New Techniques in IPA Training for Actors in the United States

PTLC2005 Dudley Knight, IPA for Actors in the USA 1
New Techniques in IPA Training for Actors in the
United States
Dudley Knight, University of California, Irvine
A Very Brief History The International Phonetic Alphabet has been an integral part of
speech and accent training of actors in the United States for most of the twentieth
century. Use of the IPA, as distinct from other phonetic transcription systems or diacritic
notation, became common in American speech texts shortly after World War I. The
textbooks that appeared in the nineteen-twenties were not directed primarily at
performers, but rather at a more general audience who studied diction as a part of a
general educational curriculum in primary and secondary schools. The center of this
activity was in the New York City public school system, where many of the speech
teachers studied phonetic transcription with William Tilly, who in earlier years had taught
Daniel Jones in his Institute near Berlin. Tilly had joined the faculty of Columbia
University’s extension program in 1918, focusing his teaching on the training of other
teachers and on the remediation of the foreign accents of the immigrant populations of
New York. His pedagogy combined a use of narrow phonetic transcription with a highly
prescriptive modeling of what he mandated as the proper speech sounds in English.
Crucial to his approach, and to the efforts of his followers, was a strong belief in the
inherent euphony of certain vowel and consonant sounds and the utility of the “World
English” speech pattern as a vehicle for upward social mobility. Indeed, one of Tilly’s
pupils, Marguerite DeWitt, titled the subject “Euphonetics” in her book EuphonEnglish in
America (1926).
Tilly’s preferred speech pattern was essentially the RP of the late nineteenth century.
The American teachers who worked with him had already developed a pattern, largely
based on American elocutionary training, that was a slight variant, allowing for prevailing
American pronunciation choices and opting for [a] rather than [ɑ] in the pat/path
distinction. Tilly’s principle contribution to the mission of these teachers was to provide—
with narrow IPA transcription—a methodology that imparted a supposedly scientific
basis to the highly prescriptive speech pattern they were already teaching. However, to
most American ears, the “Good American Speech” pattern still sounded like an English
accent. While it attained some considerable use amongst wealthy New Yorkers in the
period between the world wars, it was generally rejected by most American speakers.
Except in the theatre.
Based largely on the work of two of Tilly’s star pupils, Margaret Prendergast McLean and
Edith Warman Skinner, the Good American Speech pattern became the dominant
pattern for speech training in the American theatre. It was an easy fit. Since the
nineteenth century, American actors had customarily affected—with varying degrees of
accuracy—an English accent when performing the classical repertoire, especially
Shakespeare. It remained an unquestioned assumption in the American theatre until well
into the nineteen-sixties that all Shakespeare should be performed in RP, or something
very close to it such as Good American Speech, which was also the title of the book
written by Margaret Prendergast McLean in 1928. Skinner’s book, Speak With
Distinction, using McLean’s book as a template, became the dominant text for theatre
PTLC2005 Dudley Knight, IPA for Actors in the USA 2
speech instruction in the United States, and it remains so today in a revised edition by
Lilene Mansell and Timothy Monich (1990).
A New Approach The Good American Speech regimen, though it produces some
changes in actors’ speech use that undeniably are beneficial, is a vestige of a bygone
era in American theatre practice, but it is an anachronism that still dominates
professional theatre speech training programs in the United States. As just one example
of the resistance to change evidenced by McLean/Skinner teachers, the IPA symbols
are still being rendered in unconnected cursive script, a transcription form that has not
been used widely by phoneticians since the beginning of the twentieth century. The IPA
symbols taught to students are limited strictly to those phonemes that represent the
preferred speech sounds of “good American speech”, and allophonic variants of these
phonemes are not introduced at all, not to mention the many other symbols that
represent the sounds of the world’s languages. Despite this, students usually emerge
from this training with the notion that they have “learned the IPA”. Phonetic instruction,
therefore, exists solely as a vehicle to inculcate the preferred speech pattern and to
wean the student actors permanently away from what are designated as the
substandard patterns of their own regional dialects. Because McLean/Skinner speech
training takes, as its only pedagogical model, a specifically focused drill in this
prescriptive accent pattern, the accent can take up to two years of weekly classes for
students to acquire.
But the theatrical market for this approach may be waning. It is no longer considered
necessary—by most American regional theatres—that actors in classical plays speak
uniformly in a conventionalized class-based accent, spoken in no region of the United
States and devised by speech teachers nearly a hundred years ago. A growing numbers
of major training programs are—however belatedly—starting to consider new options.
At the University of California, Irvine, for the past twenty years within our professional
M.F.A. acting program, I have been exploring new possibilities for speech and accent
training that can serve more realistically the needs of professional actors today. It seems
to me that the primary focus of theatre voice and speech training should be to equip the
actor to find and express the unique voice of the character the actor is playing. This
unique vocal and articulatory creation cannot be forced into a single mold of “good
speech”, nor indeed can it be imposed on an actor by anyone prior to the act of creating
the role itself. It is, by its nature, created by the actor alone as it becomes, in effect, the
idiolect of the character. It is obvious to me that the training focus cannot be on the
acquisition of a single pattern of speech, or even the rote mastery of a limited set of
accents, but rather must be on the acquisition of a complex set of articulation skills
based on the student’s awareness of subtle changes in articulator shaping and the
sounds that result. The actor needs to be freed from dependency on her or his own
habits of articulation without losing the ability to come back easily to the original regional
or social accent if the need arises, whether in life or in art. In this training process the
IPA becomes a crucial tool for enabling the student to acquire this infinitely variable skills
set by focusing on the descriptive—as opposed to prescriptive—awareness of discrete
speech actions.
At UCI, we do not commence our work by teaching the IPA symbols. Students go
through several weeks of study and exploration of vocal anatomy and physiology,
including their direct observation through video-endoscopy of their own vocal actions in a
diagnostic protocol of speaking and singing tasks. The second phase of the work
PTLC2005 Dudley Knight, IPA for Actors in the USA 3
focuses on isolation of muscle action in the muscles that alter the shape of the vocal
tract. Students explore vocalizing freely, ingressively as well as egressively, as they
explore these gradations of muscular action, with no restrictions on the sorts of
articulator actions they attempt or the sounds that result, many of which are undeniably
and deliciously bizarre. Students play with these sound actions as isolated sounds and
also in connected phonation, producing a free-form nonsense language that I term
What the students perhaps do not realize, as they engage in this seemingly unstructured
exploration, is that they are starting to free themselves from the muscle memory that
pulls them habitually back into their own accents. They are also becoming aware of the
tremendous variety of human vocal sound possibilities and, more importantly, they are
learning to exist comfortably within a world of vocal sound that is not limited to a small
set of phonemes. They are, in other words, starting to free themselves from phonetic
Students are now invited to add a little more self-analysis to their exploration. They start
to distinguish areas of the vocal tract that make a greater contribution to vocal quality
and those that make a greater contribution to the shaping of vowel sounds, which I
prefer, at this stage, to term “phthongs”, both because it is a term free of the limiting
phonemic implications of “vowels” (and also because it is amusing to say). They start to
differentiate between obstruents that are inefficient physical actions to produce in
connected speech and those that can allow for fluent action into and out of other sounds.
They start to be aware of which actions produce clearly differentiated sounds and which
do not. So some of the most beloved sounds need to be retired to the admittedly
utilitarian world of cartoon character voices and rude noises, and those that remain have
at least the potential to become part of the world of language.
At this point the work begins to focus on consonant production. Students conduct an
initial self-exploration of possible degrees and forms of voiced or unvoiced obstruent
action and then explore the possible placements for those actions, always mindful of the
necessity that the resultant sounds be perceptually differentiated. It is only now that the
IPA pulmonic consonant chart is introduced. But it is devoid of symbols. The “empty
consonant chart” is an innovation of my UCI colleague, Professor Philip Thompson.
Within the “empty consonant chart”, all that we are considering are the possible
combinations of actions and placements that might fill many—though not all—of the cells
in the table. The focus is totally on the physical, not the notational, but the students
rapidly master the sensory recipe for each sound (unvoiced/voiced, action, placement).
Inevitably the students find that they can produce a fair number of consonant sounds
that do not happen to be utilized in any of the world’s languages (that we know of). But
they are all within the linguistic realm of possibility. We then add to this repertoire the
non-pulmonic consonant actions, not limited to placements as found in existing
languages but defined, of course, by physiology. Using this full array of consonant
resources, and with their completely free shaping of the infinite gradations of “phthongs”,
the students are now able to combine them into a connected speech form that I call
“Omnish”, since it can take the speaker through all the sounds used in language—and
then some.
Omnish is extremely useful for actors and, I would submit, for other speakers as well. On
the simplest level, it is an exceptionally good articulation warmup, since it is far more
PTLC2005 Dudley Knight, IPA for Actors in the USA 4
muscular than connected speech in any individual language. If actors do no more than
learn the skills of Omnish, they usually carry this more active and detailed articulation
into the speaking of English. Secondly, speaking Omnish continues to pull the students
away from a reliance on the familiar and the habitual in language use and invites them to
enjoy the enactment of language sounds that are vastly different from their own. Thirdly,
as students try to speak Omnish fluently, it becomes an excellent diagnostic tool to
pinpoint areas of residual muscle tension within the vocal tract.
In my experience, by this point in their training students are positively eager to learn the
phonetic symbols for as many of the sounds that they have produced as are contained in
language. The process is about to move from an expansion into ever-increasing
linguistic possibilities, inward into an analysis of the defining physicality of individual
languages, accents, and idiolects. To do this, the students need to notate what the
sound actions are. The preparation in perceptual and articulation skills that has gone
before means that students generally learn all the IPA symbols and diacritics within two
or three weeks, and another week of two of training in transcription of connected speech
gives them the ability to perform accurate narrow phonetic transcription of conversational
recordings of their colleagues in class.
The Detail Model All of the speech work during the first academic year (30 weeks) at
UCI is wholly descriptive, not prescriptive. At the start of the second year, students
embark on a short but intensive study of a variable model for American Speech that I
term the Detail Model. It is based on the commonly accepted precept in language study
that an appropriate—but not excessive—increase in the amount of linguistic detail (e.g.
full articulation of consonant clusters) in connected speech will make the speaker
intelligible to a wider group of listeners across a variety of regional or social accents. The
amount of detail that one might incorporate into one’s speech on stage is subject to a
constant (indeed, moment-to-moment) process of revision dependent on the perceptual
abilities of the listeners, the acoustics of the theatre space, the idiolect of the character,
the vocal commitment of the character as defined within the play’s action at any given
moment, the vocal health of the actor at that given performance, etcetera. All of these
variables combine and recombine from moment to moment, and the requirement for
linguistic detail will constantly change. The only constant “standard” is that the actor be
generally intelligible to the audience at all times. So the Detail Model is, like the work that
has come before, a set of skills that is in the control of the speaker. Quite obviously, it is
far different in concept and pedagogy from a fixed, inflexible pattern of “good speech.”
But I submit that it is of far more utility to a professional actor who may need to work in
an outdoor theatre in a classical drama one month and on a television set playing
contemporary naturalism the next.
Defining Accents Finally, students carry their skills of self-analysis into the process of
selecting those defining structures of muscular tension and relaxation that shape the
variety of sound actions in Omnish into the physical “posture” of a specific accent.
Because the students already know, and can control, the variables that comprise an
accent posture, they are able to employ linguistic findings in characteristic accent
sounds to help them define accent posture. And because they have a working
knowledge of the entire IPA, they are able to use the wealth of research in dialectology
to acquire new accents that they may not have studied directly during their period of
PTLC2005 Dudley Knight, IPA for Actors in the USA 5
DeWitt, Marguerite (1926) EuphonEnglish in America. New York: E. P. Dutton.
Skinner, Edith Warman (1990) Speak With Distinction. Lilene Mansell & Timothy Monich.
eds. New York: Applause Theatre Books.