Anatomy of an Accent—an overview

A given accent or ideolect may be thought of as having four constituent parts. In accent
teaching or acquisition, all four should be examined, described, mastered, and integrated.
Because we’re fond of alliteration, we’ve named each of these elements with a /p/ word.
You might think of them as the four pillars of accent work.

1. People (cultural context)

2. Posture (oral or vocal tract posture)
3. Prosody (rhythm, stress, intonation; the music of speech)
4. Pronunciation (specific pronunciations; the inventory and distribution of
characteristic sounds)

The final three, Posture, Prosody, and Pronunciation, are the elements we can analyze,
break down, and teach. And it is these three that naturally form the main body of the task
of accent acquisition. The first element in the list—People, or Cultural Context—though
also crucial, is a sort of background stream of information, informing the imagination and
the whole of the endeavor.

Whether you are an actor working on an accent by yourself or a coach or teacher writing a
breakdown for a cast or a class, the four Ps are the elements to pay attention to. Focusing
on one at a time will help you focus in on what can be an overwhelming stream of
information as you listen to primary speakers.

1. People (Cultural Context)

Accent is identity. At a very deep level, the choices we have made over a lifetime (mostly
unconsciously) about how we will realize the sounds of speech reflect how we think about
who we are. There is also a sort of constant feeback loop between this felt identity and the
associations that arise in other people when they hear us speak. Though these associations
are cultural constructs—despite the fact that there is no inherent connection between
specific speech sounds and patterns on the one hand and character on the other—these
associations are nonetheless real, vivid, and endemic. They are, in fact, intrinsic to our very
nature as small-group primates. It may be (is!) wrong to decide that someone is
unintelligent (or friendly, or pushy, or hardworking, etc.) because of the accent they speak
with, but it is nevertheless the case that human beings do this habitually and pervasively.
Page 1
10.18.16 Erik Singer and Knight-Thompson Speechwork ©2016
Anatomy of an Accent—an overview

Accent acquisition is an act of imagination. No matter how accurately we may hit certain
pronunciation targets, we cannot speak believably in an another accent without being able
to imagine ourselves as someone who speaks that way. The history and culture of a place
and its people is an impossibly rich and complex subject, of course, but the deeper we delve
into it, the more opportunities we will have to engage imaginitively and empathically. Art,
fiction, memoirs, history, music, food, politics, sociology, and yes, even geography are all
areas of inquiry that will provide opportunities for this sort of imaginitive engagement.

A word of caution: there is a strong temptation to be reductive when considering the

connections between culture and language or accent. “People in Australia keep their
mouths closed so as not to let the bugs in,” “The spacious drawing rooms of the British
upper classes are reflected in the wide and carefully cultivated variety of vowel sounds in
RP,” “Welsh accents rise and fall with a gently rolling rhythm, just like the green, rolling
hills of Wales” are all examples of places we should really not go.

2. Posture
Vocal tract posture, oral posture—or just posture—can be thought of as the “home base”
for the articulators. Every accent and every language has a characteristic pattern of use, of
muscular engagement and release. It is posture that determines what the sounds in an
accent or language actually are—it is the foundation for all the speech actions that occur in
a given speech pattern. (Posture is often called articulatory basis by phoneticians).

Posture is the magic ingredient in accent work—no matter how accurate the individual
sounds are, you will never really sound like a native speaker of a target accent if you never
click into the posture. Conversely, if you can find the posture of an accent, that will do a
great deal of your work for you.

Though posture is ultimately something you must feel in your own vocal tract, it is helpful
to break it down in terms of its constituent elements. These include:

• jaw height (low or high)

• jaw protrusion or retraction
Page 2
10.18.16 Erik Singer and Knight-Thompson Speechwork ©2016
Anatomy of an Accent—an overview

• lip corner protrusion or retraction

• lip pursing (upper and lower lip approaching each other)
• cheek tensing (this can occur all the way along or be an isolated bracing right next
to the incisors and bicuspids)
• tongue arching/cupping in the front/middle/back of the tongue
• tongue spreading or ‘channeling’ (flattening out like a pancake)
• tongue bunching (thickening up like a hot dog)
• tongue bracing (usually against the upper molars)
• tongue root advancement or retraction
• raising or lowering of the velum
• widening or narrowing of the pharynx
• glottal settings (breathy, creaky, raised & lowered larynx, etc.)

Though precise attention to these variables can be extraordinarily useful, you might also
think and feel in terms of the shapes inside your mouth. Does your vocal tract feel wide or
broad in some particular way? Hollow? Narrow? Short and fat? Long and tubular? Some
people like to imagine the negative space inside their mouths in terms of an imaginary
object that might fit comfortably there. Is it like a hard-boiled egg? A tea saucer? A lump of
coal? A swizzle stick? A puddle of honey? Whatever it is, once you’ve found it—that’s the
shape of your new accent.

Just remember—oral posture is not a fixed position. It is a pattern of use, of muscular

action, tension and relaxation. Taken all in all, it is the degree, kind, quality, and patterning
of muscular engagement throughout the vocal tract. Though aspects of it may sometimes
be fairly pronounced in a given speaker, it is often quite subtle. The more you pay
attention to it, study it and practice it, the more useful it will become.

3. Prosody
The music of an accent, whether we call it prosody, intonation, or something else,
comprises both rhythm and melody. How does stress function in the accent? Is there a
typical rate of speech, or pattern of changing rates? Are there characteristic pitch
contours? How might we describe them? How are they deployed?

Page 3
10.18.16 Erik Singer and Knight-Thompson Speechwork ©2016
Anatomy of an Accent—an overview

3a. Pitch
Pitch contours—inflection patterns—will be the most salient aspect of prosody for most
accents. Despite their unarguable importance, descriptions of inflection patterns have
traditionally been sorely neglected in accent acquisition materials.

We need to pay just as much attention to intonation patterns as we do to vowel and

consonant sounds. We need to master them and integrate them into the accent. We can use
the same approach to this task that we use for mastering notable sounds: 1. identifying an
inventory of characteristic patterns, 2. describing them in detail and 3. using this specific,
detailed awareness to repeat and master the individual patterns.

We can go further, of course. Once we’ve identified some characteristic patterns, we can
look at where they tend to be used and what they may signify. (This is akin to the
distribution of characteristic vowel sounds. See below for more on this.) This may or
may not be necessary—with enough listening, the deployment and meaning of these
characteristic intonation patterns may be absorbed somewhat unconsciously, so that they
can be used by the actor in the accent without deliberate, conscious choice.

Pictures or diagrams are a useful way of describing these patterns. These could take any
number of forms, whether software generated or hand-drawn. Here is one possible
example:

220
Frequency (Hz)

all a- -bout fo- -xes

55
0 1.298
Time (s)

Page 4
10.18.16 Erik Singer and Knight-Thompson Speechwork ©2016
Anatomy of an Accent—an overview

Any such diagram should always be accompanied by an audio clip or clips, of course, as
above, whether embedded directly in a document alongside the picture or as an
accompanying audio file.

3b. Rhythm
One very useful way to focus on the way rhythm works in a language or accent is to look at
the interplay between stress and timing. (Linguists call this isochrony.) Roughly
speaking, most languages fall into two broad categories: syllable-timed and stress-timed.

Syllable-timed languages give the impression of taking an approximately equal amount of

time to speak each syllable. Speakers of syllable-timed languages will carry over this
characteristic rhythm to speaking English. Some examples of syllable-timed languages:
Spanish, French, Italian, Greek, Turkish, Polish, Hindi, Gujarati, Cantonese, Finnish,
Icelandic.

Stress-timed languages (like English) give the impression of taking an approximately equal
amount of time between stressed syllables, no matter how many unstressed syllables may
come in between stresses. Speakers of stress-timed languages are therefore accustomed to
speeding up and slowing down according to stress and syllable number. The timing is more
irregular. Some other stress-timed languages are Dutch, German, Danish, Russian, Swedish,
European Portugese, Arabic, and Mandarin Chinese (probably—it’s a little controversial).

4. Pronunciation
The specific realizations of sounds in an accent reflect an underlying system. This system
can be described in terms of the inventory and distribution of its characteristic
sounds. We can use the International Phonetic Alphabet (IPA) to precisely describe
the physical actions that produce the sounds. We can use lexical sets to describe the
distribution of these sounds across different classes of words.

Page 5
10.18.16 Erik Singer and Knight-Thompson Speechwork ©2016
Anatomy of an Accent—an overview

4a. The International Phonetic Alphabet (IPA)

The IPA allows us to describe the physical actions of various speech sounds with great
precision, something that is impossible with standard spelling. The basic rule for the IPA is
that for each unique physical action (and resulting sound) there is only one symbol, and
that each symbol represents one—and only one—unique speech action. So [k] represents
the sound at the beginning of the English word cold, and [s] represents the sound at the
beginning of the word ceiling, even though both sounds are spelled with the letter ‘c.’

Actors and students who wish to learn the IPA are strongly advised to read Dudley Knight’s
Speaking with Skill. Even with only a little bit of familiarity, however, the IPA can still be a
powerful tool for accent acquisition.

4b. The Vowel Space

We can describe vowel sounds very precisely in terms of two variables: exactly where your
tongue is inside your mouth, and whether or not your lips are rounding. We can indicate
tongue position in terms of this (slightly schematic) diagram of the space inside your
mouth:

To the left is the front of the mouth, and to the right is the back of the mouth. The roof of
the mouth is above the top of the vowel space (if the tongue touches the roof of the mouth,
we are making a consonant sound rather than a vowel sound). The body of the tongue
arches to get closer to the roof of the mouth, and it cups to get further away (thus creating
more space inside the mouth). If we draw a dotted line through the middle of this space,
then above the dotted line, we are talking about some degree of tongue arching, and below
the dotted line, we are talking about some degree of tongue cupping.

Page 6
10.18.16 Erik Singer and Knight-Thompson Speechwork ©2016
Anatomy of an Accent—an overview

A tongue position in the very middle of the space, where the tongue is neither arching nor
cupping, neither front nor back—where it is resting, in fact, completely relaxed—is
represented by the phonetic symbol ə. This is the only truly relaxed sound in human
language. Every other sound we can produce requires some degree of muscular
engagement.

By combining these variables—frontness, backness, closeness, openness, and lip-

rounding—we can rather precisely describe every possible vowel quality that human
beings can produce.

There are International Phonetic Alphabet symbols for various points on this chart, but the
really important thing is getting a physical sense of the space inside of your mouth. The
advantage, for accent work, of mastering this awareness (and the corresponding system of
description) is that it is purely physical. Rather than relying solely on your ear, which for
most people can be quite inexact, you can match up what your ear hears with an awareness
of the precise physical actions involved. If you know the IPA symbols, you can also then
notate these actions (sounds) for future reference, and also read others’ phonetic
transcription.

For a video walkthrough of the vowel space, see the video at: https://vimeo.com/76179402

4c. Lexical Sets

Devised by the phonetician John Wells, lexical sets are a handy tool for describing the
vowel inventory and distribution of different accents. Groups of words within a given set,
Page 7
10.18.16 Erik Singer and Knight-Thompson Speechwork ©2016
Anatomy of an Accent—an overview

(for example, the KIT set), will all be pronounced using the same vowel phoneme within a
given accent. So in “General” American, KIT words, like ship, sick, bridge, milk, myth, women,
will all be pronounced using an unrounded, near-close, near-front vowel [ɪ]. FLEECE words
like speak, key, leave, feel, ski, regal, will be pronounced with a fully front, close unrounded
vowel [i].

There are 24 standard lexical sets, though sometimes it is necessary to invent a new one or
two to describe features of a particular accent.

keyword RP "GenAm" other examples

1. kit ɪ̟ ɪ ship, sick, bridge, milk, myth, busy, women, lift, wisdom…
2. dress ɛ ɛ step, neck, edge, shelf, friend, ready, bless, death…
3. trap æ̞ æ tap, back, badge, scalp, hand, cancel, blast, flap, matter…
4. lot ɒ ɑ stop, sock, dodge, romp, possible, quality, hop, knob…
5. strut ɐ̠ ʌ̈ cup, son, budge, pulse, trunk, blood, mother, some…
6. foot ʊ ʊ̜̽ put, bush, full, good, look, wolf, would, could, cook…
7. bath ɑː æ staff, brass, dance, ask, sample, calf, after, half, last…
8. cloth ɒ̝ ɔ̜ cough, broth, cross, long, dog, toss, wrong, cost…
9. nurse ɜː ɘ˞ hurt, lurk, urge, burst, jerk, term, worst, curse, first…
10. fleece iː i creep, speak, leave, feel, key, people, mean, leap, regal…
11. face eɪ ̯ eɪ ̯ tape, cake, raid, veil, steak, day, vapor, great, maple…
12. palm* ɑː ɑː psalm, father, bra, spa, lager, calm, Khan…
13. thought ɔ̝ː ɔ̜ taught, sauce, hawk, jaw, broad, caught, thaw, brought…
14. goat əʊ̯ ɤ̈ʊ̯ soap, joke, home, know, so, roll, most, crow, float…
15. goose uː u loop, shoot, tomb, mute, huge, view, crew, flute, duplicate…
16. price aɪ ̯ aɪ ̯ ripe, write, arrive, high, try, buy, private, minus, choir…
17. choice ɔɪ ̯ ɔɪ ̯ adroit, noise, join, toy, royal, Flloyd, broil, adenoid…
18. mouth aʊ̯ aʊ̯ out, house, loud, count, crowd, cow, proud, lousy…
19. near ɪə̯ ɪɘ̯˞ beer, sincere, beard, fear, serum, weird, leery, career…
20. square ɛə̯ ɛɘ̯˞ care, fair, pear, where, scarce, vary, stare, clairvoiyant…
21. start ɑː ɑɘ̯˞ far, sharp, bark, carve, farm, heart, Tsar, tartan, bizarre…
22. north ɔ̝ː ɔɘ̯˞ for, war, short, scorch, born, warm, horse, normal, orbit…
23. force ɔ̝ː ɔɘ̯˞ four, wore, sport, porch, borne, story, hoarse, ignore…
24. cure ʊə̯ ~ ɔ̝ː ɔɘ̯˞ poor, tourist, pure, plural, jury, endure, moor, rural…
25. happY ɪ̝ i̞ city, party, money, century, coffee, taxi, calorie, twenty…

Page 8
10.18.16 Erik Singer and Knight-Thompson Speechwork ©2016
Anatomy of an Accent—an overview

25. lettER ə ə˞ mother, brother, daughter, scissors, another, quitter…

26. commA ə ə sofa, Africa, Alaska, about, another, telephone, kilogram…
*A better choice of keyword for this set might be SPA, as many speakers, especially Americans, articulate
some variety of /l/ in the word palm, ‘coloring’ the following vowel.

Lexical Set Mergers and Splits

Southeastern British English (SEBrE) and So-Called General American (SCGA) differ not just
in their inventory of sounds, but also in their distribution. Here is an example of how we
can use lexical sets to describe this situation.

When two lexical sets are realized with the same vowel sound, we can say that they are
merged.
Example: SCGA realizes both TRAP and BATH words as [æ]. We can say that in SCGA,
TRAP and BATH are merged.

When two lexical sets are realized with different vowel sounds, we can say that they are
split (or that they remain distinct, depending on your perspective!)
Example: SEBrE realized TRAP with [æ] and BATH words with [ɑ]. In SEBrE, TRAP
and BATH are split.

4d. Pronunciation targets

To specify a specific pronunciation target, we might use an entry like this:

1. GOOSE ⟶ [aːː]

This means that GOOSE words like shoe, intrude, move, scoot will be realized with a very open,
very long, front vowel. (Granted this is a very unlikely realization for this sound.)

A consonant entry might look like this:

2. /t/ ⟶ [w]

Page 9
10.18.16 Erik Singer and Knight-Thompson Speechwork ©2016
Anatomy of an Accent—an overview

Here we’re using slant brackets // to indicate the phoneme (the underlying mental
category the sound belongs in), and we’re saying that it is realized as a [w]. So tepid, in our
example, would be pronounced [wɛpɪd] (assuming the other sounds in the word are
realized similarly to those of American English).

Page 10
10.18.16 Erik Singer and Knight-Thompson Speechwork ©2016