Вы находитесь на странице: 1из 17

Commonwealth of Australia

Copyright Act 1968

Notice for paragraph 1 35Z XA (a) of the Copyright Ac t 1968


This mat erial has been r eproduced and communi cated to you by or on behalf of th e

UniversityofWestern Sydney un der Part VB of the Copyright Act 1968 (the

A ct).

The material

in thi s co mmuni cation may be subj ect to copyright under th e Act. An y furth er

reproducti on

or communi cation of thi s material by you may be th e subj ect of copyri ght

protecti on under the Act .

Do not remove this noti ce.


Phonology and Phonetic Transcription

Many people think that learning phonetics mean s simply learning to use pho-

transcription. But there is really much morc to the s ubject than learning to

a set of symbols. A phonetician is a person who can desc ribe s peech , who lands the mechanisms of speech production and speech perception, and


knows how languages use these mechanisms. Phon eli c lranscriplion is

than a useful tool that phoneticians use in the description of speech. II is. 'ever. a very important tool . In Ihis chapler, we will be concerned with the phonetic transcrip'ion of care- ful speech - th e s tyle of speec h you u sc to show someo n e h ow CO pronounce a

word. This is called the citatio n s tyle of speec h. Tran scriptions of ci tation sty le

Me particularly useful in language documen ta ti on and lexi cography, and also as the basic phonetic observations described in phonology. In Chapter 5, we will discuss phonetic transcription of connecled speech-th e style that used in

conversation .When phoneticians transcribe a citation speech uUerance,


we are usually concemed with how the sound convey differences in meaning.

For the most part, we describe only the significant articulations rather (han the

4etails of the sounds. For exam ple. when saying the pie pronounce the consonant with the blade of th e

ridge, Olhers with the tip of the tOngue. Thi s kind of d iffe rence in articulation not affect the meaning of the word and is not usually transcribed. We will begin by considering just this simplest form of transcri ption, sometimes called a

brood trunscription.

In order to understand what we tran scribe and what we don't. it is necessary understand the basic principles of phonology. Phonology is the description of systems and pattems of sounds that occur in a language . II involves s tud y ing a language 10 determine its d istinctive sounds, that is, those sounds that co nvey a difference in meaning . Children have to do this whe n they are learning to spea k. Tbey may not realize at first that, for exampl e. there is a d iffe re nce between the consonants at the beginnings of words such as white and right. They later reali ze that these words begin with two distinct sounds. Eventually, they leam to d istinguish aU the sounds that can change (he meanings of words.

English word lie, some peo· tongue against the alveolar



Phonology and Phonetic Transcription

When two sounds can be used to differentiate words, they are said to belong to different phonemes. There must be a phonemic difference if two words (such as white and right or cat and bat) differ in only a single sound. There are, however, phonetic variations that cannot be used to distinguish words, such as the differences consonants at the of the word pop.

the lips must opcn be a puff of air

the final consonant,

it is not

it happened word you said before The sound at the

end would consonants in this

They are

but the differences between them cannot be used to change

For the first before the


a puff of air, but lips for hours, if

bilabial stops.

you could say pop

the meaning of a word in English. They both belong to the same phoneme. We cannot rely on the spelling to tell us whether two sounds are members of different phonemes. For example, the words phone andfoam begin with the same sounds, although they have different spellings. To take a more complex ex-

ample, the words key and car begin with what we can regard as the same sound, despite the fact that one is spelled with the letter k and the other with c. But in this case, the two sounds are not exactly the same. The words key and car begin

with slightly

sounds. If you whisper

consonants in these

two words.

hear the difference,

be able to feel

that your

roof of the mouth for each word.

between members slightly different, English. They are

both members of the same phoneme. We noted other small changes in sounds that do not affect the meaning in

Chapter 1, where we saw that the tongue is farther back in true than in tea, and the n in tenth is likely to be dental, whereas the n in ten is usually alveolar. In some cases, the members of a phoneme are more different from one another. For example, most Americans (and some younger speakers of British English) have

a t in the middle of pity that is very different from the t at the end of the word

play. You can say

just the first in this word without but still hear the I

pit. The one

This example there may be very subtle

of a phoncrne.

at the beginning of

but it is not changes the meaning

more like a d. Consider


(try doing

say the whole word

typically voiceless,

and very

I in lay. Say the I of lay, and you'll

hear that

voiced. examples that a phoneme single sound, but a

name for

There is a group 01 group of I sounds

that occur in English. It is as if you had in your mind an ideal t or I, and the ones that are actually produced are variations that differ in small ways that do not affect the meaning. These groups of sounds-the phonemes-are abstract units that form the basis for writing down a language systematically and un- ambiguously. (Peter Ladefoged's book Vowels and Consonants has an extended

discussion of the relationship between written language and phonology in which

k speCulates lhatlhe development of phonemic analysis was partly due to the

-.mg systems used by European linguists.} We often want to record all-a nd only- the variations between sounds that

are called pItonDnjc

-.scriptions. Languages that have been written down only comparatively ~ot ly (s uch as Swa hi li and mos t o f th e o ther languages o f Africa) have. lid) phonemic spelling syStem. There is very little difference betwcco a written

sentence . wh ile the

.Iling has remained basical ly the same . phonemic tran sc ripti ons of Engli s h

8lIIC different from wrinen texts.

a difference in meanin g. Tran sc riptio ns of this kind

-sian of a Swah ili se ntence and a phonemic transcriptio n of th at because English pronunciation has c hanged over the centuries


We can begin searching for phonem es by conside rin g th e contrasting co nsonant

munds in Englis h. A good way is to find se ts of words

cumpJe. aU the words that rhyme with pie and have onl y a sing le consonant at IIIe beginning. A set of words in which each differs from all the others by o nly -.e sound is called a minimal set. The second column of Table 2. 1 li sts a se t of


words begin with sequences oflwo or mo re of the sounds

already in the minimal set. Some of the words in the li s t begin w ith two conso-

(thigh . thy, shy). but they each begin with a sin gle consonant sound .

SIIy, for example. does nOt contain a sequence of two co nsonant sounds in th e way that spy and try do. You can record these words and sec the sequences in spy md try for yourself. Some consonants do not occu r in words rh yming wi th pie. If we allow us in g the names of the letters as words. then we can find another large set of ca nso--

DanIS beg innin g words rh y ming with pea. A li s t o f s uc h word s is

third column or Table 2.1. (S peakers or Briti s h En g li sh wi ll have to reme mber that in American Engli sh, the name or the last leiter of the alphabet belongs in this set rather than in the se t o f words rhyming with bed.) Even in this set o f words, we are still mi ssing some consonant sounds that ca n· crast with others only in the middles or at the ends of words. The letters ng often represent a s ingle consonant sound that does nOl occur at the beginning of a word. You can hear thi s sound at the end of the word rang. where it contrasts with other nasals in words s uch as ram and ran, though the vowel sound in rang is a little dif- ferent in most varieties of English. 1bere is also a contrast between the consonants in the middles of mission and vision. although lhere a re very few pairs of word s

shown in the

k.ind. There are obviously many othe r words that rh yme with pie, s uch as

~. try, spry, but these

that rhym e. Take. for


that are disti nguis bed by thi s contrast in English . (One suc h pai r for some s peakers involves the name of a chain of islands--A leurian versus allusion.) Words illus-

trating these consonants

Mos t o f th e symbol s in Table 2. 1 are the same letters we use in spelling these

words, but there are a few differences. One differen ce between spelling and

are given in the fourth col umn o f Table 2.1 .

DTn 2

Phonoiogy and Phorlf'lic

DTn 2 Phonoiogy and Phorlf'lic ------ ----- ------ Symbols for transcribing English consonants. (Alternative
DTn 2 Phonoiogy and Phorlf'lic ------ ----- ------ Symbols for transcribing English consonants. (Alternative

------ ----- ------

Symbols for transcribing English consonants. (Alternative symbols that may be found in other books are given in parentheses.) The last column gives the conventional names for phonetic syrnbol~ in the firs! colurnn.




lowercase p






lowercase k



lowercase iJ



lowercase d


lowercase g








lowercase n



cng (or angrna)



lowercase f



lowercase v









lowercdse :,



lowercase z




esh (or long 5)




(or yogh)



lowercase I



lowerc.ase w



lowercase r

j (y)







Note dlso tile fOllowing:

Note dlso tile fOllowing:





d3 (dz)



phonetic usage occurs with the letter ,which is sometimes used to represent a (k] sound. as in cup or hacon, and sometimes to represent an [s] sound, as in cellar or receive. Two c's may even represent a sequence of l k] and [s] sounds in the same word, as in accent, access A symhol that sometimes differs from the corresponding letter is [g], which is used for the sound in guy amI gues~ but never for the sound in age or the sound in the name of the letter g. A few other symbols are needed to supplement the regular alphabet. The pho- netic symbols we will use arc part of the set approved the Int(;rnalional Pho~ • •

netic Association, a body founded in 1886 by a group of leading phoneticians from France, Gellnany, Blitain, and Denmark. The complete set of IPA symbols is in the char! on the inside covers of this book. It II discussed detail later in this book. Because we often need to talk about the symbols, the

names that have been given to them are shown in the last column of Table 2.1. •

with the tail of the letter 9 descending below the fine, Some people call this


The velar nasal the end of rang is written with (IJ], leHer n combined

The Transcnption of Consonants

symbol eng; others pronounce it angma. The symbol l9 J. an upright VersiOIl of me Greek letter theta, is used for the voice l ess dental fricati ve in words such as

elh, is derived

from an Anglo-Saxon letter. It is used for the corresponding voiced sound in

words sucb as thy, then. them, b~athe. Both these symbols are asceode.rs (letter'S that go up from the line of writing rather than descending below it). The spellin& system of the English language does not distinguish between (9) and r(I). They are both written with the le Iters Ih in pa irs such as thigh. thy.

voiceless pala to-alveolar (post-alveolar) fricative LJ j (long.1)

dligh, thin, thimble, ~ther. b~ath, mo uth. Th e sy mbol L0 J. caUed

The sy m bo l for the

in shy, sheep, rash is both an ascender and a descender. It is like along. straight-

ened s going both above and below the line of writing. 1lle correspooding voiced symbol [3] is like a lo ng z descending below the line. This sound occurs in the middle of words such as vi.1iQn. measure, leisure and at the beginning of foreign words such as the French JeOl1, genda rme. and foreign names such as Zsa ZuI.

In earlier editions of this book, the sound at the beginning of the word rye was sy mbolized by I J 1. an upside-dow n letter r. This i s th e correct IPA symbol for this sound but as lhe two major dictionaries of American and British English pronunciation (see "Further Reading") use a regular [r] for this sound. we have done so here. ft is unfortunate that different books on phonetics use different forms of phonetic transcription. Thi s is nOI because phon e lician s cannot agree on whic h symbols to use, but ralher because different styles of transcription are more ap- propriate in one circulUStance than in another. Thus, in this book, where we are concerned with general phonetics, we have used the lPA symbol (j J for the ini·

tial sound in yes, yet. ~asl because the IPA

sound, the vowel in the French word tu. Another reason for using (j 1 is thai in many languages (German. Dutch, Norwegian. Swedish. and olhers) this letter is used in words such asja, wh ic h are prono unced w ith a sound that in the English spelli ng sys tem would be written with th e leiter y. Books that are co ncerned only with the phonetics of English often use I y I where this o ne uses [jJ. Some

books on phonetics also use [~] and [zl in place of the IPA symbols {fl and [J I.

respec ti ve ly. Th e first and last so und s in bo th chu rch a nd judg e are tran sc ribed

with the digraph sy mbols [tI J and

a seque nce of a stop followed by a fricative (he nce the lPA sy mbols for them are

digraph s). yet they funct ion in English as if they are really a si ngle un it. com- parable in some ways to other stop consonants. You can see that a word such as chOQ.fe might be said 10 begin with (tI) if you com pare your pronunciation of the phrases whire shoes and why choose. In the first phrase, the [t J is at the e nd of one word and the II I at the beginning of the nex t; but in the second phrase, these two sounds occur together at the beginning of the second word. The difference

between th e two phrases is

fri cate in why choose has a more abrupt fricative onsel, and the timing of the stop and fr icative is more r igid than is the liming of the sequ e nce in white s hoes.

Also , for so me speakers, the final [t] of white may be said with s imultaneous

reserves the sym bol I y J for another

[d3J. These a ffri cate sound s are pho ne ti cally

one of the timin g o f the arti culatio ns invo lved. The af-



1 8 C H A P T E R 2 !anscription [t] in the affricate [tJ


1 8 C H A P T E R 2 !anscription [t] in the affricate [tJ
1 8 C H A P T E R 2 !anscription [t] in the affricate [tJ

[t] in the affricate [tJ that demonstrate this versus he cheat5 versus my chop. There are illustrating the same point for the voiced counterpart [d3] found in jar, gentle. age, because no English word begins with [3]. Some other books on phonetics transcribe [tf] and [d3] (as in church and judge) with single symbols, such as [c] and [J]. These transcriptions highligbl the fact that affricates are single units by using a single letter to transcribe the~ We will see that some linguistic segments have two phonetic elements example, vowel i is usually helpful to rpr'lrF"'" elements in When we wish to make we are writing consonant cluster, the is used to tie the affricate in why choose

cluster, the is used to tie the affricate in why choose [tf] to distinguish [tJ] in
cluster, the is used to tie the affricate in why choose [tf] to distinguish [tJ] in

[tf] to distinguish

[tJ] in white shoes. The

begins words an initial vowel (recall Chapter 1 of the difference between flee east and fleeced) is written pho

cally with [?], a symbol based on the question mark. So flee east is pronou [fli?ist], whilefleeced is [flist]. The status of glottal stop as a consonant in English is questionable because its distribution is limited. Where other nants may appear in a variety of positions in words (e.g. note the [k] in cat, back, active, across, etc.), glottal stop only occurs word initially before in American English, In London Cockney, glottal stop also appears

vowels in words I [t J. In American

"glottalized" SLOp, or more usually

taneous glottal There is one

consonant contrasts of English. In most forms of both British and An English, which does not contrast with witch. Accordingly, both why and we Table 2.1 are said to begin simply with [w]. But some speakers of English trast pairs of words such as which, witch; why, wye; whether, weather. speakers will have to transcribe the first consonants of each of these words with [hw]. Note that, phonetically, the [h] is transcribed before [

hu/ton where other dialects final [t] in words like

[kret? D. to be considered in the

that it is the first part of each of these words that is voiceless.


The transcription

more difficult

of English differ more in lheir use of vowels than in their use of eVB.,V Second, authorities differ in their views of what constitutes an appropriate -

differ in their views of what constitutes an appropriate - VOWELS vowels (the vowel phonernes) in


vowels (the vowel phonernes) in of consonants for two reasons.

(the vowel phonernes) in of consonants for two reasons. scription of vowels. Taking the same approach
(the vowel phonernes) in of consonants for two reasons. scription of vowels. Taking the same approach

scription of vowels. Taking the same approach in looking for contrasting vowels as we contrasting consonants, we might try to find a minimal set of words that

The TranSCription of Vowels

o nly in the vowel sounds. We could. for example. look for monosyllables that begin with I h J and end with ld Jand supplement this minimal set with other lists of monosyll ables that contrast only in their vowel so unds. Tab le 2.2 shows five such sets o f words. You should listen to the recordings of these words OIl tile CD while reading the following discussion of the vowels. We will consider one fonn of British and one form of American English. The major difference between the two is that spea kers of American English pronounce

I r I sounds after vowels. as well as before them. whereas in most fonns of BritI sh

speakers distin-

Eng li s h, I r J can occ ur o nly before a vowe l. Ameri can English

guish between words such as heart and hOI not by making a diffe rence in vowel quality (as in Peter Ladefoged 's form o f Briti sh Eng li sh). but rather by pronou nc- ing h eart with an [r I and hor with the same vow e l but without a n [r ] foll owing it . In he~. hair. hire, these speakers may use vowels s imilar to those in he. head,


I r I. Mos t s peakers of Brit-

ish Engli sh d isting ui sh these words by usi ng differe nt diphthongs- movements from one vowel to another within a single sy ll able.

respectively. bot in eac h case with a (ollow in g

Svmbols fOf transcribing con trastingable. respectively. bot in eac h case with a (ollow in g vowels in Engl ish.

vowels in Engl ish. Column 1 applies to

for the phonetic

many speakers of Amerian Engl ish, Column 2 to mosl speakers 01 British

English . The la SI column gives the conyentional names



in tM !irst column unless otherwise noted.










lower<ase i small capital I lowefca5e t epsilon


























"', d


(a rd

script 0 turned smpt a open 0 UPSIlon lowercase 0 lowercase IJ turned y






















0 0






, ,












" ~ .0






lowercase a (+I)




(as noted



" "

(. jhov


(as noted above) {as noted above] (as noted above) (as IlOted above)

" "



" "








Note .11:>0:








(as noted aooye)


(02 .2


Th~ Transcnplion of vowels


Two symbols that are not ordinary letters of the alphabet. Ie J and I re I. are

used for the vowels in head and hod. respecti vely. The first is based on the Greek

le tte r epsilon and

referred to by the names epsilon and ash. M os t Americans u se the s am e vowel SQund in the words Man and hot and can use o ne form of the letter o. They would transcribe these words as [bon J

and s peakers of British English who

do nm pronounce [r] sounds after a vowel distinguish between these words by the qualities of the vowels and hav~ to use two differeD! form s of the leuer o. They would transcribe these words as [ hOI Jand l hot J. Mo st s peakers of British forms of Eng li s h. and many Ameri ca n spea kers. distinguish between pairs of words such as cat, caught; not, naught. The symbol I :> J, an open lener 0, ma y be used in the seco nd of each of these pairs of word s

a nd in words s uch as bawd, boughl, law . Many Midweste rn and Far Western American speakers do not need to use thi s symbol in any of these words. as they do not distinguish between the vowels in words such as COl and caughl . They may have different vowels in words in which there is a foHowing [ r 1sound,

such as horse, hoarse, but if there is no opposition between COl. caught or not, Mughl. there is no need to mark thi s difference by using the symbol f :> I. Doing

principle of

showing just the differences between phonemes. Another special symbol is used for the vowel in hood, could. good. Thi s sy mbol . I u J, ma y be thought of as a letter II with the ends curled o ul. The vowel in hoe, dough. code is a diphthong. For most American English speakers, the flrst e le ment is ve ry si milar to sounds that are wrinen in Spanish or Itali an with the letter o. Many speakers of English from the southern parts of Britain use a different sound for the flrst element of the diphthong in these words.

which we will symbolize with I ~ I. an upside-down leiter e called schwa. We will discuss this sound more fu lly in a late r sectio n. The final element of tile diphthong in words such as hoe and code is somewhat similar to the vowel [ u I in hood. An up s ide-down letter v. 1A J. i s us ed for the vowel in word s s uch as bud. hut . Thi s sy mbol is sometim es ca ll ed wedge. Another symbol. [31. a reversed form o f the Greek letter epsi lon , is used for the sou nd in pen, bird. cun as prono unced by most speakers of British English and those speakers of American English who do no t have an I r J in these words . In most forms of Am e rican Eng li s h. the

the second on lhe le tters a and e joined together. 1lley may be

and I hot I. But some

East Coast Ame ricans

so would simply be showing extra phonetic detail , straying from the



fu ll y combined with the vowe l. and the sy mbol 13' J is used . The tillie hook



indicates the r-co loring of the vowel. The next three words in Table 2.2 contain diphthongs composed o f elements


that have been discussed already. The vowel in hide I hald) begins with a sound

between that of the vowel in COl

mov es toward the vowel [I J a s in hid I hid J. The sy mbol [a J is u se d for the flrst part of this diphthong. The vowel in how [au 1begins with a similar sound but mov es toward Lu J as in hood. The vowel in boy I b:>1 I is a co mbination of the

so und I :> I as in bawd and llJ as in hid.

[kre t J and that in hard [hod J or I hard J, and


Phonology and Phonetc Transcription

Most Americans pronounce the remaining words in Table 2.2 with one of the other vowels followed [r], while most BritislI English speakers have additional diphthongs in these words. In each case, the end of the diphthong is [;:,], the same symbol we used for the beginning of the diphthong in hoe for most British English speakers. We will discuss this symbol further the next para- graph. Some (usually old-fashioned) British English speakers also use a diph- thong in words like poor, cure that can be transcribed as [u~]. Some people have a diphthong a;:,] in words as hire r had Others pronounce these words as two syllables (like higher, liar), transcribing them as [fal~, hm~]. The words in Table 2.2 are all monosyllables except for ahoy. Consequently, none of thenl contains both stressed unstressed vowels. By the most com-

mon unstressed vowel is [~j, the one we noted at the end of some of the diph- thongs in British English. It is often called by its German name, schwa. It occurs

at ends words such as soda I 'souL-:l. 'soudd in the

middles words

such as emphasis, demonstrate ['emfgSIS, 'demgnstreltJ, and at the beginnings of words such as around, arise rg'raund, g'rarz]. (In all these words, the symbol [']

is stress that been placed hefore the syllabic carrying the main stress. Stress should always be marked in words of more than one syllable.)

In British English, r ~] is usually the sale component of the -er part of words such as broTher, brotherhood, simplerbrA50, 'sImpi;)]. In forms of

American English with r-colored vowels, these words are usually ['brA5;:}',

'brA5;:}'hud, 'srmpld'-l As with the symbol [3'-], the small hook on r d'- J symbolizes the r-coloring. Both and are common vowels, [~] occurring fre- quently in unstressed monosyllables such as the grammatical function words the, a, to, and, but. In connected speech, these words are usually [5~, ~. t;), ~nd, b;)t]. Some of other vowels also occur in unstressed syllables, but because of differences in accents of English, it is a little more difficult to say which vowel occurs in which word. For example, nearly all speakers of English differentiate between the last vowels in Sophie, or patter But some accents have the vowel [i] as in heed at the end of Sophie, pity. Others have [J 1as in hid. Similarly, most accents make the vowel in the second syllable of taxis different from that in Texas. Some have i] and some [I] taxis. everybody

pronounces Texas as r'teksgs]. (Note that in English, the letter x often represents

f the sounds [ ].) Compare your pronunciation of these words with the reeord-

CD on the CD and uecide which unstressed vowels you use. This is an appropriate moment to start doing some transcription exercises. There are large munber of them at end this you have grasped the basic principles, you should the of exercises.



So far, we have been using the consonant and vowel symbols mainly as ways of representing the contrasts that occur among words in they also be thought of in a completely different way. may regard them as shorthand

Consonant and Vowel Charts

descriptions of the articulations involved. Thus. Lp J is an abbreviation for vo;c~­


arranged in the form of a chart as in

Figure 2. 1. The places o f articulation are s hown across the top o f the chart, start-

ing from th e mOSt forward arti c ulat ion (bi

made in the back of the mouth (ve lar) and in the throat (glollal). The manners

of articulation are s bown o n the vertical axis of the chart. By co nve nti on, the

is shown by putting the voice less symbols to the

left of the voiced symbols. 1be symbol rw J is shown in two places in the consonant c hart in Fi g ure 2. I . This is because it is articulated with both a narrow in g of the lip aperture. which makes it bilabial. and a raising of the back of the tongue toward the soft palate,

labia l ) and going toward those sounds

l~s$'bilab;alstop and JI J is equivalent The consonant symbo ls can then be

to voic~d alv~o/ar lateral

voiced-voiceless di stinction

which makes it ve lar. The affri ca te

rately in the tab le even thougb they are co ntrasti ve sounds in English . Noc.e that

if we were to include them in the tabl e. we would have the problem of deciding

th e fricative

element) or in the alveo lar column (the place of the stop e lement ). The interna- tional phonetic alphabet avoids the inaccuracy that is inevitable when the stop

the affricate hav e d ifferent place of arti culation

whether to put them in the palato-alveolar co lumn (th e place of

) a re not li s ted sepa-

sy mbo ls [tf J and (d3

ele ment and fricative e leme nt of


by li stin g only stop and fricati ve symbo ls in the consonant chart .

Figure 1 . 1

A p honetic chart 01 the English consonants we have deal t with so far , whenever there are two symbols within a !oingle cell , the one on the leh repr~ts a vokeless sou nd . All other symbol s represen t voiced soonds.

also the consona nt I h I. which is not on thi s chart , and the affrica tes I [f. dJl. which art sequences of symbol s on tht chart .


rwaJ (stop)




ii frica ti ve







Place of anicula[ion

approximant la[era] (approximant ) Place of anicula[ion " 6-== 0 _ •• ;; • - •




















































Phonology and Phonetic Transcription

Figure 2.2

A vowel chart showing the relative vowel qualities represented by some of the symbols used in transcribing English. The symbols [e, a, 01 occur as the first elements of diphthongs.


[e, a, 01 occur as the first elements of diphthongs. front high mid-high mid mid-low central






u \ \ \ \ \ I \ I u ~-------L---- \ I 0 \

























The symbols we have been using for the contrasting vowels may also be re- garded as shorthand descriptions for different vowel qualities. There are prob- lems in this respect in that we have been using these symbols somewhat loosely, allowing them to have different values for different accents. But the general val- ues can be indicated by a vowel chart as in Figure 2.2. The symbols have been placed within a quadrilateral, which shows the range of possible vowel qualities. Thus, [i] is used for a high front vowel, [u Jfor a high back one, [ I ] for a mid- high front vowel, [e] for a raised mid-front vowel, [e ] for a mid-low, and so on. The simple vowel chart in Figure 2.2 shows only two of the dimensions of vowel quality, and if they are taken to be descriptions of what the tongue is do- ing, these dimensions are not represented very accurately (as we will see in later chapters). Furthermore, Figure 2.2 does not show anything about the variations in the degree of lip rounding in the different vowels, nor does it indicate anything about vowel length. It does not show, for example, that in most circumstances, [ i] and [u] are longer than [ I ] and [ u ]. The consonant and vowel charts enable us to understand the remark made in Chapter 1, when we said that the sounds of English involve about twenty- five different gestures of the tongue and lips. The consonant chart has twenty- three different symbols, but only eleven basic gestures of the tongue and lips are needed to make these different sounds. The sounds [p, b, m ] are all made with the same lip gesture, and [t, d, n] and [k, g, I)] with the same tongue gestures. (There are slight differences in timing when these gestures are used for making

the different sounds . but we will neglect them here.) Four more gestures are requ.ired for the sounds in me frica ti ve row. three more for the (cenlJ'al) approxi- mants . and another one for the lateralllpproximant . making eleven in all 1be vowel chan has fourteen symbols. each of which may be considered to require. separate gesture. But. as we have seen, accents of English vary in the number of vowels that they distinguish, wh ich is why we said that English requires tJbovt twenty-five different gestures of the tongue and lips. All these sounds will also require gestures of the other three main com~ nents of the speech mechanism-the airstream process. the phonation process. and the oro-nasaJ process. The airstream process involves pushing air out of the lungs for all the sounds of Eng li s h. The phonation process is responsible for the gestures of the vocal folds that distinguish voiced and voiceless sounds, and the oro-nasal process will be active in raising and lowering the velum so as to distingu.ish nasal and oral sounds.


At the beginning of this chapter. we discussed another reason why it is only

English. the symbols have the

approximately true that in our transcription s of

valUe!i shown in Figures 2.1 and 2.2. In the style of transcription we have been using so far, we have used sy mbols that show just the contrasting sounds of Engli s h. the phonemes. From this point on. we will use slash lines II to mark off symbols when we are explicitly using them to represent phonemes. As we have noted, some of the phoneme sy mbols may represent different sounds when they occur in different contexts. For example. the symbol It I may represent a wide varie ty of sounds. In tapluep I. it represents a voiceless alveolar

stop. But the I t I in eighfh/ellSI may be made on the teeth. because of the influ - ence of the following voiceless dental fricative 161. This I t I is more accurately called a voiceless dental stop, and we will later use a spec ial symbol for tran-

sc ribing

is accompanied by a g lottal stop. and we will also be us ing a special symbol for thi s so und. As we saw. for mOSt Americans and for many younger Brit ish English speakers. the I t I in calty I'k<etil symbolizes a voiced. not a voice less. sound. All these different sounds are pan of the It I pho neme. Each of them oc-

c urs in a specific place: It I refore 161 is a dental StOp. I t I before a word final Inl

is a g lottal stop, and I t I aftcr a vowel and before

s top. None of these variations is different enough to change the meaning of a word in English. Note also that a ll of these variations occur in citation speech and are not simply the result of failing to "hit the target" when speaking quicldy. Simi larl y, other symbols represent different sounds in different contexts. The symbols III and Irl normally stand for voiced approximants. But in words such as ply I plall and lry I tral /. the influence of the prec eding s tops makes them voiceless. Vowel sounds also vary. The I i I in heedl hid l is us ually very differenl from the iiI in heel/hill, and much longerthan the Ii I in heal.

il . In most forms of both Britis h and American EngliSh, the III in bitten

an un stressed vowe l is a voiced


~?TER 2

Phonology and Phonetic Iranscriptlon

Many of the variations we been can describcd in tcrms of simple statements about regular sound patterns. Statements of this kind may be considered rules that apply EngJish words. In most forms of American English. for example, I I becomes voiced not only in catty, but on occasions when it occurs immediately after a vowel and before an unstressed vowel (for

.). In English nearly all kinds, it

is also a rule that whenever I tl occurs before a dental fricative, it is pronounced as

a dental stop. We can show that this is a different kind of It! by adding a small mark under it [t (As this symbol not representing phoneme, it is placed between [j.) The same is true of Id/, as in width [wlgo]; Inl, as in tenth [tqJ.8]; and Ill, as in wealth [w£!8]. In all these cases, the mark

] may be added undcr the mbo indi that represents a dental articulation. Ali these transcriptions are placed between square brackets, as they are phonetic transcriptions rather than phonemic transcriptions. Small marks that can be added to a symbol rnodify ils value are knovvn as diacritics. They provide a useful way of increasing the phonetic precision of

a transcription Another diacritic, r0 ], a small circle beneath a symbol, can be

used indicate that the symbol a voiccless sound. Earlier, we notcd that the I I I in play is voiceless. Accordingly, we can transcribe this word as

[p! ell Similarly, ply and try can

When we describe sound patterns that o(:cur in English, want be able to say that in some sense there are always the same underlying sounds that are changed because or tile contexts in which OCCUL The phonology a language is the set of rules or constraints that describe the relation between the underlying sounds, the abstract units called phonemes described at the be- ginni of this chapter, and the phonetic forms Ihat can be observed. When we transcribe a word in a way that shows none of the details of the pronun- ciation that are predictable by phonological rules, we are making a phonemic transcription. The variants of the phonemes that occur in detailed phonetic transcriptions are known as al]ophones. They can be described as a result of applying the pho- nological rules the We now discussed~ome the rules for different allophones of the phoneme I tI. For example, we know that in most varieties of American English, I tl has a voiced allophone 'vvhen it occurs between a stresscd vowcl and an 1Instressed voweL We have also illustrated rules that make /rl and /II voiceless when they occur after Ip, t, kl. (These rules need more refinement hefore they can considered to generally applicable.) In addition applying rules lhat describe particular allophones of the nemes in a transcription, there is another way we can show more phonetic detail. We can use more specialized phonetic syrnbols. example, we noted that the

vowel III is longer than the vO'vvel/l I, as in sheep versus ship. This difference in length is always there as long as the two vowels are in the8fUTle phonetic context


transcribe this difference in length by adding a length ma·.o the longer of the

example, in matter, utter,


be written [p! a1 1 and [


(betwcen the same sounds and wilh the same of Slress, etc. We


two sounds. The (PA provides the symbo l' : J to show !hat the preceding symbol represents a longer sound. Accordingly, we cou ld lnlnSCribe lhe two sounds as I i: I and III. We would sti ll be rep resenting only the underlying phonemes in !his particular accent of English, bUl doing so wim greater phonetic precision. Another eumple of using more precise phonetic symbols to show more phonetic detail has to do with me transcription of English Ir/. We mentioned !hat in previous editions of this boo k, we used the upside-down r J J 1to write the r sound of English. This was done because the lPA symbol rr J indicates a trilled r and not the approximant r of English. One principle of the International Phonetic Alp habet is to use the most co mmon form of the letter for !he most common phonetic property associated with that letter. Because trilled r is more common in languages of the world than is approximant r. the IPA uses the un· usual symbo l f J] for the unusual r sound found in English. So, you can use J J J to give a more precise transcription of the Eng li sh I r l. Students sometimes also make the mi stake of thinking that allophones are written with diacritics while phonemes are written wi th simple phonetic sym- bols. Consider. though, the pronunciation of the word letter. For most speakers

of American English. the re is no r t J sound in thi s word. In stead, the medial conso nan t sounds like a very s hort rd 1- It is different enough from (d 1 (co mpare seedy and set! Dee) that the lPA has a unique symbol for the tap allophone of It I and Id/. The alveolar tap sound in letter is written with the symbo l {r]. a letter derived from the letter r, Note. therefore, that transcription of allophones may

use si mple phonetic symbols as

we ll

as sy mbol s with diacritic marks.


'The term broad transcription is often used to designate a transcription that

a narrow transcription is more speci fic sy mbol s or

uses the simplest possible set of sy mbol s. Conversely,

one that shows

by representing some allophonic differences.

trip would be I pli zl and ItIlp/. A narro w (bu t still phon emic) transcription could be I pli :zl and Imp!. Thi s transcription would be phonemic as long as we always used I i:1 wherever we would otherwise have had IiI. In thi s way, we would not be showing any all ophones of the phonemes. A narrow al lophonic transcription would be [pli :z J and rttlP I. in which rI J and ({ I. are allophones of I I I and Ir/. Every transcription should be cons idered as having two aspects. one of which is often not explicit. 'There is the phonetic text itse lf and. at least implicitly. there is a set of conventions for interpreting the text . These co nve ntion s are usually of

two kinds. First. there are the co nven tion s that ascribe general phonetic values to the sy mbols. It was these conventions we had in min d when we said earlier that a symbol could be regarded as an approximate specifica ti on of the articulations involved. If we want to remind people of the implicit statemen ts accompany-

ing a transcription. we can make them expli cit.

othe r thin gs being equa l. I iI is longer than III. perhaps stating at the beginning of the transcription I i i =l i: /. We cou ld a lso make explicit the rules that spec ify the allophones that occur in different circ umstances. a topi c we will return to in Chapter 4.


phonetic detail. eithe r by usin g

A broad transcript ion of please and

We cou ld, fo r in stance, say that,


Phonology and Phonetic Transcription

few occasions,


be said to

the existence

accounting for


theoretically possible in the

of a nan-ow transcription so detailed that it shows all the rule-governed alterna-

tions among the sounds. A transcription that shows the allophones in this way is called a completely systematic phonetic transcription. In practice, it is dif- ficult to make a transcription so narrow that it shows every detail of the sounds

some occasions.


not imply the existence of rules

counting for allophones because, in the circumstances when the transcription was made, nothing was known about the rules. When writing down an unknown

language or when transcribing the speech of a child or a patient not seen previ-

does not

what rules

In these circurnstances, the

hols indicate only value sounds. This transcription impression istie transcription. We hope this brief survey of different kinds of transcription makes plain that

there is no such thing as the IPA transcription of a particular utterance. Some- times, one wants to make a detailed phonetic transcription; at other times, it is

convenient to

phonemic transcription.

one wants

a particular


as vowel

other times.

are not of concern and details transcriptions take many fonns.

consonants are



(Printable versions of all the exercises are available on the CD.)

A. Each of the following words contains an error in transcription of vowel

There is

one possible

but because

differences in


sometimes alternative possihle corrections. :tYlake

correct transcription in the space provided after each word.

1. man-made


should be

2. football



tea chest

['titfest 1


[ltomkret J



[ 'tiptou]



[ re'v::lId]



[rd'man 1


['bedr;)m 1



10. manage