To Marta

Contents

Preface

How to use this text

Introduction

Part 1: Making sounds

1 Silence

2 The space that shapes sound

3 From silence to sound

4 The muscles that shape sound: the tool kit

5 “Phthong” shaping

6 Obstruents: obstructing the flow

Part 2: Finding language

7 Exploring the limits: Outlandish

8 Obstruents within language

9 The physical actions of obstruents in language

10 The empty obstruent chart

11 Making your own language

An interlude: a daily articulation warm-up

Part 3: Phonetics
 12 Writing it down

13 Writing the consonants

14 Sound to word

15 Writing the vowels

16 American vowel phonemes

17 New vowels for diphthongs

18 All the vowel phonemes

19 Combinations

20 Mid-central offglide diphthongs

21 Polysyllabic words

22 Connected speech

Part 4: The skills of intelligibilty

23 Formal and informal speech

24 Consonant skills

25 Vowel skills

26 Putting it all together

Audio Tracks

Footnotes
Acknowledgements
eCopyright
Preface

Principles, perhaps
Whether they are called techniques, systems, approaches, methods,
explorations, or any other label, and whether their practitioners admit to it or
not, anything we teach in the arts is based on some set of underlying
propositions. If these propositions are examined regularly, poked at
occasionally to enliven them, soundly thrashed if required, rearranged when
necessary, dusted off, polished, and generally looked after, the owner can
offer them – all shiny and nearly new – to the partaker as principles. If they
are neglected or ignored they gradually atrophy into inert assumptions. I hope
that the following list will bear scrutiny as the principles on which this text is
based and that the text will bear witness to the principles.

1. This work tries to respond to the genuine needs of the actor or vocal
performer as related to speech. The text attempts to stay true to the
actual, practical needs of the actor as distinct from applying a system
based on the convenience of the teacher.
2. This work is based on the observation that the vocal and speech
needs of the actor within performance are constantly changing and
are never fixed. The actor, as an enactor of human behavior, is a
scavenger of all behaviors and therefore of all speech actions. They are
all useful to the vocal performer.
3. This work recognizes that biases about the beauty of individual
speech sounds are endemic and inevitable. All speakers possess them,
no matter how broad their linguistic experience or eclectic their
approach to art – or, for that matter, how virtuous their aesthetic politics.
In themselves, biases are usually innocuous as long as the possessor
does not take them too seriously; however, when combined with other
social influences, they can inflate themselves into larger and always
questionable judgments about the relative worth of accents or dialects.
Then the actor is in dangerous territory.
4. This work tries to place a firewall between these inevitable biases –
including those of the author – and the pedagogy itself. It does so not
 because of “political correctness” but because such biases – however
subtle their manifestation in the teaching of speech – are always limiting
to the actor if they start to set barriers against the explorations that the
actor can make.
5. The only “standard” that we can set for speech training is
intelligibility. By “standard” I mean a constant criterion that should be
adhered to always. Actors must always be understood easily by their
listeners. Everything else is optional: beautiful sound, interesting
accents, speaking trippingly on the tongue – all are optional; highly
desirable in many instances, of course, but not a constant criterion. Or,
to put it in its opposite context, speech that is merely intelligible and
nothing more is likely to be uninteresting, but it fulfills the most
fundamental requirement of human communication and the only
requirement that remains constant in all circumstances. Artful speech
that is unintelligible serves no purpose at all.
6. This work is based on the development of useful skills that the actor
can own. No training program can teach an actor all the accents she or
he might need throughout a career. No set of classes can acquaint an
actor with all the vocal demands that a career will elicit. But focused
training can provide tools that the actor can use to shape the unique
voice of every character that the actor will ever play.
7. This work embraces complexity in its content. Complexity nourishes
art. Reduction of the complexity of speech choices reduces the art.
8. This work embraces contradiction as an essential tool of teaching. It
is always interesting and useful to learn how to put more activity into
speech actions. It is also always interesting and useful to learn how to do
less. Getting stuck anywhere is never interesting.
 How to use this text

Do everything.

Don’t just read it – do it!

Figure 1. Mijnheer Rembrandt van Rijn explores lip corner protrusion.

Introduction

We all know how to talk. If you – as the reader of these words – are not the
victim of some disorder of the nervous system or injury to the muscles that
we use to speak, you can be confident that you can talk in a way that
communicates clearly to those listeners with whom you converse most
frequently
We use a number of different techniques to communicate our messages to
others: our tone of voice, our facial expressions, our vocal inflections, our
timing, our loudness or softness, even the vocal “quality” we use, such as
nasality or an abrasive sound. But the actions that communicate the most
basic information are the words themselves, those combinations of vocal
sounds that any speaker of a particular language understands to represent
very specific things. A breathed “h” followed by the “ow” sound (something
that alone might express a sudden pain) followed by the hiss of an “s” sound,
when combined in English, represent a place to live, a residence, a “house.”
Different speakers might say the word “house” with a number of different
accents of English and still be sure that the people who speak the same way
and who listen to them all the time would understand them, especially if they
are speaking in an informal conversation.
But suppose the conditions change. Suppose that you have to
communicate to people who speak our common language but do not share
your particular accent, or the slang you use, or the pace at which you talk. Or
suppose that your listeners not only do not themselves speak the way you do
but also haven’t heard many other people who speak like you.
Another set of possibilities: suppose that you need to command the
attention and understanding of a group of people while other people are
talking, or perhaps you need to speak unfamiliar words to a large group with
a clarity that also allows your listeners to take in also your thoughts and
attitudes about those words and about your listeners: in the first case, you
might be a teacher or an administrator; in the second case, you might be a
classical actor.
One choice that we can make when people don’t understand us easily is
to shut up. Many of us do just that; some people stop speaking when they are
around people who talk differently from them and therefore don’t understand
them. Another tactic might be to fight back: people might aggressively
accentuate their own speech patterns to make sure that people who are
outside their familiar accent group don’t understand them. But if we want to
engage our listeners, whatever their accents might be, if we want to be able to
communicate fully and effectively with other speakers of our shared
language, we need to open ourselves up to speech actions that can be
commonly understood by all speakers of that language, no matter what that
accent might be.
The only real test for the effectiveness of our individual speech patterns is
represented by the somewhat clunky term intelligibility. It doesn’t mean that
we have to produce speech sounds that impress listeners as “beautiful”, or
“cultured”, or “cultivated.” It means that we have to be able to produce verbal
sound actions that everybody who speaks our language can understand easily.
If they can understand us easily on this core level, they will also be free to
pay attention to all those other messages of tone or gesture that we are
sending woven through and around the words themselves.
As speakers in any context, we need to speak intelligibly no matter what
the accent comprehension skills of our listeners might be. If we are actors on
the stage or even in film or television, we need to establish that same
intelligibility for all listeners within whatever accent or vocal characterization
we might be using onstage. Obviously, with accent variability from both
actor and audience, it follows that the intelligibility bar gets raised
considerably for actors and that the demands on the speech skills of the actor
are much greater than for most other speakers.
In the past, the increased demands for intelligibility – often termed
“clarity” – onstage were met by imposing a fixed accent pattern of “good
speech” for actors. Today we are beginning to recognize that such fixed
accent patterns – based as they are on the theatrical conventions and aesthetic
standards of a bygone era – are no longer appropriate for the speech training
of the professional actor.
One reason why a fixed “good speech” pattern doesn’t work very well is
that the appropriate level of intelligibility from speaker to listener does not
always rise in one direction toward ever more detailed speech actions. The
fact that we need to be able to provide our listeners with a lot of detailed,
sharply differentiated speech actions does not mean that it is always
appropriate for us to speak this way all the time, either onstage or off. The
speaker who is locked into a highly formalized pattern may find that, just
because of the excessive effort being made, she or he is sending messages to
the listener that were not intended.
Rather than relying on a fixed pattern of “good speech”, the approach to
speech and accent training that you will find in these pages is founded on the
observation that actors are best served by developing a set of precise physical
skills for shaping sound into articulate speech. The ways that the actor will
use these skills are as unique to the particular actor and the particular role, as
are the vast number of individual impulses and conscious choices that an
actor makes in the course of building a performance. Because the skills are
still separable from one another, they allow the actor to vary them to meet
any requirement in stage, film, or voiceover work.
Part 1
Making sounds
1 Silence

War is diplomacy carried on by other means.

Von Clauswitz
Silence is argument carried on by other means.
Che Guevara

To understand sound we must know silence. To hear sound fully we must

hear silence.
An easy aphorism to write, but silence is a difficult thing for most of us to
experience these days. What passes for silence in our lives is an ongoing stew
of hums, whirrs, intermittent buzzes, and occasional roars from the machines
that populate our world; squeals of car tires, the blare of car horns, TVs and
radios left on in an endless ambient drone as a weird substitute for the
enfolding presence of human conversation, the endless fuzzy sigh of air-
conditioning units spreading dehumidified air (and mold, and spores, and
germs) throughout offices and homes. For many people born into
industrialized societies since the early nineteenth century, silence is – well –
almost unheard of.
In the last fifty or so years, the escalation of background sound has been
matched by the decibel overload of foreground sound. Dire reports are
published all the time about the widespread hearing impairment (particularly
loss of high frequency overtones) in young people because of the excessive
loudness (measured in decibels) of the sounds to which the ears are exposed;
personal stereos with headphones are frequently cited as the worst culprit, but
only one among many.
Contrast this trend with the writings of anthropologists who study
indigenous peoples in the jungles of Amazonia or Indonesia. They report that
among hunter-gatherer cultures it is still common for people to be able to
hear and interpret extremely faint sounds: the rustle of animals moving
through the fallen leaves on the forest floor a half-mile away, the flutter of
bird wings that tell both distance and direction of flight, the change in insect
sounds that warn of a change in weather several hours hence. While some
observers of these skills are apt to ascribe them to some sort of extra-sensory
power, it seems clear that nothing these people are doing requires any power
beyond an unimpaired hearing ability combined with the skill of attentiveness
and the experience to interpret very small sound differences.
These abilities are not beyond us, even if we have spent many hours
connected to our iPods. But we have some work to do. Over the years we
progressively hold more and more tension in our skeletal muscles – the
muscles we consciously use to move our internal structure of bone and
cartilage around. We do so as an unconscious way of bracing against the
shock of a loud – or just a new – sound. These held or “residual” muscle
tensions desensitize us, so that eventually we no longer even seem to hear the
annoying ambiences that accompany us throughout the day.
It’s a perfectly legitimate way of coping with contemporary life, and to
some degree we all do it. But we pay a price. We are armoring ourselves
against the rest of the world. Specifically, we are no longer fully available to
the world of sound around us; we are less likely to really pay attention to
other people speaking to us unless they are speaking loudly, or even better,
are speaking loudly on television or radio or podcast.
People who enjoy classical music but are used to hearing it on the
speakers or headphones of the home-entertainment centers in their living
rooms often express disappointment when they actually go to a concert; the
orchestra just doesn’t sound as brilliant and powerful as it does on the CD at
home. People who go to films a lot often complain about difficulties in
understanding the actors when they attend a live theatre performance even
when the actors are speaking at a reasonable volume with full clarity of
articulation. Unfortunately such complaints often lead to decisions by theatre
managements to put body microphones on the actors to amplify their voices,
or to engage in general “sound enhancement,” so that the human voice now
reaches the audience’s ears from loudspeakers in a general wash rather than
in specific, subtle, direct communication from the individual actors’ bodies. I
believe I am not too far wrong when I speculate that a growing number of
audience members at live theatre events are becoming willing to sacrifice
much of what makes live theatre a unique art form in order to recreate the
feeling that they are really sitting in a movie theatre being bathed in the
pumped-up surround sound: the Multiplex Experience.
So the ability to let go of residual tension in the skeletal muscles becomes
crucial in finding our way back into a fully sensitized, available body; a body
that can listen. I have to emphasize that the process of becoming sensitized,
or available to stimuli, or “finding neutral”, in no way means that as actors or
people we are becoming passive or non-analytical. On the contrary, our
sensory availability simply gives us more data to evaluate. Famed director
Peter Brook puts it eloquently in describing this state of vibrant, alert
awareness within which to find an equally vital form of silence:

In the search for the indefinable, the first condition is silence, silence as the
equal opposite of activity, silence that neither opposes action nor rejects it. In
the Sahara one day, I climbed over a dune to descend into a deep bowl of
sand. Sitting at the bottom I encountered for the first time absolute silence,
stillness that is indivisible. For there are two silences: a silence can be no
more than the absence of noise, it can be inert, or at the other end of the
scale, there is a nothingness that is infinitely alive, and every cell in the body
can be penetrated and vivified by this second silence’s activity. The body then
knows the difference between two relaxations – the soft floppiness of a body
weary of stress telling itself to relax, and the relaxation of an alert body when
tensions have been swept away by the intensity of being. The two silences,
enclosed within an even greater silence, are poles apart.
Threads of Time, pp. 106–7

The listener is fully open to the silence, and the energy of listening comes
from the attentiveness of every cell of the body. But there would seem to be
another crucial element. The vitality of the stillness that Brook perceives in
the empty desert proceeds at least in part from the space which is always fully
open to the potential for sound, an environment which expands the presence
of the listener by continuously encouraging a greater and greater
attentiveness. So there is a profound difference between the living silence that
Peter Brook found in the vastness of the Sahara Desert and the silence that
one might encounter in an anechoic chamber – a chamber in which all walls,
the floor, and the ceiling are completely damped acoustically, so that no
sound waves can reflect. Sounds produced in anechoic chambers are useful to
acoustic engineers who wish to study only those frequencies emanating from
a sound source without the complication of reflected sound from the
environment. But the sound that a human voice produces in an anechoic
chamber is a voice crying in a void, pure but strangely inhuman. The silence
that envelops one in such a place is the silence of the dead.
In a silence alive with the potential for sound, our awareness continues to
expand and our acuity of perception increases. Emerging from this attuned
awareness of silence, even the tiniest sound can have texture, imagistic
power, and can evoke the entire world within which the sound lives.
This is some of what Catherine Fitzmaurice, noted voice teacher and a
trained and acute listener, heard during a stay in Moscow and Petersburg:

Some sounds heard in Russia

Fricatives, deep-throat vowels; militia sirens; laboring buses; horses’ hooves
at night; endless traffic; endless talk in four-hour Tolstoy dramas; long
applauses becoming unison clapping; in his house, Scriabin recorded,
playing his own piano compositions; beggar children asking, asking;
loudspeakers in public places; pressed voices singing folk songs with a little
yelp at the end of the phrase; a young woman on her cell phone giggling
during a performance of Giselle at the Bolshoi Ballet; more Scriabin in the
Moscow Conservatory great hall; high heels clipping on uneven pavements;
a youth pissing in a bottle and laughing in the back row during a theatre
performance in English; gun shots at two in the morning; someone
(unanswered) banging on the door to my apartment at two in the morning;
the phone (unanswered) ringing at two in the morning; metal whining during
a high-speed train trip back from St. Petersburg; the urgent clatter of feet on
the marble floors of the Moscow metro; loud passionate arguments between
big men with love in their voices and ending with three noisy kisses, shoulder
slaps, and a grunting bear hug.

If we can listen out of an alive, sensitized personal silence, there is a lot in the
world worth hearing. If we then can feel our own sounds moving through our
bodies with the same alert availability, we can feel subtle changes in vocal
sound and action. This is where our journey from silence into vocal sound –
the assertion of our humanity – needs to begin.
Figure 2. The human skull, front view.

Figure 3. The human skull, lateral view.

Figure 4. The vocal tract: median section.
2 The space that shapes sound

On page 9 is a lateral – or side – view of a human head, seen as though the

head were cut in half. (Not exactly into two equal halves, though; if it were
cut exactly in half, it would be called a sagittal section.) This view allows us
to see the structure that we call the vocal tract, the space that shapes sound.
The vocal tract begins at the vocal folds (often misnamed the “vocal cords”
or worse misnamed the “vocal chords”) contained within the larynx. The
larynx is a structure of cartilage and connective tissue that houses the vocal
folds and hangs downward in the neck from the hyoid bone.

Figure 5. Larynx: anterior and posterior views.

The hyoid bone is the only free-floating bone in the human body: it isn’t
knit to, or jointed to, any other structure of bone or cartilage. But a number of
muscles and ligaments do attach to it, holding it relatively in place, even
though it can move easily up and down within a limited range of motion.
When we breathe in, we inhale the air through the nose or the mouth back
and downward through the larynx – past the vocal folds – and into the
trachea, the “windpipe” that brings the air into the bronchi and then into the
lungs.

The pharynx
When we breathe out, starting from the lungs a flow of breath can move up
the trachea through the larynx, from the larynx past the vocal folds and into
the space that we call the throat, but that more accurately is called the
pharynx. First, the air flow moves into the laryngopharynx, the space
immediately above the larynx, with the pharyngeal wall in back and sides,
and the root of the tongue (with a few interesting attachments) in front. The
air flow then moves into the oropharynx, the space at the back of the throat
that you can see if you shine a light into your open mouth.

Figure 6: Larynx, trachea and bronchi.

 Figure 7. The vocal tract.

At this point the air can continue straight upward into the nasopharynx,
the space that leads directly to the nasal cavity. We use this route when we
breathe through the nose.
But it can also turn and move into the oral cavity, the mouth, and then
out between the lips.
And obviously, the air can also move just as easily inward from the nose
or mouth toward the lungs.
As the lateral view on page 8 shows us, the nasal cavity is surprisingly
large. It is divided into two side-by-side spaces – two nostrils – by the
septum, a thin vertical sheet of bone and cartilage.
The oral cavity can be large also. But its size can be varied by how
dropped open the jaw might be. Then, too, the oral cavity is largely filled by
a thick group of interrelated muscles called the tongue.

By the way, what’s the difference between bone and cartilage?

Bone is rigid, self-repairing, composed largely of calcium phosphate and

contains an outer structure of compact bone with an inner core of
marrow.

Cartilage is resilient, flexible, variably elastic, only minimally self-

repairing (hello, arthritis!) and slippery. This last is useful in joint
articulations, because most bones are capped with cartilage at the joints.

The tongue
The non-medical terminology for the surface portions of the tongue includes
the tip, the blade, the dorsum (or body) of the tongue, and the root. The
dorsum is further divided into three parts, imaginatively labeled the front, the
middle, and the back.
 Figure 8. Left profile of oral cavity, showing the roof of the mouth and the tongue.

The tip and the blade of the tongue are the parts that we easily can stick
out of our mouths. They are the extremely flexible portions of the tongue, and
can engage in activity that is quite different from the activity in the dorsum of
the tongue at the same time, a very useful quality in speaking.
The tip, blade and dorsum of the tongue are the parts that you can see if
you open your mouth and look at your tongue in the mirror. The root of the
tongue is the surface portion way in the back that you cannot see because it
curves downward to form the front wall of your laryngopharynx. You can
move the root of the tongue, however, pulling it back or protruding it forward
to decrease or increase the size of the pharynx.

Figure 9. The oral cavity.

 The dorsum of the tongue is less nimble than the tip or blade, partly
because it is much thicker and also because it is directly attached to the base
of the oral cavity. But it is still capable of a variety of actions in order to
initiate swallowing and to participate in the process of shaping vocal sound
through articulation.
The only tricky part about naming the parts of the tongue is to remember
that the front of the tongue is behind the tip and the blade.
In defining the areas of the tongue surface, there are two other non-
medical terms that we can employ. The midline of the tongue, a slight
depression moving down the middle of the tongue from tip to back, is easily
seen if you look at your own tongue. The edges of the tongue are exactly
that: a continuous edge around the tongue dividing the upper (superior)
surface of the tongue from the lower (inferior) surface.
At this point, it is not necessary to consider the individual muscles that
enable the tongue to move in such complex ways. We will do so when we
isolate the tongue actions.

Figure 10. The tongue: superior surface.

Behind the dorsum of the tongue – in back of the back, as it were – the
surface of the tongue turns downward toward the hyoid bone and the rest of
the larynx. This is the root of the tongue, and it defines the anterior (or front)
limit of the laryngopharynx.

The epiglottis
At the very base of the tongue root, and attached to it, is a spoon-shaped
piece of cartilage covered with mucous membrane called the epiglottis. In its
resting position it is raised up out of the way of the flow of air through the
vocal tract. When a person initiates a swallowing action with the tongue, the
epiglottis swings down and covers the top of the larynx, so that when we
swallow food, that piece of food is safely moved back past the larynx and
into the esophagus, the tube located directly behind the larynx and trachea
that carries the food down to the stomach. After the swallowing action is
completed, the epiglottis swings back up and out of the way. The epiglottis
can contribute to a very few speech actions in some languages.

Figure 11. The epiglottis.

The roof of the mouth

In other words, the top of the oral cavity.
The gateway to the oral cavity is the lips. The muscles that control lip
action are not only crucial to the act of speaking, they are also what we call
muscles of expression. While the lips are in the act of speaking, they can
also send simultaneous visual messages simply by their shape, messages that
may modify or even contradict the words that are spoken – as can the rest of
the muscles of expression, which include all the muscles you can use to move
your face around.
Behind the lips are the teeth, and right behind the upper teeth, on the roof
of the mouth, is a slight thickening of the gum tissue that is termed the
alveolar ridge. A lot of articulation targets this narrow area, so it is important
to know where it is.
 Figure 12. The roof of the mouth.

Behind that alveolar ridge is what we commonly call the “hard palate”.
Technically, it is simply called the palate. It is a bony structure that arches
over the front half of the oral cavity. Obviously it does not move, but the
tongue moves toward it or touches it frequently in articulation.
Behind the palate is the area we usually call the “soft palate”. The correct
term, however is the velum. Because the velum is composed of muscle
covered by mucous membrane, it is capable of movement. The lower surface
of the velum, the part that we see on the roof of the mouth, can lower or raise
to shape the acoustics of vocal sound or to assist in the articulation of some
vowels and consonants. The upper surface of the velum can raise to close off
the passage of air to the nasal cavity, or it can lower to allow sound to move
into the nose, as all Americans do when they say “m”, “n” or “ng”.
The rear portion of the velum extends back past the limit of its attachment
to the sides of the skull and so becomes very flexible. Consequently it droops
down at the back of the oral cavity to form the uvula, a conical structure
hanging downward. The uvula is important in triggering the gag reflex, but
even as a passive participant is also crucial in the production of certain
speech sounds.

The glottis
The glottis is the lowest limit of the vocal tract, the opposite end of the vocal
tract from the lips or the nostrils. It is defined as being the space between the
vocal folds.
The vocal folds themselves are infoldings of mucous membrane that also
contain tiny muscles inside them and fibrous ligament on their inside edges.
They can thicken or thin themselves and they can be brought together or apart
by the action of other muscles that are attached to the cartilage structures of
the larynx. When they are apart, the air can move past them unimpeded.
When they are brought together, the action of the air flow interacts with the
unique elasticity of the vocal folds to produce a very rapid oscillation of the
folds. This, in turn, chops up the flow of air into pressure waves. These
pressure waves flow through the vocal tract producing more complex wave-
forms. Only when these complex pressure wave-forms reach the ears (or even
more specifically, the brains) of listeners, do we interpret them as what we
choose to call “sound”, the sound – in this case – of the human voice.
So “sound” is a perceptual term. While we will obtain a lot of crucial
information for ourselves about speech through listening to the changes in
sound that we can make, we will also learn a great deal by the tactile
sensations of the pressure waves flowing through the vocal tract. We learn by
listening, but we also learn by literally feeling the voice.
If this were a text about vocal production, we would explore the action of
the vocal folds in much more detail. But while the vocal folds – and the
glottis between them – do play an important part in the shaping of voice, they
are limited to the shaping of only a few distinct sounds in their contribution
to human speech action.

Figure 13. The vocal folds and their action.

 Now that we have defined – in a very general way – the territory through
which the flow of air can pass, we can start to explore the process of moving
from silence to sound.

Terms to know

Vocal tract
Vocal folds
Larynx
Hyoid bone
Trachea
Bronchi
Pharynx
Laryngopharynx
Oropharynx
Nasopharynx
Nasal cavity
Oral cavity
Nostril
Septum
Tongue
Tip
Blade
Dorsum: front, middle, back
Root
Midline
Edges
Epiglottis
Esophagus
Muscles of expression
Alveolar ridge
Palate
Velum
Uvula
Glottis
3 From silence to sound

Forget that you know how to talk.

Of course, I have just asked you to do the impossible, but give it a try
anyway. The fact that we do know how to talk (and that we know we know)
will come in handy along the way, but for now try to let it go.

Rediscovering the familiar

The process we will be going through in this work is to rediscover how we
shape simple vocal sound into complex language. We are going to become
precisely aware of some physical actions we take to shape sound that we have
been doing for most of our lives with great ease but with little or no sensory
awareness. So try to put yourself in the role of someone who is just starting to
explore the language process – a very young child, perhaps. But as you
explore like a child, feel free to bring your adult abilities to analyze and
interpret information along with you.
We will start very simply, by preparing the space in which the shaping of
sound will take place.

TOOLS
A hand mirror. Always hold the mirror just above mouth level. Make
sure that you have plenty of light to see inside your mouth.

Exploration 1
Sit or stand. Float your head up as though a string were attached to the top of
your head, gently pulling it upward toward the ceiling. Relax your shoulders
and arms.
The amateur yawn. Yawn the back wall of your throat (your pharynx) as
open as possible, from the back of your nasal cavity all the way down to your
vocal cords. Notice, as you do this action, that you probably are pulling the
root of your tongue back at the same time and opening your jaw very wide.
Audio track 1: The amateur yawn.
The professional yawn. Yawn again, but this time leave the jaw relaxed
open, but not forced open, during the entire yawn. After you have yawned the
pharyngeal muscles in the back and sides of your throat open, try to stretch
the tongue root forward, away from the back wall of the pharynx. The tongue
root is the part of your tongue that lies between the back of your tongue and
your larynx. The rest of your tongue will protrude also when you stretch the
tongue root forward, but the muscular work is all being done way in the back
of the throat. You will feel a strong stretch along the sides of the pharynx if
you are advancing the tongue root sufficiently. Relax out of the yawn, but
notice that the feeling of openness in the pharynx is still there.
Audio track 2: The professional yawn.
The inevitable yawn. Inducing a yawn deliberately, as you have just done,
almost always causes people to yawn for real. This is all right. But if it
happens to you, stretch your tongue root forward so that it becomes an
Inevitable Professional Yawn.

Exploration 2
Breathing through the nose. Leaving your jaw relaxed open, lightly close
your lips so that no air can get in or out of the mouth. Breathe normally
through your nose. Make no attempt to breathe more strongly. Simply note
the feel of the cool air moving into and through your nasal cavity. You can
also feel the cool air flowing down across the top surface of the velum (the
soft palate) and down through the upper part of your pharynx, the
nasopharynx. The cool air that you inhale (you sense it as cool assuming
you are doing this exercise at a room temperature lower than your own
normal body temperature of 98.6 degrees F!) is easier to feel than the warmed
moistened body-temperature air that you breathe out through your nose.
Breathe more energetically through your nose a few times.
Audio track 3: Nasal breathing.
Notice that if you inhale very strongly it is hard to get the air in and out
efficiently. Partly this happens because your nostrils may actually be pulled
partially closed on the inhale because of the lower air pressure within the
nostril. Another reason is that the turbinates (or conchae) in the nasal cavity
(curved bony projections covered with mucous membrane that filter, moisten,
and warm the air) are also obstructing any fast airflow. Thirdly, the tissues of
the nasal cavity actually expand and contract to make it easier to breathe first
through one nostril and then through another.

Figure 14. The nasal cavity and the sinuses.

When we are breathing easily and normally, the restricted airflow through the
nose presents no problem. But if we are going to use this airflow to
communicate to other human beings, the time and effort involved in inhaling
through the nose while speaking would be prohibitive. So most of the time,
when we are speaking several sentences in a row we inhale through the
mouth even though we inhale through the nose most of the time when we
listen to our conversational partner speak.
Breathing through the mouth and nose. Breathe through the nose as
described above with the jaw dropped open in a relaxed position but with the
lips lightly sealed together. Continue breathing through your nose as you let
the lips drop apart so that air can now enter and leave your body through your
mouth also. Even though the mouth immediately takes over most of the
airflow, you can still feel the breath flowing in and out at the nostrils. When
you breathe through both the mouth and nose at the same time, you are
completely relaxing the velum. Enjoy this, because the velum is about to go
to work.
Breathing through the mouth. In order to go from breathing through the
mouth and nose simultaneously to breathing through the mouth alone, you
will need to raise the upper surface of your velum (soft palate) to close off the
airflow to the nose. Many of us are not exactly sure what this action is, even
though most of us do it unconsciously when we speak.
To feel the specific action of the velum, continue the breath flow through
both the mouth and nose; then close the lips so that the breath flow moves
through the nose alone, then on an exhale (breathing out) close off the breath
flow to the nose also. You have just raised the velum to close off the
velopharyngeal port, the place that divides the nasal cavity from the
nasopharynx.
There is a complication, though: there are two completely different places
where you can close off the airflow under these conditions; one is the velum
raising to close the velopharyngeal port (this is the action we want to take);
the other is the space between the vocal folds, called the glottis, closing as it
does when we start to cough. By contrast, the velopharyngeal closure and
opening by the velum feels rather like snorting. If you are closing at the
velum and continue the airflow out with your lips still together, your cheeks
will puff out, but if you are closing off the airflow at the glottis no air will
come into the mouth.
When you can feel the specific action of the velum as it raises to close off the
airflow through your nose, allow the lips to drop apart and exhale through the
mouth. You now are breathing entirely through your mouth.

Tidal breathing. Sit or stand with the spine in alignment and the head
floating upward at the end of your neck toward the ceiling or sky as though it
were a helium-filled balloon. Exhale fully and then inhale fully. Do not close
off your breathing at your throat, nose or anywhere else. Then simply allow
the breath to release out as much as it “wants to”, while keeping the glottis
and the rest of the vocal tract completely open. Just stop breathing, but make
sure that you are not “holding your breath” by bracing any of your breathing
muscles.

By the way, what does the breath really “want”?

It wants to exist at the same air pressure as the air outside the body and
 will move on its own (if there is nothing to stop it) until air pressure
equilibrium is achieved. This current tidal breathing exploration allows
you to do this. Its only minor disadvantage for the maintenance of life is
that when this equilibrium is reached, the air stops moving and you stop
breathing in and out. So your exploration of air pressure equilibrium will
not last very long.

After about five to ten seconds of breathing inactivity, your body will start
breathing again. It will be a very gentle in-and-out breathing action with
pauses in between the breaths. Allow the breathing action to happen – don’t
make it happen. If you want to pull a lot more air into your lungs on the
inhale, you could do so. If you want to push a lot more air out on the exhale,
you could do that. But at the moment you are doing neither: you simply are
allowing the breathing cycle to happen. This easy, gentle, non-controlled
breathing is called tidal breathing. While life exists in your body, this
normal breathing is as inevitable and cyclical as the ocean tides.
Another thing that this exploration shows us is that the breathing mechanism,
when completely at rest, finds naturally a place of repose approximately
halfway between full inhale and full exhale. This may be a revelation to those
who may have believed that breathing relaxation is achieved with all the
breath exhaled. Not so.

Exploration 3
Your first vocal tract noise. With your jaw dropped open in a relaxed
position, breathe comfortably and easily through your mouth. If your mouth
starts to get dry, swallow to lubricate the passage and continue your easy,
normal tidal breathing through your mouth.
You will probably notice that there is little or no sound from the air passing
through this open “pipe” from your lungs to your lips, either on the inhale or
on the exhale.
Now, without moving anything in your mouth or throat (this is very
important), start to pull more air into your lungs on the inhale and push more
air out on the exhale. Use your torso muscles to get the air in and out, but DO
NOT TIGHTEN YOUR THROAT as you perform this action. And keep your
jaw relaxed open.
Audio track 4: Increasing breath flow.
As you increase the speed and volume of air flow, you will start to notice that
you can feel the air moving through your throat and mouth. You should be
able to register the feeling pretty evenly with all the tissues of the mouth and
throat. You will also start to hear a rushing sound of the air (or, more
accurately, a rushing noise of air) as you breathe. You will hear it as equally
loud on both the inhale and the exhale.
Explore this “breathy” noise for five breath cycles. Then relax and return to
easy tidal breathing. If, at any time when increasing the air flow, you feel at
all dizzy, stop the exercise until the dizziness passes.
This experiment tells us a few very useful – if very basic – things. It tells us
that producing any sound or noise though the mouth or nose always requires
some sort of airflow. It tells us that merely breathing normally through an
open, unobstructed mouth and throat produces little or no noise at all. It tells
us that in order to produce any sound or noise through an open passage using
only a flow of unvoiced air, we will need to breathe a lot harder to produce
even this most simple and crude kind of noise. It tells us that this breathy
noise is equally possible while breathing in (inhaling, an ingressive flow) as
when breathing out (exhaling, an egressive flow). Finally, it tells us that we
are still a long way from having a full set of spoken language skills, but it
points the way that we must go.

Voicing
In order to produce more variety of sound (and to avoid passing out
frequently from over-breathing), we will need to do one or both of two
things. If the flow stays unvoiced (just air moving through a pipe) we will
need to be able to obstruct that flow to a degree that will produce a distinct
noise even when we are not breathing more heavily; also we will need to
obstruct the flow in different places to produce different noises.
Or, we will need to voice the flow. This means bringing the vocal cords
together with enough tension to allow the flow passing between the folds to
be chopped up into very rapid pressure waves. We can still obstruct the
voiced flow if we want to, but we do not need to do so to produce different
sounds. Instead we can vary the sounds simply by shaping acoustically the
unobstructed (but voiced) flow.
We will appreciate the feeling of voiced sound all the more if we explore
the sounds that we can produce on the way to voice.

Exploration 4
Heavy breathing, creaky voice, phonation. The vocal folds themselves are
articulators we can use to obstruct the flow as long, because they haven’t
gone into their rapid vibratory action that produces pressure waves of voice.
If you bring the vocal folds together a bit, you start to hear and feel a
localized noise as the breath flow becomes obstructed. The noise is that of the
classic Darth Vader imitation, or on an even more banal note, the classic
telephone heavy breather. Breathe gently, however, unless you contemplate
phone-breathing as a career. Notice that you can do your heavy breathing
both ingressively and egressively.
Audio track 5: Partial glottal closure.
Bring the folds further together – but loosely, so that the sound changes from
heavy breathing to a slow, low-pitched creak, with the vocal folds flopping
loosely against one another in a very slow vibratory pattern. With a little
practice you can produce the creaky voice both egressively and ingressively
with ease. Be aware that this creaky voice should be distinguished from the
infamous (to speech teachers) glottal fry. The glottal fry is a somewhat
similar sound but is usually softer and is produced with a lot of vocal-cord
tension, giving it a nice sizzling quality, usually at the end of spoken phrases.
It tells voice and speech coaches that you lack sufficient breath support in the
abdominal muscles.
Audio track 6: Creaky voice and glottal fry.
Try the Creaky Voice ingressively as well as egressively. It is worth
practicing, partly because the resulting sound is so wonderfully odd and
partly because, if you learn to do it well, it means that you have let go of
excess residual tension in the muscles of the vocal folds.
Audio track 7: Ingressive creaky voice.
If you bring the vocal folds together a little more tightly and then allow the
cords to become slightly tauter, the slow creak suddenly changes into a very
rapid vibration that produces a sound of a specific pitch with resonance
overtones above and below the fundamental pitch. This is the action of full
voice, or phonation.
Audio track 8: Phonation.

Resonance
If we imagine the area located between the larynx and the lips as a very
flexible pipe that can change its shape or diameter to produce different
language sounds, then we are focusing our attention on the vocal tract. The
vocal tract can be shaped by the muscles that attach to the walls of this tube
to produce different voiced sounds without obstructing the flow. A voiced
flow of vibrations is a coherent sound, as distinct from the unvoiced airflow
that would be more properly defined as noise.
Coherent sound vibrations passing through the vocal tract can be shaped
acoustically through resonance: as the size of different areas of the vocal
tract change, certain parts of the complex voiced pressure wave are amplified
or cancelled. The exact process, while very interesting in itself, is not
necessary for us to explore in any detail here; but we do need to know that
resonance within the vocal tract accounts for three things in our voices:
amplitude (or loudness), vocal quality, and the formation of the acoustic
patterns that we recognize as vowels.
Let us explore what shapes in the vocal tract have various effects on these
three aspects of resonance.

Exploration 5
The sound of stupidity. Begin as in Exploration 1. Make sure that your jaw
is relaxed open and that the tongue is very relaxed also, with the tip of the
tongue just touching the back of your lower teeth. Once you have established
a normal, unobstructed, silent breathing rhythm, bring the vocal cords (also
called vocal folds) together to produce a voiced sound on the out-breath. It
will sound something like “uhhh”. A nice dumb “uhhh”. If it sounds like
“ahhh” it means that you are retracting your tongue root. Relax the tongue
completely: “uhhhhhhh.”
Audio track 9: Relaxed vocal tract sound.
This is the only completely relaxed sound in language. So it follows that any
sound that you make that differs from “uhhhhhhh” requires some muscular
work.

Exploration 6
Vocal gurning. To “gurn” is to make strange faces, and the more bizarre they
are the better. Occasionally in tabloid weekly papers (particularly in Britain)
one can see photos of the winner of some gurning contest. Usually it is some
ancient codger who has no teeth and can pull his lower lip up over his nose or
produce some equally weird facial distortion.
Try gurning. Just by itself it is a wonderful exercise program for your face.
Smoothly and easily stretch or contract every muscle of expression in your
face that you can persuade to participate.

Figure 15. The gurning tradition: Les Grimaces, Louis-Léopold Boilly (1823).

Extend your gurning into the vocal tract so that you are engaged in the same
tensing/stretching action inside the oral cavity (including the tongue), the
pharynx and – as much as you can – up into the nasal cavity. Now start to
make voiced sound. Without favoring one sound over another, allow yourself
to make faces and surprise yourself with the sounds that come out. Or in, for
that matter, if you try gurning with an ingressive sigh.
Audio track 10: Vocal gurning.
Anything goes. Any sound is permissible. But keep the sounds voiced:
vibrating air, not just air. If you need a tactile reminder of which is which, put
one hand lightly on your larynx and feel whether there are vibrations
happening all the time.
Slow-motion gurning. Now slow the movements of your facial muscles.
Keep the vocal energy going and the gurning as extreme as before, but do it
all in slow motion. The sounds will now be changing more gradually, so that
you can analyze them more readily. Analysis and gurning might seem to be
incompatible, but actually you can do both easily.
What kinds of sound changes are happening as you shape the vocal tract in
this rather extreme way? Probably everything that can change is changing.
You are making different – and very likely strange – vowel sounds. But
sometimes your actions briefly obstruct the flow; when this happens you are
actually doing consonants. (But more about this later.)
Audio track 11: Slow-motion vocal gurning.
Also, you may notice that the kind of tone you are producing is changing,
from twangy to hollow, from thin to ample, even from softer to louder, just
by the shaping of the tube from your vocal cords to your lips. These are
changes in the amplitude or loudness of the voice and (even more important
for us right now) changes in the tone quality. We are not using the word
“quality” here to mean good or bad; we are using it to indicate a unique tonal
resonance.
All the articulator-shaping we can do contributes to forming our tone quality,
but, at the risk of generalizing, most of our quality and amplitude change is
caused by changes in the pharynx, the part of our vocal tract that goes up the
back of the throat from the glottis past the mouth to the nasal cavity.
Focusing on vowels. Try gurning again on a voiced egressive sigh. Change
the gurns slowly. This time, let the back and side walls of the pharynx relax
so that they are no longer part of the process. In addition, try not to obstruct
or impede the flow of voice on your sigh by bringing the articulators too
close together. Within these limits, let the gurning be as free-form as
possible.
Audio track 12: Acoustic vowel shaping.
Every so often in mid-gurn, at least once per sigh, freeze the articulators. Feel
and hear the single sound result that is coming out. Note that sometimes the
vowel sounds you are producing will be familiar ones and sometimes they
will be very unfamiliar indeed.
Audio track 13: Move/freeze/move.
Practice this exercise for at least five minutes. Start getting used to these new
sounds and register the way they feel in your mouth, throat, or nose. As you
experience these new sounds, note that the physical “posture” of the vocal
tract is determining the sound; this is the opposite of what usually happens,
which is to imagine a sound and then shape the articulators to produce it.
Continue to explore the possibilities of shaping the flow without getting in
the way of it. As you do so, keep letting go of any muscle actions that do not
directly affect what we would consider “vowel-sound change”. Keep playing
between freezing the articulators to focus on a new sound possibility and
feeling the gradations of sound change as we move through the free-form
shaping actions on a voiced sigh.
Gradually, as we play attentively through this more focused form of gurning,
we find that we can live comfortably and confidently with all these new
sounds and sound actions, and become less and less dependent on the
relatively meager set of “vowels” with which we started.

Exploration 7
Toning. Repeat the same focused shaping of a voiced sigh that you found in
Exploration 6 above, but explore the feel of the sounds while lying
comfortably on your back, directly on a wood floor if possible. Feel the flow
flowing through all of your body as you explore the multitude of sounds you
can sigh by changing the shape of the tube from the larynx to the lips in every
way possible without obstructing the flow. Freeze the articulators and explore
the resultant sound flow for at least thirty seconds. Then move on to
something else. Continue to let action lead sound rather than vice versa. Let
yourself be surprised by the sounds that result.

Note as you feel the different sighed sounds focusing their vibration energy
in different parts of your head and torso.

Vibration energy?
The term “vibration” has a variety of interpretations in voice and speech
work, and this can cause some confusion. The most literal definition of
vibration is a purely mechanical one: it is the oscillation of an object back
and forth through a mid-point of equilibrium. A clock pendulum is a
decent – and slow – example; a tuning fork, such as a musician uses to
determine pitch, is a good and fast example. At the other extreme we find
the purely metaphorical: “He’s got bad vibes.” Or, quadrupally positive,
the Beach Boys’ “Good, good, good, good vibrations.” In between we
often find gradations and combinations of these two possibilities. Musical
sound begins with mechanical oscillation, whether it is the oscillation of a
guitar string plucked with a finger, or an oboe’s two reeds oscillating
against one another, or the oscillating action of the human vocal folds in
the larynx.
In all these examples, the mechanical oscillation is converted into
pressure waves in the air. These, too, oscillate from condensation (air
molecules pushed together) to rarefaction (air molecules pulled apart). It
is reasonable to call these “waves” rather than “vibrations”, and some
voice scientists refuse to term them vibrations, ever. But when these
pressure waves encounter the ear drums of a living creature, the process
turns back into a mechanical and physiological oscillating action. In the
same way that we perceive sound through the mechanical response of
oscillation at the eardrum, so we feel sound waves – within the vocal tract
or through sympathetic resonance in the body – as a mechanical
oscillation response of mucous membrane, bone and cartilage. Or, to put
it simply, yes, we do feel “vibrations” and can perceive “vibration
energy”.
Exploration 8
Egressive or ingressive. Standing or sitting, with the head floating upward
and your spine lengthened, explore the shaping of voiced flow as before. This
time, however, instead of the usual unvoiced inhale in preparation for the
next voiced sigh, explore staying on voice as the breath flow moves
egressively (outward) and then also ingressively (inward).
Ingressive speech has a long and noble history. In the Renaissance, there
apparently was a tradition of young lovers using ingressive speech to disguise
their voices during romantic trysts. Somehow, imagining Shakespeare’s
Romeo saying “But soft! What light through yonder window breaks?” or
Juliet’s “O Romeo, Romeo. Wherefore art thou Romeo?” ingressively takes a
little of the bloom off the romance, but perhaps a really authentic production
would do it that way. It probably will happen. According to some scholars,
the tradition of disguising the voice by ingressive flow continues in various
cultures to this day.
Audio track 14: Ingressve speech.
What is not a matter of debate is that ingressive sounds are a part of today’s
languages from Swedish to Japanese to Argentinean Spanish, and others as
well. In singing, ingressive vocal production (induction?) is often heard as an
“extended vocal technique”. In voiceover work, ingressive techniques are
used in producing unusual character voices or in imitating animal sounds –
dog barks, horse whinnies and the like. And as with the Creaky Voice done
ingressively, full ingressive phonation is an excellent way to release residual
tension in the vocal folds and also to explore efficient rib action,
diaphragmatic action and abdominal release on the inhale.
Another thing to explore on ingressive voicing is the feel of the tube
resonance in the sub-glottal pipe (sub-glottal simply meaning “below the
vocal folds”) of the trachea, or windpipe, leading from the larynx down to the
bronchi (the pipes to each of the two lung)s. In my – admittedly minority –
view, the enhancement of amplitude and low overtone resonance experienced
during ingressive speech argues for the possibility of genuine resonance in
the thoracic cavity. For most voice scientists, genuine resonance is limited to
the vocal tract itself.
 Terms to know

Turbinates (conchae)
Velopharyngeal port
Cavity resonance
Ingressive flow
Egressive flow
Unvoiced/voiced flow
Heavy breathing
Creaky voice
Glottal fry
Phonation
Gurning
Resonance
Amplitude
Vocal quality
Vowel formation
Consonants
Toning
Oscillation
Condensation
Rarefaction
Sub-glottal
4 The muscles that shape sound: the tool kit

Consonant articulators
Articulation is the process of shaping the flow of air or vocal vibration so that
it forms differentiated sounds that can be clustered in sequence to convey
meaning. This is a somewhat dry way of putting it, because we are
considering only the physical actions themselves without investing these
actions with the ideas, the impulses, and the passions that can make
intelligible speech a fully human, rich, and beautiful experience.
But passionate words or complex ideas that are shaped by articulators
frozen into narrow habits of use are always diminished. By allowing
ourselves to rediscover the physical basis of individual speech actions, we
can move past the arbitrary limitations that we have often placed upon our
own ways of speaking our language. We gain the skills of shaping sound.
We can make a distinction between articulators, which are the parts of
the vocal tract that can move through muscle action to vary the sound, and
points of articulation, which are fixed, unmoving places in the vocal tract
toward which the articulators move to obstruct the flow.
Articulators
Jaw
Lips
Tongue (tip, blade, front, middle, back, root)
Velum (or soft palate)
Epiglottis
Vocal folds (or glottis)

Points of articulation
Teeth
Lips
Alveolar ridge (gum ridge behind upper teeth)
Palate (or hard palate)
Velum (or soft palate; it is also an articulator)
 Uvula (the “little grape” – in Latin – hanging
from the back of the velum)
Pharynx

Vowel articulators
The big difference between the articulation of consonants and the articulation
of vowels is that consonant articulators always act to obstruct the unvoiced or
voiced flow, but vowel articulators shape the unobstructed flow (usually
voiced), thereby changing the acoustic properties of the sound. These same
vowel articulators also shape the unobstructed flow to produce changes in
voice quality and amplitude.

Articulators
Jaw
Lips
Cheeks
Tongue
(primarily the front, middle, back, and root)
Velum
Pharynx

(Note that the pharynx is a point of articulation for consonants but that each
of the three sections of the pharynx – the nasopharynx, the oropharynx, and
the laryngopharynx – can be active articulators is shaping vowels, and even
more important in shaping vocal quality and amplitude.)
 Figure 16. The facial muscles that shape speech.

Warming up the muscles

Put most simply, muscles do one thing: they tense. This is an electrochemical
action (it produces an actual electrical charge, measurable in microvolts) that
is triggered by voluntary or involuntary brain function. When the muscles
tense, they bulge – this is what the body-builder does when he or she poses. If
the muscle is tensed more, it shortens along the direction of the muscle fibers;
this action is what the body-builder (or anyone, for that matter) does when
she or he changes from one pose to another, moving the skeleton around.
A better term for the bulging action is bracing, and a better term for
shortening is contracting. Both the bracing action and the contracting action
are important in the physical action of vocal production, and both are also
important in shaping the vocal tract for speech actions.
When you contract a muscle, you draw the structures to which the muscle
is attached closer to one another. These structures could be bones or cartilage,
or they could be other muscles.
There is a second thing that muscles can do that derives from the first:
they can let go of their tension and relax. When muscles relax, they thin and
lengthen. A muscle with the greatest range of action is one that can move
from relaxation to full tension and then can release back to full relaxation. In
today’s society, most of us go through life with unhealthy amounts of
residual tension in our muscles, which means that our range of muscular
action is limited.
Many of the human facial muscles and the muscles that change the shape
of the vocal tract attach at one end to the bones of the skull and at the other
end to other muscles. This is why our faces – and our voices – are capable of
so many variations. These muscular attachments are accomplished by
tendons, bands of strong, flexible fibrous tissue that attach to the muscle,
then to bone or cartilage, or to other tendons. No muscles attach directly to
each other without the helpful presence of a tendon at each end of the muscle.

By the way, what’s the difference between

a tendon and a ligament?
A tendon attaches muscles to bone, cartilage or to other muscles.
A ligament attaches bones to other bones. It provides a flexible
attachment that permits movement, as in joints. Bones can also attach
directly to one another in rigid, complex structures. When they do, they
are said to be “knit together”.
Both tendons and ligaments are made of strong, fibrous tissue that is
largely composed of collagen.

So a warm-up for the facial muscles and the articulator muscles should
involve tensing (thickening and shortening) and relaxing (thinning and
lengthening). Here are some possibilities.

1. Use both hands. Spread your fingers apart. With the tips of the fingers
massage the face with a circular action and moderate pressure, just
enough to know that you are moving the muscles around underneath the
skin and on top of the skull. Start behind the hairline in your scalp and
move down to massage your forehead. Massaging the tension out of the
frontalis muscle that covers the forehead is an excellent – and speedy –
way to let go of tension throughout the body. Continue to massage
around your temples and around your ears, then massage the sides of the
nose and cheekbone. (Use this massage to help define “the skull beneath
the skin”.) Continue the circular massage to the jaw hinge and jaw
 muscles in front of your earlobes and then complete the massage at the
cheeks and around the mouth.

2. Let all your facial muscles relax. Let your jaw drop open. Close your
eyes. With your fingers spread apart, draw the tips of the fingers directly
downward on your face, with almost no pressure, from your hairline to
your jawbone. Repeat at least five times. When you have finished, try to
sense the weight of your face as it is attached to your skull.

3. Close your eyes. Place the palms of your hands over your eye sockets.
Do a very easy circular massage with light pressure. You are massaging
the orbicularis oculis muscles that ring the eyes sockets.

4. Gurn. Gurn slowly. Then gurn swiftly. As you do so, note that you are
contracting some facial muscles and – by so doing – stretching others.

Articulator isolations
These exercises, some of which may seem rather odd, are a very important
part of our work and should become a part of your warm-up routine. You
should practice them with the same sense of form and attentiveness to the
experience that you would bring to the basic discipline of learning a musical
instrument or training muscles for a sport. These actions form the skills basis
for everything that follows. The object here is to explore the mobility of the
articulators in isolation from one another. It is very important that you pay
attention to using the muscles we are focusing on actively. It is equally
important that you pay attention to not using muscles if they are irrelevant to
the focused action.
You should first do each of these exercises without sound, after which it
is useful to do them again while releasing a voiced sound, hearing – and
feeling – what sound changes are produced and (equally interesting) what
sounds stay the same.

Bilateral symmetry
With only a comparatively few exceptions, all the members of the animal
kingdom, from tigers to tarantulas, form their physical structure of bone
and muscle according to bilateral symmetry. Imagine a line down the
 middle of your body: the structures on one side of that line mirror the
structures on the other. So you have a left arm and a right arm, a left
nostril and a right nostril, and so on. There may be minor variations in the
mirrored structures, but the principle applies. So when we talk about a
muscle on one side of your face, or a bone on one side of your head,
there’s also a similar bone or muscle on the other side. Note, though, that
bilateral symmetry applies only variably to internal organs: you have a
left kidney and a right kidney, but you don’t have left and right livers.
And a number of sea creatures exhibit radial symmetry.

The jaw
We all know what the jaw does: it opens downward away from the skull and
also raises upward to meet the skull when our teeth come together to chew.
When we think about the jaw, we are really considering the “jaw bone”, more
accurately termed the mandible. Anatomically, the mandible is not a part of
the skull, even though it functions in relationship to the skull in all its actions.
The mandible contains all the lower teeth and its principal function in
movement is to bite and chew food.

Figure 17. The jaw bone or mandible.

Place the fingers of both hands on the outside of your mandible at the front of
the chin, with your thumbs underneath the chin and the rest of your fingers on
top. Gently feel the bone as it curves back on the right and left sides of the
chin along the body of the mandible. Eventually you will feel an angle
where the edge of the mandible turns upward. This area of the mandible is
called the ramus (Latin for “branch”) of the mandible. The ramus ends in
two very differently shaped processes or bony projections. The rear one is
the condylar process, which is a rounded knob covered with cartilage. It acts
as the “joint” for the jaw, fitting into a shallow depression in the temporal
bone of the skull. However, the condylar process needs to be tied to the skull
at this shallow temporomandibular joint with ligaments, so that the joint is
secure, but also capable of a great variety of movement.
There is another unique feature of the temporomandibular joint: in between
the condylar process and the temporal bone is a small disk composed of
fibrous cartilage called the articular disk. It allows the joint to take two
actions simultaneously when the jaw opens: the first is the expected rotation
of the condylar process, such as occurs in most joints, like your hips; the
second action is the sliding of the condylar process forward as the jaw opens.
There is a great advantage in this double action: it allows the mandible to
drop open, a little further away from the skull than would be possible if there
were only a rotational action. It also helps to keep the pharyngeal area of the
vocal tract open and unconstricted.
 Figure 18. The articular disk.

Explore this sliding action. Place your index fingers at the very top of the
mandible, right in front of your ears. Easily drop the jaw open (don’t force
it!) and feel that the condylar process of the mandible moves very slightly
forward. When you bring the mandible up to close it, feel the bone move
back.
Dropping the jaw open is a relaxation action. Gravity does the work for
you. You can open the jaw further, but to do so requires muscular action and
all the muscles that pull the jaw open also retract the jaw at the same time,
which constricts the laryngopharynx and the larynx itself – not a useful result
for open, free vocal production. Forced opening of the jaw also can cause
strain at the temporomandibular joint.
Since gravity drops the jaw open very efficiently, it follows that the main
muscles that attach to the mandible are used to raise the jaw upward against
gravity. These muscles are very powerful, since they are used primarily to
bite and chew food.
Place the fingertips of both hands on your cheekbones. You are feeling
bone structures of your skull that actually arch away from the rest of the skull
so that some muscles and tendons can function underneath. The anatomical
name is the zygomatic arch even though part of the always-versatile temporal
bone also forms the arch along with the zygomatic bone. Now move your
fingertips slightly downward and close your jaw. You can feel muscles brace
into your fingers; these are your masseter muscles, the primary external
muscles of human jaw closure.

Figure 19. The masseter muscle.

 They are short and thick, with broad attachments around the angle of the
mandible at one end and into the zygomatic arch at the other. These are
powerful muscles that can exert several thousand pounds of bite pressure
measured in pounds per square inch (psi). Because they are powerful, they
can store a lot of residual tension, which is one reason we often keep our
mouths clamped shut when we’re frustrated and why many of us grind our
teeth at night while we sleep.
So it is useful to be able to let these muscles go. Massage the masseter
muscles with your fingertips, using a circular action and moderate pressure.
As you do this, direct the muscles to let go. You should think it, not speak it
(lest you need to use your jaw to give the direction!), but it should be an
actual word or phrase and it should be framed in a positive form: “Let it go”
or “Relax” or “The jaw bone feels heavy” but not “Stop tensing” nor “Don’t
clench your teeth.” We tend to understand autosuggestion (suggestions to
yourself) only as positive formulations, so “Don’t tense” can translate as
“Tense!”
There are internal muscles that also raise the mandible; these are the
medial pterygoid muscles that have a similar attachment, but on the inner
surface of the mandibular angle. They work with the masseter muscles to
raise each side of the mandible like a pair of slings. The other muscles that
you see in Figure 20 are the lateral pterygoid muscles that serve to pull the
mandibular joint forward, thus helping to open the jaw and also to protrude
the jaw. When tensed individually, they also move the jaw from side to side,
with some participation by the medial pterygoids. All lateral jaw actions are
more useful for the chewing process than they are for the process of
speaking.

Figure 20. Medial and lateral pterygoid muscles.

 The final paired muscles that help the jaw bite with force are the
temporalis muscles. These muscles attach to the front tip of the ramus of the
mandible at the knifeblade-like coronoid process. This is a very narrow or
focused attachment, using a tendon that extends from the coronoid
attachment upward through the space between the zygomatic arch and the
rest of the skull. The muscle fibers then fan out in a very broad attachment
onto the temporalis bone, the parietal bone and the sphenoid bone of the
skull.

Figure 21. Temporalis muscle.

Place your fingertips, with the fingers spread apart, onto your temples on
either side of your skull so that you can feel your temporalis muscles,
beginning this exploration with your jaw relaxed open. Now bring your
mandible up to occlusion, with your teeth together as though you were going
to bite down (well, “bite up” is more accurate, isn’t it?) onto something. You
probably don’t feel a lot of action at the temporalis muscles while the jaw is
closing, because your masseter muscles are doing most of the work. But as
soon as you bite “up” and apply pressure at the teeth, you will feel the
temporalis muscles pop out, showing that they have braced and thickened.
So we have a pair of muscles with a very broad attachment at the top,
meaning that the attachment is going to be very secure – a lot of muscle
tissue but with short muscle fiber length. This means that a lot of force can be
utilized over a very short distance, and a tendon attachment that focuses and
multiplies this muscles force into a part of the mandible that can act like a
lever – using the versatile joint at the condylar process as a fulcrum – that
further amplifies the biting power. No wonder we are omnivores. No wonder
we wear our molars down throughout life. No wonder people who grind their
teeth in frustration get tension headaches focused at the temples!

The great all-purpose jaw release exercise

1. Crook your thumbs at the first joint so that they can bend around the
angles of the mandible on either side. Place your thumbs there; feel the
thumbnail resting against the large diagonal muscle that extends from
your skull behind the ears down and forward to your collarbone. Very
gently pull the jawbone slightly forward. Leave your thumbs where they
are, resting in that depression between the muscle and the ramus of your
mandible.

2. Now place your spread fingertips on the temporal muscles as before.

Apply slight pressure with your fingers and with the thumbs to
encourage the muscles to let go and the mandible to stay unretracted.

3. Slide your hands downward until the thumb-tip of each hand reaches the
angle of the mandible.

4. Turn your hands, pivoting at the wrists, so that they follow the line of the
body of the mandible forward and down. Your fingers will briefly
massage your mass-eter muscles on the way. Your thumbs continue to
use the line of the mandible as a guide.

5. Continue the gentle pressure with your fingers as you complete the action
of stroking the body of the mandible forward and down, right out to the
tip of the chin. Repeat the entire action at least five times.
This exercise encourages a relaxation of the jaw forward as well as
downward, which is the natural opening action of the mandible. It massages
your temporal and masseter muscles, encouraging them to release any
residual tension. As you do this exercise, continue to focus autosuggestion on
telling your jaw to let go.

Explore the range of jaw action

 Advisory: Do not attempt this next set of five exercises if you have any
pre-existing problems with your temporomandibular joint!

1. Slowly bring the jaw up to occlusion of your teeth for biting, as you did
when you were exploring the temporal muscle action. Let it relax open
again. Make sure that you do not retract the mandible. Imagine the
“energy flow” of gravity as being downward and very slightly forward.

2. Bring the jaw up again, but this time bring the upper front teeth and the
lower front teeth very gently together. This is the action that the jaw
generally takes when it raises in the process of articulating speech
sounds. Relax the jaw as before.

3. With the jaw relaxed open, gently protrude the mandible. Do not force it.
Relax.

4. With the jaw relaxed open, gently retract the mandible. Do not force it.
Relax.

5. With the jaw relaxed open, gently move the tip of the chin to the right.
Do not force it. Then move the tip of the chin leftward. Return to center.
Combining these actions, you can move your mandible in a circular action,
which is very useful for chewing but not particularly useful for speech. In
fact, you can move your mandible so that your chin describes three different
circles on three different axes. What might they be? Can you demonstrate
them? And lest we forget, do not force it.
Now that you have explored the anatomy and physiology of the jaw bone
(the mandible) I can reveal to you that in most of the work that follows, the
jaw’s primary responsibility will be to do nothing whatever except – and this
is very important – remain relaxed, but not forced, open. In the English
language, the jaw is required to raise for only six consonant sounds. They are
“s” as in “see”, “z” as in “zoo”, “sh” as in “show”, “zh” as in “Zsa-Zsa”, “ch”
as in “church”, and “dzh” as in “joy”. It does not need to raise upward for any
other consonants or for any of the vowels or diphthongs that form speech in
English. From a descriptive standpoint, of course, we all use our jaws a lot
more frequently in almost all our speech sounds to lessen the activity of the
other articulators. But precisely because we are in the habit of sharing our
tongue and lip actions with the jaw, I will ask that we explore the separation
of the actions by allowing the other articulators to enact their full range of
motion. Variable jaw positioning is important in forming accents, but in most
of our work here we will want to let the jaw go, and only use it for those six
consonants – which in English is often enough.
Maintaining a relaxed, open jaw position as “home base” for the
mandible while speaking has another qualitative advantage: it allows for a
more naturally open vocal tract, which enhances complex resonance of the
voice; most of us are fairly good at keeping our mouths shut – literally if not
metaphorically – but actors and other professional voice users need to be able
to let our very individual voices flourish.

The lips and cheeks

The simplest way to describe the muscles that control the lips is to think of
them as resembling a child’s drawing of the sun. The ball of the sun here is a
simple circle, but to emphasize the sun’s importance we have drawn this
circle as a double line. Radiating out from the “sun” are its “rays of light”,
lines going straight out in different directions from the circle.

Figure 22. A child’s sun.

The muscles of the lips work in a similar way. The circle of the “sun” is
the muscle that goes all the way around the lips; it is called the orbicularis
oris muscle. Actually this muscle is divided into an inner ring, located at the
“roseate” part of the lips (the part that we usually identify as the lips), and an
outer ring that fills much of the rest of the upper- and lower-lip area.
(Interesting note: the discovery that the orbicularis oris muscle has inner and
outer rings is a fairly recent one. One doesn’t often find recent discoveries in
the field of human anatomy.) The “rays of the sun” are attached at one end to
one or both of these rings, and at the other end to skull bones or to other
muscles, allowing the “ray” muscles, through contraction and release, to
change the shape of the rings – the lips.

Figure 23. The facial muscles that control the lips.

That is the simplest description – perhaps a little too simple. After all, we
know very well that we have an upper lip and a lower lip, both of which can
move independently, and that there are lip corners at the right and left sides
of the mouth that are controlled by more muscle “rays” than the rest of the
simple circle we have described before. So the “simple circle” can be pulled
into many different forms by the muscles around it.
 Figure 24. Lip action possibilities.

Warming up the lips

Hook the little finger of each hand around each lip corner. Let your lips relax.
Gently stretch the lip corners apart and forward to stretch the orbicularis oris
muscle, as in Figure 25. Gently pull the corners upward with your fingers;
then gently pull them downward.
You can also exercise the orbicularis oris muscle by tensing against the
action of the fingers with your orbicularis oris muscle. Try to pull your
fingers toward one another by using your lips muscles.
Unhook your fingers, and blow through the lips in a loose trill, first unvoiced
(the classic horse imitation) and then voiced (the classic “I’m cold”
expression).
 Figure 25. Lateral lip stretching.

1. Lip isolations
Begin all of the following exercises with lips relaxed and jaw dropped.
Between isolation exercises it is a good relaxation to blow through the lips,
unvoiced and voiced, keeping the lips and the cheeks very loose.

Figure 26. Lip corner retraction.

a Lip corner retraction (single). Pull one corner of your lips, on either the
right or the left side, directly back toward the back or ramus, of the
 mandible. Isolate the action. Then release. Try it again using one lip
corner, then the other, alternately.
b Lip corner retraction (double). Pull the corners of the lips directly back
toward the back of the jawbone. The effect produced is that of a hideous,
mirthless grin, stretching your lips tight across your teeth. If you wish to
obtain the full psychotic effect, bug your eyes out simultaneously. Don’t
pull the corners up into an ordinary, commonplace, traditional grin. Relax.
Repeat.

Figure 27 Right and left lip corner retraction.

The muscles we use to retract the corners of the lips are the risorius muscles.
They attach in front to the orbicularis oris muscle at the corners of the lips;
in the back they attach to the ramus of the mandible. When engaged, they
draw the lip corners directly toward the back of the mandible.
You can also exercise the risorius muscles by gently pushing your index
fingers into your cheeks just behind the orbicularis oris muscles, then pulling
the lips corners back as in Exercise b, providing resistance to the risorius
action.
 Figure 28. The risorius muscle.

c Lip corner protrusion. Push the corners of the lips directly forward but
not intentionally inward toward one another. Relax. Repeat. Make sure that
you isolate the action to the lip corners only and do not involve the rest of
the lip. (See Figure 29.) When you do this action, you are tightening the
outer circle of the orbicularis oris muscle but leaving the inner circle
relaxed; as a result the inner circle is pushed outward.
d The “aside”. Simultaneously engaging the outer circle of the orbicularis
oris muscle and one of your risorius muscles, protrude one corner forward
and pull the other corner back at the same time. Alternate. This exercise
has the effect of moving your entire mouth around to one side of your face,
which is useful when you wish to confide something to the person next to
you without other people reading your lips. (These other people may,
however, notice that your mouth has seemingly disappeared.) This action
is also useful if you are imitating a flounder, although expert flounder-
imitators supposedly can also migrate both eyes over to one side of the
skull, a more challenging task.
 Figure 29. Lip corner protrusion.

e Pursing the lips, 1. With the jaw relaxed open and the lips naturally
dropped apart, bring the upper and lower lips together. Release them apart.
Repeat. Do not take any action with the lip corners. When you do the
action of pursing, you are tensing the inner ring of the orbicularis oris
muscle while bracing the muscles that attach to the lip corners.
f Pursing the lips, 2. Without using the corners of the lips, curl the lips
inward around the teeth and into the mouth as far as possible, like a bad
imitation of a person with no teeth. From this position, swing the lips out
and away from the teeth as far as possible, like a trumpet. Then swing
them back into the mouth. Repeat. When you perform the action of lip-
curling inward, you are tensing the inner ring of your orbicularis oris
muscle without bracing any of the other muscle “rays” extending from the
periphery. When you swing the lips outward, you are tensing the outer ring
and also bracing some of the other muscles that raise and lower the edges
of the lips.
 Figure 30. Lip curling.

g Isolating the upper and lower lip. Begin as with (e), but after the lips are
stretched in, swing just the upper lip out, leaving the lower lip in. Then
follow with the lower lip, isolated, swinging out. With both lips out, swing
only the upper lip in, and then the lower lip. Repeat.
h Changing the pattern. Explore variants of (g), leading with the lower lip
rather than the upper lip, going to complete lip relaxation between actions,
repeatedly working one isolated lip, etc.
i Full lip-rounding. Push the corners of the lips forward as in (b) and keep
them forward; then purse the lips until they touch (as in a kiss), with the
lips fully closed. Relax the pursing action but keep the corners of the lips
forward. Repeat. The overall effect is that of the traditional fish imitation,
generally embellished by its best practitioners – four-year-olds – with the
hands at the ears to imitate gills. (This may remain optional here.) After
five repetitions relax the corners of the lips. Repeat the entire action. Full
lip-rounding is a combined use of the outer and inner rings of the
orbicularis oris muscle.
 Figure 31. full lip rounding.

j Bilabial speech action. Do a few “p” and “b” sounds, feeling the muscular
action of bringing the lips together in a pursing action until the flow is
stopped; feel the build-up of air pressure behind the closure and then the
release of the muscles aided by the release of the compressed flow as the
flow is “exploded” out of the mouth.
k Lip corners up. The comic mask. Pull the lip corners directly upward
toward the very front of your cheekbones. To enact this action you are
using the levator anguli oris muscles. This is not a smile that you will
want to use on your friends. Relax the muscles.
l Lip corners down. The tragic mask. Pull the lip corners directly downward
(Figure 32). Here you are using the depressor anguli oris muscles. Relax.
If you wish to explore tragi-comedy, try raising one lip corner and
lowering the other.
 Figure 32. The tragic mask.

m The genuine smile. This time allow the lip corners to move diagonally
upward toward your cheekbones (Figure 33).
Here you are using your zygomaticus muscles. Of course, in order for this
to be a genuinely genuine smile, you also need to have friendly feelings for
its object.

Figure 33. The genuine smile.

n Lip corners diagonally downward. Unlike the tragic mask, where the lips
corners are pulled directly downward, if you pull the lip corners diagonally
 downward, you will note a lot of muscle fibers tensing in your neck
(Figure 34). This is because you are now using your platysma muscles,
which attach to the lip corners and then fan out in thin sheets on either side
of the neck to a broad attachment at your collarbone (or clavicle).

Figure 34. The platysma pull.

There are other “rays” extending from the orbicularis oris muscle that can
raise or lower portions of the upper or lower lip. Although they are important
in the communication of facial expression, or when you want to imitate a
rabbit or Elvis, they have less importance in the articulation of speech, and I
will not complicate your life further by exploring them here.
I will, instead, complicate your life with the following:

Cheek isolations
a Cheek muscle release: with the jaw still relaxed open (important!) purse the
lips so that they almost touch. Leave the cheeks absolutely relaxed. Blow
air out between the lips so that the cheeks puff out; then inhale with equal
force so that the cheeks pull in against the teeth. Repeat two or three times;
no more than that, however, since you may start to hyperventilate. If you
start to feel dizzy, stop breathing for a few moments.
 Figure 35. Buccinator muscle.

b Thicken the tongue to “fill up” one side of the mouth but not the other, so
that airflow can escape only around one side and past one cheek. Blow air
out with moderate force. This will produce a slow, rather sloppy cheek
trill, either unvoiced or voiced. Do the same with the other cheek. (Hint:
you may need to advance the lip corner of the “trill” side slightly.) Then
shift the tongue to the center and try to trill both cheeks simultaneously.
This may take a little practise, but the amusement (or possibly horror) of
your friends will quite obviously justify the time spent.
c With the jaw relaxed open, tense (or bunch) the cheek muscles against the
teeth, and then release. Repeat five times. Remember that you are tensing
and releasing the muscles only, and not using the muscles to move the
cheek area in any direction. When you take this action you are tensing your
buccinator muscles.
d Isolate the tensing and releasing action that you did in (c) to the area of the
cheeks just behind the lip corners. The cheeks will be tensing into the
incisors and/or bicuspids. As always, make sure that you are not using your
jaw in this action.
e Buccinator isometrics. Place the tip of your tongue against the inside of
your left or right cheek. Push the cheek outward to the side with your
tongue muscles. Then, using the buccinator muscles of your cheek, pull the
cheek back in, against the pressure of the tongue action. Do the same
exercise with the other cheek.
The tongue
As the next two illustrations show, the tongue is not a single muscle, but
many muscles. Some of these muscles exist wholly within the tongue itself
and are termed intrinsic muscles. Some of them extend from the tongue to
other skeletal structures, such as the skull or the hyoid bone. These are
termed extrinsic muscles.

Figure 36. Intrinsic tongue muscles.

 Figure 37. Extrinsic tongue muscles.

Tongue isolations
a “Stretching” the tongue. Protrude the tongue-tip and blade straight out of
your mouth as far as you can stretch them comfortably. Actually, what you
are doing is pushing the tip and blade out of your mouth by tensing (and
shortening) the back part of an important tongue muscle that occupies a lot
of the rear portion of the oral cavity. As Figure 37, above, illustrates, the
fibers of this muscle are attached broadly to the body and the blade of the
tongue at one end and are attached to the hyoid bone and – primarily – to
the mandible. This muscle is called the genioglossus muscle and in
protruding the tongue forward you are using specifically the rear or
posterior portion of the genioglossus.

Retracting the tongue back into you mouth uses the forward or anterior
fibers of the genioglossus muscle.
 Figure 38. The tongue: coronal section.

b Raise the tongue tip. With the tongue protruded from your mouth, curl the
tip upward. To accomplish this action you are tensing (contracting) the two
superior longitudinal muscles of the tongue (Figure 39). As the name
suggests, the fibers of these muscles – located to the right and left of the
mid-line – run from the tip of the tongue toward the back of the tongue just
below the upper surface (in other words, just below your taste buds. Relax
the tongue.

Figure 39. Raising the tongue tip.

c Lower the tongue tip. Curl the tongue tip downward. Here you are using
the inferior longitudinal muscles of the tongue (Figure 40). Again these
muscles are paired – right and left – and the muscles fibers run front to
 back. The muscles are located just above the under-surface of your tongue.

Figure 40. Lowering the tongue tip.

d Curl the tongue tip to the right. You are now tensing and contracting the
right superior longtudinal muscles and the right inferior longitudinal
muscles (Figure 41).

Figure 41. Tongue tip to right.

e Curl the tongue tip to the left. As will come as no surprise to anyone who
is inclined to recognize patterns, you are now tensing and contracting the
left superior longitudinal muscles and the left inferior longitudinal
muscles (Figure 42).
 Figure 42. Tongue tip to left.

f Channeling and bunching. The jaw is relaxed open. Easily slide your
tongue out onto the lower lip – mostly so you can see it better. Stretch the
side edges of the tongue apart by flattening the tongue; that is, you are
bringing the superior and inferior surfaces of the tongue toward one
another (Figure 43). In so doing, you are contracting your verticalis
muscles; as the name implies, the fibers run vertically. When this action is
taken, there is a natural tendency for your tongue to “channel” slightly,
with the side edges raising a bit and the midline lowering.

Figure 43. Chaneling the tongue.

Relax the verticalis muscles and the tongue will rediscover its
comfortable plumpness. Now you’re going to do the opposite action: bring
the same side edges of the tongue in directly toward one another, thickening
the center of the tongue along the midline (Figure 44, opposite page). Relax.
Here you are tensing and releasing your transversus tongue muscles that run
from one side edge of the tongue to the other.
In bunching the tongue, do not curl the side edges of the tongue upward
into what can perhaps be described as a rolled fleshy cannolo (plural:
cannoli) as you try to pull those side edges toward one another. Repeat each
of these two actions in sequence. The effect is that the midline of the tongue
lowers when you are channeling the tongue and raises when you are
bunching the tongue.

Figure 44. Bunching the tongue.

g Arching and cupping the tongue. Front: the jaw is relaxed open. Keep the
tongue tip at rest behind the lower teeth. Arch the front of the tongue
upward toward the alveolar-palatal area to form an “eeeee” vowel.
Whisper, but do not voice, the “eeeee.” Then cup the front of the tongue
sharply downward – keeping the cupping action as forward-placed as
possible – to form a “flat A” sound; as in, appropriately, the word “flat”.
Again whisper the sound. Repeat the two actions in sequence, first very
slowly moving from “eeeee” to “aaaaa”, and then moving energetically
from one to the other. Now do the entire sequence voiced. Keep the jaw
relaxed open throughout this exercise. Here, as you probably realize, you
are using your inferior longitudinal muscles to arch the tongue and your
superior longitudinal muscles to cup the tongue.
h Lowering or cupping the back of the tongue. With the tip of the tongue
comfortably behind the lower teeth, isolate the action of lowering the back
of your tongue directly downward, then releasing it. Because the superior
longitudinal muscles do not function efficiently to lower the greater bulk of
 the back of the tongue, the action enlists the hyoglossus muscle, which
pulls the back of the tongue downward toward this muscle’s attachment to
the hyoid bone.

Figure 45. Hyoglossus muscle.

In many speech books and in some speech science texts, you will find the
assertion that when the hyoglossus muscle pulls the back of the tongue down,
there is also a reactive movement backward by the root of the tongue, thus
constricting the pharynx. This is often true as an observation, which is why
we hear a lot of people utter a constricted sound when they say “ahhh”. Test
this yourself: do your best “dentist’s office ahhhh” and hear if your voice
thins out. But while this constriction can happen, it does not need to happen.
If you use your posterior genioglossus muscle to pull the tongue root forward,
this constriction can easily be avoided. (Remember the “professional yawn”!)
i Arching and cupping the tongue, back. Perform the same action as in
Exercise g, but this time isolate the back of the tongue arched up first to
form a long “U” sound, “ooooo”, and then with the back cupped down to
form a broad “A” as in “aaaahhhh”, as in Exercise h. Repeat the sequence
from “ooooo” to “aaaahhhhh” first very slowly and then quickly. The
rounding of your lips on the “ooooo” helps in the formation of the sound
but should not be the focus of your attention; stay focused on the tongue
action. Keep the jaw relaxed open throughout this exercise.
j Advancing and retracting the tongue root, 1. Without moving the jaw
 from its relaxed open position, protrude the tongue forward from the back,
letting the tongue-tip and blade slide out onto the lower lip. Then pull the
tongue root directly back to constrict the laryngopharynx. Try not to tense
the tip, blade, or front of the tongue; isolate the action as far back as
possible. Relax the tongue-root forward. Repeat. For the retraction of the
tongue-root, you are using your styloglossus muscles, which attach to little
prongs pointing downward from your temporal bone at the back and
bottom of your skull and then loop around the underside of the tongue like
a sling.

Figure 46. Styloglossus muscle.

k Advancing and retracting the tongue root, 2. Repeat the previous

exercise but this time, allow the tip of the tongue to stay behind the lower
teeth. Isolate the action by limiting it to the tongue root itself.
l Tongue-tip to alveolar ridge and to palate, 1. Keeping the rest of the
tongue as relaxed as possible and the jaw relaxed always open, bring only
the tip of the tongue up to touch the gum ridge (the “alveolar ridge”) just
behind the upper teeth. Relax the tongue fully away so that the tip drops
down to just behind the lower teeth. Repeat. Now curl the tip of the tongue
upward and back so that the tongue tip touches the middle of the hard
palate. Can you do this without thickening the rest of the tongue? Relax the
tongue fully: does the tip quickly relax down to a position just behind the
lower teeth?
m Tongue-tip to alveolar ridge and to palate, 2. This time, allow the entire
front edge of the tongue, including the tongue-tip, to raise and seal off the
airway at the alveolar ridge behind the upper teeth. Make sure that you still
isolate this action. Do not involve the jaw in the action. Using a flow of air
from the lungs, explode the tongue away from the alveolar ridge back to its
resting position, producing a “t” stop-plosive sound. Repeat with a voiced
flow, producing a “d” stop-plosive sound. Curl the tongue-tip and the edge
of the tongue back to seal off the airway at the palate. Again, use a flow of
air to explode the tongue away from the palate producing, this time, an
unfamiliar-sounding “t”. Repeat with voicing, producing an equally
unfamiliar “d” sound. Finally, explore as many possible placements of the
tongue-tip and edge between the alveolar and palatal placements as you
can. Continue to isolate the action to only those parts of the tongue that
must work to complete the action effectively.
n Rolling the ball. Imagine that your tongue is cupped to hold a large marble
or – a more edible image – a grape, on the blade of your tongue. Slowly
roll the imagined marble back along the mid-line of your tongue as far
back as you can go before you would swallow it, were it real. Then slowly
roll it forward, keeping the tongue fully cupped. Repeat. Don’t pull the
whole tongue back: just roll the ball.
o Moving the mountain. Arch the front of your tongue as though you were
about to say “eeeeee”. Slowly roll that arch back until the arched back
portion of the tongue is just under the velum (the soft palate). Do not
retract the tongue in this (or the previous) exercise: the tongue-tip should
stay right behind your lower teeth.
p Bracing. Channel the tongue, with the midline low and the edges high.
Bring the side edges of the tongue up to meet the inside of the upper teeth
along the molars and rear bicuspids. Brace the tongue outward against the
teeth, pushing laterally. The probable result will be that the midline raises
slightly toward the roof of the mouth. Focus the effort of the bracing
selectively from the bicuspids to the rear molars.
Practice each of these isolations of the lips, cheeks, and tongue repeatedly,
making sure that you focus on only one activity and that you do not tense
irrelevant muscles. It is especially important that you try to keep the mandible
relaxed open at all times, so that the action is focused into the other
articulators. It is more important that you practice the isolations frequently,
even if you do not do so for a long period of time in any given work session.

Terms to know

Articulators
Points of articulation
Bracing
Contracting
Tendon
Ligament
Frontalis muscles
Orbicularis oculis muscles
Mandible
Body
Ramus
Condylar process
Coronoid process
Temporal bone
Temporomandibular joint
Articular disk
Zygomatic bone
Zygomatic arch
Masseter muscles
Medial pterygoid muscles
Lateral pterygoid muscles
Temporalis muscles
Parietal bone
Sphenoid bone
Temporal bone
Occlusion
 Orbicularis oris muscle,
inner and outer rings
Risorius muscles
Levator anguli oris muscles
Depressor anguli oris muscles
Zygomaticus muscles
Platysma muscles
Buccinator muscles
Intrinsic, extrinsic
Genioglossus muscle,
posterior and anterior portions
Longitudinal muscles
superior and inferior, left and right
Verticalis muscles
Transversus muscles
Hyoglossus muscle
Styloglossus muscle
Clavicle
5 “Phthong” shaping

In exploring the ways we can obstruct the unvoiced or voiced flow to form
consonants, we were able to feel a vibratory buzz – or just a rush of air – in
specific, very limited areas of the vocal tract. The shaping we are about to re-
explore – because we’ve done it before – will not have that specific focus of
feeling within the mouth, but in exchange we will be able to feel the
vibrations throughout the body.
I would like you to try to ignore the word “vowel” as a label for this
unobstructed voiced flow for awhile, because the word “vowel” suggests the
standard “a”, “e”, “i”, “o”, “u” (and sometimes “y” and “w”) that we use in
English. Don’t let yourself be limited by this limited selection of sounds.
There are infinite gradations of sound change possible here and they all are
just as worthy as any others.
Instead we can use the linguistic term monophthong. It’s a longer word
than “vowel”, but it’s more interesting to say. (I especially like the “phth”
sequence.) “Monophthong” means a single voiced, unobstructed sound.
While we are on the subject, we are probably more familiar with the term
diphthong, which means a glide from one monophthong to another. We use
these all the time in words like “voice”, “crime”, “house” and so on.
We also use the occasional triphthong, sliding from one sound to another
to a third, as in words like “sour” or “fire”.
So the general term that might tie all these more specific descriptions
together is phthong. This is a word that is not usually included in the
linguist’s lexicon, but it certainly is easy to remember. It comes from the
Greek word φδογγος, meaning “sound”. Oh, by the way, voice the “th”.
You can explore this work (and this new word) in a standing position, or
sitting, or lying down. But make sure that you are “actively relaxed”, that
your entire body is available to receive and transmit vibrations.

Exploration 1
Focused gurning on tone. You will recall that we previously explored
gurning, making strange faces by using all our muscles of expression. Now
we are going to use our awareness of articulator isolations to explore the
kinds of unobstructed shaping of sound that might be used in language.

Begin by relaxing your jaw open. Leave it open throughout this work. It is
easy enough to reintroduce jaw action into the mix of actions, but for now we
need to feel the contributions that the other shaping actions are making
without the mediation of having the jaw raise and lower to vary the size of
the oral cavity.
Begin an unvoiced flow of breath, breathing egressively. (But not
aggressively! Keep the flow gentle.) Begin to shape that flow by slowly and
attentively moving one or more of the “vowel articulators”. Make sure that
you are never obstructing the flow. Just get used to exploring different
shaping actions without having to hear a defined sound result. Don’t go back
to gurning – use only the vowel articulators. Your lips and cheeks should be
the only outside parts of your face that are moving. Everything else is
happening inside.
Now let the flow change easily from unvoiced to voiced. But keep your
focus on the feel of the movement of the vowel articulators (excluding – for
the time being – the jaw): the muscular circle of the lips and all the muscles
that radiate from it, the cheeks, the body of the tongue, the velum, the
pharynx. Allow the movements of the articulators to be slow and easy as you
release vibrations on a series of long voiced sighs.
Audio track 15: Focused gurn shaping.
Allow yourself to listen attentively to the resulting sounds. Note the actions
that seem to change only the quality of the sound; those that seem to change
the loudness; those that seem to change the monophthong sounds (familiar or
unfamiliar); and those that seem to change more than one of these categories.
After a bit of exploration, most people generally observe that quality
change happens most strongly in pharyngeal actions, that monophthong
change is produced most strongly in the mouth through the action of the
tongue, lips, and cheeks, and that amplitude (loudness) change can be
affected by any of the articulators. But the actual mix is very individual, and
all the articulators make some contribution to all the categories.
Go through some “quality” changes on the same monophthong.
Audio track 16: Acoustic voice quality.
Play with sounding nasal, denasal (as though you have a head cold), hollow,
constricted, and so on. Work anywhere in your pitch range as you do this
exploration.

Exploration 2
Isolate phthong change. Sound an easy egressive voiced sigh with your jaw
relaxed open. Be attentive to the “vocal quality” – it doesn’t matter what it is,
but one of your main tasks in this exploration will be to keep that vocal
quality the same throughout the work.
Continue to release a series of voiced sighs and start to shape the sound
by shaping the articulators – always without obstructing the flow. Keep the
vocal quality consistent and also the amplitude so far as possible. Isolate the
change to the monophthong sounds you are producing. Try to eliminate
actions that don’t actually produce some monophthong change.
Audio track 17: Phthong shaping.
Probably you will note through the feel of the actions that you are using a
more limited set of actions to shape the flow. For one thing, you will note that
the tip and blade of the tongue do not contribute much to the variation in the
phthongs; the work is centered in the dorsal portion (front, middle, or back)
of the tongue. So the tip and blade can stay relaxed. Explore using the dorsal
actions fully, though. Let the sound always be changing slowly and smoothly
through sounds that are new and unfamiliar as well as familiar. You will also
note that – with the jaw remaining relaxed open – your lips move from
slightly spread at the corners to fully rounded.

Exploration 3
Enter the maraphthong. This isn’t a serious linguistic term, but it’s a pretty
good description of what you’re doing right now. It’s more than a
monophthong. It’s more than a diphthong. It’s more than a triphthong. It’s a
marathon glide though constantly changing sound actions on a series of
voiced sighs. It’s a maraphthong.

Exploration 4
Move/freeze/move. Isolate your action to maraphthong sound change (so
you keep the same sound “quality” and the same amplitude. Focus your
attention on the physical actions of the vowel articulators as the sound flow
changes on a series of easy long-voiced sighs. After about ten seconds, freeze
the articulators in place and be aware of the feel and sound of the single
monophthong voiced sigh. Hold the articulators in position and register the
emotional quality of the sound. Where does it seem to resonate within your
body? Where does it seem to resonate within your psyche? Enjoy the sound.
Get comfortable with the sound. You will probably need this sound –
whatever it is – in your work on accents. Resist the temptation to adjust the
articulator action and therefore the resulting sound toward a more familiar
sound product. Of course you may find that you have frozen the articulators
into a familiar sound initially, which is fine. Just make sure it doesn’t happen
all the time.
Audio track 18: Move/freeze/move phthlong shaping.
After you register the isolated sound result by listening and feeling the
vibrations, go back to sounding your maraphthong. After another ten seconds
or so, isolate the articulator action again. And so on.

Exploration 5
Move/freeze lips/move. Begin the exercise as in Exploration 4. But this time,
when you freeze the articulators, freeze the lips only and continue exploring
the action of the dorsal part of your tongue. What effect does this selective
immobility of the lips have on the experience of the tongue action? What is
the effect on the resulting sound? Then go back to the maraphthong and after
a few seconds, freeze the lips again in a different position. Repeat.
Audio track 19: Move/freeze lips/move phthlong shaping.

Exploration 6
Move/freeze tongue/move. Begin again with your maraphthong exploration
but this time, when you freeze the articulators, freeze the tongue action only
and continue to explore lip rounding or spreading. Note the effect on the lip
action, if any, and on the sounds that result. Then resume the maraphthong
and after a few seconds, freeze the tongue action again. Repeat.
 Audio track 20: Move/freeze tongue/move phthong shaping.

Exploration 7
Sharing the monophthong. If you are working in a group, form a circle.
(Warning: If you are alone, do not try to form a circle, as injury may result.)
Let one person in the circle begin a maraphthong glide on a voiced sigh. Then
let that person freeze the articulators into a monophthong release. As this first
person holds the monophthong sound, a person next to the first person should
try to shape her or his articulators so that the same sound is reproduced. Once
the monophthong has been “passed” from the first person to the second, the
maraphthong-to-monophthong action of shaping a new sound is repeated
around the circle until each participant has explored receiving a sound,
exploring for a new sound, and sharing that sound with the next person.
6 Obstruents: obstructing the flow

Instinctively we know what makes an “obstruent” sound different from a

“phthong” sound in human speech, even if we haven’t met these terms
before. But we may not know why our instinct knows this. In “phthongs” we
flow vibrations out of our mouths in an easy unobstructed flow of voiced
sound. We may change the shape of the vocal tract (from the vocal cords to
the lips) somewhat like the twists and turns of a river: the form changes (in
speech when we change one vowel sound to another) but the flow goes on
with nothing restricting it.
“Obstruents” are different. When we produce obstruents we move one or
more articulators so that they get in the way of the flow.
There is a more familiar term for this class of sounds: consonants. But
for the time being we are going to use the more archaic term “obstruents”
because it is more descriptive of the physicality of the action and also
because it is useful for us to think of these actions as something new and
unique, not the same limited set of consonant actions that (if English is your
primary language) you have already learned so well.

“Obstruents” vs “consonants”: some more argument

There are some other reasons to use the term “obstruent” instead of the
term “consonant” at this stage of our work. “Consonant” derives from
the Latin word consonans meaning “sounding together”. In one of its
forms, “consonant” is defined in the Oxford English Dictionary as
“Agreement or pleasing combination of sounds; harmony, concord.” But
the physical action here is one that gets in the way of the flow: it is
intrusive, it impedes, it obstructs. Obstruent is a more accurately
descriptive term.

So obstruents are defined as articulator actions that get in the way of (i.e.
obstruct) the flow – sometimes a lot, sometimes only a little; sometimes so
little, in fact, that the obstruents are almost vowels, such as “w” or “y” or “r”
or “l” in English.
Try these sounds a few times like this: wah wah wah, yah yah yah, rah
rah rah, lah lah lah. Can you feel that your tongue is impeding the flow only
very slightly? Where do you feel the slight change in the voiced flow within
the vocal tract and especially on the tongue?
Sometimes the articulators obstruct the flow so much that it is blocked
altogether for an instant. As in “b” or “d” for example.
Say – and make sure you can feel – the sounds: bah bah bah, dah dah
dah. Tie them together on one flow of breath: bahbahbahdahdahdah. The
action of the flow continues but the stream is blocked until its energy builds
up enough to explode past the barrier.
Sometimes the obstruction of the flow is strong enough so that the
consonant’s quality lies in the flow trying to get past the obstruction. “Z” or
“v” or “s” or “sh”.
Say out loud sahsahsahsahsah then zahzahzahzahzah. The “ah” is always
voiced, but note that you produced the “s” sound without the vocal cords
vibrating and “z” with the vocal cords vibrating. So you experience that some
of these consonant obstructions of the flow, as in the “s” sound you just
made, are strong enough to make their own unique sounds without voicing
(vocal fold vibrations) at all.
You can test this. Put one hand lightly on your “Adam’s apple” at the
front of your larynx. Take a deep breath and on one flow out from your lungs
release a long energetic sssssszzzzzzsssssszzzzzzsssssszzzzzz, turning the
vibrations on and off and on again. You can feel with your hand when the
vibrations are on and when they are off.
On the following pages you will receive and follow a recipe – a pretty
simple one – for making all the obstruent sounds in the world’s languages
that are powered by a flow of breath energy from your lungs, which is the
way speakers around the world form most of their obstruent sounds – though
not quite all of them. I am keeping this simple, because I want you to explore
freely the possibilities for making as many different obstruent actions as
possible.

Ingredients
1 A voiced or unvoiced flow of air moving from the lungs through the
vocal tract and out your mouth – or sometimes your nose.
2 An action by the articulators to obstruct the flow, a lot or a little.
Complete stoppage would be one extreme of action, a tiny intrusion
would be the opposite extreme, and the rest of the actions would be
everything in between. But your listener would have to be able to
differentiate the sounds these actions produce.

3 A place in your vocal tract (anywhere between your vocal folds and your
lips) where the action happens.
Before we can combine the ingredients, we need to know a little more about
them, so that when we start to speak obstruents we can go beyond the ones
we are used to and actually find some new sounds. The flow can be either a
voiced flow of vibrations (remember “zzzzzz”) or an unvoiced flow of air (as
in “ssssss”). That’s the only variable in the flow.
Now just combine the ingredients. (I said this was a simple recipe.) In
other words, start saying some obstruents out loud. To make them easier to
say, put a voiced “ahhhhh” phthong after every obstruent. (This is the so-
called “broad ‘a’”, the “ahhhh” sound the dentist tells us to make to get our
mouths open.) We also remember that some of our obstruent sounds are
formed using a voiced flow of vibration, others using an unvoiced flow of
breath.
Put together strings of obstruents on one flow release of breath with the
obstruents separated by the “ahhhh” vowel. Say them slowly but with energy.
These usually aren’t going to sound like words as we know them. You are
exploring the feel of the articulation actions themselves. There are various
names for what we are doing – “doubletalk”, “gibberish”, “nonsense words”,
and so on.
Audio track 21: Free-form obstruents.
Even if it feels and sounds strange, keep going. Keep the flow energetic.
Ignore those puzzled stares from co-workers, fellow students, or loved ones.
Don’t be limited by the familiar “consonant” sounds you’re used to making.
Explore the possibilities of new obstruent sounds even if you have never
heard them before in any language.
When you have attained maximum variety and audacity of obstruent
formation – staying with the phthong “ahhhh” as you do so – turn your
attention to the actions your articulators are taking. How close together are
they getting? Are they interrupting the flow or just getting in the way a little?
Do the articulators vibrate against one another repeatedly or touch and
bounce away? What path through the vocal tract does the flow – voiced or
unvoiced – take? What actions make a consonant sound different from
another consonant that is similar to it in other ways? What do these
differences sound like and feel like?
Finally, expand your attention to an awareness of the place (or even
sometimes the places) in the vocal tract where the action is happening.
Consonants can form anywhere between your lips and your vocal folds (and
include both); the only requirement is that the action can obstruct the flow to
some degree.
The goal here is to come up with as many different ways of obstructing
the flow with your articulators as you can, even if some of the sounds that
result are rather nasty. Play freely with these articulatory actions, much in the
same way that a very young child explores different vocal sounds on the way
to shaping them into language. But play with an adult’s sense of attentiveness
and an adult’s sense of rigor. Vary the tempo of this strange obstruent
mélange. Vary your pitch inflections. Vary your energy level, but don’t let
the energy progressively subside. When it stops being play, stop doing it. But
come back to it.
Part 2
Finding language
7 Exploring the limits: Outlandish

We now are ready to put together our work in “Obstructing the Flow” and
“Phthong Shaping” so that we can produce a sequence of sound actions that
resemble a very odd language. It isn’t language really, because it doesn’t
have any direct meaning, except the variable meanings that might be called
up by the sounds themselves.
The only limits to this strange “language”, which we will henceforth call
“Outlandish”, are what you can physically enact (without injury!) as vocally
produced sound: voiced or unvoiced sound that is produced using a flow
shaped or obstructed by your articulators. Outlandish is – and should be -
outlandish. Compared to Outlandish, Klingon in Star Trek is a language for
wimps.
As you speak it, you will be able to collect some useful information about
sounds that fit easily into connected speech actions, and those which are so
difficult to execute in connected speech that they (and you, and your
listeners) would be happier if they stood alone. But as you speak Outlandish,
your articulators will get some excellent exercise, and you will be able to
express yourself with a kind of perverse eloquence that is fascinating in itself.
Audio track 22: Outlandish.
You will also start to get some useful information about how “phthongs”
(monophthongs, diphthongs, triphthongs, and beyond) and Outlandish
obstruents interact in connected speech. Putting obstruents together in a
sequence without any phthongs in between is called a obstruent cluster.
(Well, it’s actually called a “consonant cluster” but you have parked the term
“consonant” in an unused portion of your brain, haven’t you?) You will start
to sense the limitations of how long a given obstruent cluster can go on
before what you are saying stops sounding like a language at all. You have to
throw in a phthong or two fairly frequently. We aren’t going to analyze this
obstruent–phthong ratio from a linguistic perspective; we are just
experiencing it in practice.

Exploration 1
Group outlandish. Imagine that the participants in this exercise are all
citizens of the Lost Continent of Outlandis, where all governance is
conducted by the group, and all debate at meetings consists of one-word
questions and one-word answers.
One Outlandian begins by asking a one-word question in Outlandish to
another person in the circle. (We know it’s a question because it has an
upward inflection.) The person so addressed repeats the question – this is an
important tradition! – and then replies with a decisive one-word answer
(downward inflection) in Outlandish. The answerer then asks an Outlandish
question to another Outlandian in the circle. The process continues until all
participants have asked a question and received an answer, at which point,
according to the traditions of Outlandis, all discussion ends and it is simply
assumed that a decision has been reached. This business method goes a long
way toward explaining why Outlandis is a lost continent.

Exploration 2
Outlandish anecdotes. Tell an Outlandish anecdote lasting at least thirty
seconds to the group, or to yourself in a mirror, or into a recording device.
You should know what the story is, but under no circumstances should you
rehearse. Just let the sounds happen. Try to maintain variety of sound
production. Or, to put it the opposite way, don’t allow yourself to become too
fond of only a few sounds. Make sure that you are using both voiced and
unvoiced obstruents.

Exploration 3
Outlandish singing. If you or your fellow Outlandians discover that you are
still too fond of interminable strings of obstruents, there’s nothing like a good
song to restore a sense of balance between obstruent and phthong. So sing!
Sing the Outlandish National Anthem. Sing some other country’s national
anthem in Outlandish. (There will be a small prize for the most “melisma” –
improvising phthongal vocalization around the melody.) Croon. Rap.

After exploring Outlandish and the physical limits to which it takes us, we
are ready to refine our exploration toward those actual language sounds used
on the real continents of the world.
8 Obstruents within language

The selection of speech actions that occurs when we move from Outlandish
to the repertoire that human beings use in the languages of the world is one
that has entirely to do with obstruents. In moving from Outlandish to
language, many of the most succulent obstruent sounds reluctantly leave the
scene, to be brought back only when we need a special “rude noise”. We let
them go because for the most part they are simply inefficient to say when a
person is putting sounds together in connected speech. In contrast, all those
infinite gradations of “phthongs” (“vowels” to the rest of the world) that we
have experienced during the preceding explorations do indeed find their way
into languages, or into dialects of languages, or into the speech habits of
individual speakers.
If you have been paying close attention to the ease or difficulty of various
speech actions in Outlandish – and if you haven’t, now is the time to review –
you may have reached a few general conclusions about the consonant sound
actions that do and don’t combine fluently with other sounds. Here are some
of mine, but you may have more.

1 After five consecutive obstruents in a cluster, a “phthong” is very

appealing, even if you are speaking Outlandish. If you speak nothing but
obstruents, you are an Obstruentian, not an Outlandian, and Obstruentians
tend to explode at an early age, a fate you may wish to avoid.

2 Ingressive sound actions are useful in speech only very occasionally, and
only if the other actions are all egressive. A few languages do use
ingressive sounds to convey specific meanings, but ingressive
vocalizations have their greatest utility in singing.

3 Most obstruent sounds are produced easily with the flow moving forward
along the midline of the tongue. While there are obstruents where the
flow moves around the sides of the tongue (“laterals”) this lateral action
initiates no further back than the velum.

4 Sounds, voiced or unvoiced, that can only be produced by a radical

increase in airflow don’t usually occur in languages, but do remain in
 “rude noises”, where radical increase in airflow energy is the whole point.
Consistency in airflow helps fluency.

5 Actions using the lips (and there are quite a few in language) only work
well in connected speech if the middle third of each lip is active and
mostly if the edges of the lips are used. That trill sound, if you can
execute it – where your lower lip is slapping against your nose – may
look terrific, but it probably won’t be used regularly in language.

6 For most people, the tip of the tongue can indeed curl back to touch the
velum area, but it is not efficient in connected speech because it’s hard to
get to the next sound action.

7 The lips do not really want to spend a lot of time behind the teeth. There is
only one sound action in the world’s languages that begins with the lips
(actually, a lip) behind the teeth; and it doesn’t stay there long.
You may well come up with other conclusions or even possibly with
evidence that counters the foregoing. But we can now explore seriously what
the physical actions are that work effectively in language.
9 The physical actions of obstruents in language

The actions
All articulate sound begins with the stream of unvoiced air or a stream of
voice – airflow broken into pressure waves – moving through the vocal tract
(the tube between the vocal folds and the lips), either in or out, inhaled or
exhaled.
Obstruents, the vocal sounds that carry the denotative meaning of words,
are always formed by a physical action of the articulators that – in some
manner – gets in the way of this flow of air or vibration, not just shaping it, as
do vowel actions, but interfering with that flow, like the “shhhh” we hear
when a rock gets in the way of a flowing whitewater river. Some obstruent
actions intrude on this flow only slightly, and some do so strongly, as with
our first category:

Stop-plosives
Also called plosives or stops. The action always begins with the physical
impulse for a flow of air or vibration, even if we haven’t yet produced a
sound. That flow is literally stopped by bringing two articulators together
completely, or by bringing an articulator fully into contact with a fixed
unmovable point of articulation in the mouth, so that no flow can get past the
point of closure. However, the muscular action that produces the flow does
not stop. (For most English obstruent sounds, this would be the action of
exhaling, with the action initiated in the torso. But not all sounds are
English.) As soon as this happens, the air either becomes more dense (the air
molecules compress closer together) or gets thinner (the air molecules rarefy
– move further apart), because of the continuing muscular action producing
the flow and the natural elasticity of the tissues of the vocal tract. So there is
either pressure at the point of stoppage, or a vacuum. Either way, the “dam”,
the point of closure, is going to burst. When it does, we have the completion
of the action of the stop-plosive, in a sudden movement (release) of the
compressed air (or voiced sound) through the just-opened space within the
vocal tract.
Nasals
We have defined the vocal tract as consisting of a curved tube from the vocal
folds in the larynx through the mouth to the lips. But of course there is
another option for the voiced or unvoiced flow: through the nasal cavity. In
most English consonant sounds, and all English vowel sounds, the nasal
cavity does not receive this flow. This is because the upper side of the velum,
or soft palate, is usually raised to create its own stoppage, preventing the flow
from getting past the nasopharynx. But on certain obstruents – those whose
physical action starts out with the initial stoppage of a stop-plosive – the
velum is lowered so that the flow impulse can escape immediately through
the nasal cavity. Since the nasal cavity itself is pretty much immobile, the
shaping of the differentiated sound has to be defined by the point of stoppage
of the oral cavity (the mouth) during the production of these sounds. Where
the stoppage takes place defines the particular acoustic quality of the nasal
obstruent. As a result, the nasal obstruents are all voiced sounds, since there
cannot be enough differentiation between unvoiced nasal obstruents to be of
linguistic use. Nasals are also the first obstruents that can be classified as
“continuants”, since they can be produced continuously from the beginning
of the breath impulse to the end.

Are nasals really obstruents?

Well, we certainly are obstructing something: the oral cavity. In fact, we
are stopping it altogether. So, yes.
But we aren’t obstructing the voiced airflow at all, because it is escaping
unimpeded through the nose, unless you have a cold. So, no. Still, the
International Phonetic Association classes Nasals as “consonants” that
require obstruction – i.e. obstruents. Bear this in mind when we start to
consider “r”s, “l”s “w”s and “j”s, all of which many phoneticians say
are not in any way obstruent.

Trills
These are all voiced in spoken language, although all of them could be
produced as unvoiced sounds. Each involves bringing an articulator to the
point of closure, as in a plosive, but simultaneously increasing the energy of
the flow, so that the articulator(s) are forced rapidly open and shut, producing
their own pattern of vibration (but one much slower than the vibratory action
of the vocal folds).

Taps or flaps
The same physical action as in a trill, except that the articulator is
immediately dropped away from the extreme intrusion into the pressure flow
after one (tap) or a few (flap) vibrations.

Fricatives
The largest category of obstruents. All fricatives are continuant sounds,
which can be produced throughout the entire flow of air or vibration. They
are also the only obstruent group that can produce differentiated,
linguistically useful, sounds – both unvoiced and voiced – in all places of
articulation in the vocal tract. As the name suggests, fricatives are sounds that
derive from a kind of friction: the narrowing of the passage of air or voice
flow to the point where there is genuine impedance to the pattern of the flow.
So the articulator is brought very close to another articulator or to a point of
articulation, without actually closing the space off. One can register the focus
of a fricative, unvoiced or voiced, very clearly within the vocal tract

Approximants
These are formed very much like fricatives, but with the articulator intruding
only very slightly into the flow of air or vibration. If a fricative is the rock
jutting out of the whitewater torrent with the water whooshing around it, the
approximant is your toe dipped into the same stream. You can feel it, but it
doesn’t get in the way of the flow very much. In most cases in connected
speech they are voiced because the obstruction is so slight that the unvoiced
version does not produce a differentiated sound. But there is still an
obstruction: you can prove it to yourself when you enact approximants
because – like the toe in the stream – you can just feel the vibratory reaction
in the articulator, usually the lips or the tongue, at the point of placement. The
tingle of vibration is so slight that most linguists do not class approximants as
obstruents, preferring terms like “sonorants” or “semi-vowels.” I maintain
that even a tiny obstruction is an obstruction.

Laterals
Fricative and approximant. All other sounds can be classified as “central” –
that is, the flow proceeds directly down the middle of the vocal tract. But
laterals (the “l” sounds) are partially formed by bringing the tongue up to the
roof of the mouth, requiring the flow to divide laterally and go around the
sides of the tongue. If this action impedes the flow considerably, it is a
fricative; if it intrudes only slightly, it is an approximant.

Affricates
There is only one unvoiced/voiced pair of affricate sounds in American
English. In other languages there are more sets. Actually affricates introduce
no new actions because each is composed of a stop-plosive in which the
articulator leaves the point of articulation slightly more slowly than usual, so
that the explosion of air or voice registers briefly as a fricative sound.

All the sounds produced so far are pulmonic. This means that the flow is
moving out of the body through the vocal tract in a muscular torso action that
is causing the lungs to exhale. All British and American English speech
sounds are pulmonic sounds. However, this is not our only option, and many
languages incorporate sounds that are wholly or partially non-pulmonic. The
only constant physical requirement is that there is a flow in some direction
through some part of the vocal tract that can be shaped physically into
differentiated sound.

Clicks
These are all unvoiced. They all start with the jaw raised so that the mouth is
closed, and the general action of a click is to bring the lips or tongue-tip or
tongue-blade or tongue dorsum (body) into complete closure (a “stop”) and
then to open the jaw quickly with the back of the tongue and the velum (soft
palate) or uvula still completely sealed to the roof of the mouth (velaric
closure), so that air is drawn suddenly into the oral cavity. The articulation
energy is put into keeping the differentiating articulators sufficiently tense at
the point of articulation so that they lag behind a bit when the jaw is opened,
and then spring open instantly to allow the air to fill the oral cavity.

Voiced implosives
These are all voiced sounds, which naturally implies that the flow is coming
from deeper in the body than the vocal folds in order to produce that voicing.
And so it is. But it is still not pulmonic, because the flow is being pulled by a
vacuum created in the oral cavity, not pushed by muscular action from the
torso. The action of voiced implosives is much the same as for the clicks:
there is a complete closure, or “stop,” at some point in the oral cavity, from
the lips to the uvula. The velum is raised to stop airflow through the nasal
cavity in either direction. However, in voiced implosives the tongue and
velum are not sealed together as in clicks, but the vocal folds come together
to produce voice, so that, as the jaw is dropped, the voiced flow is pulled up
from the lungs.

Before we add the last element, you can practice just this part of the recipe.
Bring the lips together (but it could be anywhere else in the oral cavity), raise
the velum to cut off the flow from the nasal cavity, put the vocal folds in
“voicing position”, and then drop the jaw strongly while simultaneously
pulling the larynx down, using the extrinsic laryngeal muscles. You can
check this by putting your hand lightly on your Adam’s apple. As the jaw
goes down, you should feel the larynx drop. You will produce an interesting
sound reminiscent of a baby alligator still in its egg. (And you never know
when, as an actor, you may need to use this.) There is one last step to produce
the actual voiced implosive. While you are dropping the jaw, allow the lips
(at least in this illustrative example) to drop apart: this last action causes the
voiced airflow to be pulled by the vacuum in the oral cavity inward in a
plosive action at the lips, producing an imploded rather than exploded sound
product.
Ejectives
We are all aware that the muscles around the larynx can be used to pull the
larynx up and down. We have just experienced the downward action in the
voiced implosive. In English sounds this action is irrelevant, and usually
signifies a residual tension pattern in effecting pitch change, especially when
singing. But in some languages the sharp upward action of the larynx actually
serves to push air out of the vocal tract, without the lungs or the torso
generally becoming involved at all. In order to accomplish this, the vocal
folds (and probably the false vocal folds also) must be kept tightly closed, so
that this is considered a glottal closure. All the sounds that result from this
ejective action are necessarily unvoiced sounds because the glottis cannot
move. All except one are essentially ejective plosives; one is an ejective
fricative (though more are physically possible). It is possible to have ejective
fricatives as well as ejective stop-plosives into the obstruent combinations
that form the sounds of language and also the sounds that could be found in
language. We can always add phthongs between single or clustered
obstruents to form full language actions.

Points of obstruent focus

Bilabial. An action focused at both lips, by bringing to lips to or toward
one another.
Labiodental. An action focused at the lower lip and upper teeth by
bringing the lip to, or toward, the upper teeth.
Dental/alveolar/postalveolar. The tip of the tongue at the upper teeth,
or at the gum ridge (“alveolum”) right behind the upper teeth, or just
behind the gum ridge on the very front of the palate.
Retroflex. The name means curled back, so that the tongue is curled up
and back to give a focal point in the middle or back of the palate. This is
the furthest focus back in the mouth that can be accomplished by the tip
of the tongue.
Palatal. This is accomplished by focusing the front of the tongue (which
is behind the tip and the blade) toward, or to, the palate.
Velar. The middle of the tongue is focused toward, or to, the velum (the
soft palate).
Uvular. The back of the tongue is focused toward, or to, the uvula,
 which hangs down from the back of the velum. In practice, unless a trill
is desired, the tongue focuses toward the very back of the velum.

Figure 47. Labiodentals. (Cartoon by Bethany Carlson from SpecGram, Vol. CLXI, No. 2 (March
2011), reproduced by permission.)

Pharyngeal. The root of the tongue is retracted toward the back wall of
the throat, the pharynx.
Epiglottal. The tongue root is retracted, almost as though one is
beginning to swallow. This brings the epiglottis down in back, where
subglottal air pressure can produce either stop-plosive or fricative
sounds.
Glottal. The vocal folds are brought toward one another, or together, in
a speech action differentiated from their rapid vibration in voiced sound.

Now we are ready to explore our recipe, to combine these elements:

1 Unvoiced or voiced –

2 The placement, from bilabial to glottal, and finally –

3 The action from most obstructive (stop-plosives) to the least obstructive

(approximants)

There is one final element that can enter our recipe on occasion. Most of
the time when we speak, the flow of unvoiced or voiced air travels forward
along the midline of the tongue, sometimes narrowly, sometimes broadly. On
some sounds, though, we raise the tip upward, or bunch the midline of the
dorsum of the tongue upward, so that the flow is divided into two streams
that flow around the sides of the tongue. These are the lateral actions, and
when they occur we need to note that fact in our recipe also.
10 The empty obstruent chart

The way that the International Phonetic Association has chosen to represent
obstruent (aka consonant) IPA symbols is by placing them on a set of graphs.
The so-called “pulmonic” symbols are displayed on a standard graph with an
X-axis and a Y-axis. Each row represents a different physical action used in
obstructing the flow; each column represents a different placement or focal
point for that obstruction within the articulators.
The rows of different actions start at the top with the most obstructive
action (stop-plosives) and end at the bottom row with the least obstructive
(the lateral approximants). The columns begin at the left (as though in a left
profile of the mouth) with the most forward articulators – the lips – and end
on the right with the articulators at the base of the vocal tract, the glottis – the
space between the vocal folds.
Each cell in the graph, therefore, represents the possible site of one
physical action combined with one placement of that action within the vocal
tract. In the actual IPA chart, symbols are placed in many – but not all – of
these squares. Some squares are empty because the action/placement
combination they would represent is considered by the IPA to be physically
impossible to produce; some squares could easily be filled with a symbol, but
no symbol has been placed there simply because the IPA has not noted its use
in any existing language – yet.
Some cells in the IPA pulmonic obstruent chart have two symbols in
them, one toward the left side of the rectangle and one toward the right side.
The symbol on the left is unvoiced; the symbol of the right is voiced. For
now, assume that you might be able to produce both a unvoiced and voiced
sound for each cell. This won’t always be true, but give it a try.
We don’t need the actual symbols yet. All we need to know are these
ground rules for the graphs. If you look at the pulmonic chart, you will note
that it contains no symbols at all in the squares. However, the information
along each axis of the graph will give us all the information we need to fill in
the sound actions. We can discover for ourselves which sound actions are
possible to enact and which ones can also be easily used in language, even if
we don’t yet know the language into which they might fit.
Exploring the empty obstruent chart
Starting with the pulmonic chart, and starting with the stop-plosive actions,
explore out loud the unvoiced and voiced sound possibilities in each
placement. Continue this exploration with all the remaining action/placement
combinations in the pulmonic chart. Which action/placement combinations
can be performed easily and also provide a clear differentiation of sound for
the listener? If the sound meets those two criteria, it probably (though not
certainly) has a place in some existing language.

Explore each pulmonic obstruent as though it were the initial obstruent in a

one-syllable word, followed by the phthong “aaah”,

There are still going to be a few sound actions missing from the pulmonic
chart, including an unvoiced/voiced pair that is common in English. For such
sounds there is a category in the International Phonetic Alphabet entitled
“Other Symbols”. Most of them involve what we would consider co-
articulations, sounds that are produced by using more than one placement of
articulators, simultaneously. In the case of the pair of sounds from English,
the co-articulation involves the lips. Notice that all the “bilabial” actions in
the pulmonic chart are performed by “pursing” the lips – that is, bringing the
margins of the lips evenly together or close to one another. The common
English sound pair we are looking for here uses that pursing action, but
combines it with lip-corner advancement (or protrusion) to produce a full lip
rounding. This information should be sufficient to allow you to find the
sound pair. We can leave the other sounds until we need to learn all the
symbols themselves.
 Figure 48. The empty obstruent chart. (Please see www.langsci.ucl.ac.uk/ipa/ for original charts.)

The non-pulmonic chart

The graph for the “non-pulmonic” obstruents is simpler as represented in the
empty obstruent chart of page 95. There is only one row across the top
indicating the three actions for initiating and obstructing the flow, and then
the three columns tell us where these actions are placed within the vocal tract
in actual language. There are probably more placements possible, so you can
explore vocally beyond the ones shown in the graph. Make sure that you
carefully review the actions as described in the preceding section, since these
are unfamiliar to most American speakers.

Practice speaking the following recipes for obstruents. Follow them with any
phthong you wish. Not every recipe is here but most of them are. Some of
these sounds are not found in language, but they are physically possible to
enact.

voiced bilabial stop-plosive voiced post-alveolar lateral

approximant
unvoiced velar fricative voiced post-alveolar fricative
voiced retroflex nasal voiced post-alveolar tap
unvoiced labio-dental fricative unvoiced bi-labial fricative
voiced post-alveolar approximant voiced velar stop-plosive
unvoiced velar stop-plosive voiced bi-labial implosive
unvoiced uvular fricative voiced velar approximant
voiced glottal stop-plosive unvoiced alveolar fricative
voiced post-alveolar trill voiced labial-velar approximant
unvoiced bilabial trill voiced bilabial approximant
unvoiced alveolar stop-plosive unvoiced lateral palatal fricative
unvoiced lateral velar fricative unvoiced labiodental
approximant
unvoiced velar ejective voiced velar trill (The IPA says
you can’t!)
voiced dental fricative voiced bilabial fricative
voiced labio-dental approximant voiced palatal stop-plosive
unvoiced palatal fricative voiced epiglottal stop-plosive
unvoiced retroflex stop-plosive unvoiced labio-dental fricative
voiced uvular trill unvoiced pharyngeal fricative
unvoiced labial-velar fricative voiced labial-palatal fricative
unvoiced post-alveolar fricative unvoiced bilabial ejective
voiced pharyngeal fricative unvoiced postalveolar click
voiced labio-dental flap voiced alveolar stop-plosive
unvoiced alveolar fricative voiced retroflex lateral
approximant
voiced palatal nasal unvoiced dental fricative
voiced retroflex approximant voiced uvular implosive
11 Making your own language

Omnish and beyond

You now have a full repertoire of the distinct vocal sounds that are used in
language. They are not the only vocal sounds that we can make: as we know,
even the IPA chart omits some very easy-to-perform sounds. For example,
the Voiced Bilabial Trill [ B ] made it onto the chart in the 1989 revision
because linguists observed its use in two languages of New Guinea. But we
have been using it as a communicative vocalization to indicate cold
(“brrrrrr”) for possibly thousands of years in many languages. The Unvoiced
Bilabial Trill that we often use as a lip release exercise (or to imitate a horse)
is not (yet) heard as a phonemic component of language, and so is absent
from the IPA chart of obstruents. You have proved that it is possible in the
preceding exercises. So we can include the unvoiced version in the work that
follows.
We can now begin to explore the feel of these sound-actions when used in
combination. This is play, pure and simple, and like all the best forms of play
it will inform as well as entertain. It will tell you a good deal about where in
the vocal tract you can open your articulators to more active and demanding
use. It will tell you what sound combinations are easier or more difficult to
enact. No real language contains the entire IPA, so we know that what we are
attempting here – using all the possible sound actions – is an exercise, not a
replication of any particular language.

1 “Gibberish”. This is spoken language – here American English – without

sense; that is, without meaning attached to any of the “words”. Say a
“nonsense word” using the obstruent and phthong sounds we use in
English. “shleeft, rawv, aglackle, mostawhech, toofler, omzayidge,
bopongah” are just a few possibilities. Practice single “gibberish words”
until you are sure that you are using only the sound actions we use in
English. Now continue this action into connected speech, improvising a
conversation with a partner using American English sounds, but not in the
usual American English sound combinations. Avoid speaking any actual
English words. Keep the jaw relaxed open as much as possible and keep
the other articulators active. Feel the sounds and be aware of expressing
 emotional content through the sound combinations. But do not “translate”
actual sentences into your “Gibberish”. Let impulse take over: observe
your partner when she or he is speaking; listen to your partner attentively,
to the sounds your partner is making and to the way your partner is saying
them. Then let impulse take over and let your own sound combinations
respond. You could do this same excercise for any language, but the
sound combinations would be different.
Audio track 23: Gibberish.

2 Individual Omnish. Relax the jaw open. Begin to speak consonant and
vowel combinations using all the language sounds possible in the IPA
categories of action and placement, including non-pulmonic consonants,
whether or not they actually exist in the languages of the world. You will
discover that this requires much greater dexterity of articulation than in
the previous exercise. After a bit of practice you will discover that you
have moved from “Gibberish” to what we will henceforth call “Omnish”,
from the Latin word omnis meaning “all”. Make sure that you observe the
restrictions for articulation actions that we find in language. Don’t go
back to Outlandish.
Audio track 24: Omnish.

3 Paired Omnish I. With a partner listening closely to you, speak

energetically in Omnish for one minute. At the moment you are not trying
to convey specific meaning as you speak, you are just trying to use as
many of the sounds of world language as you can in a coherent sequence.
Your partner will tell you if you are achieving enough variety – or too
much, for that matter. For example, you may repeat several sound actions
again and again but fail to move on to other ones. Or you may leave out
whole categories of sound actions. Or you may achieve a great variety in
consonants but very little in vowels. Or you may find that you are doing a
lot of consonant sound-actions that are not found in language at all. Speak
Omnish again – just for the sounds – and see if you can correct these
limitations so that you are speaking all the actions and all the placements
of world language within that one minute.

4 Paired Omnish II (lento). Speak Omnish very slowly. Still pay no

 attention to “meaning”, just to the sounds themselves. Feel and enjoy the
new combinations of sound actions. Register in your own mind the
actions and placements of the articulators as you are going through them.
Get feedback from your partner about variety of sound-action choice as
before. If you persist in leaving out certain actions or placements, do
some focused practice favoring those sounds, and afterward open out into
full Omnish again.

5 Paired Omnish III (presto). Speak rapidly and crisply as though you
were acting in a sophisticated comedy written in Omnish. Your partner
will tell you if your variety or fluency is suffering because of your
accelerated speech.

6 Group Omnish I. The group should form a circle. The person who begins
will speak a single Omnish “word” of no more than three syllables in a
clear, energetic voice to the next person in the circle. The speaker should
speak the “word” slowly as though s/he is trying to teach the sound to the
listener. The next person will repeat the “word” back and – if correct –
will then generate a new “word” for the next person. The person listening
will try to receive the necessary information about the speaker’s
articulator action and placement from the sound itself and from whatever
s/he can observe about the speaker’s articulator action.

7 Group Omnish II (added meaning). The person who begins should

advance toward another person somewhere in the circle and say
something (but no more than a sentence or two) to that person in Omnish.
Try to convey a specific verbal message in Omnish. Let the physical
carriage and gestures support the words. Go further: let the sounds of this
rich language shape and move the body, and let yourself feel the sounds
center where they will through the entire body. The person spoken to
should reply appropriately to the first speaker (not repeating what the first
speaker said), and then find someone else in the circle and say something
to that person. Continue until everyone has been spoken to and has
spoken. Keep the exchange moving along, and keep it energized and
varied.

Analyze the results, whether from the individual or the group exercises in
Omnish. Are the speakers starting to feel fluent using all the sounds in
language? Do some sound combinations feel easier to execute than others?
However comfortable it feels – and it will feel comfortable with a little
attentive practice – Omnish in a very real sense is an “unreal” language.
Why? How do you experience the feel of the language as you speak it? Aside
from the greater variety, how does it feel different from Gibberish, or just
speaking English, or from speaking any other language you might know?

8 An Omnish oration (adding more meaning). As a leading citizen of the

nation of Omnia (its national motto translates from the Omnish as “All
Things to All People”) you speak fluently a language (Omnish) that
contains all the sounds that might be found in the IPA. Since you are a
candidate for office in the lowest legislative chamber (of twelve) in the
Omnian legislature, devise a short (two minutes approximately) campaign
speech, spoken entirely in Omnish, that contains the following elements:
a A short opening joke.
b A scathing rebuttal to the latest charges from your opponent.
c A moving tribute to His/Her Serene Omniscience, the
King/Queen/President/Facilitator/Dictator/Chair of Omnia.
d About fifteen seconds of upbeat generalities about your legislative
program.
e A big finish (balloon drop optional).
Other members of the class should portray your audience and should feel
encouraged to respond vocally in Omnish as well.

9 The Omnish Poet Laureate. This is an exercise for two persons working
together. One person plays the role of a famous Omnian poet visiting an
English-speaking country. The other person is the poet’s translator.
The translator conducts the proceedings and begins with by introducing the
poet in English. The poet then begins reciting a poem in Omnish. After a
line or two, the poet pauses as the translator provides a translation into
English. This continues throughout the rest of the poem.
Under no circumstances should this exercise be rehearsed in advance.
 One of the many virtues of this exercise is that the translator and also the
poet will be encouraged by the form of the performance to listen to each
other closely and to pick up cues from the other person as to the progress
of the poem. These cues will relate primarily to the sounds employed, but
will also be found in intonation patterns, timing and gestural signals.

10 Serious Omnish. In exploring Omnish it is useful to perform the new

sounds as actively as possible because the feeling of the action is easier
for us to register. Because these new sounds lie outside our familiar
American English repertoire, many of them will seem odd or comic
precisely because of that unfamiliarity. Don’t let this fact deter you from
trying them out. Dive right into this new ocean of sounds and splash
about in them. Revel in their unfamiliarity. Let them seem funny – or
ugly even: we all have our biases about language sounds and we all make
judgments about them all the time. Get all this out of your system.
At the same time, if we stay too long in the ocean we may get swept away
by the tide and never get back to firm ground. If we find these new
language sounds strange, it is only because we limit ourselves to certain
sounds when we speak American English – sounds which, when we speak
them, sound equally odd or amusing to speakers of other languages.
Language sounds are not objectively funny or ugly. It is only our cultural
or personal biases that make them seem so. All of the sounds in Omnish
are used in one or another of the world’s languages to express the entire
range of human thought and emotion. They are all taken seriously by
somebody.
To move past the barrier of stereotyping language sounds, choose a serious
dramatic text in English that is deeply meaningful to you. “Translate” it
into Omnish. Don’t actually write down any sounds or memorize a
“script”; just let your impulse say in Omnish what the words are in
English. Perform it for your class and try to communicate the text in
Omnish with as great a degree of sensitivity or profundity as you find in
the original.

The skills of Omnish

Just as we practice and gradually perfect the skills of speaking any language,
so we can practice and perfect the skills of Omnish, a task made considerably
easier by the fact that we don’t have to deal with precise vocabulary or
grammar. There is a real benefit in becoming a fluent Omnian, however.
Omnish, if done correctly, is possibly the best articulation warm-up that we
can try. It provides greater muscular activity than any repetition of “tongue-
twisters” and it does so in a way that works to free you from the habitual
patterns of articulation formed by your own native accent without negating
that accent. Tongue-twisters, however enjoyable they are to say, focus on
only a few sound combinations of English or another language (most
languages have tongue-twisters). They don’t really invite you to embody new
possibilities. Furthermore, Omnish is more physically demanding than any
individual language; it strengthens articulators and frees them
simultaneously.
In order to give yourself this effective warm-up, you have to make sure
that you are really including over (say) a minute of Omnish speaking a
complete repertoire of pulmonic and non-pulmonic obstruents as well as all
the possible gradations of phthong-shaping. This takes practice. Here are a
few ways to do it:

I As you are speaking Omnish, focus on a few actions or points of

placement and make sure that you are including them in the mix. You can
actually refer to your Empty Consonant Charts (pulmonic and non-
pulmonic) as you are speaking. If you are in a class, your teacher may call
out actions or focal points as a coaching technique. Remind yourself to use
both unvoiced and voiced obstruents. Some actions and placements that
commonly are ignored are:
a Palatals
b Retroflex
c Trills
d Uvular
e Nasal (including nasal phthongs)
f Pharyngeal
g Approximants
h Clicks
 i Voiced implosives
j Ejectives

II If you find that you are starting to sound as though you are speaking a
particular language in “double-talk” – nonsense words to feign another
language – you are focusing too much on certain sounds. This usually
means that there is a held tension pattern in your vocal tract that needs to
be released. It can also mean that you are using a limited repertoire of
phthongs.

III Really pay attention to keeping your speech actions in the world of
Omnish. It is very tempting to slip back into the extravagance of
Outlandish. Be careful about this.

IV Just as in fine dining, where the waiter brings you a little sorbet between
courses as a “palate-cleanser” to free you from the taste of the sautéed
sweetbreads and to ready you for the roast buffalo, so you can cleanse
your palate (not to mention your velum) in Omnish by throwing in a
word or short phrase in English just the way you’d speak it every fifteen
seconds or so, as though the Omnian you are embodying is a really good
mimic of the real you. This simple technique is really useful for getting
you away from Outlandish.

Outward/inward
The progression from Gibberish to Omnish is a progression into a greater
variety of articulator action. We have been opening out our possibilities. The
next step is to reverse the process and move from more variety to less. Our
purpose is to focus in on the physicality of real languages. For the time being
we will leave Outlandish out of the picture; it is a very useful exploration for
character voices or for imitations of sounds that go beyond human speech,
but because we are going to explore the feel of accents and languages as they
exist in the real world we will focus back in starting from Omnish.

11 Finding a new language from the feel of the sounds: “Double-talk”.

Create for yourself a physical focus of tension and relaxation in the
 articulators, using basic variables, such as:
a Jaw height.
b Jaw advancement or retraction.
c Tongue-arching or cupping. Forward or back, high or low.
d Tongue root-advancement or retraction.
e Channeling (widening and flattening) or bunching (narrowing and
thickening) the tongue.
f Relaxation or tensing of the cheeks into the teeth.
g Advancement or retraction of the lip corners.
h Pursing or relaxation of the lips.
i Raising or lowering the velum.
j Bracing the side edges of the body of the tongue against the teeth.

Remember that not all these variables will necessarily be in play. Especially
at the start, it is easier to focus on only a few and hear what effect they
produce on sound change. What you will come up with, though, is something
that every language and every dialect of every language has: a characteristic
vocal tract posture. This is the pattern of muscular tension and relaxation
within the vocal tract that provides the “home base” to which the muscles
would naturally return. The posture has an effect on everything connected
with articulation: what obstruents you use, what phthongs you use, and the
resonance patterns of your speech.
Less is more. If you overdo these changes, you will always come out with
a result that sounds more like an individual with a speech pathology issue
than someone with an accent. It is useful to try to make minimal changes in
posture and then to see what effect these have on your speech. Understand
that every change, however small, will change the sound pattern if you allow
it to do so. This takes some trust in letting the physicality lead the process. It
also takes some practice.
Start with the single phthong that the posture produces. Let it release with
easy but supported vocal energy. Then shimmer the posture. This means
that you should start to produce micro-movements of the articulators
 around that phthong without losing the posture as a “home base” for the
action. Then gradually increase the activity of the shimmer to add
obstruents.
Now begin to speak connected sound combinations fully – a new sort of
gibberish. Just let the sounds come, and feel how the actions of the
articulators – especially the lip and tongue actions – tend to start self-
selecting their own preferred sounds. Feel and enjoy the sounds that your
articulators seem to want to say; gradually you may let go of the sounds
that your articulators don’t seem to want to say very frequently.
Make sure that you do not hold the articulators tightly or rigidly in these
postures. If you do you, may not be able to talk at all! The sound actions
your articulators take can move away from the chosen posture; they should
simply favor moving back toward them and favor sounds that are nearer to
these articulator postures. You should still be able to speak as fluently and
easily as you did in Omnish – more easily, actually, because your
articulators will favor some sound actions over others.
You have now formed a new (non-sense) language physically in a way that
is much more real than Omnish or Outlandish, and one that probably has a
very definite and characteristic feel or focus to it. It may sound to you like
some “double-talk” version of an existing language that you have heard, or
it may sound completely new.
Finally, keeping the same focus, the same interplay of muscular tension
and relaxation that you just used for your newly created language, begin to
speak in English. If you are clear about maintaining the focus, you will
definitely produce some sort of “foreign accent”. Maybe you will
recognize it; maybe you won’t. It may even turn out to sound more like an
accent pattern from a dialect of English. But you are still experiencing the
dynamic tug-of-war between the original language focus (and sounds) and
the English focus (and sounds) which is characteristic of all foreign accent
speaking.
Audio track 25: Create an oral posture language.
Audio track 26: Create an oral posture accent.

12 Finding an accent. Create a new vocal tract posture, quite arbitrarily, and
 see what happens when you start to speak through it in English. Is it a
recognizable – or nearly recognizable – accent of English from wherever
English is spoken as a primary language? Or is it another foreign accent?
What physical changes do you need to make to fine-tune it to a
recognizable accent in English or a foreign accent?

13 Finding the “posture” of an accent. Do an imitation of an accent, in

English or from another language, which you can mimic but which is not
your own. Speak in that accent for at least a minute. As you listen to
yourself, pick out the most important or unique sound changes, the
sounds that are strongly different from the way you usually articulate
vowels and consonants in your own everyday speech. Now repeat each of
these unique sounds several times. As you do so, feel what each one of
them tells you about the comfortable posture of that action in your mouth.
What is the larger pattern of muscular tension and relaxation in the
articulators that requires the sound to be produced in that way?
And now the big question: what do these individual patterns of tension and
relaxation have in common for each unique sound action? They might, for
example, all work better if the jaw is raised quite high and if the tongue is
tensed with the front section strongly cupped. Or another combination
might be with the jaw very slightly raised, the tongue tip retroflex, and the
buccinator muscles of the cheek slightly braced. Continue to speak in the
accent, but stay with the feel of the focus, rather than thinking constantly
about the succession of individual sound changes.
Individual sound awareness can remain a very useful corrective when the
focus of an accent starts to go astray, but allowing yourself to feel the
overall posture of the accent allows you to focus on the many other
important elements of the acting process. Finding the posture of the accent
helps you to act through the accent rather than getting lost in its
mechanics. A postural approach allows the accent to become a physical
embodiment of your character’s speech – just as your entire bodily posture
is an embodiment of your character’s movement onstage. So you can focus
on behaving as the character, not on enacting the accent.
A world of possibilities opens to us.
But alas, this is not a book that will focus on accent acquisition. So, while
it is a wonderful enterprise to use the physical skills that you have acquired to
explore many different accents and different language pronunciations based
on the characteristic posture of each accent, we will focus in to a narrower
area for the remainder of this text. Our language will be English, and more
specifically American English. We will apply all these skills to the variable
possibilities of speaking American English to others.
We have actually acquired a formidable array of physical skills in
articulation. Now the question arises: how do we communicate those
gradations of sound-change to others in an efficient manner, and how do we
record them for ourselves?
An interlude
A daily articulation warm-up
1 Stand or sit so that your spine is aligned and your neck is released into
length. Your head should be balanced at the top of the spine at your
occipital joint. Relax your jaw open into a nice, dumb expression. Let all
your facial muscles relax.

2 Massage your muscles of expression, starting on your scalp, working

downward with circular fingertip massage. Massage the orbicularis oculi
muscles with the palms of your hands. Then continue with the tips of
your fingers around your nose, around your ears, into your cheeks and
lips, ending at the masseter muscles of the jaw. Massage the masseter
muscles of the jaw with moderate vigor, reminding yourself to let the jaw
release open with a forward-and-down “energy flow”.

3 Spread the fingers of both hands and gently stroke directly downward on
the face with your fingertips from scalp to jaw bone. Repeat at least five
times.

4 Repeat the “Great all-purpose jaw release exercise” at least five times.

5 With the jaw relaxed open – its “home base” for speaking – focus your
attention on your natural, easy tidal breathing. Feel the even sensation of
the air moving ingressively and egressively through the vocal tract. It you
feel the air flow focusing at any point in the vocal tract, focus your
attention on letting that area release open.

6 Send a slightly greater volume of egressive and ingressive unvoiced air

flow through the vocal tract and use the articulators to engage in
“unvoiced phthong shaping”. Note if you start to feel any obstruction of
the flow at all, and adjust the articulator action to avoid it.

7 Send egressive voiced flow through the vocal tract. Without obstructing
the flow, explore moving everything that will shape the sound. Do not
 change the “voice quality.” Do not change the amplitude deliberately.
Focus on “phthong shaping”. Lead with the physical action: appreciate
the resulting changes in sound product but do not look for any specific
phthongs. Start with a slow-motion use of the articulators and then move
the changes faster; finally, slow down again.

8 Start to obstruct the voiced egressive flow. Explore all the possibilities for
obstruction, from stop-plosives to approximants and everything in
between, while you continue all the phthong possibilities that you had in
6. You are now doing Omnish.

9 Speak for at least one minute in Omnish in the slow-motion version.

10 Speak for at least thirty seconds in fast Omnish.

11 Speak for at least one minute in Omnish at a normal conversational rate.

Relax.
Part 3
Phonetics
Figure 49. The International Phonetic Alphabet. (Please see www.langsci.ucl.ac.uk/ipa/ for original
charts.)
Introductory

Phonetics and phonemics: the obstruent version

The International Phonetic Association has assigned one written symbol to
represent each differentiated sound that is used in language. One symbol =
one sound. This makes phonetics completely different from the way we use
symbols when we spell a word, even though many phonetic symbols are the
same ones that we are familiar with in spelling.
If we allow that voicing or unvoicing is a physical action, we could go
further and say that each phonetic symbol equals a single, unique physical
action.
But the International Phonetic Association, in determining what actions
get to be represented as symbols of the IPA, has established another parallel
category: each IPA symbol must represent a single, unique unit of meaning
within the larger combinations that comprise any given language. The study
of phonetics is the study of articulator actions as they are used in languages
and dialects of language in all their detail and subtlety of differentiation. This
attention to detail causes it to differ from the broader study of phonemics or
phonology in that scholars who focus on these latter terms are interested in
articulation actions at the level of meaning. A good example of the difference
as it relates to obstruents would be the “k” sound we produce at the start of
the word “keel”. We know that if we change the action to produce a “p” the
actual meaning of those clustered sounds that we call a word would change,
and would become “peel”. Or we could change the action again and have
“feel” or “seal” or “reel”. Or we could let the obstruent sounds remain the
same and change the phthong between them, and the meaning would change:
“keel”, “kill”, “call”, “cool”. (Remember that those “k” and “c” spellings
represent the same sound.) So in each case we would be making a phonemic
change, and each one of those individual sounds that combine to form a word
would be called a phoneme. When we change the symbol within word units,
we change the meaning. So the symbols exist as representations of phonemes,
not as their more dialectically detailed subdivisions, which are termed
allophones.
But if we feel the placement of the “k” sound as we go through that
progress of “keel”, “kill”, “call”, “cool”, we can register – by where the
dorsum of the tongue touches the roof of the mouth – that the placement
changes. The “k” in “keel” is certainly touching the back of the palate. The
“k” in “cool” clearly is well back on the velum, almost at the uvula. Such
allophone changes, though small, will be of great use to us in developing a
regional or foreign accent, and they are all phonetically different from one
another, which is why we study detailed phonetics in this text. But
phonemically those allophone sounds are the same because they all convey
that “k” contribution to the meaning of the word.
Sometimes we notate these subtle phonetic differences by using a
different phonetic symbol, and sometimes we use smaller symbols above or
below the line (called diacritics) to show the physical change of articulation
in the mouth, nose, or throat. Sometimes, also (and mostly with vowels) we
devise a map of the phonetic “territory” of a phoneme, so that the viewer (and
the listener) can perceive visually where the allophones of that phoneme
might live.
These two terms, phonetic and phonemic, are not always mutually
exclusive in their application, but they obey different rules. Two similar but
slightly different sounds that might be phonemically distinct in one language
– actually changing the meaning of the word – might have a phonetic (but not
phonemic) significance in another: despite being different physical actions
producing slightly different sounds, they “mean” the same thing within the
language and often are not perceived as being different at all by the listener
and even by the speaker. We tend not to perceive differences if they are not
important to us. Conversely, sounds that – in English, let’s say – we would
perceive as exactly the same, in another language may have a phonemic
difference. In some Indian subcontinent languages, the difference between a
voiced retroflex stop-plosive and a voiced alveolar stop-plosive is phonemic
and changes the meaning of a word; in English, it’s just a “d”. The listener
might think that the speaker of the retroflex version “has an accent”, but we
wouldn’t question the meaning.
In practise, the fact that there are two parallel descriptions of what a
phonetic/phonemic symbol is – a unique physical action or a meaning unit –
doesn’t get in our way too often. That’s the good news …

Cartoon theories of linguistics

 Figure 50. Phonemics.

Figure 51. Phonetics.

(Cartoons by Trey Jones from SpecGram, Vol. CLIII, No. 4 (September 2007), reproduced by
permission.)
12 Writing it down

We now possess awareness and detailed experience of the physical actions

and the physical placements of obstruents that produce sounds that we can
hear as different from one another. We have also, in a more free-form way,
explored the unobstructed shaping of voiced sound flow that we have been
calling “phthongs”. In the practice of Omnish, we have not limited these
configurations of obstruents and phthongs to the sounds that are actually used
in the languages of the world, but explored ways to limit them to the sounds
that could be found in language.
But we know that Omnish is not where speech sounds live for speakers in
the real world. Eloquent as Omnish is in its way, it doesn’t actually mean
anything that the speaker and the listener can agree on. In order to
communicate commonly understood meaning, we have to combine the
obstruents and phthongs into syllables and then into words. These words
convey conventionally understood meanings within a given language – sound
combinations that are filled with meaning. In moving into the articulation of
actual language, we will find it very useful to be able to describe these more
complex actions to others, and we will want to be able to write these
descriptions down so that we can keep a record of these actions for our own
use and also share our analysis with others.
We already have a recipe for the production of pulmonic and non-
pulmonic obstruents, and a fair sense of what the articulators, especially the
tongue and lips, do to shape phthongs; but trying to write down our
perceptions using our current vocabulary would be extremely difficult and
tiring. If we had to write “voiced bilabial stop-plosive plus strongly cupped
(or open) unrounded front phthong plus unvoiced alveolar stop-plosive” in
order to signify to the reader that we have written the English word “bat”, the
effort would discourage us from trying to communicate anything.
So it is obvious that we will need to reduce these lengthy descriptions of
sound actions to single symbols that will convey the same information in a
much more economical way. Our first conclusion might be that we already
have just such a set of symbols: the Roman alphabet that we use every day in
writing English.
But writing “bat” as “bat” has its problems, too; big ones. Assuming most
readers of this book will be speakers of American English, we probably have
a general agreement as to how those three letters (i.e., symbols) – “b”, “a”,
and “t” – should be pronounced, at least in this word. But if we want to
communicate the actual sounds that we are using to a person who does not
know spoken English well, we could be giving that reader some non-precise
information. The letter “a”, for example, could be pronounced as a “short a”
as in “ham”, or as a “broad a” as in “calm”, or as a “short e” as in “any”, or
even as a “short i” as in “average”,
The classic illustration of the imprecision of English spelling (the
technical term for “spelling” is orthography) was suggested by the famed
Irish playwright George Bernard Shaw, who was an early advocate of
simplified spelling. He pointed out that the word “fish” could as easily be
spelled “ghoti”, using the sounds found in the words “cough”, “women”, and
“nation”. English orthography is full of letters that have multiple sounds
attached to them, as well as individual sounds that can be pronounced in
multiple ways. If we really want to communicate what we are doing when we
speak – let alone what we are doing when we speak in an accent – we will
have to come up with some system on notation that is more useful: a system
of symbols that assigns a single symbol to every physical action of
articulation that produces a sound that can be used in language. This is where
the International Phonetic Alphabet comes in. And on page 114 we see what
it looks like in its latest revision.
This is the complete chart of the International Phonetic Alphabet,
hereafter the IPA. Your first reaction upon gazing at that chart might well be
that it is incredibly complicated. But take heart: it really isn’t. For one thing,
if you’re still practicing your Omnish, you’re already doing nearly everything
on the chart and you know the action recipe for each one of the sounds
represented by all those symbols. For another thing, there aren’t really that
many symbols to learn. Take a close look at these charts: the entire Roman
alphabet is contained within them, and it forms a large percentage of the total
number of IPA symbols. I might even go so far as to claim that the most
important skill you can bring to learning the IPA will not be learning new and
unfamiliar symbols, but instead ignoring those ambiguities that are found in
English spelling – because in the IPA each symbol stands for only one unique
articulator action and placement. This is worth repeating, prominently:

Each IPA symbol stands for only ONE

 unique articulator action and placement
So, if we can ignore the different sounds for the spelled vowel “a” or the
consonant “r”, to take just two examples, we can easily use the Roman
alphabet symbols that are already contained in the IPA charts.

Sounds to syllables
We all understand – intuitively and through our experience with language –
what syllables are. This knowledge is valuable and will save us a lot of time
and effort because the subject is still a source of debate among linguists and
voice scientists. So any attempt at a technical definition of a syllable will
probably be deficient in some way, certainly deficient to some expert in the
field. Far better that we rely on our intuitive understanding and leave it at
that.
We do know that a syllable can be as short as a single vowel, as in the “a”
in “about”. A syllable can form one monosyllabic word, as in “put” (with a
very few sounds) or “clasped” (with more sounds). A single syllable can also
be one part of a very long word, as in the Mary Poppins invention
“supercalifragilisticexpialidocious”. If we count the syllables is that word, we
would all agree that there are fourteen of them, even if we haven’t defined
what a syllable is.
13 Writing the consonants

Now that we are focusing on writing down the symbols, we can let go of our
word “obstruents” and go back to the more familiar term “consonants” to
describe voiced or unvoiced sound actions that obstruct the flow. We start
with consonants, rather than with “phthongs” (“vowels”) because consonant
actions physically are very clearly defined and differentiated from one
another. Vowels (or phthongs) are more elusive – and therefore even more
interesting. We will get to them soon.

What do we already know?

Take a good close look at the pulmonic consonant chart on the next page,
now filled, or nearly filled, with phonetic symbols. Your first reaction may
well be dismay. Our eyes are naturally drawn to the unfamiliar; and a lot of
these symbols, at first glance, appear very unfamiliar indeed. But if you focus
your attention on finding symbols that you already know, you will make a
comforting discovery. So take another look.

Figure 52. IPA pulmonic consonant chart.. (Please see www.langsci.ucl.ac.uk/ipa/ for original charts.)

Good news. In bold face below, you will find that twenty-four of these IPA
symbols are ones with which you are very familiar as either lower or upper
case letters of the Roman alphabet that we use in printed English.
 Figure 53. Bolded Roman alphabet symbols in the pulmonic IPA chart.

I can’t claim that every one of them is used in exactly the way we do when
we spell English words, but most of them are. So, not only do you know,
from the work you have already done, how to produce all the actions
represented by all the symbols and then some, you also know over a third of
the symbols already
More good news. We actually know even more. If you look further, you
will find that almost all of the remaining symbols represent some variation on
Roman alphabet symbols. Sometimes the symbols have little hooks or other
embellishments attached to regular Roman letters. Sometimes the Roman
letters are turned or flipped around. But we can still see what they are; and in
every case we can get some very valuable clues about what the sounds are
that these symbols represent.
In bold face, here are all the symbols that are Roman alphabet letters or
are derived from Roman alphabet letters. Now your real search through it
will be trying to find the unfamiliar symbols.

Figure 54. All Roman alphabet-related symbols bolded.

Getting specific about the symbols
We will start with something very familiar, in the upper left-hand box, or
cell, of the chart, the bilabial stop-plosives, as we can tell from the
descriptions on the X (horizontal) and Y (vertical) axes. (Today, in a graph
like this we would describe the “X-Axis” as a line, and the “Y-axis” as a
column.) What general information about writing phonetic symbols can we
get from just these two symbols? Don’t be afraid to give very simple, obvious
answers and don’t look at the next page until you have some.

Some answers
1 The symbols are printed, not written in cursive script. There have been
systems for phonetic transcription that used unconnected or connected
cursive script, but they have not been used by phoneticians for nearly a
hundred years.

2 The symbols are written as lower-case, not upper-case capital letters.

There are no “capital letters” in phonetic symbols because we are only
concerned with representing sound actions; we don’t need to bother about
marking proper names, or the the start of a sentence, or – as in German –
nouns.

3 The symbol in the left half of the cell is unvoiced. The symbol in the right
half is voiced, even though in other respects they are pronounced the
same.

4 The symbol on the left / p / as printed has a vertical line that goes below
the line on which the symbol “sits”. The symbol is therefore called a
descender. Look at the other symbols in the pulmonic chart to see which
ones are also descenders. The part of the symbol that descends does not
need to be a vertical line, of course; it can be curved or slanted as well.

5 The symbol on the right has a vertical line that goes above the general
height of most lower-case letters. The symbol is called – unsurprisingly –
an ascender. Try to find other ascenders in the pulmonic chart. Then try
to find some symbols that are both ascenders and descenders. And finally,
 find some symbols that are neither, which is most of them.

As we start to print phonetic symbols, we shall use the old line system that
we used in childhood (well, I did) to learn to print the Roman alphabet that
we use in English, just to make certain that the symbols have the right
relationship to each other in a line of transcription. We have first the base line
on which the letter sits; above that is a dotted line to indicate the common
height of many lower-case letters. Once we learn the proper proportion of the
symbols, the dotted line can disappear, as it does by the second grade of
grammar school.

Next pick some other ascenders from anywhere in the pulmonic chart and
write at least four of them on the line below.

Write four descender symbols.

Write four symbols that fit between the baseline and the dotted line above,
that are neither ascenders nor descenders. These will include symbols that
look like capital letters, but they are all lower-case capitals.

Next, find four symbols that are both ascenders and descenders and write
them appropriately above and below the lines.

And finally, write a sequence of at least ten consonant symbols with each
symbol different as to ascender/descender characteristics from the one that
precedes or follows it.
Now we know the way in which the printed lower-case symbols are written,
their size and placement relative to one another. We already know well the
recipe that tells us what the physical action is that is represented by the
sound. So we can proceed through the rest of the symbols in the pulmonic
consonant chart easily.
To help us get into the idea that we are writing sound actions, not
“spelling” words, here are a few examples of consonant symbol use. These
all derive from the governing principle that one symbol equals one unique
sound action, and only one:

The voiced velar stop-plosive represented by the symbol / ɡ / looks like one
form of the letter “g”. When we spell words in English with the letter “g” we
could either mean a “hard g” as in “gag” or a “soft g” as in “ginger”. But we
know from our recipe for sound action that this symbol / ɡ / represents only
the “hard g” form. The “soft g” will have to be represented in another way,
with other symbols. This is why we need to add consonant symbols to the
existing Roman alphabet for phonetic transcription.
We recognize the symbol / c / as a familiar looking character in English
spelling that can represent two different sounds in English, just as we found
with / ɡ /. However the unvoiced velar stop-plosive, what people often call a
“hard c”, is already represented by the symbol / k /. Similarly, the unvoiced
alveolar fricative, often called the “soft c”, is represented by the symbol / s /.
So the / c / has lost its usual jobs. Fortunately it is utilized in the IPA to
represent a sound action not usually found in English, the unvoiced palatal
stop-plosive.
Perform this sound action. It sounds a lot like a / k /, but there is a slight
difference in the sound produced. In addition, there is a very slight breathy
offglide after the plosive action that sounds a little like a “y” (as in “yes”),
only unvoiced. Why does this offglide tend to happen with / c / but not with /
k /? There is a good reason, and it relates to the anatomy of the roof of the
mouth. And I’m not going to give you the answer right away. Write your
reason opposite or on a separate sheet of paper.
Assuming you have either figured out a plausible reason for that offglide or
else simply given up, we may proceed onward. Take another look at the
entire pulmonic chart and also the non-pulmonic consonant chart and see if
you can notice other patterns in the way the symbols are written that might
give you useful information. Write your conclusions below and please don’t
turn the page until you have done so.

Some things we find

1 All the symbols in the “Retroflex” column of the pulmonic chart have a
hook depending from some point at the bottom of the symbol and curving
to the right. So we know that / η / is the Roman alphabet letter “n” with a
retroflex modifier. The action, then, is a voiced retroflex nasal, a sound
that we find in many Indian subcontinent languages. Curl your tongue tip
back and speak it. The same principle applies to all other retroflex
sounds. So speak all of them in the column, from top to bottom. The form
of the symbol gives you all the information you need. (Except notice that
the “r” is written upside down. We will deal with this very soon. But it’s
still clearly an “r”, especially if you’re standing on your head.)

2 All three symbols in the “Tap/Flap” line have a rightward-extending hook

from the top of the symbol that is still within the lower-case limits; that is,
the symbols are not ascenders. Speak them, putting them between two
phthongs.
3 All the voiced uvular symbols are printed as lower-case capital letters.
That means that they are capital letters written between those two lines
that we were using earlier. So if it’s a lower-case capital letter, you know
that you are using a voiced uvular sound. The IPA can afford to do this
because capital letters do not need to be used to indicate grammatical or
syntactic features, such as the start of a sentence in English or a noun in
German. Their lower-case status in the IPA confirms their non-
grammatical use. You know what the letter is in English, though. So
speak it as a uvular consonant, using this information.
Look at the non-pulmonic consonant chart opposite. All the “Voiced
Implosives” have a right-curving hook growing upward from the top of the
symbol, making them all ascenders; some have descenders also. So you know
that all those sounds are voiced implosives. You know them already; so speak
them, from top to bottom.

Figure 55. Non-pulmonic consonants, IPA.. (Please see www.langsci.ucl.ac.uk/ipa/ for original charts.)

Also in the non-pulmonic consonant chart, the “Ejectives” column shows

Roman alphabet symbols that are followed by an apostrophe. An example
would be / p’ /. Speak all the ejectives in the column. Can you enact other
possible ejectives that aren’t in this list?
“Clicks” have to be learned individually as symbols, but they do share the
characteristic that they are all both ascenders and descenders.
 The answer to the question on page 128
The reason the palatal [ c ] has that little fricative offglide, whereas the
velar [ k ] explodes cleanly, is that the velum, being flexible and
containing muscle, can help the plosive action by springing actively away
from the tongue. The palate, in contrast, cannot move, so it is harder for
the front of the tongue-body to get away from the palate. You can get the
front of the tongue away from the palate quickly and cleanly, but you
really have to work to do it.

The problem of “h”

The IPA defines the phonemes / h / and / ɦ / which represent the unvoiced
and voiced “h” that we find in English, as glottal fricatives, and places them
appropriately in the Glottal Pacement column in the Pulmonic Consonant
Chart. These two phonemes appear in English words always as prevocalic
consonants that always precede a vowel in a syllable. The “h” appears in
spelled words after a vowel, of course, but they are not sounded; or even
noised.
We can test this definition. Relax your jaw open and bring your vocal
folds toward one another enough to obstruct the unvoiced flow, but not into
approximation where you would produce a creaky voice or voiced phonation.
You are now doing an unvoiced glottal fricative, our IPA definition. You
have already explored this action early in this book when you explored
“heavy breathing”. Now very carefully use only this glottal fricative method
of producing an “h” in the words “haul”, “how”, “have”, “hem”, “hit”,
“heel”. If you have genuinely done an unvoiced glottal fricative, I suspect
that the “h” you are producing bears little resemblance to the way that you
actually say “h” in words and phrases.
Audio track 27: Unvoiced glottal fricative.
But if we aren’t doing an unvoiced glottal fricative here, what are we doing?
Go back to the way that you usually say a consonant “h” and feel what seems
to be happening as you move from the “h” into a vowel. You can use the
same list of six words to practice. Clearly something very different is
happening to produce the “h” sound.
In fact, when you enact the “h” in normal speech, you are preaspirating
the vowel. That is, you are first shaping your articulators to form the vowel –
whether it is the “ee” or the “oo” or any vowel in between – then you are
increasing the airflow to produce an even turbulence throughout the oral
cavity that produces an unvoiced noise. (Remember the “Making your first
sound” exploration?) Only then do you start the voicing to produce the
vowel. So in a very real sense the “h” is just a modification of the vowel that
follows it, not a single, unique consonant action. It certainly is not a glottal
fricative.
Speakers have recognized the variable action of “h” for a long time.
Ancient Greek represents this preaspiration as a diacritic, a mark modifying
the action of a phoneme. So the Greek word for the indefinite article “the” is
written as “ó”, pronounced “ho”, with the little diacritic mark above the “o”
standing for the preaspiration of the “h”. In many ways, this is a much saner
way to represent what is really going on.
The so-called voiced glottal fricative / ɦ / is even more eccentric, because
it is physically impossible for the vocal folds to produce simultaneously
phonation – voicing – and a fricative action, for which the folds cannot be
brought together. Only a little attentive exploration reveals that what we are
doing when we produce a “voiced glottal fricative” is actually a breathy
voiced realization of the vowel that follows, with the same increased airflow
that we used before.
Audio track 28: Voiced glottal fricative.

Obviously the symbols / h / and / ɦ / should not be located where they are in
the pulmonic consonant chart; and very possibly not in the pulmonic chart at
all. Or would it be better to go back to the solution in ancient Greek that
doesn’t use a letter for the “h” sound, but gets by very well indeed with just a
diacritic mark?
There are a couple of reasons why it might be useful to keep the symbols,
even though they defy the usual IPA requirement that each phonemic symbol
should represent a single, unique physical action or placement. The first
reason is decently mundane: we are used to the “h” to represent what we do
when we utter a preaspirated vowel. Using a diacritic over a vowel would
make reading phonetic symbols a little more difficult. The second reason is
that even though the placement of the action of “h” is as various as there are
phonemes and allophones of “h”, there really is a unique physical action that
ties all of them together: the increase in the velocity of the egressive
airstream that is needed to produce the preaspirated noise.
So / h / and / ɦ / remain in the IPA, having been granted clemency. They
really need a new name for their action, however, since they clearly are not
fricatives. I suggest calling them aspirants, since it is the increase in the air
flow that makes them unique. Besides, the term aspirants has a hopeful,
optimistic tone, which is not a bad attitude for learning phonetic symbols.

Everything else

Turning, inverting and reversing

One of the earliest stated goals of the International Phonetic Association in
forming the International Phonetic Alphabet was to use the Roman alphabet
as much as possible in forming phonetic symbols. (We might have hoped that
one of the first decisions would have been to avoid giving the organization’s
name and its principal product the same acronym – IPA – but alas, no.) As
we know, the IPA (the organization) used the entire Roman alphabet. It also
used the capital versions of some of them, in lower case, for voiced uvular
consonants and the voiced lateral velar approximant. But if we look again at
that column of uvular consonants, we notice that the voiced uvular fricative
/ ʁ / is not only a lower-case capital “R”, it is also an inverted version of the
voiced uvular trill / R /, meaning that it is flipped upside down without
turning. It’s upside down, but it’s still facing in the same direction.
Some letters are turned (rotated) to make them upside down, which
means that the letter is now facing backwards. Good examples are the voiced
alveolar approximant / ɹ /, which is the way many – but not all – Americans
pronounce the consonant “r”, and one of my personal favorites, the voiced
palatal lateral approximant / ʎ /, which you might think is a Greek letter
lambda. It isn’t, however. It is a “turned y”. A lambda would be a reverse
image of the “turned Y” and would look like this: “λ”. We can find a good
example of symbol reversal in the pulmonic consonant chart if we compare
the glottal stop-plosive / ʔ / with its reversed version, the voiced pharyngeal
fricative / ʕ /.
These turned, reversed, and inverted symbols aren’t many in number, and
about ten seconds of attention paid to each one will fix their correct position
in your mind.
If it’s turned, it is upside down and facing backwards.
If it’s reversed, it is right side up, but facing backwards.
If it’s inverted, it is upside down, but it’s not facing backwards.

And finally …
Some of the remaining Roman alphabet modifications can be learned easily
with little memory devices. For example, the voiced velar nasal / ɧ / looks
rather like an “ng” put together into one letter. The same voiced nasal action,
produced further forward on the roof of the mouth, specifically on the palate,
is written with the “g” hook on the “forward” or left leg of the “n”, as / ɲ /.
The easily spoken voiced palatal approximant / ɾ / becomes tighter, more
constricted in its voiced palatal fricative version, and it is represented with a
symbol, the “curly-tail j” that seems to be attacking itself / ʝ /. The “Fish-
Hook r” / r / looks like a fish-hook; but also sonically it looks just like what it
is: a voiced alveolar tap that is a part of a voiced alveolar trill / r /, so the
tap symbol looks like an incomplete trill symbol. The slightly abrasive noise
of the unvoiced velar fricative / x / is sounded as larger and more florid in
its uvular version, and it is fittingly represented by a descender symbol of
more generous proportions / χ /.
The generous proportions of the / χ / are actually the normal proportions
of this symbol’s other identity – as the Greek letter “chi”. This brings us into
the category of symbols borrowed from other orthographies, specifically
Greek and Old English (or contemporary Icelandic). There aren’t a lot of
these and they usually correspond closely in their own orthographies to the
sounds they are meant to represent in the IPA (the alphabet).
Most of the Greek and Old English symbols are found in the fricatives
row on the pulmonic consonant chart. We find them paired in cells where
Roman letters to represent these sound actions just don’t exist, such as the
unvoiced and voiced bilabial fricatives / ɸ / and / β /. We can see them
paired in cells where the Roman letters to represent the sounds do exist but
are imprecise or inefficient, such as the unvoiced and voiced dental
fricatives / θ / and / /. The / χ / reclines in solitary Greek grandeur, paired
with an inverted Roman partner.

The source for the nomenclature of all these phonetic symbols is The
Phonetic Symbol Guide, second edition by Geoffrey Pullum and William
Ladusaw (University of Chicago Press). The book is a surprisingly
entertaining read and is the definitive reference for all phonetic symbols,
living and dead, past and present, major and minor. If you want to be able
to bandy about terms like yogh, esh, ash, ram’s horn, slashed o, and fish-
hook r, all of which can enliven any social conversation, this is the book
to consult. The histories of the symbols are diverting too: when things get
confusing, the authors frankly admit to the confusion, a refreshing
approach, e.g. “Confusingly, the symbol we call Script V looks somewhat
more like a Greek upsilon than IPA Upsilon does; and the version of
Upsilon used in Smiley 1963 looks much more like Script V than like
Upsilon. We regret not being able to terminologize better in this area, but
we cannot see a better alternative. (Some phoneticians refer to Upsilon by
the name Bucket, but it looks more like an urn to us)” (page 185).

There are very few other symbols left. They have been adapted or
invented by various phoneticians throughout the years, submitted to the IPA
(the organization) and approved. Their origins are sometimes complex,
sometimes murky, and sometimes just wacky. In the unvoiced and voiced
post-alveolar fricatives / ʃ / and / ʒ / we can see some remnants of an “s” in
the “esh” symbol / ʃ /,1 and a bit of “z” in the “yogh” symbol [ ʒ ], which
occurs in Middle English – but as a different sound. Approximant symbols
like the “script (or cursive) v” / υ / (and you thought it was a Greek upsilon or
at least some form of “u”, didn’t you?) or the “long-legged turned m” [ ɰ ]
(which sounds rather like the name for a water bird or a stick insect) are far
less clear in their history and derivation. The “long-legged turned m”, the
symbol for the voiced velar approximant, may not even be found as a
unique uncombined sound in language. But there it is in print; and approved
by the IPA too. And you can speak it.

Other symbols
No graph here. No relative position determined by voicing or action or
placement. This is just a list, and its relative disorder bespeaks its status as
the orphanage of the International Phonetic Alphabet. This category is the
place to put all the symbols for consonant actions that the International
Phonetic Association in its collective wisdom couldn’t quite fit into either the
pulmonic or non-pulmonic consonant charts. Some of the decisions to
consign sound actions to the “other symbols” list seem obvious enough;
others are more baffling … er, interesting.

Figure 56. Other consonant symbols. (Please see www.langsci.uctac.uk/ipa/ for original charts.)

The first two symbols in the left column are very familiar to us as American
English speakers. Why are they not in the pulmonic consonant chart? The
second of these two symbols is very familiar indeed. It is a / w /, as in the
English orthography of the words “wish” “would” and “away”. What are we
actually doing when we physically enact this sound action? Do you
remember the articulator isolations and our separation of the actions for lip
corner protrusion and lip pursing that we explored early in the work? If we do
both these actions simultaneously, we produce full “lip rounding”, such as we
use in the unobstructed vowel sound “oo”. If we continue the pursing action a
little further, so that it very slightly intrudes into the voiced flow, we produce
a voiced approximant sound. And there we have it: / w /. Or do we?
We notice something else in our recipe description of this symbol. The
IPA has listed it as a voiced labial-velar approximant. Don’t mistake this
for a labio-velar approximant. “Labio-velar” would mean the lower lip
moving toward the velum. If you were actually to succeed in getting your
lower lip to arrive in the vicinity of your velum (your soft palate) you might
require medical intervention to get it away again. What “labial-velar” means
is that the lips are rounded into an approximant obstruction of the flow, as we
have just described above, but in addition the middle portion of the dorsum of
the tongue is arched up toward your velum to simultaneously produce the
voiced velar approximant / ɰ /.
Just say that one sound / ɰ / to remind yourself of the action. So you are
doing two separate articulation actions at the same time. Try doing both
sound actions – at the lips and at the velum – simultaneously. This is our first
experience of a co-articulated sound, a really popular feature of the “Other
Sounds” category. You are basically doing two articulation actions
simultaneously to produce a unique combined sound result.
As it happens, in this particular case, I don’t believe that this is really an
accurate description, so I disagree with the IPA. More specifically, I don’t
agree that both these actions – at the lips and at the velum – need to happen to
produce the / w / that we know and love. Keep your tongue well away from
your velum and just do the fully rounded, mildly intrusive lip action on a
voiced flow. See what I mean? The / w / sounds just fine to me without the
velar action. There’s one vote right there.
The same thing goes for the first symbol / ʍ /, a turned (not inverted –
look closely) “w” that represents the speech action of an unvoiced labial-
velar fricative. Note that it is now a fricative, because unvoiced flow usually
requires more obstruction to produce a unique noise without your having to
increase the airflow.
Now I allow that a lot of people may raise the tongue toward the velum as
a part of speaking the phonemes / w / and / ʍ /. Perhaps most people do so.
But some people don’t, and don’t need to. In my view, the velar co-
articulation of / w / and / ʍ / is secondary to the lip action and is possibly
irrelevant altogether.
There is something unique about the lip action of / ʍ / and / w /, however
– something that doesn’t fit into the categories in the pulmonic consonant
chart. Compare these fully rounded actions to the unvoiced and voiced
bilabial actions of / ɸ / and / β /, and the difference becomes clear: in the
bilabial fricatives / ɸ / and / β / the lips are pursed, only with no lip corner
protrusion – you’re just bringing the edges of the lips closer together. In the
fully rounded / ʍ / and / w / the corners are protruded and the lips are pursed
simultaneously. But this combined lip action chart doesn’t fit into the actions
represented along the Y-axis of the pulmonic consonant chart. To get it in
there would require a new category of action: lip corner protrusion. Since
these are the only symbols that require it, the IPA may well have felt that the
effort wasn’t worth the trouble for only two symbols.2
We know enough about our recipe for consonant production to be able to
explore the rest of these symbols with some confidence. We now know that
the voiced labial-palatal fricative / ɥ / is a co-articulated combination of a /
w / and a / ʝ /. It’s the sound that the French language uses in the middle of
the word “lui” or “huit”.
And now the epiglottals. The famed linguist David Crystal once
described trying to practice speaking epiglottal fricatives and stop-plosives
aloud while standing on a station platform in the London tube (subway, to
Americans). After only a few seconds of enthusiastic epiglottic exploration
he noticed that everyone else on the platform had moved as far away from
him as possible. If these sounds have the power to disperse crowds, clearly,
performing them has to be fun. The epiglottis is the spoon-shaped piece of
cartilage covered with mucus membrane that is attached to your tongue root
and is poised above your larynx; it drops down to cover the larynx whenever
you swallow so that your food is directed to your esophagus and then your
stomach, not down your trachea to your lungs. So all epiglottal sound actions
begin with that swallowing action – except that you don’t engage the top
surface of your tongue as you would when you swallow food. You can tell
that there is a real difference from pharyngeal fricatives in the sound when
you perform an unvoiced or voiced epiglottal fricative / H ] or/ ʢ /. In
performing the epiglottal stop-plosive / ʡ / – you’ll notice that the IPA is
vague as to whether it is unvoiced or voiced – make sure that you are doing
the action using the epiglottis and not the glottis; make sure there’s an
egressive flow, too.
The unvoiced and voiced alveolo-palatal fricatives / ɕ / and / ʑ /, as the
name implies, involve a fricative channel that obstructs the flow along the
midline of the tongue from the top (or laminal) part of the tongue-blade
(moving upward toward the alveolar and post-alveolar area) to the front
portion of the dorsum of the tongue (moving upward toward the front of the
palate). Note that the tongue tip is not slightly retracted, as it is for the post-
alveolar fricatives / ʃ / and / ʒ /.
The lateral alveolar flap / ɺ / is pronounced with the tongue-tip held
against the alveolar ridge throughout. Then the side edges of the tongue are
quickly flapped – using the pressure of the flow – against the side alveolar
ridges to produce a very quick closure and release. The action is very similar
to something we will explore in a few pages – the action of lateral plosion.
The difference is that here the action is looser, so that the tongue flaps rather
than being held and exploded. In English, it might be used in the
pronunciation of the “dl” sequence in the singularly euphonious proper name
Dudley, especially if the speaker is slightly drunk at the time, though its
principal application with perfect sobriety is in the language KiChaka.
Our last “Other Symbol” is / ɧ /. This is another co-articulated sound
action, consisting – as the IPA chart flatly states – of a combined and
simultaneous / ʃ / and / x /, an unvoiced post-alveolar fricative and an
unvoiced velar fricative. You now have the recipe for this co-articulated
sound, so try it. It is the symbol “that dare not speak its name”, or at least a
name that the IPA dare not speak. Ladusaw and Pullum, always to the rescue,
identify this orphan as the hooktop heng. It is only used in a few dialects of
Swedish, but it’s a good exercise to perform, especially because it is tricky to
keep the tongue lowered enough in between the obstructions of the / ʃ / and
the / x / to keep the flow unimpeded in between.
One other thing: the “Other Symbols” category might be an excellent
residence for the “aspirant” phonemes / h / and / ɦ /.
Audio track 29: The pulmonic consonant chart.
Audio track 30: The non-pulmonic consonant chart.
Audio track 31: Other consonant symbols.

Consonant transcription practice 1

Write each phoneme three times in the space provided after its printed
version. Speak each sound action as you write it in an initial position
followed by a single phthong to form one syllable.

Symbol practice exercise

PLOSIVES

NASALS

TRILLS

TAPS & FLAPS

FRICATIVES
LATERAL FRICATIVES

APPROXIMANTS

OTHER SYMBOLS

CLICKS

VOICED IMPLOSIVES
EJECTIVES

Consonant transcription practice 2

Now we will practice our consonant IPA transcription of individual symbols
by writing the appropriate phonetic symbol for each voicing/placement/
action recipe Again, speak the sound action represented by the symbol as you
write it:

Voiced bilabial fricative _ _ _ _ _ _ Voiced retroflex flap _ _ _ _ _ _ _ _ _ _

Unvoiced uvular stop-plosive _ _ _ _ Voiced palatal approximant _ _ _ _ _ _

Voiced labiodental nasal _ _ _ _ _ _ Unvoiced postalveolar fricative _ _ _ _ _

Voiced alveolar stop-plosive _ _ _ _ Voiced retroflex nasal _ _ _ _ _ _ _ _ _

Unvoiced labiodental fricative _ _ _ _ Unvoiced uvular stop-plosive _ _ _ _ _

Unvoiced palatal fricative _ _ _ _ _ _ Unvoiced dental fricative _ _ _ _ _ _ _

Voiced bilabial nasal _ _ _ _ _ _ _ _ _ Voiced uvular trill _ _ _ _ _ _ _ _ _ _ _

Voiced palatal lateral approximant _ _ _ Voiced labiodental approximant _ _

__

Voiced velar fricative _ _ _ _ _ _ _ _ _ Unvoiced palatal stop-plosive _ _ _ _

_

Unvoiced dental click _ _ _ _ _ _ _ _ _ Unvoiced velar fricative _ _ _ _ _ _ _

_

Voiced dental fricative _ _ _ _ _ _ _ _ _ Voiced alveolar fricative _ _ _ _ _ _

__
Voiced alveolar approximant _ _ _ _ _ _ Voiced labiodental fricative _ _ _ _
__

Voiced postalveolar fricative _ _ _ _ _ _ Voiced palatal nasal _ _ _ _ _ _ _ _

_

Unvoiced alveolar lateral fricative _ _ _ Voiced pharyngeal fricative _ _ _ _

__

Voiced bilabial stop-plosive _ _ _ _ _ _ Unvoiced velar ejective _ _ _ _ _ _ _

_

Voiced velar implosive _ _ _ _ _ _ _ _ Unvoiced glottal fricative5_ _ _ _ _ _ _

Voiced labiodental nasal _ _ _ _ _ _ _ Unvoiced epiglottal fricative _ _ _ _ _

_

Unvoiced pharyngeal fricative _ _ _ _ _ Voiced velar approximant _ _ _ _ _ _

_

(Unvoiced)3 glottal stop-plosive _ _ _ _ Unvoiced bilabial click _ _ _ _ _ _ _

_

Unvoiced alveolar stop-plosive _ _ _ _ Unvoiced postalveolar ejective _ _ _ _

Voiced alveolar4 tap _ _ _ _ _ _ _ _ _ Voiced labiodental flap _ _ _ _ _ _ _ _

Unvoiced labial-velar fricative _ _ _ _ _ Voiced palatal fricative _ _ _ _ _ _ _

_
Your teacher, if you have one (if not, enlist a friend), can now speak
individual consonant sounds for you to transcribe. The usual way to speak
consonant actions is to use the consonant in an initial position in a syllable,
followed by a phthong (usually “ahh”. The consonant is then repeated in a
medial position as a two-syllable utterance, with the consonant preceded by
and then followed by a phthong. Write the appropriate symbol on each short
line, paying attention to ascenders and descenders. We include more spaces
than there are consonant symbols so that you can do this practice in more
than one session.
Audio track 32: Individual consonant phoneme quiz.
I have alluded to the two parallel definitions of a symbol in the International
Phonetic Alphabet: as a unit of meaning called a phoneme, and as a unique
physical action. These two definitions maintain a fairly simple relationship
throughout most of the consonant graphs and even in the “other symbols” list.
But with two of the approximants, things get more complicated.
“L” and “r” both have several symbols that represent very different
physical actions that we, as English speakers, would still recognize as the
phonemes “l” and “r.” (In other languages they might represent phonemic
difference.) But in order for us actually to put these letters between the
brackets that tell us that this is not a spelled letter, but a phonemic symbol,
we have to make a choice. Most of the time, we bow to predominant usage or
the preferences of speech teachers: with “l” phoneticians and speech teachers
will usually opt for the so-called “clear l,” so that the phonemic symbol
becomes / ɫ /; this, despite the fact that many Americans pronounce even
initial “l” as the so-called “dark l” or / ɫ /. The physical difference between
the clear and dark “l” is a slight one: both place the tongue tip at the alveolar
ridge, but the dark “l” / ɫ / uses more cupping of the blade and front of the
tongue, common in American speech because most Americans relax their
jaws open more than speakers do in Britain.
To American speakers, other allophones of the phoneme / l / would
certainly sound like accented pronunciations, but they would still be
recognizable as “l” sounds.
The “r” is more complicated still. There are recognizable allophonic
variations of “r” that extend from the articulation at the lips – the Elmer Fudd
“cwazy wabbit” substitution of / w / for the “r” – all the way back to the
uvular trill / R /. This variation in allophones covers a lot of vocal tract
territory. So which symbol might we use to indicate the phoneme?
One solution that many phoneticians used years ago (and phoneticians
still use today for “broad” or less detailed phonetic transcription) is to use the
most recognizable form of the “r”, the same standard lower-case printed “r”
that we are using in this sentence. It has the advantage that if we understand
that we are simply using this “all-purpose r” to mean the phoneme, then we
have a symbol that is exactly the one that we have been using for spelling “r”
words. So our commonly accepted phonemic symbol is / r /, yes?
No. The [ r ] has the disadvantage that as an allophone (indicated by its
placing here in brackets, rather than between slashes), it represents the voiced
alveolar trill that is found in some languages (Amsterdam Dutch or Russian
for example) and in some accents of English (Welsh, for example), but a trill
is hardly all-purpose. Instead, the preferred phonemic symbol for anything
other than the broadest transcription is / ɹ /. As you know, this is specifically
the symbol for the voiced (post)-alveolar approximant “r”, a pronunciation
that is regularly employed in English Received Pronunciation (“RP”) – often
called Standard British – and one that is commonly advocated by American
speech teachers as a preferred pronunciation. So there is some history and
ideology behind this choice of phonemic representation.
The apparent problem is that many, if not most, Americans don’t
pronounce “r” using the alveolar placement; certainly not all the time. Yes?

The 800-pound pirate in the room

And no again: that’s a problem but it isn’t the problem. The problem is that
out of all the marvelous possibilities for allophones of “r” contained in the
IPA (the Alphabet) -
Audio track 33: Allophones of “r”.
[ w ] voiced labial-velar approximant (substitution; can also be another
phoneme)

[ ʋ ] voiced labio-dental approximant (substitution; can also be another

phoneme)

[ ɹ ] voiced post-alveolar approximant

[ r ] voiced alveolar trill

[ ɾ ] voiced post-alveolar tap

[ ɻ ] voiced retroflex approximant

[ ɽ ] voiced retroflex flap

[ ʁ ] voiced uvular fricative

[ ʀ ] voiced uvular trill

– there is no IPA symbol that represents the unique action that is used by
many Americans to form the “r” in a prevocalic (before the vowel in a
syllable) position; and that is used by most Americans – and many other
speakers – to pronounce the “r” in a postvocalic (after the vowel in a syllable)
position.
Why not? As with the / l / phoneme, the explanation is logical, but logic
tinged with a healthy helping of history and ideology. The International
Phonetic Association’s justification for ignoring the “American r” is that it
has no separate existence as a phoneme. Remember that, while the list of “r”
allophones above would all be recognized by any English speaker as sharing
a general “r” quality and meaning, in certain other languages these same
discrete actions and sounds would be recognized as conveying different
meanings – making them different phonemes, in fact.
The IPA (the Association) has a good reason for wanting to draw the line
somewhere in assigning symbols; there are so many possible variations of
consonant and vowel production, especially in accents and in the patterns of
individual speakers, that the IPA, early in its existence, established a principle
of parsimony, requiring that only phonemes could be assigned symbols, lest
the IPA (the Alphabet) sink of its own weight.
But qualifying as a phoneme mostly means arriving early on the phonetic
transcription scene. In initially describing the Welsh language, a phonetician
might note that the speakers regularly use a voiced alveolar trill / r / in both
prevocalic and postvocalic positions. So that’s the phoneme for Welsh, which
gives the trilled / r / the right to exist as a phoneme in the International
Phonetic Alphabet. But we also know that the voiced alveolar trill is used just
as regularly in some accents of English (including the Welsh accent of
English) where it is considered allophonic because it is not, in English, the
approved phoneme for “r.”
Let’s make the argument the opposite way. If Welsh as a language did not
exist and the alveolar trilled “r” did not exist as a sole pronunciation in any
other language, then all the speakers trilling away in Scotland or in
Amsterdam might find no assigned phonetic symbol to represent what they
were doing. The “American r” is actually in the imagined predicament I have
just outlined. As a potential phoneme, it has been preempted by the British
RP post-alveolar approximant.
But what actually is the “American r?”
“The dog’s name”
“Ay mocker. That’s the dog’s name,” says Juliet’s Nurse, speaking of the
first sound in Romeo’s name. And for years in American elocutionary texts,
this Shakespearean reference was summoned up to show how inherently ugly
the “r” sound can be when actually spoken by actual Americans. The nurse,
though, had good reason to characterize the “r” in this manner, because in
Shakespeare’s time, speakers probably pronounced the “r”, in all positions, in
a way that sounded closer to a contemporary American pronunciation. In
London speech, prevocalic “r’ did not center on the post-alveolar / ɹ / until
the eighteenth century; and it was not until the 1780s that London speakers
slowly lost the postvocalic “r” altogether.

In much regional English speech – especially West Country – and in

some areas of the United States, speakers use some version of the voiced
retroflex approximant / ɻ / to produce prevocalic and sometimes postvocalic
“r”. It is this retroflexion that reproduces the classic piratical “Arrrrrrrr” of
Treasure Island fame. In older American phonetic transcriptions from the
1930s, the / ɻ / was the inevitable symbol for “r” in mid-western and western
American speech. It was used because it is the nearest substitute for the real
thing. But this symbol / ɻ / does not accurately indicate the action or the
sound that most Americans use, especially in post-vocalic positions within a
word.
What, then, is this action and its resulting sound? Even if we do not have
a phonetic symbol for it, we certainly can describe it. The sound that most
Americans (exempting some populations of the Eastern Seaboard) use for
postvocalic “r” is what we often call “r-color”. Technically, it is variously
termed a “molar”, braced”, or “bunched r”. And many American use it as
their “all-purpose r”, even in prevocalic positions.
Audio track 34: The braced “r”.
I favor the term braced r, as it is closest to the the physical action, although
the other terms help to complete the description and the degree of bracing can
vary. Unlike the alveolar / ɹ / or the retroflex / ɻ /, the tip and blade of the
tongue are relaxed down with the tip of the tongue behind the lower teeth and
the jaw relaxed open. The middle of the dorsum or body of the tongue is
raised and the side edges of the tongue are braced vigorously against the
inside of the upper teeth in the area of the rear bicuspids and first molars.
Considered in terms of its orientation to the roof of the mouth, this action is
taking place in the palato-velar area. The muscular bracing of the tongue has
the effect of further tensing and thickening the midline of the tongue,
bringing it even closer to the border of the palate and the velum. By varying
the degree of bracing the degree of rhoticity (“r”-ishness) of the sound can
be varied also. So an American can say –

beer, weird, fear, Sears, bare, stair, chairs, large, charm, scarf, bark, torn,
court, lore, bore, boor, moor, toured, purr, turf, learn, purse, churn, fire,
mire, sour, flour

– with a lot of rhoticity or with just a little by varying the bracing. The
“braced r” is a very efficient tool for the dialect actor, because one can easily
calibrate the degree of rhoticity. It is also a crucial part of the posture of
various Midwestern American accents and, again, subtle gradations in
rhoticity help tremendously to localize an accent.
Audio track 35: Gradations of post-vocalic “r”.
But sadly, there is no “braced r” in the IPA nor is there any diacritic to notate
variations in rhoticity, because it hasn’t staked out its territory as a phoneme
in any language. Since American English has now superceded (for better or
worse) RP English as the model for ESL (English as a Second Language)
study worldwide, it would seem appropriate for the IPA to revisit this issue. I
would even be willing to see a newly devised “braced r” symbol present
amongst the “disordered speech” symbols that have been added to the
phonetic alphabet for the use of speech pathologists. Certainly the notion of
American post-vocalic “r-color” as being pathological would appeal to the
English RP zealots and also to many speech teachers in the USA. And
dialecticians would applaud, for very different reasons.
There is more to say about rhoticity in a post-vocalic position, but it is
best dealt with along with a consideration of vowel transcription. So the
suspense can mount.
Answers to consonant quiz (page 146)

1. / b /.

2. / k /.

3. / ʒ /.

4. / l /.

5. / s /.

6. / θ /.

7. / v /.

8. / ɲ /.

9. / q /.

10. / m /.

11. / ɓ /.

12. / x /.

13. / ʔ /.

14. / ʁ /.
15. / ɫ /.

16. / ɖ /.

17 / ʃ /.

18. / ʍ /.

19. / t’ /.

20. / ħ /
14 Sound to word

“Suit the action to the word”

When Hamlet says this to the players he is, of course, referring to physical
gestures when he uses the word “action”. But we know that each phonetic
symbol represents not only a unique sound product, but also a unique
physical action of the articulators that allow the sound to result.
So far we have considered these sound actions more or less in isolation
from one another. But now we can start to place them into a sequence that
will produce a larger unit of meaning: a word.
Let us consider a typical sequence. If we start with the phoneme / b / and
then follow it with the vowel phoneme / i / (a phoneme we will consider
soon) and finish with the phoneme / d / we have produced a sequence of
actions and resulting sounds that in English symbolically represents a
variably small globular object. We have formed the word “bead”.
We notice a few characteristics of this particular word, the first word we
have described completely in phonemes. For one thing, it is entirely voiced.
Each of the three component sound actions is a voiced one. If we changed the
first phoneme from / b / to / s / we would create an entirely different word
that is partially unvoiced and partially voiced. The result would represent a
variably small object that might be globular or not, but one that could grow
into a plant if placed in the ground and watered.
As we move from individual sounds to words, the articulators are
learning the first steps of a complex dance inside the oral cavity. The word
“bead” starts with a voiced bilabial stop-plosive that explodes into the “close
front vowel”6 that in turn is terminated by the voiced alveolar stop-plosive.
Both stop-plosives are stopped and exploded as notated phonetically. But in
getting the articulators from one action to another within a word, the speaker
may need to change the actions to allow an easy sequence. This is especially
true when we articulate words of more than one syllable that require some
syllables to be stressed more than others.

Transcription practice
A good way to learn the consonant symbols easily is to write them and speak
them as syllables or one-syllable words. The underscore line in the words
below represents any phthong (that is, any vowel or diphthong) that you
might wish to speak between the consonants. Be adventurous with the
phthongs but be accurate with the consonants. We will start in familiar
territory with sound actions that are found in American English. Feel free to
consult the consonant charts as needed.

Choose each syllable from the phonetic symbols in the pulmonic consonant
chart. Some will sound like words you know; some will be nonsense
syllables. Do not intrude any other consonant sounds other than the ones
printed here:

[b _ _ t] [f _ _ s] [v _ _ n]
[d _ _ ɡ] [m _ _ z] [p _ _ ɫ]

[k _ _ ʃ] [ʒ _ _ p] [j _ _ θ]

[h _ _ ð] [b _ _ n] [ʃ _ _ ŋ]

Now we will add more pulmonic chart symbols that may or may not be found
in American English. Don’t be shy about checking the chart:

[ʂ _ _ k] [ɸ _ _ ɭ] [ɡ _ _ ɖ]

[ɹ _ _ β] [ç _ _ z] [ɲ _ _ ɢ]
[x _ _ t] [ʝ _ _ β] [ɱ _ _ θ]

[ɳ _ _ ʒ] [h _ _ ɟ] [ð _ _ q]

The next step is to add non-pulmonic consonants and consonants from the
IPA list of “Other Symbols”:
[w _ _ b] [ɓ _ _ ŋ] [ʍ _ _ χ]
[ʀ _ _ θ] [p _ _ ɣ] [! _ _ ɴ]

[ɥ _ _ ʃ] [ʑ _ _ ɱ] [ɬ _ _ ʒ]

[ʔ _ _ ɻ] [ɻ _ _ ç] [ʙ _ _ ʛ]

[ɮ _ _ b] [ _ _ n] [ʋ _ _ l]

[ʁ _ _ ʈ] [ɣ _ _ ɓ] [k _ _ ð]

Two more additions to the mix: consonant clusters (more than one consonant
in sequence) and multiple syllables. Whether or not the sound sequences you
make mirror actual words, they will still sound more like words, even if
they’re nonsense.

Two new ideas and two new techniques

The presence of consonant clusters – two or more consonants not separated
by a vowel in the same syllable – has (somewhat sneakily) introduced into
our transcription of our first words, two new categories of consonant
combination. The first is that of the affricates – two successive consonant
actions in a syllable, the first of which is a stop-plosive, the second a
fricative. Exploding the stop-plosive into the fricative can make the
succession of two actions sound like a single new sound of limited duration.
In English, there are only two affricates in common use. They use the
same action and placement and they are an unvoiced/voiced pair. We have
used them in the preceding list. The first is the unvoiced post-alveolar
affricate, represented in phonetic transcription as / tʃ /. So we know the
affricate combination is composed of a “lower-case t” / t / and an “esh” / ʃ /.
But the resulting sound combination is really more of a sound mix, because
the plosive action of the / t / is still happening as the fricative / ʃ / is formed.
So the affricate result may seem to be a new unique sound rather than a
consonant combination.
The same observations apply to the voiced post-alveolar affricate,
represented phonetically as / dʒ /. The “lower-case d” / d / explodes into the
“yogh” / ʒ /, producing an apparently new sound result.
In other languages, many other affricates are possible. As just one
example, the stop-plosive/fricative combination / ts / occurs in many
languages – Greek is one, as in the Greek names Tsouras or Tseckares. In
these positions, at the start of a stressed syllable and followed by a vowel, the
“ts” combination is truly an affricate, whereas in an English words like
“that’s” or “sets” it is considered a consonant sequence.

Transcribe the consonants in the following words, indicating the phthong by

a subscript line:

chew _ _ _ _ _ _ _ _ _ _ peach _ _ _ _ _ _ _ _ _ _

edge _ _ _ _ _ _ _ _ _ _ jam _ _ _ _ _ _ _ _ _ _ _ _

ratchet _ _ _ _ _ _ _ _ _ _ surgeon _ _ _ _ _ _ _ _ _ ___

magic _ _ _ _ _ _ _ _ _ __ situation _ _ _ _ _ _ _ _ _ _ _ _

natural _ _ _ _ _ _ _ _ _ _ enjoin _ _ _ _ _ _ _ _ _ _

The second category is syllabic consonants. These are consonant clusters

(with no vowels in between) that – in themselves, through the physical action
of moving from one consonant to another – can form separate syllables. Not
all consonants can achieve this feat. The main consonants in American
English that can become syllabic are the nasals / m /, / n /, and [ ɧ ], and the
approximants / l / and / ɹ /. Syllabic consonants are often used by speakers
when moving from a stressed syllable to an unstressed one; it is the lack of
stress emphasis that allows us to take the vowel away and go directly from
one consonant to another.
A syllabic consonant is notated with a diacritic mark, in this case a
stubby vertical line written directly under (the linguistic term is subscript)
the symbol: / ṃ /, / ṇ /, / ḷ /. It will take a little practice to distinguish syllabic
consonants from the same consonants used in a non-syllabic way with a
vowel in the syllable. A syllabic consonant might be the only sound in an
unstressed word, the / n / in “rock’n’roll” being a good example: we know it
stands for the word “and”, but we need only the one nasal sound to convey
that meaning.
Sometimes we can find two syllabic consonants in a row: “animal”
[_nṃṅ /. Notice that here the / ṇ / isn’t syllabic because it starts the syllable
and is not preceded by another consonant. As with other words of this type,
“animal” doesn’t have to be said with syllabic consonants; you could put
vowel sounds in between the consonant actions.
Here is a practical example of the use of syllabic consonants. If we wish
to transcribe the two-syllable word “chasm”, we would write the consonant
symbols [ k_zm ]. As we can easily see, the consonant sequence / zm / forms
its own new syllable, without the need to have a vowel in the syllable at all.
As I have noted, it is a common practice to add a subscript diacritic, a short
vertical line, below the syllabic consonant to identify it as such, as in / k_zṃ
/. This diacritic is not always necessary, but sometimes it is, because a
sequence of consonants might be spoken as either syllabic or non-syllabic.
The noted science-fiction writer Gregory Benford wrote a novel titled Cosm
in which the “zm” sequence is syllabic, but if one were effusively to describe
the novel as having “cosmic significance” the “zm” sequence would not be
syllabic, because the medial consonants “z” and “m” would be drawn into the
separate syllables “cos” and “mic”.
Here are a few of the many possible words in English that can be spoken
using syllabic consonants. Say them slowly and feel the actions you take to
bring the two consonants together, without any vowel sound in between:
level, pedal, even, cabin, seventy, cobble, rattling, bottled, collect, veneer,
prism, Petaluma, eagle, correct, aberration, happen, eaten, satin, schism,
potent, second, able, abalone, ballistic, galactic, homonym (two syllabics),
cinnamon (two syllabics), rattle ’em (two syllabics)
 Audio track 36: Syllabic consonants.

Transcribe the following words, using the format that we have established.
There may be more than one way to do the transcription (meaning that there
is more than one way to say the word!)

eaten _ _ _ _ _ _ _ _ _ _ _ _ _ _

enemy _ _ _ _ _ _ _ _ _ _ _ _ _ _

happening _ _ _ _ _ _ _ _ _ _ _ _ _ _

fatal _ _ _ _ _ _ _ _ _ _ _ _ _ _

seven _ _ _ _ _ _ _ _ _ _ _ _ _ _

fashion _ _ _ _ _ _ _ _ _ _ _ _ _ _

broken _ _ _ _ _ _ _ _ _ _ _ _ _ _

Manilla _ _ _ _ _ _ _ _ _ _ _ _ _ _

huddled _ _ _ _ _ _ _ _ _ _ _ _ _ _

camera _ _ _ _ _ _ _ _ _ _ _ _ _ _

corona _ _ _ _ _ _ _ _ _ _ _ _ _ _

garage _ _ _ _ _ _ _ _ _ _ _ _ _ _

medalist _ _ _ _ _ _ _ _ _ _ _ _ _ _

recollect _ _ _ _ _ _ _ _ _ _ _ _ _ _

And the techniques

If you were paying attention to your own vocal tract action as you performed
the exercises you just completed, you may have noticed that there are also
two new physical techniques that speakers often use to speak syllabic
consonants. (And possibly you were doing them.) They allow you to move
directly from one consonant to another with no vowel in between.
Nasal plosion. Form an / n / in your mouth – just as you would a / t / or / d /,
so that you cannot release any air or vibration through your mouth. Let the / n
/ sound flow through your nose and feel the vibrations tingling through the
pharynx, the back of your throat, and then through the nasal cavity and out
your nostrils.
Keep the flow going on another sighed / n /, but this time try to stop the
flow somewhere inside the nasal cavity, by raising the top surface of the
velum to the top of the nasopharynx, and let the air pressure build up as
though you were preparing to expel something through your nose. Note
where you feel the contact of the closure. Then let the / n / explode through
the nose, as the plosive action helps to bring the velum sharply away from the
nasopharynx. (Note: Speech work isn’t pretty. Deal with it.) Repeat several
times until the action of closing and opening the nasal cavity is clear to you
(or the Kleenex is full, whichever happens first).
Nasal plosion is notated phonetically by a diacritic, which modifies the
stop-plosive that is exploded into the nasal consonant. The small diacritic “n”
(for “nasal”) is written above the baseline of the transcription (superscript)
and after the plosive symbol. Thus “bidden” becomes [ b_dnṇ ] Or “seven”
as it is often pronounced, could become [ s_bnṃ ].
Make sure that you are not substituting a glottal plosion for a nasal one.
Do so by trying it both ways, doing the nasal plosion on both an unvoiced
and voiced plosive: [ tnn ] and [ dnn ], and then on a glottal plosive [ ʔn ].
Can you feel that on the glottal plosion you are not raising or lowering the
upper surface of the velum at all? All the control is at the glottis.
Practice the same actions on [ m ] and on [ ŋ ]. Can you think of any
words where you might use a syllabic / ŋ /?
Pair / m / with / b / and pair / n / with / d /: alternate nasal plosion into the
/ m / with oral plosion into “buhhhh” on one extended flow of vibration: [
bnmmmm ] going right into “buhhhhh”. Then do the same thing with an
extended / n / and “duhhh”. Notice that you can’t make the changeover work
if you do the closing and opening actions simultaneously: you have to raise
the top surface of the velum to stop off the nasal cavity in order to build up
pressure for the oral plosion, and similarly stop off the oral articulators to
build up pressure for the nasal plosion.
Here are some practice words for nasal plosion. Perform them first
unvoiced and then with voicing to feel the physical action before you concern
yourself with sound product.
eaten, beaten, Seton, pittance, kitten, retinal, fatten, catenary, button, curtain,
cotton, Laughton, oaten, Putin, hidden, bidden, Haydn, redden, madden,
badinage, adenoid, sodden, student, couldn’t, seven, eleven, making,
wrecking
Audio track 37: Nasal plosion.
Then write the consonants in each word phonetically. Replace the phthong
with a subscript line:

Lateral plosion. Lateral plosion is another articulator strategy for getting

from a stop-plosive to the syllabic consonant [ ḷ / quickly and easily. In the
word “settle” one can stop the [ t / and explode it laterally into the [ l / by
leaving the tip of the tongue touching the alveolar ridge and dropping the side
edges of the tongue sharply, assisted by the compressed plosion of the
airflow. Phonetically this action is represented by the superscript diacritic [ 1
] modifying the previous consonant, so that “settle” would be transcribed as [
s_t1l ] or as [ s_d1l ]. Obviously this means that voiced lateral plosion is
equally possible, as in the word “middle”, which would be transcribed –
lacking only the vowel – as [ m_d1l ].
Here are some practice words for lateral plosion. Perform them first
unvoiced and then with voicing to feel the physical action before you concern
yourself with sound product.
beetle, spittle, vittle (also spelled “victual”), settle, cattle, cuttle, title, portal,
mettle, needle, middle, meddle, addle, curdle, puddle, coddle, caudal, modal,
feudal
Audio track 38: Lateral plosion.
Then write the consonants in each word phonetically. Replace the vowel with
a subscript line:

Go through the phonetic words in the consonant cluster practice list on page
157 and see if you can identify the affricates and the syllabic consonants –
with or without nasal and lateral plosion. Write only those words phonetically
below, this time using the appropriate diacritics; again replace the phthongs
with a subscript line. As before, say each word silently and then again voiced
before you write it to be certain of the physical action. We are now writing
syllables and words phonetically and we can perceive that the clearly
differentiated consonants provide most of the “meaning framework” for our
language use. But we now must complete the process by adding those
elements that provide even more of the sense, and most of the heart, to our
words. These are what we have been calling the “phthongs”, but we can now
pin them to specific meanings within words. So they may now be termed
vowels.
15 Writing the vowels

Until now, we have explored the vowels (formerly “phthongs”) very freely as
simply the acoustic result of our physical shaping of an unimpeded flow of
voice. All the gradations of sound change – unfamiliar as well as familiar –
are crucial to us in speaking accents. But as fellow speakers of American
English, we realize that some of those acoustic results are familiar to us and
some are not.
We know that “seat” and “sit” are different words in American English –
and British English too – even though the consonants that begin and end the
word are exactly the same. These vowel sounds actually change the meaning
of the word. So, from our experience with consonant phones and phonemes,
we know that the “ea” in “seat” and the “i” in “sit” represent different vowel
phonemes – vowel sounds that are sufficiently different from one another
that the members of a language population will accept them as different and
capable of carrying meaning. If that one phoneme in the word is different, the
whole word is different.
The symbols in the IPA Vowel Chart represent vowel phonemes just as
the symbols in the consonant chart represent (or strive to represent)
consonant phonemes in the languages of the world. We shall get back shortly
to transcribing the phones, those crucial gradations of sound that we use in
accent acquisition, but for now let us consider the simpler category of
phonemes and the symbols that denote them.
 Figure 57. IPA American Vowel Chart. (Please see www.langsci.ucl.ac.uk/ipa/ for original charts.)

The chart above is called a vowel quadrilateral because it is a four-sided

figure. In a way it is similar to the consonant graphs that we have
encountered before, because the symbols have a special relationship to one
another based on physical positioning within the mouth. There is a sort of
“X–Y” axes format as well. We see that across the horizontal “X-axis” is
written “Front Central Back” and down the vertical “Y-axis” is written
“Close Close-Mid Open-Mid Open”.
The first thing that we notice about this graph, as distinct from the
consonant graphs or charts, is that the left side of the vowel quadrilateral is
slanted sharply downward toward the right. There is a reason for this. The
vowel quadrilateral is an attempt to provide a schematic representation of the
relative position of the articulators while engaging in the acoustic shaping of
the vowels.

Figure 58. Vowel chart within the mouth.

We know from all our experience with “phthong shaping” that the way to
physically shape the acoustic properties of vowel sounds is to arch upward or
cup downward the body (dorsum) of the tongue while keeping the tip and
blade relaxed and passive to the process. The arching or cupping can take
place forward toward (or away from) the palate or back toward (or away
from) the velum. The left side is slanted as a mild attempt to show that –
because of the anatomy of the oral cavity – “front vowels” are produced
further forward in a closed position and a bit further back in an open one.
So the vowel chart represents, roughly, a left-facing profile of the oral
cavity. Vowel phoneme symbols near the top of the chart show the highest
position of the arched (closed) tongue. The lowest position show the most
cupped (open) tongue positions, all in a somewhat arbitrary and imperfect
relationship to one another. The front vowels are arched or cupped in
relationship to the palate and post-alveolar areas; the back vowels are arched
or cupped in relationship to the velum and uvular areas. Note that in our
illustration above the tongue in a resting position lies directly through the
middle of the quadrilateral, exactly at a symbol lying in the center of the
chart. This symbol is a turned lower-case e or in our recipe a mid central
vowel –

/ə/

– and it represents the first voiced sound that we produced at the very start of
our work: the “sound of stupidity”, the “uhhhh” that we all utter in the USA
when we don’t know what we’re going to say next. The name of the symbol /
ə / is the “schwa”. The schwa is the only completely relaxed vowel sound
that we produce. Everything else requires some tongue effort.
Audio track 39: Schwa.

Relax your jaw open. Relax your tongue completely. Sigh out a schwa

[ əəəəəəəəəəə ].

Then write the schwa several times and say it as you write it:

Every other vowel that we will form will require some muscular effort of the
tongue – further forward or further back in the oral cavity, higher and more
closed toward the roof of the mouth or further way and open. Enjoy your
relaxed schwa while you can.
Vowels require other muscular effort as well. Look again at the vowel
chart on page 166. Most – though not all – of the symbols are paired. Often
there is a symbol that you might recognize as an English letter paired with a
symbol that is quite unfamiliar. The difference is that the symbol on the left is
unrounded and the symbol on the right is rounded. This refers to lip-
rounding. So if a vowel is unrounded, the lip-corners stay relaxed and the lip
edges stay relaxed apart. In fact, in the high or closed vowels that are
unrounded, the lip-corners may actually be retracted very slightly (hello,
risorius muscle). In the rounded closed vowels, the lip corners are fully
protruded and the lips are pursed just short of obstructing the flow.
From the complete relaxation of the / ə / we move to the full muscular
use of the closed front and back vowels, unrounded and rounded. Simply
experiencing the pair of front closed vowels and the pair of back closed
vowels will tell us much of what we need to know about the physical action
of vowel formation. Here are the two close front vowels:

/i / / y/
A probably unnecessary reminder: when the IPA describes these as “close” it
doesn’t mean “closed”. The articulators are not closed together, nor are they
close enough together to obstruct the flow, because then we would have
consonants, not vowels. But they are as close together as vowels can get.
The left symbol / i / is formed by arching the front of the body of the
tongue strongly upward toward the front of the palate; the lips remain
unrounded and may even be slightly retracted. The voiced flow moves
unimpeded through this shaping. Our vowel recipe for this shaping is an
unrounded close front vowel. The resultant sound is familiar to American
English speakers and may be found in our first phonetic transcription of a
complete word “seat” / sit /.

Speak the following words as they are transcribed phonetically and note that
many of them would have different spellings when written in standard
English orthography:

Audio track 40: Unrounded close front vowel/ i /.

Now transcribe phonetically the following English words:

read _ _ _ _ _ _ _ _ _ _

bleak _ _ _ _ _ _ _ _ _ _

peek _ _ _ _ _ _ _ _ _ _

zeal _ _ _ _ _ _ _ _

keep_ _ _ _ _ _ _ _ _ _

ease _ _ _ _ _ _ _ _ _ _

sea _ _ _ _ _ _ _ _ _ _

wield _ _ _ _ _ _ _

Review the physical sensation of this vowel by holding the correct tongue
and lip position and then breathing unvoiced through the oral shape you have
created. Breathe a little more strongly than you would to speak so that you
can feel the air on the tongue and the roof of the mouth. Feel it with both an
egressive and ingressive airflow. Now do the same with a voiced egressive
airflow. It is important to rely even more on the feel of the vowel shaping
than on the resulting sound.
Leave the tongue in the same arched position. Without moving the tongue,
protrude the lip corners and purse the lips into a full lip rounding that still
does not obstruct the flow. Send an unvoiced airflow egressively and
ingressively through this shaping.
Do the same on a voiced egressive flow. The sound result should be a
rounded close front vowel. It is represented by the phonetic symbol / y /.
We were in very familiar territory with / i /, but the sound / y / is not heard in
American English generally, even though we use the symbol in our spelling
all the time. But in French, this is a very common vowel, found in such words
as “tu” / ty / or “rue” / ʁy /. It is also common in German and in other
European languages that feature lip-rounding on vowels.
 Try going into the / y / in a slightly different way. Start with a voiced / i /
to establish the arched tongue position. As you keep sounding the / i / on a
long sigh, go into the full lip rounding for / y /. Hear the sound change, but
feel it also. As you round your lips, make sure that you do not, out of
muscular habit, pull the arch of the tongue back at all. We are used to
performing a rounded close back vowel in English: this is not it. Keep that
tongue arch forward. Maintaining the close front arch of the tongue, round
and unround the lips moving between / i / and /y /:

/i/ /y/ /i/ /y/ /i/ /y/

Write four “nonsense words” using / y / as the vowel(s). Then speak them:

Audio track 41: Rounded close front vowel / y /.

Close back vowels

/ɯ/ /u/
The tongue arching for the two close back vowels / ɯ / and / u / is similar to
the action for close front vowels / i / and / y /, but with one important
difference. The arching action is enacted with the back portion of the tongue
body (dorsum), not the front, and the arching action moves upward toward
the velum.
We can start our exploration of the unrounded and rounded close back
vowels with the more familiar rounded version, the rounded close back
vowel, which is the right-hand symbol in the pair. The action is represented
by the symbol / u / and is the sound we make in American English when we
say “Ruth” or “bloom” or “route”. Phonemically, these various spellings are
always represented by / u /.
As with the close front vowels, the tongue tip and blade remain relaxed
with the tongue tip just behind the lower teeth. The jaw should remain
relaxed open. All the work should be done with the body of the tongue.
 To form / u /, arch the tongue up toward the back of the velum and
simultaneously round the lips (protrusion plus pursing) fully. Breathe
unvoiced through this shaping to feel it before you sound it. Then add voicing
to an egressive release of flow

[ uuuuuuuuuuuuuuuuuuuuuuuuuuuu ]

Speak the following words as they are transcribed phonetically:

Transcribe phonetically the following English words:

doom _ _ _ _ _ _ _ _

sleuth _ _ _ _ _ _ _ _

drew _ _ _ _ _ _ _ _

goo _ _ _ _ _ _ _ _

pool _ _ _ _ _ _ _ _

blue _ _ _ _ _ _ _ _

soon _ _ _ _ _ _ _ _

flew_ _ _ _ _ _ _ _ _ _
Audio track 42: Rounded close back vowel / u /.

The unrounded close back vowel / ɯ / is not found generally in American

speech, although it is becoming more frequently heard in some regional
accents, especially among younger speakers, as a substitute for / u /. It has
been a familiar feature of Northern English accents for a long time.
Arch the tongue toward the back of the velum as for / u /. Leave the lips
unrounded. Speak the / ɯ / on a long sigh:
 [ ɯɯɯɯɯɯɯɯɯɯɯɯɯɯɯɯɯɯɯɯ ]
Make sure that the tongue arching does not slowly roll forward.

Audio track 43: Unounded close back vowel / ɯ /.

Keeping the tongue arched toward the back of the velum, alternate lip
rounding and unrounding, starting with an unvoiced egressive airflow to feel
the shaping. Then do the same thing with a voiced sigh:

[u] [ɯ] [u] [ɯ] [u] [ɯ]

Now write four “nonsense words” using [ ɯ ] as the vowel(s):

Speak the following transcribed English words that use [ ɯ ] as substitute for
[u]

[hɯz] [tʃɯz] [slɯs] [pɹɯv] [ɡlɯm] [tɯ]

Open vowels: from arching to cupping

We may be surprised that the efficient speaking of open vowels requires a lot
of tongue muscularity. This is because we often speak open vowels by
opening the jaw wide—which is fairly easy to do—rather than by using the
tongue. Opening the jaw beyond the point of natural relaxation is not a very
good idea, though. It can contribute to tempero-mandibular disorder (TMD),
for one thing. It also causes us to associate the constriction of the pharynx
that accompanies forced jaw opening with the normal formation of some
open vowels, a big mistake that even voice scientists can (and sometimes do)
make. Open vowels do not require a constricted vocal tract. The secret of
successful open vowel production is to use the tongue, not the jaw.
Of course the jaw needs to stay relaxed open.
Open vowels use tongue cupping instead of the tongue arching that we
have been using for the close vowels. As before, the tongue tip and blade
remain relaxed with the tip just behind the lower front teeth.
Open front vowels

/a//ɶ/
These two symbols are found at the lower left of the vowel quadrilateral. The
unrounded version / a / is certainly a familiar symbol, a lower-case printed
“a”. In American English spelling (orthography) the “a” symbol can stand
for a lot of different sounds: “cat”, “calm”, “garbage”, for example. In
phonetic transcription, as we know, it stands for only one. But what is that
one? For most American speakers, it is in none of the three words I just
quoted …
The easiest way to get at the [ a / is to consider the symbol that is right
above it in the vowel quadrilateral, represented by the symbol

/æ/
This symbol looks like an “a” and an “e” pushed together. It is that is taken
from Old and Middle English and its official name is the “aesc”, pronounced
“ash”. (It has been around for well over a millennium, and is still used in the
orthography of Icelandic.) It is just slightly less open than the / a /, so the
feeling in shaping the two symbols is quite similar. It is a very familiar sound
in American English speech.
To form the / æ /, keep the tongue-tip relaxed behind the lower teeth, the
jaw relaxed – but not forced – open, and then cup the very front portion of the
body of the tongue strongly downward into the curve of the mandible below
the lower teeth. Keep the lips unrounded. Breathe strongly ingressively and
egressively a few times (don’t hyperventilate!) to feel the shaping and then
sigh a voiced egressive release

[ æææææææææææææææææææææææææ ]
This is what we often call a “short a’ or “flat a”, the vowel that Americans
use some version of in words like “cat”, stand”, “map”, “dance”, “class”,
“half”, “brash”, etc.
 Say these words using this vowel, and then write them phonetically:
fat, sash, clap, vast, chance, calf, mast

Now try to lower the cupping even more. Your well-attuned oral sensors will
note that, in doing so, the curve of the mandible at the front of the jaw
requires that the cupping action of the tongue rolls slightly back as you cup
the tongue further. Breathe silently as before to feel the positioning and then
voice the resulting sound. If you have followed the recipe fully, the sound
will be / a /. the so-called “intermediate a” or “Italian a”.
It is the lowest—most open—front vowel you can make. We find it in
American English in the old-fashioned Southern accent single-vowel
pronunciation of the personal pronoun “I”. It is also the first element of the
diphthong that most Americans use to say that same personal pronoun “I”.
But, in isolation, many American speakers don’t really recognize it. You
have explored a lot of new territory in your phthong shaping; here’s one place
where that experience is very useful.
The phonetic symbol for this sound action, as we know, is / a /. Don’t be
influenced by any previous ideas about a Southern accent, but speak these
phonetically transcribed words as the symbols indicate. Keep the cupping
action really strong without opening the jaw excessively.:

Audio track 44: Unrounded fully open front vowel / æ / and / a /.

We will save the writing of this phonetic symbol for its use in two
diphthongs.
Here is the symbol for the rounded open front vowel:

/ɶ/
This is an interesting symbol because for some time the International
Phonetic Association considered it a more or less theoretical symbol that did
not actually occur in any existing language. One can understand why this was
so. Lip-spreading (the active form of unrounding) and lip-rounding lessen
considerably as we move from the top (close) area of the vowel quadrilateral
to the bottom (open) area. So the comparative physical actions and the
comparative sound results between [ a ] and [ ɶ ] are both negligible.
However, recently Austrian German has come to the rescue, discovering in
the proverbial nick of time that the word for rope “seil” is pronounced [ Sɶ ].
The wedge-shaped colon symbol, by the way, is a diacritic indicating that
the symbol preceding it is lengthened; if you’ve got it, flaunt it.
In producing this symbol, however, we learn something useful about lip-
rounding. To move from [ a ] to [ ɶ ], all you need to do is produce the
strong tongue cupping of [ a ] and then protrude the lip corners very slightly.
No pursing required. Try writing six nonsense words phonetically using [ ɶ
].

and then speak them.

On a voiced sigh, move between [ a ] and [ ɶ ]

[a] [ɶ] [a] [ɶ] [a] [ɶ]

Feel and hear that the difference in sound result is negligible. This is a good
illustration of the fact that lip unrounding or rounding, while important in the
shaping of vowel phonemes, is a secondary action to tongue position – and
especially so with the “open vowels”.

Audio track 45: Rounded fully open front vowel [ ɶ ].

Open back vowels

/ɑ/ /ɒ/
The script a and the turned script a. The unrounded symbol on the left is
not only familiar as another way of writing “a” in English orthography, it is a
very familiar sound in American English speech.
Think – and feel – back to the isolation exercise called “Rolling the ball”.
If you start with the strong front-cupping of the tongue for [ a ], you can
easily “roll the ball” – cupping it all the while – back to a cupping performed
by the back of the body of the tongue. To describe it in a slightly grosser way
(and why not, I say), cup the tongue as far back and as strongly down as you
can without engaging the gag reflex and throwing up. Do not retract the
tongue root as you do this; if you do, you will constrict the pharynx and
create an unnecessarily constricted version of the sound result. The lips
remain unrounded – and unprotruded, for that matter. Do not force the jaw
open! You are now producing the unrounded open back vowel.
Breathe ingressively and egressively through this shaping; it is very open,
but it is not relaxed: it is a very muscular action. Now sigh egressively
through this shaping to form the sound result [αααααααααααααααααααα],
the proverbial sigh of relief. Or of discovery. (The two are surprisingly
similar, aren’t they?)

Speak the following American English words as transcribed phonetically:

[kɑm] [lɑt] [ʃɑp] [hɑt] [vɑz] [ɑ] [dɑk] [ʃɑk] [sɑm]

“calm” “lot” “shop” “hot” “vase” “ah” “dock” “shock” “psalm”

Audio track 46: Unrounded fully open back vowel / α /.
Our word options here are limited. Many [ a ] vowels occur in “ar”
combinations and many in words of more than one syllable that use vowels
you haven’t yet explored. This understood, try writing a few more one-
syllable words with the vowel [ a ]:

The symboll [ ɒ ] is termed by linguists the turned script a . We know that

doing a little lip-protruding will not produce a sound result of great acoustic
significance. Even though the rounded back open vowel [ ɶ ] is paired with
the [ α ], it is of insignificant sound difference unless the cupping of the
tongue is lessened slightly. This is why many phoneticians commonly render
the vowel as placed just slightly higher on the vowel quadrilateral. For the
time being, though. practice the action as a very slight lip corner protrusion
from [ α ] to [ ɒ ]:

Now do the same action again and this time, try a slight relaxation of the
cupping on [ ɒ ] to render the tongue position less fully open.

Audio track 47: Rounded fully (or open) open back vowel [ɒ ].

Filling in the vowel chart

So far our exploration of the IPA’s vowel quadrilateral has focused on the
pairs of unrounded and rounded vowels that are found at the corners of the
vowel quadrilateral. These are the most extreme muscular actions that we
take with the tongue and with the lips in shaping the acoustic properties of
vowel phonemes. With these skills you can explore all the other actions and
their illustrative symbols, using the vowel recipe of physical action. It is
another three-part recipe.
1 Unrounded or Rounded.
2 Front, Central or Back.
3 Close, Close-Mid, Open-Mid, Open, and sometimes Fully Open.
Or any shaping in between. The total space within the vowel quadrilateral
represents – in an imprecise, schematic way – the action of arching or
cupping of the front, middle or back of the body (or dorsum) of the tongue. It
does not involve the tip or the blade.

To give us a more familiar framework without forcing us into the familiar to

the exclusion of everything else, we can now learn quickly the symbols for
the commonly accepted American vowel phonemes. This exploration will
allow you to experience actions – and sound products of those actions – that
we all recognize as conveying definite, unambiguous meaning when
contained within a word in American English. They do not, however,
necessarily represent the phonemes as they are used in every single American
accent.
 Rather than learn all the American vowel symbols right now, we will stay
with the monophthongs, the actions where the articulators are held without
moving, to shape the voiced, unobstructed flow into a single sound. We will
learn the diphthongs – compound vowel actions – later.
As you look at these monophthong symbols, you will see that most of
them are Roman alphabet letters that we already use in English. In fact, we
have already found some of them in the work we have just completed.
As you speak these vowel phonemes, do not worry about minor
variations in your pronunciation of the vowel, as long as you and your
listeners can clearly understand the correct word in the following list of
minimal pairs:7
peat-pit pit-pet
pet-pat pat-pot
pot-pawt pawt-put
put-poot putt-pert
Audio track 48: Minimal pairs.
Audio track 49: Minimal pairs using allophones.
Admittedly a very few of these words are nonsense words in English. We use
them here to keep the consonants that begin and end the words the same
throughout. If there is any doubt about which word you are saying, alter your
shaping of the flow to make the distinction clearly intelligible.
Here, then, are the monophthongal phonemes of American English. They
are easy to learn and easy to speak, because you have already had a lot of life
practise in doing so. But do not get too comfortable with this inherited skill:
the really interesting work for the actor is lies ahead in the myriad of
gradations of these shapings: the throng of phthongs that cram the vowel
quadrilateral like the proverbial stuffed goose. In other words: the
allophones, the ingredients of phonetic distinction, the place where accents
live.
16 American vowel phonemes

In the front

/ i / written / i /
“Lower-case I”. Tongue arched high and forward, toward the mid-palate. Lip
corners unrounded and perhaps slightly retracted. As in “FLEECE,”8 “mead,”
“niece,” etc.
Audio track 50: / i /.

/ I / written / I / or / l /
“Small Capital I”. Tongue arched slightly less high and forward, toward the
mid-palate: Lip corners slightly retracted. As in “KIT,” “give,” “chill,” etc.
Audio track 51: / ι /.

/ ɛ / / written / ɛ /
“Epsilon”. Tongue cupped slightly down and slightly less forward: Lip
corners very slightly retracted. As in “DRESS,” “set,” “vest,” etc.
Audio track 52: / ε /.

/ æ / written / æ /
“Ash”. Tongue cupped very slightly further down and slightly less forward.
Lips unrounded but corners not retracted. As in “TRAP,” “BATH,” “patch,”
“half,” etc.
Audio track 53: / æ /.

In the back
/ u / written / u /
“Lower-case U”. Tongue arched high and back, toward the front of the
velum. Lip corners forward, lips rounded. As in “GOOSE,” “lose,” “chew,”
“true,” etc.
Audio track 54: / u /.

/ ʊ / written / ʊ /
“Upsilon”. Tongue arched less high and back, toward the front of the velum:
Lip corners forward, lips slightly less rounded. As in “FOOT,” “good,”
“should,” “push,” etc.

Audio track 55: / ʊ /.

/ ɔ / written / ɔ /
“Open O”. Tongue cupped slightly downward in back. Lip corners slightly
less forward, lip corners protruded. As in “THOUGHT” (sometimes),
“CLOTH” (sometimes), “caught,” “bought,” “gone,” “all,” etc. But note that
many Americans pronounce these words with a realization closer to the more
open phoneme / α /.

Audio track 56: / ɔ /.

/ ɑ / written / ɑ /
“Script A”. Tongue cupped actively downward in back: Lip corners relaxed,
lips not rounded. As in “LOT,” “THOUGHT” (sometimes), “CLOTH”
(sometimes), “calm,” “father,” “cot,” “lobster,” etc. But note that not all
Americans pronounce all these words using this symbol (i.e. physical action).

Audio track 57: / ɑ /.

In the center
We begin with the most relaxed central or mid vowel, the only fully relaxed
vowel sound. All the vowels we have considered so far require some tongue
use (or effort): arched or cupped, front or back. And also some lip use,
corners spread or forward, lips rounded. Not this one.

/ ə / written / ə /
“Schwa” or “Turned printed E”. Tongue completely relaxed in the mouth:
Lips completely relaxed. As in “uhhhh,” “duhhhh,” the unstressed vowel in
“COMMA,” “relative,” or “away,” or possibly an extremely de-energized (in
articulation) pronunciation of “up,” “love,” “thumb,” etc.

Audio track 58: / ə/.

The two other mid-vowels commonly used in American speech are the
subject of some discord amongst phoneticians. I prefer to place them in the
vowel quadrilateral as many American speech teachers use them, rather than
their IPA placement which is more appropriate for RP British pronunciation.

/ ɜ / written / ɜ /
“Reversed Epsilon”. The tongue is not so much arched as bunched (tensed or
braced slightly). The lips remain relaxed. If, while sounding this symbol, the
tongue is thickened slightly more, the vowel will start to move into the area
of focus for a retroflex approximant consonant [ ɻ /, producing a brief “R-
color” riding on the end of the vowel. This is represented by the attachment
of a [ ˞ / to the vowel, producing [ ɝ ] for the “R-color” version of [ ʒ ]. As in
“NURSE,” “bird,” “learn,” “worst,” “tern,” etc. A variety of spellings, and all
these words may be said with or without “R-color.” I place this sound and
symbol slightly higher (more close) in the vowel quadrilateral than the IPA
does.

Audio track 59: / ʒ /.

/ ʌ / written / ʌ /
“Turned V”. In its American form this vowel occurs just below, and slightly
back of, the schwa. The IPA gives it more back placement, as an unrounded
(slightly lip-spread?) form of [ ɔ ]. In American use the only difference
between it and the schwa in sound is that when we stress the syllable in
which the sound appears there is a tendency to flatten the tongue very
slightly. So there is a tiny change of physical action, and also a minute
change in the sound. We might set the tentative transcription rule then, that
the [ ʌ ] is used in the stressed syllable in a word like “corrupt,” while the
schwa [ ə ] is used in the unstressed syllable, “corrupt,” (since we pronounce
the written “o” like “uh” [ ə ]). As in “STRUT,” “one,” “such,” “flood.”

Audio track 60: / ʌ /.

Here is the way many American phoneticians position the central vowels
within the vowel quadrilateral:

Figure 59. American vowel quadrilateral.

Write the following words in phonemic transcription. Say each word several
times before you try to transcribe it. Say it several times after you transcribe
it too. Try to write the word the way you say it, using the phoneme that does
the job best.
think _ _ _ _ _ _ _ _ _ _ _ _ _ _
pulled _ _ _ _ _ _ _ _ _ _ _ _ _ _
gone _ _ _ _ _ _ _ _ _ _ _ _ _ _
shrimp _ _ _ _ _ _ _ _ _ _ _ _ _ _
crude _ _ _ _ _ _ _ _ _ _ _ _ _ _
perks _ _ _ _ _ _ _ _ _ _ _ _ _ _
refreshed _ _ _ _ _ _ _ _ _ _ _ _
manatee _ _ _ _ _ _ _ _ _ _ _ _ _ _
phlegmatic _ _ _ _ _ _ _ _ _ _ _ _ _ _
probable _ _ _ _ _ _ _ _ _ _ _ _ _ __
wheel _ _ _ _ _ _ _ _ _ _ _ _ _ _
grassy _ _ _ _ _ _ _ _ _ _ _ _ _ _
vexed _ _ _ _ _ _ _ _ _ _ _ _ _ _
gauze _ _ _ _ _ _ _ _ _ _ _ _ _ _
robs _ _ _ _ _ _ _ _ _ _ _ _ _ _
wool _ _ _ _ _ _ _ _ _ _ _ _ _ _
above _ _ _ _ _ _ _ _ _ _ _ _ _ _
on _ _ _ _ _ _ _ _ _ _ _ _ _ _
rationale _ _ _ _ _ _ _ _ _ _ _ _ _ _
nauseous _ _ _ _ _ _ _ _ _ _ _ _ _ _

(Note: Don’t just transcribe words off the paper. Say them. Feel them. Then
transcribe your own physical actions and the sounds that result.)

Here are one or more ways of transcribing the above words:

θɪŋk ʍil

pʊld ɡɹæsi

ɡɔn ɡɑn vɛkst

ʃɹɪmp ɡɔz
krud ɹɑbz

pɝks wʊl

ɹifɹɛʃt əbʌv əbəv

mænəti ɔn ɑn

flɛɡmætɪk ɹæʃənæl

pɹɑbəbl̩ nɔʒəs nɔʃəs nɑʃəs

Your own transcription may not match any of these. But your job is not to
match what is printed here: it is accurately to represent the way you speak
these words. The “right answer” in phonetic transcription is accurate
representation, not meeting a standard way of saying something. It is
descriptive, not prescriptive. At the same time, it might be a useful exercise
to try these pronunciations anyway.
17 New vowels for diphthongs

… and it’s [ dɪfθɔŋz ] not [ dipθɔŋz ]

A diphthong is a sequence of two vowels in which the first is stressed and
the second is unstressed, so that to the hearer the sound slides swiftly and
almost imperceptibly from the first to the second. There are a few new vowel
phonemes in American English that are used primarily in diphthongs.

/ eĭ / written / eĭ /
The vowel phoneme [ e ] is placed just below the [ ι ] in the vowel
quadrilateral. As a pure vowel, it is the sound that we hear in French in “été”,
and occurs frequently in English dialects but not in most American speech. It
forms the first element – the first of the two sounds – of the diphthong that
we often call a “long a”, as in the words “FACE”, “take”, or “play”.
We may not even have thought of this compound sound as a combination
of two vowels rather than a single one before. The tongue starts with the [ e ]
but before the sound ends the tongue raises very slightly to the [ ι ]. It may
seem that the sound is actually sliding higher up to the [ i ], and perhaps for
you it is. Know the difference, however: for most people the fact that the
second element is unstressed means that the action won’t get quite high
enough for an actual [ i ]. But the fact that the tongue is headed in that
direction means that we may think that we are producing an [ i ], when
actually we are reaching a high version of an [ ι ].
The “breve” diacritic over the second element [ ĭ ] means that the symbol
it modifies is shortened and pronounced very lightly. We generally notate a [
˘ ] over the second element in a diphthong to show that the two vowels are
not spoken with equal stress.
So in [ eĭ ] the front of the tongue is arched quite high and forward, with
the lips slightly spread, and near the end of the sound the tongue arches a bit
higher but a bit less forward. As in “say”, “rain” “shape”, “prey”, “lay”,
“vague”. etc. [ seĭ ] [ ɹeĭn ] [ ʃeĭp ]
Alternate saying the diphthong [ eĭ ] and the vowel [ e ] until you can feel and
hear the difference between them.

Now try doing this in words. The diphthong will probably sound like – or
close to – how you say the word in your normal speech. The pure vowel
pronunciation will make the word sound like you are speaking with another
accent from your own, Irish perhaps, or French. Try each pronunciation
twice: once stretching the vowel sound(s) out, and once saying the word as
quickly as possible.
“Same” [ seĭm ] [ sem ]

“Take” [ teĭk ] [ tek ]

“Fail” [ feĭl ] [ fel ]

“Chase” [ t∫eĭs ] [ t∫es ]

“Waste” [ weĭst ] [ west ]

Audio track 61: / e / / eĭ /.

/ aĭ / written / aĭ /
“I think, therefore I am.”
This diphthong is the one we use to assert ourselves in American speech.
Quite literally: “I” (or “eye”, or sometimes “aye”). The so-called “long i”.
But as with the “long a” of [ eĭ ] this is not a single sound: it is a diphthong, a
sequence of two vowels. As in “PRICE”, “like”, “type”, “climb”.
If we want to distinguish this diphthong clearly from other diphthongs or
single vowels, the first element is what is often called an “intermediate a”. It
is written as / a / just like the lower-case “a” that you see again and again in
this text. But it is not pronounced like those spelling illustrations. Many
Americans, even those who use this sound in this (and another) diphthong
every day, have difficulty initially in hearing the vowel in isolation as
different from the vowel sounds around it. But if we approach it physically,
the production of the sound is quite easy. And as we produce it consistently,
we will get used to its unique sound.
 Start with the vowel / æ /. Then cup the front of your tongue even lower.
Make sure that you keep the tip and blade of the tongue relaxed: the work is
happening just behind them. Make sure also that you don’t pull the focus of
the action back so that you are doing a “broad a” or / α /. Keep the cupping
action of the tongue focused forward.
Sigh this new sound on a long / aː /. Hear and feel it without letting it slip
into a more familiar sound. In American dialects we use slight variations of
this isolated [ a ] frequently in the so-called “Boston a” of “Paak the caa in
Haavud Yaad”. Or the genteel old-fashioned Southern of “maaghty faane”.
Now we can use the sound in our diphthong. We start slowly with the / a /
and gradually raise the front of the tongue from its extreme cupping to a high
arch. The lips go from an unrounded open position to a slight spread at the
corners. So the sound slides from / a / to a very short, light / I /.
For now it is sufficient that we write the diphthong as / aĭ /. But because
the tongue is making a bigger cupping-to-arching action on / aĭ / than on / eĭ
/, there is even more of a tendency to slide past the / ɪ / and into an / i /. See if
you can tell which you naturally tend to do. Then see if you can control that
arching of the tongue to produce either one.
There can be other pronunciation differences, too. Many Americans start
this diphthong from / α / or even / ɔ /. So “time” is often said as [ thɑĭm ] or
sometimes [ thɔĭm ].

Try saying the alternative pronunciations of each element of the diphthong in

some words:

light, chime, fight, while, nice, eye, vie, whine, slice,

shine, wild, kite, sighed, nigh, ripe, kite, pie.

Now read some words phonetically:

Audio track 62: / aĭ / with allophones.

/ aʊ̆ / written / aʊ̆ /
In isolation this is the sound of pain: “Ow!” Or as in “MOUTH”. “loud”,
“round”.
Most Americans start this diphthong from / a /, the same “Intermediate a”
(or “lower-case a” as it is termed in phonetics) that we just explored in / aĭ /.
But here the movement of the tongue from cupped to arched goes slightly
back, rather than staying forward. And rather than a slight spread of the lips
from the first element to the second, the lips round almost fully. The resulting
second element is [ ʊ ], the sound in “good”, or “push”.
As with / aĭ /, because the cupping-to-arching action is considerable we
often have a tendency to overshoot the mark and take the second element to [
u ], the fully rounded high back placement (as in “two” or “soon”). Say [
taʊ̆n ] and then [ taŭn ]. Can you say both easily while still keeping the
second element very short?
An extremely common variant of this diphthong is [ æʊ̆ ], the bane of
prescriptive speech teachers because it starts the diphthong from the “flat a”
(in phonetics, the “ash”). If this is the way you pronounce words like “out” or
“down”, it is not necessary for you to change just to please those few whose
ears might be offended; however, you must be able to say the other variants,
and to feel the difference in your mouth. To pull students’ tongues away from
[ æʊ̆ ], speech teachers often mandate that the diphthong be pronounced [ ɑʊ̆
], with the full sonority of the “broad a” (in phonetics the “script a”). To
some, including me, this may sound a trifle old-fashioned, but it does keep
the [ aĭ ] and the [ ɑʊ̆ ] completely distinct from one another.

So try saying the following words in their most common variant forms:

Audio track 63: / aʊ̆ / with allophones on “house”.

“house”
“now”
plow”
“bough”
“count”

[Just for a change. That strange corner symbol after the glottal stop in [ ʰaːəʔ
] means that it is not exploded, just stopped. And the “h” means that there’s
a little puff of air after the unvoiced stop-plosive. We’ll have more of these
decorative symbols soon.]

/ Oʊ̆ / written / Oʊ̆ /

The expression of surprise. Or comprehension. Or perhaps surprised
comprehension: “Oh”. It is what we sometimes call a “long o” as in “GOAT”
or “show” or “hope”.
This diphthong is produced with two “back vowels”. In its idealized form
it begins with the pure back vowel / o /. Some speech teachers maintain that
Americans pronounce the pure vowel (non-diphthong) / o / in words where
the “o” receives lesser stress or is otherwise shorter in duration, such as
“rotisserie” or “cocaine”. I disagree. The action of moving through the
diphthong in / oʊ̆ /, from / o / to / ʊ /, is very slight: you arch the tongue a
tiny bit higher, and purse the lips – which already should have the lip corners
well forward for the / o / – just slightly. We can do this easily even on the
shortest syllable and, if one listens closely, that slight tensing during the
sound happens almost always in American speech.
The proof is always in the action. Try the alternatives: say / oː / as an
extended sound in the following words. Make sure that you don’t move the
articulators at all during the / o / vowel sound:

Row, know, old, post, vote, most, tone, below, abode, mould,
location, total, lotion, cope, sewed, baroque, shoulder, plosive, demote.

Now, toward the end of the release of sound, add that slight pursing of the
lips and that slight elevation of the tongue arching that will take the vowel
into a diphthong / oʊ̆ /. Repeat the same words again as diphthongs.

Audio track 64: / oʊ̆ / with allophones.

Now say the following sentence as naturally as possible:

“I told you to go over to the boat.”

Are you really starting the / oʊ̆ / diphthong with a pure / o /? Possibly; but
probably not. Most Americans slide this diphthong in from a mid-vowel.
(Most likely a mid-vowel we haven’t explored yet.) It’s what we call an on-
glide. So our using / oʊ̆ / to indicate this diphthong is sort of a convention, a
way of consistently identifying the phoneme / oʊ̆ / even though phonetically
each of us might say it slightly differently.

/ ɔĭ / written / ɔĭ /
We know this one is a diphthong. We know that the sound is changing. Even
the most unperceptive listener (which you aren’t, of course) can hear that
weary forward slide from one element to the other: “Oy” – as in the classic
Yiddish sigh of resignation “Oy vey” (at least when it’s spoken by a non-
Yiddish speaker, or maybe a non-New Yorker). Our perception is aided by
the fact that here the diphthongal journey is not only from more open to more
closed, but also from back to front.
There is much possible phonetic variation for this diphthong, as there was
with the ones we have explored above. One might change the first element
from / ɔ / to / o /, as in [ oĭ ]. One might slide on through the / ɪ / of the
second element onto [ i ]. One might even intrude a lip-rounded [ w ] into the
diphthong, as in [ oŭ.wIː ].

Say the following words as naturally as possible, hearing and feeling what
you do:
CHOICE, joy, coy, toy, annoy, boy, coin, voice, ploy, coil, toil, oil, foil,
oyster.

Audio track 65: / ɔĭ / with allophones.

As with the other diphthongs, this is a notation of the phonemic / ɔĭ /, the

approximate way that English speakers say the diphthong to convey a
particular defining unit of meaning in a word, “joy” [dʒɔĭ ] rather than “jay”
[dʒeĭ ].
Try speaking the following sentence using your own usual pronunciation
of the / ɔĭ /:

“Joyce’s boy-toy loitered noisily in the foyer.”

Now try it again definitely using / ɔ / and / ɪ /. (Do the two elements
singly, then slide from on to the other: / ɔĭ /.) Now definitely as [ o ]. Now
using [ oɪ̆ ]. Explore feeling and hearing the difference, even though
phonemically it is the same diphthong.
18 All the vowel phonemes

Unrounded front vowels

We already know all of these, but here they are again:

Speak them again, just for review. Remember that we encountered some of
them as the first elements of diphthongs. Speak them all as extended pure
vowels. Feel the action of the front of the tongue-body moving from arching
to cupping.

Rounded front vowels

We have already met two of the rounded front vowels in our initial
exploration of the corners of the vowel quadrilateral. Here are all of them,
from close to open:

The front of the tongue is arched toward the alveolo-palatal region as for / i /.
The lips are fully rounded as for / u /.

The front of the tongue is arched as for / ɪ /. The lips are rounded as for / ʊ /.

The front of the tongue is arched as for / e /. The lip corners are protruded
(with slight lip-pursing) as for / o /.

The front of the tongue is slightly cupped as for / ε /. The lip-corners are
protruded with no lip pursing, as for / ɔ /.
The front of the tongue is strongly cupped, as for / a /. The lip corners are
very slightly protruded, as for / ɒ /.

Unrounded back vowels

We have met two of the unrounded back vowels already. Here are all of
them, from close to open:

The back of the tongue is arched toward the velum as for / u /. The lips are
slightly spread, as for / i /.

The back of the tongue is slightly arched toward the velum as for / o /. The
lips are very slightly spread as for / e /.

The back of the tongue is slightly cupped as for / ɔ /. The lips are unrounded
as for / ε /.

The back of the tongue is strongly cupped as for / ɒ /, indeed, arguably more
than / ɒ /. The lips are unrounded as for / a /.

Rounded back vowels

We already know these. Here they all are, in order from close to open:

The back of the tongue is arched high toward the velum in a fully close
position. The lips are fully rounded, both lip-corner protrusion and full
pursing, but without obstructing the flow.
The back of the tongue is arched, but slightly less close. The lip-pursing is
very slightly relaxed, but lip-corner protrusion remains full.

The back of the tongue is only slightly arched. Lip corner protrusion is
almost as strong as for / ʊ /.

The back of the tongue is very slightly cupped. Lip-corner protrusion is only
very slightly relaxed from / o /.

Again, in the vowel, quadrilateral this is the “rounded” version of / ɑ /.

However I, and many American phoneticians, consider that the actual
realization of this phoneme can only happen as a distinguishable vowel from
/ ɑ / if you slightly uncup the back of the tongue. There is only a very slight
lip-corner protrusion in this phoneme, which would not alone change the
sound product appreciably. We hear this sound in “short o” sounds as they
are heard in a very few American accents, as well as in English RP (Standard
English) and in “Good American Speech”, the accent that still is often
considered appropriate for classical plays in the United States.

The central vowels

Moving the mountain again

In the articulator isolation exercise “Moving the mountain” we explored
rolling the arch of the tongue from the alveolo-palatal area of the roof of the
mouth to the back of the velar area. The key to this exercise is that the tongue
itself does not protrude or retract; like a wave passing through the ocean, the
crest of the arched tongue dorsum raises and then relaxes sequentially, thus
making the “wave” or the “mountain” appear to move back or forward.
 We have now experienced this selective arching and cupping in the
formation of the closed and open vowels. To carry this experience into the
formation of the central vowels, you can begin with the tongue arched as for /
i /. Roll the arch (but don’t retract the tongue) halfway back toward / u /.
When the tongue arch is midway between the two extremes, it is in the
correct position for the close central vowels.

The lips are spread as for / i /.

The lips are rounded as for / u /.

For the half-close central vowels, lower the arch of the tongue slightly, to
the same degree of arching as found in / e / and / o /, but equidistant between
them.

The lips are very slightly spread as for / e /.

The lip corners are protruded with very slight lip-pursing as for / o /.

For the half-open central vowels, the tongue is slightly cupped, to the same
degree as found in / ε / or / ɔ /, but equidistant between them.

The lip-corners are unrounded as for / ε /. It is commonly realized as [ ɝ ] in

American speech. We have already had a full description of this phoneme as
the nurse vowel.
The lip corners are protruded with no lip pursing as for / ɔ /.
There is one open central vowel. The tongue is strongly cupped, to almost
the same degree as found in / a / and / ɒ /, but equidistant between them.

The lips are wholly unrounded.

Audio track 66: all the other IPA monophthongs.
19 Combinations

Mid-central offglide diphthongs

Without adding any new vowel symbols or sounds, we can explore ways of
combining these basic American phonemes into more diphthongs that play
important roles in American speech. The most important category of these
combined actions is the group of diphthongs that – in traditional transcription
– always end in the “schwa”, the totally relaxed “uhh” sound that we have
(perhaps unjustly) described as the “sound of stupidity”. The phonetic symbol
is, as we already know, / ə /. Here, the / ə / is the object of the off-glide and
so receives weaker stress than the symbol that precedes it.
There are five mid-central offglide diphthongs in American English. The
first one is:

/ ɪə̆ /
as in “near”, “steered”, “veer”, “mere”, “tears”, “shear”, “beer”, “here”,
“tiers”, “peer”.
As written, we would pronounce these words with a primary emphasis on
the first element of the diphthong and then simply continue the vowel (the
phthong) action into the complete tongue relaxation of the schwa / ə / as an
offglide.
Practice each of these words slowly and pay particular attention to that
offglide relaxation, making sure that you always end with complete
relaxation and – at the risk of boring repetition – a relaxed mandible. Then
practice the words at a normal speed.
But it surely has not escaped your notice that despite the fact that these
words have several different spellings, they all have one thing in common:
they all have the letter “r” following the diphthong.

Re-enter rhoticity
We have already discussed rhoticity (r-ishness) as it relates to consonant
transcription. In this group of diphthongs almost all the words have post-
vocalic rhoticity. We even will encounter the classic pirate “arrrrr” as a rhotic
mid-central diphthong.
This rhoticity, in a sense, hangs on the mid-central vowels, and for a very
good reason: the tongue position is very close to the “braced-r” position; only
a slightly greater degree of tongue-bracing tension, and the rhoticity comes
roaring through.
As we know – and regret – there is no phonetic symbol for the “braced r”.
But the International Phonetic Association has allowed rhotic speakers in
America, England, and elsewhere a diacritic symbol that can be appended to
the weak form second element of the diphthong to indicate some
indeterminate degree of rhoticity. If we are using the schwa, as above, the
compound symbol looks like this:

[ɪ ]

So now you can speak the word list on the preceding page adding that third
element of tongue-bracing to produce some degree of rhoticity.

The rhotic zones

We know that the “braced r” is a palato-velar approximant. This
recognition, along with our knowledge of other approximant “r”
phonemes, allows us to plot a set of three approximant “rhotic zones”
along the roof of the mouth from front to back. They would separate other
phoneme approximants in the following form:

Other vowel options

For most American speakers, the triple action (strong vowel/weak vowel/
rhotic bracing) that you enacted in the slow version of [ ɪ ] sounds a trifle
odd, mostly because the bracing has to take a considerably journey from the
complete relaxation of the schwa to the braced approximant rhoticity. For
most Americans, the second element of the diphthong would be spoken in a
slightly more close position, with the tongue very slightly arched in the
middle of the dorsum. From this point, only a tiny bit of additional bracing
would bring the tongue into the palato-velar rhotic zone.
The traditional symbol for American speakers is / ɜ /, the “reverse
epsilon”. While this symbol is rendered on the IPA vowel chart as an
“unrounded half-open central vowel, its “phonemic territory” can extend
upward to the “unrounded half-close central vowel” position (so, in our
physical terms, it is slightly arched, instead of slightly cupped). It was
proposed as such to the IPA in the 1930s by the eminent American
phonetician John Kenyon, but this proposal was rejected. Nonetheless,
American phoneticians have followed this practice since that time. Kenyon
also proposed this position for monophthong words like – well – “words”.
Or:

first, purse, shirt, learn, girth, murk, turn, swerve,

curb, flirt, shirred, gurn, dearth

If the same mid-central offglide diphthong was written in this manner

with added rhoticity, it would look like this:

[ɪ ]

There is another option, though, because if you look at the vowel

quadrilateral again, you see that there is another symbol occupying that
unrounded half-close central vowel space. We have already established that
many of these symbols are theoretical anyway, according to the Handbook of
the International Phonetic Association. But if we think of these places on the
vowel quadrilateral as being roughly equivalent to physical action (and to
relative acoustic placement), the appropriate symbol to use for the second
element of this diphthong would be the “reverse lower-case e” or / ə /,
making the appropriate symbol look this this:
 [ɪ ]

One could very well use this symbol for all the monophthongs in the previous
word list as well, if one were to choose the option of picking the closest
phonetic symbol to the appropriate physical placement, rather than expanding
the phonemic territory of the phoneme / ɜ /. Its only disadvantage is that it is
not recognized as an English phoneme, but then so few of the central vowels
are recognized as actual phonemes in any language!
All of these second-element (or weak-vowel) features can be carried
though in the other four mid-central offglide diphthongs, so we do not have
to repeat all the explanation each time. Instead, we will simply present all
three transcription options.
The second mid-central offglide diphthong is:

[ ɛə̆ ] [ ɛ ] [ɛ ] [ɛ ]

with the first element slightly cupped in the front of the tongue’s dorsum, but
very forward in placement. As in the words:

square, care, stair, chair, there, their, lair, eyre. glare, dare

The third mid-central offglide diphthong is:

[ ʊə̆ ] [ ʊ ] [ʊ ] [ʊ ]

with the back of the tongue’s dorsum arched toward the velum in a position
between half-close and close and the lip-corners protruded and rounded in the
first element. As in the words:

cure, tour, poor, sure, Coors, boor, lure

The fourth mid-central offglide diphthong is:

[ ɔə̆ ] [ ɔ ] [ɔ ] [ɔ ]
with the back of the dorsum very slightly cupped in the first element in a
half-open position and the lip-corners protruded with no pursing. As in the
words:

north, force, shore, more, lore, four, pour, store, gore, nor, war, oar

The fifth and final mid-central offglide diphthong is:

[ ɑə̆ ] [ ɑ ] [ɑ ] [ɑ ]

with the back of the tongue dorsum cupped strongly down in a fully open
position with the lip wholly unrounded. As in:

Aaarrrrrrrrrrrrrr

(At last.) But also as in:

start, large, farm, cart, hard, carve, charge,

part, mark, shard, carp, farce

Audio track 67: The IPA dipphthongs / ɪə̆ / / ɛə̆ / / ʊə̆ / / ɔə̆ / / ɑə̆ /.

Triphthong
A triphthong is a sequence of three vowel sounds in a single syllable. The
first element (sound) of the triphthong is the only one that is stressed and
elongated. The other two are sounded as a sequence of shortened, unstressed
offglides.
The two triphthongs in most common use in American English are [ aĭ
] and [ aʊ̆ ].
 [ aĭ ] [ aĭ ] [ aĭ ]

As with diphthongs, a triphthong begins with a stressed element. In both the

common American English triphthongs, this stressed first element is [ a ], the
lowest or most open of the front vowels, in which the front of the tongue is
cupped strongly. The vowel sound raises (closes) rapidly to an unstressed [ ĭ ]
and then relaxes to the central mid-vowel [ ě ] with r-color. Some illustrative
words are:

fire, mire, sire, tire, wire, choir, hires, liars, prior, inquire, flyer

where the triphthong is spoken as a single syllable, as in “sire” [ saĭ ]. Many

speakers, following the practice beloved by rock music lyricists, will
commonly change a triphthong into a two-syllable word; so that “fire” is
pronounced as [ fa .jɝ ] instead of [ faĭ ]. When that happens, the front of
the tongue (not the tip or blade) is raising so much that the vowel / I / doesn’t
just raise to / i /, it gets so close to the roof of the mouth that the action turns
into a consonant, the voiced palatal approximant / j /, because there is a very
slight obstruction of the flow.

Audio track 68: Triphthong/two syllable contrast / aĭə̆ /.

To feel and hear the difference, say the word list again, speaking each word
alternately using the / j / and not using it, as in:
 [ aʊ̆ ] [ aʊ̆ ] [ aʊ̆ ]

Most Americans say the first sound of this triphthong as [ a ]. Many

Americans use [ æ ], and many speech teachers would prefer us to use [ α ].
Whichever you choose, the effect for a triphthong is the same: the first sound,
or “element”, is stressed and lengthened, with a closure of the tongue on the
first offglide followed by a relaxation of the tongue to the “schwa” [ ə ]. All
three sounds in succession are spoken within one syllable. All of the
following words can be spoken on one syllable as triphthongs:

sour, shower, power, cowered, coward, tower, our, bower

It may come as a surprise to some speakers that the words spelled with “-
ower” can be spoken in one syllable, since many Americans habitually speak
such words with two syllables. Here the lip rounding from [ a ] to [ ʊ ] is
increased so that there is more lip-rounding than would be necessary for even
an [ u ], and the voiced rounded bilabial approximant [ w ] is formed,
dividing the action into two syllables. Neither the one-syllable (triphthong)
version nor the two-syllable version is more correct, but we should have the
skills to speak the words either way.
There are other potential triphthongs in American English. “Mayor” or
“payer” could be spoken as the triphthong words [ meĭ ] and [ pʰeĭ ]. But
“mayor” most often becomes a a homophone for “mare” [ mɛ ], or is spoken
in two syllables [ mei.jɝ ].

Audio track 69: Triphthong/two syllable contrast / aʊ̆ə̆ /.

20 Mid-central offglide diphthongs

Phonetics and phonemics: the vowel version

We know that a phoneme is a speech action that conveys a unit of meaning
within a language; it is that commonly understood agreement among speakers
of English, for example, that “sit” has a different meaning from “set” or “sat”
or “sought” or “suit”, all of which, whatever their spelling, have a single, but
different, vowel phoneme between the same consonants. We also know –
from our free-form explorations of “phthongs” – that there are nearly infinite
variations within these vowel phonemes. What is a way to transcribe an
accurate version of vowel placement? There are three, actually.

1. Substitutions
Because words exist within larger linguistic structures like phrases and
sentences that help to identify meaning, a phoneme may retain its unit of
meaning even though its action/sound variation takes it into the territory of
another phoneme. Take the two words “get” and “git”. In isolation, what do
they tell us about their meaning? We might guess correctly that “get” is a
verb and that its meaning most likely is “obtain”. We might even know that
“git” is a noun that is UK slang for “contemptible idiot”. We do not use the
word much in the USA, despite our adequate supply of idiocy. But we also
know, because many of us say it that way, that “get” [ ɡεt ] is very frequently
pronounced as “git” [ ɡɪt ]; it is even spelled that way often, as in the cowboy
song classic “Git along, little dogies”. We still know it means “get” though.
(And we all know that a “dogie” means a motherless calf.)
Such substitutions of one phoneme for another to show pronunciation are
extreme examples of what I term “phoneme intrusion” to describe a phoneme
that starts to sound like another one but still retains its original unit of
meaning.

Advantages. When we are writing vowels and diphthongs phonetically it is a

great convenience to be able to use these simple substitutions for accents: [
kɛnt ] for [ kænt ] (“can’t”), [ fo st ] for [ fɝst ] (“first”), and so on. Most
speech teachers and many phoneticians represent accents of English or the
pronunciation of other languages in just this way: it’s not only easy to
transcribe, it’s easy for the reader to understand.

Disadvantages. The trade-off for this convenience is some degree of crudity.

Accents that are taught using only this limited number of phonemes often
sound unrealistic because they frequently (though not always) fail to
represent the phonetic or allophonic subtleties that most accents possess. But
first, we must examine …

The territory. When we look at a vowel quadrilateral, divided into sections by

lines and populated by symbols in suspiciously orderly array, we may get a
hint that we are looking at a theoretical construct. And so we are. When
famed English phonetician Daniel Jones established this version of a vowel
chart (modifying earlier variants) in 1917, he did so by placing most of the
front and back vowels in equidistant intervals from close to open. The few
remaining vowel phonemes are also placed with some symmetry in between
these so-called cardinal vowels. Jones constructed this formal graph to define
the physical territory of the vowel chart, limited to its quadrilateral shape by
the limits of vowel articulation within the oral cavity. By so doing, he was
able to define physically the territory within the chart by providing specific
points of physical reference, all related to degree of tongue arching/cupping
or lip rounding/unrounding, the two chief ways of shaping acoustically the
unobstructed flow to form vowels. He did not place the cardinal vowels in
these positions because he was in any way suggesting that words using these
phonemes must be pronounced at the cardinal positions.

2. Diacritics
The second of the three ways that we can move these theoretically placed
phonemes into more accurate positions within the vowel quadrilateral is by
using diacritic symbols to adjust the phonemes into their possible allophones.
The transcription on page 232 of “clothes” uses diacritics to render the
phonetic variety more accurately both in consonants and in vowel
transcription.
 There’s only one small problem, though: you won’t be able to decipher
these diacritics at this stage of your work. Still, I invite you to marvel at the
sentence below in all its diacritical glory.

[spi̠k ðə̞ spi̠tʃ a̽ĭ pɹ̥e̝ĭ jü | əz a pɹə̃nãʊ̆st ɪ̽t tu̟ ju̽ | ttɹ̥ɪ̤pɪ̝ŋli ɑ̤ ðə tʰʌ̰ŋ ]

A few chapters further on, you will be able to read it trippingly on the tongue.
For now, it suffices that you know that diacritics can modify both consonants
and vowels.

Advantages. Using diacritics allows you to be more varied than does simple
phoneme substitution. It allows you to give quite an accurate representation
of an accent. That is why transcribers, when listening to native speakers of an
accent, use phoneme substitution with diacritic modification more than other
methods.

Disadvantages. Most of these diacritics only give the reader/speaker a

general direction for this subtle physical adjustment of the phoneme, but do
not specify exactly where the vowel allophone is located within the
quadrilateral. So one speaker might interpret the diacritic as leading to one
position while another speaker – equally correctly, based on the evidence –
might interpret it as leading to a slightly different position. This is getting
very subtle indeed and may be more than accent learning requires, but it is
not yet completely accurate. It also fails to show us readily if there are any
clusterings of vowel realization that might give us useful physical evidence of
the overall posture of the accent within the vocal tract.

3. Moving the phoneme

This is the last of the methods for transcribing vowel allophones. It requires
that the reader/speaker of the transcription consult a separately created vowel
quadrilateral that is not a part of the actual transcription in order to note
exactly where the phoneme is placed at variance from the cardinal
placements.
Here are the vowel placements for a native speaker of Swedish, as
recorded in the Handbook of the International Phonetic Association:
 Figure 60. Vowel quadrilateral, Swedish, IPA Handbook (Cambridge University Press, 1999).

We can see that all the phonemes are located away from their expected
positions on the cardinal chart, some only slightly and others considerably.
We can see also that there is a definite concentration of vowels in
comparatively close and fronted positions. All the commonly used rounded
front vowels are present also. If these are the vowel positions that are easy for
a Swedish speaker to use in informal speech, it suggests something about the
characteristic posture of the articulators when speaking the Swedish vowels.
Try to speak each of the vowel allophones as recorded on this chart. Start
with [ iː ] and move down the front from close to open, then to the open
central vowel allophone [ a ] and then up the back from open to close [ u ].
Finally explore the difference between the mid-central rounded vowel [ ɵ ]
(usually shown as a mid-close central rounded vowel) and the mid-central
schwa [ ə ] that usually appears in this position.

Advantages. This is a more precise way of representing the vowel allophones

than either of the other methods. It helps us understand patterns of vowel
placement that help us to define the physicality of the accent, whether from
an English-speaking dialect, from a primary speaker of another language, or
from a native speaker of another language who is speaking in English. It also
shows, quite dramatically in this example, the degree to which an allophone
may start to occupy the territory of another cardinal phoneme. These
boundaries are very permeable. (This permeability, by the way, is another
problem in accurate transcription when we are trying to use only vowel
substitutions.)
Disadvantages: It is not a technique that lends itself to real-time
transcription, as we might try in listening to a contributor performing an
accent, or even listening to a recording that you could repeatedly hear. We
would have to prepare the allophonic vowel chart quite separately from our
actual transcription, which would in itself be a Catch-22 problem because our
phonetic transcription would depend on the existence of the vowel chart. This
technique is best utilized by researchers who can gather data from a number
of samples and who have the time to subject the samples to acoustic analysis.
Finally, at a certain point of departure from its cardinal position, it might be
more practical to simply write another phoneme rather than trying to extend
the phonemic territory of a phoneme to occupy some distant region of the
vowel quadrilateral. The real question is how committed a transcriber is to
that unit of meaning that a phoneme represents: if meaning rather than
physical placement is paramount, “get” is better represented by a diacritic or
by a relocation of the phoneme than by writing it as “git”. If simple
physicality is at a premium, “git” gits the message across very well.
From a practical standpoint, we should be prepared to use all these
transcription methods, depending on the needs and environment. Narrow
transcription of accents is not needed in many cases. Narrow transcription is
crucial in others. What is crucial for the transcriber is making sure that the
reader/speaker knows what the ground rules are.
21 Polysyllabic words

Onset of vocalization
When we whisper something to someone, we know that all the sounds we’re
producing are unvoiced; that is, an extra-strong airflow shaped by the
articulators is doing the work. In whispering we never bring the vocal folds
together to form the vibration action that we call “voice”.
By contrast, when we speak “aloud” we often think that we are now fully
“on voice”, with the vocal folds always vibrating to send our voices
throughout a space to all our listeners. But this is not so. If it were, then
Hamlet’s advice to the players – “Speak the speech, I pray you, as I
pronounced it to you – trippingly on the tongue” – would actually sound like
this: “Zbeag the zbeedge I bray you az I bronounzed id do you, dribbingly on
the dung.”
In connected human speech, where we speak one or many words in a
single uninterrupted breath, the vocal fold vibration – the voicing – turns on
and off many times during the simplest sentence. Of course all the vowels
and diphthongs are voiced, but most consonants can be paired in either
voiced or unvoiced action, producing sounds very different from one another.
In the unvoiced stop-plosives / p /, / t /, and / k / the flow of air often
explodes the articulators apart forcefully, producing a puff of air. If you put
the palm of your hand three or four inches in front of your mouth and whisper
“puh” energetically you will feel that puff of air sharply on your hand. Now
with your palm the same distance away, say a fully voiced “buh”. You will
feel little or nothing.
Say the word “team”. Say it again, and notice that as the / t / explodes
there is a tiny fraction of a second where that compressed puff of air has to
escape before the voice “turns on” to produce the vowel sound “ee” / i /. It’s
not “t” followed by “ee”, it’s “t” followed by a puff of air followed by “ee”.
Phonetically this action is termed aspiration and in American speech the “t”
in team is usually (but not always; these are not hard and fast rules)
pronounced as an aspirated “t”. The way to notate this small aspirate pause
before that next voiced sound kicks in is by writing a small superscript [ h ]
above and to the right of the stop-plosive that produced the aspiration. So the
word “team” as it is pronounced by most speakers would be written
phonetically as [ tʰim ].
The same thing happens with / p / and / k /. The words “path” and
“cough” as spoken by many American speakers would be [ pʰæθ ] and [ kʰɔf
] when written in phonetics.
What has happened, then, is that the aspirate puff of air has pushed the
onset of vocalization a little further toward the end of the word.
So do we always put a [ ʰ ] after every unvoiced stop-plosive that we say?
Are we always producing that aspirate puff of air? No. There are a few
conditions, based on physical action, that usually govern aspirate unvoiced
stop-plosives / p /, / t /, and / k /:

1 The stop-plosive will usually be followed by a vowel, diphthong,

triphthong, or in some cases a pause (when it is at the end of a word) in the
same syllable for aspiration to occur.

2 The syllable is almost always stressed for aspiration to occur. It doesn’t

have to receive the primary stress, but it has to be stressed somewhat. Or to
put it the opposite way: if the syllable is unstressed, the stop-plosive – even if
followed by a vowel – probably won’t be aspirate. So in the word “compel”
the / k / would probably not be aspirate because the “com-” is not stressed.
But the / p / would be aspirate because the “-pel” is stressed: [ kəmpʰɛɫ ].

3 When an unvoiced stop-plosive occurs before another consonant, whether

in the same syllable, in different syllables, or from the end of one word to the
beginning of another in connected speech (even if the next word begins with
a vowel), the stop-plosive is not generally aspirate. The word “inept” would
thus probably be spoken as [ ınεpt ] or possibly [ ınεpth ]. The word
sequence “take vows” would be said [ tʰeĭk vaʊ̆z ], where the [ tʰ ] is aspirate,
but the [ k ] is not. The word sequence “at odds” would probably be spoken [
æt αdz ] with no aspiration.

4 There are four consonants in American English which are affected by an

unvoiced stop-plosive preceding them in a syllable. They are / j /, / w /, / ɹ/
and / l /, all approximants. Like other consonants, these sounds are spoken
immediately after the stop-plosives without any intervening puff of air. But
unlike other consonants these approximants allow the aspirate puff of air
from the stop-plosive to take place, in effect blowing out their voicing. Two
of these approximants, / j / and / w /, have their own unvoiced fricative
symbols, / ç / and / ʍ/ respectively. So “puke” and “quince” would usually
be said as [ pçuk ] and [ kʍɪns ]. The voicing doesn’t fully come in until the
vowel.
The other two approximants have to be treated differently. / l / does
indeed have an unvoiced fricative form, [ ɬ ]. It is pronounced with the side
edges of the tongue strongly obstructing the lateral flow of air, producing a
very hissy unvoiced “l”. This is the sound we hear in Welsh “double-l”
words, such as Dylan Thomas’s fictional town “Llareggub” in his verse play
Under Milk Wood. (Never one to ignore a dirty joke, Thomas concocted the
fictitious Welsh place-name from “bugger-all” spelled backwards.) For most
American speakers, however, the unvoiced “l” is not spoken so tightly after
an unvoiced stop-plosive. So instead of the [ ɬ ] it is more accurate to write
the ordinary [ l ] with a small subscript circle directly underneath it; this
diacritic indicates that the voiced symbol it modifies is unvoiced: [ ]. Thus
“plod” could be notated as [ p ɑd ] or “applaud” as [ əp ɔd ]. The last of these
approximant symbols, [ ɹ ], has no unvoiced fricative form as a separate
symbol, so once again we employ the unvoicing diacritic underneath it to
produce the notation [ tɹ̥ʌdʒ ] for the word “trudge” or [ kɹ̥ɪpɫ̩ ] for “cripple”.

5 There is an exception to this. If the stop-plosive-plus-approximant

combination is preceded in the same syllable by an “s”, then the onset of
vocalization is pulled back toward the beginning of the syllable, and the
approximant is voiced again. Thus “ squeal” is spoken – and notated – as [
skwil ] not [ skʍil ], and “strip” would be spoken as [ stɹɪp ] [ stɹ̥ıp ]; note
the difference in the voicing or unvoicing in the following word pairs:

pray – spray pre – spree cream – scream trap – strap,

queaze – squeeze cue – skew plain – explain
6 The conditions we have just explored are not inflexible rules; they are
perceptions about common usage. Individual speakers may vary from these
guidelines, voicing or unvoicing sounds, intruding an aspiration between a
stop-plosive and a vowel, or not doing so, in ways that confound our attempts
to organize speech actions into neat right-and-wrong categories. The
important thing is to be aware of the possible and probable actions, to listen
for them in the speech of others, and to feel them in our own speech.

American vowel, diphthong, and consonant practice

Write the following words in phonetic transcription

1 Place
2 Noisy
3 Broken
4 Blame
5 Grope
6 Wouldn’t
7 Affairs
8 Grievous
9 Flange
10 Snicker
11 Busy
12 Whipped
13 Languor
14 Evasive
15 Pulley
16 Abrasion
17 Northern
18 Dirigible
19 Pazzazz
20 Caffeine
21 Corpulent
22 Neither
23 Dowry
24 Voracious
 25 Pauses
26 Snafu
27 Shopworn
28 Clouds
29 Cowered
30 Attire
31 Chicanery
32 Curdled
33 Toughs
34 Elevate
35 Winch
36 Choir
37 Reigning
38 Appearance
39 Chopped
40 Result
41 Stench
42 Mask
43 Showers
44 Detour
45 Garbage
46 Enough
47 Placate
48 Marshal
40 Soften
50 Luge

American vowel, diphthong, and consonant practice

With sample transcriptions

Speak each of these transcriptions. Compare the pronunciation of the word
with the way you would pronounce it and compare your own transcription
wiith the version here. Each answer given here represents only one possible
pronunciation. Do not interpret it as the only “right” way to say the word.
1 Place [ p eĭs ] 26 Snafu [ snæfu ]

2 Noisy [ nɔĭzi ] 27 Shopworn [ ʃɑpwɔɚˇn ]

3 Broken [ bɹoʊ̆kən ] 28 Clouds [ k aʊ̆dz ]
4 Blame [ bleĭm ] 29 Cowered [ kʰaʊ̆ɚˇd ]
5 Grope [ ɡɹoʊ̆p ] 30 Attire [ ətʰaĭɚˇ ]
6 Wouldn’t [ wʊdⁿn̩t ] 31 Chicanery [ ʃɪkʰeɪ̆nɜɹi ]
7 Affairs [ əfɛɚˇz ] 32 Curdled [ kʰɝdˡɫ̩d ]
8 Grievous [ ɡɹivəs ] 33 Toughs [ tʰʌfs ]
9 Flange [ flændʒ ] 34 Elevate [ ɛləveɪ̆t ]
10 Snicker [ snɪkɜ ] 35 Winch [ wɪntʃ ]
11 Busy [ bɪzi ] 36 Choir [ kʍaɪ̆ɚˇ ]
12 Whipped [ ʍɪpt ] 37 Reigning [ ɹeɪ̆nɪŋ ]
13 Languor [ læŋɡɝ ] 38 Appearance [ əpʰɪ ɹəns ]
14 Evasive [ ɪveĭsɪv ] 39 Chopped [ tʃɑpt ]
15 Pulley [ pʰʊli̞ ] 40 Result [ ɹɪzʌlt ]
16 Abrasion [ əbɹeĭʒn ] 41 Stench [ stɛntʃ ]
17 Northern [ nɔɚˇɝn ] 42 Mask [ mæsk ]
18 Dirigible [ dɪə̆ɹɪdʒəbɫ ] 43 Showers [ ʃaʊ̆ɚˇz ]
19 Pazzazz [ pəzæz ] 44 Detour [ ditʰʊɚ ]
20 Caffeine [ kʰæfin ] 45 Garbage [ ɡɑɚbɪdʒ ]
21 Corpulent [ kʰɔɚˇpjʊlənt ] 46 Enough [ ɪnʌf ]
22 Neither [ niðɚ ] 47 Placate [ pl̥eɪ̆kʰeɪ̆t ]
23 Dowry [ daʊ̆ɹi̞ ] 48 Marshal [ mɑɚ˅ʃɫ̩ ]
24 Voracious [ vɔɹeɪ̆ʃəs ] 49 Soften [ pʰɔzɪz ]
25 Pauses [ sɔfn ] 50 Luge [ luʒ ]

Nonsense words
Speak these words that employ only English phonemes. Don’t try for fluency
immediately; speak each word slowly. Accuracy is more important than
fluency. Bu after you are really confident with each word, speak it in a
normal conversational pace:

Now try these words that use phonemes that are not all found in English. Use
the same method as above.

Don’t be afraid to practice these very slowly, referring to any of the charts or
explanations you need. Put the words together and then say them several
times until they become more comfortable to speak.

Now we will take another major step, moving from single words – real or
nonsense – to phrases and sentences; what linguists call connected speech.
This step brings in new elements, because we don’t always speak words in
connected speech in the same way that we do when we say them individually.
You can easily prove this. Speak the sentence “What do you want me to
say?” first as separate individual words with an inhaled breath between each
word, and then as a connected sentence on one breath. Notice what happens
to the words.
 For this next step, we have to move from “broad” or general phonetic
transcription to “narrow” or specifically detailed phonetic transcription.
22 Connected speech

Moving toward “narrow transcription”

“Rumours about the new play had been fanned by Shaw through London.”
(Michael Holroyd, Bernard Shaw, Vol. 2)

On the next page will be found several different phonetic transcriptions of

this sentence, using slightly different pronunciation choices or different
degrees of detail in the transcription. In order to represent these three
different pronunciations, we will use symbol changes from one phoneme to
another and also diacritic symbols to modify individual phonemes into their
allophonic variations. But first, here are the diacritics used, each creating an
allophone of a phoneme symbol:

[ ə̆ ]
“Breve.” The symbol modified is given no stress and is very short.

[ɝ]
“Rhoticity sign.” The symbol modified ends with a slight tongue-bunching
into “r-color” – the “braced r”.

[ ɪ̝ ]
“Raising sign.” The symbol modified is pronounced with the tongue arching
(or uncupping) slightly higher toward the roof of the mouth. Also referred to
as being a more “close” sound.

[ ]
“Lowering sign.” The symbol modified is pronounced with the tongue
cupping (or de-arching) slightly lower toward the floor of the mouth. Also
referred to as being a more “open” sound.

[ d̥ ]
“Under-ring.” The symbol modified is pronounced unvoiced.

[t ]
“Corner.” The symbol modified is stopped but not exploded.

[ ]
“Subscript bridge.” The symbol modified is pronounced with a tongue-tip
placement on the back of the upper teeth – “dentalized”.

[ ˈbæt ]
“Vertical stroke (superior).” The syllable modified is given primary stress.

[ ˌbæt ]
“Vertical stroke (inferior).” The syllable modified is given secondary (but
still some) stress.

Here are the pause symbols:

[ ]
A short pause, roughly equivalent to a comma.

[ ]
A longer pause, roughly equivalent to a period.

Armed with this information, say each of these versions of this sentence. The
variations are more or less random, although the final version does tend
toward a particular accent. Do you have all the phonetic information here that
you need to define the accent clearly?

Audio track 70: Rumours – three versions.

Speak each sentence slowly and meticulously, letting the phonetic symbols
guide you. Then repeat it several times, slowly moving toward fluency
without sacrificing the flow of the specific sounds in each version.

Now we can try another pair of short sentences:

“What do you want? Did you settle on a pronunciation yet? And some
different ways to say them:9

Audio track 71: What do you want – five versions.

Notice the diacritics with which you are unfamiliar. Consult the list below
and try again.

New diacritics

[ æ̃ ]
“Tilde.” The symbol modified is pronounced with nasality, accomplished by
dropping the upper surface of the velum and letting the vibrations flow
through the nose as well as the mouth.

[ t̬ ]
“Subscript wedge.” The symbol modified is pronounced with some voicing.
This commonly occurs in American speech when an unvoiced stop-plosive –
(especially [ t ] – is said in an unstressed syllable, as in “pretty” [ pɹ̥ɪt̬i̞ ]. The
[ t ] becomes voiced almost – but not quite – to a full [ d ].

[ ɑ̽ ]
“Superscript cross.” The symbol (always a vowel) thus modified is
pronounced toward the mid-central part of the vowel quadrilateral. In other
words, it is relaxed toward the “schwa” [ ə ]. This tendency is extremely
common in American speech.

[ iː ]
“Length mark.” The sound of the symbol modified is lengthened. This can
happen with vowels and with continuant consonants. In print this symbol
consists of two filled-in wedges pointing at one another; in writing a colon
(two dots) will do just as well.

[ iˑ ]
“Half-length mark.” The symbol modified is lengthened slightly. As in [ wiˑ
sɔː ðə mæn ] – “we saw the man”.

A few other diacritics in common use for American English:

[ ɑ̈ ]
“Superscript umlaut.” The vowel symbol thus modified is centralized; that is,
moved toward the center line without being more open or close. Note the
difference from the mid-centralized vowel symbol [ ̊ ̽ ] which is relaxed
toward the schwa [ ə ].

[ ɑ̤ ]
“Subscript umlaut.” The voiced symbol thus modified is murmured; that is,
pronounced with a breathy voice quality. Any vowel or any voiced consonant
can be murmured.

[ ɔ̰ ]
“Subscript tilde.” The symbol thus modified is pronounced with a creaky
voice. This diacritic can be used to indicate “glottal fry” in a speaker,
although in the narrowest sense a “creaky” voice is not a tense sound at the
larynx, as the glottal fry certainly is.
[ ʊ̟ ]
“Subscript plus.” The vowel thus modified is placed further forward
(advanced) in tongue position.

[ t̺ ]
“Subscript turned bridge.” The symbol thus modified is an action enacted
with the apical portion of the tongue; that is, with the tip.

[ t̻ ]
“Subscript box.” The symbol thus modified is an action enacted with the
laminal portion of the tongue; that is, with the blade.

[ ɛ̠ ]
“Underbar.” The vowel thus modified is placed further back (backed) in
tongue position.

[ ʌ̘ ]
“Advancement sign.” The current IPA usage of this symbol (of which I am
not overly fond) is to indicate a pronunciation with an advanced tongue-root
(the tongue-root pushed forward). In the good old days it was used to indicate
a generally forward placement of the sound, for which purpose the “subscript
plus” is now employed. It would seem to me more logical to have the
subscript plus sign and underbar used to indicate tongue-root position. But
did they consult me? No.

[ ʌ̙ ]
“Retraction sign.” Currently used by the IPA to indicate tongue-root
retraction (the tongue-root pulled back toward the pharynx). For my
complaint about this usage, see above. Consistent tongue root retraction is
common in much American speech.

[ t̺ ]
“Subscript turned bridge.” The symbol thus modified uses the tip (or apical –
from “apex”) portion of the tongue to place the obstruent action at or near the
roof of the mouth.

[ t̻ ]
“Subscript box.” The symbol thus modified uses the blade (or lamina –
meaning thin plate) of the tongue to place the obstruent action at or near the
roof of the mouth.

Having collected the information that allows us to explore more “narrow”

(detailed) transcription, we can now explore it further in recording actual
ways that we might speak a text. In this case, it is a classical one. But it could
be contemporary as well.

Broad and narrow phonetics

hæmlɪt ənd hɛkjʊ̽bə sɛvɹəɫ tʰaɪ̆mz ]
Hamlet has just seen the Player King brought to tears while performing a
speech about Hecuba’s hysteria at the fall of Troy. Now alone, he says:

Audio track 72: O, what a rogue – broad version.

These lines have been written in fairly broad phonetic transcription. Fully
broad transcription would have nothing but the phonemes. To define
pronunciation somewhat, we used the “all-purpose r” that we discussed on
page 149 for prevocalic “r” sounds and we noted “r-coloring” as a diacritic
to the phoneme [ ǝ ]. We put the “breve” symbol [ ˘ ] over the weak
vowel sound in diphthongs, half-pause and full-pause marks which serve the
function of commas and periods, and we noted syllabic consonants such as [
ṇ ]. We wrote the words with no other modifications, giving full value to the
consonant and vowel sounds.
Writing this way has several advantages. For one thing, it’s
comparatively fast and easy to notate. It gives a clear general picture of how
the lines sound when spoken. But it also leaves out a lot of information that
might make your speaking of it more precise, and the pronunciations tend to
the generic. When you are reading it aloud (and you should always read
phonetic transcription aloud), if you are really reading the phonetic sounds as
written (make sure that you fully explode those stop-plosives!) and not just
translating, you may sound a little stilted, a little idealized. Obviously, even
in this broad transcription some decisions about pronunciation have been
made and, while they may help you to recognize the words, they may not be
the way you would prefer to pronounce them, or even the way you could
pronounce them comfortably, no matter how supple your articulators might
be.
And some decisions haven’t been made – exactly what words to put the
stress on, for one thing. Or what our actual speech actions are when we string
all these great words together. After all, how we say the words in isolation
can be very different from how we will say them in a phrase or sentence. So
let’s revisit the text and put some more details in. This time we’ll add stress
marks: [ ˈ ] for primary stress, and [ ˌ ] for secondary stress, coming before the
syllable they refer to. And no mark at all means that the syllable is
unstressed.
We’ll also change some of the symbols, such as [ r ] to [ ɹ ], to reflect a
more specific physical action (in this case, the all-purpose “r” turning to
become the voiced post-alveolar approximant). We’ll let some of the vowels
change because they occur in unstressed syllables. We’ll note aspiration after
unvoiced stop-plosives [ pʰɛzn̩t ]. We’ll also allow some of the vowel sounds
to lengthen [ ɑː ], and some consonants to become unvoiced [ ]. Some of the
consonants will be stopped but not exploded [ t ], or dentalized – said on the
teeth [ t̪ ] – or both.
All these changes are done in the interest of representing more accurately
the real speech actions that occur in these lines. Now remember that this will
not necessarily be an “ideal” reading, just a more specific one. Go through it
slowly, even though some of the changes will make more sense to you when
you do them at a normal rate of speed. After you feel all the sound actions
precisely, flow the phrase at a normal rate of speech.

Audio track 73: O, what a rogue – second version.

Remember, you still may not ordinarily pronounce the phrases the way they
are transcribed above, but it is definitely one way you could pronounce them.
This is a more narrow, or detailed, transcription. It has the disadvantage of
being slower – and a bit harder – to write; it has the advantage of giving you
a lot more information about the physical actions the speaker is taking in
articulating.
We are still being rather detailed also in our pronunciation. (And nothing
wrong with that, say I.) If, however, we were to let our speech actions
become a little more – how shall we say – fluid and keep our transcription
just as detailed, we might come up with something like this:

Audio track 74: O, what a rogue – third version.

Now what I have written is a more conversational version of essentially the
same accent as the versions that have preceded it. You’ll notice that a certain
amount of linguistic detail by the speaker has been left out, even though there
is actually more phonetic detail to the transcription. But we can go further
using only the vowel symbols that we have learned so far – and a couple of
other diacritics – and take the speech into another sound pattern entirely.
Again, go through the sounds slowly, until you are comfortable with all the
changes – and have figured out what all those diacritics mean – and then see
if you can read it aloud at a normal speed. See if you can experience where
all those sounds focus in the mouth: the pattern of tension and relaxation in
the articulators that makes this sound pattern inevitable for you when you
speak. That is the “feel” of the accent.

Audio track 75: O, what a rogue – fourth version.

What are the new diacritics used, and what do they signify? While we
wouldn’t need or for that matter want to use this kind of narrow transcription
all the time, it is precisely in this degree of specificity that the subtleties of a
person’s individual accent can be noted for ourselves and communicated to
others. Be aware that we have now notated some individual inconsistencies in
the way an individual phoneme might be spoken in different contexts. That is
what happens in real life when a person speaks. So now we have truly entered
the world of phonetics, not phonemics; we are becoming aware of changes in
speech that don’t just relate to units of meaning, but rather to the unique
sounds that are produced in a dialect or accent pattern.
We are also in a world where there are no longer simple “right answers”
for making a correct transcription. A good transcription in this context would
be extremely observant and detailed, and the physical actions observed would
be notated clearly and would be likely ones for a speaker to do from a
physiological standpoint. A poor transcription might have the articulators
taking actions that they wouldn’t or couldn’t manage. It might notate actions
that were inconsistent with the accent as a whole, or just write the symbols or
diacritics inaccurately. It might use one symbol when the writer meant
another.
But in a narrow phonetic transcription there may very well be more than
one way to express these physical actions. Phonetics ultimately is a skill, not
a formula. Or to put it in only slightly more grandiose terms, phonetics is an
art, not a science, though it uses aspects of scientific observation and
methodology. So our goal is always the most artful transcription, not just the
correct answer.
Naturally you don’t need the following transcription, but here it is anyway,
written in that amusingly imprecise orthography, English spelling:
O, what a rogue and peasant slave am I!
Is it not monstrous that this player here,
But in a fiction, in a dream of passion,
Could force his soul so to his own conceit
That from her working all the visage wann’d,
Tears in his eyes, distraction in his aspect,
A broken voice, an’ his whole function suiting
With forms to his conceit? And all for nothing,
For Hecuba!
(Hamlet, II, ii, 555-8)

One final note about the transcriptions. In the sixth line, I have taken the
reading of “distraction in’s aspect” letting the “in his” contract to one
syllable, and keeping the line ten syllables long. A lot of actors do it this way
(Olivier, for one). It is perfectly reasonable, though, to say the line as written,
using eleven syllables and letting the line end with a less stressed syllable, a
so-called “feminine ending”.

Clothes
I love this text. It’s great because Mark Twain is a great writer. It’s great
because it is rhetorically elegant. It’s great because it’s true. And, for our
purposes, it’s great because it has such a wonderful variety of sound use in
the English language. Speak on.

What would a man be, what would any man be, without his clothes? As soon
as one stops and thinks over that proposition, one realizes that without his
clothes a man would be nothing at all; that the clothes do not merely make
the man, the clothes are the man; that without them he is a cipher, a vacancy,
a nobody, a nothing.

Titles – another artificiality – are a part of his clothing. They and the dry-
goods conceal the wearer’s inferiority and make him seem great and a
wonder, when at bottom there is nothing remarkable about him. They can
move a nation to fall on its knees and sincerely worship a man who, without
the clothes and the title, would drop to the rank of a cobbler and be
swallowed up and lost sight of in the massed multitude of the
inconsequentials....

A policeman in plain clothes is one man; in his uniform he is ten. Clothes and
the title are the most potent thing, the most formidable influence, in the earth.
They move the human race to willing and spontaneous respect for the judge,
the general, the admiral, the bishop, the ambassador, the frivolous earl, the
idiot duke, the sultan, the king, the emperor. No great title is efficient without
clothes to support it.

Is the human race a joke? Was it devised and patched together in a dull time
when there was nothing important to do? Has it not respect for itself? I think
my respect for it is drooping, sinking – and my respect for myself along with
it. There is but one restorative – clothes! Respect-reviving, spirit uplifting
clothes! Heaven’s kindliest gift to man, his only protection against finding
himself out: they deceive him, they confer dignity upon him; without them he
has none. How charitable are clothes, how beneficent, how puissant, how
inestimably precious! Mine are able to expand a human cipher into a globe-
shadowing portent; they can command the respect of the whole world –
including my own, which is fading. I will put them on.
Mark Twain, ‘The Czar’s Soliloquy’
Remember: Don’t translate. Do the articulatory actions and feel the sounds.
Audio track 76: Clothes.
Part 4
The skills of intelligibilty
23 Formal and informal speech

This section of the book moves across the line that I have very purposely set
up between descriptive and prescriptive learning, between the information
that describes what connected speech actually is and the information that
mandates what (under defined circumstances) speech actions should be. But
all the skills that you have learned until now will be of crucial importance
here and will make the acquisition of any prescriptive pattern easy.
We know – because we do it all the time in real life – that we constantly
use different strategies to vary our communication with others, depending on
whom we are talking to and the demands of the occasion. So our connected
speech never gets stuck in just one set of techniques for stringing sounds
together in sequence to form phrases and sentences.
But there is a difference between the physicality of formal and informal
speech. We use more formal speech, sometimes instinctively, when we need
to communicate to people who might not understand us. They might not
share our accent, or they might not hear too well, or they might be listening to
us in a challenging acoustic environment, such as a theatre or a lecture hall.
Or it might be the communication of very complex text to a listener (although
complete formality of speech may not be the best stratagem for this). Or it
might be a challenging situational requirement, such as an airplane pilot who
does not speak English as a first language communicating with the control
tower while landing a plane, or a doctor from another country giving the
specifics of a prescription to a nurse in an American hospital – both high-
stakes environments for intelligibility.
We use informal or conversational speech when we are speaking fluently
to other people with whom we feel comfortable in environments where there
are not undue acoustic or textual challenges and where we do not need to
impress. The performance of many theatrical works would fall into this
category, too. Even in Shakespeare the actor’s main goal should not be to
impress the audience with one’s speech skills; the objective is to make the
characters live on stage and to share their life with the audience.
There is another consideration that in a perfect world wouldn’t exist, but
in our real world does. This is the element of social prestige. For many
generations there has existed a social bias in favor of some degree of formal
speech. It has diminished considerably in the last fifty years, but I, alas, am of
an advanced age that remembers when it was in full flower. During my time
in grammar school I well recall being corrected constantly because my
speech was considered substandard, especially when, at the age of six, I
arrived at school in New England from New Orleans. And many of my
fellow students, some of them native New Englanders, suffered this social
stigma far more and far longer than I did.
For generations, in America, prescriptive speech training was a major part
of elementary, secondary, and college education, although that is no longer
true. Stigma in speech pattern today is much more about whether you are
speaking in an accent that is easily accepted by the people around you, but
prestige based on class is still an active ingredient.
There are many good reasons for learning the skills of formal American
English speech. But social prestige is not one of them.

Formal speech actions

You know all of this already – in fact, you have done some of it. Our work
here consists in focusing our attention on our arrangement of these sequential
speech actions.

1 Try this experiment. Say the following sequence of phonemes as wholly

separate actions, with an in-breath between each one. For the vowels,
naturally, say the correct phoneme, even though there may be some variation
among speakers. For the stop-plosives and the affricates, follow the
consonant with a vowel. Use your entire breath flow, unvoiced or voiced, on
each consonant and vowel action and be very aware of the feel of forming,
feeling and then hearing each sound, whether it is unvoiced or voiced. With
the stop-plosives and affricates, make sure that you feel the stoppage and also
the explosion. Keep your jaw dropped comfortably open on all phonemes
except the phonemes / s /, / z /, / ʃ /, / ʒ / and the combined affricate
phonemes / tʃ/ / and / dʒ /. This is a lot to think about, so do it slowly.

/b/ /ə/ /t/ /ɪ/ /f/ /j/ /u/ /m/ / aʊ̆ / /e

/ /ð/ /ɪ/ /t/ /æ/ /z/ /m/ /ɛ/ /n/ /i/ /ə
/ /v/ /j/ /ʊ / /p/ /l/ / eɪ̆ / /ɝ/ /z/ /d/ /u
 / / aɪ̆ / /h/ /æ/ /d/ /æ/ /z/ /l/ /i/ /f/ /d
/ /ð/ /t/ / aʊ̆ / /n/ /k/ /r/ / aɪ̆ / /s/ /p/ /
oʊ̆ / /k/ /m/ / aɪ̆ / /l/ / aɪ̆ / /n/ /z/

2 Now do several successive phonemes on one breath. Do this very slowly,

extending all the phonemes except the stop-plosives and affricates, which no
longer need that added vowel. Breathe where you need to; even if you begin
to sense words forming sometimes, do not feel that you have to breathe at
word endings. Keep each phoneme fully formed, as before.

3 Keeping the phonemes fully formed, as before, do more phonemes on one

breath. Make sure that you aren’t changing any of the phoneme actions. By
this point you surely will have started to perceive larger linguistic units
emerging from the sounds – words and even phrases. Feel free to breathe
after the ends of words or phrases.

4 If you can still keep each of these phonemic actions formed fully in exactly
the same way that you did before, say the entire sentence in two breaths. By
now, you may be very aware that you are speaking a sentence from Hamlet’s
advice to the players: “But if you mouth it as many of your players do, I had
as lief the town crier spoke my lines” (Hamlet, III, ii).

5 See if you can say the sentence at normal playing speed while still retaining
the unaltered identity of each of those phonemes, even in connected speech.
This is not easy to enact, and the results may occasionally sound stilted or
strange, but do not compromise. Because I have asked you to speak
phonemes, there may be some variation in the actions you take for them in
some of the consonants and definitely with the vowels and diphthongs – and
that one triphthong. For this go-through, that is all right.

Congratulations: you have just produced an extremely instructive parody of

formal speech. Even the most rigorous and doctrinaire speech teacher in
creation today would not ask you to speak exactly like this, although in years
long gone by voice and speech teachers, then called elocutionists (along with
their pronunciation-specialist cousins, orthoepists) came close.
 What sounds odd about this speech pattern? Be careful that your answer
is not “It doesn’t sound like the way I talk.” After all, the person next to you
in class, doing the previous exercise, may not have sounded quite the same as
you. No, the thing that sounds odd about it is that the individual phonemes,
spoken in more and more closely connected speech, have not altered
themselves to allow easy movement of the articulators from one action to
another, so the flow of connected speech sounds stilted and over-articulated.

What is useful about the speech pattern you just enacted? The most obvious
thing is that everyone can understand easily what you are saying, even if the
articulation seems stiff, odd, and lacking in individual character.
Our imaginary “doctrinaire speech teacher” would have yet another
perspective about the oddity and the utility of the speech product in the
preceding exercise. He or she might allow the student to use altered speech
actions to carry her or him with some fluency from one sound to another
(within limitations), thus lessening the oddity, but that same teacher would
want the student also to eliminate any differentiation within each consonant
and vowel phoneme so that each articulation conformed to a required
standard that is set by the speech teacher.
The practical problem with this view is that, as long as each speech sound
as spoken by real speakers remains within the realm of generally understood
phonemes, there is no real problem for the listener in understanding what the
speaker is saying. So if intelligibility is the main objective, minor differences
within the phoneme – the kinds of things we find in regional accents all the
time – do not really matter as long as the appropriate phonemic detail is
communicated to the listener by some means.

We can conclude that differentiation between phonemes and the

communication of the linguistic information contained within each of them
(by some means) is the main characteristic of formal speech actions. Try
speaking the sentence you used before in a larger portion of Hamlet’s speech.
Keep the phonemes fully articulated. Do it twice – the first time following the
standards that we established before. Then do it again and try to get all the
phonemic information in, but allow yourself to move comfortably from one
sound to another. (Note: there is no exactly right way to do this. Experiment.)
Speak the speech, I pray you, as I pronounced it to you, trippingly on
 the tongue. But if you mouth it as many of your players do, I had as lief
the town crier spoke my lines. Nor do not saw the air too much with
your hand thus, but use all gently; for in the very torrent, tempest, and,
as I may say, whirlwind of your passion, you must acquire and beget a
temperance that may give it smoothness.
(Hamlet, III, ii)

Your second experiment shows the feel and sound of formal speech moving
in the direction of fluency, and therefore moving in the direction of informal
or conversational speech. If you carry the physical strategies that you use in
fluent formal speech further, your speech will start to become much less
formal. Eventually, using the same strategies, you will start to lose phonemic
detail; and your intelligibility may suffer in some environments.

Fluency strategies
You should not be misled by the binary distinction of “formal/informal”.
Most speech in real life, as well as onstage or in front of the camera, lives
somewhere between these extremes, negotiating between them: phonemic
detail on the one hand and fluency on the other. Both are desirable in any
speech interaction with a listener: we must be understood, but we also must
be able to achieve this intelligibility with minimal effort.
You have already experienced both the clarity and also the awkwardness
of excessively formal speech and have instinctively made some
accommodation to fluency. But what actually are you doing when you move
toward fluency? There are several ways that we change our speech actions to
allow our articulators to move easily from one phoneme to another. The two
categories that govern most such actions are called coarticulation and
assimilation.

Coarticulation
In learning the pronunciation of the phonetic symbols we encountered several
interesting examples of coarticulation, by which we mean forming a unique
sound product in consonants by simultaneously obstructing the flow at two
places in the vocal tract rather than one. Sometimes we accomplish this
simultaneous action by using two different articulators; sometimes we do it
by using the amazing versatility of the tongue muscles. One example of using
different articulators is / ɥ /, which is described by the International Phonetic
Association as a “voiced labial-palatal approximant”, formed by producing a
palatal approximant / j / and also rounding the lips to produce a simultaneous
/ w /; it is used in French in words like huit or lui. Another more challenging
one, using only the tongue, is the endangered “hooktop heng”, endangered
because it is found only in one dialect of Swedish and the speakers of that
dialect have – for the most part – given it up (not surprisingly, once you have
tried to say it). It is formed by producing a / ʃ / and a / x / simultaneously, as
you may recall from our study of the IPA. In African languages, we find the
simultaneous enactment of / k / and / p / to form the combination / p / and
there are several other such stop-plosive combinations.
English has such simultaneous actions as well, but sometimes they are
hidden because we are so used to the sound products of the language. It may
come as a surprise to us that the / w / is considered by the Association to be a
“labial-velar” coarticulation. We are quite used to the “labial” part of it – with
some useful lip corner advancement – but the “velar” description may cause
some question. It certainly does to me.
The lateral approximant / l / is a better example of a coarticulation. The
IPA chart describes it simply as an “alveolar lateral approximant”, with no
other specification. However, if we consider the muscular action of the
tongue in keeping the side edges well down to direct the flow laterally around
the tip and blade, while the tip is raised to contact the alveolar ridge, we will
sense – with our keen knowledge of the sense and position of the entire
tongue – that the dorsum of the tongue is thickened and its superior surface is
raised as a natural muscular response to the curling action of the tongue tip
and blade. So the middle of the tongue is thickened or arched toward the
velum in a coarticulation that is described as “velarization” [ lˠ ]. The
assumption, then, by most phoneticians is that is any / l / is considered to
have some degree of velarization or arching of the middle of the tongue
toward the velum, in addition to the contact by the tip of the tongue to the
alveolar ridge. But in much American and also in much UK pronunciation of
English, there is no contact of the tip of the tongue to the alveolar ridge in the
so-called postvocalic / l / positions. What we are left with, in fact, is the
arching of the middle portion of the dorsum of the tongue to an approximant
version of the voiced velar fricative; as a symbol it is [ lˠ ]. An even more
adventurous way of representing it phonetically would be as the voiced velar
approximant / ɰ /. In accents where the tongue root is retracted, such as
Cockney, the action becomes uvular; since there is no symbol to indicate
“uvulization”, the nearest we can come is to honor the tongue retraction by
writing – in narrow transcription – the symbol as [ lˁ ].
The phonetic realization of the phoneme / l / for final “l” has been a
source of controversy among phoneticians for years. Some have suggested –
wrongly – that the final “l” is always simply retroflex [ ɭ ]. Others have
suggested that it becomes the voiced uvular fricative [ ʁ ]. Still others have
suggested that it becomes a vowel, such as / oː /.
As you explore the less formal speaking of text, you will find lots of
examples of coarticulation. Because this is a book for actors and speakers,
not for phonetics students, we need not explicate all the possible varieties, but
informal coarticulation does have two basic sub-categories that deserve some
brief mention.
Prearticulation (the linguistic term is “anticipatory coarticulation”). This
is the practice of moving an action from one point of articulation to another,
so that you can say the next phoneme more easily in connected speech.
Prearticulation would be used if we said the word “income” as [ ɪŋkʌm ] or
“utmost” as [ ʌp moʊ̆st ]. When we say “obvious” as [ ɑb̪ viʲ.əs ] with a
dentalized “b”, we are stopping the / b / as a labio-dental action to make the
transition to the labio-dental / v / easier.
We do postarticulations too: actions where a consonant action changes
in the second phoneme of a consonant cluster because of the placement of the
one just before it. When we say “agree” in American speech, we are almost
certainly using the “braced” or “molar” r (for which there is no symbol in the
IPA so we will settle for the phoneme [ r ]) before the / i /, not because of the
vowel that follows, not because it needs to be there but because the preceding
consonant [ g ] guides the approximant that follows into a placement that is
enacted near it with the same portion of the tongue. Postarticulation happens
more commonly with the phoneme / r / than with most other consonants.
We do use these coarticulations across phonemes all the time, especially
in consonant combinations, because they contribute to fluency without
seriously compromising intelligibility, except in those fairly rare cases – rare
even onstage – where each and every consonant phoneme must be made
completely clear by differentiating that action totally from every other action.
We can do allophonic prearticulations from consonant to vowel as well.
When we say “at all” we are probably doing a somewhat more retroflex
version of the “t” than we might use in “at ease”, in order to prepare the vocal
tract for the easy movement to the back vowel of “all”, whether that vowel is
[ ɔ ] or [ ɑ ] in American speech. Or consider the difference in lip position on
the “s” in the following words: “seep, sap, sop, soap, soup”. For most
speakers, the lips move from fully unrounded progressively to fully rounded.
This doesn’t have a great effect on the identity of the phoneme (we still know
well that it is an “s”) but it illustrates the physical action of pre-articulation
effectively because the lips have to do what they have to do.

Assimilation
Assimilation is the monarch of informal speech. Most of the accommodations
to efficiency of enactment that we make in informal speech can be included
in this generous category. It is the practice of contracting one or more sound
actions into another simpler action. We use assimilation all the time and in
various forms. A very simple but good example is what we do with
individual consonants when we say the two “t” phonemes across two
consecutive words in “at that”. We commonly stop the first / t / on the back
of the upper teeth (phonetically [ t̪ ]) rather than on the alveolar ridge and
then we release the voiced dental fricative / ð / from the same place, so that
the two consonants move efficiently from one to the other. It isn’t quite as
linguistically detailed as exploding the “t”, but this still conveys the linguistic
information of both the phonemes, even though the first phoneme is stopped
but not exploded and is also enacted on the back of the upper teeth rather than
on the alveolar ridge, the better to release the / ð /. There is an obvious
resemblance to the previous category, coarticulation, but it is slightly
different because it crosses word boundaries.
We do a similarly simple but not identical action when we divide the
stoppage and the plosion of the / t / over two “t” sounds in a row when we
say “at times”. While it does not involve a change of placement, it does
involve the division of a compound two-part action, the stop and the plosive,
to convey the linguistic information for each; one “t” is stopped, the other
exploded. We do this a lot in English as a practical sequence of actions with
“t” sounds in successive words, but we do not do it in doubled consonant
words such as “butter”. In Italian, though, it is formalized in the characteristic
sound of doubled consonants such as the doubled [ t ] in “stretto”, where the
first “t” is stopped and the second exploded (the linguistic term is
“geminate”, from “Gemini”, the twins). And “stretto”, by the way, is a good
description for many assimilated sounds, since it means “compressed or
shortened”.
Assimilation can also involve the loss of phonemes in consonant clusters;
sometimes the loss of combined consonant and vowel or diphthong
combinations; and sometimes – in very conversational speech – in the loss of
one or more entire syllables or even clauses. It can also involve changing one
or more consonants to other ones.
A phonetics text would go through these in considerable detail. But for
our purposes here, a few illustrations will suffice.

Yod coalescence. The “yod” is the name for the voiced palatal approximant /
j / as in “yes” or “beautiful”. When we say an alveolar stop-plosive / t / or / d
/ in a final position in a word and it is followed by a / j / in an initial position
in the next word, as in “that you” or “would you”, it is common to let the / t /
or / d / expand back from the tip to include the blade (laminal portion) of the
tongue forming the post-alveolar fricative / ʃ / or / ʒ /, thus producing in
sequence the unvoiced affricates / tʃ / or / dʒ /: “that you” [ ðæt ju ] turns
into [ ætʃu ] and [ wʊd ju ] turns into [ wʊdʒu ] or sometimes [ wʊdʒ ].
This assimilation is very common in American English speech, which is
fairly lax in all its accents when compared to the relative linguistic tenseness
of Standard British English or RP. But even RP can do the same (or a similar)
assimilation when it and American English speakers who employ the so-
called “liquid u” as in “tune” [ tçun ], “Tuesday” [ tçuzdi ], “duke” [ djuk ],
or “deuce” [ djus ] assimilate it to [ tʃun ], [ tʃuzdi ], [ dʒuk ], or [ dʒus ].
The assimilation could proceed further with the assimilation of “What do you
want?” to [ wəʃəwɑ̃ʔ ]. This phenomenon, also called “palatalization”,
occurs across word with a final / s / or / z / going into an initial / j /. “Bless
you” becomes [ blɛʃu ], “curse you” becomes [ kʰɝːʃu ], and “please you”
becomes [ pl̥iʒu ].

Eth / ð / dropping. It is very common, in informal American speech, to

change the phoneme / ð / to / d / or more narrowly [ d̪ ]. So “with daring”
could change to [ wɪˈd dɛ ɪŋ ] and “but this” could change to [ bət dɪs ].
There is an element of coarticulation in this assimilation change also: these
two main categories are not totally distinct from one another.

Schwa / ə / absorption. This is a term originated by J. C. Wells. We have

already encountered the phenomenon in a context that is used even in most
formal speech: the syllabic consonant, where “even” [ ivən ] becomes [ ivn̩ ]
or with some prearticulation and nasal plosion added [ ibⁿm̩ ]. Its most
extreme informality could be [ iːm̩ ]. In an initial position in a connected
speech word, the schwa can be absorbed into a syllabic / l / in the famous
quote from Dr. Frankenstein “It’s alive!” This can be realized as [ ɪtsl̩aɪ̆v ]. If
the doctor is really in a hurry to examine his monster, he might even say [
sl̩aɪ̆v ]. In a medial unstressed syllable, as in “innocent”, the schwa would be
absorbed to [ ˈɪnsn̩t ]

Glottal / ʔ / substitution. The most common glottal substitution is to replace

“t” in a medial or final position. This substitution is sometimes enacted as a
complete replacement, as often occurs in American speech in a final stopped
position: “What?” is realized as [ ʍʌʔ ] or [ ʍʌʔ ]. (Naturally, you haven’t
confused the glottal symbol [ ʔ ] with the question mark, have you?)

The glottal [ ʔ ] can also be coarticulated with the stop-plosive in a final

stopped (but not exploded) position. It occurs with / t / but even more
commonly with / p / or / k / as in “sip” [ sɪp ʔ ] or “take” [ tʰeɪ̆k ʔ ]. While
any of these words could be enacted with only the stopped (unreleased) stop-
plosive, and, while they would sound almost exactly the same as in the
examples above, the “double-stopping” is characteristic of accents in English
that exhibit tongue retraction, such as Cockney or many American accents.

Consonant cluster assimilation. A very common action in informal

American speech is to eliminate final – or even medial – “t” sounds when
they are preceded or followed by another consonant, frequently by / s / or / n
/. “That’s all” becomes [ dæsɔɫ ]. “Most interesting” becomes [ ˌmoʊ̆sˈinə
ˌɹəstɪŋ ]. “Her fasts are costly” might be assimilated to [ hɝfæsɹ̩kʰɑsli ].
Successions of stop-plosives can undergo assimilation: “beasts” becomes [
bis ], tastes” becomes [ tʰeɪ̆s ]. / d / is often dropped after / n /. “Band” could
either be enacted as [ bæn ] or even as [ bæ̃ ], especially if the word is in a
final position in a phrase.

Nasalization. Often informal American speech we find / nt / or other post-

vocalic nasal consonants or combinations are assimilated into a nasalization
of the vowel, often where the word ends in an unreleased stop. “I can’t go”
can be spoken as [ aɪ̆ kʰæ̃ɡ ɡoʊ̆ ]. “Advance” could be [ dvæ̃s ]. “Vaunt”
could be [ vɔ̃ʔ ], “sink” could be [ sɪ̃k ], “limp” could be [ ].

/ h / dropping. In connected speech we drop the / h / very frequently. Earlier

in this book I explained my objection to the International Phonetic
Association’s recipe for this symbol as an “unvoiced glottal fricative”. In
English, at least, we can discover with very little experimentation that few if
any people actually produce this noise at the glottis. Instead it is a
preaspiration of the vowel that follows, its distinguishing physical feature
being an increase in pulmonic airflow. I prefer to call it an unvoiced
aspirant. This increase in airflow means that, in addition to being always
prevocalic, the / h / must always begin a syllable that is either fully or
partially stressed. If the airflow decreases, as it always does in an unstressed
syllable, / h / is going to be dropped unless the speaker makes a special effort
to say it. So “Did he leave?” spoken informally usually becomes [ ˈdɪdiˈliv ].
“Take him away” would become [ ˈtʰeɪkm̩əˈweɪ̆ ]. “Leave her to heaven”
might be spoken as [ ˈlivɝt̬əhɛvn̩ ] if the Ghost in Hamlet was in an
unusually nonchalant mood. Note that the / h / remains in “heaven” because
the syllable is stressed.

/t / voicing. This is a phenomenon that is very characteristic of American

speech in most regional and social dialects. When / t / appears as the first
letter in a prevocalic position in an unstressed syllable, as in the words
“pretty”, “city”, or “gratitude”, American speakers give voicing to this
unvoiced alveolar stop-plosive – to some degree, at least. Some British
commentators consider that Americans voice the / t / fully and substitute / d /
as though American speakers were saying “sadisfy” or “naughdy”. John
Honey, makes this point forcefully:
Some features of American English … deserve to be resisted by
anyone concerned for intelligibility and clear communication. One
of these is the American treatment of the “t” sound; … If it appears
between vowels, as in “letter” and “city” it becomes almost a “d”,
making it difficult to distinguish between writer and rider, bidden
and bitten.
(John Honey, Language is Power, page 248)

There is some evidence for this, since Americans often rhyme the two
spellings, as in the banner I saw recently in a printing shop proclaiming
“Lordy, Lordy, Al’s turned forty.” Or the Hooter’s employee (and how do
you pronounce the “t” in the restaurant title?) who successfully sued her boss
for emotional injury because she had won a sales contest that promised, she
thought, a “new Toyota” as a prize. Instead she was presented with a small
figure of a sage character from Star Wars: the Empire Strikes Back. But if
Americans truly substitute the phoneme “d” for the phoneme “t” in the way
Honey suggests, there would be much more confusion even among
Americans about this pronunciation ambiguity, and there isn’t. Clearly
something else is going on that may escape the ears of some British listeners
– and some Americans too.
Many phoneticians suggest that the adjustment of the “t” moves to an
alveolar tap / ɾ / with American speakers, so that “dirty” would become [
dɝɾi̞ ]. Aside from the curious juxtaposition of two allophones of / r / in this
word, here are reasons to consider that this may not be completely accurate,
though it certainly is better than the substitution surmises. A small – but
growing – number of phoneticians, with whom I agree, consider that a better
transcription of the voicing action on “t” would be a / t / with the subscript
wedge voicing diacritic, indicating that the “t” is partially voiced: [ t̬ ] as in
“today” [ t̬ədeɪ̆ ], “Saturday” [ sæt̬ɝdɛɪ̆ ], “Fruit of the Loom” [ fɹut̬ ə ðə
lum ], “fatty” [ fæt̬i ], or “etymological” [ ɛt̬əməlɑdʒɪkl̩ ]. I submit
cautiously that Americans may be able to distinguish between “bidden” and
“bitten” based on the difference between semi-voicing and full-voicing.

Contractions. Many of the preceding verbal actions are, in effect,

contractions: enfoldings of speech actions one into another in which both
consonants and vowels can be dropped. Their purpose is to provide enough
linguistic information to the listener while, at the same time – and in the
interest of fluency – not overdoing the action. There are a few other
contractions that we have ignored until now only because we know them so
well. They have a long (see Shakespeare) and noble history even in the most
elevated of English pronunciation. Though contractions are not limited to
them, two unstressed or “weak form” words in succession provide a
particularly healthy environment for contractions. We know them well:
“I’ve” for “I have”, as in the phrase “I have a plan”, where “have” and “a”
are both unstressed. “you’re” for “you are”, “I’m” for “I am”, “she’s” for
“she is”, “what’s” for “what is”, “they’ve” for “they have”, “that’ll” for “that
will”, “isn’t” for “is not” and so on and so on. So useful are these
contractions in all but the most formal speech that most of them are used
even if the first word in the pair is stressed within a phrase. I’ve used them
frequently in this text to maintain its informal, though hopefully informative,
tone.

It is incorrect to characterize informal speech as “lazy” or “sloppy”. On the

contrary, it is efficient; conforming to the observation that speakers in
informal conversational settings instinctively employ minimum (but
sufficient) effort to maintain intelligibility, unless they have some overriding
reason to conceal their meaning. But minimum effort is not an absolute; the
effective minimum for one speech environment might be inappropriate for
another. We can also see that there are many different articulatory strategies
that speakers can use to be understood always by a listener. It certainly is an
error to assume that for all purposes on stage, even within an individual play,
an actor has to set a single unvarying level of speech action. The real skill of
articulation is learning to sense the kinds of actions that are appropriate to the
complex, moment-to-moment conditions of performance and to execute them
fully and efficiently.

Formal and informal

Initial practice: back to Hamlet

Have a partner for this exercise. One of you will be the formal speaker and
one the informal speaker. The formal speaker, using the criteria for formal
speech (even though we have not yet examined speech skills in detail) will
speak the passage pronouncing every phoneme fully, even if this feels
excessively articulated. Then the informal speaker will speak the same
passage, using as little effort in articulation as possible while maintaining
minimal intelligibility (and keeping the volume up).
O, it offends me to the soul to hear a robustious periwig-pated fellow
tear a passion to tatters, to very rags, to split the ears of the
groundlings, who for the most part are capable of nothing but
inexplicable dumb shows and noise. I would have such a fellow
whipped for o’erdoing Termigant. It out-herods Herod. Pray you,
avoid it.
(Hamlet, III,ii)

Discuss what the speech strategies were that each of you used. What
information can you trade about intelligibility or appropriate articulatory
action?

The formal speaker should do the speech again, moderating the pronunciation
of the text toward less formality, but maintaining as much formality as will
make every phoneme very intelligible to the listener. The informal speaker
should explore putting in the necessary phonemic detail – but as little as
possible – to make the reading not only intelligible, but comfortably so.
Switch roles. Go through the same process, except that this double run-
through will be informed by the last one.

This is our first pass at the formal–informal continuum. When we have

completed our exercise work on the speech skills of consonant and vowel
action we shall try it again, even if you were brilliant this time.
24 Consonant skills

You should be aware by now that the point of view of this book is that formal
speech is not intrinsically superior to informal speech, as long as the informal
speech is completely intelligible to the listener. However, we have to
consider the reality that informal connected speech is a skill in which most
American speakers, and most American actors, have a lot more expertise and
practice. Since a key premise of this book is that actors can use everything, it
is important that we spend some time exploring in greater detail phonemic
(i.e. meaning) differentiation through phonetic (i.e. specific and fluent)
means. There are real benefits to the actor in exploring the specific actions
that provide linguistic detail in our speech.
We already know that complete formality sounds odd because it usually
involves more linguistic detail than we as listeners might need. There are
other messages that might be contained within such strict formality: the
implication that the listener is hard of hearing, or disinclined to listen, or
simply stupid. These implied messages limit the fluidity of verbal
communication and especially the possibilities for characterization by actors.
But formalism does have its positive uses in actions skills. While we
would never go to the extremes of formality (except when performing in an
outdoor theatre with terrible acoustics, a high wind, and a train going by) we
need to develop our skills in moving toward formality, if only because many
of us have never been there. Speech informality is something that we all are
really good at already, whereas the skills of speech formality require actions
that many students have never explored.
It is a generally accepted precept in language teaching that listeners
unfamiliar with the words they are hearing will comprehend better if the
words are spoken with a high level of linguistic detail, and with all the
phonemes, consonant and vowel, highly differentiated. As language students
become more familiar with the vocabulary and grammar, some degree of
informal speech strategies can be introduced by the teacher in the interest of
fluency. But even a diligent student, newly immersed in the language
environment of a given country, may have great difficulty in understanding
native speakers of the language when they are conversing with one another.
In the same way, speakers of a common language with different dialect
backgrounds will find that they all understand a speaker better if a greater
degree of linguistic detail is employed. The actor who is speaking complex
text to a diverse audience will find that its group comprehension is usually
bettered by judicious increase in the detail of certain speech actions,
especially if the acoustic demands of performance are also more challenging.
There is a pitfall in focusing exclusively on formal skills, however
separately they may be taught. If all our practice moves inexorably in the
direction of formality and increased activity, we can lock ourselves into an
inflexible pattern of formality, even if we are trying to keep the skills separate
from one another. So it is crucial to practice all the word lists or sentences in
two directions – toward detail and then toward informality. By practicing
these skills on this kind of continuum, we will have the added benefit of
honing our perceptual skills so that we know when a particular set of sounds
is or is not communicating effectively to a listener. Eventually this
knowledge becomes instinctive, so that an actor does not have to think about
any sort of skills set, but feels instead – in every moment – the unique way in
which her character wants to communicate with the other characters onstage
and with the audience too.

Specific skills

1 All consonants except [ s ], [ z ], [ ʃ ], [ ʒ ], [ tʃ ], and [ dʒ ] in American

English can be formed with the jaw remaining relaxed open, and with greater
muscularity of the other articulators. This is not a new skill for you as we
have been exploring this convention throughout our phonetic work.
Furthermore it is a natural posture for most American accents. But it also
makes good sense in performance. If the jaw is relaxed open, there is more
room in the oral cavity for the sound to come from. Getting the jaw out of the
formula for articulation allows us to be more varied and specific in our use of
the other articulators. For many accents, variation of jaw position will be very
important but, for now, keep it relaxed open.
For the six consonants listed above, the jaw needs to be raised and
brought forward very slightly toward the upper front teeth. Relax the jaw
open again immediately after the sound is completed, so that the relaxed jaw
remains “home base” for articulation. Speak the following words and
sentences slowly and then at normal pace, being very careful to raise and then
release the jaw back to home base as needed. Monitor your jaw with your
fingers lightly at the sides of your mandible if you wish.
situate, fashion, Sicily, cashew, seizes, judges, notion, satchel, nausea,
versus, contusion, Atchison, contagion, seizure, vicissitude, ration,
sideshows, church socials, jerseys, bastion, Xerxes, prestidigitation
She casually spoke with the teacher. They chided the speaker for his
foolish notions. Imagine a scenic design with no visual elements on
stage. The judge showed his agitation.
Audio track 77: Jaw action.
Repeat. But this time, explore what happens if you keep the jaw raised
toward the upper front teeth all the time. What works better and what works
worse?

2 Just to see how it feels, try pronouncing all the consonants in the following
word list with full, precise action. Do this very slowly, with a lot of attention
to the complete enactment of each phoneme; then repeat it at normal speed.
Be wary of following the spelling slavishly: there are spelled compound
consonant sounds but there are no silent letters in the words.

plinths, drifts, crusts, groped, lengths, schlepped, thanked, plagues,

clucked, bragged, cthonic, angst, clapped, shelves, dephlogisticated,
marsupial, bloviate
Audio track 78: Pronouncing all the consonants.
The internists inadvertently performed a lobotomy instead of a
phlebotomy. Rustication precludes professional diligence. The cleric
attempted to placate his infuriated parishioners.

3 Explore making voiced/unvoiced distinctions in the following consonants:

[ ʍ ] – [ w ], [ tʃ ] -[ dʒ ], [ s ] -[ z ], [ t ] -[ d ], [ θ ] – [ ð ]
Carry this through to the following word lists:

[ ʍ ] – [ w ] which witch, when wen, why Wye, whether weather, whale

wail, whit wit, whirred word
 [ tʃ ] – [ dʒ ] charred jarred, choke joke, chain Jane, chaw jaw, etch
edge, larch lodge
[ s ] – [ z ] seal zeal, sit zit, sue zoo, peace peas, Miss Miz, Fess fez,
cuss cuz, eased east, cause slides, slays sleeves, close close, picks pigs,
drags sacks
[ t ] – [ d ] feat feed, meat mead, rote rode, suit sued, patty paddy,
Pinter Pindar, pant panned, faint feigned, post posed, petty Pedi,
mottle model, betting bedding
[ θ ] – [ ð ] ether either, teeth teethe, sooth soothe, cloth clothe, this’ll
thistle, Theseus seethes, breathe breath, wreathe wreaths
Why worry when we wonder which way we wander? The judge
charged the jury in chambers. He’s ceaselessly espoused his causes so
zealously. Matty petted her pretty parakeet and tried to distract
attention from her pouty “puddy tat”. Think this uncouth threat
through, and then breathe methodically.
Audio track 79: Unvoiced/voiced distinctions.
As you can tell easily, the distinction between unvoiced and voiced
phonemes changes the meaning of the words in the word pairs completely.
Speak the word pairs with full attention to a maximum distinction between
unvoiced and voiced. Then say each sentence using the same maximum
unvoiced/voiced distinction. Say each sentence four more times, gradually
moving all the paired consonant phonemes toward full voicing. Then speak
the sequence again, gradually moving the paired consonant phonemes toward
full unvoicing; the last exploration may sound strange to you.
We know, from our discussion of informality, that / t / semi-voicing to / t
/ voicing is the most common voicing in American speech. It deserves some
extra consideration.
Voiced/unvoiced challenge. The British novelist Martin Amis wrote a story
in which the main character, a young boy, is speaking “in sarcastic
Americanese” about a tragic event. Can you read the passage exactly as
written? It requires voicings of all unvoiced consonants, including those that
Americans never voice. Because it assumes that Americans voice all
unvoiced consonants, not just “t” sounds, it is “sarcastic”. But it is also
informative about the British perception of American speech.
Note that the “th” sounds are the only written letters for which there is no
orthographic way of indicating whether the phoneme is unvoiced or voiced. It
could be / θ / or / ð /. But in context, it is certainly / ð /. When Amis knows
that a written / s / would be pronounced in a particular word as / z /, he
doesn’t change the spelling. Note also that Amis changes the spelling of the
“short o” sounds, represented in Standard English (or “Received
Pronunciation”) as / ɒ /, to “a”, even though to an American that might
suggest a more fronted vowel.
Try to read the passage exactly as written. Then, translate the text back
into the way Americans actually speak, noting now the feel of the unvoicings
or voicings define the meaning.

A derrible thing habbened do me on my haliday. A horrible thing, and a

bermanend thing. I won’d be the zame, ever again.
Bud virzd I’d bedder zay: don’d banig! I’m nad zuvvering vram brain
damage – or vram adenoidz. And I gan wride bedder than thiz when I
wand do. Bud I don’d wand do. Nod vor now. Led me egsblain.
I am halve Englizh and halve Amerigan. My mum is Amerigan and
my dad is Englizh. I go do zgool in London and my bronunziazhion is
Engizh – glear, even vaindly Agzonian, the zame as my dad’z.
Ameriganz avden zeem zurbrised do hear an eleven-year old who zbeegs
as I zbeeg. Grandaddy Jag, who is Amerigan, admids thad he vinds id
unganny. As iv zuj an agzend reguirez grade ganzendrazhion even vram
grownubs, led alone jildren. Amerigans zem do zuzbegd thad the Englizh
relagz and zbeeg Amerigan behind glosed doors. Zhouding oud, on their
redurn, “Honey, I’m home!” My other grandvather veld divverendly:
Englizh, do him, was the more najural voice. Zo thiz zdory iz vor them,
doo, as well vor Eliaz. I dell id thiz way – in zargazdig Ameriganeze –
becaz I dond wand id do be glear: do all be grizb and glear. There iz thiz
ztrange resizdanze. There iz thiz zdrange resizdanze.
(Martin Amis, “What Happened to Me on My Holiday”,
from Heavy Water and Other Stories, pages 199–200)

4 There are several – four, actually – physical strategies for enacting the
voiced alveolar lateral approximant / l /. The first, usually used in a
prevocalic position in a word, is the so-called “clear [ l ]”, produced by
placing the tip of the tongue and the very front of the laminal portion onto the
alveolar ridge and then dropping the side edges of the tongue only as far as
needed to produce the approximant action. The mid-line of the laminal
portion of the tongue is also raised slightly.
The second type that is often used by American speakers in prevocalic
and postvocalic positions within a word is the “dark [ ɫ ]”. It is often thought
that in order to produce the [ ɫ ] a certain amount of retroflexion is involved,
with the tongue tip retracted to a post-alveolar position. I submit that this is
not really the case. The difference between [ l ] and [ ɫ ] is that [ ɫ ] is
produced with only the tip of the tongue on the gum ridge, but in the same
location as the [ l ]. The blade of the tongue is lowered sharply (and the jaw is
relaxed open).
The third type of / l / that we find in both prevocalic and postvocalic
positions in American speech is the retroflex [ ɭ ]. Here the tongue-tip is
curled back to a post-alveolar or even alveolar-palatal placement.
We have discussed previously the fourth strategy, the velarized [ lˠ ],
where only the velar coarticulation of / l / is used. This is common in
American speech but is only used in a postvocalic, generally final, position in
a word.
Other allophones of / l / may be ignored here because they are not
employed commonly by American speakers, although the unvoiced lateral
fricative [ ɬ ] might be used in a word sequence like “at least” or “Atlantic” if
lateral plosion of the / t / is used into a stressed syllable.
The common wisdom among many speech teachers is that the clear [ l ]
should be used in both prevocalic and postvocalic positions. This is an overly
rigid doctrine and often leads to a lack of fluency. When / l / precedes a front
vowel it would seem very reasonable. When a front vowel moves into a
postvocalic “l” it would seem efficient to use a [ ɫ ]. When the / l / follows a
back vowel, we almost always move into some degree of an [ ɭ ]. If we avoid
excessive retroflexion, we can produce a version of / l / that communicates
the phoneme very easily to an audience. Because the retroflexion naturally
causes a greater degree of velar arching, the (minimally retroflected) [ ɭ ] is
useful for final / l / placement as well as for use in an acoustically challenging
environment, such as a large outdoor theatre or a crowded classroom of
teenagers. Dropping the tongue-tip to produce the wholly velarized [ lˠ ] is an
important skill in informal speech but does not communicate as clearly to
large groups as any of the other three options.
Explore the words in the following list, placing the / l / in initial, medial,
and final positions. Try the first three allophones of / l / in each word. Add
the velarized [ lˠ ] for the final / l / words. What sounds most clearly defined
and – at the same time – fluent for each word and sentence? Where does your
tongue “want” to touch the roof of the mouth in these words and sentences?
Can you modify any of your tongue action preferences in the interest of
intelligibility?
lee, lean, lip, lit, lift, left, lest, lab, land, lashed, locked, lopped, lobs,
law, lost, lawn, laud, look, Lou, lose, loop, lewd, luge, loose, lay, lame,
lie, life, lice, loud, lounge, lout, low, loaf, lone, lope, loin, Lloyd, leer,
lair, lure, lore, Lahr
silly, pellet, ballad, polyp, taller, pullet, ruling, sailing, pilot, howler,
molar, boiler
meal, seal, feel, wheel, reel, eel, deal, keel, Beale, veal, zeal, fill, pill,
sill, sell, tell pal, Val, Baal, call, ball, mall, tall, wall, shawl, trawl,
pull, fool, tool, dole, shoal, file, style cowl, foul, coil, boil, rule, school,
pool, ghoul, stool, spool, drool
Lulu, Lily and Lolita were pleased to call on Pearl. All letters were
delivered to the ballet. Ulla mulled over all the allegations.
Audio track 80: Initial, medial, final / I /.

There is another issue related to / l / that needs to be discussed. When [ɫ] or [

ɭ ] is used in a final position after a close vowel like / i / or / u /, we have a
tendency to intrude an approximant plus vowel before the final / l /, so that
“feel” [ fiɫ ] becomes [ fijəɫ ] and “fool” [ fuɫ ] becomes [ fuwɫ ]: a one-
syllable word becomes a two-syllable word. Why does this happen?
Consider the physicality of the problem and the solution. Usually the
front of the tongue is arched high to form / i / and then is lowered to facilitate
the rise of the tongue-tip to form the / l /. The problem is that this happens
before the tongue-tip has reached the point of contact with the alveolar ridge,
which is the crucial physical action to establish the / l /. When going into a
final [ l ] from a high (close) vowel, either [ i ] or [ u ], make sure that you
raise the tip of the tongue to contact the alveolar-palatal point of articulation
before you lower the front or middle of the tongue. This sequence of actions
will allow you to say “feel” or “fool” as [ fiːl ] or [ fuːl ] instead of [ fiːjəl ] or
[ fuːwəl ].
When going into a final [ l ] from a high (close) vowel, either / i / or / u /,
make sure that you raise the tip of the tongue to contact the alveolar-palatal
point of articulation before you lower the front or middle of the tongue. This
sequence of actions will allow you to say “feel” or “fool” as [ fiːɫ ] or [ fuːɫ ]
instead of [ fiːjəɫ ] or [ fiːjəɫ ].

5 The dreaded / r / problem. Americans say / r / in a prevocalic position in

one of three ways: as a post-alveolar approximant [ ɹ ], as a retroflex [ ɻ ], or
as a “braced” or “molar” / r /, for which there is no phonetic symbol. Most
Americans use either the second of these options, or the third, or a
combination of both. All of these three strategies convey the phoneme to a
listener. But many Americans have scant experience with the physical
production of the [ ɹ ]. It is imperative to gain this skill, not only because it is
very useful in American speech, but also because it is used in a number of
other accents.
As a reminder, the action is simple: with the jaw relaxed open, the tip of
the tongue is curled upward – and therefore the blade also – leaving the
dorsum of the tongue relaxed. The tip intrudes minimally into the voiced flow
in a post-alveolar position. You will feel the effect of the pressure waves (e.g.
vibrations) at the tip of the tongue. To release the sound, let the tongue-tip
relax back to a position behind the lower front teeth.
Explore the simple physical sequence slowly as an unvoiced action. Then
explore it voiced. Make sure that you are isolating the action to the tip and
blade of the tongue.
Here is another way to get to it. Speak an extended [ z ]. As you do so,
gradually relax your jaw open and let the tongue-tip be naturally drawn back
to a post-alveolar position by that action. Relax the tip very slightly to an
approximant rather than fricative position.
Now speak the following word list of / r / in initial and medial positions.
Use first the post-alveolar approximant [ ɹ ], then the voiced retroflex [ ɻ ]
and finally the braced or molar / r /.
reap, reed, rip, wrist, wreck, raft, rod, wrong, rook, rule, rug, rain,
rind, round, roil, robe, ruse, rake, rind, rural
pray, bribe, try, drain, creep, grove, free, crave, drone, throat,
average, three, shriek, arrest, direct, parade, carouse, charade, barren,
carrot, parrot, narrow, marrow, lariat, errand, chorus

Now explore the following medial and final / r / using any of the three
physical actions open to you: voiced post-alveolar approximant, retroflex,
and braced.
rear, rare, Ruhr, roar, bear, fear, tour, core, far, jar, stir, fur, whir,
occur, defer, learn, churn, usurp, lured, toured, world, purse, barn,
charge, heart, park, parch, lard, shard, charm, farce, bored, hoard,
scorn, weird, beard, pierce, fared

Finally, speak the following sentences, being aware of which of these three
actions is clearest and most fluent.

Rory ruined Fleur’s transistor radio.

Carol reread the résumé of Irving’s daughter.
Dragan resented Trevor’s relentless boring perorations.
Audio track 81: / r / in variation

6 Efficient pronunciation of / s /. This consonant action is a problem for many

speakers. There are many variants of this unvoiced fricative and some of
them, such as lateral release, would need to be addressed by a speech
pathologist. But there are a few techniques that can help with the most
common issues.
In America roughly fifty per cent of speakers enact the / s / with the
tongue-tip up and the airstream focused at the alveolar ridge. And about fifty
per cent of speakers pronounce / s / with the tongue tip down; here the air
stream releases over the blade of the tongue with a focus that is post-alveolar.
In the past, speech teachers would insist that all speakers use the “tongue-tip
up” version. While this technique has some advantages in connected speech,
it is also true that this basic modification of tongue-tip action is fairly arduous
and may not be worth the time and effort if the speaker is producing a clear /
s / with the tongue-tip down – and many people do.
No matter what physical action the actor takes in forming it, the phoneme
/ s / should always be produced with as small an aperture (formed at the
midline of the tongue) as possible, and also with as gentle an airflow as
possible. The placement should be alveolar, or as near as possible to alveolar.
If all of these actions are done accurately, the / s / will not hiss or sound
mushy. If the tongue focus is dentalized [ s̪ ] the result for most people will
be a very thin, hissy sound product. If the focus is retracted too far (a danger
with the tongue-down version) the airstream will become too broad by the
time it washes down the back of the upper front teeth and the noise will be
mushy. A lot of this fronted/retracted focus is dependent on an individual’s
oral anatomy, especially with regard to the size, shape, and spacing of the
teeth. So experimentation is in order. The most important criteria are: narrow
airstream and minimal airflow.
Practice using a minimal flow of air on the / s /. Very often, in speaking,
we send much more air than we need through whatever aperture the tongue is
forming with the roof of the mouth for all unvoiced fricative consonants. Say
the word “heat” a few times just to make sure you can start the / i / without
using a glottal attack on the vowel. Then move to “eat” without any glottal
attack / ʔ /. Then say “seat” with as little airflow on the / s / as possible. Also
make sure that you keep the stream of the airflow as narrow as possible,
whether you pronounce the / s / customarily with tongue-tip up or tongue-tip
down. Alternate saying “eat” and “seat”; then say them in an irregular
pattern: “eat eat eat seat seat eat seat eat seat seat eat seat eat eat eat seat”,
etc. Note if a listener (who knows what to listen for) can tell which word you
are saying.
see, seat, seam, cease, sit, sift, silt, set, said, send, sat, sand, sack, sap,
sock, sod, sob, solve, sop, psalm, saw, sought, soft, Salk, so, sew, soap,
sewed, soak, soul, soot, sue, soon, suit, soothe, serve, search
speed, steel, ski, spin, stint, skill, spell, stet, sketch, spat, stand, scam,
spot, stop, scotch, spawn, stall, scald, spoke, stole, scold, stood, spool,
stool, school, sphinx, sprawl, strand, scrape
assist, recent, fasten, associate, Mississippi, assists, resets, necessity,
Cecil, Cicely, Cecily, Sicily
seeps, deeps, creeps, leaps, sips, lips, tips, reps, taps, chaps, gaps,
hops, chops, tops, drops, loops, troops, groups, hoops, chirps, burps,
ups, cups, seats, meets, Keats, fits, writs, pets, bets, lets, mats, pats,
rats, chats, lots, pots, knots, cots, thoughts, puts, hoots, boots, hurts,
asserts, huts, butts, shuts, baits, fates, mites, rights, lights, louts, touts,
shouts, boats, totes, coats, votes, seeks, cheeks, geeks, reeks, sticks,
ticks, picks, wrecks, checks, necks, packs, sacks, lacks, knocks, socks,
rocks, balks, talks, chalks, looks, books, hooks, rooks, Luke’s, works,
lurks, tucks, mucks, bucks
niece, peace, lease, crease, hiss, miss, kiss, Bess, less, Tess, dress,
chess, brass, pass, gas, hostile, Haas, Haagen-dasz, loss, boss, toss,
puss, loose, moose, curse, hearse, fuss, muss, pace, race, face, ice,
mice, trice, house, mouse, louse, voice, choice,
crisp, lisp, asp, clasp, hasp, rasp, least, beast, pieced, mist, whist,
wrist, list, best, chest, vest, rest, past, massed, last, fast, bossed, tossed,
lost, cost, roost, boost, first, worst, bust, trust, baste, taste, riced, iced,
roust, joust, roast, boast, toast, voiced, hoist, joist, risk, flask
crisps, lisps, asps, clasps, beasts, feasts, wrists, fists, schists, tests,
bests, pests, casts, masts, costs, frosts, roosts, thirsts, lusts, bastes,
tastes, ousts, Fausts, posts, roasts, coasts, hosts, toasts, hoists, risks,
asks, flasks, casks, husks, rusks
imps, chimps, stamps, comps, romps, lumps, chumps, hints, mints,
rinks, thinks, thanks, banks, shanks, pronks, lunks, dunks
Standing aside, Cicely spoke seriously with Cecily. Stanton Stassen
sold Bruce a gross of husks and some soap-suds that suddenly
disappeared. Silas the moose passed gas in gusts.
Amidst the mists and coldest frosts With barest wrists and stoutest
boasts, He thrusts his fists against the posts And still insists he sees the
ghosts.
 Audio track 82: / r / initial, medial, final.
7 Word linking. A reminder that the formal / tj / and / dj / – when an actor is
linking words in complex text – may be preferable to the yod coalescence
into the affricates / tʃ / and / dʒ /. Practice this word list using both the / j /
and then using the afficates.
seat you, fit you, get you, let you, at you, pat you, that you, blot you,
bought you, taught you, put you, suit you, rate you, sight you, fight you,
clout you, coat you, what you, butt you
feed you, lead you, bid you, kid you, wed you, pad you, laud you, hood
you, brood you, raid you, jade you, ride you, bide you, crowd you, load
you
Audio track 83: Word linking / dj /
An allied issue with linking words (and within a few words too) in formal
speech is the enactment of word sequences in which the first word ends with /
l / and the next word begins with / j /. Because the / j / is dorso-palatal, it is
tricky (though formally required) to get the tongue into proper placement if
the speaker is coming out of an alveolar / l /. But, in informal speech, a
reasonable substitute is the velarized [ lˠ ] where the tongue-tip stays down
and makes easier the arching of the front of the tongue into the approximant /
j /. So the sequence [ lj ] turns into the sequence [ lˠj ]. Practice the list both
ways.
steal you, heal you, will you, bill you, sell you, tell you, call you, maul
you, pull you, curl you, hurl you, dull you, lull you, ail you, fail you,
pile you, rile you, cowl you, foul you, dole you, coil you, boil you,
William, value, psyllium, Valium, thallium, Coolio
Audio track 84: Word linking / lj / alveolar and velarized.

8 Glottal stop-plosives. Despite the IPA’s placement of / ʔ / on the pulmonic

chart as an “unvoiced glottal stop-plosive”, it is obvious that the IPA is not
entirely accurate here. It will not take much experimentation to feel that the
glottal plosion that occurs when we say “uh-oh” is voiced, not unvoiced. A
genuinely unvoiced glottal stop-plosive would be a light cough. We know
that glottals are used all the time by American speakers; often before a vowel
in an initial position in words such as east, air, owe, up, or ooze. They also
occur in medial and final positions within a word, usually as an unsounded
stop [ ʔ ] to replace / t /.
Sometimes there are good reasons to use a voiced glottal stop-plosive: it
is a very effective way to emphasize strongly a word that begins with a
vowel. But the need for glottal emphasis is comparatively rare in American
speech, formal or informal. The only words in American English where
glottal plosion conveys phonemic content are “uh-uh” and “uh-oh”. (Try
these both ways, with and without glottal plosion, and you’ll see what I
mean.) Those two vocalizations excepted, glottal plosion is never needed for
any reason other than strong emphasis, so it becomes a phoneme separated
from meaning and thus a phoneme without any purpose at all. Sometimes,
too, there are good reasons to use a glottal stop as substitution for another
stop-plosive: it still conveys phonemic detail while maintaining fluency. But
if it becomes a habit, you will use it too often when other actions might
produce more detailed connected speech.
The real problem with glottal plosion and glottal stops (when used in
English speech – in some dialects of German they are required) is that they
commonly show that you are holding residual tension in the delicate muscles
of your vocal folds, a sign that you have insufficient easy support in your
breathing mechanism. It is a crucial skill for effective vocal use as well as
speech to be able to initiate a word with a vowel without exploding it from a
tensed glottis and to be able to end a syllable without always using a glottal
stop. Actually it is a crucial set of skills.
The first skill is going onto an initial vowel from a relaxed glottis. Begin
by bringing the vocal folds close together, but not to the point of “creaky
voice” or phonation. Breathe out easily without increasing the airflow. The
effect is a little like a very gentle unvoiced “h”. From that feel of the airflow,
go into the vowel. You should do this slowly at first and then gradually phase
the initial aspiration out. You can explore this on huh huh huh huh huh. Do it
first with on separate breaths and then do the entire sequence of five on one
long release of flow. Whenever you start a sentence with a vowel, you will
have to use this technique.
When you speak a sentence, if you are moving from a final consonant to
an initial vowel, you can carry the final consonant through to the next word,
as though the next word actually began with a consonant, as in “Sit over on
the left aisle.” An allied technique is to use the so-called “liaison r” where the
post-vocalic / r / strengthens to feel like an initial “r”, as in “over on” in the
previous sentence.
Your expertise in phthong shaping will assist you in mastering the next
skill: dealing with a final vowel or diphthong moving into a word that also
initiates with a vowel. Vowel-to-vowel blending, in other words. It is easy to
move from one vowel phoneme to another without intruding a glottal plosion.
There are only two rules for this. First, don’t be afraid to elide one vowel
to another without an intervening attack on the word; if you do it clearly we
will understand that the syllable is changing because syllabic change is also
marked by sub-glottal pulses of energy. Second, don’t feel that you have to
intrude an approximant in order to start the initial vowel with a semi-vowel;
you can speak “we are” as [ wi.ɑ ] by simply dropping the arch of the
tongue away from the / i / to the / ɑ / without having to intrude a / j / to form
[ wijɑ ].
It is also common for American speakers to stop an unvoiced stop-plosive
in a final position while simultaneously coarticulating stoppage at the
appropriate placement for that particular stop-plosive, as in stop, lip, hope,
wrap that, set, boot, black, take, and so on. It is always clearer to enact the
full stop-plosive action in final stop-plosives
There is one other skill for eliminating unneeded medial glottals, but we
will deal with it in a page or two.
Practice the following words and sentences using these techniques to
eliminate unnecessary glottals; then (as always) put the glottals in
aggressively. Feel the difference between the two, because you will need both
skills as an actor.
each, east, if, ill, Ed, add, as, odd, on, all, ought, ooze, earn, earth, up,
aim, aisle, out, owe, oil, ear, air, ore, are, ease, eel, eke, eaves, each, if,
is, itch ill, ink, ate, ape eight, oh, ah, ash assist, oolong,
beat, seat, wit, met, wet, set, rat, brat, fat, lot, sot, cot, taught, fought,
bought, rote, vote, put, “beaut”, butte, newt, boot, scoot, hoot, sate,
bait, light, right, mite, bout, shout, pout, rout, moat, vote, stoat, coat,
groat, quoit, court, fort, short, cart, part, start, art, mart seep, peep,
 sheep, reap, heap, cheap, sip, drip, flip, chip, pep, step, rap, gap, lap,
sap, cap, lop, flop, cop, shop, stop, mop, top, hop, soup, loop, hoop,
coop, dupe, “turp”, “perp”, twerp, sup, cup, tape, shape, cape, type,
ripe, cope, hope, grope, soap, lope, dope, tarp, flake, make, take,
choke, bike, Luke, dark, lurk, seek
muse of, take out, read on, rake over, shake out, as is, on / off, write in,
add on, pull out, peace on earth
for a, care about, peer out, pour out, know all, bore in, tear into, more
oats, fear adversaries, stare at
say on, try out, be opposed, see all, buy out, through obstacles, no
entrance, pay off, row away, why argue, now entertain

When glottises attack

Elliot ate an apple and offered Andrew another.
Each and every avenue is open at eight o’clock.
I am in agreement on every aspect of our association.
Alan’s attitude is overly obnoxious.
Exercise is an important and energizing activity.
Amanda is in Alabama at an annual event.
Actually I am aware of all errors in Adam’s arithmetic assignment.
Every evening in autumn our area orchestra attempts to entertain an
uninterested audience of adolescents.
Esmeralda earned an audience’s enthusiastic applause after exhibiting
an outrageous and ergonomically effortful acrobatic undertaking.
“Aren’t all of us – and I above all – existing endlessly in an over-arching
aura of old (indeed ancient) outworn, outré, Aristotelian ontology?”
Iggie inquired innocently.10

Audio track 85: Glottal stop-plosives.

9 Aspiration in informal and formal speech.

In formal speech, unvoiced stop-plosives [ p ], [ t ], and [ k ] are aspirate [ ph

], [ th ], and [ kh ] before vowels (including diphthongs and triphthongs) in
the same stressed syllable. Some examples are “appalled” [ əˈpʰɔːld ],
“timely” [ ˈtʰaɪ̆mli̞ ], “occur” [ əˈkʰɝː ]. In these cases the voicing of the
vowel waits for the aspirate gust to go by before asserting itself.
If the unvoiced stop-plosive is followed in a stressed syllable by an
approximant consonant, especially [ ɹ ], [ l ], [ w ], or [ j ], there is no pause
between the stop-plosive and the approximant, but the aspiration in effect
blows out all or part of the voicing of the approximant. So “crew” becomes [
kɹ̥u ], “plane” becomes [ pl̥eɪ̆n ], “cue” becomes [ kçu ], and “quail”
becomes [ kʍeĭl ].
There is an exception to this delayed “onset of vocalization”. If the
unvoiced stop-plosive [ p ], [ t ], or [ k ] is preceded in the same syllable by
an [ s ], the stop-plosive is not aspirate. This reads as complicated, but it isn’t
really. Just feel when the voiced vibrations – (the [ ɫ ] sounds) – come in
while you say first the word “team” and then the word “steam”.
In all other cases, the stop-plosive is not usually aspirate in American
English. When followed by a consonant (other than an approximant) as in
“Hapgood” [ hæpgʊd ] or in an unstressed syllable as in “naughty” [ nɔti̞ ],
the aspiration does not occur. The same is true, for that matter, at the end of a
word, whether in the middle of a sentence “at all” [ ət ˈɔːl ] or at the end of a
statement “Not!” [ nαt ]. In the first case, the word “at” is unstressed, and in
the second case the emphatic statement might well close with a stopped
consonant that is not exploded. In this last example, even if a little aspiration
crept in to honor the exclamation mark, it wouldn’t convey any more
linguistic information.
Explore the following word lists and sentences, feeling the differences in
physical action of aspiration. And yet, again, try the opposite: eliminating
aspiration (which occurs in many languages), adding aspiration, from plosive
to plosive, etc.
pea, peep, peach, peak, piece, pique, pip, pit, pig, pin, ping, pitch, pith,
pill, pick, pet, pep, pest, pen, peg, pap, pat, pack, pan. pang, path, pass,
pal, pop, pot, pod, pock, posh, pall, pawed, pawn, put, push, pull, poop,
pool, Pooh, purr, pert, perm, purse, perk, purred, pup, putt, pug, pun,
puff, pay, paste, pain, pail, pine, pipe, pike, pie, pied, pow, pout, pound,
pouch, pope, poach, post, poke, pole, poise, point, peer, pierce, pair,
pared, poor, pour, pores, port, porch, par, part, park
tea, teak, tease, teeth, teach, tip, tick, tiff, tin, till, tech, ten, test, tell,
tap, tack, tan, tang, tag, top, tot, tog, tock, tall, taught, toss, took, two,
too, tool, toot, turn, Turk, terse, tuck, touch, tuft, taste, tame, tale, take,
time, tyke, tile, tight, type, tide, tout, town, towel, tote, tone, toast, toll,
toy, toil, tier, tear, tour, torn, tarred
key, ken, keel, kit, kick, kin, king, kiss, kitsch, ketch, ken, cap, cat, cab,
cask, cash, calm, cod, cop, con, call, caught, cause, caulk, cough,
could, coo, cool, coop, kook, curse, curr, curd, curl, cup, cut, come,
cave, case, cape, cane, kind, kite, cow, count, couch, cowl, cope, coat,
code, coast, coke, cone, comb, cove, coach, coal, coy, coin, Coit, coil,
care, cairn, Coors, core, corn, coarse, car, cart, carve, Carl priest,
prince, primp, press, pram, prom, prop, prawn, prove, pray, pry, pride,
prow, prowl, pro, probe, prone, prayer, tree, treat, trip, trim, trick,
trill, tress, tread, trek, trap, tram, track, trash, trod, trawl, true, trade,
traipse, train, try, tripe, trope, troll, Troy, Cree, creep, creed, crease,
crimp, krill, Chris, creche, creft, crash, cram, crock, crawl, cross,
croft, crook, crew, crude, crave, crate, cry, crime, crown, crowd,
crouch, crow, crone, croak
plebe, plead, please, pleat, plinth, pled, pledge, plaid, planned, plaque,
plop, plot, plume, play, plane, plate, place, ply, plight, plough, ploy,
clean, cleat, cleave, clique, click clip, cleft, cleanse, clap, clan, clad,
claque, clog, clod, clop, claw, cloth, clue, clay, claim, climb, cloud,
clown, clothes, clone, close, cloy, clear, Clare, chlor-, Clark,
cue, cute, quease, queen, quick, quill, quit, quest, quell, quack, quad,
qualm, quail, quake, quite
speed, spin, spend, spell, spat, span, spot, spawn, spool, spurt, spurn,
spun, Spain, spate, spade, spy, spine, spike, spout, spouse, Spode, spoil,
spear, spare, spoor, spore, spar, steed, steel, steep, still, stint, stitch,
step, stead, stand, staff, stack, stag, stab, stop, stock, stall, stalk, stood,
stone, stow, stoat, stole, ski, scheme, skeet, skip, skiff, skin, skit, skill,
sketch, scam, scat, scab, Scot, scald, school, scoot, scum, skull, scud
skirt, skate, scale, skein, sky, scout, scowl, scour, scold, scope, spree,
 sprit, sprint, spring, spread, sprat, sprang, sprawl, spray, sprain, spry,
sprout, stream street, strip, string, strict, stress, stretch, strand, strap,
strop, straw, strew, stray, strain, straight, strive, strike, Stroud, stroll,
strove, stroke, scream, screed, script, scrim, scratch, scram, scrod,
screw, Scrooge, Scroop, scrape, scrounge, scroll
cupcake, actor, nocturnal, aptitude, nectar, sector, action, noxious,
section, napkin, raptor, peptide, batboy, lactose, stripling sitcom,
fruitcake, lapdog, captain, Atkins, octave, chatroom, April, sacred,
coupling, reckless, atlas, septum, puptent, acrid, sacrament, macrame,
sapling
“Speed the Plow” is a cold but charismatic play about corruption.
Training troops will drain Congress’s capital. Playing college lacrosse
causes casualties.
Audio track 86: Aspiration of unvoiced stop-plosives (first lines).

10 Syllabic consonants.

The major syllabic consonants [ ṃ ], [ ṇ ], and [ ḷ ] always occur as

unstressed syllables. As a general rule, if you can use a syllabic consonant
(thus eliminating a voiced vowel between the preceding consonant and the
unstressed syllable’s consonant) you should do so: [ ˈæpl̩ ] not [ æpəl ]. This
shows a fluent economy of action, and also ensures that the syllable with the
syllabic consonant is really unstressed. Or, to put it another way, if you are
always putting a vowel into these syllables, you are probably stressing all
your syllables too much. It is acceptable – actually it is preferable – to go
from a stop-plosive to a syllabic consonant by means of “nasal plosion” or
“lateral plosion” – nasal plosion if you are exploding the stop-plosive into a
nasal consonant such as [ ṇ ] or [ ṃ ], and lateral plosion if you are exploding
the stop-plosive into an [ ḷ ]. It is simply the most efficient way to reach the
syllabic consonant without intruding an unneeded vowel. Thus “hidden” is
said as [ hɪdⁿn̩ ] not as [ hɪdən ], and “cattle” is said as [ kʰætˡl̩ ] not as [
kʰætəl ].
Be wary of turning an approximant – especially [ l ] or [ ɹ ] – when each
follows another consonant in a cluster, into a syllabic consonant. Syllabic [ ḷ ]
and [ ɹ ] are definite possibilities and are particularly tempting when you are
emphasizing a word by elongating it. Unless the speech of the character
demands it, “Please!” should not become [ pʰl̩ːˈiːz ] and “Gross!” should
avoid the familiar – but now archaic – valley girl [ ɡɻ̩ːˈe̞ːˌʉːs ], where it
stretches magisterially to three syllables.
Practice these word lists both using syllabic consonants, and also using,
or withholding, nasal or lateral plosion where necessary. Then explore other
informal – or more formal – options.
prism, chasm, venom, plenum, capitalism, communism, season,
seasonal, poison, poisonous, cousin, nation, raisin, Byzantine,
Casanova, evil, hassle, facial, rational, awful, Ethel, weasel, nozzle,
pestle, nestle, bristle
[ tⁿn ] [ dⁿn ] – eaten, hidden, certain, patent, sudden, bidden, midden,
sadden, kitten, Seton, redden, maiden, mitten, fatten
[ tˡl ] [ dˡl ] – little, settle, riddle, puddle, battle, metal, brittle, gauntlet,
addle, fiddle, cattle, shuttle, Natalie, Italy, middle, pedlar, atlas,
peddle, whittle, Petaluma, pantaloon, cantaloupe, vittles, boodle,
cuttle, fertile, modal, riddle, settling, whittling, addling, curdling
please, plead, pleat, clean, cleat, cleave, glean, fleece, sleet, cliff,
glyph, blend, bless, cleanse, cleft, fleck, slept, schlep, plaid, clad, glad,
blot, clop, clock, claw, Claude, clause, plume, bloom, clue, gloom,
flume, slew, play, blade, claim, glade, flame, plight, blind, climb, glide,
plough, blouse, cloud, glower, blow, clone, gloat, float, slow, ploy,
cloy, Floyd, preen, priest, preach, breed, breach, tree, treat, treed,
dream, creed, crease, green, Greek, grief, free, frieze, three, print,
brim, tryst, drift, crisp, grist, frill, thrill, shrift, prove, prune, broom,
brood, true, truce, troop, drool, droop, croon, crude, cruise, groom,
groove, fruit, through, shrew, shrewd
eaten, beaten, written, cotton, mitten, fatten, rotten, button, Satan,
potent, mutton, patent, potentate, curtain, frighten, batten
little, whittle, brittle, settle, fettle, kettle, metal, nettle, battle, cattle,
 chattel, bottle, brutal, fertile, myrtle, cuttle, natal, fatal, vital, total,
portal, mortal
Eden, bidden, hidden, midden, didn’t, leaden, deaden, sadden, hadn’t,
madden, oughtn’t, couldn’t, shouldn’t, maiden, Haydn, louden, rodent
needle, beadle, wheedle, fiddle, twiddle, middle, medal, saddle, coddle,
model, caudal, dawdle, boodle, curdle, puddle, muddle, huddle, sidle,
bridle, adenoidal, colloidal
Audio track 87: Syllabic consonants; nasal and lateral plosion.

11 About final stop-plosives.

The plosive part is the problem. This is the most important skill for formal
speech. As such, it is worthwhile to practice the specific skill of stopping a
stop-plosive, unvoiced or voiced, and then exploding it fully without
appearing to over-articulate when you do so. When this is enacted
excessively in performance the actor’s speech appears stilted and affected.
When it is called for, but not enacted, the intelligibility of the performance
will suffer grievously.
First, let us remind ourselves of what we should already know: a stop-
plosive is a double action consonant, the most extreme obstruction of the
airflow, in which the flow is completely stopped for a brief moment at the
point of complete obstruction wherever it might be, from the lips to the
glottis. e need to remember that the flow is interrupted only briefly – so
briefly, in fact, that the air molecules behind the point of closure are
condensed just enough to give them – along with the elasticity of the tissues
of the vocal tract – a certain springiness that encourages plosion. But the flow
itself never really stops, whether unvoiced or voiced.
I mention this because the problem for many actors and other speakers’
enactment of the stop-plosive is doing too little or too much. It is my personal
suspicion that the pervasive tendency to do too little comes from a fear of
doing too much. Learning to enact just enough should solve that problem and
– as with most of the work in this part of the book – you already have the
skills to do this.
With all stop-plosives the really important thing is that you take both
parts of the term seriously: except where we know that we can engage in
assimilation where the stop-plosive is stopped only, we should practice
diligently the ability to stop the flow and then to explode it. We are usually
brilliant at stopping; we are less practiced at exploding. So with unvoiced
final stop-plosives, practice exploding the final / p /, the / t /, or the / k / as
fully as you can in the word lists below. Then explore the degree to which
you can minimalize the aspiration that follows the plosion by reducing the
flow that that been used to create the stop-plosion action.
With voiced final stop-plosives the problem is slightly different, though
louder. If you force too much voiced flow through the voiced stop-plosive,
you will produce a brief vowel sound after the voiced stop-plosive action, the
so-called “shadow vowel”. The solution, as before, is to really practice
moderating the subglottal pressure downward so that you can feel and hear
the completion of the stop-plosive action, but not hear that “shadow vowel”
after it.
Another action that helps for both unvoiced and voiced final stop-plosives
is to make sure that the release of the articulators away from the point of
closure is very active. You can assist the activity of the release by keeping the
jaw relaxed open so that the action is limited to the lips or tongue and does
not include the jaw. So the formula is: less flow and more articulator activity.
Practice the formula with the word lists and sentences below.
deep, leap, sheep, keep, seep, lip, sip, ship, tip, hip, nip, tape, drape,
nape, shape, pep, cap, lap, nap, wrap, sap, chap, top, lop, cop, fop,
shop, mop, hop, lope, mope, soap, cope, loop, stoop, dupe, group, type,
hype, Skype, wipe, ripe, cup, sup, slurp, chirp
seat, meet, bleat, greet, feet, neat, pit, sit, wit, hit, bit, lit, fit, sate, bait,
rate, wait, fate, date, set, let, wet, met, bet, get, whet, pat, cat, sat, fat,
gnat, chat, pot, not, cot, hot, got, plot, spot, caught, naught, bought,
thought, sought, coat, boat, moat, groat, vote, put, mute, boot, suit,
shoot, coot, hoot, might, bite, sight, light, fight, shout, lout, pout, bout,
quoit, adroit, cut, but, what, mutt, pert, shirt, curt, flirt, Burt
seek, beak, leak, wreak, teak, pick, tick, stick, lick, nick, sick, take, lake,
wake, make, steak, beck, trek, check, wreck, peck, sack, back, tack,
mac, rack, mock, dock, shock, flock, hock, knock, talk, balk, walk,
stoke, coke, joke, nook, cook, brook, shook, took, nuke, Luke, fluke,
Mike, bike, tyke, psych, strike, dike, hike, muck, tuck, suck, shuck,
truck, perk, lurk, jerk, clerk, cirque
 nib, rib, Jeb, reb, cab, stab, jab, Mab, ab, job, sob, cob, fob, rob, gob,
knob, robe, lobe, tube, lube, cube, rube, vibe, jibe, kibe, rub, tub, nub,
sub, stub, hub, curb
seed, feed, weed, mead, lead, reed, keyed, heed, bid, rid, Sid, grid, hid,
made, raid, payed, shade, led, said, shed, Ted, red, head, mad, bad,
sad, clad, add, grad, plod, sod, shod, cod, mod, wad, hod, god, rod,
Maude, awed, jawed, laud, load, road, code, bowed, hoed, good,
should, cooed, mood, rude, clued, sued, shooed, who’d, food, slide,
ride, hide, chide, cowed, bowed, loud, crowd, cloud, mud, bud, dud,
cud, Fudd, bird, stirred, herd, word, blurred, shirred, whirred
league, pig, dig, jig, whig, big, vague, leg, beg, egg, peg, dreg, bag,
sag, nag, rag, lag, jag, flag, gag, hag, bog, slog, jog, fog, clog, Prague,
hog, nog, cog, brogue, rogue, dug, mug, chug, bug
A big dog sat right on the dock. Peg Wright watched the bird chirp.
Greg Stark asked the food clerk, “Is that called a sub or a zep?” The
mutt got rid of the rat, but the skunk routed the dog, the cat, and the
chimp. “I’d not like to sit,” Pat said. “Make that hat fit,” shouted Rob.
Audio track 88: Final stop-plosives (first lines).
Speak the sentences above several times, first separating all the words with a
short pause between each, so that all of them are “final” stop-plosives. Then
try the sentences again with connected flow, keeping full stoppage and
plosion for each stop-plosive, but making sure that you are neither over-
aspirating the unvoiced final plosion nor inserting a “shadow vowel” after the
voiced plosion. Now do the sentences again, but explore what actions you
can use to achieve fluency without sacrificing any of the phonemic content.
25 Vowel skills

We begin with a series of “vowel calisthenics”. You should already have a lot
of flexibility with vowel shaping, but drill in these exercises will help you to
define vowel phonemes fully for maximum intelligibility in American
English.

*Not the diphthong [ eɪ̆ ].

Target each sound, using the tongue and lips, but not the jaw, which should
remain relaxed. Start slowly, holding each sound position for several
seconds; then speed up, moving actively from one sound position to another.

Figure 61. EE-EY-AH.

Audio track 89: Vowel calisthenics: EE-EY-AH etc.

2 The miaow. Jaw relaxed open, lips lightly together. Begin the release of
sound on a [ ɲ ] hum. Release the lips apart into an [ i ] with the tongue
bunched high toward the very front of the palate, then let the flow of sound
action slide smoothly down through all the front vowels, then back and open
to the [ α ], and then smoothly up to the closed [ u ]. Feel the action of the
lips and cheeks supporting the sound differentiation.
 Figure 62. The miaow.

Audio track 90: The miaow.

3 Tense – lax. Start with the tongue in a completely relaxed position, in the
mid-vowel [ ə], the “schwa”. Tense into the target vowel in its most
differentiated form. Then relax back into the “schwa”.

Figure 63. Tense – lax.

Audio track 91: Tense – lax.

4 Adjust tongue and lip position precisely to target the “center” of each
phoneme.
(a) Move cleanly from one phoneme to another.
(b) Slide slowly from one phoneme to the next; feel and hear the gradations
of sound; note also where you feel and hear the phoneme change.
 Figure 64. Peripheral phonemes and allophones.

Audio track 92: Peripheral phonemes and allophones.

5. Separation of tongue action and lip action (rounded or unrounded).
(a) Keep the tongue position constant and change the phoneme by modifying
the rounding action of the lips.
(b) Keep the lip position constant and change the phoneme by altering the
tongue position within the mouth. Note: jaw position does not need to
change; compensate by greater muscularity of lip and tongue action.
 Audio track 93: Separation of tongue and lip action in vowel production.

Specific skills

Vowel phoneme targeting

One generalization that we can make about vowel phonemes is that
intelligibility increases when the vowel phonemes are clearly distinct from
one another. American speakers have a pervasive habit of mid-centralizing
their vowels – relaxing them all toward the schwa / ə /. For informal speaking
in conversation, and for much speaking onstage, such mid-centralization is
perfectly appropriate, as long as the speaker makes enough distinction
between phonemes still to be understood. However, it is very important that
actors master the skill of active phoneme differentiation, which involves
rather more muscularity in the articulators than most of use are used to.
Your task, in the next exercise, will be to attempt to find the “essence” of
each phoneme, the voicing of the allophone that most completely defines the
phoneme and makes it sound utterly unlike any other phoneme. This means
that while you are voicing the phoneme, you may have to search around in its
allophonic territory to find the one allophone that – in your opinion – most
clearly represents the essence of that one phoneme.
It should come as no surprise to you that we are not talking about science
here, although there are scientific ways of measuring your sound’s allophonic
positioning within the vowel quadrilateral. This is a judgment call. Since two
heads are often better than one, this is a useful drill to do with a partner.
Start by performing the vowel or diphthong on an extended vocal release,
modifying it until you decide that this is the most distinctive, essential
version of the vowel phoneme. (If it’s a diphthong, focus on the essential
nature of the first element of the diphthong,) Maintaining the physical posture
of that precise allophone, start through the word list, extending the vowel on
each word to define vocal tract posture, even with the other phonemes around
it. Then see if you can speak the word list at normal conversational speed
without compromising the clear, distinctive articulation of the phoneme.

/i/
ease, east, each, emu, evil, eland, eagle
wheedle, needle, keep, steam, bees, shield, niece, wreak, weasel
clean, cream, fleece, cheese, shrieks, bleeds, grieve, sneak, screeds, please
be, see, three, she, knee, bree, we, tree, free, glee, ski, Lee, plea, Cree
He’s seen a bee in the tree for three weeks.
Leave the fleeces between each sheet.
Heed me: she teems with greed and steamy need.
Audio track 94: / i / targeting.

/ɪ/
in, if, it, is, ink, ill, itch, imp, is’t, isn’t, into, issue, ichor
bin, fin, chin, wrist, gift, hit, miss, zinc, fizz, pick, whip, knit, shin
blince, quince, swift, grip, brim, crisp, glitch, print, plinth, trimmed
“This is it,” he admitted grittily.
Isn’t it silly to shift things mysteriously?
The grisly grifter was ill with the vicissitudes of iniquity.
Audio track 95: / I / targeting.

/ eɪ̆ /
Abe, ache, able, ace, aid, ape, age
pace, mace, rate, change, fake, came, bait, mail, nape, jail, chain, taste
crave, place, grain, place, trade, flame, raged, praised, Blake, waist
say, pay, may, lay, pray, grey, flay, dray, shea, way, belay, stray
Nate made it safely to the base.
Grey whales change their way of spraying daily.
Abe Blake played the ace with taste and grace.
Clay and May raided Shea Stadium.

Audio track 96: / eĭ / targeting.

/Ɛ/
elf, end, every, any, Evan, Elbert, Edgar, Evan, Ezra, episode shell, well,
fen, tell, reb, rep, wreck, set, deft, left, lend, vet, pet, tech
glen, pressed, fret, checked, treks, Shrek, cleft, prepped, dressed, feh
Then and there, Becky texted her best friend forever, Emily.
Gemma’s ostensibly gentle jest evidenced her genuine stress level.
If Senna had any sense, every elf could have entered better directions to
effect the hex.
Trek’s tresses tend to blend with Drek’s dresses.

Audio track 97: / Ɛ / targeting.

/æ/
ask, ash, and, act, asp, aft, angle, average, ask, apt, azure, active
tap, Mack, sack, lap, sham, rat, rang, pack, lag, stab, patch, fad
trap, clang, shaft, past, blast, graft, craft, rapped, stand, glass, razz
Patrick actually planned to pass the piano handily to Manny.
Vacuous vapid vanity advances annually.
Shaq acted with absolute command of his craft.
Audio track 98: / æ / targeting.

/ aɪ̆ /
I, aisle, icon, ayvar, iPod, idol, iMac, ice, I’ve, Ike, Ayn, eyes
nice, pike, rhyme, vile, tyke, shine, bite, vice, chide, jive, file, light
tripe, polite, blight, shrives, flights, climbed, strides, flies, drives
nigh, fly, why, sty, dry, cry, my, fry, shy, lie, high, vye, try, ply
Sly shysters might fly from crime-fighting.
I’ll try to stymie his flight from his crime.
Why chide my pirate bride for my bright rhyme?

Audio track 99: / aĭ / targeting.

/ aʊ̆ /
ouch, out, owl, ounce, ouster, outlay
town, rout, mouse, louse, lout, town, vowel, bout, bows, vows, mouse
flouts, drown, grout, blouse, flounce, stout, pounce, clout
sow, chow, plough, how, wow, now, row, bough, chow, cow

How now brown sow? I mean Frau. I mean cow.

Count out the thousand clowns.
The mouse caroused with the louse.
Around his stout and pouty snout, the crowd did shout out loud.

Audio track 100: / aʊ̆ / targeting.

/ɑ/
odd, opt, Oz, oblong, Occam, octet, ostentatious, Amish
cot, calm, top, shock, Bach, mock, Tom, fob, job, wan, tog, chock, god
plod, stop, blotch, flock, frogs, clods, globs, Barack, piranha, blasé
Ta, ma, la, Shah, rah, pa

The peroxide blonde’s jobs were not odd.

Tom honored Dali’s flock of clocks.
Don’t knock the rock and don’t knock Bach.

Audio track 101: / ɑ / targeting.

/ɔ/
awe, autism, awful, augury, auburn, Austin
pall, Paul, talk, bought, shawl, caul, bawl, fought, loss, taut, sought
taunt, Claude, trawl, paunch, slaughter, applaud, flaunt, crawls
jaw, shaw, gnaw, craw, awe, saw, caw, daw, flaw, haw, law, maw, paw

All the daws taunted the trawler with their awful caws.
The mawkish brawlers appalled the tallest vaulter.
Shaw altered his talk on his cause.

Audio track 102: / ɔ / targeting.

/ oʊ̆ /
oat, oaf, old, oast, ohm, own, oak, over, opal, Omar, O-ring, ogle, ogre
boat, coat, goat, home, boat, foal, rope, vote, poke, vogue, home, cove
bloke, trove, clothes, encroach, drove, ghosts, coasts, toasts, broached
no, know, row, flow, pro, slow, though, Mo, grow, stow, go, doe, Poe

Joe’s toes slowed his moseying over to his clothes.

Flo’s pose closed the show.
The hoped-for explosion of the bloated crows from the hose flow was a no-
go.

Audio track 103: / oʊ̆ / targeting.

/ʊ/
put, should, push, good, pull, would, took, could, look, nook, bull, wool
crooks, looks, plural, burrow, Murray, flurry, brook, bushed, pulpit
Could the bull’s good looks be hooded by wool, or couldn’t it.

Audio track 104: / ʊ / targeting.

/u/
oops, ooze, outré, oud, oof
loop, pool, stoop, shoes, cooed, roof, chews, sues, food, boom, noose
plume, groups, fruit, troop, cruise, truths, sluice, brood, gloom, truce
boo, hoo, too, sue, stew, shoe, view, rue, pew, do, voodoo, Prue, flue

He perused the news in search of elusive truth.

Who knew the clue to the aloof spruce goose’s cruise?
They enthused too effusively about the blues tune.
Audio track 105: / u / targeting.

/ə /, or / ʌ /
up, under, away, alone, above, utter, unfit, Othello, arrive, ump, us
tuck, sup, dug, of, what, muck, shove, rum, ton, gull, mull, chuck, love
pluck, drugs, crux, plump, grunge, shrunk, gloves, drubs, crumb, flux
duh, uh

“None of us can trust the government,” the drunken thug grudgingly

uttered.
Never trouble trouble till trouble troubles us.
His musk was unpleasant but his one tusk was unbearable.

Audio track 106: / ə / or / ʌ / targeting.

/ɝ/
earn, Irving, Ernest, irk, Erwin, Irma, early, urbane, erstwhile
lurk, pert, shirt, blurt, term, turn, swerve, turf, verve, stirred, work
purr, fur, shirr, myrrh, sir, cur, burr, her, grrrr, whir, chirr, aver

Sherman served his first term with courage.

Murray Zerbe earned mercy from the clergy.
The surly nurse cursed as the girl blurted her worst words.
 Audio track 107: /ɝ/ targeting.

/ɪ /
ear, earwig, earring, eerie, Erie, irritate, irreversible, irredeemable
Near, fear, tear, sere, sheer, weird, gear, cheer, leer, dear, peer
Spears, fleer, Greer, Sears, dreary, bleary, Trier, steer, clear

“Don’t interfere! Steer clear of here,” said Greer.

Cheers, dear! We’re drinking near beer.
The dreary, teary, bleary peer appeared with a beard.

Audio track 108: / Iɚ / targeting.

/ɛ /
air, ergo, errant, Ernani, Heiress
share, pare, fare, dare, care, wear, lair, hair, bear, tear, there, where
stair, Blair, flair, chairs, glare, Claire, prayer, agrarian, malaria, Astaire

Who cares where there is a pair of mares?

Blair stared and Cher glared.
At the bear’s lair, the heir shared the fare.

Audio track 109: / ɛɚ / targeting.

/ʊ /
Ur-Faust
Poor, sure, lure, boor, moor, dour, tour, Coors, your
McClure, manure, spoor, mature
The dour tourist was a boor.
You’re sure the spoor is from the manure on the moor?
McClure got the sure cure.

Audio track 110: / ʊ / targeting.

/ɔ /
oar, or, Orpheus, organ, orbit, orphan, Orly, orgone, ordain, Orson
Mort, fort, forge, course, pork, corn, lord, force, short, gorse, Morse
implore, engorge, stores, floor, drawer, chortle, corpuscle, vortex
pore, four, door, lore, gore, sore, shore, core, more, wore

Rory bore the force of the roaring nor’easter.

Morty Blore ignored the four short horses.
Gorgeous Lorna Corey ordered more corsets from the store.

Audio track 111: / ɔɚ / targeting.

/ɑ /
“r”, art, ardent, Arthur, arson, arch, arbitrate, archive, argue, Armani
part, cart, Mars, bark, card, large, charm, farce, park, darn, shark, hard
Killarney, alarm, Carmen, narcissism, varmint, barker, spark, stars
Czar, mar, par, bizarre, far, car, bar, scar, spar, gar, jar, tar

Arthur Hardy argued with the parson.

The bizarrely scarred czar parted with his marvelous art.
Charmaine Gardner was alarmed by the large “varmint”.

Audio track 112: / ɑɚ / targeting.

The targeting of the phoneme pronunciation should not be considered the
“correct” pronunciation. You should also be able to enact any of the myriad
allophone subtleties around this “essence” of the phoneme. It is useful to try
these lists again, exploring how far you can move away from the “essential”
version while still maintaining the identity of the phoneme, Of course, when
you do this, you are exploring accent variation and flexibility of
pronunciation, both of tremendous value to the actor.

Other vowel skills for American actors

It is the position of this book that allophonic variation of vowels is perfectly
acceptable as long as it does not interfere with easy intelligibility. For actors
there are also the considerations of appropriateness of accent for a play, as
well as the discovery of the unique vocal use of the character.
We all know, however, that many speakers of American English have
developed certain commonly shared pronunciation habits. These are not
negative in themselves – indeed they may be an important part of an actor’s
skills – but if unexamined they can keep the actor from attaining the
flexibility of articulation that should be her or his goal. Here, then, are a few
exercises to help the actor identify these habits and to free herself from
dependence on them at the same time that they can still be accessed as
needed.
[ ε ] or [ æ ] are often raised (or more closed) by speakers before a nasal
consonant in the same syllable or before an / r / in the next syllable. People
take this action with the tongue because they are anticipating, or
“prearticulating”, the nasal consonant or “braced r” that follows. While a
common characteristic of American speech appropriate for many accents, as
a habit it can become a problem if the listener hears the vowel as another
phoneme, such as “send” sounding like “sinned”, “panned” like “penned”,
“garish” like “gearish”, or “Farrell” like “feral”. Practice deliberately doing
this raising action on the vowel with the word list, then practice keeping the
vowel sound low. Make sure that the jaw does not raise on the vowel.

/ɛ/
them, hem, them, Lem, gem, end, send, bend, lend, enter,
renter, gender, sense, when, wren, blend, vent, Kent, pensive,
pending, offensive, blender, friend, Zen, when, yen, defend,
sentence, genuine, wrench
Benny sent Emma the seventy cents she’d lent him to rent a
pen-and-pencil set.
merry, cherry, merit, peril, ferris, Erin, sheriff, tarot, verity, Nehru
Errol merited the fairest share of the derelict’s inheritance.

Audio track 113: / ɛ /raising

/æ/
ham, jam, tram, cram, gram, ram, sham, dram, pram, rampant,
shambles, Gramercy, trample, flamboyant, amble, Campbell,
ran, fan, man, van, ban, hand, land, bland, grand, brand,
grant, plant, ant, rant, lance, chance, pants, advance, banter,
fancy, advantage, shanty, mantle

Sadly, Stan Cameron’s fancy banter can’t answer Hank

Crandall’s scandalous rantings from Paramus.
Aaron, barrel, carol, tarry, Karen, Paris, Harrods, Jared, narrative,
clarity, Laramie, parallel, barrister, polarity, Faraday, saraband,
charity, garrison, polarity, marry, baron
Clarence Darrow’s garrulous narrative parried the harridan’s arrogant
barrister.
Audio track 114: / æ / raising.

The vowel [ ɔ ] is often lowered (opened) to [ ɑ ]. For [ɔ ] be sure to protrude

the lip corners. Note: The merging of the vowel [ ɔ ] into the vowel [ ɑ ] is
characteristic of the majority of American speakers; linguists refer to it as the
“caught-cot merger”. Practice this list both ways, making the vowel
distinction between “caught” and cot”, and then using the merger to “cot”, so
that you are equally comfortable doing either one.

awe, alter, augment, awkward, all, Auden, Ausable, augury,

cause, paws, lawn, pawn, Paul, fought, bought, caught,
sought, taught, naught, wrought, lost, cost, frost, song, long,
wrong, soft, coughed, moss, dross, tall, call, stall, pall, fall,
brawl, crawl, trawl, shawl, cloth, broth, wroth, saw, flaw, craw,
raw, Shaw, gnaw, law, jaw, paw, draw
Though often called a costly, gaudy, awful loss, Broadway always
offers its all.

Figure 65. Smalltalk. (Cartoon by Bethany Carlson from SpecGram, CLX, No 4, January 2011,
reproduced with permission).

Audio track 115: / ɔ / lowering

[ ʌ ] or [ ə ] raised to [ ʊ ] before [ l ] or [ ɫ ]. Sometimes speakers achieve
the same effect by using a syllabic “l” [ ]. As always with this work, practice
the pronunciation with both vowel phonemes so that you feel equally
comfortable with both.

pulse, Tulsa, dull, gull, null, cull, hull, gulley, evulse, mull,
gullet, mullet, sullen, cruller, ulterior, ultimate, vulture, culture
as distinct from pull, bull, full, pulley, bullet
The sullen sultan indulged in a cruller with the dullard from
Tulsa.

Audio track 116: / ʌ / or / ə / raising.

26 Putting it all together

The texts that follow will allow you to explore your ability – and even more
important, your instincts – to vary the degree of linguistic detail
appropriately. There is a sort of progression to this chapter that will
encourage you to explore the development of your range of skills in
increasingly challenging contexts. Sometimes the challenges are informal,
sometimes formal, but the progression is from less connected structures to
more connected ones.
There are lots of things to explore in these texts. What makes a single
word clear to your listener? What makes a phrase and sentence clear? What
makes the larger structure of a set of sequential thoughts clear as your
intellect and feelings flow through and inform the shaping of these words?
What other factors vie with “clarity” in the actor’s mission of complete
communication with your listener? Why might “clear” or “clarity” be an
blurry way of describing what articulation skills must do and what they can
do?
If, as I assert, the only basic “standard” for speech skills is intelligibility
in any environment, a more generous use of these skills can take you further:
into a complex expressivity, whether it is as an actor, or as a speaker in any
other human interaction.
All the skills you have developed in your work thus far enable your
artistic impulses to find a variety of responses in the articulators that shape
speech. This book is about English; more specifically, American English. But
these skills will serve any accents of any language.
As an actor, but actually as a speaker generally, your only responsibility
is to serve the listeners who are hearing you in a variety of acoustic
environments in real time – no playbacks allowed. So the best way to play
with all these variables is to have a listener who is prepared to really listen
and to be really honest in her or his response.

Partnering
If you start as listener – listen carefully to your partner reading a few
sentences of any of the following texts. If you wish, take phonetic notes on
problem areas; that is, areas where your partner’s speech actions may
interfere with easy intelligibility. Try to perceive what in your partner’s
articulatory actions might need to be changed. Then refer to the drill words
and drill only those sounds that are causing difficulty. Return to the text and
see if your partner can fit the new actions into the text reading. Continue
through the text in this manner. When the text sounds appropriately detailed,
go on to another text or become the speaker.
At the beginning of this work it is very useful to switch from being
listener to being reader and back fairly frequently, until your stamina in each
role increases.

If you start as reader – focus on the physicality of the speech actions, on the
way the actions feel. Keep the actions specific, so that an increase in
muscularity doesn’t take you into a general residual tension within the
muscles of articulation. Take nothing for granted.
Feel also how your awareness of the ideas and images you are expressing
changes with this increase in linguistic detail. This is crucial: if you increase
linguistic detail as an exercise that is separated from the character, the
increase in “clarity” will happen at the expense of the clarity of thought and
intention in your acting.
Using the format above, go through these texts. You are never bound to
accept all the judgments of your partner, but bear in mind that she / he is a
crucially valuable “objective ear” listening to sounds that you have been
doing for years; time and habit often dull self-perception. Always bring your
awareness back to the physical: it’s less about how you sound than about
what you are doing.

For both of you – it is very useful to record portions of the readings and the
drill as needed. If possible there should be an audio recorder at every session.
Keep your diagnostic audio files with you and use the unrecorded portion for
drill practice. Listen to my reading only as a reference for specific sounds or
sound combinations.

Junk mail articulation

I actually received this verbal treasure as an e-mail, back in the days (say, ten
years ago from this writing) when cybercrime had to go to these extremes to
get a message through. Don’t ask me, or even your dictionary, what many of
these words mean. This text exists in that luscious area between nonsense and
comprehensible language. Have at it. Notice the periods, aka full stops: they
create an amusing artifice of sentences. Honor them. Pronounce the words as
you see fit, but speak your pronunciation with confidence and commitment.
Notice how some words really seem to want to attach themselves to the
words around them to form larger units of meaning.
precanceling catface imprudentness damson philharmonics wooler isogenous.
tewsome mayapples latitant pitsaw eupneic marijuanas discoloring outlearn
hyperresonant sappiest. linearization unsoluble. theandric crystallography
unhasty psychoanalytical plumped strowing biocentric filmsets jambolan.
undaub marijuanas wheelhorse onychophyma graminifolious bedazzle.
apometabolous tommed qoph gustation catechols undoctrinal misbill omelets
brevit armholes. phoenicians haha strand grabble kiblas oysterhood
pedagogues scrawler preparation dogfights. slopeness unglorifying
mosquitocide sages assisting leucites bolus contractedness hyperresonant
kreeps. jovilabe puerperalism drifting pillowslips apometabolous synclitic
individualize cicutoxin marascas simianity. intraleukocytic unsesquipedalian
buttocker lenticula enameler precooks gelated hooping decaspermal. merles
reflectionist marijuanas glaucophanite sphinges unsesquipedalian sparged.
extrapelvic pillowslips somatotyper subsulphid unpunishing resorbent.
fourcher putouts genuflex coannexes skirl pyritology breathingly. pinchback
tunnels simultaneousnesses whitecapper presympathize photonephograph
cashaws weaponries damson unstung. smeltery pst unhasty decivilization
trousering pyritology louty pneumotomy styptical periclitation enzymologist.
thraldom cathedralic moonset angarias lifo ultraenthusiastic oviduct felon
arachnophagous syllabled. gauntries redshirted mutt cosentient encephalitic
chariotway. cumbrance chondroitic catechols blennioid. enfettered when
snowmobile lenticula wardsman bulletproof. interavailable spiffing weighted
pullers tunnels citrin recommendable fasciculation acylogen. ultraenthusiastic
unperpetrated pyrosises programer unpermeated colliquation undergaoler.
contravention rehandling compunctionary aspartyl tastier sheather temser
nonregent. geognostical dicranoid vibracular mosquitocide ciboule lutianid.
feldspathization eyedropperful wineries rotundas anethol.

Now we move a step close to intelligibility supported by context. An actual

mobster of the 1920s and 1930s, Dutch Schultz – celebrated in story, novel
and feature film – is recorded stenographically on his deathbed. He has, or
had in his healthier moments, a full fluency in the English language, but a
bullet or two or twelve has gotten in the way. This text has actual syntactic
elements, so we’re moving toward easier comprehensibility. What do you
need to do to make the wisps of sense clear to your listener? What is starting
to tie the thoughts together? When do the incongruities tear that progression
apart? Don’t answer my questions intellectually: understand your impulses.

The last words of Arthur Flegenheimer

aka Dutch Schultz

Oh, oh dog biscuit. And when he is happy he doesn’t get snappy. Please
please to do this. Then Henry, Henry, Frankie, you didn’t meet him. You
didn’t even meet me. The glove will fit, what I say. Oh! Kai-Yi, Kai-Yi. Sure,
who cares when you are through? How do you know this? Well then, oh
cocoa know, thinks he is a grandpa again. He is jumping around. No hoboe
and phoboe I think he means the same thing … Oh mamma I can’t go on
through with it. Please oh! And then he clips me. Come on. Cut that out. We
don’t owe a nickel. Hold it instead hold it against him … How many good
ones and how many bad ones? Please I had nothing with him. He was a
cowboy in one of the seven days a week fights. No business no hangout no
friends nothing. Just what you pick up and what you need. This is a habit I
get. Sometimes I give it up and sometimes I don’t … The sidewalk was in
trouble and the bears were in trouble and I broke it up. Please put me in that
room. Please keep him in control … Please mother don’t tear don’t rip. That
is something that shouldn’t be spoken about. Please get me up, my friends,
please look out, the shooting is a bit wild and that kind of shooting saved a
man’s life Please mother you pick me up now. Do you know me? No, you
don’t scare me. They are Englishmen and they are a type I don’t know who is
best they or us. Oh sir get the doll a roofing. You can play jacks and girls do
that with a soft ball and play tricks with it. No no and it is no. It is confused
and it says no. A boy has never wept nor dashed a thousand kim. And you
hear me? … All right look out look out. Oh my memory is all gone. A work
relief. Police. Who gets it? I don’t know and I don’t want to know but look
out. It can be traced. He changed for the worst. Please look out. My fortunes
have changed and come back and went back since tha … They dyed my
shows. Open those shoes … Police mamma Helen mother please take me out.
I will settle the indictment. Come on open the soap duckets. The chimney
sweeps. Talk to the sword. Shut up you got a big mouth! Please help me get
up. Henry Max come over here. French Canadian bean soup. I want to pay.
Let them leave me alone.11

Junk mail syntax

Here is another gift to my computer’s inbox. It represents a divergence from
sense rather than an approach to it. Whoever wrote it was highly motivated to
compose a persuasive business letter – the better to scam the reader. But an
apparent unfamiliarity with certain characteristics of English creates
problems.
Can you surmount these difficulties by finding a confidence of expression
without changing a single word or the cues of any of the punctuations or
capitalizations? Go ahead: be eloquent. Make these millions of email
recipients send you that personal information.

Dear Winner

We Apologize, for the delay of your payment and all the Inconveniences And
Inflict that we might have indulge you through.

However, we are Having some minor problems with our payment system,
this is Inexplicable, And have held us stranded and Indolent, not having the
Aspiration to devote our 100% Assiduity in accrediting foreign payments.
We Apologies once again from the Records of outstanding winners due for
payment With {ONLINE CYBER PROMOTION} your name and Particular
was discovered as next on the list of the outstanding winners who are Yet to
received their payments.
Emails were selected anonymously through a Computer ballot system from
over 35,000 companies and 70,000 individual E-mail addresses all over the
world and your e-mail address emerged as the winner of the 11 selected email
address. This program is promoted and sponsored by Orient software
corporation (Orient Networks) in collaboration with The Online Cyber
International.
I wish to inform you now that the square peg is now in Square whole and can
be voguish for your payment is being processed and will be released to you
as soon as you respond to this letter. Also note that from our record in our
File, your outstanding winning payment is S$950.215.00 (NINE HUNDRED
AND FIFTY THOUSAND, TWO HUNDRED AND FIFTEEN DOLLARS).
Payment will be made to you in a certified bank draft or wire transfer into a
nominated bank Account of your choice, as soon as you get in touched with.
Mr. John Nicole
Provide him with the following details, as this will enable him to process and
release of your cash prize without any delay.
Your Full Name:.................................................
Telephone and fax Numbers:......................................
Residential Address:............................................
Your urgent reply will help him process the release of your price money.
Mr. John Nicole.
Will effect the speedy release of your cash prize to you within 7 working
days.
Yours Sincerely,
Mrs. Jane Phillips, Vice President

The Old Trouper

by Don Marquis
At last we have sense complete: the brilliant poetry written by “archy” (and
ghost written by journalist Don Marquis in the 1920s). Archy is a cockroach
who claims to be the reincarnation of a free-verse poet. He climbs atop Don
Marquis’s typewriter (if you need a reminder of what typewriters were, just
check the internet) and then hurls his exoskeletal body onto each key to type
one letter. Since he can’t hit the shift key simultaneously, everything he
writes is in lower case with no other punctuation. So what we have here is
sense that lacks the punctuation cues that help us to tie syntactic units
together. Your have to provide these syntactic elements by varying timing
and inflection as well as articulation activity.
One final comment: every American actor ought to know who Richard
Mansfield, Edwin Booth, Augustin Daly, Joseph Jefferson, Edwin Forrest,
and Helena Modjeska were. They shaped American theatre. They all had it
here. Mehitabel, by the way, is an alley cat who considers herself the
reincarnation of Cleopatra.

i ran into mehitabel again

last evening
she is inhabiting
a decayed trunk
which lies in an alley
in greenwich village
in company with the
most villainous tom cat
i have ever seen
but there is nothing
wrong with the association
archy she told me
it is merely a plutonic
attachment
and the thing can be
believed for the tom
looks like one of pluto s demons
it is a theatre trunk
archy mehitabel told me
and tom is an old theatre cat
he has given his life
to the theatre
he claims that richard
mansfield once
kicked him out of the way
and then cried because
he had done it and
petted him
and at another time
he says in a case
of emergency
he played a bloodhound
in a production of
uncle tom s cabin
the stage is not what it
used to be tom says
he puts his front paw
on his breast and says
they don t have it here
the old troupers are gone
there s nobody can troupe
any more
they are all amateurs nowadays
they haven t got it
here
there are only
five or six of us oldtime
troupers left
this generation does not know
what stage presence is
personality is what they lack
personality
where would they get
the training my old friends
got in the stock companies
i knew mr booth very well
says tom
and a law should be passed
preventing anybody else
from ever playing
in any play he ever
played in
there was a trouper for you
i used to sit on his knee
and purr when i was
a kitten he used to tell me
how much he valued my opinion
finish is what they lack
finish
and they haven t got it
here
and again he laid his paw
on his breast
i remember mr daly very
well too
i was with mr daly s company
for several years
there was art for you
there was team work
there was direction
they knew the theatre
and they all had it
here
for two years mr daly
would not ring up the curtain
unless i was in the
prompter s box
they are amateurs nowadays
rank amateurs all of them
for two seasons i played
the dog in joseph
jefferson s rip van winkle
it is true i never came
on the stage
but he knew i was just off
and it helped him
i would like to see
one of your modern
theatre cats
act a dog so well
that it would convince
a trouper like jo jefferson
but they haven t got it
nowadays
they haven t got it
here
jo jefferson had it he had it
here
i come of a long line
of theatre cats
my grandfather
was with forrest
he had it he was a real trouper
my grandfather said
he had a voice
that used to shake
the ferryboats
on the north river
once he lost his beard
and my grandfather
dropped from the
fly gallery and landed
under his chin
and played his beard
for the rest of the act
you don t see any theatre
cats that could do that
nowadays
the haven t got it they
haven t got it
here
once I played the owl
in modjeska s production
 of macbeth
I sat above the castle gate
in the murder scene
and made my yellow
eyes shine through the dark
like an owl s eyes
modjeska was a real
trouper she knew how to pick
her support I would like
to see any of these modern
theatre cats play the owl s eyes
to modjeska s lady macbeth
but they haven t got it nowadays
they haven t got it
here

mehitabel he says
both our professions
are being ruined
by amateurs
archy

Jesus Hopped the A Train

by Stephen Adly Guirgis

Here are two short pieces of dalogue by contemporary playwright Stephen
Adly Guirgis. (They are from the same scene, but are not sequential.) The
need for intelligibility is as acute as for any other form of text; and that
intelligibility must serve some listeners who may not be aware of all the
language use in this text. At the same time, there are obviously limits on the
degree of linguistic detail that can credibly be employed. How do you deal
with the spelling variations? Must they be followed exactly, or do they point
you in the direction of an overall oral posture for the speaker? Does the
language use of the character and the knowledge of the locale take you
toward an accent?

MARY JANE: Our jury, Angel, they want to acquit you because it’s right,
not because it’s wrong. They don’t think you deserve “Life” or anything
close to it. No one blames you for Reverend Kim’s death except the State
of New York! And what is that? It’s an institution! It’s a set of rules set
up to apply to each and every circumstance, as if they’re all the same.
They are not all the same. The jury knows that. And they will clear you
of all charges, from murder on down, because they understand what
happened here beyond the “technicalities” and they empathize. Not
because of your dazzling smile, but because, under the same
circumstances as you, they might have done the same damn thing
themselves. You made a statement, Angel. And they are going to back
that statement up. Your testimony will supply reasonable doubt. And
that’s all they want. And that’s what I think. Make sense?

ANGEL: We useta, me and Joey, we useta sneak out our house on Sunday
nights, jump the turnstiles. And we would hop down onto the subway
tracks, walk through the tunnels, lookin’ for shit, makin’ adventures,
playin’ like we was G.I. Joe … Pick up a empty can a Hawaiian Punch or
some ol’ beer bottle for fake walkie-talkies, and we’d have our snow
boots on so we could be astronauts. And we would pretend we were the
last two survivors on earth and that we came from the future … stupid …
the future … like in that Planet of the Apes movie with the two guys?
Only we had no weapons, juss chocolate milk. And we’d get so lost in
our games and our discoveries and our made-up stories … so many
stories: lookin’ for ghosts, lookin’ for apes, lookin’ for fortunes, runnin’
from rats, talkin’ ’bout girls, talkin’ ’bout Thelma from Good Times,
talkin’ ’bout daydreams, talkin’ ’bout Bruce Lee versus Evel Knievel.
Talkin’ in words that wasn’t even words … and … and it would always
surprise us when we saw the lights … even though we could feel the train
coming, but it was the lights. The closer those lights came, rumble of the
tracks, sound a the conductor’s horn blarin’ at us, we’d get so excited
we’d freeze – two seconds of freezin’ cold … hypnotized … holdin’
hands, waitin’, waitin’, then: Bang! We’d jump off the rails, hug the wall,
climb back up the platform, start runnin’ – runnin’ – tearin’ ass clear
 across back to Riverside or Cherry Park. One time … one particular time,
when we was holdin’ hands right before we both jumped off the rails,
somethin’ happened, and we couldn’t let go, couldn’t untangle ourselves
from each other, and we were inside that light, and … we both saw
skeletons and radiation, and we was paralyzed in a way that I juss can’t
explain, till somethin’ blew us apart, juss blew us, and we landed safe.
We didn’t move for a long time. We was cryin’, and Joey ripped his
brother’s coat … We wasn’t speakin’ till we got to our block and Joey
said that it was the light that ripped us apart and saved our lives … Joey
said, “Jesus hopped the A train to see us safe to bed.”

The Guerdon

a parody of Henry James by Max Beerbohm

Novelist and caricaturist Sir Max Beerbohm wrote this parody in the early
part of the twentieth century, one of several that he published on the style of
the famous novelist Henry James after learning that James was to receive the
Order of Merit from the King of England. The comic premise of the piece is
that neither the Lord Chamberlain, who is presenting the list of recipients of
this honor, nor the King himself, seems to know qute who Henry James is.
Welcome to this imitation in extremis of James’s often tortuous syntax.
Carry the thoughts through without losing vocal and intentional energy
somewhere in the middle. These are longer modes of expression than you
may be used to: accept the challenge to develop your communicative
stamina.

That it hardly was, that it all bleakly and unbeguilingly wasn’t for “the likes”
of him – poor decent Stamfordham – to rap out queries about the owner of
the to him unknown and unsuggestive name that had, in these days, been
thrust on him with such a wealth of commendatory gesture, was precisely
what now, as he took, with his prepared list of New Year colifichets and
whatever, his way to the great gaudy palace, fairly flicked his cheek with the
sense of his having never before so let himself in, as he ruefully phrased it,
without letting anything, by the same token, out.
“Anything” was, after all, only another name for the thing. But he was to
ask himself what earthly good it was, anyhow, to have kept in its
confinement the furred and clawed, the bristling and now all but audibly
scratching domestic pet, if he himself, defenseless Lord Chamberlain that he
was, had to be figured as bearing it company inside the bag. There wasn’t, he
felt himself blindly protesting, room in there for the two of them; and the
imminent addition of a Personage fairly caused our friend to bristle in the
manner of the imagined captive that had till now symbolized well enough for
him his whole dim blind ignorance of the matter in hand. Hadn’t he all the
time been reckoning precisely without that Personage – without the greater
dimness that was to be expected of him – without, above all, that dreadful
lesser blandness in virtue of which such Personages tend to come down on
you, as it were, straight, with demands for side-lights? There wasn’t a
“bally” glimmer of a side-light, heaven help him, that he could throw. He
hadn’t the beginning of a notion – since it had been a point of pride with
him, as well as of urbanity, not to ask – who the fellow, so presumably
illustrious and deserving chap in question was. The omission so loomed for
him that he was to be conscious, as he came to the end of the great moist
avenue, of a felt doubt as to whether he could, in his bemusement, now
“place” anybody at all; to which condition of his may have been due the
impulse that, at the reached gates of the palace, caused him to pause and all
vaguely, all peeringly inquire of one of the sentries: “To whom do you
beautifully belong?”
The question, however, was to answer itself, then and there, to the effect
that this functionary belonged to whom he belonged to; and the converse of
this reminder, presenting itself simultaneously to his consciousness, was to
make him feel, when he was a few minutes later ushered into the Presence,
that he had never so intensely, for general abjectness and sheer situational
funk, belonged as now. He caught himself wondering whether, on this basis,
he were even animate, so strongly was his sense of being a “bit” of the
furniture of the great glossy “study” – of being some oiled and ever so handy
object moving smoothly on casters, or revolving, at the touch of a small red
royal finger, on a pivot. It would be placed questioningly, that finger – and
his prevision held him as with the long-drawn pang of nightmare – on the
cryptic name. That it occurred, this name, almost at the end of the
interminable list, figured to him not as a respite but as a prolongment of the
perspirational agony. So that when, at long last, that finger was placed, with
a roll towards him of the blue, the prominent family eye of the seated reader,
it was with a groan of something like relief that he faintly uttered an “Oh
well, Sir, he is, you know – and with all submission, hang it, just isn’t he
though? – of an eminence!”
It was in the silence following this fling that there budded for him the
wild, the all but unlooked-for hope that “What sort, my dear man, of
eminence?” was a question not, possibly, going to be asked at all. It fairly
burst for him and blossomed, this bud, as the royal eye rolled away from his
into space. It never, till beautifully now, had struck our poor harassed friend
that his master might, in some sort, be prey to those very, those inhibitive
delicacies that had played, from first to last, so eminently the deuce with
him. He was to see, a moment later, that the royal eye had poised – had, from
its slow flight around the mouldings of the florid Hanoverian ceiling,
positively swooped – on the fat scarlet book of reference which, fraught with
a title that was a very beam of the catchy and the chatty, lay beside the
blotting-pad. The royal eye rested, and the royal eye even dilated, to such an
extent that Stamfordham had anticipatively the sense of being commanded to
turn for a few minutes his back, and of overhearing in that interval the rustle
of the turned leaves.
That no such command came, that there was no recourse to the dreaded
volume, somewhat confirmed for him his made guess that on the great grey
beach of the hesitational and renunciational he was not – or wasn’t all
deniably not – the only pebble. For an instant, nevertheless, during which the
prominent blue eye rested on a prominent blue pencil, it seemed that this
guess might be, by an immense coup de roi, terrifically shattered. Our friend
held, as for an eternity, his breath. He was to form, in later years, a theory
that the name really had stood in peril of deletion, and that what had saved it
was that the good little man, as doing, under the glare shed by his
predecessors, the great dynamic “job” in a land that had been under two
Jameses and no less than eight Henrys, had all humbly and meltingly
resolved to “let it go at that.”
From Parodies, an anthology ed. Dwight McDonald
(Random House, 1960)

Guerdon – reward
Colifichets – (Fr.) trinkets, knick-knacks

Coup de roi – (Fr.) kingly blow

Exercise in pronunciation
by William T. Ross (1890)

Next, we have a selection that carries formality to the point of absurdity. This
was considered overly elaborate speech even in the nineteenth century; and
many of these words, not to mention their pronunciation, would have been
unfamiliar to speakers more than a hundred years ago.
Naturally, you will want to throw some of these words into your next
casual conversation. As to their meaning, it you have any questions the
dictionary awaits. If you have any questions about any pronunciations, I have
provided a phonetic version following the orthographic version. The
pronunciations are American, not British, although in the case of proper
names of British persons or places, the pronunciations are British – perhaps
with a bit of an American accent.

One enervating morning, just after the rise of the sun, a youth, bearing the
cognomen of Galileo, glided in his gondola over the legendary waters of the
lethean Thames. He was accompanied by his allies and coadjutors, the
dolorous Pepys and the erudite Cholmondeley, the most combative aristocrat
extant, and an epicurean who, for learned vagaries and revolting
discrepancies of character, would take precedence of the most erudite of
Areopagitic literati.
These sacrilegious dramatis personae were discussing in detail a
suggestive address, delivered from the proscenium box of the Calisthenic
Lyceum by a notable financier, on obligatory hydropathy as accessory to the
irrevocable and irreparable doctrine of evolution, which has been vehemently
panegyrized by a splenetic professor of acoustics, and simultaneously
denounced by a complaisant opponent as an undemonstrated romance of the
last decade, amenable to no reasoning, however allopathic, outside of its own
lamentable environs.
These peremptory tripartite brethren arrived at Greenwich, to aggrandize
themselves by indulging in exemplary relaxation, indicatory of implacable
detestation of integral tergiversation and esoteric intrigue. They fraternized
with a phrenological harlequin who was a connoisseur in mezzotint and
falconry. This piquant person was heaping contumely and scathing raillery
on an amateur in jugular recitative, who held that the Pharaohs of Asia were
conversant with his theory that morphine and quinine were exorcists of
bronchitis.
Meanwhile, the leisurely Augustine of Cockburn drank from a
tortoiseshell wassail cup to the health of an apotheosized recusant, who was
his supererogatory patron, and an assistant recognizance in the immobile
nomenclature of interstitial molecular phonics. The contents of the vase
proving soporific, a stolid plebeian took from its cerements an heraldic
violoncello, and assisted by a plethoric diocesan from Pall Mall, who
performed on a sonorous piano-forte, proceeded to wake the clangorous
echoes of the Empyrean. They bade the prolix Caucasian gentleman not to
misconstrue their inexorable demands, whilst they dined on acclimated
anchovies and apricot truffles, and had for dessert a wiseacre’s
pharmacopoeia.
Thus the truculent Pythagoreans had a novel repast fit for the gods. On the
subsidence of the feast they alternated between soft languor and isolated
scenes of squalor, which followed a mechanic’s reconnaissance of the
imagery of Uranus, the legend of whose incognito related to a poniard
wound in the abdomen, received while cutting a swath in the interests of
telegraphy and posthumous photography. Meanwhile, an unctuous orthoepist
applied an homeopathic restorative to the retina of an objurgatory spaniel
(named Daniel) and tried to perfect the construction of a behemoth, which
had gotten mired in a pygmean slough while listening to the elegiac soughing
of the prehistoric wind.
From Voice Culture and Elocution
by William T. Ross (New York: 1890)
What follows is a fairly broad phonetic transcription of the passage above.
For general pronunciation of individual words, broad transcription is usually
good enough. Although the pronunciations are deliberately American when
there is a choice between an American and a British pronunciation, there may
still be variants from the way you might say a word. In reading it, pay special
attention to the stress diacritics, which – lest we forget – occur before the
syllable modified. There are also a few allowances in the phonetic notation
for fluency, so tiny informalities may intrude. In performance, see how
informal you can make it without (a) losing full expression of the character
who might talk like this and (b) losing intelligibility.
As a transcription exercise, you might want to try doing a narrow
(detailed) transcription, based on what you hear when another speaker
performs one or two sentences of the text.
Arthur the Rat
The test passage

This short cautionary tale was written by Professor W. Cabell Greet of

Columbia University in the 1930s. Its purpose was to create a text that
included all the sound combinations of American English. It is a really good
text – though by no means the only one – to explore accent variation, so you
may want to keep this text at your service when you record native speakers
for that new accent that you need to learn.

Once there was a young rat named Arthur, who could never make up his
mind. Whenever his friends asked him if he would like to go out with them,
he would only answer “I don’t know.” He wouldn’t say “Yes” or “No” either.
He would always shirk making a choice.
His aunt Helen said to him, “Now look here. No one is going to care for
you if you carry on like this. You have no more mind than a blade of grass.”
One day the rats heard a great noise in the loft. The pine rafters were all
rotten so that the barn was rather unsafe. At last the joists gave way and fell
to the ground. The walls shook and all the rats’ hair stood on end with fear
and horror.
“This won’t do,” said the Captain. “I’ll send out scouts to search for a
new home.”
Within five hours the ten scouts came back and said, “We found a stone
house where there is room and board for us all. There is a kindly horse
named Nelly, a cow, a calf, and a garden with an elm tree.” The rats crawled
out of their little houses and stood on the floor in a long line. Just then the
old one saw Arthur.
“Stop,” he ordered coarsely. “You are coming, of course.”
“I’m not certain,” said Arthur, undaunted. “The roof may not come down
yet.
“Well,” said the angry old rat, “we can’t wait for you to join us. Right
about face! March!”
Arthur stood and watched them hurry away. “I think I’ll go tomorrow,” he
said calmly to himself. “But then again, I don’t know. It’s so nice and snug
here.”
That night there was a big crash. In the morning some men, with some
boys and girls, rode up and looked at the barn. One of them moved a board
and he saw a young rat, quite dead, half in and half out of his hole. Thus the
shirker got his due.
Mondegreens
An exploration of intelligibility

“Mondegreens” are misunderstood song lyrics or lines of poetry. The origin

of the word “mondegreen” comes from a newspaper columnist named Sylvia
Wright, who introduced the term in the nineteen-fifties. In the words of
fellow journalist Jon Carroll:

As a child she had heard the Scottish ballad “The Bonny Earl of Murray” and
had believed that one stanza went like this:

Ye Highlands and Ye Lowlands

Oh where hae you been?
They hae slayn the Earl of Murray,
And Lady Mondegreen.

Poor Lady Mondegreen, thought Sylvia Wright. A tragic heroine dying with
her liege; how poetic. When it turned out, some years later, that what they
had actually done was slay the Earl of Murray and lay him on the green,
Wright was so distraught by the sudden disappearance of her heroine that she
memorialized her with a neologism.”

The most famous mondegreen is “Gladly, the cross-eyed bear” for “Gladly
the cross I’d bear.” Vying for second place are the Jimi Hendrix lyric,
“Excuse me while I kiss the sky” heard as “Excuse me, while I kiss this guy”,
and the Creedence Clearwater song “There’s a bad moon on the rise” heard
as “There’s a bathroom on the right.” Our pledge of allegiance (preceding the
1950s addition of “under God”) can be heard as:

“I pledge a lesion to the flag, of the United State of America, and to the
republic for Richard Stans, one naked individual, with liver tea and just this
for all.”

(Probably few of us today remember Richard Stans. Sic transit Gloria mundi,
or, as Sylvia Wright might have understood it, “Sick friends are gory on
Monday.”)
Some other mondegreen pairs
Midnight at the Oasis Midnight after you’re
wasted
Feliz Navidad Police naughty dog
My Baby Likes the Western Movies My Baby’s Like a Wet
Sock Moving
The girl with kaleidoscope eyes The girl with colitis
goes by
Is your figure less than Greek? Is your finger less
than clean?
The bright blessed day, the dark sacred night The bride blessed the
day, the dog said
goodnight
Doctor, lawyer, beggar-man, thief Dr. Laura, you pickled
man-thief
Taking care of business Tape it to a biscuit
Dirty deeds and they’re done dirt cheap Dirty deeds and
they’re done to sheep
Don’t it make my brown eyes blue Donuts make my
brown ice blue
Got a lot of love between us Got a lot of lucky
peanuts
Hold me closer, tiny dancer Hold me closer, Tony
Danza
Count the headlights on the highway Count the head lice on
the highway
I’m gonna break my rusty cage and run I’m gonna braid a
rustic Cajun rug
That deaf, dumb, and blind kid That deft thumb of
 lightning
In spite of my rage, I am still just a rat in a cage The spider marines,
Siam’s steel-chested
rabbit arcade
She pays my ticket when I speed She paints my chicken
when I sleep
My mind is racin’, but my body’s in the lead My mind is bacon, but
my body’s Sizzlean
All we are saying is “Give peace a chance” Oh, we are sailing,
yes, give Jesus pants.
It’s just a spring clean for the May queen It’s just the sprinkles
for the bakery
So don’t knock down my door Sew donuts to the
door
Two tickets to paradise Two chickens to
paralyze
Almost paradise All those parrot eyes
Viva Las Vegas People love bagels
The emperor’s new clothes The antlers are too
close
I wanna be sedated I want a piece of date
bread
Someone saved my life tonight Someone shaved my
wife tonight

With a partner as listener, choose a mondegreen pair from this list, or a pair
that you already know. Speak each version alternately with enough linguistic
detail so that there is no ambiguity or misunderstanding about any of the
words. But is the rendition fluent? Can you keep the ambiguity at bay while
saying the mondegreens easily and fluently? Then explore the modifications
in articulator action that start to blur the distinction between the two. Find the
actions that allow complete ambiguity. Explore all these possibilities in
different acoustic environments, e.g. outdoors/indoors, distant/close,
with/without background noise, et cetera.

Mondegreens from the books ‘Scuse Me While I Kiss This Guy, He’s Got the
Whole World in His Pants and When a Man Loves a Walnut, all compiled by
Gavin Edwards, Fireside Books (Simon & Schuster, 1995, 1996, 1997).

So ...
So keep exploring. Every possibility for speech use is open to you when you
passionately and intelligently explore your own natural actor’s curiosity
about how people speak. Your curiosity – freed from the restrictive mandates
of someone else’s idea of “good speech” – gives you complex and specific
skills to master and that mastery will be the work of a lifetime. Becoming
“perfect” at any speech pattern is a false goal because language when it is
used by real people in real life is always a flexible and malleable thing,
shaped by a multitude of factors. The good news is that by continuing the
journey of exploration I have outlined in this book you can become very,
very good at finding within yourself the unique voice of the character you are
playing. Or perhaps, if you are not an actor, your own unique voice. You’ll
keep developing a deeper and more productive relationship to the shaping of
speech. And that’s more than good enough.
Audio Tracks

Speaking with Skill was originally published with a supplementary DVD,

containing audio tracks referred to in the text. To listen to these tracks, please
follow the links below, or type their URLs into your browser.

Audio Track 1 http://youtu.be/3gL9wINewrE

Audio Track 2 http://youtu.be/tridj_22jOM
Audio Track 3 http://youtu.be/iZbzrD3dgWA
Audio Track 4 http://youtu.be/nU5OTLMElfs
Audio Track 5 http://youtu.be/g8XHAISmQQk
Audio Track 6 http://youtu.be/aIXJe8SQg5E
Audio Track 7 http://youtu.be/nwf86KtI4gI
Audio Track 8 http://youtu.be/lW0vKu9gv2c
Audio Track 9 http://youtu.be/xEFm9ATi0dE
Audio Track 10 http://youtu.be/lu10ET_ATlU
Audio Track 11 http://youtu.be/XGC3nxjyfas
Audio Track 12 http://youtu.be/AKD1dkiN-1Y
Audio Track 13 http://youtu.be/kamvc_jR8DU
Audio Track 14 http://youtu.be/rcqpG6yY-EM
Audio Track 15 http://youtu.be/ZLSJZDfKVtA
Audio Track 16 http://youtu.be/CKaWB8wmwP4
Audio Track 17 http://youtu.be/5RXGuzwQ64E
Audio Track 18 http://youtu.be/VclG5AdEcSQ
Audio Track 19 http://youtu.be/4VnDVHKqJL8
Audio Track 20 http://youtu.be/4-fkt6Je6lo
Audio Track 21 http://youtu.be/ghyk4JA9jVc
Audio Track 22 http://youtu.be/lofM_5mBm_k
Audio Track 23 http://youtu.be/EgkLDcizW0M
Audio Track 24 http://youtu.be/-cwrT-flWtM
Audio Track 25 http://youtu.be/T5MiQ-BXFcU
Audio Track 26 http://youtu.be/Q8LSXpxjckI
Audio Track 27 http://youtu.be/YqZN1rfQUCk
Audio Track 28 http://youtu.be/idQzuScLJqk
Audio Track 29 http://youtu.be/UDurZazoRaI
Audio Track 30 http://youtu.be/saglTm3H0xk
Audio Track 31 http://youtu.be/j0iK6C_jTQg
Audio Track 32 http://youtu.be/0m-p7uk6V2E
Audio Track 33 http://youtu.be/8PeiP9oWNSg
Audio Track 34 http://youtu.be/nylfC7b4sEQ
Audio Track 35 http://youtu.be/auQtfurPuHU
Audio Track 36 http://youtu.be/9L9XYAul15w
Audio Track 37 http://youtu.be/w3r_xV8vFk4
Audio Track 38 http://youtu.be/B2_FX7aP2yg
Audio Track 39 http://youtu.be/S1ZWTRRgssg
Audio Track 40 http://youtu.be/P_CwHtWxdj4
Audio Track 41 http://youtu.be/zsICJVQ5l1w
Audio Track 42 http://youtu.be/_tmvwSuWr94
Audio Track 43 http://youtu.be/xVjnxNolsnQ
Audio Track 44 http://youtu.be/O5oqVM8sUTQ
Audio Track 45 http://youtu.be/QmaSwh0VFkw
Audio Track 46 http://youtu.be/uvr85f38K-A
Audio Track 47 http://youtu.be/kJpMkSeQkWY
Audio Track 48 http://youtu.be/YTFn46AkSPo
Audio Track 49 http://youtu.be/E951wqT-XM8
Audio Track 50 http://youtu.be/LFA7VFzcdR8
Audio Track 51 http://youtu.be/7R-QcTYkYEY
Audio Track 52 http://youtu.be/CnAqYYvoyp0
Audio Track 53 http://youtu.be/CLXiBbD6DlY
Audio Track 54 http://youtu.be/6oaT-DFqdYc
Audio Track 55 http://youtu.be/IwUV-pERFKo
Audio Track 56 http://youtu.be/bw9PJYa_LUY
Audio Track 57 http://youtu.be/nBSOHwG-UTE
Audio Track 58 http://youtu.be/-edl4g2WxT0
Audio Track 59 http://youtu.be/9R9_qzh_JNI
Audio Track 60 http://youtu.be/uVSdSbR0Wvk
Audio Track 61 http://youtu.be/agH2v3ULr2c
Audio Track 62 http://youtu.be/JoDeNBAu7Jw
Audio Track 63 http://youtu.be/dxxw4g6qJ-E
Audio Track 64 http://youtu.be/PSg-0r46-qI
Audio Track 65 http://youtu.be/XSWmigveHuM
Audio Track 66 http://youtu.be/njoiT0V5zHM
Audio Track 67 http://youtu.be/h8-ctDp6C7w
Audio Track 68 http://youtu.be/zdk6jtS8-k4
Audio Track 69 http://youtu.be/CSJG2ANdWIE
Audio Track 70 http://youtu.be/uGw8CSWdmq8
Audio Track 71 http://youtu.be/4UPDf5Ub5yA
Audio Track 72 http://youtu.be/a88ZAaaumiY
Audio Track 73 http://youtu.be/Gfx8LpqL1pU
Audio Track 74 http://youtu.be/--c_4WhTJdk
Audio Track 75 http://youtu.be/eEgK9CH4yTw
Audio Track 76 http://youtu.be/KMw6fcVxQXA
Audio Track 77 http://youtu.be/UTjBEOlSo7Y
Audio Track 78 http://youtu.be/vVc2DQ6Udgw
Audio Track 79 http://youtu.be/xNeNRM3jttQ
Audio Track 80 http://youtu.be/gQcwq53NBg8
Audio Track 81 http://youtu.be/eVFYk9Nfn9s
Audio Track 82 http://youtu.be/9yVs4b04GXw
Audio Track 83 http://youtu.be/vbLyEmov9lg
Audio Track 84 http://youtu.be/bUbFC4zIegw
Audio Track 85 http://youtu.be/Xp15IvrB5X4
Audio Track 86 http://youtu.be/UvaOlv0BKBc
Audio Track 87 http://youtu.be/2uGhqtGlBmM
Audio Track 88 http://youtu.be/dffOBSXrIVk
Audio Track 89 http://youtu.be/o2SDXC5m1-Q
Audio Track 90 http://youtu.be/Iktxjfx1DdM
Audio Track 91 http://youtu.be/FSENKVHVAm4
Audio Track 92 http://youtu.be/fprXA2fAGfo
Audio Track 93 http://youtu.be/TmFHxMdmRQU
Audio Track 94 http://youtu.be/P9mA1McMacw
Audio Track 95 http://youtu.be/ThvdYPALsLg
Audio Track 96 http://youtu.be/6cCePtQutDA
Audio Track 97 http://youtu.be/kMgvPqUznn4
Audio Track 98 http://youtu.be/jrkdTSxTAFw
Audio Track 99 http://youtu.be/x7NaYP6zPvg
Audio Track 100 http://youtu.be/TCwv7Aee1HY
Audio Track 101 http://youtu.be/PzEOLs05qew
Audio Track 102 http://youtu.be/a2XFXc-fSoA
Audio Track 103 http://youtu.be/1nTe3ozu2cY
Audio Track 104 http://youtu.be/j5NqJY0gBOI
Audio Track 105 http://youtu.be/oTwYxtsDXzc
Audio Track 106 http://youtu.be/d_OkGU09K4c
Audio Track 107 http://youtu.be/dYEtc3kWoeA
Audio Track 108 http://youtu.be/0huNQ_RwvxU
Audio Track 109 http://youtu.be/AHiY2o8qJiM
Audio Track 110 http://youtu.be/v9q-oR_BaxI
Audio Track 111 http://youtu.be/U7lVf8hbVGA
Audio Track 112 http://youtu.be/vOaLWZMp7hQ
Audio Track 113 http://youtu.be/jVPqkcs1tmk
Audio Track 114 http://youtu.be/H-jWkDIqcUg
Audio Track 115 http://youtu.be/jf49Iwq3rDQ
Audio Track 116 http://youtu.be/Q0ZESLpckgA
Footnotes

1. This symbol does appear as the “long s” used in typesetting into the
eighteenth century, but the “long s” didn’t represent this sound.
2. In which case, one wonders why the pair of lateral fricatives, ensconced
comfortably in the pulmonic consonant chart, are not listed as “Other
Symbols” too. The influence of the powerful Welsh lobby, perhaps?
3. Designated as unvoiced by the IPA (the Association) but I question this.
4. Designated as alveolar by the IPA, but I consider it to be commonly
postalveolar in both American and British speech. The same is true of [ ɹ].
5. I vigorously disagree with this designation by the IPA.
6. It’s the sound we make when we say a “long e”, often spelled “ee” or – in
this case – “ea”. The phonetic symbols for vowels will arrive very soon;
do not despair.
7. “Minimal Pairs” are one-syllable word pairs where the difference in the
word meaning is defined by only one phoneme change.
8. Illustrative words in capital letters represent the “lexical sets” devised by
J. C. Wells in his book Accents of English. Rather than just trying to
communicate vocally a single vowel phoneme “I’m saying ’eee’,” it is
more accurate and convenient to say “I’m saying the ’fleece’ vowel.”.
9 The symbol [ ʝ ] is called a “curly-tail J” and is a voiced palatal fricative:
pronounced much like [ j ] but with the tongue tighter toward the palate.
Here it suggests the speaker’s heavy emphasis of the word.
10. Except for the last two, the sentences above are taken from The
Professional Voice, edited by Robert Sataloff, M.D. (Singular Publishing,
1992).
11. Transcribed in the hospital by F. J. Long, stenographer of the Newark
Police Department, after Schultz was fatally shot in October 1935.
Printed in Parodies edited by Dwight MacDonald, as an inadvertent
parody of Gertrude Stein.
Acknowledgements

For a book whose gestation has been as long as this one, I would have to
express appreciation to an unseemly number of people; so I dare not try to
name all the names at the risk of leaving someone out. As a general
statement, then, I am firstly indebted to everyone whom I have taught since
the early 1970s. That’s a lot of debt, but it’s also true. As soon as I carried
into the classroom my slightest suspicions about the way that speech was
being taught in the United States and in English-speaking countries generally,
I received fundamental guidance from the students with whom I was
working. They, not I, provided the environment, the keen interrogation, and
the excitement for exploring sound change within which my questions could
grow and answers slowly emerge. Further, every professional actor with
whom I have worked along this journey has provided new impetus to me,
either as a fellow actor or as a colleague in voice/text/dialect work in
professional productions. But I must express my particular gratitude to all my
students in the MFA acting program at the University of California, Irvine;
and to all the professional colleagues who have participated in the Knight-
Thompson ‘Experiencing Speech’ workshops. Their intelligent questions and
explorations of this work – and this text – as it has developed over the past
years have been the most immediate crafters of the contents of this book.
I am indebted to my colleagues Eli Simon, Richard Brestoff, Cam
Harvey, Cynthia Bassham, Keith Fowler, and Annie Loui in the Department
of Drama, Claire Trevor School of the Arts, at the University of California,
Irvine, for their support and friendship. I am particularly indebted to
Professor Robert Cohen for showing me through his own writings how well a
textbook can be constructed; and for being the most tactful of goads to me to
complete this work. Most of all, I thank my UCI colleague Professor Philip
Thompson for his close collaboration in all aspects of this book. This text,
and the work it represents, has been fundamentally shaped by his
extraordinary gifts as a teacher and colleague. The approach to speech and
accent training that we now call Knight-Thompson Speechwork
(www.ktspeechwork.com) is as much Philip Thompson’s creation as it is
mine. For the errors in this book, however, I proudly claim full credit.
 I am grateful to the editorial and production staff at Bloomsbury
Publishing, especially Claire Cooper, Judy Tither, and Simon Trussler, for
their unswerving professionalism and editorial acumen. Jenny Ridout has
guided this project from the start, and I owe her special thanks. I thank
Wendy Rasmussen for her excellent anatomical illustrations; and also the
UCI actors who posed heroically for the muscle isolation photos: Kyra
Zagorsky, Karin Jarnefeldt, Nathan Crocker, Tyler Seiple, Paul Culos and
Jennice Butler.
As a student myself, I am indebted to Kristin Linklater for shaking up my
ideas of what speech training should or should not be. I am most profoundly
grateful to Catherine Fitzmaurice, who has been – and who remains – my
teacher, friend, and mentor. I also appreciate the astute comments of my
colleagues in the Voice and Speech Trainers’ Association on the progress of
this work. Professor Charles A. Knight has provided a lifelong model of
excellence in scholarly writing.
Finally, and most of all, I am indebted to my wife, artist Marta Whistler,
whose contributions to this book, and to my life, surpass enumeration.
 First published in 2012

This electronic edition published in 2012 by Bloomsbury Publishing Plc

Bloomsbury Publishing Plc

50 Bedford Square
London WC1B 3DP
www.bloomsbury.com

All rights reserved

You may not copy, distribute, transmit, reproduce or otherwise make
available this publication (or any part of it) in any form, or by any means
(including without limitation electronic, digital, optical, mechanical,
photocopying, printing, recording or otherwise), without the prior written
permission of the publisher. Any person who does any unauthorised act in
relation to this publication may be liable to criminal prosecution and civil
claims for damages

Every effort has been made to trace copyright holders and to obtain their
permission for the use of copyright material. The publisher apologizes for
any errors or omissions and would be grateful to be notified of any
corrections that should be incorporated in future reprints or editions of this
work.

Copyright © Dudley Knight 2012

The author has asserted his rights under the Copyright, Designs and Patents
Act, 1988, to be identified as the author of this work.

A CIP catalogue record for this book is available from the British Library.

eISBN: 978 1 4081 5715 2 (e-book)

This book is sold subject to the condition that it shall not, by way of trade or
otherwise, be lent, resold, hired out, or otherwise circulated in any form of
binding or cover other than that in which it is published and without a similar
condition, including this condition, being imposed on the subsequent
purchaser.

Sign up for newsletters to be the first to hear about our latest releases and
special offers