Wells 1982)

Phonetic variation in GE





evidence of contact influence from St. Kitts Creole (SKC), but it is only
weak evidence at best. I have been able to identify possible sources in the
British Isles and North America for some of the unique features described
In this section I will describe aspects of GE segmental phonology. This
description is based on two half-hour recorded conversations, each between
a male and a female resident of Gustavia, all over 55 years of age. I have
also used notes from five other unrecorded interviews I conducted, two
texts from Maher (1987, 1996), and nine short texts from Parsons (1933).
I have not included data from the Shrimpton (1994) texts from 1804.
My comparison of these texts with other evidence of GE (Decker 2004a)
indicates that they do not represent the speech of Gustavia.

Segmental phonology

The vowels and consonants used in GE are quite similar to those in

the English of Ireland, Scotland, northern England, and possibly North
America. The vowels exhibit several significant differences from or
Antiguan Creole.
3.1.1 Lax vowels
GE lax vowels tend to be found in closed syllables. When there is either no
onset or coda there is more possibility for variation in the pronunciation
of the lax vowel.
Vowel length is a strong indicator of creole speech that is different from
SE. SKC lax vowels tend to be shorter in duration than the tense vowels.
GE vowels exhibit very little difference in the length of lax and tense vowels.
Table 9.1 provides a summary of the phonetic variation described in the
following paragraphs.

The English of Gustavia, St. Barthelemy


kit //
In most Caribbean English Creoles (CEC) the kit vowel only occurs in
closed syllables, but in GE it occurs in any syllable shape. In closed syllables
the kit vowel seems stable, having little variation. For example, there is
no change in the vowel in all samples of /wd/ with from all four language
helpers. However, in other environments there is considerable variation.
For example, in an open syllable such as the beginning of because, behind,
believe, and before the kit vowel can be [   ]. The first syllables in
each of these examples are unstressed. Bolinger (1989) says that features
such as vowel height and roundness become less specified in unstressed
vowels. This may explain some of the variation. The kit vowel in a syllable
rhyme, preceding the velar nasal //, as in bring or thing, has similar
variability: [   ]. Cruttenden (2001: 107ff ) says this vowel, in some
SE words, has a similar range of variation [     ] in British
The pronunciation may also be influenced by suprasegmental
features like sentence-level stress. For example, between my four language assistants I heard the following variants for the definite article:
[d  d  d  d  d]; and for the locative preposition in, there are
examples of [ n  n  n]. Sentence-level stress may or may not fall on
these high-function words depending on the emphasis of the sentence,
and this may explain some of the variation.
dress //
The dress vowel // is fairly stable in closed syllables. One informant
tended to produce a slightly higher vowel [ ]. In open syllables there
is more variation with a number of high-function words being pronounced close to the face vowel /e/. For example, they was transcribed
as [dn  d  d  d e  d e]. This represents variation from a more
creole-influenced nasal dress vowel [] to an acrolectal face vowel [ e].
Cruttenden (2001: 130) says that Middle English [a] split in Early
Modern English to [e] and [], which then became [e] in Present RP
English. The [] has been preserved in many regional dialects in England
in words like make, take, and catch. In GE these words have the dress
vowel //, as is typical in CECs.
trap //
The trap vowel // is found in both closed and open syllables. In my
data there are a few occurrences of a more central variant [] in open
syllables. The presence of the trap vowel in GE is noteworthy since it is


ken decker

not found in basilectal SKC, but it is common in SE. If GE had developed

out of SKC, or Antiguan, this trap vowel would not be present. This is
one bit of evidence that there has not been restructuring in GE.
palm //
In GE the palm vowel // is fairly stable. In open syllables it sometimes
varies as far forward as []. There are examples of the strut vowel where
we usually hear the palm vowel, such as /wz/ pronounced as [wz] was.
One of the features of CEC phonology is a distinction between a lengthened palm vowel // and an unlengthened vowel //. This distinction is
not very prevalent in GE but there were a few examples. These may be the
result of contact with neighboring CEC varieties, but it doesnt seem to be
sufficient evidence of an earlier creolization of GE.
strut //
Unlike neighboring CEC varieties the strut vowel // occurs in GE,
but usually as a diaphoneme of other vowels. For example: cut was usually
pronounced by my informants, as [kot  kt], but there are a few examples
of [kt]. Similarly, come is usually heard as [kom], but can also be heard
as [kom
 km]. It may be possible that this variation is a result of an
incomplete footstrut split process that began during the transition
from Middle English to Early Modern English.
The strut vowel can also be found in words usually pronounced with
the palm vowel, such as /wz/. Even though around and again are consistently pronounced with the palm vowel as the first syllable, about is usually
heard with the strut vowel in the first syllable.
There are also a few words in which I believe the strut vowel is the
target pronunciation, not a diaphoneme. For example, us is not a pronoun
used in CEC but it is used in GE. There are only a few examples of it in any
of the data, but it is always pronounced with the strut vowel //. Since
this is not a vowel found in CECs it has not been influenced by Caribbean
foot //
The foot vowel // occurs very infrequently in the data. It is a very stable
vowel occurring in closed syllables, except when the onset or coda of a
one-syllable word is dropped, as in /d/ would or /k/ could.
One language helper consistently produced [t] for the preposition to,
while the other language helpers produced [tu]. Other examples of this
variation include: [ful  fl] full, [ud  d] good, and [lukn  lkn]
looking. Cruttenden (2001: 122) says that the pronunciation of some words

The English of Gustavia, St. Barthelemy


Table 9.2 Phonetic realizations of the tense

vowels and diphthongs of Gustavia English
Lexical set (Wells 1982)

Phonetic variation in GE


e   e 
o  o      o
a  o  ou



in seventeenth-century northern English, including good and look, had a

variable pronunciation between [u] and [].
3.1.2 Tense vowels and diphthongs
Tense vowels and diphthong vowels occur in open and closed vowels in
GE. Some of these vowels are quite stable, having very little variation
in pronunciation. In most CECs, tense vowels are usually lengthened in
opposition to the shorter lax vowels, but in GE they have no significant lengthening as opposed to the length of lax vowels. Table 9.2 below
provides a summary of the variation described in the following paragraphs.
fleece /i/
The fleece vowel /i/ is quite stable, occurring in open and closed
syllables. It is more like North American varieties of English rather than
varieties found in the British Isles. Cruttenden (2001: 106) says this vowel
has been stable since 1500. A few words spelled with <ea> shifted from
[   ] to [i] by the end of the seventeenth century. There are
a few occurrences where the fleece vowel is lengthened in the stressed
syllable of a multi-syllable word, for example /tis/ teachers and /izj /
easier. This instability of the <ea> words in Early Modern English may
account for this feature in GE.
face /e/
In GA and RP English the face vowel is a diphthong /e/, but in GE, and
some CECs, it is a monophthong /e/, as in /seln/ sailing. It is found in
both open and closed syllables. The /e/ pronunciation is similar to varieties
of English in lower northern and north Midland England, Ireland, and
Scotland. There are a few words for which I have samples with and without


ken decker

the diphthong /e/, for example /mebi  mebi/ maybe. (See comments in
the dress vowel section above concerning the [  e  e] variation.)
The pronunciation of the diphthong may be a result of convergence with
SE pronunciation learned through school instruction.
Two of my informants each said name as /nm/. The // diphthong
is the basilectal SKC pronunciation of the face vowel. This may be an
indication of convergence with SKC, but it is weak since it only occurs in
one word.
goat /o/
In GA and RP English the goat vowelis a diphthong /o/ or // respectively. In GE, and some CECs, the goat vowel is a monophthong /o/,
as in /nobd/ nobody. The /o/ pronunciation is similar to some varieties
of English in lower north and north Midland England, Ireland, Scotland,
and Wales. Cruttenden (2001: 136) says that the shift from Old English
[] to present-day // in open syllables reached a quality near [o] in
the sixteenth and seventeenth centuries. This may be an indication of the
divergence of this vowel from what became GA and RP.
In rapid speech I have examples of it becoming more central as [],
losing labial rounding becoming [], sometimes with a slight drop to [],

or dropping as low as []. There are also examples of it becoming nasalized

when followed by a nasal consonant, as in /kom/ come, or an implied nasal,
as in /do/ dont.
Cruttenden (2001: 114) says that Present RP English // derives from
Middle English //. In the nineteenth century this vowel shifted to [].
Possibly this provides an explanation for the variability in GE.
One of my informants pronounced boat as /bot/. The /o/ diphthong
is the basilectal SKC pronunciation of the goat vowel. This may be an
indication of convergence with SKC, but it is weak since it only occurred
in one word.
goose /u/
The goose vowel /u/ is quite stable, occurring in open and closed syllables. The term /skun/ schooner was the only word in which the goose
vowel was consistently lengthened by all informants. The unlengthened
pronunciation is more like North American varieties of English rather than
varieties found in the British Isles.
mouth /a/
The mouth diphthong /a/ is quite stable in GE, occurring in open
and closed syllables. GA and RP varieties of English also pronounce

The English of Gustavia, St. Barthelemy


the mouth vowel as /a/. All CECs also have a mouth vowel diphthong that begins at a low or midpoint sliding to a high, back rounded
point. As part of the Great Vowel Shift, Middle English [u] began
diphthongization in the fifteenth century. It reached the present pronunciation in the eighteenth and nineteenth centuries (Cruttenden 2001:
Coopers (1979: 65) description of SKC gives /ou/ as the pronunciation for the mouth vowel. In a few words one of my GE informants
raised the pronunciation as high as /ou/ when the diphthong was the
first syllable of a word. Another informant pronounced it as /o/ when
preceding a velar nasal //. It is possible that there may have been some
convergence with SKC that has created this variability, but it is weak
choice //
In many CECs the choice vowel // has become unrounded and merged
with the price vowel //. There are only a few examples of this shift in
GE: /bz/ boys, /pnt/ point, and /ws/ voice. Cooper (1979:52) does not
list // as a basilectal diphthong in SKC, but he lists /chais/ choice as an
example of the // diphthong.
However, in some cases GE has also done the opposite. Throughout
the data there are examples of the price vowel // beginning at a low
rounded point, merging with the choice vowel //. My older informants,
and samples from Maher and Parsons, tend to use a raised and rounded
beginning pronunciation //. This results in pronunciations such as: /sd/
side, /ln/ island, /wl/ while, and /fn/ find. The younger informants
used // or a slightly raised //.
The Atlas of English Dialects (Upton and Widdowson 2006: 18) identifies
this pronunciation for find in parts of the West Country, from the West
Midlands to the Isle of Wight, Essex, and Kent. Wright (18981905) also
identified this pronunciation for time in Ireland.
Other than Gustavia, I have only observed the raised pronunciation //
for the PRICE vowel in St. Maarten. Roberts (1988: 92) mentions this
as a feature found in Bajan. Neither Wells (1982), nor Holm (1988), nor
Hancock (1969, 1986) mention this as a feature of CECs.
price //
Aside from the // > // and // > // shift described above, the price
vowel // occurs in many words, just as in SE. It is very stable, found in
open and closed syllables.


ken decker

3.1.3 Consonants
There are several features that are generally considered representative of the
phonologies of CECs (Baker 1999, Holm 1988, Wells 1982). Some of these
features occur in GE, but only a few of them are as strongly descriptive of
GE as for the CECs.
TH-stopping describes the merger of the interdental fricatives // and
// with /t/ and /d/ respectively. Wells (1982: 565) says that this feature
is so strong in CECs that the fricatives only occur in the speech of the
most educated. This feature is prevalent in GE, but not exclusive. In the
texts from Parsons and Maher, <th> does appear in a few lines.3 In my
recordings there are no examples of the fricatives. Cruttenden (2001: 184)
says that this merger often happens in southern Irish speech. As with many
African languages, the dental fricative sounds are rare in many languages
globally. Therefore, this feature does not need to be traced to Irish influence,
neither does it indicate substrate influence from Africa. The development
of this feature could simply be considered an innovation parallel to that of
the CECs.
Word-final consonant cluster reduction
SE words that end with a consonant cluster usually do not have the final
consonant in CECs. Word-final consonant cluster reduction is found many
English varieties. In my previous article (Decker 2004a: 232) I reported that
this feature was not frequent in GE. However, with the newer text and a reevaluation of my earlier data I find that a majority of word final consonant
clusters are reduced. Several of the typical consonant cluster reductions are
shown in the following examples. The /nd > n/ reduction can be seen in
example (1). There are two examples of the /kt > k/ reduction in example
(2). The /nt > n/ reduction can be seen in example (3), but in example (4)
the process has gone further by eliminating the nasal consonant also and
nasalizing the preceding vowel. The /st > s/ reduction is seen in example
(5). Notice how the /tn/ combination in example (5) is not reduced; this is
because there is a syllable boundary between the two segments. The /bz/
combination in example (5) is not reduced because there is a morpheme
boundary between the segments.
(1) /d skn tm/
the second time

The Parsons and Maher texts are not transcribed in phonetics.

The English of Gustavia, St. Barthelemy


(2) /d fk dt wi hd kntk wd dis pipl/

the fact that we had contact with these people
(3) /diz ovmn pipl/
these government people
(4) / hd t go t se tms/
I had to go to St. Thomas
(5) /ow tkn s tk ot stn vbz/

our (way of ) talking just takes out certain verbs

The loss of preconsonantal and postvocalic /r/ in the prestigious RP dialect
began in the fifteenth century, but other varieties retained it (Upton and
Widdowson 2006: 31).4 Rhoticity is a variable feature in GE. In most cases,
postvocalic /r/ has been dropped as in example (6). However, there are a
few examples of postvocalic /r/ and rhotic vowels //, as in example (7).
(6) /bt tn jz ft/
about ten years after
(7) /de r ln/
they are learning
The speech of Barbados is fully rhotic (Wells 1982: 570). The scarcity of
post-vocalic /r/ in GE may indicate the lack of any significant linguistic
input from Barbados.
VW confusion
Some varieties of CEC have variability in the use of /v/ and /w/. (See Baker
1999: 329.) In GE there are examples of /v/ shifting to /w/, as in example
(8) below. There are also just as many examples of /w/ shifting to /v/, as in
example (9) below. My youngest informants did not exhibit this feature.
(8) /yu gt mn ws/
you have a mans voice
(9) /vl vn di mrkn siplen . . . /
well, when the American seaplane . . .
Holm (2000: 162) raises the possibility that this variation may result from
substratal influence from West African languages. Wells (1982: 568) says that

/r/ = [].


ken decker

it is uncertain whether this feature has been influenced by eighteenth- and

nineteenth-century London Cockney or an African substratum lacking
/v/. It is also possible that the absence of a consistent model in either the
superstrate or substrate resulted in a continued variability in GE.
Palatalization of velar plosives
Palatalization is common in many varieties of English, including GE.
This occurs in words like: cute, mule, and popular. In CECs this feature
is expanded to include voiceless and voiced velar plosives, /k/ and //
preceding the lengthened palm vowel // (Roberts 1988: 57). This feature
is found in SKC in /kj t/ cart and /j dn/ garden (Cooper 1979: 54).
However, palatalization of velar plosives before the palm vowel is very rare
in GE. It was only noted for certain in one word in Parsons texts, seen in
example (10).
(10) . . . and cyarried her home
The velar nasal
In CECs, words from SE with the mouth vowel /a/ preceding an alveolar
nasal /n/ have shifted to a goat vowel/o/5 followed by a velar nasal //.
Thus, /mantn/ mountain becomes /motn/. My older informants were
consistent in their production of the CEC, as in: /to/ town, /kro  kr/
crown, and /pronos/ pronounce. My younger informants were more inconsistent, for example producing /raon  ron  r/ around and /to/

town. We do not know if this feature was present in Parsons or Mahers

texts because they are not transcribed in phonetics.
I have not been able to find any justification for this shift in either
English or African languages.


Baker (1999), in his attempt to identify diagnostic features that identify

CECs, lists 138 features, of which 82 are lexical items. There are very few
lexical features that suggest any connection of GE with CECs. One of
Bakers features is the use of various forms of /fi  fo  fa/ as an infinitive
marker. Holm (1988: 168ff.) proposed that this feature may be a result

Wells (1982: 572) says it is a strut vowel //, but I have always heard it higher and with rounding
of the lips.

The English of Gustavia, St. Barthelemy


of convergence between British English and African languages. There are

several examples of fah in Parsons and Mahers texts, but they all have the
same prepositional role of for. In my data there are a few examples of /f/
used like a CEC infinitive marker; see examples (11) and (12).
(11) /dm jst go y no f ple n di pulz/
they used to go, you know, to play in the pools
(12) /z snd hi kstj um f hm f rs f sel ms/
is sending his costume, for him to dress for the sailors mass
Most CECs have a pronominal system that is somewhat different from
English. There is no difference between subject or object pronouns, nor
are there possessive pronouns. However, GE has all of these features. There
are only a few examples of pronouns being used in nonstandard ways. In
example (11) above, the word /dm/ they, instead of /de  de/, is a rare
example of a CEC pronoun being used by one of my GE informants.
Another CEC use of they is in the phrase they have/had, meaning there
are/were, which functions as an impersonal introductory discourse function. Allsopp (1996) noted this feature is used throughout the Caribbean.
The use of awi for the first-person plural pronoun we is typical of eastern
Caribbean creole (Baker 1999: 331). One of my older informants used this
pronoun once.
Forms of yer and yerry are found in many CECs. There is one occurrence
of /jr/ hear from my older male informant in example (13). There is some
debate as to the origin of the pronunciation (Baker 1999: 326), but Wright
(18981905) says this is a dialectal pronunciation found in northern and
western regions of England.
(13) / jr f di gdn/
I heard it was in the garden.
In Parsons texts there are several words that are not used in Modern
SE. There are several occurrences of the verb reach meaning to arrive. It
occurs quite commonly in northern England and Scotland (Wright 1898
1905) and throughout the Caribbean (Allsopp 1996). Press is a Scottish
English word meaning a clothes cabinet (Wright 18981905). Allsopp
(1996) says it is found throughout the eastern Caribbean. A mare, spelled
mere by Wright (18981905), is used throughout northern England and
Scotland to mean a small pond, but it is not reported elsewhere in the


ken decker

In the speech of my older informants there are numerous examples,

and variations, of /twz/ as a past-tense marker or topicalizer. I will
discuss its function in the next section. The variations in form include
/tw  t  tz  z  wz/. Maher (1987) records /wz/ and twasnt.
According to Wright (18981905), twas was very common in eighteenthcentury England.
There is one example of each of these kinship terms: tree young men sons
sons and gyirl chil daughter. Allsopp (1996) suggests that these terms
may be a calque from West African languages.
The only other noteworthy thing that can be said about the lexicon is
that my two older informants used more French later in the conversation
as they encountered more topics for which they did not know the English
words. There are also a few examples of French words in Parsons texts.
This is evidence of the increasing importance of French to the lives of the
GE speakers, and erosion of their English fluency. The English of my two
younger informants had more similarities with SE in pronunciation and
the speakers dexterity with English lexical items. This may be because they
received better English education in school than the older pair.


The English of Gustavia has become a fairly stabilized koine. There are
nearly eighty to ninety years separation between Parsons texts and mine,
but there is very little difference in the morphosyntax of the speech. However, as I pointed out in my previous article (Decker 2004a), there is some
variation in the speech. Furthermore, there is very little difference between
GE and SE sentence structure. In this section I will only discuss a few
variations between GE and SE.
3.3.1 Noun phrases
The basic word order of the noun phrase in CECs is essentially the same as
in English, but there are differences in pluralization and possessive phrases.
In CECs pluralization is accomplished with the use of the morpheme
dem either before or after the noun. In nearly all of the data GE marks
plural like SE. There are a few words that do not have the plural suffix -s.
Following is an example from a Parsons text that has a more CEC plural
(14) de tree big boys dey got up
the three boys got up

The English of Gustavia, St. Barthelemy


While dey is not necessarily a creole word, and there is a plural -s on boy,
the dey is in the position of a creole plural marker and unnecessary for an
English sentence.
GE marks possession, in most cases, just as SE. Unlike CECs, GE uses
possessive pronouns or a possessing noun with possessive suffix -s preceding
the possessed noun. There are a few cases in which the possession marking is
absent in my GE data. Example (15) is from a Parsons text and example (16)
is from my informant: the CEC-like possessive construction in example
(16) uses a pronoun // unmarked for gender as some CEC third-person
singular pronouns.
(15) it was his moder doings
(16) /k di mj son z n t hi bot/
because the mayors son has gone to his boat
3.3.2 Verb phrases
Holm (2000: 174) says there is no particular set of syntactic features [that]
will identify a language as a creole without reference to its sociolinguistic history, it is also true that the structure of the verb phrase has been
of primary importance in distinguishing creole varieties from non-creole
varieties. One of the primary features is that most creole verbs remain
uninflected regardless of time reference or subject agreement. The GE verb
phrase is not like the CEC creole verb phrase, but there is evidence that
GE may have been influenced by neighboring CECs.
Past tense
The CECs are known for the tense, mood, and aspect (TMA) structure
of the verb phrase. The so-called past tense in CECs, which Bickerton
(1975) called an anterior tense, functions differently from English past
tense. There is virtually no evidence of the CEC TMA structure in GE.
Most GE verbs are inflected as in SE. However, there are many verbs that
are not marked for SE past tense nor are they marked for CEC anterior
tense either.
To get a sense of the variation I counted all of the verbs in Parsons
texts that could be marked for past tense as indicated by the context. I
excluded the copula form (was), modal verb (would have), and irregular
verbs that have a unique past participle (gone). There were 198 verbs that
could be marked with the -ed suffix for past tense, 65 (33%) were not
marked correctly for SE. There were 91 irregular verbs and 18 (20%) of
them did not use the proper past-tense form.


ken decker

According to Cooper (1979: 81) the past reference in SKNC6 depends

more heavily on context than on the actual presence of the preverbal
formative bin.7 Example (17) shows an SKC utterance with an unmarked
past tense.
(17) mi wok haad dis maanin aredi (from Cooper 1979: 83)
I worked hard this morning already
(18) When de moder return she found de pig (from Parsons)
When the mother returned she found the pig
Example (18) shows a GE sentence with one of the verbs unmarked for
past tense other than the context. It is possible that all of the unmarked
and uninflected verbs are functioning like SKC verbs. I dont have a good
explanation for this variation. It might be as a result of influence from
neighboring CECs or there may be some sort of relexification.
There were several examples of the use of /dd/ as a past-tense marker,
as in examples (19) and (20). In my previous article (Decker 2004a) I
proposed that this might mark past perfect tense. The use of did + VERB
for simple past tense was common in seventeenth-century southwestern
English varieties (Winford 2003: 315). Allsopp (1966) lists several uses of
did as part of the past tense verb phrase in several CECs: simple past,
imperfect, and perfect. This only occurred in the speech of my oldest pair
of informants.
(19) /hi dd ln t/
he lent it  he had lent it
(20) /dei dd fl p d bi/
they filled the beach  they had filled the beach
There were also several examples of the use of /doz/ as a habitual aspect
marker, as in example (21). Allsopp (1966) says this is found throughout
the eastern Caribbean. This only occurred in the in the speech of my oldest
pair of informants.
(21) /i doz prons t d/
she pronounces it good

St. Kitts Nevis Creole.

Cooper (1979: 82) explains that bin is not actually a past tense, or anterior tense marker in SKNC,
bin functions primarily as a sequencer: it relates one action, event or state of affairs to another
action, event or state of affairs.

The English of Gustavia, St. Barthelemy


There are only a few examples of these CEC-like preverbal TMA markers. Most verbs are marked for past tense. There are several examples from
each of Parsons storytellers where they say the same verb correctly inflected
in one sentence but incorrectly in another sentence.
Copula deletion
In some CECs (Decker 2004b: 106; Holm 1988: 175ff.) there is a feature of
copula deletion in descriptive clauses, when the subject is modified by an
adjectival phrase. (See example (22) from Belize Kriol.) There were several
examples of copula deletion in Parsons and my texts, see examples (23) and
(24). However, these are not all descriptive clauses with adjectival phrases.
Cooper (1979) says nothing about copula deletion in SKC.
(22) Di froot aal swibl op.
The fruit is all shriveled up.
(23) I (am) sure dat . . . (from Parsons)
(24) /i no pltn/ (from my data)
I (am) not a politician
Subjectverb number confusion
Due to the fact that most CEC verbs are not inflected, neither is there
subjectverb agreement marking on the copula. However, in GE most
copula verbs are marked for subjectverb agreement. There are only a few
examples in Parsons texts in which the copula is not in the correct form
for subjectverb agreement.
(25) De one dat you is (are) goin to marry
(26) Dere were (was) a neighbor

4 Conclusion
The history of Gustavia, St. Barths, was quite different from the developments on plantation islands like St. Kitts or Antigua. In Gustavia,
non-English speakers and non-SE speakers worked alongside one another.
There were probably many varieties of English, but within the confines of
the small geographic area there was sufficient opportunity for a koine to
GE is a unique variety of English, but it is not as divergent as the creole
varieties found on some neighboring islands. There has been little change


ken decker

in GE in the last hundred years. This is best exhibited by the degree of

similarity between Parsons texts from the 1930s, Mahers texts from the
1980s, and my texts from the 2000s. The data support Mufwenes (2008:
84) thesis that, One is hard pressed to find in creoles any grammatical
features that have not been selected from the nonstandard varieties of the
relevant vernaculars or in their substrate languages, although these have
not been replicated faithfully. There are a few grammatical features that
are similar to CEC structures. However, they only occur in a few locations
in the speech of a small number of the informants.
I have suggested that the phonological variation in GE vowels can be
explained by the changes in the phonology of English during the seventeenth century. However, if GE developed in the late eighteenth century,
the developments in England would seem too early. I have also shown
how some pronunciations that developed in the seventeenth century are
still retained in some regional varieties to this day. So it is not impractical
to expect that in the eighteenth century there were migrants from rural
areas of England who traveled to the Caribbean and provided the building
blocks of GE.
I believe that GE represents an early stage of contact between Africans
and Europeans. The contact occurred in a context in which slaves and owners lived and worked closely; a context which offered sufficient opportunity
for the slaves to acquire English. The presence of some SE features and
the absence of certain CEC features are further evidence that there has not
been a creolization or decreolization process. There are very few features,
if any, that might suggest African substrate influence. In each case there
would have been convergence between similar features in the substrate and
superstrate languages. This would be a typical phenomenon in any contact
situation and not a unique creolization process.
Gustavia English will probably only survive for a couple more decades
at best. Most of the remaining speakers are elderly and the middle-aged
speakers have more opportunities to use French on a daily basis. A further comparison of GE with the English of Saba and Statia may reveal

Aceto, Michael. 2007. Statian Creole English: an English-derived language emerges
in the Dutch Antilles. World Englishes 25: 41135.
Allsopp, Richard. 1996. Dictionary of Caribbean English Usage. Oxford University

The English of Gustavia, St. Barthelemy


Baker, Philip. 1999. Investigating the origin and diffusion of shared features
among the Atlantic English Creoles. In Philip Baker and Adrienne Bruyn,
St. Kitts and the Atlantic Creoles. London: University of Westminster Press,
Bickerton, Derek. 1975. Dynamics of a Creole System. Cambridge University Press.
1981. Roots of Language. Ann Arbor: Karoma.
Bolinger, Dwight. 1989. Intonation and Its Uses. Stanford University Press.
Byrne, Francis. 1984. Fi and fu: origins and functions in some Caribbean English
based creoles. Lingua 62: 97120.
CIA. 2012. The World Fact Book. www.cia.gov/library/publications/the-worldfactbook/geos/tb.html (accessed 12 November 2012).
Cooper, Vincent O. 1979. Basilectal Creole, decreolization, and autonomous language change in St. Kitts-Nevis. Unpublished PhD dissertation, Princeton
Cruttenden, Alan, ed. 2001. Gimsons Pronunciation of English, 6th edn. London:
Decker, Ken. 2004a. Moribund English: the case of Gustavia English,
St. Barthelemy. English World-Wide 25(2): 21754.
2004b. The Song of Kriol: A Grammar of the Kriol Language of Belize. Belize
City: Belize Kriol Project.
Fleischmann, Ulrich. 2005. Black Culture, White Discourse and Creole History:
A Study on Interpretations of American Slavery. www.larramendi.es/i18 n/
catalogo imagenes/grupo.cmd?path=1000198 (accessed 6 November 2012).
Hancock, Ian. 1969. A provisional comparison of the English-based Atlantic
creoles. African Language Review 8: 772.
1986. The Domestic Hypothesis, diffusion and componentiality: an account of
Atlantic anglophone creole origins. In Pieter Muysken and Norval Smith,
eds., Substrata Verses Universals in Creole Genesis: Papers from the Amsterdam Creole Workshop, April, 1985. Amsterdam: John Benjamins, 71
Holm, John. 1988. Pidgins and Creoles, vol. 1: Theory and Structure. Cambridge
University Press.
2000. An Introduction to Pidgins and Creoles. Cambridge University Press.
Jeffry, Daniella. 1997. Understanding the language situation in Saint Martin: the
historical perspective. Unpublished manuscript.
and Malin Runsten. 2001. St. Barthelemy Sveriges sist koloni.
Kohler, Asa
pdf (accessed 10 December 2002).
Maher, Julianne. 1987. Transcription of an interview. Unpublished manuscript.
1996. Fishermen, farmers, traders: Language and economic history on
St. Barthelemy. Language in Society 25: 373406.
2010. The roots of linguistic conservatism in St. Barthelemy. Paper presented at
the 18th Biennial Meeting of the Society for Caribbean Linguistics Conference, University of the West Indies, Cave Hill, Barbados, 913 August 2010.


ken decker

Mufwene, Salikoko. 2001. The Ecology of Language Evolution. Cambridge

University Press.
2008. Language Evolution: Contact, Competition, and Change. London:
Parsons, Elsie Clews. 193343. Folklore of the Antilles, French and English, 3 vols.
New York: American Folklore Society.
Roberts, Peter. 1988. West Indians and Their Language. Cambridge University
Shrimpton, Neville. 1994. A Preliminary Note on Some Afro-English (English Creole)
Texts from Saint Bartholomew. Umea, Sweden: Umea University, Department
of English.
Shrimpton, Neville and Philip Baker. 1995. Buddy Quow, St. Kitts and St. Barts.
In Philip Baker, ed., From Contact to Creole and Beyond. London: University
of Westminster Press, 8196.
Upton, Clive and J. D. A. Widdowson. 2006. An Atlas of English Dialects. Oxford,
Oxford University Press.
Wells, J. C. 1982. Accents of English, vol. 3: Beyond the British Isles. Cambridge
University Press.
Williams, Jeffrey. 2003. The establishment and perpetuation of Anglophone white
enclave communities in the Eastern Caribbean: the case of Island Harbour.
In Michael Aceto and Jeffrey Williams, eds., Contact Englishes of the Eastern
Caribbean. Amsterdam: John Benjamins, 95119.
Winford, Donald. 2003. An Introduction to Contact Linguistics. Oxford: Blackwell.
Wright, Joseph. 18981905. English Dialect Dictionary. London: Oxford University

ch a p ter 1 0

Anglo-Paraguayan English
Danae M. Perez-Inofuentes


Paraguay is a land-locked country in the heart of South America bordered

by Argentina to the south, Bolivia to the north, and Brazil to the east.
Six and a half million Paraguayans live on 400,000 square kilometers.
Among linguists, Paraguay is of particular interest because of its societal
bilingualism. Its official languages are Spanish and Guarani, that is, Guaran
Paraguayo, a koine of a Tupi-Guarani language that underwent considerable
restructuring due to language and dialect contact. Even though it was long
regarded as backward and a hindrance to progress, Guarani is spoken by the
majority of the Paraguayans today, and since its recognition as one of the
two national languages in the 1967 Constitution, it has become a symbol
of national identity (Melia 2011). Thus, while Spanish is principally used
for official purposes and secondary education, Guarani is the preferred
language of 60 percent of the Paraguayan households. Most Paraguayans
speak a variety of jopara, a language consisting of both Spanish and Guarani,
which offers a showcase of language mixing (Zajicova 2009).
The Paraguayan diglossia is rooted in history. Due to Paraguays relative
remoteness in the South American hinterland, the European immigration progressed slowly in colonial times. Instead of receiving immigrant
families, Paraguay was rather attractive to individual adventurers trying
their luck. Accordingly, most immigrants were single men, and their offspring with Paraguayan women resulted in a stable Guarani-speaking mestizo population (Melia 2011: 4257). The War of the Triple Alliance in
1870, however, halved the Paraguayan population to 336,000. In order to
overcome the subsequent economical stagnation, the Paraguayan government intended to stimulate European immigration anew by offering tax
relief and fertile lands at low prices to immigrant communities. Hence,
many immigrant groups such as Mennonites, German imperialists, or
French agriculturalists sought a future in Paraguay around the turn of the


danae m. perez-inofuentes

century (Warren 1985). Given that some of these enclaves maintained their
heritage languages, developing new varieties of them, languages from all
over the world are now spoken natively in Paraguay. As a result, Paraguays
particular heteroglossic situation and societal bilingualism are unparalleled
One of these immigrant groups is particularly interesting for sociolinguists. Like most other Latin American countries such as Argentina and
Brazil, Paraguay experienced considerable immigration from the British
Isles during the nineteenth century. Important enterprises such as the railway and other transportation companies as well as influential financial
institutes were British-run. Nevertheless, the Anglo-Paraguayan community differed from other English-speaking communities in Latin America
due to its extraordinary inception. While it consisted of no more than a
hundred British subjects in 1884 (Warren 1985: 16), it grew significantly in
1893 when a group of over four hundred Australians arrived in Asuncion.
They were a group of socialists who had left their country on the brink of
civil war in Queensland. In response to the great Shearers Strike in 1891,
they planned to set up a communist society without social stratification as
an example to show the world that capitalism was outdated. The country
they chose to put this project into action was Paraguay (Kellett 1997).
Even though these conditions seemed ideal for a new, contact-influenced
English variety to emerge, English was not maintained as a communal language among the Australian immigrants in Paraguay. The utopian project
failed, and the colony soon disintegrated. Nevertheless, descendants of the
original settlers still live in Paraguay, and their individual relationships to
their heritage language vary considerably. This chapter is a first attempt to
describe this hitherto unknown English-speaking community in Paraguay.
After giving an account of the sociolinguistic history, I will describe some
aspects of the internal development of the English language in this unique
case of trilingual language contact, paying particular attention to contactinduced phenomena. The data presented here are part of an ongoing
research project that started in 2011 with several weeks of fieldwork.

The sociolinguistic history

The Australian colonists acquired an extensive piece of land close to the

town of Coronel Oviedo, approximately 120 kilometers from Asuncion,
and named it New Australia. According to the aim of the project of setting
up an independent and self-supporting community, they chose a particularly remote location where neither a river nor a road would connect

Anglo-Paraguayan English


them with the local population and market. This condition was of pivotal
importance as all participants had committed themselves to two years of
teetotalism until the community gained a foothold; geographic isolation
would keep possible temptations away. The most important condition,
however, was to stick to the color line, and accordingly, all members
of New Australia were English-speaking whites and supposed not to get
married to non-whites (Livermore 1950).
The settlers in New Australia stemmed from different social backgrounds
and origins. Their leader was William Lane, a young journalist from Bristol
and editor of Australias first union-owned journal The Worker. Lane was
the initiator of the New Australian project and aimed to demonstrate that
socialism was the highest form of human society (Kellett 1997: 9). He
explained his conception of socialism and mateship in his 1892 novel The
Workingmans Paradise as follows:
. . . mates is them wots got one pus. If I go to a shed with Jack an were
mates an I earn forty quid and Jack gets sick an only earns ten or five
or mebbe nothin at all we puts the whole lot in one pus . . . If Jacks got
the pus an I want half-a-crown, I says to Jack, says I, Jack, gimme the
pus. (Miller 2007: 130)

Each member contributed at least 60 to the common pool of the New

Australia Co-operative Settlement Association to buy land, cattle, and
tools, and an independent currency was used within New Australia.
The New Australian community consisted of two groups of people.
The majority of the settlers were the bushmen, that is, shearers and
other agricultural workers from Queensland and New South Wales. Lane
was convinced that these courageous and rough looking but tender
workers were ideal participants to live up to his ideals of socialism and
brotherhood and to face the privations and hard work awaiting them in
Paraguay (Kellett 1997: 11). Nearly all the bushmen were young bachelors
of little or no formal education. The other group of settlers consisted of
educated white-collar middle-class workers, such as journalists, teachers,
and tradesmen. Their common goal was to establish a settlement of at
least 1,200 families independent from the Paraguayan society a goal that
was not achieved: by the end of the century, approximately six hundred
Australian and British settlers had joined the socialist project in eastern
Paraguay (Souter 1991: 282).
At the time when the Australian pioneers arrived, the sociolinguistic situation in Paraguay offered conditions that seemed ideal to ensure
the maintenance of the communitys racial and linguistic purity. The


danae m. perez-inofuentes

communication with the Paraguayan administration was carried out in

Spanish and made possible by a number of interpreters. The communitys
immediate neighbors, however, spoke Guarani. The inferior status of the
Guarani language and people was expected to inhibit miscegenation. Thus,
even though Guarani and Spanish were indispensible to get in touch with
their immediate environment, the maintenance of the English language
seemed to be secured.
Apart from the communicative difficulties, the Australian settlers in
Paraguay encountered living conditions that challenged them additionally.
On the one hand, the clearing of the land was laborious, and three serious
droughts during the 1890s destroyed most of their first crops. Other inconveniences such as wild animals, insects, diseases, and untamed cattle made
life on the colony even more difficult. According to Henry Connelly,1
one of the original settlers, it was particularly the women who suffered
from the hardships and instigated quarrels among the settlers. On the
other hand, the groups communistic enthusiasm was not to last. Lanes
totalitarian leadership and commitment to abstinence from alcohol consumption, mateship, and the maintenance of the color line were soon
subject to debate in New Australia, and the bachelors and bushmen, in
particular, pursued interests that diverged from those of the middle-class
families. In addition, some men soon bartered goods and alcohol with
the native families that inhabited their land. As a consequence, the colony
was divided after a few months. Lanes loyalists, mainly educated middleclass families, reinitiated the experiment in another location called Colonia
Cosme, whereas most of the bushmen and bachelors remained with the
cattle stock on the original site of New Australia (Livermore 1950). Today,
the two communities still exist, and none of them maintained English as
a communal language. However, the sociolinguistic developments of the
two villages diverge considerably. In what follows, I give an account of each
community individually.

The Cosme community

Sixty-two utopians stood by William Lane after the divide and moved with
him from New Australia to the particularly isolated spot known as Colonia
Cosme. Thirty-six of them were single men or married men without their
wives, and there was one single woman, eight married couples, and nine
children (Souter 1991: 110). Among them was Lanes brother John as well as

In a document written in 1924, which is part of a collection of papers kept in Asuncion by Padre
Santiago, James Feehan, who put it at my disposal for the purpose of this research. It consists of
roughly 15,000 words and was transcribed faithfully in the 1950s by Ricardo Smith in Nueva Londres.

Anglo-Paraguayan English


most of his closest friends such as his advisor and translator Arthur Tozer
from London, William Wood from Queensland, and William Saunders
from California. In 1896, William Lane traveled to Australia and the British
Isles to recruit more people willing to join his socialist project in eastern
Paraguay, and a few more families as well as Mary Cameron (later Dame
Mary Gilmore) arrived in Cosme. In 1903, Colonia Cosme had a population
of ninety-seven (Souter 1991: 205).
Having left all their economic capital in New Australia, the Cosme
community had meager financial resources to start with. However, since
almost all its members were educated middle-class people, the social capital was considerable. Cosme had a school with a library of more than
1,200 books (Souter 1991: 188), and John Lane and Mrs. Mary Cameron
were in charge of teaching the children history, geography, evolution, and
science. The Cosme community organized gatherings and dances, read
poetry, and performed plays in the social hall. In accordance with their
goal of setting up an English-speaking society apparently, Lanes idea
was to include Paraguay some day to the British Empire (Souter 1991: 169)
the Cosme colonists put special emphasis on transmitting knowledge
of the English literature to the next generation. In addition, the Cosme
Monthly and the Cosme Evening Notes published regularly the news and
announcements of the community. This endocentrism notwithstanding,
the settlers also intended to improve their skills in Spanish. The following
poem illustrates their efforts:
Maria hado un lambo poco
Con vestido as blanco as papel
Y todos partes que Maria va
El lambo vaed ciertamente tambien.
(cited in Souter 1991: 156)

Mary had a little lamb

With dress as white as paper
And everywhere that Mary goes
The lamb was sure to go, too.

This Spanish translation of Mary had a little lamb is undoubtedly a creative

masterpiece. However, its sheer unintelligibility suggests that the first generation had limited knowledge of Spanish. In short, the Australian settlers
in Paraguay spoke neither Spanish nor Guarani and were committed to
maintain their English heritage language.
However, the communitys slow economic progress and the demographic
imbalance induced many community members to seek a better life elsewhere. William Lane himself withdrew disillusioned in 1899 and settled in
New Zealand. Other families also left Paraguay as they realized that their
teenage children were about to marry into the Paragayan society. Many
settlers, among them Mary Gilmore (nee Cameron), went to Argentina
either to join the prospering Anglo-Argentine community, or to earn their


danae m. perez-inofuentes

passage to Australia (Whitehead 1997). Private correspondance indicates

that many of them stayed in contact with the Anglo-Paraguayan community. Colonia Cosme and the socialist experiment were officially dissolved
in 1909.
The most prolific families that stayed in Cosme thereafter were the
Woods and the McLeods from Queensland as well as the Titilahs from
Scotland. The second generation that was growing up in this carefree
setting enjoyed the plays and the fancy dress parties at the social hall
as well as their freedom out in the woods. Some of the boys born on
Colonia Cosme, however, spent time abroad when they fought for the
British Empire at Gallipoli in World War I, thereby assuming their British
citizenship. At home, the children spoke English only, but when playing
with their Paraguayan peers, they were regularly exposed to Guarani and
Spanish. Hence, the second generation of the Cosme community was
trilingual in English, Spanish, and Guarani. A 75-year-old descendant of
the Wood clan explains that all the children had the bad habit of language
mixing, referring to code-switching and the frequent use of loanwords in
all three languages. Her cousin recalls that his father, who had grown up in
Cosme, spoke all three languages fluently, but that he had a gringo accent
when he spoke Spanish.
However, the next generation also chose to desert Cosme. The reason for
this exodus was not only the lack of prospects, but also the limited amount
of property to be transferred to the large numbers of descendants. The
young Cosme emigrants found good jobs in British-run companies such
as the railroad company in nearby Villarrica, a booming town with a large
British community (Warren 1985: 24). The Wood girls, for instance, worked
as flight attendants, translators, or clerks at the American embassy, while
their cousins managed British-owned cattle ranches in northern Paraguay.
All these jobs had one thing in common: The command of English was
indispensable. Thus, given their modest financial resources and limited
access to further education, the English language was their only means
to gain social mobility and, in fact, accelerated the exodus from Colonia
The last incentive to leave Cosme was the visit of Gavin Souter in 1966.
While working on his book A Peculiar People (1991, first published in 1968),
Souter visited Cosme and was surprised at the state of the English language. At that time, Cosme had approximately three hundred inhabitants
(Souter 1991: 269). On his return to Australia, he connected the Wood
family with their relatives in Australia, whereupon two young men left
Paraguay for Australia. Soon after, other descendants of the McLeod family

Anglo-Paraguayan English


As a result of this exodus in the course of the twentieth century, the

descendants from Cosme live scattered all over the country today. Many of
them settled in the big cities of Asuncion, Concepcion, and Encarnacion
and speak Spanish as their first language with basic skills in Guarani and
English. Thanks to the achievements of their parents, they are relatively
well off, which gives them the opportunity to study English as an L2
at school or abroad, above all in the US. Their motivation to do so,
they claim, is due to their cultural heritage as well as to enhance their
competitiveness. It is therefore possible to affirm that the Anglo-Paraguayan
case is similar to the Anglo-Argentine community of Buenos Aires where
only upper-class descendants have the financial means to maintain English
as an L2 (Cortes-Conde 1996: 119). In Cosme, the English language is
practically non-existent today. There is only one fluent English speaker,
and daily communication takes place in Guarani. In brief, Spanish is the
first language of the urban descendants from Colonia Cosme, whereas
Guarani, noticeably an indigenous language, is now the principal language
of communication in Colonia Cosme itself. English is reappearing as an L2
among the next generation of Anglo-Paraguayans who have the financial
means to study it.

The New Australia community

In the meantime, New Australia progressed in a very different way. In

fact, no in-depth research was conducted in New Australia, and due to
the rapid decline of its Anglo-Saxon and socialist identities after Lanes
departure, historical accounts did not pay particular attention to this
community. Later studies of oral history only followed the most visible, English-speaking descendants principally from Cosme (Whitehead
1997). However, New Australia is the location where this research project
After the divide, less than eighty persons remained on the original site
of New Australia (Souter 1991: 144). Their new leader was one of Lanes
closest friends and greatest critics, the Irishman Gilbert Casey. However,
in contrast to the utopians in Cosme, New Australians were not without
means. The generous support from the Paraguayan government, added to
the funds raised by William Lane, had provided New Australia with a fair
capital to start off, and it quickly gained access to the local market. In the
course of the dissolution of the New Australia Co-operative Settlement
Association, the stock was distributed in equal shares, and new settlers
from Britain and Australia brought additional capital. Among these new
arrivals were the Cadogans from Australia and two English families, the


danae m. perez-inofuentes

Smiths and the Kennedys. The growing community maintained contact

with other British subjects in Paraguay and engaged in British traditions
and cultural practices such as cricket, tennis, and football, and some New
Australians became members of the nota bene liberal Colorado party
of President Gonzalez. Through these activities, some of the New Australians soon belonged to the local upper class. In addition, a few young
New Australians fought against Bolivia in the Chaco War in 1932; Douglas
Kennedy, for instance, died in a military camp in Bolivia. Only a handful of New Australians left for Argentina, and private letters show that
those who did so kept in touch with the Anglo-Paraguayan community.
Unlike the people living on Colonia Cosme, the population in New Australia seemed to identify with their host country and assimilated readily
into Paraguayan society. Today, the entire district has over five thousand
Along with the communitys rapid integration into the Paraguayan society, the English language lost vitality as a communal means of communication. However, the colonists did try to establish English as the official
language of the community. Henry Connellys 1924 document confirms
that the community asked for English teachers and Anglican priests at the
British consulate, but their request was not granted. Consequently, the
children of New Australia were instructed in Spanish, and no Anglican
priest would christen them in English. In addition to these vain efforts, the
communitys geographic and demographic characteristics further advanced
the process of language shift. New Australia was not a tight-knit community but rather a group of independent ranches scattered over a vast estate.
Each party was busy with its own cattle and crops, and there was no central
village where the settlers could meet regularly. This individualism soon
ended in the breakup of the tradition of getting together at the galley to
have a chat or listen to the band playing music (Souter 1991: 133), and there
was no tradition of performing plays or reading poetry together. Given
these loose connections between the settlers, communication in English
did not take place on a daily basis, and the settlers were obliged to learn the
local languages to get along. The foundation of the village of Nueva Australia in 1942 (today Nueva Londres) was too late to revitalize the English
Further, due to the devastating effects of the War of the Triple Alliance
between 1865 and 1870, the majority of the Paraguayan population consisted
of women, children, and elderly people (Warren 1985: 1516). Most of
the New Australians, in contrast, were young bachelors. They soon took
Paraguayan wives; at times also more than one, and according to the padre

Anglo-Paraguayan English


from Nueva Londres, the majority of the districts inhabitants trace their
roots back to the Australian settlers. The Australian men told their wives
to avoid speaking Guarani with their children since it was considered
inferior. The childrens Paraguayan peers, however, only spoke Guarani,
and as soon as the New Australian children went to school, they were
exposed to all three languages at once and started to mix them. As in
Cosme, their fathers disapproved of language mixing and discouraged
them from speaking Guarani at all. Elderly descendants of the Kennedys
in Nueva Londres confirm that they learned English from their fathers,
Spanish from their mothers, and Guarani from their friends, and that
only the children of two English-speaking parents had a good command
of English. In the third generation of New Australians today, there is
no knowledge of English left, and their first language is Guarani. The
descendants of less wealthy families with lesser formal education prefer
Guarani to the extent that they need assistance to communicate in Spanish.
As a consequence, New Australians lost their heritage language and shifted
from English to Guarani within just two generations. Despite the founders
efforts to avoid it, Guarani is now the communitys principal means of
However, thanks to the social and financial advantages of certain families, above all the Smiths and the Kennedys, the fourth generation expresses
interest in revitalizing the English language by studying it at school and
university. The only remaining native speaker in Nueva Australia, an 86year-old descendant of the Smith family, after forgetting it nearly completely, is now happily able to have a chat in his mother tongue with one
of his granddaughters. It is therefore possible that the financial advantages of the descendants will enable the next generation to learn their
heritage language as an L2, as the Cosme descendants have done. Again,
the extent to which this revitalization is a matter of ethnic identity or
the status of English as a professional skill remains to be seen in future
To close this brief summary of the sociolinguistic history of the AngloParaguayan community, it is noteworthy to point out again that practically
all the rural descendants of New Australia, be it in Cosme or in Nueva Londres, prefer Guarani to Spanish and English. In other words, the remaining
descendants of the originally almost six hundred settlers not only shifted
from English to another prestigous local language Spanish, in this particular case as in the Anglo-Argentine case (Cortes-Conde 1996), but
to a local indigenous language. The Anglo-Paraguayan encounter hence
stands out as the first known case where a well-organized English-speaking


danae m. perez-inofuentes

immigrant community shifted from English to an indigenous language (see

Crystal 2000). However, as English did not disappear without a trace, the
next section describes some aspects and particularities of Anglo-Paraguayan

3 Features of Paraguayan English


The data

Prior to this research, no linguistic fieldwork has been conducted on

Paraguayan English (hereafter PAE). The description sketched for the purpose of this chapter is based on approximately one hundred minutes of
audio recordings in English. The six informants belong to the extended
Wood clan from Cosme and give an impression of how language shift has
advanced in Paraguay. About fifty minutes consist of interviews with three
second-generation men born between 1899 and 1910, who were recorded2
in the 1980s by Anne Whitehead in the course of her oral history project
Paradise Mislaid (1997). They lived in Paraguay all their lives except for time
they spent abroad as British soldiers in World War I. Another fifty minutes
were recorded with two third- and one fourth-generation speaker during
my fieldwork in Paraguay in 2011. One third-generation speaker was born
in 1935 in Cosme, where she acquired English natively from her parents.
She grew up in Asuncion and lived in Peru for a couple of years when she
was in her twenties and in the United States in her forties. The other thirdgeneration informant was born in 1947 on a British-run ranch. His first
languages were Guarani and Spanish, and he improved his limited nativeEnglish skills during an exchange year in the US. The fourth-generation
speaker, an agronomist born in 1980 in Asuncion, learned English at school
and lived in the US and Japan for a couple of months. The distinction
between these two groups of different generations belonging to the same
family is useful to gain an insight into the changes that PAE underwent
due to language contact and shift.
It is important to recall that the original group of settlers was very
heterogeneous. On the one hand, there were the two clearly distinct groups
of middle-class and working-class immigrants. On the other hand, New
Australia consisted of participants who came from different parts of the
English-speaking world. The principal families that stayed in Cosme after
its dissolution were the Woods and the McLeods from Queensland, and the

These recordings were presented by the Australian Broadcast Corporation (ABC) in 1990.

Anglo-Paraguayan English


Titilahs from Scotland. In the New Australia district, there were, among
others, the Caseys and the Murrays from Ireland, and the Cadogans from
Australia and the Smiths and the Kennedys from England. William Lane
himself came from Bristol, but as he had lived in the US, he was said to
have a slight Yankee twang (Souter 1991: 18). This mixture suggests that
the founders dialects may have varied considerably. However, Connellys
1924 document provides data on the first generation of Anglo-Paraguayans
(see note 1 above). Connelly was a sheep shearer from Queensland, and his
orthography and style give us a number of clues about the original settlers
vernacular speech. I make reference to this text in order to contribute
additional evidence to this description.


In what follows, I first provide an impressionistic account of the six informants speech by describing their use of Wells (1982) set of vowels. The first
items in Table 10.1 represent the more frequently used vowels in each case. I
then describe two consonant features that are highly distinctive in English
dialectology, rhoticity and intervocalic T. Finally, I point out some of the
contact-induced particularities of PAE to give a more detailed impression
of this unique trilingual setting.
The comparison of two different generational groups illustrates that
their accents differ considerably. The vowel system of second-generation
speakers of PAE is consistent except for the variation of the bath and
goose vowels. The variation of the bath vowel, however, concerns only
the item dance pronounced with //, whereas all other words, such as master
or half, are consistently pronounced with //. The regular use of diphthongs
indicates that second-generation PAE was non-rhotic. This assumption is
confirmed by Connellys manuscript, which contains words such as warter
water, conferdance confidence, or oppersition opposition. The spelling
with hypercorrect /r/ suggests that Connelly did not pronounce postvocalic
/r/. This observation is corroborated by Lanes novel, which reproduces the
workingmens speech with words such as pus purse. Finally, it is noticeable
that intervocalic T is always voiceless among second-generation speakers.
This consistent pronounciation of second-generation speakers contradicts
the assumption that the founders varieties were heterogeneous. Rather, the
data analyzed here suggest that the Cosme community spoke a consistent
variety of English close to British Standard English with little dialectal


danae m. perez-inofuentes
Table 10.1 Anglo-Paraguayan English vowel system

Group i
(2nd generation)

Group ii
(3rd and 4th generation)










Another feature of second-generation speakers can be added here.

According to one informant, his father used onomatopoeic expressions
when working with animals. One of the most frequent expressions to
attract the animals for whom certainly no Standard English was necessary was /kjx/. This indicates that some of the first-generation speakers
used velar fricatives in their vernacular variety of English, an assumption that can be supported by the actual pronunciation of the Scottish
family name Titilah as /titilak/. Adding this feature to the description of
first- and second-generation speakers outlined above, it seems reasonable to
assume that some of the founders of PAE spoke British English with certain
influences from different parts of the British Isles, such as Scotland. The fact

Anglo-Paraguayan English


that the second generation spoke a standard-like variety of British English

suggests that the first generation was actively committed to transmitting
Standard English to the next generation.
The speech of group II, in contrast, displays more variation. Most of the
variation occurs with vowels that do not exist in Spanish. For instance, the
opposition between the kit vowel and the fleece vowel is neutralized,
which makes homophones of minimal pairs such as live and leave. This
neutralization reflects the process of the shift of English as an L1 to L2.
Moreover, the predominant use of monophthongs with postvocalic /r/ is
indicative of rhoticity, which certainly originates from the speakers stays
in the US. Likewise, except for certain personal names such as Peter, which
were probably transmitted within the family, the speakers of group II
use voiced intervocalic T. In sum, whereas the speech of the first two
generations of PAE speakers was characterized by a consistent, standardlike British vowel system, non-rhotictiy, and voiceless intervocalic T, the
third and fourth generations display considerable variation in their vowel
system, use of postvocalic /r/, and voiced intervocalic T. These changes are
the result of language shift.
Apart from the features described above, a few other observations concerning language contact can be made. Given the limited set of final
consonants in Spanish and the particular aspiration of implosive /s/ in
Paraguayan Spanish, the speech of the third and fourth generations tends
to simplify final consonant clusters as in past pronounced as pass. The
final /s/ is frequently omitted, and words such as because are reduced to
/biko/. Further features that are particular to Hispanic English and occur
occasionally among the speakers of group II are the prosthetic /e/ in words
with initial consonant clusters beginning with /s/ as espanish, the bilabial
realization of the labiodental fricative /v/ as // as in have pronounced as
/h/, and the lack of aspiration after plosives as in times as /taimz/. As far
as interferences from Guarani are concerned, the most prominent feature
is the glottal stop before word-initial vowels as in English and self-esteem.
However, these features are only observed in the speech of the third and
fourth generations, which supports the claim that PAE has become an L2
On the basis of this sketch of the speech of three generations of the
Wood family from Cosme, we can observe that whereas the founders
dialect displayed many features of British English, the third and fourth
generations show a tendency towards North American English as well
as L2 varieties of English. This parallels the case of the Anglo-Argentine


danae m. perez-inofuentes

community, which has shifted to English as an L2 with many characteristics

of North American English among the youngest generation (Jefferies 2010).


The only features that lend themselves to morphological analysis at this

point of my research are verbal endings and agreement. The manuscript of
working-class New Australian Henry Connelly displays a high variation in
the present-tense paradigm, that is, final /s/ often occurs with all persons
such as women appeals to us and I loves them. This feature is also reproduced
in Lanes novel in direct speech such as I says and we puts (Miller 2007:
130) and indicates that some of the original settlers probably stemmed
from northern England (Trudgill 2010: 40). This assumption is, in fact,
consistent with the origin of the velar fricative /x/ mentioned above. The
fact that none of the audio recordings of second-generation speakers contains instances of generalized present-tense /s/ indicates, once again, that
the well-educated founders of Cosme spoke a more Standard-like variety of English with their offspring in spite of being speakers of different
vernacular dialects.
The speech of the third and fourth generations does not have one single
instance of regularized final /s/. There are, however, instances of irregular
verbal endings, be they from L2 acquisition or L1 influence. It has yet to be
determined if the omission of third-person /s/, as in that expression mean, is
due to the aspiration of final /s/ in Paraguayan Spanish or a typical feature
of an L2 variety of English.


Connellys written document gives evidence of the earliest Spanish and

Guarani items introduced in the PAE vocabulary. It contains a considerable
number of loanwords from both languages such as the term montie bush,
scrub stemming from Spanish monte, which soon became part of the
PAE vocabulary. Similarly, the item camp also reported in Falklands
Island English, although with the meaning of settlement (Britain and
Sudbury 2010: 219) is a calque from Spanish campo field and was used
in expressions such as in the camp in the field/country. Many words that
do have English counterparts appear in Spanish in Connellys document
such as stacion train station (noticeably hypercorrected without wordinitial /e/), informe report, or novillo calf. Also words referring to local
issues are typically Spanish and Guarani, for example, mate cocido mate

Anglo-Paraguayan English


[morning] tea, bombilla drinking straw, and lexical items that describe
the Paraguayan landscape as tuyu swamp.
At the same time, certain typical Australian words have survived over the
generations. Thus, the Anglo-Paraguayans use tucker (pronounced without
postvocalic /r/) and no worries conscient of the fact that these items identify
them as Australians. Last but not least, the descendants of New Australia are
proud of their family names since they distinguish them from the rest and
legitimize their identification with their heritage. Their pronounciation,
however, varies from the original English version, and Bates, Drakeford, or
Smith, for example, are realized as /bateh/, /raifo/, and /esmit/, respectively.

4 Conclusion
This chapter introduced a hitherto unknown English-speaking community
in Latin America, the Anglo-Paraguayan community, which had its origin
in the socialist New Australia project of William Lane. The communitys
extraordinary inception with the aim of setting up a white, Anglo-Saxon,
and socialist community in isolation offered ideal sociolinguistic conditions for a new English variety to arise. Once in Paraguay, however, their
English heritage language was exposed to the unique Paraguayan diglossia, which challenged its maintenance. After the settlement divided into
Nueva Australia and Cosme, each community evolved differently. Cosme
remained in isolation before it was deserted by the next generations. New
Australians, on the other hand, adapted quickly to their environment leaving their heritage language behind. Only vestiges of the English language
are left nowadays in the two villages.
The English language itself has undergone considerable changes in the
course of the twentieth century. The description of the most outstanding
features of PAE outlined here shows that English has become an L2 variety
in Paraguay. Despite the heterogeneous nature of the founders varieties,
the second generation from Cosme seemed to speak a relatively homogeneous, standard-like variety of English. The third and fourth generations,
in contrast, display many L2 features of English as well as considerable
L2 influence. This development allows the conclusion that PAE did not
survive as a native language in Paraguay, but it did establish itself as an
L2 variety among the urban descendants of Cosme. In Nueva Londres,
where it is non-existent today, it might reappear as a L2 among fourth- and
fifth-generation descendants.
The Paraguayan case of language shift is of special linguistic interest
and has great potential for further research. The fact that English as a


danae m. perez-inofuentes

global and prestigious language disappeared at the cost of a low-prestige

indigenous language makes it unique and begs for further investigations
in ethnolinguistics and the ecology of language (Haugen 1972). English
became a means for social mobility in Cosme, a community with limited
financial resources, and contributed significantly to the abandonment of
the village. In prospering Nueva Londres, on the other hand, English was of
no use for the second generation and disappeared quickly. A comparison of
the two communities is therefore interesting as it sheds light on the decisive
factors that underlie the process of language shift. Given that in this context
the community shifted from the more prestigious international language
to the indigenous local language, concepts such as English as a Killer
Language can be viewed in a more differentiated manner as claimed by
Mufwene (2008: ch. 12), and general assumptions on language prestige and
the instrumental value of languages may be rethought.

Britain, David and Andrea Sudbury. 2010. Falkland Islands English. In Daniel
Schreier, Peter Trudgill, Edgar W. Schneider and Jeffrey P. Williams, eds.,
Lesser-Known Varieties of English: An Introduction. Cambridge University
Press, 20923.
Cortes-Conde, Florencia. 1996. Is stable bilingualism possible in an immigrational
setting? The Anglo-Argentine case. In Ana Roca and John B. Jensen, eds.,
Spanish in Contact: Issues in Bilingualism. Somerville: Cascadilla Press, 11322.
Crystal, David. 2000. Language Death. Cambridge University Press.
Jefferies, Julian. 2010. Anglo-Argentine English. In Daniel Schreier, Peter Trudgill,
Edgar W. Schneider and Jeffrey P. Williams, eds., Lesser-Known Varieties of
English: An Introduction. Cambridge University Press, 195206.
Haugen, Einar. 1972. The Ecology of Language. Stanford University Press.
Kellett, John. 1997. William Lane and New Australia: a reassessment. Labour
History 72(2): 117.
Livermore, Harold V. 1950. New Australia. Hispanic American Historical Review
30(3): 290313.
Melia, Bartomeu. 2011. Historia de la lengua Guaran. In Ignacio Telesca (coord.),
Historia del Paraguay, 3rd edn. Asuncion: Taurus, 42551.
Miller, John Maurice [William Lane]. 2007 [1892]. The Workingmans Paradise.
Charleston: Bibliobazaar.
Mufwene, Salikoko. 2008. Language Evolution: Contact, Competition, and Change.
London: Continuum.
Souter, Gavin. 1991 [1968]. A Peculiar People: William Lanes Australian Utopians
in Paraguay. Brisbane: University of Queensland Press.
Trudgill, Peter. 2010. Investigations in Sociohistorical Linguistics: Stories of Colonisation and Contact. Cambridge University Press.

Anglo-Paraguayan English


Warren, Harris Gaylord. 1985. Rebirth of the Paraguayan Republic. The First Colorado Era, 18781904. University of Pittsburgh Press.
Wells J. C. 1982. Accents of English, 3 vols. Cambridge University Press.
Whitehead, Anne. 1997. Paradise Mislaid: In Search of the Australian Tribe of
Paraguay. Brisbane: Queensland University Press.
Zajicova, Lenka. 2009. El bilinguismo paraguayo: usos y actitudes hacia el guaran y
el castellano. Madrid: Iberoamericana.

ch a p ter 1 1

Gullah West
Texas Afro-Seminole Creole
Ian Hancock


The language of the Afro-Seminoles is an English-related creole whose

origins go back four hundred years to the west coast of Africa. Sharing its
origins with Sea Islands Creole (SIC, usually called Gullah or Geechee), it
continues to be spoken by a dwindling elderly population of fewer than
three hundred in south Texas, central Oklahoma and northern Mexico.1
Called Seminole (shim-i-no-li) in the Brackettville, Texas, community and
Mascogo in the sister community in Nacimiento, Coahuila, Afro-Seminole
Creole (ASC) dates to the time of separation from Florida, following that
territorys being sold to the United States in 1821.
Most of its living speakers in Texas were born in the closed environment
of the Fort Clark Indian Reservation; since 1917 their families have been
living in the town of Brackettville (Brackett) in close contact with other
Americans, and few people younger than about sixty-five are fluent in the
language. But because of the independence of the Afro-Seminoles, and
their earlier geographical and cultural isolation from the larger society,
their language has preserved far more of its original character than has
Sea Islands Creole, spoken on the Atlantic seaboard, to which it is closely

The beginnings of Guinea Coast Creole English

Afro-Seminole Creole is only one of a large group of related languages

spoken today in Africa, North and South America, and the Caribbean, all
of which descend in part from a common earlier ancestor, which we may
call Guinea Coast Creole English. This is no longer spoken itself, nor do

That Suzanne Romaine should claim (2001: 160) that the language is extinct is puzzling, since she
has not ever visited the community, and must have relied solely on my own publications for her
information about Afro-Seminole Creole.


Gullah West: Texas Afro-Seminole Creole


we have records of it, but on the basis of its modern descendants, and the
historical facts we have gathered, to some extent we are able to reconstruct
the circumstances of its origins, and even to know what it might have
sounded like.
There are records from 1553 onwards of groups of British sailors going to
live on the West African coast more or less permanently. Some of these were
criminals, some political exiles, and others were simply attracted by local
African life and preferred to stay rather than return to Europe. Whatever the
reasons, these sailors were all men, and nearly all between the ages of 15 and
30. They are referred to in modern writings as lancados, a Portuguese word
meaning men who were thrown from the ships Portuguese because the
first lancados, and indeed the first Europeans, to settle in West Africa were
Portuguese. The lancados we are concerned with in this case came from
all parts of Britain and spoke a great many different dialects of English
(standard English was still emerging so no one spoke that); there were no
radios or newspapers; literacy and schooling were privileges of the wealthy,
and contacts with speakers of other dialects were few. Those joining a
ship in port for the first time would have presented some problems of
communication to their new shipmates.
In the course of time, the sailors developed amongst themselves a kind
of English they could all understand. They did this by keeping those words
and constructions they had in common, and discarding whatever extreme
dialect forms might have hampered communication. This process, called
leveling, is something like creolization, except that because of the kinds of
speech involved, which were all dialects of the same language instead of
totally different languages, the common denominator level they reached
was far less different from the grammars of their different English dialects
than it would have been in a truly multilingual situation.
This leveled English was even more distinctive because it was used on
board ship, and as a result had a strong nautical flavor. Each sailor speaking
this Ship English could of course also speak his natural home dialect, though
if he did, it might have been difficult for his fellow crewmen to understand
him properly. When these sailors settled down on the Guinea Coast of West
Africa, they married African women and in the lancadoAfrican households
that they established, the Guinea Coast Creole English slowly developed.
This kind of social arrangement existed between about 1580 and 1630,
after which time the English started to get their slaves directly from Africa
instead of from the Dutch in the West Indies. When the English became
recognized as slavers, they were naturally no longer welcomed in the same
way, and had to build castles and forts to protect themselves if they wanted


ian hancock

to stay in West Africa. But by this time, a couple of generations of AfroEuropeans the first Creole people had grown up, and their language and
new society had become well established. The sailors continued to speak
Ship English, adding to it from their home dialects because they were now
no longer at sea, and adjusting it to the speech of their wives, who probably
spoke Serer, or Wolof, or Mandinka, or Temne, and who were also learning
to speak like their men. They kept the vocabulary of the mens speech even
though the Africans greatly outnumbered the Europeans mainly because
creolized English was useful all along the Guinea Coast from settlement
to settlement, while the African languages were spoken over fairly small
areas and made trading difficult outside the home area. There was also a
Portuguese creole spoken along the same coast, and it is still spoken today
in parts of West Africa. Even Africans from elsewhere who came to live and
work in the Creole communities (people called grumetes or grumettos or
laptots) learned to speak Creole, and since, in many cases, it was the Creoles
rather than the Europeans who kept the slaves imprisoned on the coast
before transportation, it was Creole, not English, which they also learnt
to speak. It was essential that they knew at least some of that language,
because whenever possible, slaves who spoke the same language were kept
apart from each other. Creole was all they had in common. The Africans
learnt Creole so that they could speak to each other, not so they could
speak to their captors.
In the early years of the slave trade, slaves were kept waiting on the
coast for a year or even longer before shipment across the Atlantic, and
then that voyage could last for many weeks. This gave the earliest arrivals
plenty of time to acquire a good knowledge of Creole, and even when
the volume and efficiency of the slave trade increased, so that the newly
arrived slaves would not have had time to learn it, they still learnt some
from the slaves they were put to work with once they reached the Americas. Since they also worked with indentured whites, especially in Barbados and North America, bondservants who were usually Scottish or
Irish and who spoke their own regional kinds of British English, and
since metropolitanization was already having an effect on the creole taken
across the Atlantic, it is probably safe to say that Black English, which
that speech has become today, never did have a wholly creole origin.
Because of the geographical isolation of the Sea Islands, and the comparative absence of whites there, and because of the continual (and illegal) arrival
of creole-speaking West Indian slaves in the area until scarcely more than
a century ago, and the fact that slaves arriving on the Atlantic seaboard
were not all sent to other parts of the United States, Gullah does not

Gullah West: Texas Afro-Seminole Creole


have the same history which has produced Black English elsewhere in the

3 Beginnings of Gullah
The British took most of their slaves to Barbados, which they settled in
1627, before distributing them to their other colonies. By 1795, well over
half of the c. 2,000 Africans in South Carolina, which was founded in
l670, were from Barbados, though after 1698 they were being brought in
more and more from Africa directly. South Carolina originally covered a
huge area, which even included much of what is today Florida. Georgia was
then Creek Indian country, and was considered to be free territory. When it
became a colony by charter in 1732, it immediately tried to prohibit slavery,
but because of pressure from South Carolina its attempt was unsuccessful.
Up until 1749, Georgia had been getting its own slaves from Carolina,
but after that date began to import them from elsewhere. Unlike Carolina, Georgia continued to bring slaves in from the West Indies, and
until a halt was drawn to the importation of West Indian and African
slaves in 1770, they were arriving from Jamaica, Antigua, Barbados,
St. Croix, St. Kitts, St. Martin, St. Vincent, Montserrat, Nevis, Martinique,
Guadeloupe, Grenada, and Cuba, a pattern of settlement quite different
from that in South Carolina.
It is possible that Barbados, because of its history and settlement, never
did develop its own dialect of Creole English or, if it did, it was a highly
anglicized variety. But the other islands had their own creoles, and slaves
from Nevis, St. Vincent, and elsewhere must have experienced little difficulty in communicating with each other on the North American plantations. Gullah appears to have grown out of a leveling of all of these. In
some respects, the process of leveling, which produced Gullah, was not
unlike the leveling which produced Ship English.
Gullah has characteristics found in several of the Caribbean island creoles, but it isnt exactly like any single one of them. It also has features
in common with Guinea Coast Creole brought in later with West African
slaves, and which do not turn up in Afro-Seminole, which separated from
the main body of Gullah speakers at an earlier time before that happened.
The West African creole most closely associated with Afro-Seminole Creole
is Krio, spoken in Sierra Leone and Gambia. Krio is a direct descendant
of Guinea Coast Creole English, but like Gullah is also something of
a hybrid, having overlays from Nigerian Pidgin, Jamaican Creole, and
almost certainly Gullah too, brought into Sierra Leone with the resettled


ian hancock

Nova Scotians. The picture gets complicated, for while there are reasons to
suspect that some of the Nova Scotians who arrived in 1787 spoke Gullah
(though no actual proof ), there were also Krio (and Mende and Vai) speakers coming into Carolina from Sierra Leone. Certainly Krio and Gullah
share far too many similarities for it to be merely coincidence.

4 Sources of the African Maroons

Numbers of black and Native American escapees from the English plantations in the seventeenth- and eighteenth-century Crown colonies of Carolina and Georgia were able to find refuge in Spanish Florida, where they
were allowed to establish autonomous communities around St. Augustine and where they were known as cimarrones, a word meaning, roughly,
fugitives.2 Already by 1821 there were 34 Seminole settlements in northern
Florida, three of which were African. According to Giddings (1858: 3), the
word Seminole was first used to refer to the black escapees into Florida,
and was only later applied by the Creeks to the Indian fugitives.
In 1817 General Andrew Jackson and his army were sent to northern
Florida to subdue the Seminoles and seize the land from Spain. They
killed livestock, burned crops, and destroyed the black forts along the
Apalachicola and Suwannee Rivers. At that time, the Seminoles, who
numbered an estimated 7,000, were welcomed by the Spanish government
since they served as a buffer between themselves and the English. A royal
decree from Spain dated 10 October 1699 promised protection:
. . . a todos los desertores negros de los ingleses que huyeron a San Augustin
y que se convirtieron al catolicismo
[to all Negro deserters from the English who fled to St, Augustine and
became Catholics]

Porter (1971: 164) writes of the Spaniards who were caught in Georgia and
imprisoned for enticing slaves to leave Carolina and go to Florida. Those
maroons did not, however, always join up with the Indian fugitives, though
some did, especially later when aggression from the north became more

Although the popular association of the word Seminole today is with the Indian population in
Florida, according to Giddings (1858: 3) it was first used to refer to the African escapees into that
region, and was only later applied by the Creeks to the Indian fugitives. Seminole has generally
been supposed to derive from a Native American word cima meaning a type of wild grass, but
more recently another etymology in the Arawak word smaran meaning bow and arrow has been
proposed by Jose Arrom (1986). The Indians themselves pronounced cimarron as cimalon or cimanol
transposing the m and the l.

Gullah West: Texas Afro-Seminole Creole


severe. The migration of Africans to the fort in Florida had stopped by the
mid 1760s:
Spanish power in Florida, moribund for a score of years, had been extinguished . . . the British were at last in control and runaway Negroes from
South Carolina and Georgia could no longer find refuge under the walls of
St. Augustine. (Porter 1971: 171)

This did not mean an end to the settlement of Africans in Florida; it

merely meant that the fugitives were establishing their own independent
communities separate from both the Indians and the Spaniards:
As late as 1774, blacks were apparently not living among the Seminole Indians. As slaves continued to escape from the American colonies, settlements
of blacks sprang up in Florida, but their relations with the Indians were not
always good. (Littlefield 1977: 5)

This was going on even into the early 1820s. In a letter written at that
time, Charles Pinckney (17571824), one of the drafters and signers of
the Constitution of the United States (Powers 1998), complained about the
numbers of slaves escaping from South Carolina into Florida, which by then
had become US territory. Thus the black maroons, or Afro-Seminoles, were
seeking refuge in Florida between about 1690 and the 1820s; that they were
mainly from Georgia during the earlier part of that nearly 140-year timespan, and that most Georgian slaves were West Indian rather than directly
African, supports the argument for a Caribbean origin of Afro-Seminole
Creole. One clue to the early makeup of the Afro-Seminole population is
provided by the words Joo and Joomaican, who are remembered as having
been present during the early period.
At the time that Florida became US territory, slavery was still legal,
and raids to capture free Africans (as well as Indians) created considerable
problems for Governor Jackson in his efforts to develop the new territory,
including further bloody conflicts; in December, 1835 Major Francis Dade
and his troops were ambushed by 300 Seminole warriors near Fort King
(Ocala), starting the Second Seminole War, an episode leading to the mass
removal of Seminoles to Indian Territory in what is today Oklahoma. By
1834, 3,824 Indians had been removed to the west. The war lasted until
1842, by which time 4,420 Seminoles had surrendered and been sent west.
From 1855 to 1858 the Third Seminole War (also known as the Bowlegs
War) took place, when Billy Bowlegs and his family were captured and
deported to Indian Territory. Only about 300 Seminoles almost all of
them Indians remained in Florida, where they had been granted five


ian hancock

million acres of land further south in the Everglades. The first Indian
Seminoles from British territory were Oconee people from Milledgeville,
Georgia, who moved into Florida in 1750, over half a century later than the
first African escapees. These were joined by the Muskogee (cf. Mascogo as
a Seminole ethnonym), and following them were the Apalchicola, Chiaha,
Hitichi, Sawokli, and Tamathli, all of whom lived in the Chattahoochee
River area in western Georgia, and all of whom spoke dialects of Hitichi.
In 1767 they were joined by the Maskogee-speaking Eufala from Alabama,
and in 1788 other Maskogee-speaking groups also joined them. Following
the Creek War in 181314, the number of Indian Seminoles tripled because
of new arrivals from Georgia and Alabama the Yuchi from Georgia, the
Alabama (from Alabama), the Yamassee, and the Apalachee. Today, the
Indian Seminoles in Florida speak two quite distinct languages, both of
them Mushkogean: Muskogee and Mikasuki. Groups of Back Seminoles
left Florida for other places as well; some went to the Bahamas (Wood
1980; Howard 2002), some were reportedly in Guanabacoa, Cuba, as early
as 1820, and others were invited to stay with the Cherokee. Still others
decided to remain in Florida.
4.1 Oklahoma
In 1849, some of the Oklahoma settlers applied to the Mexican government
for permission to go and live there, possibly because they believed they
would be more at home in a Hispanic environment and perhaps could
speak Spanish, but particularly because almost as soon as they had arrived
in Indian Territory, the US government declared them legally to be slaves,
while slavery had already been abolished in Mexico some twenty years
before. A group of about 500 Black and Indian Seminoles left Oklahoma in
the late fall of 1849, crossing Texas where they were joined by two hundred
Kickapoo Indians in the Brazos river valley near Waco, and crossing into
Coahuila, Mexico, in July 1850. At first the Black Seminoles settled in
Moral, not far from the Texas border, while the Indian Seminoles settled
separately at La Navaja and the Kickapoo at Guererro. Later the Black
Seminoles moved a hundred miles further into Mexico to Musquiz, soon
after that moving a few miles away to El Nacimiento de los Negros, with
a few families going instead to Matamoros. The Kickapoo moved to the
nearby colony of El Nacimiento de los Indios, but practically all of the
Indian Seminoles decided to return to Oklahoma (Opala 1980).
Slave raids continued even in Nacimiento, however, led mainly by
US Army Captain Warren Adams who was especially concerned with
recapturing slaves who had escaped from Texas; by now some 3,000 were

Gullah West: Texas Afro-Seminole Creole


living as fugitives in the Sierra Madre mountains. While the effects of these
raids hurt the Seminoles, much greater losses resulted from a smallpox
epidemic brought back from an encounter with the Comanches in 1857,
which left 74 people dead.
During their move west the Seminoles also encountered other Indian
languages such as Cherokee and Biloxi; in Mexico and Texas they interacted
with speakers of Kickapoo, Lipan, and other languages; the word ma:sko:ki
(Haas 1940:49, Loughridge 1964) is the Creek self-designation; people of
African descent are called (s)tilusti in that language. The Mascogos or Black
Seminoles today do not speak any Indian languages, although individuals
knowing some words and expressions were alive into the 1970s and have
been recorded (e.g. kokka-yenna where are you going? kwa-he (Im going)
home). But the fact that they were employed as interpreters for the US
Army a century before that is evidence enough that they were familiar with
various Native American tongues.


In Mexico, the Black Seminoles met another Creole-speaking group who

were already there. These were the Black Creek who, like the AfroSeminoles, were originally Africans, who had become acculturated to the
Indians they lived with without losing their creole language. They were the
Africans who lived with the Upper Creek in Georgia, and who had also
been sent west to Indian Territory. While the Afro-Seminoles, who lived
with the Lower Creek and others in Florida, left Tampa Bay by boat for
New Orleans and traveled to Indian Territory via the Mississippi River, the
Black Creek reached Oklahoma overland. They were brought to Coahuila
and left there by their Indian owners, who had been negotiating for land
for them since 1834. In addition to these two groups, the community was
also being joined by state-raised men and women escaping from slavery
in Texas via an underground railroad leading south into Mexico. Such families as the Gordons and the Shields descend from these fugitives. Although
members of the Brackettville and Nacimiento communities recognize their
various origins and are pretty well aware of which family is Black Creek and
which is Seminole or one of the smaller contributing groups, the commonest designation used by everybody, especially with outsiders, is Seminole.


In 1870 following negotiations with Mexico, the American government

sent US Cavalry Captain Franklin Perry to Nacimiento to recruit the
Seminoles, because of their reputation as fighters and because of their


ian hancock

familiarity with Native Americans, to come and help the US Army drive
the Plains tribes out of west Texas so that settlement there would be less of a
problem for the whites. The Seminoles agreed, and garrisoned themselves
under the leadership of General Bullis in Fort Duncan at Eagle Pass in
Maverick County, and Fort Clark at Brackettville in Kinney County, in
south Texas. They were successful, and continued to serve the United
States until they were discharged in 1914. For three more years they lived
on their own reservation at Fort Clark, but this was taken from them,
and since 1917 they have lived across the highway in Brackettville. Some
returned to Nacimiento, and others have gone to live in the neighboring
towns of Del Rio, Eagle Pass, Ozona, and elsewhere. Some even live now
in California, Missouri, and New York, and still make trips from time to
time to Brackettville at New Year, on Juneteenth, and for Seminole Day in
mid September.
The Seminoles were never informed of their rights as American Indians,
and later attempts to be included on the Seminole Register and to obtain
land of their own were ignored. As Woodhull says,
General Bullis was greatly honored, and his name and fame are held in
reverence by the people of the Southwest frontier, but his scouts have been
disbanded and their families have been moved off the Reservation at Fort
Clark. They are not entitled to consideration as Indians, because they did
not register under some provision of Congress, of which they knew nothing,
and they get no consideration as negroes. (1937: 127)

On September 16, 2007, a delegation from Oklahoma led by Representative Angela Molette (Tuscaloosa Ohoyo) officially confirmed the Black
Seminoles as the United Warrior Band of the Seminole Nation (one of the
so-called Five Civilized Tribes) in a ceremony in Brackettville, Texas, at
which Seminole Negro Indian Scout Association President William Warrior was sworn in as tribal chief.

5 Earlier work on the language

The existence of the Seminoles language was kept from outsiders until
1976. Joe Dillard, whose Black English was for many years the standard
work on African-American speech, reported that his field trip to Brackettville tended to confirm the notion that the dialect of the Black adults
is essentially that of Black English everywhere in the United States (1972:
182). And Kenneth Wiggins Porter, who had surely worked more closely
with the people over a period of thirty or more years than any other

Gullah West: Texas Afro-Seminole Creole


outsider, expressed embarrassment and surprise at having worked with

the Seminoles for so long without ever having learned of their language
(personal communication, July 1976). He had earlier called their speech
perfectly understandable English (Haynes 1976: 3); thus two short stanzas in what he called Seminole Speech were published in the Texas press
some years ago (Evans 1990: 7), but which in fact are written in an Uncle
Remus-like English using impressionistic orthography, rather than in Seminole, of which Porter by his own admission was unaware:
An den, ah wuz a trablin Ouah faith an prayuhs dey wuh too weak
Wid a hoss atween mah knees To sabe the body whole.
Back to Nacimiento Ol Satan stroy de libin man
To de springs and cypress trees! But he could not tetch de soul!

In actual Seminole, these verses would be

En den ah binnuh trabble We fait en preh dem bin dess too weak
Wit hoss tween me knee Fuh sabe de whole a we body
Back gen duh Nassum-yennuh Ole Sadant stroy de natchul man,
Duh de worruh en cypuss chree! But e nubbuh tetch e soul!

In his unpublished book on the life of John Horse, Porter (1947) likewise
includes several samples of what he considered to be Seminole Creole, thus
he has John Horse say:
I specs mebbe deys Injuns likes em almos as much as de wite folks! Suah
seems lak deys got mighty scace, anyhow . . . seems lak dey day aint no
gophuhs lef in dis whole country Ah kn lay mah hans on! . . . dat boy heah
again . . . de one wid de gophuhs.
Lessn de head ob dis-yeah snake catches onto he tail befo he cross de
ribbuh dis trick aint gwine tuh wuhk! Lawdy lawdy! But dis am one smaht
trick! To mahch we across behin dat camp, across de ribbuh, back aroun,
an obuh again! Smaht as dat fiah-trick yestiddy and not such hahd wuhk!
Wonduh who tought dem up? Abraham? Dat ole John Caesar? Mebbe
Jumpuh? Osceola hissef? Whoebbuh it wuh, he mighty smaht man smaht
lak Ahd lak to be. Ah reckon dem wite folksll stay in dey camp now
lessn dey cides go on back wheh dey come f om an wheh dey belongs!
Lawdy lawdy! Wisht I had one ob dem fah-seein things so Ah cud see de
faces ob dem ossifuhs as we mahch past!

This is remarkable, in light of the fact that Dr. Porter lived with the
Seminoles, on and off, for over thirty years. Samples of Seminole Creole in
Sivad (1993: passim) are likewise impressionistic and remote linguistically
from actual Seminole speech, though based on published accounts of the


ian hancock
Table 11.1 Articles in
Afro-Seminole Creole
one, uh
dem, de

a, an
the singular
the plural

Ah lib don deah til ah wuz ninten, and ah don niva go bock nuh
mo. Too hod don deah.
Duh youngun dem nuh lika we way; dem nuh lika we duh talk lika
Duh tarm we gib you a dese: you lay down arm and stop de war;
you sojas go back an stay in der fots; we Indyen cross ober duh
Ouitaloochie [River]; an from dis time fort for ebber affer, we make
de Grand Ribber duh line o boundary atween de two. We promise lib
in peace and good tarm wi all white neighbor. Dat all got say.
An wuh fuh we submit? We not conquered! We whup you people one,
two, tree time. We whup you, damn, we keel you well too. Mek so
[why] e submit? We com heah gib conditions, not askum.

Features of the language


Sound system

Afro-Seminole Creole is not identical in features of its sound system

with Sea Islands Creole. For instance, it lacks the non-English allophones
described by Turner (1949) for SIC such as [], [], [], [], and others. It
does, however, evidence the palatalization of initial velars in such words as
gyal girl, gyaadn garden, gyaalic garlic, and kyandle candle. As well,
the articulation of <oi> as [ai] in such words as nize, spile spoil, jine
join, and pizn poison points to eighteenth-century English phonological inputs to ASC.
6.2 Grammar
6.2.1 Nouns
Table 11.1 lists the articles used in ASC. Nouns do not usually change for
plural by adding an -s at the end as in English; a few words like day and

Gullah West: Texas Afro-Seminole Creole


Table 11.2 Subject pronouns

in Afro-Seminole Creole
ah, me
you, hunnuh
e, him
hunnuh, yall

you singular
he, she, it

ting sometimes take a final -s, and the word chile has its own plural chirren,
but the usual way to show that there is more than one of anything is to
follow the word with dem:

De man-dem wey binnuh wuck dey

The men who were working there

(2) Ah en talk tuh me frien dem

Im going to talk to my friends
If there is a number before the noun, then the dem is not necessary:
(3) Fo uh me frien
Four of my friends
The same word dem, when placed after someones name, means that person
and his family or group of friends:
(4) We duh gwen siddung long wit Louis-dem
Were going to sit with Louis and his family (or Louis and his
(5) Kay-Kay-dem done eat up all we tettuhpoon
Kay-Kay and her friends have eaten all our sweet potato pudding
6.2.2 Pronouns
The word for I is nearly always Ah, but me is sometimes used in emphatic
constructions, and before negative nuh, especially in the expression me nuh
know I dont know. E is also the commonest word for he or she or
it, but him is used for emphasis very frequently. (See Tables 11.2 and
11.3 for subject and object pronouns.)


ian hancock
Table 11.3 Object pronouns
in Afro-Seminole Creole
you, hunnuh
um, rum
hunnuh, yall

him, her, it

The form of um [m] with an r, i.e. rum [rm], is only used when the
word before it ends in certain vowels; this is the same in West African Krio
and in Sea Islands Gullah:

Gam ([gm] = gi um) tuh rum

Give it to her


Ah cyan membuh rum

I cant remember it

6.2.3 Possession
Possession is shown through juxtaposition of nouns without any overt
marking on either the head or dependent.

Pompey dahdy
Pompeys father


John Horse hoss

John Horses horse

(10) Me ahnty neighbuh Toyota

My aunts neighbours Toyota
Hunnuh is only a plural pronoun in most related creoles, but this
is not the case in ASC. Possessive pronouns (and demonstratives) (see
Tables 11.4 and 11.5) go before adjectives. If the possessive pronoun comes
at the end of a sentence, it is followed by own:

Darra-dey cah duh we own

That car is ours

(12) E nuh look lukkuh e own

It doesnt look like his/hers

Gullah West: Texas Afro-Seminole Creole


Table 11.4 Possessive pronouns in

Afro-Seminole Creole
me, muh, my
you, yuh, hunnuh
hunnuh, yall

you singular
his, her, its
his, her, its, emphatic

Table 11.5 Demonstratives in

Afro-Seminole Creole

this, close by
that (dat when emphatic)
those, far away

The same word own can go with a few other words too:
(13) Duh who-dat own?
Whose is it?
(14) Dishyuh mus be somebawdy own
This must be somebodys
6.2.4 Prenominal modifiers
ASC like other anglophone creoles places modifiers before nouns in noun
phrases. Some groups of words that can go before nouns are delimited as
follows. Adjectives
These behave just like verbs, except that without a tense or aspect marker
they still can have a present tense. Its hard to think of adjectives having
tenses, but its one way to explain the difference between dis leaf yalluh
and dis leaf duh yalluh; the first one means this leaf is yellow, a kind of
permanent state which includes the present since it is yellow while you
make the observation about it; the second one has duh which is the aspect


ian hancock

word for action in progress, so it would mean this leaf is getting yellow,
or this leaf is yellowing.
Adjectives can be used with the other tense and aspect markers too, just
like verbs.

a. Dis leaf en yalluh

This leaf will be yellow
b. Dis leaf ennuh yalluh
This leaf is going to turn yellow
c. Dis leaf bin yalluh
This leaf was yellow
d. Dis leaf binnuh yalluh
This leaf was turning yellow
e. Dis leaf done yalluh
This leaf has turned yellow

Adjectives are made comparative by using the word mo in front of them,

or if they are just short words, by adding -uh to them. Sometimes both mo
and -uh are used together. The word for than is nuh:
(16) a. You ogliuh nuh me
Youre uglier than I am
b. You mo ogliuh nuh me
Youre uglier than I am
c. You mo tankful nuh me
Youre more thankful than I am
They are made superlative by using the word mos in front of them, or
if they are just short words, by adding -is to the end. Sometimes both mos
and -is are used together:
(17) a. You duh de odis ooman
Youre the oldest woman
b. You duh de mos odis ooman
Youre the oldest woman
c. You duh de mos tankful man
Youre the most thankful man
6.2.5 Verbs and tense and aspect
In English, the different forms of a verb are shown by adding different
endings to it, e.g. walk, walks, walked, walking. In Seminole, the basic
form of the verb does not change like this. Instead of adding endings to it,

Gullah West: Texas Afro-Seminole Creole


separate words are placed in front of it. This is typical of creole languages
In creoles, and therefore in Seminole, more importance is attached to
the nature of something happening than to the actual time it happens. In
Seminole there are two words to express the nature of the action (called its
aspect) and two words to express the time of the action (its tense). These
can be combined with each other to make more complicated structures.
The aspect markers in ASC are duh, uh3 that indicates an action is in
progress or happens usually or habitually, and done that indicates an action
is completed.
It is important to remember that these have no reference to time. Duh
can be used with the tense words to indicate continuous action in the past
or future, and done can refer to something that will be completed at some
time in the future that hasnt even begun yet, or else was completed before
some time in the past, also by being used with the tense words.
Tense in ASC is marked by preverbal forms bin, which indicates an
action before now, and en/gwen, which indicates an action in the future.
If action now (i.e. in the present) is expressed, it is done so with duh
since if the time of the action is now, that action must be in the process
of happning. When the verb alone is used, the time it refers to is past.
This is not true of some verbs, which refer to actions, which seem to be
independent of time, like know or want.
As in other creole varieties, the preverbal tense and aspect markers can
combine to create more complex and/or finely tuned distinctions. The
different combinations and meanings of these can be understood more
easily in the following examples.
i. With no tense or aspect markers:
(18) a. Ah chry fuh do um
I tried to do it
b. Ah tell de man dis mornin
I told the man this morning
c. Molly joog me good wit e pin
Molly stuck me hard with her pin
d. Ah know how fuh shet um
I know how to shut it
e. Wuh else yuh wan?
What else do you want?

Uh is the form of duh that is used after the tense word bin, listed below (binnuh = bin duh).


ian hancock

When the verb alone comes after the word fuh, and there is no subject
pronoun, the fuh means to:
(19) Ah bin too bex fuh talk tuh rum
I was too angry to speak to her
When a subject pronoun comes before fuh and a verb, then fuh means
must or should:
(20) Ah fuh talk tuh rum
I should talk to her
(21) Ah bin fuh talk tuh rum
I should have talked to her
(22) Hunnuh nuh fuh jrink da worruh
You mustnt drink that water
ii. With tense marker bin:
(23) Meck e bin churray um?
Why did he throw it away?
(24) Dem bin nyus fuh talk Simanole
They used to speak Seminole
(25) Dem bin pit e dahdy een jail4
They put his father in jail
(26) Dem wale me
I was beaten; they beat me
(27) Dem tief e car
her car got stolen; they stole her car
iii. With tense marker en:
The future word en has several other forms, such as gwen, gwine, ennuh,
gwunnuh, and so on. The pronunciation without the g- seems to be the
most common, and probably existed in the creoles from very early on. In
Trinidad Creole the future word go has another form oh, and in Saramaccan
Creole spoken in South America, the only form now is oh. Even in American
Black English, Im gonna do it has the variant pronunciation Im onna
do it and even Im uh do it.

This sentence would also be the translation of his father was put in jail, because there is no passive
in Seminole.

Gullah West: Texas Afro-Seminole Creole


(28) Hunnuh en fin we deh

You will find us there
(29) De sperrit-dem en kyah youway
The spirits will carry you off
(30) Hunnuh gwine dead too
You-all will die too
iv. With aspect marker duh:
(31) Ah duh chry fuh do um
I am trying to do it
(32) Molly duh cratch e so foot
Molly is scratching her sore leg
(33) Ah fuh duh talk tuh rum
I should be talking to her
(34) Dem duh jouk um
Theyre teasing him
(35) Him duh go tuh school
He is going to school
Notice that in the last example, he is going to school can have two
different meanings, as in English. It can be the answer to where is that
boy on his way to now? and also to "what is he doing these days? Some
creole languages have different constructions for each of these.
v. With aspect marker done:
(36) You done bruck um fuh chrue now
Youve really broken it now!
(37) Ah done tiyah fuh read
I have become tired of reading
Some verbs used with done can be translated with become as well,
when there is no object following.
(38) E done fix
It has become fixed
(39) E done cook
It has become cooked


ian hancock

Compare the forms in (38) and (39) with those in (40) and (41).
(40) E done fix um
He has fixed it
(41) E done cook da poke
He has cooked that pork
vi. Bin with done and duh:5
(42) All me peepil binnuh talk um
All my people used to speak it
(43) Dem binnuh shout een de chuch
They were singing in the church
(44) E bin done tell me bout you befo
She had told me about you before
vii. En with done and duh:
(45) By dis time tumorra hunnuh en done spen two whole week yuh
By this time tomorrow, youll have spent two whole weeks here
(46) Hunnuh en uh see me, nuh worry
Youll be seeing me, dont worry
6.2.6 Auxiliaries
Seminole has taken some other auxiliaries from English, including must,
could, and would, and their combinations mussa, coulda, and woulda.
(47) Ah nuh bin know seh ah could do um
I didnt know I could do it
(48) Ah shonuff would like fuh go too
Id sure enough like to go too
(49) Ah coulda tell you dat easy
I could have easily told you that
(50) E woulda spile fuh chrue
It would really have spoiled
Two other verbs with characteristic pronunciations in Seminole are ha
(have) and leh (let):

Bin duh is usually pronounced binnuh in ordinary speech

Gullah West: Texas Afro-Seminole Creole



Dem chillen nuh ha nuttin fuh do

Those children have nothing to do

(52) E ha fuh git back fuh school

She has to get back to school
(53) Leh we go, bubbuh!
Lets go, sonny!
6.2.7 The be verb
Be here covers all the different forms of that verb is, am, are, was, were,
being, and so on. In Seminole, there are different ways of saying this.
i. Be between nouns is duh in the present tense, (gw)en be in the future,
and binnuh in the past:
(54) Him duh de odes one aroun yeh
Hes the oldest one around here
(55) Mr. Toughtry bin duh lyer
Mr. Toughtry was a lawyer
(56) E bin wan fuh be lyer
He wanted to be a lawyer
(57) Duh da e en be
Thats what hes going to be
Duh is also used as a highlighter when certain words in a sentence need
to be emphasized. It this case, they come at the beginning as in (58ab) or
with question words as in (59ac).
(58) a. We wan talk tuh John
We want to speak to John
b. Duh John we wan talk tuh
Its John we want to talk to
(59) a. Duh wisseh hunnuh duh gwine?
Where are you going?
b. Duh who-dat bin call me name?
Who called my name?
c. Duh wuh dem bin tell hunnuh?
What did they tell you?
Unlike the other creoles (except SIC), Seminole grammar will not allow
for verbs to be brought forward in the same way; both West African Krio


ian hancock

and Jamaican Creole can say duh buy you buy im or duh tief you tief im?
(did you buy it or did you steal it), but in Seminole it would have to be
you buy um or you tief um?
ii. Be in the sense of exist or be in a place (like Spanish estar), is dey:
(60) Hunnuh book dem dey pun da cheer
Your books are on that chair
(61) Muskittuh bin dey ebbawey
Mosquitoes were everywhere
(62) Ah en dey een me room
Ill be in my room
6.2.8 Negatives
There are many examples of negative sentences in the earlier pages. Negative
constructions are usually made by putting nuh (or no or nah) right after
the subject noun or pronoun:
(63) Me nuh sabby um
I dont know him
(64) Me oncle nuh know
My uncle doesnt know
(65) Me ahnty nuh bin wan fuh know
My aunt didnt want to know
(66) En ah n en tell um
And Im not going to tell her
When a sentence has two parts, i.e. a subject and an object, both are made
negative, so it is correct Seminole grammar to say we nuh see nuhbawdy en
we nuh bin eat nuttin, we didnt see anyone and we didnt eat anything.
The verbs could, would, coulda, woulda, and kin (can) dont have
negatives with nuh; the negative forms of these verbs are couldn, wouldn,
couldna, wouldna, and cahn or cyahn.
The aspect marker done, when made negative, is not nuh done but
nabbuh: E nabbuh shem he hadnt seen her.
6.2.9 Conjoined utterances
Words and sentences can be joined together in different ways to make
longer, more complicated constructions. Sometimes two complete sentences can be put side-by-side with a joining word, sometimes a sentence

Gullah West: Texas Afro-Seminole Creole


can be put inside another sentence, and sometimes a sentence can be put
after a noun or a verb. When sentences are joined in any of these ways,
they need joining words.
i. A sentence following another sentence:
(67) Me duh gwine en you fuh tay yuh
I am going and you must stay here
(68) Josie wan leff um dey buh you wan fuh teck um wit you
Josie wants to leave it there but you want to take it with you
(69) Ah en eat now, been you nen dey home befo six
I will eat now, since youre not going to be home before six
(70) E say e nuh know how e en fine room fuh e seddown, nummuh e
en seddown somewey
She said she didnt know how shed find room to sit, except that
she was going to sit down somewhere
ii. Following a noun:
(71) Dishuh yaze wey de doctor bin fix still nuh right
This ear that the doctor fixed still isnt right
(72) Dem piece uh ood wey dey onneet da stove en ketch fire ef you
nuh moobe um
Those bits of wood that are under that stove will catch fire if you
dont move them
iii. Following a verb or an adjective:
(73) Duh chrue seh all two de man drown?
Is it true that both men drowned?
(74) Ah bin yeddy seh duh lie
I heard that it was a lie
(75) E ax um seh duh wuh you wan?
He asked him what is it that you want?
6.2.10 More about fuh
Two different uses of the word fuh have been given already, namely as the
indicator of a verb when it has no subject (to run, to jump, to eat, etc.) and
as a word meaning must or should when the verb does have a subject
(I must run, I must jump, etc.): fuh run, ah fuh run; fuh jowmp, ah


ian hancock

fuh jowmp, etc. Fuh can be used this way by itself, as in these examples, or
together with ha (have), to give haffuh: Ah haffuh run.
Together with bin, bin fuh means should have, as in dem bin fuh go
they should have gone, but sometimes it is mistakenly used as though it
meant the same thing as bin duh (binnuh).
In sentences of the kind something for you to eat, or a song for them
to sing, which need a to before the verbs in English, there is no need to
use fuh:
(76) a. Sometin fuh you eat
Something for you to eat
b. One song fuh you sing
A song for you to sing
It is also not necessary to use fuh after wa (want):
(77) Ah wan go
I want to go
6.2.11 Dey pun and studdeh
Dey pun means to be engaged in some action, as in e dey pun fool, hes
acting the fool (at this time). The word studdeh can also have a similar
meaning, and signifies that the action of the verb is repetitive or continuous:
E studdeh binnuh watch de gyal he was steadily watching the girl.
6.2.12 Adverbs
Regular adverbs have the same form as the adjectives they are related to; it
is only their position within an utterance that makes them adverbs.
(78) a. De poodie gyal duh sing
The pretty girl is singing
b. De gyal duh sing poodie
The girl is singing prettily
Some other adverbs, not derived from adjectives, are shown in
Table 11.6.
Adverbs are also whole phrases which tell you how, why, where or when.
(79) a.

Behime de do
Tru de do-mout
Puntop we roof
Wit e pent-bresh
Kezz e bin wan fuh
Nice de winduh

Behind the door

Through the doorway
On our roof
With his paint-brush
Because he wanted to
Near the window

Gullah West: Texas Afro-Seminole Creole


Table 11.6 Non-derived adverbs in

Afro-Seminole Creole
meck, meck-so
wisseh, wey
wuh-time, win
soon een de monin

the other day
early in the morning



Tags are little words you stick on the end of a sentence to give it a particular
tone. Two common tags in Seminole are enty and nuh:
(80) Nuffuh peepil bin deh dey, enty?
Plenty of people were there, werent there?
(81) Dem en come back, enty?
Theyll come back, wont they?
Enty can also come at the front of a sentence as in example (82) or used
in combination with nuh as in example (83).
(82) Enty dem yie bin shet?
Werent their eyes closed?
(83) Nuh loss um, nuh
Dont lose it, will you


Aspects of the structure and lexicon of ASC have been described elsewhere
(Hancock 1977, 1980a, 1980b, 1986, 1993, 1998), but a few characteristics
of the language may be noted here. ASC lacks much of the Mende and
other African-derived lexicon found in SIC. Nevertheless, ASC contains
about forty words of African provenance, some half of which are traceable
to KiKongo/KiMbundu, the balance to languages of the Guinea coast.
On the other hand it has a number of words of American Indian and
Spanish origin not found in SIC. Some are given here; further examples
and discussion are found in Hancock (1998).
With matches in Bantu: oolah bedbug, pingy cooking pot, cootie
stunted pig, teemuh dig a hole, zoondoo a hammer.


ian hancock

With matches in Twi: Cuffy male given name, Cudjo male given
name, kunkie a tamal.
With matches in Upper Guinea languages: boontuh buttocks,
chikka-bode teeter-totter, tabby mud daub, chooklah girlfriend,
ninny breasts.
The English items match in the main those found in other anglophone
Atlantic creoles, pointing to both place (southwestern England) and time
(the eighteenth century).
From SW English dialects: weekaday weekday, mole fontanelle,
yeddy hear, leff leave, broke break, loss lose, ees yeast, ood
wood (see Hancock 1994).
From Scottish English dialects: pit put, snoot snout, wurrum
worm, graytuh grate, bresh brush.
From Spanish: banyuh wash, kwahah make cheese, matatty
grindstone, soakettuh mud, beeoleen violin, calpintero
woodpecker, treego rice, huckle adobe hut.
From Native American languages: suffki corn porridge, stammal
ground corn, poleyjo hominy, polijotee a corn-based drink.
There are words for which no etymology has so far been found, such as
babba carry on the back, or skiffy vagina (though cf. Krio bamba carry
on the back and Bahamian Creole skiff young woman).

Similarities with other creoles

It has been argued by Haynes (1976) that ASC is not a creole at all; by Leap
that it is Indian English (referred to in Haynes 1976); and by Drechsel
(1976) that it is relexified Mobilian Yama (a Choctaw-based pidgin). The
latter espoused the general polygeneticist argument that it is a product
of local origin and development and not the result of diffusion from a
common anglophone creole base. But it is a creole and, predictably, most
like SIC. It differs from that language in a number of ways, probably due to
retention of features lost or obsolescent in the latter, where the postnominal
plural dem for example now functions as an & Co. marker only: John
dem John and his family/group but additionally in Seminole book dem
Early SIC texts show no as a preverbal negator, and this is the only means
of negating in ASC: E nuh shem he didnt see her. Sea Islands Creole now
generally negates with aint (e ain shum). This hasnt happened in ASC
since the future marker here is en (< gwen < gwine < going). Similarly,
nuh hasnt become ain in Jamaican Creole either, as it has in Trinidad and

Gullah West: Texas Afro-Seminole Creole


elsewhere, because in Jamaican the past tense word ben (bin in Seminole)
has become en.
As in archaic Jamaican Creole (but not recorded for SIC), there are the
forms warrah and darrah for what and that; like the Caribbean but
not the African anglophone creoles, ASC has the construction Ah gi um
worruh fuh e jrink (I gave him water (for him) to drink, but cf. Krio Ah
gi am watta foh (leh e) drink).
Atlantic anglophone lexical items widely found in related creoles include
lukka like, as, nuff plenty of, nummuh only, shoes shoe, yaze ear,
teet tooth, wisseh where, do-mout doorway, big-yie envious, cutyie give a glance of anger, yie-worruh tears, moon menses, day-clean
daybreak, han arm and hand, foot leg and foot.

8 State of the language today

Unlike SIC, whose conservative forms are becoming lost due to anglicization, ASC is disappearing because it is not being transmitted to subsequent
generations. Nor will its speakers reveal their knowledge of ASC to casual
outside enquiry; it is a very private language. While standing behind a
bench in the cemetery at a dedication ceremony some time ago, I heard a
lady some distance away hailing a friend who was sitting right in front of
me, in English. When she reached her, she stooped to kiss her cheek, and
repeated the greeting in her ear, this time in Seminole.
The oldest fluent speakers are now in their sixties, and while younger
people can understand much of it, they cannot reply in the same language.
Children can neither speak nor understand it. Influence from English
is evident in the Texas community, though not in Nacimiento, where
Spanish is now the main language of the village. But more African words
are remembered and used there, and older pronunciations of some items,
such as choo-eh spill, now trowway (< throw away) in Brackettville, and
the name of the language itself. In both communities, speakers claim that
their parents and grandparents spoke an even more remote variety, which
even they had trouble understanding. Interference from English is evident
in the texts, below, where plural -s is evident in respecks and days.
There is ongoing discussion of seeking funding to establish an annual
summer school in Brackettville, in order to teach the language and history
to the present generation.

Appendix 1: Texts
Goot eebnin, goot eebnin, how hunnuh, how hunnuh duh do? Ah hope
hunnuh duh do fine, Ah duh do awright . . . now ah des come yuh dis


ian hancock

eebnin fuh see how hunnuh duh do en fuh pay me respecks tuh de dead.
Da duh me frien Silas Hall, en e dead yeah, e dead, you know e bin fuh
dead. Now nuh lahf, nuh lahf, hunnuh en die too, hunnuh en dead one
o dese days, see de man right deh? E duh dead; ef you nuh believe me
go obuh deh en joog um, e nentannup, e nen hulluh, e nen say nuttin.
Hunnuh gwine dead too!
Sistuh Phyllis, duh one ting ah wan fuh know; duh wisseh you bin git da ole
deep baid out deh fum? Dole baid so deep e jeh like uh man fuh leddown
een e grabe, e ha fuh stan up fuh tun obuh een um!

All a we binnuh lib yuh fum de fus time da Cunnel McKenzie call we
fuh come. Hunnuh mussuh bin yeddy seh duh we wey done clay-out dem
Injin; whey-ever dem bin dey, duh we bin clay dem off dis lahn.
One ole king bin yeddy seh dem bin fuh pit one mo nyunger king een
e place, en e biggin fuh sorry bout um. E wan fuh keep e trone, so e call e
sojuh-dem fuh go kill all de peepil een de nation, fuh nuh leh nubbawdy
nuh dey fuh show de nyew king how fuh do e wuck. So de sojuh-dem teck
dem gun en dem go kill all de peepil. Nubbawdy nuh bin leff fuh show
de nyung king how fuh do e job. Atuh dat de fus king sen fuh de odduh
one seh him fuh come tuh him house eebnin time, en e fuh bring one fat
bohog long wit um. Time de nyung king yeddy dis e sen answer back suh
steaduh him fuh go to de fus king house, him fuh come tuh him house,
en e en gi um de bohog fuh present.

Appendix 2: Seminole history falsely represented at Fort Clark

A document and pictorial display entitled Area attractions, in the lobby
of the Fort Clark Springs Hotel (which used to be Fort Clark itself ), and
which is also displayed publicly in the town of Brackettville, contains the
following information for visitors which both distorts and trivializes the
Seminole contribution to Texan history:
SEMINOLE SCOUTS: Serving during the frontier era, the scouts were
the descendants of slaves stolen from Southern plantations by Florida
Seminoles. The US Government hired 150 as scouts for the Army to trail
hostile Indians of the Southwest. A group settled in Brackettville around
Fort Clark, and their descendants remain as area farmers and ranchers. The
old cemetery is on a country road about three miles south of town.

While it is true that the ancestors of the Seminoles were stolen by white
people from Africa, they were of course not stolen by Indians from the
plantations but escaped voluntarily to join them in Florida, themselves

Gullah West: Texas Afro-Seminole Creole


becoming Seminoles, a word simply meaning fugitives. As stated above,

the name Seminole was in fact applied to Africans many years before it
was applied to the Indian escapees. The fact that they created their own
maroon (i.e. non-slave) society is what makes the Seminoles unique among
African American populations in the United States, and what qualified
them to participate in the 1992 Smithsonian Institutions Maroon Festival
on the Mall in Washington, DC. Nor did their ancestors simply settle
in Brackettville around Fort Clark but were specifically invited to serve
as scouts for the Army out of Eagle Pass and Brackettville by the US

Arrom, Jose Juan. 1986. Cimarron: apuntes sobre sus primeras documentaciones y su
probable origen. Serie monografica no. 18. Santo Domingo (DR): Ediciones
Fundacion Garca-Arevalo.
Dillard, Joe. 1972. Black English: Its History and Usage in the United States. New
York: Random House.
Drechsel, Emanuel. 1976. Pidginization and creolization in North American Indian
languages: Mobilian Jargon and Afro-Seminole Creole. Unpublished report
to the National Science Foundation.
Evans, Christopher. 1990. A scouts honor. The Fort Worth Star Telegram, March
25, pp. 78.
Giddings, Joshua R. 1858. The Exiles of Florida: Columbus: Follett. Reissued Baltimore: Black Classic Press, 1997.
Gonzales, Ambrose E. 1922. The Black Border: Gullah Stories of the Carolina Coast.
Columbia: The State Co.
Haas, Mary. 1940. Creek vocabulary. Unpublished mauscript.
Hancock, Ian. 1975. Creole features in the Afro-Seminole speech of Brackettville
Texas. Caribbean Linguistic Society Occasional Paper, no. 3.
1977. Further observations on Afro-Seminole Creole. Caribbean Linguistic Society Occasional Paper, no. 7.
1980a. The Texas Seminoles and their language. Working paper of the AfroAmerican Studies and Research Center of the University of Texas at Austin,
1980b. Gullah in Texas. In Joe Dillard, ed., Perspectives on American English.
The Hague: Mouton, 30533.
1986. On the classification of Afro-Seminole Creole. In Michael Montgomery
and Guy Bailey, eds., Language Variety in the South: Perspectives in Black and
White. Tuscaloosa: Alabama University Press, 85101.
1993. Mortars and metates. In Peter Seitel, ed., Festival of American Folklife.
Washington, DC: Smithsonian Institution, 5961.
1994. Componentiality and the creole matrix: the south-west English contribution. In Montgomery, ed., 94114.


ian hancock

1998. History through words: Afro-Seminole lexicography. In L. Fiet and

J. Becerra, eds., Caribbean 2000: Identities and Cultures. Rockefeller Foundation Publication, University of Puerto Rico, 87104.
Haynes, Lilith. 1976. Candid chimaera: Texas Seminole. Term paper, Department
of English, New Mexico State University, Las Cruces.
Howard, Rosalyn. 2002. Black Seminoles in the Bahamas. Gainesville: University
Press of Florida.
Littlefield, Daniel F. 1977. Africans and Seminoles. Westport, CT, and London:
Greenwood Press.
Loughridge, R. M. 1964. English and Muskokee Dictionary. Okmulgee: Baptist
Home Mission Board.
Montgomery, Michael. ed. 1994. The Crucible of Carolina: Essays in the Development of Gullah Language and Culture. Athens, GA, and London: University
of Georgia Press.
Opala, Joseph. 1980. A brief history of the Seminole freedmen. Occasional paper
no. 3 of the African and Afro-American Studies and Research Center, Austin:
University of Texas.
Porter, Kenneth Wiggins. 1947. Freedom over me: the story of John Horse (Gopher
John, c. 18121882) Seminole Negro chief and his people in Florida, the
Indian Territory, Mexico and Texas. Unpublished typescript.
1971. The Negro on the American Frontier. New York: Arnos Press.
Powers, Bernard E. 1998. A founding father and Gullah culture. National Parks
11/12: 269.
Romaine, Suzanne, 2001. Afro-Seminole Creole. In John Algeo, ed., The Cambridge History of the English Language, vol. 6: English in North America.
Cambridge University Press, 160.
Sivad, Doug [Douglas Davis]. 1993. African Seminoles. In Ron Sakolsky and
James Koehnline, eds., Gone to Croatan: Origins of American Dropout Culture.
Edinburgh: The AK Press, 26370.
Turner, Lorenzo Dow. 1949. Africanisms in the Gullah Dialect. University of
Chicago Press.
Wood, David E. 1980. A Guide to the Seminole Settlements at Red Bays, Andros,
18171980. Nassau: The Bahamas Government Printing Department.
Woodhull, Frost. 1937. The Seminole Indian scouts on the border. Frontier Times
17(3): 11827.

part iii

Asia and the Pacific

ch a p ter 1 2

Palmerston Island English

Rachel Hendery


After 140 years of near-total isolation, the inhabitants of Palmerston Island,

a tiny atoll in the Cook Islands group, have developed a unique linguistic
and cultural identity that draws on both English and Polynesian backgrounds. They consider themselves English in many ways ethnically,
culturally, and linguistically yet also have strong ties to the rest of the
Cook Islands, and to New Zealand.
Palmerston, then uninhabited, was settled in the early 1860s by a small
group that included the Englishman William Marsters, his three Cook
Island wives, a Portuguese-speaking man named Jean-Baptiste Fernandez,
and a small group of other Cook Island men and women, who may or
may not have remained beyond the initial few years (see Hendery 2013
for more details of the islands early history). In 1877 when the first missionary visited, he noted that there were around thirty people on the
island (Gill 1877). The inhabitants today each trace their ancestry back to
Marsters and one of his wives and are monolingual speakers of Palmerston Island English. In JulyAugust 2009 when I conducted my fieldwork there the population consisted of 13 adult women, 13 adult men,
and 28 children. Five of the people currently on the island are not originally from Palmerston: this includes the teacher (from New Zealand,
now married to a local man), two Rarotongan women married to local
men, and the two missionaries, from another island in the Cook Islands
The island has been very isolated throughout its history. There is no
regular transport to or from it, and it is 400 km away from the closest
other inhabited islands. Up until around eight years ago, there were no
This research was supported under the Australian Research Councils Discovery Projects funding scheme
(Project number DP110103714). Rachel Hendery was the recipient of an Australian Research Council
Postdoctoral Fellowship.



rachel hendery

moorings, so the visitors were few, as it was dangerous to anchor for long.
During the twentieth century, sometimes many years passed without any
contact with the outside world. Nowadays around thirty yachts in total
visit each year, during the AugustSeptember cruising season. Most of
these are visitors from the USA or Europe. They are permitted to stay no
longer than three nights, and they sleep on their boats, coming ashore for a
few hours each day, usually meeting only one or two Palmerston families.
There is no television signal on the island, and until 2011, only one satellite
telephone and internet connection, used mainly by the teacher. Recently
Telecom has provided internet connections in all homes, but it remains
to be seen whether these are used and maintained. Some families listen to
radio, and most watch DVDs that they obtain from the visiting yachts,
so there is some potential outside influence on the language through these
The small population and the isolation of the island mean that it is
possible to (a) record all Palmerston Islanders, and (b) track all external
influences on the language, making Palmerston Island a wonderful opportunity for studying the development of a linguistic and cultural identity in
small mixed-origin communities. During four weeks of fieldwork in 2009,
I spoke with all of the then-inhabitants of the island except for one elderly
man, who is blind and deaf.
Previous descriptions of Palmerston Island English have appeared in
Ehrhart (1996), and Hendery and Ehrhart (2011, 2012). Ehrhart (1996)
was based on fieldwork she conducted in the early 1990s, and as discussed
in Hendery and Ehrhart (2012) and as will be seen later in this chapter,
there seems to have been some change in Palmerston English in the past
twenty years. Hendery and Ehrhart (2011, 2012) describe the morphology
and morphosyntax of Palmerston English, with particular attention to
contact influence and angloversals. This present chapter will provide a
more extensive description of the variety than previous work.

2 Sociolinguistic description
People on Palmerston Island today divide themselves into three groups,
tracing their descent through the patriline back to each of Marsters wives:
Akakaingara, Matavia, and Tepou. The island itself, and all of the uninhabited islets around the lagoon, are also divided in three, each belonging to one of the three families.1 Family membership determines land

The term family on Palmerston Island refers to one of these three groups; for the smaller nuclear
family, the term household is used.

Palmerston Island English


inheritance, council representation, hunting and gathering rights, marriage possibilities, and to some extent, day-to-day socialization patterns.
Perhaps because the islanders require marriage partners to belong to different families (meaning that children usually have parents that are each
from different groups, and women change family affiliation at marriage)
there are no detectable linguistic differences between the three groups.
Linguistic variation on the island appears to be much more individual, and
also to some extent determined by social networks, in particular the bush
people/beachfella distinction.
Palmerston Islanders refer to those who live in the cleared sandy main
settlement area at the north end of the island as beachfellas. Bush people,
on the other hand, are those who have built houses further south, among
the palm trees that cover the rest of the island. In practical terms, the
distance between the most distant houses is only a few hundred metres. In
social terms, however, distinctions are made between the two groups, and
each believes the other to have different cultural and linguistic practices.
My own observations bear this out. Counting the frequency of various
typical Palmerston Island features in my data (pro-drop, lack of formal
plural marking on semantically plural nouns, use of -s marked verbs with
non-third-person subjects, and use of bare verb forms with third-person
subjects), and controlling for interlocutor and genre, I found a small but
statistically significant difference in the frequency with which the two
groups speech display these features (p < 0.05).
In terms of their orientation to other varieties of English, Palmerston
Islanders see UK English varieties as having high prestige, but most of
their contact with other varieties is with New Zealand English (towards
which they seem to have a neutral attitude) and Cook Island L2 English
(towards which their attitudes are rather negative). These linguistic attitudes are mainly due to their historical connections. Palmerston Islanders
view themselves as distinct from the rest of the Cook Islands, and are
proud of their association with England. Public holidays are Queen Victorias birthday and Dukes Day, commemorating the day the Duke of
Edinburgh visited the island.
The school and the church play central roles in island life. Prescriptive norms are disseminated through the school, and as the teacher is a
New Zealander, the children are encouraged to use standard New Zealand
English in written and formal spoken English. Some households explicitly
discourage their children from using certain standard features they have
been taught at school (especially the use of interdental fricatives, which
are not normally present in Palmerson Island English). The missionaries
who run the church are native speakers of Cook Island Maori, and are


rachel hendery

making an effort to teach this language through classes held in the school,
and through its use in prayers and hymns. This is a recent development,
and it is possible it will lead to more passive understanding of Cook Island
Maori among the population, as well as perhaps a more positive orientation
towards the language and towards Cook Island L2 English.
There is a large amount of linguistic variation on the island, phonetic,
morphological, and syntactic. This includes both inter-speaker and intraspeaker variation. While much of it is variation between more standard
English features and Palmerston Island English features, some is among
various nonstandard alternatives. Generally I will discuss such variation
below in the description of the relevant features. For the sake of characterizing Palmerston Island English as a distinct entity, it is necessary to
abstract across the variation and to some extent ignore the use of standard
English features even when they are the most frequent variant for some
speakers. Where necessary, therefore, I will adopt the (for Palmerston rather
artificial) concept of basilect and acrolect. This allows abstraction of the
pronunciations and grammatical features that are furthest from those of
Standard Englishes. There are no speakers of Palmerston Island English
who use this basilectal variety all the time, however, and I have never witnessed even a single conversation where these basilectal features were not
heavily mixed with more standard English features. It should also not be
assumed by the reader that the features described here are necessarily older:
there is in fact evidence that variation existed right from the earliest decades
of the islands settlement.

3 Features


3.1.1 Consonants
Palmerston Island English has, at a minimum, the consonants listed in
Table 12.1. In basilectal Palmerston English there is no //, //, no /v/ or /h/.
Standard English // and // correspond to Palmerston /t/ word initially
and /s/ word finally. Standard English /v/ corresponds to Palmerston /w/
in all positions. Standard English /h/ corresponds either to a glottal stop,
or to zero: the glottal stop is common, but apparently optional. Standard
English voiced stops and fricatives are devoiced in the usual Palmerston
None of these correspondences is absolute: for any word there is a range
of pronunciations along a continuum from Standard English to basilectal


Palmerston Island English

Table 12.1 Minimal consonant inventory for
Palmerston Island English









Palmerston English. This means that one can analyse e.g. the voiced stops,
fricatives, and /v/ as having marginal phonemic status. They appear in free
variation with their voiceless counterparts, but only in the set of words that
have the voiced sounds in Standard English. The same cannot be said for
/h/. Glottal stop, /h/, and zero onsets appear to be in free variation in all
words that begin with a vowel. The /r/ is trilled.
Palmerston Island English has a phonotactic constraint whereby consonant clusters in syllable codas are avoided. Clusters that would appear
in Standard English tend to be reduced in Palmerston English to a single
consonant. The velar nasal is not limited to syllable-final position to the
same extent as it is in Standard English, but can be found word initially in
proper names and in words that have their origin in Cook Island Maori.
3.1.2 Vowels
Variation in Palmerston Island English vowels seems to be on a continuum
from a system that clearly has its roots in a Northern English dialect through
to something more like that of Cook Islanders who are L2 English speakers.
Several Palmerston Islanders have vowels closer to those of Standard New
Zealand English. There are often striking differences between the vowels
of speakers from the same household, who grew up together and who have
the same degree of contact with New Zealand. Monophthongs
Length rather than height appears to be the main distinction between
the fleece and kit vowels. While fleece tokens on average are higher
than kit tokens, there is a great deal of overlap between the two sets,
and any individual token of either set could fall within the average range


rachel hendery

of the other. Length, on the other hand, always distinguishes the two
sets, with fleece tokens generally lasting approximately twice as long as
In traditional Palmerston Island English there is no split between the
strut and the foot lexical sets: both have the vowel []. This was presumably the pronunciation inherited from William Marsters, and suggests
he came from the Midlands or further north in the UK. This is a very
salient feature to the community. Speakers will always produce examples
such as /km/ or /bt/ when asked about how their grandparents spoke.
Most speakers today have the strut/foot split, but a small number of
lexical items that have remained in the foot set while other varieties of
English have assigned them to strut: bucket is always one of these, and
depending on the speaker the exceptions can also, for example, include
come, but, and cup. For some speakers, comer newcomer/outsider has the
foot vowel while come has strut. The phonetic realization of the foot
vowel for Palmerston Islanders today is [] rather than [], but they produce something closer to [] when giving examples of their grandparents
speech. Several older speakers have [] for all words in both the strut and
foot sets.
There is a bath/trap distinction [a]/[], and unlike the strut/foot
pair, no indication that the split is recent. This helps narrow down Marsters
place of origin, as this split occurred in southern England. This supports
the widely accepted story that he came from the Midlands (probably
Leicestershire or Birmingham).
The nurse vowel in Palmerston Island English is much more fronted
than it is in many other varieties of English, usually realized as []. Length
rather than height is the main distinction between this vowel and that of
the dress set. Diphthongs
The diphthongs of Palmerston English that are most different from those
of other English varieties with which it is in contact (Cook Island English,
Australian English, and New Zealand English) are those of face, goat,
and mouth. Face sometimes receives a monophthongal pronunciation
[e]. For most speakers, however, it is realized as []. The goat set has
an open rounded vowel, usually followed by a schwa off-glide: [ ]. The
mouth diphthong has a great deal of variation, with some speakers regularly producing a very New Zealand-English-sounding close vowel, but
most using a pronunciation more like conservative RP English: [a]. Even
in speakers who otherwise show strong New Zealand English influence

Palmerston Island English


in their vowels, there is no trace of the near/square merger that New

Zealand English has undergone in recent years.
3.1.3 Stress
Stress in Palmerston Island English sometimes differs from the stress patterns of Standard English. All of the cases of this I am aware of are ones
where Palmerston English has shifted the stress to the penultimate syllable, sometimes from the final syllable such as in introduce or canoe and
sometimes from the antepenultimate syllable, as in octopus.

Morphology and syntax

3.2.1 Pronouns
Palmerston Island English uses the resources of Standard English, combined with several innovative forms to construct a pronominal system
that essentially mirrors the Polynesian systems of the region. An inclusive/exclusive distinction is encoded in the first-person dual, with the term
yami being used for first-person inclusive. This form is not obligatory, and
is only infrequently used by Palmerston Islanders today. The alternative
first-person dual terms are we two, found in my data only with inclusive
sense, and (the) two of us, used for both inclusive and exclusive. In the
plural, we all seems to be used with an inclusive reading, while us lot has
both inclusive and exclusive use.
Dual and plural are usually distinguished, encoded by the use of two,
lot, or all. The plural pronoun yous is also found, but seems to be a recent
borrowing from New Zealand English, as it is not found in Ehrharts
recordings from the early 1990s. A new dual form for third-person mixedsex pairs has arisen among young people, himshe, illustrated in (1). My 16and 18-year-old informants use it often, and it seems to be in common
usage with younger children too. I did not observe it among the adults.
(1) Himshe standing over there.
They are standing over there.
The pronoun it is sometimes found with human referents, as in example
(2) I think they call it Papa. Papa Ruau. Because Papa Ruau came from
the Northern group and it got involved with the English.
I think they call him Papa. Papa Ruau. Because Papa Ruau came
from the Northern group and he got involved with the English.


rachel hendery
Table 12.2 Pronominal forms in Palmerston Island English





1 incl
1 excl

I, me

we two, yami, two of us, we, us

two of us, we, us
you two, you
himshe, them two, they, them

we (all), us (lot), we, us

us lot, we, us
you lot, yous, you
they, them, big lot of them

he, she, it, him, her

Standard English object pronouns me, him, her, us, them, and the various
Palmerston extensions of these (two of us, us lot, big lot of them, etc.), are
used very frequently used as subjects, as shown in (3).
(3) Me busy right now.
I am busy right now.
Table 12.2 gives a summary of all pronominal forms in use on Palmerston Island, including those that are identical to Standard English. Selection
among the alternatives seems to be governed mainly by register, and secondly by age of the speaker (yami is possibly used more by older speakers
although I do not have enough tokens of it in my data to be certain of this;
himshe is used exclusively by young people.)
Pronouns do not have to be explicitly expressed in Palmerston Island
English. Subject and object arguments can be elided when their referents
are clear from context. Pro-drop is very frequent, especially with dummy
subjects such as in existential constructions (There is . . . ). Constructions
with its are ambiguous, as consonant cluster reduction means that its
would often be realized as [is] in any case. One might speculate that this
could have been a contributing factor in the development of pro-drop on
Palmerston Island together with influence from Cook Island Maori and
perhaps Fernandezs Portuguese. Examples of pro-drop are given in (4).
Either of the arguments of ditransitives can be elided.
(4) a. Its really fun when hear them speaking.
Its really fun when you hear them speaking.
b. And now is thousand and thousand baby around.
And now there are thousands and thousands of babies around.
The frequency of pro-drop seems to be correlated with gender. The
highest users of pro-drop are all men (including one who in my corpus
leaves 37 per cent of his subjects unexpressed, and two others who use
pro-drop for more than 10 per cent of their subjects 12 and 19 per cent

Palmerston Island English


respectively). The lowest users, six of whom use it less than 5 per cent of
the time, are all women.
3.2.2 Nouns
Formal marking of pluralization on nouns is optional in Palmerston Island
English. When nouns are marked as plural, -s is generalized to any noun
(and has no /z/ allomorph as it does in Standard English). (This allomorph
would not be expected, in any case, due to the devoicing of final stops
described above. A /s/ allomorph does exist, however.) Nouns that have
irregular plurals in Standard English sometimes receive the -s plural in
Palmerston English as well as the irregular plural marking, so that we
find singular/plural alternations such as childchildrens, womanwomens,
The form of the indefinite article is always //, never /n/. The definite
article is /d//t/, often reducing to /dt/ before vowels. Definite and
indefinite articles do not seem to have the same distribution as in Standard
English. A full account of Palmerston Island English article usage remains
a matter for future research. A few generalizations that can be made at
this stage are that definite articles are often used with years (I left here in
the 1992); Palmerston English has some mass nouns that are count nouns
in Standard English, and vice versa, which affects article selection; and
quantifiers in general interact with articles differently from how they do in
Standard English.
3.2.3 The noun phrase
The structure of the noun phrase is by and large the same as in Standard
English. The only differences are in the area of possession and relativization.
Possession usually takes the form: possessed of possessor, even in long
strings such as the cousin of the family of my grandmother. Palmerston Island
English prefers the s-clitic construction only for proper names, which rarely
occur with the of construction. Either construction is possible for other
animate possessors. A bare juxtaposition of two noun phrases is another
common possessive construction, especially frequent with proper names.
It is possible to use the order possessedpossessor in Palmerston Engligh
even without the of construction. This is most common in possessive
compounds such as blood pig pigs blood or shit chicken chicken shit,
but can also be found with the -s construction such as in example (5). As
this word order is the normal possessive order in Cook Island Maori it
seems likely that its predominance in Palmerston Island English is due to
substrate influence or subsequent contact effects.


rachel hendery

(5) two barrel diesel Korinakos

two barrels of diesel belonging to Korinako
There are several common relativization strategies in Palmerston Island
English. One is a zero-marked strategy. Examples of this are given in (6)
(6) a. There was someone already been here before.
There was someone who had already been here before.
b. That was Gill says this. (Ned Marsters, Burland interview 1959)
That was Gill who said this.
As we can see from these examples, unlike Standard English, Palmerston
Island English does not restrict this strategy to clauses where NPrel functions as an object. This is presumably related to the general permissability
of pro-drop, even in simple clauses.
Zero-marking can also be found in combination with pronoun retention,
however, producing a paratactic relative clause:
(7) See, my father he was a man he believes in Him.
See, my father was a man who believes in Him (Ned Marsters,
Burland interview 1959)
Pronoun retention in the relative clause occurs in the other RC constructions of Palmerston English too. There are also examples in my corpus
where the entire NPrel (i.e. not only a pronoun) is retained.
(8) you chuck their feed which their feed will be on the hook and then
you grabem.
You throw their food (into the water), which will be on the hook,
and then you grab them.
In the earlier attestations of Palmerston Island English such as John
Burlands interview with Ned Marsters in 1959, we also find relative clauses
marked with what. I did not find these in 2009. Besides the zero-marked
strategy, I found only constructions with who, which, and that. Unlike some
varieties of English, Palmerston Island English allows that with animate

Verbs Verb agreement

In Palmerston Island English we find that all persons and numbers can
occur with verbs unmarked for agreement, or with verbs marked with -s, as


Palmerston Island English

Table 12.3 Present-tense verb paradigm for Palmerston Island English

1 incl
1 excl




I goes

you goes
he/she/it go

yami goes
the two of us goes
you two goes
them two goes

we/us goes
us lot goes
you lot goes
they/them goes

Table 12.3 shows. Ehrhart (1996) has analysed the system as one in which
the bare verb is used for third- person singular, and -s is used for all other
person and number categories.
This may have been the general system in the early 1990s when Ehrhart
conducted her fieldwork, but it is not possible today to find speakers
who consistently follow this pattern. The speakers who use -s on nonthird-person-singular verbs do so, at most, 30 per cent of the time. Many
speakers never use -s for these. The distribution is not accounted for by
e.g. the Northern agreement rule, although this may have contributed to
the origin of the Palmerston system. The frequency of 30 per cent is much
lower than the frequencies of other Palmerston Island English linguistic
features, so it seems unlikely that this frequency could be accounted for
simply by register variation or accommodation.
It seems likely that the apparent inconsistencies are the result of competing systems: the traditional one described by Ehrhart competing with the
Standard English system. Because the two systems are exactly the inverse
of each other, this explains why any person/number seems able to appear
with or without -s. Interestingly, there is a third pattern that some speakers follow consistently: bare verbs for all persons/numbers. These speakers
seem to have resolved the conflict between the two systems by removing
all agreement markers entirely. Ehrhart did not find this bare verb system
among the speakers she worked with in the early 1990s, so it appears to be
a recent development. Tense
A bare or -s marked verb in Palmerston Island English can have past, present
or future interpretation depending on the context. Some examples of these
forms with past and future reference are given in (9).
(9) a. Yesterday we eat fish.
Yesterday we ate fish.


rachel hendery
b. We want to make sure everything is ready for tomorrow morning:
just hop on the boat, pile everyone inside and off we goes to the
We want to make sure everything is ready for tomorrow morning:
well just hop on the boat, pile everyone inside and off well go to
the islet.

Alternatively, future time reference can be expressed with will/ll, going

to, or gonna. Past time reference can be expressed with the Standard English
simple or perfect past constructions, including a perfect construction without an auxiliary, which I will discuss under aspect below.
Preterite forms are often slightly different from those of Standard
English. Some irregular forms have been regularized, and consonant cluster
reduction means that final -d is often not present. In what might be another
strategy for overcoming the phonotactic constraint by resyllabification
the past tense is sometimes doubly marked by the addition of the -ed morpheme to verbs that already contain it. This seems to occur in both the
preterite and in participle forms. Some examples of this double marking
are shown in (10).
(10) a. A man passeded him with his goat.
A man passed him with his goat.
b. This island is blesseded by God.
This island is blessed by God.
The avoidance of a consonant cluster cannot be the only reason for this
form, however, as we also find, for example, tooked, shooked, and stucked,
among other forms, where the addition of the /t/ suffix creates a cluster
that would not exist in Standard English. The only examples I have of
these latter forms, however, occur before words beginning with a vowel,
allowing resyllabification.
Another common way to express past tense in Palmerston Island English
is by use of the preverbal marker been. This is a construction found frequently in creoles and other Pacific English varieties.

a. She been record you?

Has she recorded you?
b. We been go in a group.
We went in a group.

This construction seems to be most frequent in cases where standard

English would use a perfect aspectual form such as She has recorded you or


Palmerston Island English

Table 12.4 Standard English and Palmerston Island English verbal and
adjectival predicates
Standard English
subject be adjective
He is tall
subject verb
He looks

Palmerston Island English

subject be adjective
He is tall
subject be verb
He is look

subject adjective
He tall
subject verb
He look

I have eaten one, but as (b) shows, this does not hold exclusively, which is
why I have included this form as a tense rather than an aspect.
A complication of this form is that there is also a corresponding presenttense construction consisting of an inflected present-tense form of be plus
the bare verb, illustrated in (12).

a. Yeah Im use this New Zealand slang, eh?

Yeah, I use New Zealand slang, eh?
b. Hes steal it first and aks after
He steals it first and asks afterwards.
c. Glad these things is clean their place now.
Im glad they have cleaned their place now.
d. Both of them are look.
Both of them (would) look.

These examples all have different aspectual and modal readings so it

is difficult to know what the function of this construction is. As the
Palmerston Island English lexicon is very multifunctional, it is possible
that this construction is simply a natural outcome of collapsing some of
the distinctions between adjectives and verbs. In Table 12.4 I show how the
verbal and adjectival predicates of Standard English differ from each other,
and how these differences have been levelled out in Palmerston Island
English, first by the acceptability of a copular-less adjective predicate, and
secondly by the existence of the be + bare verb construction. The parallel
between this and verbal predicates without be is enhanced by the degree to
which the present-tense verb in Palmerston English resembles the infinitive,
since agreement is optional even with the third person singular. Aspect
The only innovative element with a clearly aspectual function in Palmerston
English is the clause-final finish, which as in many other Pacific Englishes


rachel hendery

and English-based creoles has a completive meaning. In Palmerston Island

English this element seems to be new Ehrhart did not find it when she
conducted her fieldwork in the early 1990s. It is optional, and not very
common. While I heard it more frequently than this, it only appears four
times in my recorded corpus. One of the four instances is not prototypical,
as it seems to be a paratactic juxtaposition of two clauses: Ill leave you and
is/its finish.2 This example is interesting as an indication of how the finish
construction could have arisen, through juxtaposition and then reduction
of the clause is finish. Alternatively, or in combination with this, there may
have been contact influence from other Pacific English varieties.

a. I painted the boat finish.

I painted the boat.
b. I went feed my pig finish, came sit here.
I fed my pig and then came to sit here.
c. Tell a story. Ill leave you is finish.
Tell a story. Ill leave you (before you begin).

The perfect of Standard English can be used without an auxiliary in

Palmerston Island English, for example done, seen, been in (14). These only
seem to occur with perfect aspectual readings, so cannot be simply analysed
as levelling of the preterite and participial forms.
(14) a. If you done something stupid, I would say to you shei.
If you had done something stupid, I would say to you shei.
b. I never seen any of my grandparents.
I have never seen any of my grandparents.
c. I been to Australia three time.3
I have been to Australia three times.
The continuous is formed, as in Standard English with a participle in
-ing, (which, as in many varieties of English, can also be realized as -in). A
small number of verbs have a present participle in -ening. The prototype
for this is fishening, which is used by all speakers the majority of the time.
The form is found occasionally for other verbs, in particular singening,
although this is not as universal as fishening. It seems unlikely that the -en

As mentioned earlier, is and its are indistinguishable in Palmerston Island English due to consonant
cluster reduction. Is finish would also be acceptable as a full clause, however, because of pro-drop.
The form finish is expected in the sentence It is finish(ed), as consonant cluster reduction means that
finished does not fit Palmerston Island English phonotactics.
This construction should not be confused with the been + infinitive construction described above.
That construction does not have the same aspectual constraint.

Palmerston Island English


in these forms should be equated with the -en of the past participle, as
these verbs do not take a past participle in en, either in Palmerston English
or Standard English. It is perhaps more likely that they arose through the
addition of -ing to forms such as fishin and singin, making it similar to
the double-marking of past-tense verbs such as blesseded etc., mentioned
As we saw for the perfect construction, an auxiliary is optional in the
continuous construction too:

Dog is up on the bottom, jump out of window, and the boy looking
The dog is up on the bottom (of the windowsill), jumps out of the
window, and the boy is looking at the dog. Mood
There appears to be a rather infrequent subjunctive use of be in Palmerston
Island English, shown here in examples from Ned Marsters in 1959 and
from one of my informants in 2009.
(16) a. He says they all be the same, his family and Mahutas (Ned
Marsters 1959)
He says they are all the same, his family and Mahutas.
b. I try to do it the way shes saying it,so when I speak to the children
it be the way that shes wanting.
I try to do it the way she says it, so when I speak to the children
it is the way she wants it.
The other noteworthy feature of Palmerston English with regard to
mood is the development of the modal verb must. This has apparently
become an adverb on the model of Standard English maybe. This is illustrated in (17).
(17) a. Mustbe hes big.
He must be big.
b. Cos its the only boat went out, mustbe is Bob.
Because its the only boat that went out, it must be Bob.
c. It surely mustbe is Bob.
It surely must be Bob.
d. It mustbe fell in.
It must have fallen in.


rachel hendery

Syntactically it would still be possible to analyse each of the above

examples as paratactic juxtapositions of several clauses, one containing the
two verbs must be, with pro-drop accounting for the reduction of some
clauses. For example, (a) would then be derived from (It) must be (that)
hes big. Intonationally, however, mustbe appears to be a single unit, with a
heavy stress on must and unstressed be. I would therefore prefer the simpler
analysis of these examples, where mustbe is being used as an adverb.
3.2.5 Prepositions
Selection of prepositions in Palmerston Island English often differs from the
corresponding prepositions in Standard English. Some verbs that require
prepositional complements in Standard English select directly for noun
phrases instead in Palmerston Island English. Two of these are illustrated
in (18).

a. And the farmer went inside, went to look the water.

And the farmer went inside, went to look at the water.
b. Im not talking all of the island.
Im not talking about all of the island.

Some uses of prepositions in Palmerston Island English suggest that different ways of conceptualizing space apply than in other English varieties.
For example, one falls off rather than out of a tree and climbs on a tree
instead of up it, suggesting that climbing trees is envisaged as a person
pressed against a trunk, rather than sitting or standing among the branches
or in the canopy. This makes sense when one considers that most trees
on Palmerston Island are coconut palms. One falls from a window, rather
than out of a window. Some set idioms also have different prepositions in
Palmerston Island English. These include, for example, to put ones trust
into someone, and to be good on something. One sets something with fire
rather than setting it on fire.

Word order

3.3.1 Basic word order

The basic word order of Palmerston Island English is SVO. Both verbal and
non-verbal predicates exist, and there is generally no copular in nominal,
adjectival or locative predicates. Examples of the latter are shown in (19).
(19) a. She a teacher.
She is a teacher.

Palmerston Island English


b. Me busy right now.

I am busy right now.
c. They on this island.
They are on this island.
Ditransitives display the same argument order alternation as in standard
English, as shown in (20).
(20) a. We each been giving fish to them.
We have each been giving fish to them.
b. Give her a hiding.
Give her a beating.
3.3.2 Fronting
Fronting of the predicate is very frequent in Palmerston English, especially
in non-verbal clauses. This produces a word order similar to that of Cook
Island Maori, so this construction may be the result of substrate influence
or subsequent contact with Maori L2 English speakers.

a. Too small, the table

The table is too small.
b. Is hot, the sun.
The sun is hot.

Fronting, usually with pronominal resumption, can also be used for

topicalization, as in (22). A sentence can have multiple fronted topics, as
in (22b).
(22) a. The yacht, is he going to full up?
Is the yacht going to fill up?
b. I know in the island in the islets mosquito, I dont know
how he breeds.
I know . . . I dont know how mosquitos breed in the islets. (Ned
Marsters, Burland Interview 1959)
On or for can be used as a topic marker before a fronted constituent, as
in the following examples.
(23) a. On the cup, I heard about that.
As for the cup, I heard about that.
b. For me, I say um.
As for me, I say um.


rachel hendery

3.3.3 Questions
Yes/no questions in Palmerston Island English can have a standard dosupport construction, as in (24a), or simply use declarative word order
with interrogative intonation as in (24b).
(24) a. Does they eat it?
Do they eat it?/ Will they eat it?
b. You like the fish?
Do you like the fish?
Wh-questions also frequently have the same word order as declaratives. Interrogatives are optionally fronted, and do-support generally absent.
Examples of wh-questions are given in (25).
(25) a. Where you was?
Where were you?
b. What Shirley was doing there when you went?
What was Shirley doing when you went there?
c. What for you want that?
What do you want that for?
Embedded questions usually resemble those of Standard English in
their word order. This means that Palmerston Island embedded questions
and non-embedded questions can be identical. Occasionally verb inversion occurs in embedded questions, such as the first embedded question
in (26).
(26) You describe where is the dog and where you go to the fish.
You describe where the dog is, and where you go to get to the fish.
3.4 Lexicon
A striking feature of the Palmerston Island English lexicon is the flexibility
of words with regard to their word class. Many items function as nouns
and verbs, some examples of which are given in (27).
(27) a. Then you dough it.
Then you knead it.
b. You broom now: next month you broom again
If you sweep now, next month youll have to sweep again.
Some elements function both as adjectives and verbs, as illustrated in

Palmerston Island English


(28) a. We off the generator at two.

We turn the generator off at two.
b. She bright her eyes.
She opened her eyes wide.
As shown in the discussion of verbs and the description of coordination
patterns above, the syntactic distinction between adjectives and verbs is in
any case less sharp in Palmerston English than it is in many other varieties
of English. It is unclear, however, whether this is a contributing factor to
the multifunctionality of these words or the result of it.
Reduplication exists, but is infrequent and does not seem to be transparent or productive. Some examples of reduplicated forms are kaikai feast,
sing-sing sing-along, and chuck-chuck to do something enthusiastically.
The latter is illustrated in (29).
(29) You hear Bob chuck-chucking it in church today?
Did you hear Bob singing enthusiastically in church today?
Most Palmerston English words with Maori origin are in the domains of
flora and fauna, but there are a handful of other Maori words in common
use, including umu ground oven, motu islet, enua island, uapo singalong, orometua missionary, and the honorific titles mama and papa used
for the elderly and for religious leaders.
There are several words with unknown origin: these include /paki/
skinny (prototypically used for fish, but by extension sometimes used
for other animals and humans), // (an admonition used to children
who are misbehaving, perhaps from English shame) and /lk/ (a similar
admonition used for adults who are doing something stupid, perhaps from
English should look).
Some English-origin words are rare or archaic in other varieties of
English, but common on Palmerston. These include fowl, eyeglass, shanty,
poorman grapefruit, lead (for electrical cord), and pitch (used for any dark
unidentified mark or dirt).
Others have undergone semantic shift. A t-shirt is known on Palmerston
Island as a singlet, and a sleeveless vest is a short-sleeved singlet. Stoppers of
any sort, including screw-top lids, are bungs. A coloured ink is a coloured
pen. The word drink implies a hot beverage, such as tea or coffee; a cold
drink must be specified as such, and the phrase hot drink is redundant
and not used. A cup is not an umbrella term that encompasses glasses, but
rather used for a specific subset of drinking vessels. A cup must have a
handle. A mug, on the other hand, is a very large vessel for pouring from,


rachel hendery

not for drinking out of. Slippers are what other English-speaking countries
call flip-flops, jandals, or thongs.
One picks birds rather than catches them, and spills a drink (into a cup)
rather than pours it. Directions are usually given as down (towards the
north, where the main settlement is), or up (to the south, away from the
main settlement). An acceptable answer to the question Where are you
going? is simply Going up, or Going down. The phrase thank you is used to
mean something like I understand what you are saying, and the general
greeting to all visitors to the island or to a house, even to non-Palmerston
Islanders visiting for the first time, is welcome home.


Palmerston Island presents an extremely valuable case study for linguists

interested in mixed languages, variation, new dialect formation, and the
linguistic effects of isolation. It provides an interesting parallel to the
development of New Zealand English, as it was settled not long after
the colonial settlement of New Zealand, and both cases involved very
similar inputs: UK English and Maori. Comparison of Palmerston Island
English with other Pacific English varieties, in particular Pitcairn-Norfk
(as Tahitian is very similar to Cook Island Maori), may also be fruitful.
The similarities found across Pacific English varieties can lead us to a better
understanding of the mechanisms involved in mixed language formation,
creolegenesis, and contact-induced change, in particular the role played by
the typological characteristics of the input languages. Palmerston Island
English is a particularly useful addition to such comparative study because
its history is relatively short and well documented. We know exactly what
the inputs were, and the degree and type of contact that has occurred since
settlement, both of which are more limited than is the case for many other
Pacific English varieties.
The future of Palmerston Island English is uncertain. The island is, at its
highest point, three metres above sea level, rendering it extremely vulnerable
to climate change. Severe cyclones have in the past deforested the island
of coconut palms and damaged the reefs, leaving the population without
their most important food sources. With climate change, such storms are
likely to become more frequent, and rising sea levels will ultimately make
the island uninhabitable. Most Palmerston Islanders have close family in
Auckland, New Zealand, and it is more likely that they would resettle there
than in Rarotonga, where they also often have family connections but do
not speak the language. The Palmerston Island community in Auckland is

Palmerston Island English


said to maintain Palmerston English to some extent, but it seems doubtful

that this could continue indefinitely. Further research on Palmerston Island
English is therefore a matter of urgency.

Burland, John. 1959. Unpublished transcript of an interview between John Burland, Ned Marsters and Peka Marsters. Wellington, New Zealand. Alexander
Turnbull Library. MSX-8809.
Ehrhart, Sabine. 1996. Palmerston English. In Stephen A. Wurm, Peter
Muhlhausler and Darrell T. Tryon, eds., Atlas of Languages of Intercultural
Communication in the Pacific, Asia, and the Americas. Berlin and New York:
Mouton de Gruyter, 52336.
Gill, W. W. 1877. Unpublished Letter to the London Missionary Society regarding
visit to outstations July 3Aug 14. LMS archives, School of Oriental and
African Studies, London. CWM South Seas Journals Box 11, 187181.
Hendery, Rachel. 2013. Early documents from Palmerston Island and their implications for the origins of Palmerston English. Journal of Pacific History 48:
Hendery, Rachel and Sabine Ehrhart. 2011. Palmerston Island English. In Bernd
Kortmann and Kerstin Lunkenheimer, eds., The Electronic World Atlas of
Varieties of English [eWAVE]. Leipzig: Max Planck Institute for Evolutionary
Hendery, Rachel and Sabine Ehrhart. 2012. Palmerston Island English. In Bernd
Kortmann and Kerstin Lunkenheimer, eds., The World Atlas of Varieties of
English. Berlin and New York: De Gruyter Mouton, 62842.

ch a p ter 1 3

Pasifika Englishes in New Zealand

Donna Starks, Andy Gibson and Allan Bell


This chapter considers forms of English spoken by Polynesian peoples in

New Zealand. We start with a geographic description of the Pacific, situating Polynesia and its peoples and their economic and political positions
in the region and outlining the effect these positions have had on the
movement of people and on their language use. An increasing number of
Polynesian communities are located outside their home islands, in largely
anglophone nation states on the Pacific Rim, resulting in bilingualism and
language shift to English. To describe the forms of English that emerge,
we need an understanding of the similarities which bind together, and the
differences which distinguish, Polynesian people of different ethnicities,
ages and language backgrounds. As part of this process, we discuss the concept of pan-ethnic identities and language varieties, which conceal cultural
differences whilst embracing similarities. Once the sociolinguistic context
has been described, the second half of this chapter provides a comprehensive overview of the (limited) existing literature on the features of Pasifika
Englishes in New Zealand.
The Pacific Islands is a cover term used to refer to the more than
20,000 islands located in a geographic area spanning many thousands of
kilometres across the South Pacific Ocean. Much of this area is known as
Oceania, and is comprised of three main groupings of islands: Micronesia,
Melanesia and Polynesia. Micronesia is the area closest to the equator,
and includes the islands of Guam, Palau, Nauru and Kiribati, amongst
others. Melanesia refers to the group of westernmost islands. The area
reaches to the islands of New Caledonia and Vanuatu, includes Papua
New Guinea and parts of Indonesia, with Fiji at its most easterly point.
This positioning means that Fiji is geographically closer to Polynesia than
much of Melanesia, and has strong links with its Polynesian neighbours.
While we do not cover Fijian English in this chapter, it is one variety of

Pasifika Englishes in New Zealand


English in the Pacific islands that has been well described (Tent and Mugler
The Polynesian islands cover the greatest geographic expanse, roughly
triangular in shape, with Aotearoa/New Zealand, Rapa Nui/Easter Island
and Hawaii as its three corners. The geographic positioning of Hawaii,
and its status as part of the USA, has resulted in links which are more eastward than westward in orientation, leading to differences between Hawaii
and its geographically distant Polynesian relatives. American Samoa is in
a similar situation. Most of the Polynesian islands have relatively small
populations which historically have been ethnically homogeneous. While
some countries, such as Samoa, Tonga and Tuvalu, are independent nation
states, others such as Niue and the Cook Islands are in free association with
New Zealand, with its citizens having rights similar to those of New Zealand
citizens, including free entry into the country, and access to local education
for those who wish to study in New Zealand. Historically, the Polynesian
island states were self-sufficient, but in todays global economy they are
often classified as relatively poor. Some are also threatened by sea-level rises
through climate change.
As a consequence, from the second half of the twentieth century, Polynesian peoples began immigrating in large numbers to countries within the
Pacific Rim in search of better economic opportunities, especially in New
Zealand, Australia and the United States. The flow of populations between
Australia, New Zealand, the USA and the island states is often multidirectional as close-knit Polynesian families often move back and forth, visiting
their relatives in their home islands, forming a chain of migration across
the Pacific. The amount of travel from the Polynesian islands is, however,
unbalanced. While birth rates on the islands are often higher than in developed countries, popul