Comparison of Sex-Typed Motor Behavior in Male-to-Female Transsexuals and Women
David H. Barlow, Ph.D.,1 Joyce R. Mills,2 W. Stewart Agras, M.D.,3 and Debra L. Steinman4
Eight presurgical male-to-female transsexuals emitted significantly more feminine behaviors on a validated sex role motor behavior checklist than eight feminine women. Transsexuals also emitted more masculine behaviors than feminine women, although the total number of masculine behaviors in each group was far less than the total number of feminine behaviors. Indications for the management of transsexuals as well as the functions of sex role motor behavior in our society are discussed.
KEY WORDS: sex-typed motor behavior; transsexualism; masculinity; femininity.
1Department of Psychology, State University of New York at Albany, Albany, New York 12222.
2Nashville, Tennessee 37211.
3Department of Psychiatry, Stanford University, Stanford, California 94305. 4Department of Psychology, University of Rhode Island, Kingston, Rhode Island 02881.
It is widely reported that transsexuals think, feel, and behave as members of the opposite sex (Benjamin, 1966; Green and Money, 1969). This implies that there may be no systematic psychological differences between male transsexuals and women on the one hand or female transsexuals and men on the other hand. In a study of effeminate boys, Green (1974) has observed scores on psychological tests very similar to those of same-aged girls, lending some support to this hypothesis.
Similarly, Pauley (1974) has observed that female transsexuals obtain scores on attitudinal inventories that are in the normal range for men, although Money and Brennan (1968) noted that scores on the M-F scale of the Guilford Zimmerman Temperament Survey were slightly more feminine than those of "normal" men in a small sample of female transsexuals. Nevertheless, as Pauley (1974) points out, female transsexuals adapt quite readily to masculine sex roles.
Some clinicians, however, have reported that adult male transsexuals appear more feminine than women. For example, Pauley (1974) notes that female transsexuals "accept the masculine role more naturally than their male counterparts who frequently exaggerate the feminine role and portray a caricature of a woman. This hyperfeminine portrayal by some male transsexuals is immediately recognized by an overwhelming aroma of perfume, seductive behavior, attire which is inappropriate to the occasion." This interesting observation of a possible difference in the behavioral expression of mistaken gender identity between male and female transsexuals may be observed in their responses to attitudinal and interest inventories in which some male transsexuals score more "feminine" than the average for "normal" women (Walinder, 1967; Money and Primrose, 1968; Pauley, 1974).
While these differences are not statistically significant, Money and Primrose (1968) suggest that these trends indicate "a high degree of verbal adherence to female stereotypes" while cautioning that the effect may be due to social desirability. Thus, while suggestive, the results from attitudinal and interest tests do not yet clearly confirm the clinical impressions mentioned by Pauley (1974) that male transsexuals exaggerate feminine identification.
Another way to compare transsexuals and nontranssexuals is to measure sex role behavior more directly, in the tradition of Birdwhistell (1970), who examined sex role specific motor behavior, or what he termed gender signals. Recently a behavioral checklist of sex role motor behavior has been developed and validated (Barlow et al., 1979b). This checklist specifies sex-typed behavior occurring while sitting, standing, or walking. The purpose of this experiment was to explore the clinical impression that male transsexuals are more "feminine" than women, using the sex role motor behavior checklist.
Subjects included eight white women, ages 18-35, who were employed at a major medical center. These women were employed as nurses, nurses aides, medical students, etc., but the majority were nurses. Eight presurgical maleto -female transsexuals who came consecutively to our sexual problems clinic were also included. All were white with an age range of 17-26. Some were employed in diverse settings, others were unemployed or in school. Three had been cross-living at the time of their initial appointment for periods ranging from 1 month to 1 year. The remaining five, while cross-dressing frequently, had not been cross-living.
All subjects were rated on the sex role motor behavior checklist (Barlow et al., 1973, 1979b) in a manner described below. This checklist contains a number of possible sex-typed motor behaviors that occur while sitting, standing, or walking. Examples from sitting include buttocks away from the back of the chair, which is sex-typed masculine, and buttocks close to the back of the chair, which is sex-typed feminine. An example from walking is feet placed as if straddling a line, which is sex-typed masculine, and feet placed as if stepping on a line, which is sex-typed feminine (see Barlow et al., 1979b, for a complete description of this checklist). During a specific observation, either masculine-typed behaviors, feminine-typed behaviors, or in some cases both might be endorsed by a rater during a particular occurrence of sitting, standing, or walking. Reliability coefficients among independent raters are high, and data on validity, where sex role motor behaviors are separated from gender, are presented in Barlow et al. (1979b). Typically, the subject is observed for approximately 10 sec. by a trained rater who then checks appropriate behaviors on the checklist.
In this study, an observer was asked to seek out women working in the Medical Center who appeared to be particularly "feminine." Two independent raters were then directed to the working location of these women, where they surreptitiously rated sitting, standing, and walking. The reliability coefficient between these two raters on the overall checklist was 95.5%.
Since the sex role motor behavior checklist was a routine part of our assessment in the Sexual Problems Clinic, all patients coming through this clinic were surreptitiously rated by one of the same raters during the first session. People belonging to a large number of diagnostic categories came through this clinic, from rapists and exhibitionists, who typically do not emit feminine sex-typed motor behavior, to transvestites and transsexuals, who do emit some feminine sex-typed motor behaviors. At the time of the rating, the rater was unaware to which diagnostic category these patients would eventually be assigned. Assessment was completed by the clinic staff, and ratings for the first eight consecutive patients who were finally diagnosed transsexual were retrieved from their files.
The total number of masculine sex-typed behaviors on the scale in use at the time of this study for sitting, standing, and walking was 6, 4, and 5, respectively. The total number of female sex-typed behaviors for sitting, standing, and walking was 6, 4, and 6, respectively (see Barlow et al., l979b).
Feminine Sex Role Motor Behaviors
The female subjects emitted a total of 55 feminine sex role motor behaviors. This total represents 95% of the total number of feminine and masculine motor behaviors observed. The presurgical transsexual subjects emitted a total of 80 feminine sex role motor behaviors. This total represents 85% of the total number of masculine and feminine behaviors observed.
Fig. 1. Feminine sex role motor behaviors. The figure compares the total feminine sex role motor behavior scores of eight females to those of eight presurgical transsexuals. The total scores include the total number of feminine sitting, standing, and walking behaviors observed and recorded on the Motor Behavior Checklist for each individual subject.
Fig. 2. Masculine sex role motor behaviors. The figure compares the total masculine sex role motor behavior scores of eight females to those of eight presurgical transsexuals.
Figure 1 presents the individual subjects' total feminine motor behavior scores. The X feminine sex role motor behavior score for the females was 6.87. The X feminine sex role motor behavior score for the transsexuals was 9.88 (see Fig. 4). The X difference between groups was 3.01. A one-tailed sign test indicated that transsexuals emitted significantly more feminine behaviors than women (p < 0.01).
Masculine Sex Role Motor Behaviors
The females emitted a total of three masculine sex role motor behaviors. This total represents 5% of the total number of feminine and masculine motor behaviors observed. The transsexuals emitted a total of 14 masculine sex role motor behaviors. This total represents 15% of the total number of masculine and feminine motor behaviors observed. This difference was also significant by a one-tailed sign test (p < 0.02).
Figure 2 presents the individual subjects' total masculine motor behavior scores. The X masculine sex role motor behavior score for the females was 0.37. The X masculine sex role motor behavior score for the transsexuals was 1.75 (see Fig. 4). The X difference between groups was 1.38.
Difference Between Masculine and Feminine Motor Behavior Scores A difference score was computed for each individual subject by subtracting the total number of observed masculine sex role motor behaviors from the total number of observed feminine sex role motor behaviors (see Fig. 3). The overall X differences score for the females was 6.5. The overall X difference score for the transsexuals was 8.13 (see Fig. 4). In addition, the total number of masculine plus feminine behaviors observed for the females was 58, the mean being 7.25. The total number of masculine plus feminine behaviors observed for the transsexuals was 94, with a X of 11.75.
Fig. 3. Difference scores between feminine and masculine sex role motor behavior scores. Difference scores were computed by subtracting the total number of masculine motor behaviors from the total number of feminine motor behaviors for each individual subject.
Fig. 4. Overall X feminine, masculine, and difference scores. The figure compares the overall X feminine sex role motor behavior score, the overall X masculine sex role motor behavior score, and the overall X feminine minus masculine difference score of the female group to those of the presurgical transsexual group.
This study indicates that male transsexuals emit more feminine behaviors than feminine women in these particular situations. It is possible that transsexuals were attempting to behave in a particularly feminine manner, but observations were surreptitious, so no extrinsic demand, such as impressing a therapist, was in effect. Since motor behavior is one of the more prominent factors contributing to our gender impressions during initial interpersonal encounters (Weitz, 1976), it is possible that clinical judgments of exaggerated femininity are based on these behaviors. This finding has implications for the management of transsexuals. If the goal of a transsexual is to be a normal (albeit feminine) woman, then both pre- and postsurgical transsexuals could be taught to lower the number of feminine behaviors emitted to more closely match that of feminine women. Since more dramatic alterations in motor behavior have been effected in transsexuals wishing to do so (Barlow et al., 1979a), these slight alterations could be accomplished, presumably without difficulty.
Of particular interest in this study, however, is the observation that, despite the emission of excessive numbers of feminine behaviors, male transsexuals also continue to emit more masculine behaviors than feminine women. Only one out of eight transsexuals failed to emit some masculine behaviors during the observation, while only three out of eight of our "feminine" women emitted any masculine behaviors at all. It is possible that emission of occasional masculine behaviors in the context of high-frequency feminine behaviors (as well as feminine appearance, clothing, etc.) might be as much of a basis for transsexuals being "read" or identified as males as the exaggerated number of feminine behaviors. Given the importance of these behaviors to gender impressions (Weitz, 1976; Birdwhistell, 1970), this deserves to be tested.
It is, of course, possible that raters were influenced by gender in scoring the occurrence of masculine behavior, but this is unlikely since the scoring of feminine behaviors would contradict this possibility, and since gender does not seem to influence rating of sex role behaviors in a study designed to answer this question (Barlow et al., 1979a).
Finally, it is probably inappropriate to suggest to transsexuals in treatment, or anyone, that sex-typed motor behavior should be carried inflexibly and stereotypically into all social situations. For example, one who appeared extremely masculine or extremely feminine in all situations might not be as socially effective as one with more flexibility. This "flexibility" has been hypothesized to occur with sex-typed interpersonal behavior (e.g., Bem, 1974; Kelly and Worrell, 1977) and might be termed "behavioral androgyny" (Barlow et al., 1979a). Heckerman, Schoen, and Barlow (unpublished) have found some interesting examples of situationally determined sex role motor behavior in college students. More recently, a transsexual who desired to reorientate to his anatomical gender (Barlow et al., 1979a) later reported that one of the greatest benefits of treatment was an ability to be more "flexible" by behaving in either a masculine or a feminine way as was "appropriate" (based on his judgment) to the situation.
Since male transsexuals emitted more sex-typed behaviors than women when observed in this one situation, it is also possible that male transsexuals are more stereotypically rigid across all situations, further accounting for some of the clinical impressions of "exaggerated femininity." This has yet to be tested.
While continued study of sex role motor behavior should benefit transsexuals, whether their goals are surgery or, as occasionally happens, reorientation to their own gender, these data have broader implications. In the best traditions of the study of abnormal psychology, close examination of sex role behavior in transsexuals should illuminate some of the larger questions concerning the development and phenomenology of sex roles and the relationship of various aspects of sex roles to interpersonal relations and adjustment.
Barlow, D. H., Reynolds, E. J., and Agras, W S. (1973). Gender identity change in a transsexual. Arch. Gen. Psychiat. 28: 569-579.
Barlow, D. H., Abel, G. G., and Blanchard, E. B. (1979a). Gender identity change in transsexuals: Follow-up and replication. Arch. Gen. Psychiat. 36: 1001-1010.
Barlow, D. H., Hayes, S. C., Nelson, R. O., Steele, D. L., Meeler, M. E., and Mills, J. R. (1979b). Sex role motor behavior: A behavioral checklist. Behav. Assess. 1: 119-138.
Bem, S. L. (1974). The measurement of psychological androgyny. J. Consult. Clin: Psychol. 42: 155-162.
Benjamin, H. (1966). The Transsexual Phenomenon, Julian Press, New York. Birdwhistell, R. (1970). Kinesics and Context, University of Pennsylvania Press, Philadelphia.
Green, R. (1974). Sexual Identity Conflict in Children and Adults, Basic Books, New York, Penguin, 1975.
Green, R., and Money, J. (1969). Transsexualism and Sex Reassignment, Johns Hopkins Press, Baltimore.
Kelly, J. A., and Worrell, J. (1977). New formulation of sex roles and androgyny: A critical review. J. Consult. Clin. Psychol. 45: 1101-1125
Money, J., and Brennan, J. J. (1968). Sexual dimorphism in the psychology of female transsexuals. J. Nerv. Men. Dis. 147: 487-499.
Money, J., and Primrose, C. (1968). Sexual dimorphism and dissociation in the psychology of male transsexuals. J. Nerv. Men. Dis. 147: 472-486.
Pauley, I. B. (1974). Female transsexualism: Part II. Arch. Sex. Behav. 3: 509-526.
Walinder, J. (1967). Transsexualism: A Study of Forty-three Cases, Scandinavian University Books, Goteborg.
Weitz, S. (1976). Sex difference in nonverbal communication. Sex Roles 2: 175-184.