1. E l i s a b e t h G a re i s
A AMERICAN ENGLISH
WIDE VARIETY OF PRONUNCIATION TEXTBOOKS ON ARE ON
the market today, allowing teachers to select the most suitable materials for their
ESOL students. This abundance of texts would be ideal, were it not for some
inconsistencies in the presentation of pronunciation features. Thus, textbooks
often differ in the use of pronunciation symbols, the description of individual
vowels and consonants, the number of sounds characterized as diphthongs, and
the delineation of pitch and stress patterns. Given the fact that many ESOL stu-
dents move through several texts in the course of their studies, these discrepan-
cies are confusing and can create insecurity among students as well as teachers.
This article provides discussion of the most common discrepancies, exempli-
fied in five popular pronunciation texts: Accurate English, Pronunciation Pairs,
Phrase by Phrase, Sound Advantage, Speechcraft, and the Manual of American Eng-
lish Pronunciation. The texts were selected for their range from beginning to
V 43 N 2 2005 E T F
18 OLUME UMBER NGLISH EACHING ORUM
2. advanced level and for their diversity in nota- States may have reduced the perceived need
for an all-enveloping sound description sys-
tion and content. (Note: No criticism of the
tem, like the IPA.
selected texts is intended. All of the listed
books are internally consistent and represent
excellent guides for pronunciation instruc-
If discrepancies were limited to monolingual
tion.) Before discussion of the pronunciation
versus bilingual language materials, the relevance
texts, we’ll take a brief look at related discrep-
of this issue for ESOL students—especially at
ancies in monolingual dictionaries.
the beginning and intermediate level, where stu-
While this article focuses on U.S. English,
dents seldom use monolingual dictionaries—
similar discrepancies exist in publications on
would be minor. Unfortunately, discrepancies
other varieties of English. Teachers are encour-
can also be found within the realm of foreign
aged to compare the findings in this article with
and second language materials. Since deviations
their teaching materials and to explore incon-
are probably more confusing with respect to
sistencies in their own linguistic environment.
pronunciation texts than bilingual dictionar-
Monolingual dictionaries ies—with students going through several text-
books as their proficiency increases—the fol-
The idiosyncratic use of pronunciation sym-
lowing discussion uses pronunciation texts as
bols in monolingual American English dictio-
examples. Discrepancies with respect to vowel
naries has been a long-standing source of frus-
sounds will be discussed first, followed by con-
tration for language students and teachers
sonants and prosodics.
alike. Sound transcriptions differ from dictio-
nary to dictionary and bear little or no resem- Vowels
blance to the international pronunciation In general, ESOL pronunciation textbooks
alphabet (IPA), which is standard in most bilin- adhere to IPA standards to a great extent.
gual dictionaries and pronunciation textbooks. Among the vowels, for instance, the following
The vowel sound in the word tap, for example, sound symbols enjoy widespread agreement and
is consistently transcribed as /Q/ in bilingual are used consistently, without much variation:
/I/, /E/, /Q/, /´/, /ç/, and /U/. There are a few
dictionaries and pronunciation texts. In mono-
sounds and intonation features, however, that
lingual U.S. dictionaries, however, transcrip-
seem to incite such strong passions as to war-
tions usually involve a variation of the letter a
rant idiosyncratic notations. When these devia-
as the preferred symbol. The American Heritage
tions occur, they often represent not only devi-
Dictionary, for instance, uses /a(/ and Merriam-
ations from IPA notation but also diversity in
Webster’s Collegiate Dictionary (Webster’s) uses
opinion over the specific nature of the sounds
/a/—a symbol which in many bilingual dictio-
naries and pronunciation textbooks stands for
Table 1 (next page) shows a selection of vowel
the vowel sound in top, rather than tap. Similar
sounds as they are treated in various pronunci-
confusion reigns for /i/ which in most bilingual
ation texts. Also included are two monolingual
dictionaries and pronunciation texts stands for
dictionaries to illustrate the additional dichoto-
the vowel in teen but in Webster’s is used for the
my between the ESOL and monolingual realm.
vowel in tin. Likewise, /j/ stands for the begin-
Naturally, textbook authors have reasons
ning of the word year in many foreign language
for their choices. Sometimes the various choices
education materials but is used in Webster’s for
and their reasons are provided in the text; other
the beginning of the word jeer. These are just
times they are not. It is when no explanation is
a few examples. Discrepancies also exist with
given that students may become confused. The
respect to other sounds.
following is an attempt to explain the discrep-
Students often ask why monolingual Amer-
ancies in the vowel section of Table 1.
ican dictionaries don’t use the IPA. The answer
probably lies in the same U.S. insularity that /i/ and /u/
The sound /i/ is transcribed /iy/ or /iy/ in some
has resulted in so many other deviations from
international standards (including standards pronunciation texts to indicate that, especially
of measurement, weight, and temperature). In in the Eastern United States, there is a tongue
addition, the relatively minor importance of movement at the end of the sound, rendering
foreign language education in the United it a glide rather than a pure vowel. For the
E T F V 43 N 2 2005 19
NGLISH EACHING ORUM OLUME UMBER
3. SOUND TYPES COMPARISON OF PRONUNCIATION SYMBOLS
AND SAMPLE WORDS SELECTED PRONUNCIATION TEXTS AND MONOLINGUAL DICTIONARIES
AE MP PP PBP SA SC VD AH W
i iy iy iy iy i e# e#
A a A A a A A/Å
´ ´ √ √ ´ √ √ u( ´
´ ´ ´ ´ ´ ´ ´ ´ ´
u uw uw uw uw u oo ü
aI ay ay Ai ay ay aI quot;‹ quot;‹
aw aw aU/AU ou
aU aw aw Au o#
çI çy çy çi oy oy çI oi
eI ey ey ey ey e a# a#
oU ow ow ow ow o/oU o# o#
y y y y y y j y y
’ ’ ’
´’ ´r Œr ´’ ´r Œr Œ’ u^r ´r
’ ’ ’
´’ ´r Œr ´’ ´r ´r ´’ ´r ´r
AE = Accurate English • MP = Manual of American English Pronunciation • PP =
Pronunciation Pairs • PbP = Phrase by Phrase • SA = Sound Advantage • SC =
Speechcraft • VD = Voice and Diction
AH = The American Heritage Dictionary • W = Webster’s
same reason, /u/ is sometimes transcribed /uw/ of these symbols are often used interchange-
or /uw/ (Ladefoged 1982). ably in language texts.
One reason for pronunciation textbook au- The difference between /a/, /A/, and /Å/ is
thors (or editors) to eschew the glide notation that the sounds constitute a progression from
and use /i/ and /u/ instead is that they may be front to back vowel. Specifically, /a/, which is
from a part of the United States other than the farthest in the front, is used at the beginning
East and therefore prefer the pure vowel tran- of diphthongs; the low, tense /A/ is the sound
scription. In addition, there is the argument most Americans use for the vowel in balm; and
that /iy/ and /uw/ are based on phonemic (not the low, lax /Å/, which is farthest back and more
phonetic) analysis and that glide transcriptions rounded, is accompanied with a slight round-
mix vowel and consonant (or semi-vowel) sym- ing of the lips and found mostly in New Eng-
bols, therefore making them undesirable. land and British speakers, as in bomb (Eison-
The reason for the transcription discrepan- son 1992; Ladefoged 1982).
cies is therefore a combination of regional pre- In the case of the three a-sounds, the rea-
disposition, linguistic philosophy, and ulti- son for the discrepancies seems to be one of
mately personal choice. desired simplification. The symbol /a/ or /A/ is
often used as a stand-in for all variations, so
that students are spared from, perhaps unnec-
Moving down the vowel chart from /i/, the
next controversial sound is the /A/, as in palm.
The IPA differentiates between three low, albeit /ə/
similar vowels and recommends three different The issue of the schwa may be the most
symbols: /a/, /A/, and /Å/. At least the first two vexing of all discrepancies. ESOL students are
VOLUME 43 NUMBER 2 2005 E T F
20 NGLISH EACHING ORUM
4. often told that the schwa (or /´/) is the most scription of /o/ more logical) and a wider range
common sound in English, used for the major- in other regions of the United States (giving
ity of unstressed vowel sounds. However, some more support for the use of the symbol /oU/)
textbooks differentiate between the unstressed (Ladefoged 1992). Other times, it is not clear
schwa and the stressed, slightly lower and more whether variations indicate personal convic-
back sound /√/, while others use the symbol /´/ tions of the authors or whether choices are
for both. In other words, some textbooks main- made to simplify sound descriptions and limit
tain that the vowel sounds in the words above text-internal symbol variety for the sake of
and custom are different (i.e., that the stressed low- or intermediate-proficiency students.
syllable is pronounced /√/ and the unstressed Consonants
one /´/), while other textbooks recommend Pronunciation textbooks exhibit fewer dis-
pronouncing them identically as schwas for all crepancies with respect to consonants than vow-
syllables. Proponents of the latter pronuncia- els. If there are problem areas, they seem to be
tion argue that it is mostly British speakers who of relatively minor relevance as well.
use the /√/. While the difference between /√/
and /´/ may be small, one reason to retain the
ESOL students who are familiar with the
distinction is that it forces students to focus on
markedly different r-sounds in other languages
the important features of stress and vowel re-
might find the use of the symbol /r/ for the
duction in American English (Edwards 1992).
American r-sound slightly confusing. In IPA
In a related matter, a more minor deviation
notation, the symbol /r/ actually indicates a
can also be found concerning final -ed and -es
trill, as in Spanish perro (dog), for example. The
(as in rented and houses). Some texts recom-
American r-sound (variably described as a glide,
mend a pronunciation of /Id/ and /Iz/, others of
liquid, semi-vowel, or approximant) would be
/´d/ and /´z/. Yet others give a choice between
more accurately transcribed as /®/ (Edwards
the two versions, stressing that both /´/ and /I/
1992; Ladefoged 1982). Simplification seems
are common as reduced vowels. No specific
to be the reason for the substitution.
reasons for the discrepancy is apparent; the
matter seems to be one of personal choice.
Maybe the most confusing IPA deviation can
be noticed in the frequent use of /y/ for the
The last vowel discrepancy is the question
first sound in year. According to the IPA, /y/ is
of how many diphthongs exist in American
the rounded vowel sound found in the French
English. Some textbooks count five main
une or German über. For the beginning sound
diphthongs (/aI/, /aU/, /çI/, /eI/, /oU/); others
in year—which is variably described as a con-
count only three (/aI/, /aU/, /çI/) and consider
sonantal or nonconsonantal sound (Edwards
/ey/ and /ow/ glides, similar to /iy/ and /uw/.
1992)—the symbol /j/ should be used. Pre-
In other words, /eI/ and /oU/ are considered
sumably, the reason for using /y/ instead is the
full diphthongs by some, while others maintain
similarity of this symbol with the letter y, as in
that the second vowel sound is not fully devel-
year. This connection is supposed to allow stu-
oped and the whole sound is therefore more of
dents a more intuitive approach to the pro-
a glide. The book Voice and Diction goes even
nunciation of this sound.
farther and transcribes the two sounds as /e/
and /o/—with the rationale that they are non-
Pronunciation texts vary in their transcrip-
phonemic diphthongs; i.e., there are no mini-
tion of the r-colored (or rhoticized) vowels
mal pairs differentiating diphthong and pure
found in words such as curler. In different texts
/Œr/, /´r/, /Œ’’/, and /´’’/ are used to transcribe
Disagreement also abounds on the starting
the sound. The first issue here concerns the
and ending vowel of these sounds. For the be-
number of symbols in the transcription. Pro-
ginning of /aI/, for example, symbols vary from
ponents of a single symbol for the sound com-
/a/ to /A/; for the end of /aI/, the transcriptions
bination (/Œ’’/ and /´’’/)—which is also recom-
/I/, /i/, and /y/ are used. Sometimes, the choice
mended by the IPA—maintain that the sound
is influenced by regional differences. The
is a phonetic reality and should therefore be
diphthong /oU/, for example, has a relatively
limited range in the Midwest (making a tran- represented as such (Edwards 1992).
E T F V 43 N 2 2005 21
NGLISH EACHING ORUM OLUME UMBER
5. The second discrepancy concerns the issue Terminology
Discrepancies in the realm of prosodics
of stress. Some texts differentiate between
start with confusing terminology. The nomen-
stressed (/Œr/ or /Œ’’/) and unstressed er-sounds
clature used for the stress on a syllable (e.g.,
(/´r/ or /´’’/). The symbols /Œr/ and /Œ’’/ there-
the third syllable in the word informátion), for
by indicate a higher tongue position and are
example, includes the terms syllable stress and
used for stressed sounds, whereas /´r/ and /´’’/
word stress; whereas terms used for the main
have a lower tongue position and are used for
stress in a thought group or sentence range from
unstressed sounds. The word curler would
sentence stress, to phrase stress, information focus,
thus be transcribed / kŒrl´r/ or / kŒ’’l´’’/.
and—again—word stress (e.g., Eisonson 1992).
Other texts don’t differentiate between
The term word stress can therefore indicate
stressed and unstressed er and use one of the
either type of stress, depending on the pronun-
four alternatives as a stand-in for all occurrences.
ciation text, which is indeed very confusing.
Since the difference between stressed and un-
Apparently, no prescription exists. One
stressed er is minor, this widespread simplifi-
hopes that, in the course of time, a single com-
cation is perhaps understandable.
mon usage would manifest itself and become
The allophonic difference between /w/ for
Stress and intonation features
the first sound in weather and /hw/ or /„/ for
Differences in diacritics for primary and sec-
the first sound in whether is minor as well. It is
ondary stress are also common (e.g., «infor'ma-
therefore only occasionally noted in pronunci-
tion, ínformàtion). In addition, textbooks differ
ation texts. The rationale is that the differenti-
in whether they indicate secondary stress or limit
ation seems to be disappearing in most forms
themselves to primary stress notations only.
of English, especially in frequently used
Similarly, the number of pitch levels in
words, such as what and when (Eisonson
American English is either not mentioned at
1992; Ladefoged 1982).
all or differs between texts. Ranges from three
to four or more levels are common.
Of slightly more relevance are discrepancies
There are also two different descriptions of
in the description of final stops: /t/, /p/, /k/.
the rising sentence intonation used for yes/no
While some texts recommend releasing final
questions, such as Do you want coffee? In some
stops, others describe them as unreleased. Yet
texts, the pitch is described as level until it rises
others give the speaker an option of releasing or on the last syllable (see Illustration 1 below
not releasing them. At play are issues of formal- left); in others, it is depicted as level followed
ity and careful articulation, as well as sound by a fall-rise; i.e., a dip to a lower pitch level
environment. Most people, for example, don’t prior to the rise (see Illustration 2 below right).
release final stops when the next word begins The latter was previously considered British
with a nasal, as in cat nap or with another stop, but is now often used in American English as
as in the cat pushed (Ladefoged 1982). well. Interestingly, while some pronunciation
Prosodics texts describe the yes/no question intonation as
Apart from discrepancies among vowel and a simple rise (as in Illustration 1), taped speech
consonant transcriptions, differences also exist samples accompanying the books may feature
in the treatment of prosodic features, such as fall-rise patterns and thus contradict the
stress and intonation. Many of the issues are textual description (e.g., Hagen and Grogran
outside the purview of the IPA and therefore 1992). It’s possible that such internal discrep-
lack a standardizing force. ancies are simply based on a difference between
(left) Illustration 1
(right) Illustration 2
Do you want coffee?
Do you want coffee?
VOLUME 43 NUMBER 2 2005 E T F
22 NGLISH EACHING ORUM
6. author and taped speaker; or they may be a ceive the vowel sounds in the word custom to
sign of changes in language use overtaking be different even though they are transcribed as
established transcription habits. In either case, the same sound in their textbook). Alternative-
teachers and students should be aware of the ly, instructors may preempt questions by teach-
two alternatives. ing a whole unit on the more controversial
issues. Such a unit could be as basic or com-
Conclusion plex as the students’ proficiency level allows.
Many more sound and prosodic discrepan- Since regional differences and personal pref-
cies exist than have been discussed in this arti- erences (e.g., desired level of detail, formality,
cle. The reasons for discrepancies vary. What and speed) seem to be major influences in the
follows are the most common explanations phenomenon of discrepancies, it may be help-
and one example for each. ful for teachers to diagnose their own speech
patterns and choose the transcriptions that
• Regional differences. For example, many
best match their pronunciation. If an instruc-
speakers in the Eastern United States dif-
tor pronounces the controversial vowel in teen
ferentiate between the vowels in Don and
without a glide, for example, /i/ would be the
Dawn; not so in the Western United States.
symbol of choice instead of the other alterna-
• Level of simplicity or detail desired in a
tives /iy/ or /iy/. Such a personalized transcrip-
publication. For example, the number of
tion system can provide consistency in the face
content word categories—which are im-
of textbook variations and even a level of sup-
portant for the stress-timed rhythm of
port for the instructor.
English—range from the basic list of
Instructors often teach two or more differ-
nouns, adjectives, adverbs, and full verbs
ent levels of students, with different textbooks,
to expanded, more comprehensive inven-
in the same term. Rather than having to re-
tories that include demonstratives, pos-
member how each book deals with the contro-
sessive and reflexive pronouns, negatives,
versial features and risk losing track of which of
questions words, and adverbial particles,
the controversial symbols to use in which class,
such as off in take off.
the teacher can thus adhere to his or her own
• Levels of careful articulation, often com-
system. Not only does this method relieve the
bined with British versus American ten-
instructor of having to recollect the countless
dencies. For example, the medial t pre-
variant combinations in the assigned textbooks,
ceding an unstressed syllable, as in city,
it also ensures that the teacher’s transcription
can be pronounced vigorously as a /t/ or
symbols match his or her specific speech pat-
as the quick, less vigorous intervocalic
terns, making lessons on the controversial fea-
tongue flap or tap /R/.
tures more authentic and coherent.
• Speed of articulation. For example, /t/ Thus, a personalized transcription system
after /n/, as in twenty, may be articulated will allow instructors to navigate between text-
in slow but not articulated in fast speech. books more easily and ensure cohesion be-
tween their modeling and transcriptions. In
Many of these discrepancies are minor in rel-
addition, the teacher’s use of his or her own
evance or may constitute levels of detail rather
choices can become a catalyst for addressing
than true differences. Others are more confus-
the question of variation. When teachers intro-
ing but are explained sufficiently for the read-
duce the issue of textbook discrepancies to their
er to make educated choices.
students, explain the points of contention, and
It is when controversial features are unex-
model their own and the alternative speech pat-
plained in the textbook and are not discussed in
terns, they make students familiar with impor-
the classroom that an already difficult subject
tant and interesting issues in pronunciation
can become unnecessarily taxing. Instructors
theory and practice, as well as raise awareness
should therefore be aware of the major discrep-
about the range of acceptable patterns (Levis
ancies found in pronunciation texts and dictio-
1999). A teacher may use a fall-rise pitch pat-
naries and be prepared to address them in class.
tern, for example, when discussing yes/no
This can be done in a variety of ways. Teachers
question intonation but point out that the text-
may opt for a case-by-case explanation when
students question a discrepancy (e.g., they per- book features the alternative level-rise pattern.
E T F V 43 N 2 2005 23
NGLISH EACHING ORUM OLUME UMBER
7. Students may then experiment with both pat- Eisonson, J. 1992. Voice and diction: A program for
improvement (6th ed.). New York: Macmillan.
terns and decide which one suits them better.
Hagen, S. A and P.E. Grogran. 1992. Sound advan-
Such experimentation can heighten a stu-
tage: A pronunciation book. Englewood Cliffs,
dent’s sense of linguistic self-awareness and pro-
NJ: Prentice Hall Regents.
mote a spirit of adventure in the classroom. It
Hahn, L. D and W. B. Dickerson. 1999. Speech-
also helps empower the students to become re- craft: Discourse pronunciation for advanced learn-
sponsible for their own learning. In the end, ers. Ann Arbor: University of Michigan Press.
this level of insight and the realization that Ladefoged, P. 1982. A course in phonetics (2nd ed.).
there are choices may well capture at least some New York: Harcourt Brace Jovanovich.
students’ imagination and lead them on a path Levis, J. M. 1999. Variations in pronunciation and
ESL teacher training. TESOL Matters, 9 (3):16.
of continued fascination with language and
Merriam-Webster. 1998. Merriam-Webster’s collegiate
dictionary. Springfield, MA: Merriam-Webster.
Prator, C. H. Jr. and B. W. Robinett. 1985. Manu-
al of American English pronunciation (4th ed.).
Baker, A. and S. Goldstein. 1990. Pronunciation
Chicago: Holt, Rinehart and Winston.
pairs: An introductory course for students of Eng-
The American heritage dictionary of the English lan-
lish. New York: Cambridge University Press.
guage (4th ed.). 2000. Boston: Houghton Mifflin.
Chan, M. 1987. Phrase by phrase: Pronunciation
and listening in American English. Englewood
Cliffs, NJ: Prentice Hall Regents. ELISABETH GAREIS is an associate professor
Dauer, R. M. 1993. Accurate English: A complete
at Baruch College (City University of New
course in pronunciation. Englewood Cliffs, NJ:
York), where she coordinates ESOL services
Prentice Hall Regents.
for international graduate students and fac-
Edwards, H. T. 1992. Applied phonetics: The sounds
of American English. San Diego: Singular. ulty members.
E T F V 43 N 2 2005 31
NGLISH EACHING ORUM OLUME UMBER