You are on page 1of 31

International Journal of Transgenderism

ISSN: 1553-2739 (Print) 1434-4599 (Online) Journal homepage:

Clinical Aspects of Transgender Speech

Feminization and Masculinization

Shelagh Davies MSc and RSLP-C & Joshua M. Goldberg

To cite this article: Shelagh Davies MSc and RSLP-C & Joshua M. Goldberg (2006) Clinical
Aspects of Transgender Speech Feminization and Masculinization, International Journal of
Transgenderism, 9:3-4, 167-196, DOI: 10.1300/J485v09n03_08

To link to this article:

Published online: 17 Oct 2008.

Submit your article to this journal

Article views: 866

View related articles

Citing articles: 20 View citing articles

Full Terms & Conditions of access and use can be found at
Clinical Aspects of Transgender Speech
Feminization and Masculinization
Shelagh Davies, MSc, RSLP-C
Joshua M. Goldberg

SUMMARY. Societal norms of speech, voice, and non-verbal communication are often strongly
gendered. For transgender individuals who experience a mismatch between existing communica-
tion behaviours and felt sense of self, changes to the gendered aspects of communication can help
reduce gender dysphoria, improving mental health and quality of life. While peer resources are of-
ten beneficial in changing overall appearance and presentation, speech and voice modification is
best facilitated by a trans-competent speech professional. In this article we review clinical research
relating to transgender speech and voice change and discuss clinical protocols for trans-specific as-
sessment, treatment, and outcome evaluation. doi:10.1300/J485v09n03_08 [Article copies available
for a fee from The Haworth Document Delivery Service: 1-800-HAWORTH. E-mail address: <docdelivery@> Website: <> © 2006 by The Haworth Press, Inc. All
rights reserved.]

KEYWORDS. Transgender, speech, voice, gender transition, gender dysphoria

Transgender individuals may require assis- health and quality of life. With all parameters of
tance to feminize or masculinize speech, voice, communication, the goal is to allow the out-
and non-verbal communication. Changes to the side–speech, voice, and movement–to reflect
gendered aspects of communication can help what the client feels inside. While peer support
reduce gender dysphoria and facilitate gender resources can be highly beneficial in changing
presentation that is consistent with the felt overall appearance and presentation, speech
sense of self, resulting in improved mental and voice modification is best facilitated by a

Shelagh Davies, MSc, is a Registered Speech-Language Pathologist (Canada), Speech/Voice Consultant of the
Transgender Health Program and Clinical Instructor in School of Speech and Audiological Sciences, University of
British Columbia, Vancouver, BC, Canada. Joshua M. Goldberg is Education Consultant of the Transgender Health
Program, Vancouver, BC, Canada.
Address correspondence to: Shelagh Davies, c/o Transgender Health Program, 301-1290 Hornby Street, Van-
couver, BC, Canada V6Z 1W2 (E-mail:
This manuscript was created for the Trans Care Project, a joint initiative of Transcend Transgender Support &
Education Society and Vancouver Coastal Health’s Transgender Health Program, with funding from the Canadian
Rainbow Health Coalition. The authors thank Fionna Bayley, Katharine Blaker, Georgia Dacakis, and Murray Mor-
rison for their comments on an earlier draft, and Donna Lindenberg, Olivia Ashbee, A. J. Simpson, and Rodney Hunt
for research assistance.
[Haworth co-indexing entry note]: “Clinical Aspects of Transgender Speech Feminization and Masculinizatione.” Davies, Shelagh, and
Joshua M. Goldberg. Co-published simultaneously in International Journal of Transgenderism (The Haworth Medical Press, an imprint of The
Haworth Press, Inc.) Vol. 9, No. 3/4, 2006, pp. 167-196; and: Guidelines for Transgender Care (ed: Walter O. Bockting, and Joshua M. Goldberg)
The Haworth Medical Press, an imprint of The Haworth Press, Inc., 2006, pp. 167-196. Single or multiple copies of this article are available for a
fee from The Haworth Document Delivery Service [1-800-HAWORTH, 9:00 a.m. - 5:00 p.m. (EST). E-mail address: docdelivery@haworthpress.

Available online at

© 2006 by The Haworth Press, Inc. All rights reserved.
doi:10.1300/J485v09n03_08 167

trans-competent speech professional who can Jensen, & Law, 1999; Neumann & Welzel,
provide a comprehensive evaluation, design an 2004; Neumann, Welzel, Gonnermann, &
effective treatment program, and help prevent Wolfradt, 2002a; Oates & Dacakis, 1983,
vocal problems that may arise from changes to 1997; Pausewang-Gelfer & Schofield, 2000;
habitual fundamental frequency or voice qual- Van Borsel, De Cuypere, & Van den Berghe,
ity. 2001; Wiltshire, 1995; Wollitzer, 1994). Speech
As with other transgender care, we recom- masculinization for FTMs has not been as well
mend that speech services be offered in the con- studied (Oates & Dacakis, 1997; Soderpalm,
text of a complete approach to transgender Larsson, & Almquist, 2004; Van Borsel et al.,
health that includes comprehensive primary 2000). Generally, the literature surveyed stated
care and a coordinated approach to psychologi- that testosterone therapy always results in drop
cal and social issues. Speech services must be in pitch sufficient to allow FTMs to live as men
individualized based on the individual’s goals, (Andrews, 1999; de Bruin et al., 2000; Gold,
the risks and benefits of treatment options, and 1999; King et al., 1999; Oates & Dacakis, 1997;
consideration of social and economic issues. Petty, 2004; Soderpalm et al., 2004; Yang,
Palmer, Murray, Meltzer, & Cohen, 2002).
However, this is not empirically supported. A
THE IMPORTANCE OF SPEECH study of FTMs who had taken testosterone for
AND VOICE SERVICES at least one year (N = 16) found that 25% were
IN TRANSGENDER CARE sometimes perceived as female on the phone,
with 31% expressing interest in therapy to fur-
Although studies assessing transgender speech ther masculinize speech (Van Borsel et al.,
needs have thus far involved only small num- 2000). The speech needs of FTMs who do not
bers of participants,the results suggest that con- take testosterone were not discussed in any of
gruency of speech and gender identity is impor- the literature surveyed.
tant to transgender individuals in both the There is great variation in the extent to which
male-to-female (MTF) and female-to-male speech changes are undertaken or desired by
(FTM) continuum.1 For example, in a survey of transgender individuals. Some transgender per-
a broad range of transgender individuals (N = sons who desire changes to speech and voice
179)–including crossdressers and others who seek maximum feminization or masculiniza-
did not identify as transsexual–23% of respon- tion, while others experience relief with a more
dents reported a current need for speech therapy androgynous presentation. Some transgender
(Goldberg, Matte, MacMillan, & Hudspith, individuals seek to develop two speech patterns
2003). In studies of MTF transsexuals, femin- (one more masculine, one more feminine) ei-
ization of communication was rated extremely ther because they identify as bi-gendered or
important by 73% of 11 participants in one because external pressures relating to family,
study (Wollitzer, 1994), and as “very impor- employment, cultural community, or other
tant” by over half of 28 respondents in another concerns prevent living full-time in a way that
study (Neumann, Welzel, Gonnermann, & is consistent with felt sense of self. Most current
Wolfradt, 2002b). A study of FTM transsexuals transgender speech protocols do not support
reported that 88% of 16 participants considered bi-modal speech as a treatment goal, based on
masculinization of communication as impor- the belief that to effect maximalchange it is nec-
tant or more important than sex reassignment essary to have a consistent single speech pat-
surgery (Van Borsel, De Cuypere, Rubens, & tern. Switching back and forth between two
Destaerke, 2000). speech and voice patterns may be too difficult
In speech, voice, and transgender health lit- for some clients, and inconsistent use decreases
erature, speech feminization is widely recog- practice opportunities to acquire the new
nized as an important element of transgender speech and voice pattern. However, the human
care for MTFs (Andrews, 1999; Becklund- capacity to learn and speak more than two lan-
Freidenberg, 2002; Byrne, Dacakis, & Douglas, guages, develop a specific accent for an acting
2003; Dacakis, 2002; de Bruin, Coerts, & Greven, role, and develop a singing voice that is differ-
2000; Gold, 1999; Hooper, 1985; King, Lindstedt, ent than speaking voice suggests it may be pos-
Shelagh Davies and Joshua M. Goldberg 169

sible to develop bi-gender speech and voice. cal perspective, based on the evidence currently
We encourage clinicians to be open to this pos- available.
sibility and not to routinely exclude clients who In North America, many clinicians provid-
have two speech patterns as the treatment goal. ing transgender speech services do so in the pri-
We recommend that speech services be made vate practice setting, and data from clients in
available to the full spectrum of the transgender these settings are rarely published. To assist in
community. greater understanding of transgender speech is-
sues and further development of practice proto-
cols, we encourage clinicians in both university
EVIDENCE-BASED PRACTICE and private practice settings to ask transgender
IN TRANSGENDER SPEECH clients for permission to share anonymized as-
AND VOICE CHANGE sessment data with other speech professionals.
As with all research, it is important that
The practice recommendations in this article transgender clients’ involvement in research be
are based on published literature specific to fully voluntary–i.e., services should not be con-
transgender speech, supplemental interviews tingent on agreementto publish outcome data.
with four expert clinicians, and the authors’ The clinical process of feminization or
professional experience. As research in this masculinizationof speech and the voice is pred-
field is limited, some of our recommendations icated on the concept that there are “feminine”
are based on current practices or theoretical ra- speech and voice norms and “masculine” speech
tionale where the literature is inconclusive or and voice norms. Within linguistics there is
absent. generally recognition that norms of “feminine”
In our review of the literature we found a and “masculine” discourse and language are
paucity of evidence in the area of transgender socially determined phenomena that vary
speech, particularly in clinical practice. Early across cultures, regions, and historical periods.
clinical research reported single subject case There is less understanding of the ways that
studies (Bralley, Bull, Gore, & Edgerton, 1978; voice may also be shaped by social influences
Hooper, 1985; Kalra, 1977; Kaye, Bortz, & (Delph-Janiurek, 1999). This is most obviously
Toumi, 1993; Mount & Salmon, 1988; Yardley, problematic in studies which assert universal
1976); more recently small group studies have “female” and “male” speech and voice charac-
reported outcomes of speech therapy (Byrne et teristics based solely on study of English-lan-
al., 2003; Dacakis, 2000; Neumann et al., guage speakers, but in reviewing the literature
2002b; Soderpalm et al., 2004) and pitch-ele- we were also concerned by assumptions of ho-
vating surgery (Brown, Perry, Cheesman, & mogeneity relating to age, culture, class, re-
Pring, 2000; de Jong, 2003; Gross, 1999; gion, and social context between speaker and
Kunachak, Prakunhungsit, & Sujjalak, 2000; listener. While there is obvious value in consid-
Neumann et al., 2002b; Wagner, Fugain, ering existing empirical evidence relating to
Monneron-Girard, Cordier, & Chabolle, 2003; gender perceptions and attributions in develop-
Yang et al., 2002). However, further research is ment of speech feminization and masculiniza-
needed to evaluate specific techniques and pro- tion protocols, we believe it is misleading to
tocols. interpret the existing data relating to gender
In the literature and in our discussions with norms and voice as universally normative. In
clinicians we noted that decisions about prac- reviewing research findings we include discus-
tice protocols were often significantly im- sion of the limits of these findings.
pacted by budget constraints, the logistics of the
clinical setting (e.g., university-based student
clinics running from September to April), and CORE COMPETENCIES
protocols necessary for conscientious research OF THE SPEECH PROFESSIONAL
but not necessary in regular clinical practice. IN TRANSGENDER CARE
While there are administrative and logistical re-
alities that need to be considered, we felt it was While speech professionals do not need to be
important to base our recommendations on experts in every realm of transgender care to
what we felt to be optimal practice from a clini- work with transgender clients, the clinician

providing speech feminization or masculini- cols for assessment, treatment, evaluation, and
zation services is expected to have basic followup.
trans-competence. Trans-competency in clini- Some transgender individuals seek speech
cal services involves both the ability to interact services not to feminize or masculinize com-
in a respectful way with transgender individu- munication, but rather to address voice quality
als–sometimes termed cultural competence issues (such as hoarseness or raspiness follow-
(Kohnert, Kennedy, Glaze, Kan, & Carney, ing pitch-altering surgery), loss of singing
2003; Núñez, 2000)–and also clinical knowl- range following changes to habitual speaking
edge and skill relating to (a) speech and voice pitch range, or feelings of disconnection from
science, and (b) trans-specific assessment, the voice resulting from rapid hormonal or sur-
treatment, and outcome evaluation (Goldberg, gical changes. While clinicians working with
2006). transgender clients on these issues should be fa-
Cultural competence in transgender speech miliar with relevant trans-specific physical and
and voice change refers to the capacity to pro- psychosocial issues, the same clinical proto-
vide respectful and relevant services to a di- cols generally used to deal with these concerns
verse range of clients. In addition to general in other clients can successfully be used with
skill working with clients from a variety of cul- transgender clients–i.e., no special trans-spe-
tural, ethnic, class, and age groups, the clinician cific clinical protocols are needed. This article
is expected to be familiar with transgender ter- focuses on clinical protocols that are unique to
minology, diversity of gender identity and ex- speech feminization or masculinization, an
pression, the processes involved in gender tran- area that (unless associated with vocal pathol-
sition, and trans-specific psychosocial issues ogy) is considered trans-specific.
that shape clients’ goals and treatment options
(Goldberg & Lindenberg, 2001). The clinician
should also be aware of basic protocols such as TRANS-SPECIFIC PRACTICE
use of the client’s preferred gender pronoun and PROTOCOLS
name in verbal interactionsand written records.
Clinical competence in transgender speech Trans-Specific Speech Assessment
and voice change requires a solid foundation in
theory relating to adult speech and voice pro- The first step in transgender speech treat-
duction, speech and voice disorders, speech ment is a thorough assessment to guide the de-
and voice treatment techniques, and other ele- velopment of a therapeutic evaluation and
ments of speech and voice science. We recom- treatment plan. The following section dis-
mend that the speech clinician working with cusses recommendations relating to establish-
transgender individuals have at least two years ment of therapeutic rapport, recording of client
clinical experience assessing and treating typi- history and objectives, evaluation of speech pa-
cal adult speech and voice disorders prior to rameters, assessment of potential for change,
working with transgender individuals, as the determination of therapeutic goals, discussion
clinical processes of speech and voice femini- of therapeutic options, and preparation for
zation or masculinization require a high degree change. The additional evaluation required
of clinical sophistication. prior to pitch-elevating surgery is discussed in
There are few opportunities to obtain train- the section on surgical treatment protocols.
ing in speech and voice feminization or mas-
culinization, and many clinicians learn as they Building Therapeutic Rapport
work with transgender clients. This article is in-
tended to help clinicians who already have both The relationship between client and clini-
trans-awareness and experience in speech and cian begins with the first interactions. In initial
voice work to become more familiar with gen- sessions, the clinician is not only assessing the
der differences in speech and voice; the effects client, the client is also assessing the knowledge
of hormones and hormone therapy on speech and supportiveness of the clinician. A relation-
and voice; treatment options to feminize or ship grounded in mutual respect, trust, and gen-
masculinize the voice; and trans-specific proto- uine care for the client’s well-being facilitates
Shelagh Davies and Joshua M. Goldberg 171

open communication and encourages active Gelfer, 1999; Soderpalm et al., 2004). To assist
engagement in therapy; conversely, it can be in coordination of care, other health profession-
difficult to build therapeutic rapport if conflicts als involved in the client’s general and trans-
arise in initial sessions. Many transgender indi- specific care should be noted (Soderpalm et al.,
viduals have had negative experiences with 2004). Clients who present with difficulty
ill-informed or unempathetic health profes- swallowing, a dysphonic voice, or other symp-
sionals, and there may be wariness about enter- toms that may indicate voice disorder–such as
ing unreservedly into a relationship around vocal fatigue, loss of range, or throat discom-
communication–which is, by its nature, highly fort–should be referred for laryngological ex-
personal. amination (Hearing, Speech & Deafness Cen-
Because the assessment process sets the ter, 2005). All current medications, including
stage for all future interaction, it is extremely feminizing or masculinizing hormones, should
important to make the client feel respected and be recorded.
safe, and to create a feeling of positive anticipa- History of behaviours that may negatively
tion for the therapy process. Issues that speech impact speech, such as smoking (tobacco, co-
professionals need to consider in the intake pro- caine, marijuana, etc.) and drinking alcohol
cess include storage of information, privacy is- should be explored (de Bruin et al., 2000;
sues in setting appointment times, client name Pausewang-Gelfer, 1999). Because the stigma
preference, use of the client’s preferred pro- associated with substance use makes it difficult
nouns, and therapist bias and judgments about to get accurate information about current pat-
transgenderism (King et al., 1999). terns of use, it may be useful to ask if a client
“has ever . . .” rather than asking about current
Recording Relevant History behaviour at the original intake; this can be re-
visited as part of treatment planning.
Client history should include information History of behaviours that may positively
about both trans-specific concerns and also impact speech should also be explored. For ex-
general issues that are known to impact thera- ample, it may be useful to inquire about per-
peutic options and potential outcomes. While sonal, professional, and recreational use of
some transgender individuals are very comfort- voice (e.g., involvement in singing or acting) to
able talking about their history, others are more determine whether previous training could be
private. In some cases it may be appropriate to tapped during therapy. Previous attempts to
revisit sensitive questions after therapeutic rap- feminize or masculinize speech should be in-
port is well established, or to lead with general vestigated, including techniques used, dura-
questions unrelated to trans-specific issues. As tion of self- or professionally-directed ther-
with the general population, some clients re- apy, and the client’s subjective feelings about
spond well to informal intake (e.g., the question the outcome (Andrews, 1999; Dacakis, 2002;
“What brings you to see me?” may elicit a great Pausewang-Gelfer, 1999; Perez, 2004).
deal of information); in other cases a more Trans-specific history should include infor-
structured interview process or intake form mation about other feminization or mascu-
may be beneficial. Sample intake forms are linization treatments that may affect speech–
available as an online supplement at http://www. such as testosterone therapy in FTMs or facial feminization surgery in MTFs–and any noted
guidelines-speech.pdf. impact on speech following these treatments
As with any client presenting for speech ser- (Andrews, 1999; King et al., 1999; Pausewang-
vices, initial intake should include a general Gelfer, 1999; Soderpalm et al., 2004). It is not
medical history, with particular attention to his- necessary to inquire specifically about trans-
tory of nose or throat complaints, respiratory specific treatments that are unlikely to directly
ailments, hearing difficulties, voice disorders impact speech, such as history of chest, breast,
(including problems stemming from self-di- or genital surgery. Relevant areas to explore in-
rected attempt to modify voice or heavy use of clude: (a) consideration of the impact of any
voice), or any other conditions that could im- planned surgeries on the timing of speech ther-
pact speech (de Bruin et al., 2000; Pausewang- apy, (b) any factors relating to transition that the

client feels are important in terms of motivation dating, and embarrassing to be videotaped, and
and timing of speech therapy–for example, particularly to watch and discuss the tape with
wish to have speech change complete by a spe- the clinician. We recommend using videotape
cific date to facilitate job change–and (c) any only if the client is comfortable with this and
medical or psychosocial issues that the client there is strong clinician-client rapport.
feels may affect ability to engage in speech To gather objective data in an assessment, a
change (e.g., some transgender people report computer program that measures fundamental
changes to concentration and emotional lability frequency, intensity, and vowel formants is
as a side effect of hormone regimens). necessary. Kay Elemetrics’ Computer Speech
Lab 4300 and “Dr. Speech” (Tiger DRS Inc.)
Evaluating Current Speech Parameters were mentioned in the literature surveyed
Associated with Gender (Dacakis, 2002; Mount & Salmon, 1988;
Soderpalm et al., 2004). Free software pro-
Thoroughly assessing the client’s speech grams that measure fundamental frequency
gives a baseline against which to measure may be downloaded from the internet and can
change and provides information about which be useful for practice by clients who have
changes would be most useful (Andrews & computer access.
Schmidt, 1997; Kaye et al., 1993; King et al.,
1999). While voice parameters such as funda- Client’s Subjective Assessment
mental frequency and speaking frequency
range can be measured objectively, many Because the client’s goals for speech fem-
speech characteristics associated with gender, inization or masculinization relate directly to
such as melody and vocal timbre, cannot be ob- both self-perception and feelings about the per-
jectively quantified. A complete clinical im- ceptions of others, it is important to understand
pression should include the clinician’s objec- the client’s perspective and expectations in
tive and subjective findings, and also the both of these areas (Andrews, 1999; Dacakis,
client’s subjective assessment (Andrews & 2002; Oates & Dacakis, 1997; Pausewang-
Schmidt, 1997; Byrne et al., 2003; Coleman, Gelfer, 1999; Soderpalm et al., 2004). This may
1983; Dacakis, 2002; de Bruin et al., 2000; be done through informal discussion and/or
Gold, 1999; Mikos & Pausewang-Gelfer, formal measures such as standardized ques-
2001; Oates & Dacakis, 1997; Pausewang- tionnaire.
Gelfer, 1999; Pausewang-Gelfer & Schofield, If informal interview is the only tool used, to
2000; Wollitzer, 1994). facilitate later assessment we recommend that
Following standard practice in an evaluation the clinician use the same questions in pre- and
of speech and voice, audio recordings of the cli- post-evaluation. For example, the clinician
ent’s performance across a variety of tasks such could ask the client to describe three situations
as reading, picture description, and conversa- involving speaking that the client is dissatisfied
tion should be made. These recordings assist with, and three things the client would like to
the client and clinician in analyzing current change about the way she or he speaks in these
communication patterns, setting goals for ther- situations.
apy, and determining a baseline against which A standard speech questionnaire like the Vo-
to measure change. With the client’s permis- cal Handicap Index or Voice Symptom Scale
sion, the audio recordings may also be used as a (Wilson et al., 2004) can be modified to include
resource to train student speech professionals. trans-specific concerns. The Transgender Self-
If the clinician has access to digital technol- Evaluation Questionnaire, developed by the
ogy and the client feels comfortable being lead author, is available online at http://www.
videotaped, the assessment session can be
taped and the footage then reviewed with the guidelines-speech.pdf; the La Trobe Commu-
client. This may be a useful way of evaluating nication Questionnaire (Byrne et al., 2003) is an
non-verbal communication features such as example in the published literature.
gestures, movement, and facial expressions. Whether informal or formal assessment
However, many clients find it intrusive, intimi- tools are used, it can be informative to ask cli-
Shelagh Davies and Joshua M. Goldberg 173

ents to rate identity, self-perceived behaviour, tempting to conform to an external stereotype of

appearance, and speech on a masculinity/femi- femininity or masculinity (Becklund-Freiden-
ninity and male/female scale. This allows the berg, 2002; Oates & Dacakis, 1997). Finding
clinician to gain a clearer picture of the client’s this good fit requires introspection on the cli-
identity and also aids in discussion of the cli- ent’s part and an informed opinion about what
ent’s feelings about possible discrepancies be- is possible.
tween gender identity and gender expression
(Soderpalm et al., 2004). Clinician’s Evaluation
Concern about others’ perceptions often re-
lates to passability–being perceived by others Pitch. While there are several factors that to-
as a man or a woman. The desire to pass is a gether determine attributions of gender to a
complex feeling that may be influenced by the speaker, studies suggest that fundamental fre-
client’s self-defined gender; community norms; quency (F0) is primary in perception of a
beliefs and expectations of friends, family, co- speaker as male or female (Byrne et al., 2003;
workers, community peers, or others who are Coleman, 1983; Günzburger, 1993; Wollitzer,
close to the client; internalized transphobia; de- 1994). Normative data for male and female F0
gree of social support; and experiences of mis- vary across languages and dialects (Elert &
treatment (as individuals who are visibly Hammarberg, 1991; Graddol & Swann, 1989;
transgender are often more vulnerable to ha- Rose, 1991; Tom, 2004). Among English-lan-
rassment, discrimination, and violence). Be- guage speakers, the mean F0 for non-trans-
cause norms relating to social interactions and gender men and women overlaps from 145-165
speech are context-dependent, it is important to Hz (Oates & Dacakis, 1997). Studies of Eng-
know the context for speech that the client is lish-speaking transsexual women report that
particularly concerned about, such as em- bringing F0 into this range of overlap may not be
ployment or social relationships (Becklund- sufficient, by itself, to shift the gender percep-
Freidenberg, 2002). As the client begins chang- tion of listeners. For example, transsexual
ing speech and voice patterns, reactions of women with F0 of 145-160 Hz (i.e., within the
those close to the client (e.g. family, friends, co- “gender-neutral” range for English speakers)
workers, community peers) should be dis- are usually judged as male (Spencer, 1988;
cussed (Andrews, 1999; Dacakis, 2002; Oates Wolfe, Ratusnik, Smith, & Northrop, 1990).
& Dacakis, 1983; Pausewang-Gelfer, 1999). The primacy of F0 in perception of a speaker’s
For clients concerned with passability, the reac- gender in languages other than English was not
tions of strangers are important and these discussed in the literature we reviewed.
should be recorded either through informal es- Speech analysis software such as Kay
timate or formal means such as a diary. Elemetrics can be used to measure the average
While some transgender individuals may speaking pitch and pitch range across several
seek speech services because they have diffi- tasks (Dacakis, 2002; Mount & Salmon, 1988;
culty passing on the telephone or in face-to-face Soderpalm et al., 2004). Data should be re-
communication, others are more concerned corded in both hertz and semitones to facilitate
about reducing a perceived discrepancy be- clinical evaluation, using one of the readily
tween speech and identity. Assessing self-per- available conversion tables (Hirano, 1981).
ception relates to the fit that clients feel between The visual display of a software analysis pro-
their current speech and their felt sense of gen- gram can provide valuable information for a
der–i.e., how the client feels hearing herself or client about habitual and target average speak-
himself talk. The question of how well speech ing pitches, particularly in the context of dis-
fits with the client’s perception of self may be cussion about typical male and female speaking
easy for the client to answer right away, or it pitch ranges in the client’s primary language.
may come over time with experimentation, In addition to noting fundamental frequency
practice, and observation of role models. Both and frequency range, it is useful to note if the
the literature and the clinicians we interviewed pitch is higher (for MTFs) or lower (for FTMs)
discussed the importance of a “good fit” be- in a less complex task like reading than in spon-
tween the speech and the client rather than at- taneous conversation. If so, the client may al-

ready be consciously or unconsciously attempt- give guidance such as, “Listen to how your
ing voice feminization or masculinization. voice stays flat when you say . . .” or “Listen to
Although anatomy determines the upper and how your voice moves around when you say
lower limits of an individual’s pitch range, that. That is what we are looking for.”
there is evidence that F0 is largely dependent on Resonance. In the literature surveyed, the
social context (Hasegawa & Hata, 1995). Ide- term resonance was used to describe three dis-
ally, speaking pitch would be evaluated by col- tinct aspects of speech: (a) the effects of the vo-
lecting data in naturalistic situations common cal tract on the sound produced by the larynx
in the client’s day-to-day life, and also with a (formant frequencies), (b) the vocal quality that
variety of conversational partners. While this corresponds to the perception of vibrations in
wider baseline would be informative, it may be various parts of the body, or (c) the function of
impractical or prohibitively expensive to gather the nose as a resonator. There is empirical evi-
data in a public setting. dence that vowel formant frequencies signifi-
Intonation. Intonation (sometimes termed cantly influence the perception of English-lan-
“inflection”) is also considered important in guage speakers as male or female (Coleman,
gender perception, particularly when F0 is in 1983; Mikos & Pausewang-Gelfer, 2001;
the “gender-neutral” range of overlap between Pausewang-Gelfer & Mikos, 2005). Measuring
male and female norms (Becklund-Freidenberg, the “corner vowels” /i/, /u/, and /a/ may be par-
2002; Pausewang-Gelfer & Schofield, 2000; ticularly useful in assessing transgender speech
Spencer, 1988; Wolfe et al., 1990). In English, (Mount & Salmon, 1988; Spencer, 1988), as
women tend to be more variable in intonation these vowels represent the maximal range of
than men, generally using more upward glides formant frequencies in vowel productions in
and avoiding downward glides and level into- many languages (Titze, 1997).
nation patterns (Challoner, 2000; de Bruin et Among English-language speakers, vowel
al., 2000; Gold, 1999; Oates & Dacakis, 1997; formant frequency is estimated at 20% lower in
Wolfe et al., 1990; Wollitzer, 1994); as intona- adult men than adult women (Coleman, 1983;
tion varies significantly across languages, this Dacakis, 2002; Oates & Dacakis, 1983). The
should not be considered a universal norm. reasons for this are not clear, but a study of
Intonation patterns should be recorded using physiologically matched English, Hindi, and
speech analysis software at the same time that Mandarin male and female speakers (N = 40)
frequency is recorded. The visual display re- concluded that differences in format frequen-
cording can then be viewed with the client, to il- cies are due primarily to cultural and linguistic
lustrate patterns associated with gender–for ex- factors rather than sex-based anatomical differ-
ample, repeated and dramatic decrease in pitch ences (Andrianopolous, Darrow, & Chen,
at the end of a sentence is typically considered a 2001).
male speech pattern among English speakers, The role of the other types of “resonance” is
while variability in intonation is considered a less certain. Singers often refer to “chest reso-
more typically female pattern among English nance” as the full, rich sound that is produced in
speakers. Exaggerated intonation shifts may be lower notes and accompanied by a feeling of the
observed in some transgender women trying voice vibrating in the chest; “head resonance”
to mimic non-transgender women (Wollitzer, describes a brighter, forward sound that accom-
1994), and if present these should be pointed panies sensations of the voice ringing or reso-
out to the client. nating in the mouth, nose, sinuses, or upper part
It is also useful to make clinical judgments of the head. While some authors suggest that
about inflections during speech. Conversation among English-language speakers “chest reso-
or a sample of reading can be recorded, then nance” is associated with male speech while
played back with both the client and the clini- “head resonance” is associated with female
cian listening to the vocal inflections. During speech (de Bruin et al., 2000; Gold, 1999;
the subsequent discussion the clinician can as- Kujawski, 2003; Oates & Dacakis, 1997), there
sess the acuity of the client’s perceptions. If the is no empirical evidence that increasing the
client is unable to hear what the clinician per- subjective feeling of “head resonance” or de-
ceives to be important, the clinician can then creasing the subjective feeling of “chest reso-
Shelagh Davies and Joshua M. Goldberg 175

nance” increases the perception of MTF speak- for a laryngological examination if the voice is
ers as female (Dacakis, 2002). Further study is judged to be dysphonic.
necessary to see if using these perceptions in Articulation. Subjective impression may be
training voice production produces difference made about the quality of articulatory produc-
in vowel formant frequencies. tions. In the literature reviewed, several authors
Vocal intensity. Vocal intensity–the loud- observed that among English-language speak-
ness of speech–may be measured with a sound ers women tend to articulate more clearly than
level meter. In North America, the meter is usu- men but in a light manner, men tend to make
ally placed 30 cm from the lips; at this distance, harder articulatory contacts and “punch out”
norms are 68-76 dBA for adult males and 68-74 their words, men tend to drop final phonemes
dBA for adult females (Koschkee & Rammage, (e.g., “walkin” instead of “walking”), and men
1997). Despite the evidence that there is little tend to reduce or alter the production of some
sex-mediated difference in the loudness of ac- speech sounds such as voiced “th” (Andrews,
tual speech, in some regions it is a common ste- 1999; Gold, 1999; Oates & Dacakis, 1983,
reotype that women tend to speak more softly 1997). The literature includes subjective obser-
than men, and some clinicians include this vations about habitual lip, tongue and jaw posi-
speech parameter in assessment and treatment tions, without agreementabout correlationwith
planning (Andrews, 1999; Dacakis, 2002; de gender associations (Günzburger, 1993; Oates
Bruin et al., 2000; Günzburger, 1993; Oates & & Dacakis, 1983, 1997).
Dacakis, 1983). We recommend objectively Durational characteristics. Depending on
measuring intensity if the client reports it as a the durational characteristics of the client’s pri-
problem or if the clinician subjectively feels it mary language, it may be useful for the clinician
may be an issue. to observe whether the client sustains voicing
Some transgender individuals who are self- through speech sounds, words, and phrases, or
conscious about speech may adopt insufficient uses a more staccato speech style where words
and phrases are produced more separately. It
vocal intensity in an attempt to avoid public at-
has been suggested that in European languages
tention, or MTFs may speak quietly to try to women typically have a longer mean duration
“soften” the voice (Dacakis, 2002). This can re- of voicing during phrases and isolated words,
sult in difficulty maintaining desired speech and linger on occasional vowel sounds (An-
characteristics in situations where a higher vo- drews, 1999; Günzburger, 1993).
cal intensity is needed to counter high environ- Language and discourse. While there are
mentalnoise or to convey intensityof emotion. strong social stereotypes about gender norms
Voice quality. In English, most measures of and language (e.g., use of slang, size modifiers,
voice quality are not consistently associated and tag questions), gender-associated norms of
with categorization of voice as masculine or language and discourse are so dependent on an
feminine (Andrews & Schmidt, 1997). How- ever-shifting social context that findings from
ever, “breathiness” is considered a feminine trait studies done in past decades may not be reflec-
among English-language speakers (Becklund- tive of current patterns and trends (Becklund-
Freidenberg, 2002; Dacakis, 2002; Gold, 1999; Freidenberg, 2002; Oates & Dacakis, 1983).
Oates & Dacakis, 1997; Wollitzer, 1994). Additionally, there is strong interplay between
Voice quality is typically measured subjec- gendered language norms and norms relating to
tively according to the speech professional’s culture, class, sexual orientation, and age
acoustic impression, possibly with the use of (Graddol & Swann, 1989; Linville, 1998;
perceptual rating scales such as the Perceptual Moran, McCloskey, & Cady, 1995; Morris &
Voice Profile (Oates & Russell, 1997). Jitter Brown, 1994), so norms appropriate for one cli-
and shimmer data may be collected by a soft- ent would not be appropriate for another. If
ware acoustic analysis package to support the there are habits relating to modifiers, qualifiers,
clinicalimpression, but these parameters can be indirect versus direct speaking style, or other el-
hard to measure accurately, requiring a very ements of language that the client finds discom-
quiet space, rigid protocols, and finely cali- forting or that the clinician feels may contribute
brated equipment. The client should be referred to perceptions that don’t fit the client’s self-im-

age, we recommend that the clinician offer ent’s femininity or masculinity, but should also
feedback in these areas. be asked to judge whether the speaker was male
Rather than attempting to memorize lists of or female.
qualifiers or artificially adopt set phrases, we
recommend that modification of language and Assessing Potential for Speech
discourse be based on the client’s own observa- and Voice Change
tions of gender markers in the specific environ-
mental context of concern to the client (e.g., Clients vary in ability to achieve certain
work, home, cultural community, social set- pitches, match a target pitch, and follow models
ting). To facilitate the determination of contex- of intonation or articulatory productions
tually appropriate speech and voice norms, the (Becklund-Freidenberg, 2002; Byrne et al.,
client should be encouraged to weigh research 2003; Dacakis, 2000, 2002; Soderpalm et al.,
findings and the clinician’s suggestions against 2004; Van Borsel et al., 2000). Using an explor-
her or his lived experience. Clients with strong atory diagnostic process helps determine how
beliefs about “appropriate” language may ben- physically and psychologically easy or difficult
efit from clinician assistance to compare ste- it may be to effect change, and gives informa-
reotypical ideas of behaviour to the actual tion about the sort of intervention that may be
observed behaviour of peers. necessary in therapy. For example, if a client
Non-verbal communication. Norms relating has difficulty matching pitches auditorily, us-
to posture, gestures, and other non-verbal as- ing a visual pitch display will probably be nec-
pects of communication are strongly influ- essary. If the pitch range appears restricted, a
enced by cultural, class, and age norms. Gener- lower (MTF) or higher (FTM) frequency pitch
ally, in the dominant culture of North America, target would be more appropriate, and specific
maintenance of eye contact, increased smiling, exercises to increase vocal range such as
nodding and inclining toward others, increased Stemple’s vocal function exercises (Sabol,
use of hand and arm gestures, and occasional Lee, & Stemple, 1995; Stemple, Lee, D’Amico,
touching of the listener are associated with fem- & Pickup, 1994) should be considered. If the
inine communication patterns (Andrews, 1999; MTF client has a seamless transition into fal-
Gold, 1999). While it is not within the typical setto, some falsetto notes may be available for
scope of practice of a speech-language patholo- widening the upper range of vocal inflections.
gist to provide a detailed assessment of non-
verbal communication behaviours, anything Speaking Pitch
that is striking to the clinician or to the client
should be noted as part of the subjective evalua- There are a number of ways to explore aver-
tion. age speaking pitch and speaking pitch range.

Subjective Third-Party Evaluation 1. The client glissandos around in the upper

(MTF) or lower (FTM) range, without
In some cases it may be helpful to have one or moving into falsetto (MTF), then sus-
more naïve listeners provide subjective impres- tains a pitch and uses it to intone a word
sions of a recording of the client’s speech. This and short phrase. This is recorded and
may be useful when clients are particularly con- then evaluated for quality and ease of
cerned with passability, or when clients are un- phonation. This is repeated several times
able to appreciate changes that have taken throughout the range. The client is also
place. For example, one study found that MTF asked to intone words and phrases in
clients did not rate their speech as more femi- higher pitches (MTF) or lower pitches
nine following therapy, but observers did (FTM) to ensure there is room for vocal
(Soderpalm et al., 2004). To be considered inflections (Kujawski, 2003).
“naïve,” the listener should not be a speech cli- 2. An arbitrary target pitch is set by the cli-
nician or student, and should also not be famil- nician and the client matches it. Then,
iar with the client’s goals. If passability is the choosing pitches above or below that
goal, the listener should rate not only the cli- one, they decide on an initial target. It
Shelagh Davies and Joshua M. Goldberg 177

should be noted that this pitch is for prac- speech, the goal is an inflectional pattern that is
tice purposes only and can be changed at narrower but not “flat” sounding. If the pattern
any time. is consistent with the client’s goals relating to
3. A pitch that is one fourth of an octave feminine or masculine speech norms, this is
above (MTF) or below (FTM) the habit- noted; if not, the clinician can model a more
ual speaking pitch is set by the clinician consistent inflectional pattern and the client can
and the client matches it. The initial inter- copy it. The result is played back for the client to
val in Auld Lang Syne or Here Comes the hear the effect. This exercise gives information
Bride can be used to help the client under- on the client’s ability to hear and model vocal
stand the pitch change that is sought. inflections, and also gives the client feedback
4. The clinician models a frequency within about how the voice may sound if a different
the lower range of female norms (MTF) inflectional pattern is adopted.
or upper range of male norms (FTM) on a
visual display in a computer voice analy- Other Parameters
sis program, and the client produces a
pitch that stays above (MTF) or below Changes to other parameters such as tongue
(FTM) it. carriage, articulatory productions, vocal qual-
5. The client says a phrase in her most femi- ity, and vocal loudness can be considered if ei-
nine (MTF) or his most masculine (FTM) ther the client or clinician thinks they may be
voice. important to address. In the initial session, use-
ful information can be gleaned by exploring a
For MTFs with low pitch, diagnostic therapy number of speech/voice parameters from the
should be done to see if facilitation techniques trans-competent clinician’s repertoire.
enable higher pitches. Using sounds that facili-
tate efficient vocal fold vibration such as /m/, /z/, Assisting the Client to Determine
lip trilling, and tongue trilling, the client pho- Therapeutic Goals
nates in a higher pitch, either randomly or
matching a pitch set by the clinician. The goal is To help the client determine fully informed,
to produce the sound at the desired pitch with- considered, and achievable therapeutic goals, it
out any feeling of strain in the throat, emphasiz- is useful for the clinician to provide a synopsis
ing a strong feeling of vibrating or buzzing in of the client’s baseline speech and voice char-
the front of the face. For FTMs with high pitch, acteristics, physiologic limitations and esti-
it may be useful to explore facilitation tech- mated potential for change, and an informed
niques that give sensations of ease and reso- professional opinion about the parameters that
nance in the lowest register of the voice. would be beneficial to address to achieve the
If the client is unable to produce a higher or client’s stated objective (Dacakis, 2002; Hooper,
lower pitch without throat sensation or fatigue, 1985; Neumann et al., 2002b; Oates & Dacakis,
the clinician may want to start with some stan- 1983, 1997; Pausewang-Gelfer, 1999; Wollitzer,
dard voice therapy exercises to reduce inappro- 1994). For example, if an English-speaking
priate habits. Referral to an otolaryngologist MTF client presents with the primary concern
may also be indicated. that her voice is not perceived as female, it may
be appropriate to target a higher fundamental
Inflections frequency if her habitual speaking pitch is 100
Hz; if her average pitch is higher than 150 Hz, it
A short sentence is read by the client and ex- may be more appropriate to target resonance,
amined for its inflectional variation. The target inflection, and other speech characteristics that
is an inflectional pattern consistent with the are believed to have a greater influence on gen-
gender norms for the client’s language. For der perception when pitch is above the Eng-
English-speaking MTFs who are seeking to lish-language norms for male speech.
feminize speech, the goal is an inflectional pat- Table 1 summarizes aspects of speech that
tern that is wide but still natural-sounding; for are associated with sex and gender attribution,
FTM English-speakers seeking to masculinize and associated English-language norms. Norms

TABLE 1. English-Language Norms of Speech and Voice Associated with Gender

Considered Highly Salient to Gender Attributions

Element of Speech Female/Feminine Norms Male/Masculine Norms
Pitch Mean =196-224 Hz, range = 145 Mean = 107-132 Hz, range = 80
Hz-275 Hz; higher upper and lower Hz-165 Hz
limits of range

Formant frequencies Higher Lower

Intonation More variable in intonation, more More level intonation, more downward
upward glides glides

Weaker Evidence to Support Role in Gender Attributions

Element of Speech Female/Feminine Norms Male/Masculine Norms
Loudness 68-74 dBA 68-76 dBA

Breathiness Perceived as mildly breathy, softer Not perceived as breathy, harder

speech onsets speech onsets

Articulation Clear, light Forceful onsets; dropped phonemes,

reduced use of voiced “th”

Duration Longer mean duration of phrases and Staccato speech style

isolated words, lingering on vowels

should be considered as a spectrum rather than gical voice change, and recommendations to
two isolated poles, to encourage speech profes- prevent vocal fatigue or voice disorder (Dacakis,
sionals and clients to carefully consider thera- 2000, 2002; Hooper, 1985; Kaye et al., 1993;
peutic goals that fit with sense of self. Kunachak et al., 2000; Oates & Dacakis, 1997,
1983; Thomas, 2003; Yang et al., 2002).
Assisting the Client to Understand Because changes to specific acoustic voice
Therapeutic Options characteristics affect numerous perceptual vari-
ables, a well-rounded speech treatment plan
Some transgender individuals have sophisti- will target “constellations of related voice char-
cated knowledge about gender-related speech acteristics rather than independent acoustic vari-
parameters and therapeutic options, and come ables” (Wollitzer, 1994, p. 99). For example,
to the initial assessment with a clear direction raising pitch may increase laryngeal tension
they wish to pursue. Others have no knowledge and vocal tract constriction, influencing shim-
and expect guidance from a professional. Dur- mer, jitter, signal-to-noise ratio, and resonance
ing the initial evaluation it is important to assess (and thus subjective perceptions of voice qual-
the individual’s knowledge of speech and ity). For this reason, an optimal speech therapy
voice. Consumer education materials have program should target all parameters of speech,
been developed as part of the Trans Care Pro- not just those related to pitch.
ject to help promote consistent and accurate in-
formation about transgender speech change Preparing for the Process of Speech
and treatment options (Davies & Goldberg, Modification
2006). In all cases, care should be taken to en-
sure that clients understand potential benefits Speech feminization or masculinization is a
and risks relating to both non-surgical and sur- long process requiring considerable work on
Shelagh Davies and Joshua M. Goldberg 179

the client’s part. While therapy outcomes can- exercises to increase flexibility of voice pro-
not be predetermined, the estimated amount of duction (Dacakis, 2002).
daily practice time and expected duration of the Enhanced observation and awareness of
course of therapy should be discussed, as speech patterns of self and others. While trans-
should the factors that can influence the course gender individuals are often highly skilled at
of therapy (Byrne et al., 2003). As changing observing others, practice may be needed to un-
speech requires altering deeply ingrained derstand, observe, and analyze the specific
communication habits and behaviors that can components of speech (Hooper, 1985).
be difficult to modify, it may be useful to use Changes to speech. Average speaking pitch,
the “Stages of Change” model (Prochaska, pitch range, inflections, formant frequency,
DiClemente, & Norcross, 1992; Zimmerman, breathiness, loudness, articulation, tongue posi-
Olsen, & Bosworth, 2000) or other behavioral tion, language, facial expressions, and gestures
change tools to assist in anticipating and ad- may be targeted to feminize or masculinize
dressing barriers to implementing change. speech (Bralley et al., 1978; Brown et al., 2000;
If pitch-changing surgery is sought, there Dacakis, 2000, 2002; de Bruin et al., 2000;
should be discussion of the parameters of Gold, 1999; Hooper, 1985; Kalra, 1977; Kaye
speech that may still need work after surgery, et al., 1993; Kujawski, 2003; Mount & Salmon,
such as intonation and format frequency. Clients 1988; Oates & Dacakis, 1997; Pausewang-
should also be informed of the estimated heal- Gelfer, 1999; Soderpalm et al., 2004). Specific
ing time involved and the time required to stabi- objectives relating to voice modification de-
lize the new pitch (Dacakis, 2002; Neumann et pend on what is feasible to produce without
al., 2004; Wagner et al., 2003). strain, what fits with the client’s self-image,
and how important passability is to the client;
Treatment Options and Techniques some clients may be comfortable with gen-
to Feminize or Masculinize Speech der-neutral speech, while others will want to
and Voice aim for a voice that is perceived by listeners as
male or female. For clients who are concerned
Non-Surgical Treatment about “fitting in” or about passability, rather
(Speech Therapy) than adopting an artificial set of speech norms it
is recommended that clients observe communi-
Speech Therapy Goals cation patterns in their social, cultural, and
work environments to develop a context-spe-
As discussed earlier, we recommend that the cific set of norms (Oates & Dacakis, 1997).
clinician assist the client to determine therapeu- Prevention of vocal fatigue. Use of the vocal
tic goals, recognizing that transgender individ- tract in non-habitual ways can cause strain. Im-
uals have diverse identities and objectives re- portant therapeutic goals are the maintenance
garding feminization or masculinization and of efficient and easy speech, establishing ap-
that the clinician should not be directive in pro- propriate practice, and informing the client
moting specific goals. The range of therapeutic about how best to maintain vocal health
goals may include any or all of the following. (Dacakis, 2000, 2002; Gold, 1999; Kaye et al.,
Speech assessment, information, and other 1993; Mount & Salmon, 1988; Oates &
preparation for speech therapy. Some clients Dacakis, 1983, 1997; Soderpalm et al., 2004).
are interested primarily in a speech assessment
and a professional opinion on what would be in- Treatment Format
volved in changing elements of speech. Infor-
mation about therapeutic options can help with Traditionally, speech therapy has empha-
decisions regarding the timing of gender transi- sized one-to-one work to facilitate the person-
tion. One program described in the literature of- alized intervention necessary to modify and
fered three to four introductory sessions that monitor change in target behaviours. However,
provided information about gender differences speech therapy groups–typically comprised of
in communication, information about vocal hy- four to six clients–are commonly used to work
giene and prevention of voice disorders, and with specific populations (e.g., individuals

with aphasia, people recovering from traumatic Role-playing is more easily done in a
brain injury, clients with fluency disorders). group, and the opportunity to observe
Group therapy can facilitate peer support and others can give valuable insight into par-
encouragement, and reduce self-consciousness ticipants’ own practice. Additionally, the
that may be experienced when the client is group provides a safe setting to learn lis-
working alone with the therapist. tening skills, and to practice observing
It has been our experience that both individ- speech in a way that will not be intrusive
ual and group therapy are important compo- in a real-world setting.
nents of transgender speech care. We recom-
mend that both formats be made available, with The necessary repetition of training exer-
the option for a client to take part in either or cises can be done in a group as long as the thera-
both depending on therapeuticneeds and goals. pist is able to monitor the progress of all the
Components of a transgender speech ther- participants and give individual input and feed-
apy program that can be done well in a group in- back as required. The group can be divided into
clude: pairs to give practice time in both talking and
1. Education and information. Clients un-
Some interventions require one-to-one work
dergoing speech feminization or mas-
with a therapist, including: (a) determining ap-
culinization need to understand how the
propriate target pitch, (b) training target pitch if
voice is produced; how physiological dif-
ferences in male and female voice pro- the individual has difficulty matching pitches
duction system affect the voice and auditorily, (c) significantly changing individ-
listener perception; physiologic and so- ual characteristics associated with “feminine”
cial norms relating to gender and speech; or “masculine” speech, and (d) individualized,
treatment options, outcomes, and risks; specific input on anything the individual has
and techniques to prevent strain associ- difficulty understanding or doing in the group
ated with voice change. While some setting. Individualized input is especially im-
transgender individuals are extremely portant in training an efficient voice that is re-
well-informed about speech, others have sistant to vocal fatigue or dysphonia.
no knowledge or have been exposed to
inaccurate information via the internet or Length of Treatment Time
peer groups.
2. Discussion. Group format is ideal for par- Treatment time varies greatly depending on
ticipants to share observations, insights, the degree of change sought, the client’s vocal
and practical advice. In the Changing abilities, and psychosocial issues. There is no
Keys speech and voice feminization professional consensus on the optimal length of
groups developed by the lead author (dis- treatment for maximal treatment efficacy. One
cussed later in this article), participants study reported a modest correlation between
have commented on how useful they the number of therapy sessions and mean pitch
found these discussions. achieved at the end of therapy (Dacakis, 2000);
3. Speech therapy exercises. There are sev- however, another reported that client satisfac-
eral advantages to using a group setting to tion was not related to the number of therapy
offer those components of a therapy pro- sessions, and that clients tended to become frus-
gram that are required by all individuals. trated and discouraged when therapy continued
These would include relaxation exer- over a long period of time (Soderpalm et al.,
cises, basic exercises in efficient vocal 2004). It has been our experience that treatment
technique, and ear training (using listen- generally ranges from a minimum of 15 hours
ing exercises to train a heightened per- to a maximum of 1 year of weekly sessions, and
ception of differences in speech). For that shorter, more intensive treatment times en-
individuals who feel self-conscious about courage motivation and accommodate changes
doing speech exercises, participating in a to life circumstances more readily than pro-
group can have a normalizing effect. longed treatment.
Shelagh Davies and Joshua M. Goldberg 181

Psychosocial adjustment is an important part 1. Imitation of non-transgender people ob-

of changing speech. Participants may require served in daily life (de Bruin et al., 2000;
time to get in touch with what sort of voice best Gold, 1999; Hooper, 1985; Kujawski,
matches the person within. This is by necessity 2003; Mount & Salmon, 1988; Neumann,
a process that takes time and professional input 2000b; Oates & Dacakis, 1997). This in-
as to what is possible. Many transgender indi- put from the real world is useful in help-
viduals begin with the goal of having a pitch ing clients develop spontaneous speech
that is unrealistically high (MTF) or low habits that “fit” in their particular com-
(FTM); only with experimentation and practice munity.
will it become apparent that this is probably not 2. Progressively complex practice while
achievable, necessary, or even desirable. Addi- maintaining good voice quality (Hooper,
tionally, it can take time to feel that an altered 1985; Kujawski, 2003; Mount & Salmon,
voice is an authentic expression of self rather 1988; Oates & Dacakis, 1997; Pausewang-
than an artificial “mask.” If psychosocial issues Gelfer, 1999). Integration of pitch, pitch
are significantly impacting treatment, referral range, and inflections is typically done in
to a trans-competent mental health professional progressively complex practice (vowels,
may be useful. monosyllabic words, phrases, sentences;
reading, answering questions, interactive
Therapeutic Techniques dialogue). Motor learning theory sug-
gests that, initially, simple behaviours are
In an extensive review of speech literature, acquired more easily than complex ones
we did not find any published protocols for (Kent & Lybolt, 1982). However, behav-
speech therapy with FTMs. We recommend iours that are to be done together must be
that speech-language pathologists working learned together.
with FTMs be clear that they are using a trial 3. Vocal flexibility exercises to maintain vo-
protocol, and seek client permission to record, cal range and voice quality (Pausewang-
Gelfer, 1999). Vocal range and flexibility
evaluate, and publish information on the effi-
exercises are a standard part of a voice
cacy of the protocol. therapy protocol.
There are numerous published protocols for 4. Motor training (Oates & Dacakis, 1997).
speech feminizing therapy with MTFs (An- As speech is a motor act, input is most
drews, 1999; Becklund-Freidenberg, 2002; de useful when it is given at the motor-sen-
Bruin et al., 2000; Gold, 1999; Hooper, 1985; sory level. Matching a sensory target
Kujawski, 2003; Mount & Salmon, 1988; Oates (e.g., “Does your voice feel easy or
& Dacakis, 1997; Pausewang-Gelfer, 1999). stuck? In the face or in the throat?”) is a
As an example of a local protocol, the Changing more effective method of training the de-
Keys program–a mix of group and individual sired production than giving verbal in-
therapy–is discussed in Appendix A. structions such as, “Do this with your
Evaluating the design of treatment proto- jaw” (Titze & Verdolini, in press).
cols. Although treatment protocols must be 5. Identifying and altering voice qualities
flexible enough to address each client’s goals, when coughing, laughing, and clearing
physiologic parameters, and psychosocial needs, the throat (Andrews, 1999; Dacakis,
therapy should be grounded in current knowl- 2002; de Bruin et al., 2000; Oates &
edge of best clinical practice of speech and Dacakis, 1997). These vegetative and
voice therapy. In the absence of empirical evi- spontaneous laryngeal functions may be
dence testing the efficacy of specific tech- higher or lower in pitch than the client de-
niques to feminize or masculinize speech, we sires and may respond to therapeutic in-
evaluated speech therapy protocols on the basis put.
of clinical rationale–a clearly articulated, logi- 6. Experimentation with a broad range of
cal, and valid reason for choosing a specific voice styles (Gold, 1999). Experimenta-
protocol or technique. On this basis, we feel the tion with a broad range of voice styles, in-
following strategies are supportable: cluding ones that might be considered far

beyond what the client would actually 4. Referring to peer support resources. While
want to use, expands the range of possi- the level of knowledge about non-verbal
bilities, and makes smaller changes–ones communication varies greatly among peer
the client may actually use–feel less ex- support providers, individual or group
treme. peer support may offer experiential in-
sights and an arena for practice. As peer
Non-Verbal Communication: Facial knowledge often has strong currency, it
Expressions, Posture, and Movement can be important to remind clients to
weigh the suggestions of peers against
Some transgender individuals are keen ob- their own experience.
servers of non-verbal behavior and are acutely 5. Referring to a trans-competent clinician
attuned to gendered norms relating to non-ver- who has training in non-verbal communi-
bal communication. Others may require assis- cation. In some regions, workshops spe-
tance from a speech therapist. While recogniz- cifically for transgender women are avail-
ing that non-verbal communication is extremely able, such as the “Give Voice” program
important, some speech-language pathologists run by Sandy Hirsch in Seattle. Move-
feel unqualified to offer input; others may feel ment coaches in theatre training pro-
more comfortable doing so. Depending on an grams may be able to assist in finding or
individual clinician’s expertise in this area and developing local resources.
the client’s financial resources, options can in-
clude: Habituation
As with any speech therapy, habituation and
1. Focusing on strengthening the client’s
generalization of feminized or masculinized
observational skills. Experimentation and
communication is both challenging and neces-
observation are more useful than learning
sary. There is a profound difference between
and following rigid patterns of behaviour.
being able to maintain a pitch change on a pro-
2. Offering general feedback on the client’s
longed vowel in a clinical setting and sustaining
self-defined parameters for change. Based
changes throughout speech in everyday life,
on observation of community peers, the
particularly when making offhand remarks in
client can identify desired parameters for
casual conversation when self-monitoring may
change, practice these changes in the
not be as vigilant, or when the client is under
therapy session, and receive subjective
stress or fatigued (Becklund-Freidenberg, 2002).
feedback from the clinician. Parameters
Strategies to promote carryover into everyday
for change may include smiling, eye con-
life may include (a) practicing words that are
tact, facial expressions, posture, and ges-
typically part of daily conversation, such as,
tures while speaking and listening. Feed-
“Hi,” “Bye,” “Yes,” and “No”; (b) focusing
back depends on the desired goal (e.g.,
practice of conversational speech on situations
did the client smile more or less? When?)
or topics related to the client’s life; (c) simulat-
and also the clinician’s subjective sense
ing real-life situations that the client feels pose
of whether the change seemed appropriate.
the most difficulty, such as a job interview or in-
3. Offering general feedback about social
teraction in a coffee shop (Goodnow, 2001;
conventions relating to masculine or femi-
Hooper, 1985); (d) experimenting with emo-
nine expressions and movement. The
tional intensity by practicing sentences ex-
client should be informed of the cultur-
pressing joy, sorrow, irritation, and anger; and
ally-specific nature of non-verbal com-
(e) practicing with the clinician outside the
munication norms and the limits of the
clinic setting, in telephone and in-person inter-
clinician’s expertise in this area. It can be
helpful to discuss the difference between
stereotypes, norms, and observed behav- Follow-Up Sessions
ior, and to remind the client to consider
the clinician’s input in light of their own A small study of MTF transsexuals (N = 10)
experience and perspective. reported a significant correlation between a
Shelagh Davies and Joshua M. Goldberg 183

longer treatment time and stable elevation of motivation to maintain practice, a forum to
pitch over time (Dacakis, 2000). In view of this practice and to share ideas and concerns, and an
finding, follow up sessions after the initialtreat- opportunity to socialize and do specific role-
ment has finished, or facilitated support groups playing. Client-run groups can also foster the
for ongoing practice, may be important in main- client’s sense of ownership and control of
taining change. Clinically supervised followup speech and voice production, rather than feeling
also provides an excellent opportunity to gather dependent on the therapist.
much-needed data about the effectiveness of a In any self-help group there is a danger that
protocol over time. an individual may inappropriately assume a
Clinical group or individual followup ses- professional clinical role. In a speech group,
sions. There is not yet any empirical evidence this could be circumvented by providing group
regarding the optimum frequency for followup facilitation training to members, having the
sessions, the optimum content, or the criteria speech-language pathologist as guest visitor
for termination. In the absence of data, we sug- from time to time, and having self-help sessions
gest that refresher sessions be initially offered 3 along with therapist-run refresher sessions.
months after treatment and then at 4 to 6 month
intervals, or as the clinician and client deem Modification to Improve Accessibility
appropriate. and Utility to Clients with Access Barriers
Followup sessions should include a discus-
sion of successes, problems, strategies, and dif- Protocols must be flexible enough to address
ficulties the client has experienced since the diversity of service needs and issues relating to
end of therapy; a review of the core exercises of access. In the transgender speech literature re-
the program (to ensure the client is practicing viewed for this project, there was little discus-
correctly and to determine if the exercises are sion of modification to address the needs of cli-
still appropriate); and time to address any con- ents who have difficulty accessing the typical
cerns that have arisen since the end of treat- setting or format of speech service, such as indi-
ment. Ideally, followup would include re-eval- viduals who have speech, hearing, cognitive, or
uation of the same parameters measured in the
learning disabilities; are not highly fluent in
pre-treatment assessment, both to assess the
English or are not literate; or are geographically
maintenance of the desired changes and also to
evaluatethe effectiveness of refresher sessions. isolated or cannot leave a residential long-term
If the initial therapy was provided in a group care facility or prison. Without empirical evi-
setting, a group setting is a natural forum for re- dence to guide practice, we offer the following
fresher sessions. As with group format for ini- suggestions based on our experience providing
tial therapy, group format for refresher work servicestoadiverserangeoftransgenderclients.
offers valuable opportunities for clients to com- Distance services. Individuals who are physi-
pare experiences. In our experience this can be cally unable to attend speech therapy or are
most useful and encouraging, especially for awaiting speech therapy services could benefit
those in the early stages of gender transition. In- from an information package available through
dividualized followup may be more appropri- the mail or internet. This kind of “distance
ate than group format if the client has numerous learning program” is currently under develop-
concerns or unusual concerns that require indi- ment at La Trobe University in Australia (G.
vidual attention,or if the client feels uncomfort- Dacakis, personal communication, March 7,
able in a group setting. 2005). Such a distance learning program could
Client-run speech support groups. Self-help include information on the mechanics of
groups are commonly organized for individu- speech and voice production, gendered aspects
als with speech and language disorders such as of speech and voice, tips on observing and lis-
aphasia and stuttering, and may also be useful tening to conversations of men and women in
for transgender individuals who have com- the client’s own community, evaluation of
pleted clinical treatment and are seeking peer commercial speech training programs avail-
support to maintain or strengthen speech able on the Internet, and phonosurgery risks
changes. Client-run speech groups can provide and benefits–similar to the consumer education

materials described earlier (Davies & Goldberg, used with good success with other populations–
2006). for example, palotography and ultrasound have
“Telehealth” is increasingly being explored been used in work with people who are hard
for distance delivery of speech therapy services of hearing and have phonological disorders
(Duffy, Werven, & Arons, 1997; Haynes & (Bernhardt, Bacsfalvi, Gick, Radanov, & Wil-
Kully, 2005; Jessiman, 2003; Mashima et al., liams, 2005; Bernhardt, Gick, Bacsfalvi, &
2003; Myers, 2005). Speech therapy cannot be Adler-Bock, 2005; Bernhardt, Gick, Bacsfalvi,
done by telephone or email as therapy requires a & Ashdown, 2003). For transgender clients,
comprehensive evaluation, regular monitoring there are a number of software programs that re-
of the client’s performance, and specific train- cord fundamental frequency and allow the cre-
ing input. However, clients can use telephone ation of a “model wave.” The clinician could re-
or email to consult with a clinician and receive cord a desired average speaking pitch or an
general information. Video hookup connecting intonation pattern and the client could then use
a rural health unit with an urban speech pro- the visual input to copy it; alternatively, the cli-
gram can be used to train rural practitioners and nician could record the lowest (MTF) or highest
to provide a partial level of service to geograph- (FTM) desirable frequency and the client could
ically isolated clients. use the visual input to keep the speaking pitch
Multilingual services. For individuals who above (MTF) or below (FTM) this line.
do not speak the dominant language, a basic in- If a client has cognitive or learning disabili-
formation package can be translated into a vari- ties, depending on the nature of the disability it
ety of languages. Interpretationor translationof may be useful to include a loved one or care aid
more in-depth information is challenging in in the therapeutic process. This person could
speech services, as the clinician must speak the help the client establish a regular practice
client’s language well enough to be aware of schedule and give input to the exercises, under
subtleties of inflections, inflectional range, the guidance of the speech-language patholo-
word stress, and semantic and syntactic choices. gist. A different format may be useful for the
In cases where the speech therapist and client client who has difficulty processing the infor-
do not speak the same language, the only direct mation necessary to change speech habits.
therapeutic input that could perhaps be given Rather than using an approach that requires in-
would be in changing the average speaking trospection (e.g., “How does that sound? Am I
pitch. SLPs who are multilingual should be en- feeling my voice in my face?”), the clinician
couraged and supported to take trans-specific may be more directive in determining which
training, perhaps working in consultation with exercises would be most useful and could be
a more trans-experienced clinician to provide done appropriately by the client; the clinician
service in the client’s primary language. and client together would draw up a practice
If the client is partially fluent in the language schedule, and the client would simply practise
spoken by the therapist, wishes speech therapy the motor movements outlined. Individualized
in this language, and will be speaking this lan- attention is likely more effective than group
guage in everyday life, therapy delivered in the work to provide the client with more intensive
client’s secondary language can be beneficial input. To be successful, this kind of format
as the client has the opportunity of learning would require regular clinical intervention and
more feminine or masculine patterns of speech support outside the therapy room.
as she or he acquires the language. For individu-
als who are only partially fluent in the language Self-Guided Speech Feminization
spoken by the therapist, the therapeutic process
will likely be longer and will require much There are a variety of videos, websites, and
more individualized input. other materials available for self-guided speech
Access for individuals with disabilities. feminization. We cannot comment on the effi-
Transgender clients with speech or hearing dis- cacy of these materials, but we are concerned
abilities who are able to attend speech therapy that (a) many are not produced by speech pro-
sessions may find great benefit from using vi- fessionals, and (b) there are risks associated
sual input during speech therapy. This has been with attempting to change voice without pro-
Shelagh Davies and Joshua M. Goldberg 185

fessional assistance. Speech feminization or Thyroid chondroplasty may be performed at

masculinization involves substantial changes the same time as vocal surgery to reduce the la-
in habitual production and so has the potential ryngeal prominence (Brown et al., 2000;
to cause a voice disorder or aggravate an exist- Kunachak et al., 2000; Neumann et al., 2002a;
ing one. We strongly recommend that anyone Neumann & Welzel, 2004; Wagner et al.,
seeking to feminize or masculinize speech first 2003). This is a cosmetic procedure that should
be assessed by a speech-language pathologist, not affect the voice.
that a speech clinician be involved in monitor-
ing progress, and that a speech clinician be Risk-Benefit Ratio of Pitch-Elevating Surgery
consulted if there are any symptoms of vocal fa-
tigue or negative changes to vocal quality. Ad- There is a paucity of outcome data for pitch-
ditionally, we recommend that consumers be elevating surgery, particularly longitudinal data
cautious of any materials promoting a rigid set to monitor outcomes over time, but there are
of speech norms, as speech is too individually concerns that the outcome is highly variable
and culturally driven to be guided solely by a set and that initial results tend to diminish over
of generic rules. time (Koufman, n.d.). In addition, reported
negative effects of pitch-elevating surgery in-
Surgical Treatment: clude compromised voice quality, diminished
Pitch-Elevating Surgery vocal loudness, adverse impact on swallowing
or breathing, sore throat, wound infection, and
Surgical techniques to elevate pitch are scarring (Brown et al., 2000; Dacakis, 2002;
based on the physiological components of Koufman, n.d.; Lawrence, 2004; Oates &
pitch: Dacakis, 1997; Neumann & Welzel, 2004;
Petty, 2004; Thomas, 2003, 2005; Wagner et
F0 = (vibrating length of vocal folds/2) ⫻ al., 2003; Yang et al., 2002).
(mean vocal fold tension/vocal fold In general, professional opinion is mixed
density)1/2 about pitch-elevating surgery, with some clini-
cians stating that it is not a viable treatment op-
(Kunachak et al., 2000). Fundamental fre- tion (Andrews, 1999; Koufman, n.d.), and oth-
quency can thus be raised by shortening the ers recommending that surgery be considered a
folds, decreasing the total mass of the folds, or treatment of last resort for MTFs who have not
by increasing the tension of the folds (Neumann experienced satisfactory increase in voice pitch
et al., 2002a; Pickuth et al., 2000; Yang et al., following speech therapy (Lawrence, 2004;
2002). Surgical techniques to achieve this in- Oates & Dacakis, 1997; Wagner et al., 2003).
clude (a) anterior commissure advancement, Other clinicians are more enthusiastic about
(b) creation of an anterior vocal web, (c) crico- pitch-elevating surgery, suggesting that sur-
thyroid approximation, (d) induction of scar- gery can protect the voice from damage caused
ring along the vocal folds, or (e) reduction of by strain to elevate pitch through non-surgical
vocal folds by intracordal steroid injection, means (Brown et al., 2000; Neumann &
laser evaporation of the vocal fold, or compos- Welzel, 2004; Thomas, 2005; Yang et al.,
ite reduction or reconstruction of the vocal fold 2002). While there are clear risks of vocal sur-
(Brown et al., 2000; Donald, 1982; Kunachak gery and the decision to pursue vocal surgery
et al., 2000; Neumann et al., 2002a, 2002b; should be carefully considered, we feel the de-
Orloff, 2000; Pickuth et al., 2000; Thomas, cision about risk-benefit ratio and preferred
2005; Wagner et al., 2003; Yang et al., 2003). technique is best left to the patient, with input
To date, we feel that cricothyroid approxima- from both a trans-experienced surgeon and a
tion is the only method that has been assessed trans-experiencedspeech-languagepathologist.
with sufficient rigor to be considered a viable
treatment option (Brown et al., 2000; de Jong, Pre-Surgical Assessment
2003; Neumann et al., 2002b; Oates & Dacakis,
1997; Soderpalm et al., 2004; Wagner et al., In addition to the standard screening per-
2003; Yang et al., 2002). formed prior to any surgery, such as assessment

for risks relating to anesthesia and infection, as- should consult with the surgeon to determine
sessment prior to pitch-elevating surgery appropriate followup.
should include anatomical and functional as- Immediately following surgery, temporarily
sessment of the larynx, subjective auditory as- decreased pitch, diminished voice quality, and
sessment by both a speech-language patholo- edema were commonly reported in the litera-
gist and the surgeon, and computer recording ture, with spontaneous recovery in most cases.
and analysis of pitch range (Neumann & Less common complications that required
Welzel, 2004; Yang et al., 2002). Care should medical intervention included mild emphy-
be taken to ensure the patient understands the sema, neck abscess, negative response to the
risks and anticipated outcome of the technique sutures or plates used in cricothyroid approxi-
that will be used. mation (requiring removal of the material),
After finding that some subjects have and loosening of the sutures used in crico-
strained and unnaturally elevated voices fol- thyroid approximation, requiring further sur-
lowing surgery, attributed to habitually speak- gery (Neumann & Welzel, 2004; Wagner et al.,
ing at an artificially elevated pitch for sustained 2003).
periods of time prior to surgery, one surgical For most pitch-elevating surgical tech-
group reported testing for ability to phonate at a niques, it is recommended that patients not use
pitch within the masculine range as part of pre- the voice at all for one to seven days after sur-
operative consultation (Yang et al., 2002). Cli- gery, and then use the voice cautiously until any
ents who are unable to do this were felt to have discomfort due to postoperative edema has
the equivalent of a muscle tension dysphonia, passed (Brown et al., 2000; Neumann &
and were referred for preoperative voice ther- Welzel, 2004; Orloff, 2000). For the more inva-
apy to recover the ability to produce relaxed sive combined thyroid cartilage and vocal fold
phonation. reduction, two weeks vocal rest is suggested
Estrogen is associated with risk for deep vein (Kunachak et al., 2000). Following crico-
thrombosis and pulmonary embolism (Dahl, thyroid approximation, steam inhalation may
Feldman, Goldberg, & Jaberi, 2006). If the be recommended to hydrate and lubricate the
patient will be immobilized for a prolonged vocal cords, to promote healing (Brown et al.,
period during or following pitch-elevating sur- 2000).
gery, consultation with the prescribing physi- Speech therapy is recommended following
cian is necessary to discuss the advisability of surgery to help the patient adapt to and stabilize
tapering estrogen use before surgery (Bowman the new voice (Neumann & Welzel, 2004;
& Goldberg, 2006). Wagner et al., 2003). If pitch-elevating surgery
Smoking increases the risk of complications was performed before other components of
from anesthetic and impairs healing, and there speech had been satisfactorily altered, reso-
is evidence that smoking following pitch-ele- nance, articulation, and other components may
vating surgery can negatively impact on voice also need to be addressed via speech therapy
quality and pitch (Wagner et al., 2003). Patients (Dacakis, 2002; Neumann & Welzel, 2004).
should be informed of the risks associated with
smoking and of smoking cessation resources,
and strongly encouraged to not smoke prior to OUTCOME EVALUATION
or immediately following surgery.
Evaluation is a continuous process in speech
Post-Surgical Care care, with various informal and formal methods
that may be used to determine progress and
Post-surgical care depends on the specific shape the direction of future treatment(Hooper,
surgical technique employed. The surgeon 1985; Mount & Salmon, 1988; Soderpalm et
should review aftercare instructions with the al., 2004). We recommend that at minimum the
patient as part of informed consent prior to sur- baseline assessment be repeated immediately
gery. The surgeon should also be accessible for following the end of therapy, and post-treat-
questions relating to post-operative complica- ment data compared to pre-treatment findings.
tions. The patient’s local primary care provider If the client is agreeable to long-term followup,
Shelagh Davies and Joshua M. Goldberg 187

given the paucity of long-term data it would be In some cases the clients may be very satisfied
ideal for the client to be re-evaluated 6 months, with the clinician’s performance despite mini-
1 year, 5 years, and 10 years after treatment; for mal changes to speech; whatever the outcome,
transient clients this degree of followup may clients may have constructive critical feedback
not be possible, but even data at 6 and 12 months to offer the clinician regarding the ability to re-
would be a significantcontributionto the field. late information clearly and accurately, sensi-
In addition to re-evaluating objective and tivity and respect in communication, overall fa-
subjective impressions of speech as per the ini- miliarity with transgender concerns, efficient
tial assessment, we recommend that clients be coordination with other clinicians, and accessi-
invited to evaluate satisfaction with the out- bility of treatment.
come of treatment (Bralley et al., 1978; Byrne If long-term followup is feasible, in addition
et al., 2003; Kujawski, 2003; Mount & Salmon, to the standard re-evaluation of speech it may
1988; Pausewang-Gelfer, 1999; Soderpalm et be useful to inquire about clients’ continuation
al., 2004). Several trans-specific studies re- of therapeutic exercises and symptoms of vocal
ported a discrepancy between subjective satis- fatigue (Dacakis, 2000; Soderpalm et al., 2004).
faction and objective or subjective changes to Three of five participants in one long-term
voice, with some clients pleased with the out- study reported that after speech therapy had
come despite minimal objective change, and ended, they attempted further change through
others perceiving less change than that reported techniques learned from the internet or in books
by naïve listeners (Bralley et al., 1978; Dacakis, (Soderpalm et al., 2004). It may be useful to of-
2000; Soderpalm et al., 2004; Spencer, 1988). fer consumer education regarding risk preven-
This raises the question of what is considered a tion and ongoing monitoring to clients who are
“successful” intervention. Some authors inter-
interested in pursuing techniques outside a pro-
preted the findings as evidence that clients can-
fessional setting.
not accurately judge “successful” voice change
(Bralley et al., 1978); others felt that discrep-
ancy between subjective satisfaction and ob-
jective changes to voice may have stemmed
from increased passability in other dimensions
Speech and voice change services for trans-
(e.g., from hormones or electrolysis), a good
working relationship with the clinician, or sat- gender individuals are an important element of
isfaction with the availability or cost of the ser- transgender care. Treatments to feminize and
vice (Dacakis, 2000). It is also possible that cli- masculinize speech and voice can help reduce
ent goals shifted over time or that clients’ goals discomfort for the dysphoric client, improving
for speech did not center on pitch or passability, confidence and comfort in day-to-day commu-
the typicalmeasures employed for evaluation. nication interactions. As self-directed speech
Earlier we suggested that the primary goal of and voice change can result in vocal strain we
speech feminization or masculinization is to strongly recommend that professional speech
decrease discrepancy between speech and the services be included in transgender health pro-
client’s sense of self; it is, we think, highly rele- grams and made available not only to transsex-
vant to ask about the client’s feelings about “fit” uals but also crossdressers, bi-gendered people,
between speech and identity as part of post- androgynous people, and others who desire to
treatment assessment, even if the client did not feminizeor masculinizetheir speech and voice.
explicitly state this as an objective at the start of There is currently an insufficient data base to
treatment(Soderpalm et al., 2004). Another rel- determine evidence-based best protocols in
evant measure might be the client’s report of transgender speech and voice modification.We
being able to use the desired speech consis- hope that this article will both assist speech pro-
tently in the settings that were identified as the fessionals in adapting and modifying existing
targets at the outset of therapy (Kujawski, 2003; protocols to address a client’s individual needs,
Pausewang-Gelfer, 1999). and also stimulate interest in evaluation of prac-
We also encourage clinicians to invite cli- tice protocols for MTFs and FTMs. Further re-
ents to evaluate the quality of service provided. search in this area is strongly recommended.

NOTE nal of Language and Communication Disorders, 35,

1. Published transgender speech research focuses Bowman, C., & Goldberg, J. M. (2006). Care of the pa-
on transsexual women, with only a few studies involv- tient undergoing sex reassignment surgery. Interna-
ing male crossdressers or female-to-male transsexuals. tional Journal of Transgenderism, 9(3/4), 135-136.
In this document we use “male-to-female” (MTF) broadly Byrne, L. A., Dacakis, G., & Douglas, J. M. (2003).
unless otherwise noted, to describe a spectrum of people Self-perceptions of pragmatic communication abilities
who were assigned “male” at birth and who wish to in male-to-female transsexuals. Advances in Speech
feminize or de-masculinize their speech (including male Language Pathology, 5, 15-25.
crossdressers, transsexual women, and bi-gendered or Challoner, J. (2000). The voice of the transsexual. In M.
androgynous people born male). Similarly, “female- Freeman & M. Fawcus (Eds.), Voice disorders and
to-male” (FTM) refers to people who were assigned their management (pp. 244-267). Philadelphia: Whurr
“female” at birth and who wish to masculinize or Publishing.
de-feminize their speech. This breadth of terminology is Coleman, R. O. (1983). Acoustic correlates of speaker
used to promote inclusion of non-transsexual clients sex identification: Implications for the transsexual
who may seek speech feminization or masculinization voice. Journal of Sex Research, 19, 293-295.
Dacakis, G. (2000). Long-term maintenance of funda-
mental frequency increases in male-to-female trans-
sexuals. Journal of Voice, 14, 549-556.
Dacakis, G. (2002). The role of voice therapy in male-
REFERENCES to-female transsexuals. Current Opinion in Otolar-
yngology & Head and Neck Surgery, 10, 173-177.
Andrews, M. L. (1999). Voice and psychosocial dynam- Dahl, M., Feldman, J. L., Goldberg, J. M., & Jaberi, A.
ics: Gender presentation. In M. L. Andrews (Ed.), (2006). Physical aspects of transgender endocrine
Manual of voice treatment: Pediatrics through geri- therapy. International Journal of Transgenderism,
atrics (pp. 432-446). San Diego, CA: Singular Pub- 9(3/4), 111-134.
lishing Group. Davies, S., & Goldberg, J. (2006). Gender transition:
Andrews, M. L., & Schmidt, C. P. (1997). Gender presen- Changing speech. Vancouver, BC: Vancouver Coastal
tation: Perceptual and acoustical analyses of voice. Health Authority. Available online at http://www.
Journal of Voice, 11, 307-313.
Andrianopolous, M. V., Darrow, K., & Chen, J. (2001). speech.pdf
Multimodal standardization of voice among four de Bruin, M. D., Coerts, M. J., & Greven, A. J. (2000).
multicultural populations: Formant structures. Jour- Speech therapy in the management of male-to-fe-
nal of Voice, 15, 61-77. male transsexuals. Folia Phoniatrica et Logopaedica,
Becklund-Freidenberg, C. (2002). Working with male- 52, 220-227.
to-female transgendered clients: Clinical consider- de Jong, F. (2003). Surgical raise of vocal pitch in male
ations. Contemporary Issues in Communication Sci-
to female transsexuals. Paper presented at the XVIII
ence and Disorders, 29, 58.
Biennial Symposium of the Harry Benjamin Interna-
Bernhardt, B., Bacsfalvi, P., Gick, B., Radanov, B., &
tional Gender Dysphoria Association, Ghent, Bel-
Williams, R. (2005). Exploring electropalatography
and ultrasound in speech habilitation. Journal of gium.
Speech-Language Pathology and Audiology, 29, Delph-Janiurek, T. (1999). Sounding gender(ed): Vocal
169-182. performances in English university teaching spaces.
Bernhardt, B., Gick, B., Bacsfalvi, P., & Adler-Bock, M. Gender, Place and Culture: A Journal of Feminist
(2005). Ultrasound in speech therapy with adoles- Geography, 6, 137-153.
cents and adults. Clinical Linguistics and Phonetics, Donald, P. J. (1982). Voice change surgery in the trans-
19, 605-617. sexual. Head and Neck Surgery, 4, 433-437.
Bernhardt, B., Gick, B., Bacsfalvi, P. & Ashdown, J. Duffy, J.R., Werven G.W., & Aronson, A.E. (1997).
(2003). Speech habilitation of hard of hearing ado- Telemedicine and the diagnosis of speech and language
lescents using electropalatography and ultrasound as disorders. Mayo Clinic Proceedings, 72, 1116-1122.
evaluated by trained listeners. Clinical Linguistics Elert, C-C., and Hamrnarberg, B. (1991). Regional voice
and Phonetics, 17, 199-216. variation in Sweden. In Actes du XIIème Congres Inter-
Bralley, R. C., Bull, G. L., Gore, C. H., & Edgerton, M. national des Sciences Phonétiques, Vol. 4 (Université
T. (1978). Evaluation of vocal pitch in male trans- de Provence, Service des Publications, Aix-en-Pro-
sexuals. Journal of Communication Disorders, 11, vence), 418-420.
443-449. Gold, L. (1999). Voice training for the transsexual.
Brown, M., Perry, A., Cheesman, A. D., & Pring, T. VASTA Newsletter, 13. Retrieved January 1, 2005,
(2000). Pitch change in male-to-female transsexuals: from
has phonosurgery a role to play? International Jour- html
Shelagh Davies and Joshua M. Goldberg 189

Goldberg, J. M. (2006). Recommended framework for male-to-female transsexual subject. Scandinavian

training in speech feminization/masculinization. Van- Journal of Logopedic and Phoniatrics, 18, 105-109.
couver, BC: Vancouver Coastal Health Authority. Kent, R. D., & Lybolt, J. T. (1982). Techniques of ther-
Available online at apy based on motor learning theory. In W. H. Perkins
resources/library/tcpdocs/training-speech.pdf (Ed.), General principles of therapy (pp. 13-25). New
Goldberg, J. M., & Lindenberg, M. (Eds.) (2001). Trans- York: Thieme-Stratton.
Forming community: Resources for trans people and King, J. B., Lindstedt, D. E., Jensen, M., & Law, M.
our families. Victoria, BC: Transcend Transgender (1999). Transgendered voice: Considerations in case
Support & Education Society. history management. Logopedics Phoniatrics Vocology,
Goldberg, J. M., Matte, N., MacMillan, M., & Hudspith, 24, 14-18.
M. (2003). Community survey: Transition/crossdress- Kohnert, K., Kennedy, M. R. T., Glaze, L., Kan, P. F., &
ing services in BC–Final report. Vancouver, BC: Carney, E. (2003). Breadth and depth of diversity in
Vancouver Coastal Health and Transcend Trans- Minnesota: Challenges to clinical competency. Amer-
gender Support & Education Society. ican Journal of Speech-Language Pathology, 12,
Goodnow, C. (2001, February 12). Speech therapy helps 259-272.
transgender women develop a feminine sound. Seattle Koschkee, D. L., & Rammage, L. (1997). Voice care in
Post-Intelligencer. Retrieved January 1, 2005, from a medical setting. San Diego, CA: Singular Publish- ing Group.
shtml Koufman, J. (n.d.). Call for a moratorium on voice
Graddol, D., & Swann, J. (1989). Gender voices. Ox- feminization surgery for the M-to-F transsexual in
ford: Basil Blackwell Ltd. the United States. Winston-Salem, NC: Center for
Gross, M. (1999). Pitch-raising surgery in male-to-fe- Voice and Swallowing Disorders. Retrieved January
male transsexuals. Journal of Voice, 13, 246-250. 1, 2005, from
Günzburger, D. (1993). An acoustic analysis and some moratorium.htm
perceptual data concerning voice change in male-fe- Kujawski, C. (2003). The transsexual voice. Unpublished
male trans-sexuals. European Journal of Disorders manuscript, Michigan State University. Retrieved
of Communication, 28, 13-21. January 1, 2005, from
Hasegawa, Y. & Hata, K. (1995). The function of F0-
peak delay in Japanese. In J. Ahlers, L. Bilmes, J. S.
Guenter, B. A. Kaiser, & J. Namkung (Eds.), Pro-
Kunachak, S., Prakunhungsit, S., & Sujjalak, K. (2000).
ceedings of the 21st Annual Meeting of the Berkeley
Thyroid cartilage and vocal fold reduction: A new
Linguistics Society (pp. 141-151). Berkeley, CA:
phonosurgical method for male-to-female transsexu-
Berkeley Linguistics Society.
Haynes, E., & Kully, D. (2005). The use of telehealth in als. Annals of Otology, Rhinology and Laryngology,
the treatment of stuttering. Communiqué: The News- 109, 1082-1086.
letter of the Canadian Association of Speech-Lan- Lawrence, A. A. (2004). Voice feminization surgery: A
guage Pathologists and Audiologists, 19(3), 4. critical overview. Seattle, WA: Author. Retrieved
Hearing, Speech & Deafness Center (2005). Voice fem- January 1, 2005, from
inization. Seattle, WA: Author. Retrieved January 1, voicesurgery.html
2005, from Linville, S. E. (1998). Acoustic correlates of perceived
htm versus actual sexual orientation in men’s speech.
Hirano, M. (1981). Clinical examination of voice. In G. Folia Phoniatrica et Logopaedica, 50, 35-48.
E. Arnold, F. Winckel, & B. D. Wyke (Eds.), Disor- Mashima, P.A., Birkmire-Peters, D.P., Syms, M.J.,
ders of human communication (5th ed., pp. 81-84). Holtel, M.R., Burgess, L.P., & Peters, L.J. (2003).
New York: Springer-Verlag. Telehealth: Voice therapy using telecommunications
Hooper, C. R. (1985). Changing the speech and lan- technology. American Journal of Speech-Language
guage of the male to female transsexual client: A case Pathology, 12, 432-439.
study. Journal of the Kansas Speech-Language-Hear- Mikos, V. A., & Pausewang-Gelfer, M. (2001, Novem-
ing Association, 25, 6. ber). The relative contribution of speaking funda-
Jessiman, S. (2003). Speech and language services using mental frequency and formant frequencies to gender
telehealth technology in remote and underserved ar- identification of biological males, biological females,
eas. Journal of Speech-Language Pathology and Au- and male-to-female transgendered individuals based
diology, 27, 45-51. on isolated vowels. Presented at the Convention of
Kalra, M. A. (1977). Voice therapy with a transsexual. the American Speech-Language-Hearing Associa-
In R. Gemme & C. Wheeler (Eds.), International tion, New Orleans, LA.
Congress on Sexology (pp. 77-84). New York: Ple- Moran, M. J., McCloskey, L., & Cady, B. (1995). Lis-
num Press. tener age estimates of elderly African American and
Kaye, J., Bortz, M. A., & Toumi, S. I. (1993). Evalua- Caucasian male speakers. Journal of Cross-Cultural
tion of the effectiveness of voice therapy with a Psychology, 26, 751-758.

Morris, R. J., & Brown, W. S. (1994). Age-related dif- Perez, K. (2004). Voice feminization for transgender
ferences in speech variability among women. Jour- women. Denver, CO: Exceptional Voice, Inc. Retrieved
nal of Communication Disorders, 27, 49-64. January 1, 2005, from
Mount, K. H., & Salmon, S. J. (1988). Changing the vocal transgender.html.
characteristics of a postoperative transsexual patient: Pickuth, D., Brandt, S., Neumann, K., Berghaus, A.,
A longitudinal study. Journal of Communication Spielmann, R. P., & Heywang-Kobrunner, S. H.
Disorders, 21, 229-238. (2000). Spiral computed tomography before and af-
Myers, C. (2005). Telehealth applications in head and ter cricothyroid approximation. Clinical Otolaryn-
neck oncology. Journal of Speech-Language Pathol- gology, 25, 311-314.
ogy and Audiology, 29, 125-129. Prochaska, J. O., DiClemente, C. C., & Norcross, J. C.
Neumann, K., & Welzel, C. (2004). The importance of (1992). In search of how people change: Applica-
the voice in male-to-female transsexualism. Journal tions to addictive behaviors. American Psychologist,
of Voice, 18, 153-167. 47, 1102-1114.
Neumann, K., Welzel, C., Gonnermann, U., & Wolfradt, Rose, P. (1991). How effective are long term mean and
U. (2002a). Cricothyroidopexy in male-to-female trans- standard deviation as normalization parameters for
sexuals: Modification Type IV. International Journal tonal fundamental frequency? Speech Communica-
of Transgenderism, 6(3). Retrieved January 1, 2005, tion, 10, 229-247.
from Sabol, J. W., Lee, L., & Stemple, J. C. (1995). The value
htm of vocal function exercises in the practice regimen of
Neumann, K., Welzel, C., Gonnermann, U., & Wolfradt, singers. Journal of Voice, 9, 27-36.
U. (2002b). Satisfaction of MtF Transsexuals with Soderpalm, E., Larsson, A., & Almquist, S. A. (2004).
operative voice therapy: A questionnaire-based pre- Evaluation of a consecutive group of transsexual in-
liminary study. International Journal of Trans- dividuals referred for vocal intervention in the west
genderism, 6(4). Retrieved January 1, 2005, from of Sweden. Logopedics Phoniatrics Vocology, 29, 18-30.
Núñez, A. E. (2000). Transforming cultural competence Spencer, L. E. (1988). Speech characteristics of male-
into cross-cultural efficacy in women’s health educa-
to-female transsexuals: A perceptual and acoustic
tion. Academic Medicine, 75, 1071-1079.
study. Folia Phoniatrica, 40, 31-42.
Oates, J. M., & Dacakis, G. (1983). Speech pathology
Stemple, J. C., Lee, L., D’Amico, B., & Pickup, B.
considerations in the management of transsexualism:
(1994). Efficacy of vocal function exercises as a
A review. British Journal of Disorders of Communi-
method of improving voice production. Journal of
cation, 18, 139-151.
Oates, J. M., & Dacakis, G. (1997). Voice change in Voice, 8, 271-278.
transsexuals. Venereology: Interdisciplinary, Inter- Titze, I. R. (1997). Are the corner vowels like primary
national Journal of Sexual Health, 10, 178-187. colors? Journal of Singing, 5, 35-37.
Oates, J., & Russell, A. (1997). Perceptual Voice Pro- Titze, I. R., & Verdolini, K. (in press). Vocology. Iowa
file. In A sound judgment: A CD-ROM to teach per- City, IA: National Center for Voice and Speech.
ceptual voice analysis [CD-ROM]. Campbelltown, Thomas, J. P. (2003). Cricothyroid approximation and
Australia: Clear Digital Vision. laryngeal reduction: Information on risks specific to
Orloff, L. A. (2000). Disorders of pitch. San Diego, CA: this pitch altering surgery and reduction of the exter-
Division of Otolaryngology–Head and Neck Sur- nal appearance of the voice box. Portland, OR: Au-
gery, University of California, San Diego. Retrieved thor. Retrieved January 1, 2005, from http://www.
January 1, 2005, from
ent/PatientInfo/voi_pitch.html Thomas, J. P. (2005). Male to female vocal surgery.
Pausewang-Gelfer, M. (1999). Voice treatment for the Portland, OR: Author. Retrieved October 31, 2005,
male-to-female transgendered client. American Jour- from
nal of Speech-Language Pathology, 8, 201-208. index.html
Pausewang-Gelfer, M., & Mikos, V. A. (2005). The rel- Tom, K. (2004, November). Fundamental frequency
ative contributions of speaking fundamental frequency characteristics of Mexican-American speakers of Eng-
and formant frequencies to gender identification based lish and Spanish. Paper presented at the Annual Con-
on isolated vowels. Journal of Voice, 19, 544-554. vention of the American Speech-Language-Hearing
Pausewang-Gelfer, M., & Schofield, K. J. (2000). Com- Association, Philadelphia, PA.
parison of acoustic and perceptual measures of voice Van Borsel, J., De Cuypere, G., & Van den Berghe, H.
in male-to-female transsexuals perceived as female (2001). Physical appearance and voice in male-to-fe-
versus those perceived as male. Journal of Voice, 14, male transsexuals. Journal of Voice, 15, 570-575.
22-33. Van Borsel, J., De Cuypere, G., Rubens, R., & Destaerke,
Petty, B. (2004, April). The vocal instrument: How it B. (2000). Voice problems in female-to-male trans-
works and how to take care of it. Paper presented at sexuals. International Journal of Language & Com-
Transgender Voices Festival, Minneapolis, MN. munication Disorders, 35, 427-442.
Shelagh Davies and Joshua M. Goldberg 191

Wagner, I., Fugain, C., Monneron-Girard, L., Cordier, Wollitzer, L. C. (1994). Acoustic and perceptual cues to
B., & Chabolle, F. (2003). Pitch-raising surgery in gender identification: A study of transsexual voice
fourteen male-to-female transsexuals. Laryngoscope, and speech characteristics. Unpublished master’s
113, 1157-1165. thesis, University of British Columbia, Vancouver,
Wilson, J. A., Webb, A., Carding, P. N., Steen, I. N., British Columbia, Canada.
MacKenzie, K., & Deary, I. J. (2004). The Voice Yang, C. Y., Palmer, A. D., Murray, K. D., Meltzer, T.
Symptom Scale (VoiSS) and the Vocal Handicap In- R., & Cohen, J. I. (2002). Cricothyroid approximation
dex (VHI): A comparison of structure and content. to elevate vocal pitch in male-to-female transsexu-
Clinical Otolaryngology, 29, 169-174. als: Results of surgery. Annals of Otology, Rhinology
Wiltshire, A. (1995). Not by pitch alone: A view of and Laryngology, 111, 477-485.
transsexual vocal rehabilitation. National Student Yardley, K. M. (1976). Training in feminine skills in a
Speech Language Hearing Association Journal, 22, male transsexual: A pre-operative procedure. British
53-57. Journal of Medical Psychology, 49, 329-339.
Wolfe, V. I., Ratusnik, D. L., Smith, F. H., & Northrop, Zimmerman, G. L., Olsen, C. G., & Bosworth, M. F.
G. (1990). Intonation and fundamental frequency in (2000). A ‘Stages of Change’ approach to helping
male-to-female transsexuals. Journal of Speech & patients change behavior. American Family Physi-
Hearing Disorders, 55, 43-50. cian, 61, 1409-1416.


APPENDIX. The Changing Keys Program The program is subsidized by the Vancouver
Coastal Health Authority, a health governance
Changing Keys (CK) is an English-language body responsible for delivery of health services in
speech feminization program offered in Vancou- the Vancouver-Coast region of British Columbia,
ver, British Columbia, Canada. CK was created by a to make it possible for low-income transgender
speech-language pathologist (Shelagh Davies) as women to participate. Participants are asked to pay
part of the Transgender Health Program’s services. what they feel they can afford within a sliding scale
The program consists of (a) a one-hour individual of $0-$100 for the entire program. This funding
speech and voice evaluation at the start of the pro- structure was key in making the program accessible
gram, (b) weekly two-hour speech and voice ther- as there are high rates of poverty among transgender
apy group sessions, for seven weeks, (c) individual- women (Goldberg et al., 2003).
ized sessions midway through the seven weeks,
(d) speech therapy exercises to be done between
groups (“homework”), (e) a one-hour individual APPLICATION PROCEDURE
speech and voice evaluation at the end of the pro- AND CLIENT SCREENING
gram, and (f) a refresher session held 3 to 4 months
after the course has ended. CK is held at a multi- CK is advertised extensively by notices and post-
disciplinary inner city community health centre that ers to service providers who work with transgender
serves large numbers of transgender clients and women, announcements in community peer sup-
houses the Transgender Health Program and a port groups, and online announcements to commu-
transgender peer support group. nity mailing lists. Interested participants are asked
The program is limited to six self-identified to apply to the Transgender Health Program by pro-
transgender women who want to feminize their viding contact information and answering seven
voice, don’t have coverage for speech therapy questions relating to eligibility:
through Extended Health or other benefits, can
commit to coming to all of the sessions and doing 1. What is your goal for taking part in Changing
practice sessions between groups, feel comfortable Keys?
working on voice in a group setting, and are able to 2. This pilot is restricted to people who self-
read and speak comfortably in English (individual identify as transgender women/male-to-fe-
speech therapy is recommended for clients who are male. Does this fit for you?
only partially fluent in English, to allow more clini- 3. This pilot is restricted to people who don’t
cian attention). It is not necessary that participants have coverage for speech therapy through
be living as women, be taking hormones, or have Extended Health or other benefits. Do you
had surgery. Participants are asked to present as have benefits that pay for speech therapy? If
women or gender-neutral at the therapy sessions, as so, we recommend you use those benefits to
this is felt to facilitate practice of feminine voice. pay for private sessions.

4. Have you read over the dates, and can you vocal inflections, changing voice quality, and mod-
commit to coming to all the sessions? ifying characteristics of articulation.
5. What are your preferred times for the pre- and At the end of the assessment the results are dis-
post-group assessment? cussed with the client, and together the client and
6. Are you comfortable listening to information the therapist establish goals for specific therapy.
about voice and doing practice exercises The therapy process is explained to the client, the
with other transgender women in a group? expected commitment is described, and any ques-
7. Are you able to read and speak comfortably tions about the therapy program are answered. The
in English? If not, please contact the Trans- client should leave the evaluation with a clear idea
gender Health Program to discuss options. of what changes are possible and useful, and have a
(The Transgender Health Program can explore sense of the processes involved.
possible referral to a multilingual speech thera- Weekly Sessions
pist in private practice, or arrange funded in-
terpretation service for individualized sessions Six 2-hour group sessions are held weekly. Ses-
with an English-speaking speech therapist.) sions are divided into four parts: (a) checking in on
previous week’s practice, observation and carry-
over activities–what worked, what didn’t, and what
PROGRAM STRUCTURE needs to be modified in group exercises or the indi-
Initial Assessment vidual’s practice, (b) voice training, with the goals
of producing an easy, resonant voice at the target
The assessment provides an opportunity to eval- pitch and generalizing it into speech of increasing
uate the client’s current speech and voice produc- complexity, (c) exercises directed at a specific
tion habits, determine how well the speech matches topic, such as increasing vocal inflections, and (d)
the inner sense of self, consider what changes would information and discussion: e.g., pitch-raising sur-
be beneficial, and evaluate how easy or difficult the gery, gender markers in communication.
changes would be. Speech and voice parameters as-
sessed include average speaking pitch, speaking Individualized Sessions
pitch range, impression of vocal inflectional pat-
terns, and voice quality. If vocal loudness subjec- Halfway through the program there are 30-min-
tively appears to be outside normal female range it is ute individual sessions with each participant. These
objectively evaluated by measuring the average, are used to give one-to-one input into particular ar-
maximum and minimum speaking loudness. Speech eas of difficulty and to modify exercises to suit each
and voice are assessed in oral reading, picture de- client. For example, if the client has difficulty sus-
scription, and spontaneous conversation. Discrep- taining the voice at a target pitch, specific voice
ancies of parameters among tasks are noted. The as- training is given or a more suitable pitch is used.
sessment is audiotaped so it can be reviewed at the These sessions are often client-driven, with the cli-
end of the program and compared with the fol- ent providing the focus for the session.
low-up evaluation.
A subjective evaluation is also done. The client Homework
describes three specific things she would like to
change about her speech and voice or three situa- As this is a short, intensive program, participants
tions in which she would like to sound more femi- are expected to do substantial practice between ses-
nine. She also fills in a questionnaire describing sions. Although homework requires substantial com-
how her current speech and voice affect her life. mitment, clients are often highly motivated and dili-
The client is asked to identify real-life situa- gent in practice. Homework consists of three parts:
tions–ranging from easy to difficult–in which she
can practice generalizing what was learned in the 1. Basic vocal training exercises. These exer-
sessions. This helps her try things out while still cises are taught the first day of therapy and
having the support of the group. It also allows par- are to be done for 10 minutes twice a day. In-
ticipants to take ownership of the techniques as they structions are written in the course manual
are learning them and can be a powerful motivator and recorded on a CD or tape.
to continue practice. 2. Weekly topics of practice. These include
The assessment includes trial therapy to deter- specific practice of the speech parameters
mine how easily the client is able to make changes in discussed in the weekly therapy session, such
her speech and voice. Parameters assessed may in- as transferring a higher speaking pitch into
clude producing voice at a higher pitch, varying different real life situations (e.g., asking for a
Shelagh Davies and Joshua M. Goldberg 193

transfer on a bus or answering the phone), us- Verdolini developed this therapy protocol for
ing wider vocal inflections, and being an ac- treating voice disorders, using input from both tra-
tive listener. ditional singing and speaking voice pedagogy and
3. Observations. Becoming familiar with gender- current concepts in voice science and psychology.
ed differences in communication is essential Although the protocol was not developed specifi-
to making changes, but unstructured obser- cally to train transgender women, the twin focuses
vation can be overwhelming and ineffective. of ease and forward resonance sensations train effi-
Each week, participants are asked to observe cient voice production that helps protect the vocal
a specific aspect of women’s speech. For ex- folds from damage. The forward focus may also
ample, questions relating to inflection may help increase vowel formants, helping the voice to
include: How does a woman’s voice move be perceived as female (Becklund-Freidenberg,
around during speech? How are inflections 2002).
different among women? Do inflections vary The core exercise program is taught in the first
with the age of the speaker, the speaking situ- session and includes stretching, relaxing exercises,
ation, how the woman may be feeling emo- and producing the voice at a target pitch. Specific
tionally, her conversational partner? Other sounds that have been shown to encourage efficient
topics have included: Do women laugh or vocal fold vibration and maximize forward reso-
smile at different times than men? How do nance sensations are used to train the higher pitch.
women take turns in conversation? What do Voice training takes approximately 60 minutes in
women do when they are listening? What is it the first session, and then 20-30 minutes in subse-
about speech that makes it sound feminine or quent sessions. Participants are instructed on how to
masculine? monitor their practice. Difficulties that occur in
practice, such as throat tightness or effortful pro-
Final Assessment duction, are addressed in subsequent sessions and
individual instruction is given as necessary.
The parameters measured in the initial assess-
ment are re-measured and the client again com- Core Exercise Program
pletes the subjective evaluation form. The pre- and
post- measurements and the tape recordings are Relaxation
compared and changes noted. Suggestions for mod-
ification and continuation of practice are discussed. We begin with standard general relaxation exer-
The client’s input about the course is sought. cises to relax the head and neck area, jaw, tongue,
face, and mid-body respiratory muscles; additional
Refresher Session exercises may be suggested for individual partici-
pants as needed. A goal during this time is to in-
A 2 hour refresher session is held 3 to 4 months crease general awareness of how the mind and body
after the completion of the program. Participants feel at this particular point in this particular day:
bring a completed self-evaluation questionnaire tired or rested, anxious or calm, focused or scat-
and vocal parameters are reassessed. The basic ex- tered, tight shoulders, breath-holding, etc.
ercises of the program are reviewed and there is time
to discuss successes and challenges of using their Facilitating Production in the Upper Pitch
new voice in the real world. This session serves as Range
both a motivator for continued practice and an op-
portunity for the clinician to provide guidance on Using a voiced bilabial fricative (“raspberry”) or
difficulties experienced by the client. a tongue trill (Spanish “r”) the clients glide the voice
around in the middle to upper pitch range. This tech-
nique is used in both voice therapy and singing ped-
VOICE TRAINING agogy and has been described in Joseph Stemple’s
Vocal Function Exercises (Sabol et al., 1995;
One of the aims of the Changing Keys program is Stemple et al., 1994). According to Stemple, going
to develop the production of a higher speaking pitch to the end ranges of the voice has a similar effect as
range that is efficient and easy to produce–a com- stretching a muscle to end range; the exercise facili-
mon goal of most course participants. The protocol tates ease and efficiency in the middle ranges. For
used is based on the Lessac Marsden Resonant our purposes, we are exploring the sensations of
Voice Therapy (LMRVT) program, developed by producing a higher-pitched voice easily and effi-
Katherine Verdolini (Titze & Verdolini, in press). ciently.

Sensations during the exercise are carefully In accordance with the LMRVT protocol, clients
monitored. The voice should feel resonant and easy are asked to monitor two things as they practice:
in the throat at all pitches. If the throat begins to Does the voice feel easy to produce, and does the
tighten in the higher pitches, voice therapy facilita- sound feel like it is going up and out (or does it feel
tion techniques are used. The goal is to produce a like it is getting caught–in the throat or anywhere
resonant, easy sound throughout the upper pitch else)? If the client does not have these sensations of
range. Going into falsetto register is fine in this ex- ease, we do specific facilitation exercises. We then
ercise. Although the target speaking pitch should be use the LMRVT protocol to expand this sensation of
in modal rather than falsetto register, some trans- easy, resonant voice production into sounds and
sexual women are able to use the falsetto occasion- words. The goal is to generalize this easy, resonant,
ally when using a wide pitch range, and this can higher-pitched voice, first in structured speech and
sound acceptable as long as it is well blended with then into spontaneous speech in increasingly diffi-
the rest of the voice. cult situations.
After this voice training program, the voice
should not feel tired: it should feel warmed up and Extending the Higher-Pitched Voice into Speech
ready to use. If the voice begins to feel tired or if
there is throat sensation, this is a signal that some in- Generalizing the use of higher pitch follows
tervention needs to be done in the way of modifying standard speech and voice therapy protocols, start-
voice production technique. The exercise of gliding ing with easier tasks and gradually working into
around in the upper part of the voice is then ex- more challenging ones. Speech tasks are those that
panded into vowels. are common in speech and voice therapy, progress-
ing from single words to short phrases, greetings,
Producing a Higher-Pitched Voice short oral reading tasks, picture descriptions, and
structured questions and answers. While the voice
Raising the average speaking pitch is a common
is first produced at only one pitch (chanting), as
goal among group participants, and is supported in
soon as possible regular speech inflections are in-
the literature and by experienced clinicians. How-
ever, there is a wide range of clinical opinion about
Maintaining elevated pitch in a resonant voice
how to train a higher pitched voice and what pitch is
that feels easy to produce is a vocally athletic task;
optimal to target. Most clinicians agree that a goal
for English-language speakers is to train a voice that the client is sustaining a pitch that the vocal mecha-
is somewhere in the “middle range” between non- nism was not constructed to produce. It must be
transgender English-speaking men’s and women’s done efficiently, both to sound like natural female
voices–between 155 and 185 Hz. However, trans- speech and also to avoid the development of voice
sexual women frequently prefer a higher target problems. Transferring this easy, resonant, effi-
pitch, so some experimentation may be necessary to cient method of producing a higher pitched voice
establish what is both possible and optimal. into everyday life is both challenging and impor-
Once the body and voice have been “warmed up” tant.
using the previous exercises, we start the voice on a In doing these exercises the client begins to de-
2 to 3 second /m/ at F3 or 185 Hz. Because this is a velop a physical sense of how she can produce a
group program, we use one pitch for practice; in feminine voice and an aural sense of what it sounds
one-to-one sessions it would be possible to choose a like. It will necessarily sound very different from
target pitch that matches a client’s individual goals her male voice. This altered perception can be quite
and existing vocal capacity. F3 (185 Hz) is a train- disorienting and it is essential to have a time period
ing pitch, not a target for average speaking pitch: it of adjustment to play around with what is possible
is higher than what most transsexual women will and what may be the best fit with the participant’s
use in everyday speech. However, it is beneficial for personality and sense of self.
participants to experience the sensations of produc- As pitch work progresses, the average speaking
ing a voice without strain that is much higher than pitch is checked periodically. There is no expecta-
their accustomed pitch. If this pitch produces strain tion that the average pitch will remain at the target
for any participants, we lower the target pitch to one training pitch of 185 Hz, but if it drops below 155 Hz
that can be produced with feelings of ease. A num- there needs to be further work producing a higher
ber of CK participants have commented that F3 voice in sustained sounds. At this point, practice in
(185 Hz) is too low and they use a higher one when vocal inflections is begun, along with practice in
doing this practice at home. producing the higher pitches.
Shelagh Davies and Joshua M. Goldberg 195

Vocal Inflections Vocal Quality: Breathy versus Resonant Speech

This phase of treatment begins with a discussion Among English-language speakers, mildly breathy
about English-language vocal inflections and speech is associated with feminine voice. Many
associations with “femininity” and “masculinity,” transgender women have already adopted a breathy
to determine clients’ perceptions and existing voice by the time they seek therapy. Mild breathi-
knowledge. In feminizing vocal inflections among ness also has the advantage of automatically modi-
English-language speakers the goals are to decrease fying hard attacks on consonants and vowels, giv-
flat inflections, increase inflectional range, and in- ing speech a softer quality. However there is a
crease vocal flexibility (the amount the voice moves contradiction between resonant voice, which is the
around within a phrase, rather than the extent of focus of CK, and breathy voice. A breathy quality is
pitch excursions). produced with less efficiency so the voice may be
As with other topics, we start work on vocal in- more prone to vocal fatigue and not be heard against
flections by listening. For this purpose I use a tape of background noise.
eight speakers describing a picture. The speakers This contradiction is discussed as part of the
are males and females of different ages and cultural group sessions and participants generally report in-
backgrounds. We listen specifically to the vocal in- tuitively finding their own ways of dealing with
flections used by the speakers, paying particular at- voice quality issues. For participants experiencing
tention to different patterns used by men and throat pain or vocal fatigue, the resonant voice
women. Clients are also instructed to listen to con- works best as it lasts longer and is louder; other par-
versations in their community and pay particular at- ticipants feel more comfortable with a breathier
quality as it better conveys the impression they
tention to vocal inflections.
want. Some participants adopt a resonant voice in
Individuals who have habitually used little vocal
loud situations and a breathier one in quiet ones.
inflection in speech often find that expanding the in-
This ability to change vocal qualities requires good
flectional range feels embarrassing and artificial. In control over voice production and may be a reason-
exploring vocal inflectional range clients are en- able goal for some transgender women who are con-
couraged to go “over the top,” far beyond what they cerned about voice quality issues.
would realistically use in speech. This can have a
freeing effect and also allow the client to experi- Vocal Loudness
ment without being restricted by what would be
considered appropriate; refinements happen at a CK participants typically struggle more with
later stage of the program. achieving adequate loudness in a noisy environ-
Exercises initially use limited vocabulary so the ment than an inappropriately loud speaking voice.
client must use a range of vocal inflections to con- Using a resonant voice increases loudness in an effi-
vey meaning and emotional expression. As in any cient and effective way, and can be trained specifi-
standard speech therapy protocol, the complexity of cally to be used where there is a lot of background
the task increases as the person’s performance im- noise. If loudness is a concern for participants, the
proves. Carryover into everyday life can be facili- group does a vocal exercise involving repetition of a
tated by choosing a specific phrase or sentence that phrase with differing levels of loudness. The goal is
the client uses frequently. to increase the loudness by increasing resonance
Work with inflections continues throughout the sensation, not by pushing from the throat. If the res-
program, as this is an important aspect of speech and onant voice is judged as too loud, there are specific
also one that usually takes time to change and habit- training exercises that reduce loudness while main-
uate. As with pitch, clients frequently report they taining forward focus; as discussed earlier, adopt-
need to monitor these vocal parameters constantly ing a breathy quality will automatically reduce
during conversation. For this reason it is useful in loudness.
the early stages of therapy to choose specific prac-
tice times when the client will be conscious of Motor Speech Characteristics
speech and voice production and use the techniques
learned in therapy. The client is asked to begin with Hard onsets on initial vowels and consonants are
a person or place that is “comfortable” or “easy,” generally considered a masculine speech character-
and gradually extend the practice rather than con- istic among English-language speakers (Andrews,
fronting very difficult situations right away. Build- 1999; Gold, 1999). Adopting a breathy voice qual-
ing confidence in the new speech and voice is an im- ity may be enough to soften the onsets so they are no
portant part of the therapy program. longer perceived as abrupt; conversely, softening

the onsets may give a softer, breathier quality to the Topics discussed include in this phase of the pro-
voice. As it is easier to modify a general feature of gram include: (a) the use of qualifiers and tags, such
voice production than to specifically change each as “isn’t it,” “sort of,” “kind of,” “don’t you think,”
production of an initial phoneme, paying attention “I think that,” and “could you possibly,” (b) sharing
to voice quality may be the easier way to achieve a difficulties and problems as a means of establishing
cluster of goals. We discuss articulation as we are connection, (c) confirming the speaker’s emotional
experimenting with voice quality so participants messages, (d) making comments about another
are aware of the interaction. woman’s clothing or appearance, (e) direct versus
Encouraging more connected speech production indirect confrontation, (f) listening styles and be-
can also help reduce the abrupt interruptions in haviors, and (g) cues relating to conversational
speech flow that hard onsets create. This technique turn-taking and interruptions.
is similar to vocal prolongation used in fluency ther- Role playing is typically used for therapy exer-
apy; however, in this instance, the speech rate is cises that target language choice and discourse
maintained at a normal or near normal level. pragmatics. For example, if the purpose of the exer-
In working with articulation we listen to exam- cise was to use words that convey more emotional
ples of “feminine” and “masculine” patterns or lis- content, the participants could describe a picture,
ten to the speech of a group participant who already focusing on its emotional impact. For exercises fo-
uses connected speech and gentle onsets. The group cusing on development of active listening skills,
then repeats specific phrases or words, to get the participants may practice in pairs; the listener is in-
feeling of that kind of production. Due to time con- structed to encourage the speaker and actively show
straints the group generally does not spend a signifi- that she is paying attention, while the speaker must
cant length of time on articulation; if it is of particu-
seek the listener’s opinion and involvement in the
lar concern to an individual, intervention may take
conversation. If the goal was to make a casual con-
place in a private session.
nection with another woman in a public place, two
participants could role-play having a casual conver-
Language and Discourse Pragmatics
sation in a public setting, such as chatting in a lineup
at a cashier, trying on clothes, or waiting at a bus
Although language choice and speaker-listener stop. To facilitate carryover, participants are asked
interaction during conversation are influenced by to practice the issue addressed in the group in
many factors other than gender, there is a body of
real-world situations, as homework.
work in the sociolinguistics and popular literature
specific to gender influences on communication.
We discuss this literature in CK, both to give a point
of reference and to stimulate debate and observation PRELIMINARY EVALUATION
skills. Participants then see whether or not what is OF CHANGING KEYS
written in textbooks is actually happening in their
own communities. This encourages context-spe- CK is a new program, and evaluation of its effec-
cific norms that are flexible and can easily be tiveness is still in the early stages. Preliminary re-
adapted to suit the client’s personality and situation. sults suggest a range of outcomes, with some partic-
Also, since word choice and interaction in conver- ipants experiencing more significant change than
sation vary greatly from person to person and situa- others. On a self-evaluation questionnaire most par-
tion to situation, training specific behaviours is too ticipants noted positive changes in their speech and
rigid. The participants are encouraged to consider voice, and expressed increased confidence when
what women and men in their own communities are speaking. Further evaluation is needed to assess the
doing, and to determine which patterns feel com- program’s effectiveness and refine the CK proto-
fortable to them. col.