Clinical Patterns Among Male Transsexual Candidates with Erotic Interest in Males
Frank Leavitt, Ph.D.,1,3 and Jack C. Berger, M.D.2 Male-to-female transsexuals who reported an erotic interest in males showed different patterns of sexual activity. Sexual history was used to categorize a transsexual sample into three groups: 44% abstained from sexual activity (Inactive group), 19% were sexually active but avoided using their penis in sexual activity (Avoidant group), and 37% were sexually active and derived pleasure from their penis (Pleasure group). The groups were compared for differences in gender identification, developmental patterns, and personality. Transsexuals in the Avoidant group showed patterns of traits and experiences that generally conformed to characteristics of the nuclear transsexual. They were dissimilar from the other two groups on measures of feminine functioning, heterosexual history, and fetishism. Transsexuals who interact with males in ways that are viewed as more classically homosexual shared more in common with the transsexual group which abstained from sexual activity with males. Both groups displayed more masculinity in development and more evidence of emotional disturbance. The implications of these findings for diagnosis and treatment are discussed. KEY WORDS: gender dysphoria; male-to-female transsexual; homosexual transsexual. 1Department of Psychology and Social Sciences, Rush Medical College, 1653 West Congress Parkway, Chicago, Illinois 60612. 2Department of Psychiatry, Rush Medical College, Chicago, Illinois. 3To whom correspondence should be addressed. INTRODUCTION Male-to-female transsexuals meeting DSM-III criteria (APA, 1980) for the gender disorder are heterogeneous in sex object preference. By far, the majority of males seeking reassignment are attracted to males and labeled homosexual transsexuals. Smaller percentages show heterosexual and bisexual attraction. The rest are asexual (Blanchard, 1985). The homosexual transsexual label is both confusing and controversial among males seeking sex reassignment. Transsexuals, as a group, vehemently oppose the label and its pejorative baggage (Morgan, 1978). As a rule, they are highly invested in a heterosexual life-style and are repulsed by notions of homosexual relations with males. Attention from males often serves to validate their feminine status. For many biological male transsexuals, acts of intimacy with women are truncated, because sexual attraction and relations with women pose the homosexual issue. Recent research presents a different view of sexual activity among male transsexuals. In a study of 44 transsexual males (Langevin et al., 1977), 88% (35/44) allowed a male partner to touch their penis, 29% (17/40) received fellatio, and 2% (1/44) performed anal intercourse. These findings indicate that erotic pleasure in the penis is more common among male transsexuals than formerly assumed and suggest that the homosexual label may be justified in a subgroup of transsexuals. These patients also differed on variables measuring emotional stability and cross-dressing. Those patients who derived erotic pleasure from their penis were more likely to be emotionally unstable, and less likely to have switched to full-time living as women, suggesting that clinicians may be dealing with different disorders that require separate explanations. This is in line with Stoller's claim (1973) that erotic pleasure in the penis does not exist in the true transsexual. Patients with histories of deriving erotic pleasure from the penis voice complaints of gender dysphoria that do not distinguish them from other patients seeking gender reassignment. Most meet DSM-III criteria which are broadly written. Their form of sexual gratification runs so counter to the majority view of the fundamental features of the transsexual disorder as to require clarification. At minimum, it may simply mean that popular notions of transsexualism that developed on an anecdotal basis need to be revised. On the other hand, the "atypical" pattern of sexual gratification could mean that patients presenting with this history are pseudo-transsexuals who really experience other disorders. In these cases, it may be legitimate to question whether the disturbance is one of gender, sexual disturbance, or psychopathology. The purpose of this paper is to offer a broader view of male transsexuals showing erotic interest in males. The general hypothesis is that transsexual males deriving erotic pleasure from the penis represent a different diagnostic entity and should show important developmental, gender, and personality differences that distinguish them from male transsexuals who do not involve the penis in sexual activity with males. METHOD Subjects The sample consisted of all biological males seeking presurgical psychiatric evaluation for sex reassignment in a gender identity clinic; 112 males who were otherwise unselected formed the initial study pool. These subjects were classified for erotic attraction to males using the Androphilia Scale developed by Freund et al. (1982) and later modified by Blanchard (1985). A score of 7.49 or higher on the modified Androphilia scale was required for inclusion in the study. This score was set so as to exclude patients with scores more than 1 standard deviation below the mean of Blanchard's homosexual transsexual group (X 9.86; SD 2.37). On the basis of the cutoff score, 81 subjects were erotically attracted to males and defined the homosexual transsexual study pool. The 81 subjects were divided into three groups on the basis of their answers to questions concerning prior sexual contact with males. Sexual contact was operationalized along six dimensions articulated by Langevin et al. (1977). These involve males masturbating the subject, males performing fellation on the subject, males performing anal intercourse on the subject, the subject masturbating other males, the subject performing fellatio on other males, and the subject performing anal intercourse on other males. These six areas were covered by 24 questions on a questionnaire. Based on answers to these questions three groups were formed. Criterion for inclusion in the Pleasure group was admission of five sexual experiences specifying sexual activity involving use of the patient's penis in either masturbation, fellatio, or anal intercourse. Admission of five sexual experiences with males, but not involving the subject's penis resulted in placement in the Avoidant group. The remaining subjects constituted the sexually Inactive group. This resulted in 30 Pleasure subjects, 15 Avoidant subjects, and 36 sexually Inactive subjects. Subjects were also classified into the same three categories on the basis of an extensive interview. All but four subjects were assigned to the same category as derived from questionnaire data. These four subjects were not included in the analyses. The three groups were comparable in age. The Inactive group was slightly more educated than the Pleasure group (see Table I). The three groups were quite discrete even though the cutoff score was arbitrarily set. In the Avoidant group, 10 never allowed their partners to touch their penis, 2 admitted to one experience, 2 admitted to two, and 1 to three. In the Pleasure group, the lowest number of experiences was 9, and at least one third of the group reported more than 100 experiences. In the Inactive group, over 75% reported no sexual experiences with males. Table I. Demographic Data of Patients in Three Transsexual Groups Group Avoidant Pleasure Inactive (n = 15) (n = 30) (n = 36) F value p Age Mean 29.9 32.8 34.5 1.7 ns SD 4.2 7.5 9.6 Education 13.2 12.7 14.6 6.70 0.002 Months living full time as a woman 59.4 37.1 8.8 5.80 0.005 Age of cross-dressing onset 9.0 9.0 8.3 0.28 ns The BEM Sex Role Inventory (BSRI), the MF scale of the Minnesota Multiphasic Personality Inventory (MMPI), and the Draw-A-Person Test (DAP) were used to measure aspects of gender identity. The BSRI (Bem, 1977) is a self-administered 60-item scale consisting of 20 personality characteristics that are stereotypically masculine, 20 that are stereotypically feminine, and 20 that are embedded as buffer items. Each item is scored along a 7-point scale ranging from never or almost never true (1) to always or almost always true (7). Two kinds of scores are obtained. One is a simple indicator of masculinity or femininity derived by adding masculine and feminine items separately. The other is obtained by classifying patient's scores as either falling above or below normative (Bem, 1981) median-split scores (4.90 for femininity and 4.95 for masculinity) and then classifying them as androgynous (above median on both masculinity and femininity), feminine or masculine sex typed (high on the named scale, and low on the other), or undifferentiated (below median on both). The MF scale of the MMPI provides a single measure of masculinity and femininity based on respondents' answers to 60 true-false questions dealing with interest, vocational choice, aesthetic preference, and activity-passivity dispositions. High scores (> 70) on the male scale imply femininity of emotional interest. Low scores imply masculinity of emotional interest. On the DAP, the subject is asked to draw a whole person (Machover, 1949). Following this drawing, the subject is asked to draw a person of the opposite sex to the first figure drawn. The gender of the first figure drawn is recorded and taken as an indicator of gender identification. It has been shown that 80 to 90% of males (Fleming et al., 1979) draw a male figure first. Two variables were measured from responses to the Rorschach test. Adequacy of reality testing was assessed using the F + percentage and number of popular responses seen by the subject. Sexual concern was assessed using the number of sex responses elicited in evaluation. The Rorschach was administered and scored according to Beck et al. (1961). Reliability of scoring has been documented (Leavitt and Garron, 1982). General psychopathology was assessed using the MMPI. The test was individually administered and scored for 3 validity scales and 10 clinical scales. Demographic and developmental data were obtained from a questionnaire included in the comprehensive psychological examination administered by a clinical psychologist with more than 10 years experience in examining transsexuals. The questionnaire consisted of items documenting age, education, marital history, age when subject first started cross-dressing, either fully or partially, years living full time in the cross-gender role, repugnance for genitals, and history of sexual attraction to women. Cross- gender fetishism was measured by an item borrowed from the work of Blanchard (1985). "Did you every feel sexually aroused when putting on females' underwear or clothing?" Developmental history involving preference for feminine toy play, feminine play partners, and avoidance of masculine activity was obtained using items on the Gender Identity Scales for Males (Freund et al., 1977). These items are detailed in Table II. Table II. Psychosocial Characteristics of Transsexual Males Varying in the Sexual Expression of Their Erotic Preference for Males Group Avoidant Pleasure Inactive (n = 15) (n = 30) (n = 36) Variable % % % X2 P Never married 100.0 76.7 52.8 12.2 0.002 Avoids masculine activity 73.3 40.0 52.8 4.5 ns Preference for feminine toy play 86.7 53.3 50.0 6.3 0.000 Preference for female play partners 86.7 63.3 52.8 5.8 0.050 Repugnance for own genitals 93.3 16.7 36.1 24.9 0.001 History of sexual attraction to women 0.0 33.3 58.3 15.7 0.001 History of fetishistic arousal 6.7 33.3 50.0 8.8 0.010 Table III. Sex-Role Identity Scores from the BSRI, the MF Scale, and the DAP Group Pleasure Avoidant Inactive (n = 30) (n = 15) (n = 36) F value p BSRI Masculine scale 4.60 4.66 4.77 0.397 ns Feminine scale 5.41 5.49 5.32 0.861 ns MMPI MF scale 83.70 82.20 84.80 0.690 ns X2 Value DAP Drew female first 60% 80.0% 65% 3.340 ns RESULTS Gender identity scores of the masculine and feminine scales of the BSRI were analyzed in two ways following the work of Bem (1977). First, one-way analyses of variance were used to test for gender difference among the three transsexual groups. Results of separate analyses of male and female scores are presented in Table III. There were no statistically significant gender differences among the three groups. Subjects in the three groups scored below the normative median for males (4.95) on the masculine scale and above the normative median (4.90) on the feminine scale (Bem, 1981). The second analysis used Bem's median-split scoring system to classify all subjects into four sex role identity groups. Subjects were labeled as masculine (high masculine-low feminine score), feminine (high feminine-low masculine score), androgynous (high masculine-high feminine score), and undifferentiated (low masculine-low feminine score) based on this separation. Percentages of the three groups falling into the four sex role groups are presented in Table IV. The relation between type of transsexual and sex role identity was not significant, X2(6) = 3.32, p > 0.77. As seen in Table IV, the majority of transsexuals in all three groups tended to be either feminine or androgynous. Table IV. Incidence of Sex-Role Identity in Transsexuals Varying in Sexual Activity with Males Group x-role Inactive Avoidant Pleasure identity (n = 36) (n = 15) (n = 30) Androgynous n 13 4 10
% 36.1 26.7 33.3 Feminine n 13 9 15
% 36.1 60.0 50.0 Masculine n 4 1 2
% 11.1 6.7 6.7 Undifferentiated n 6 1 3
% 16.7 6.7 10.0 Table V. Mean Scores of the Three Transsexual Groups on the MMPI Scales and Rorschach Variables Group Variable Avoidant Pleasure Inactive F value p (n = 15) (n = 30) (n = 36) MMPI scale L 54.4 50.1 49.9 1.81 ns F 51.0 60.1 57.5 2.40 0.05 K 56.9 56.1 57.1 0.09 ns Hs 48.2 58.2 56.0 2.47 0.05 D 64.9 64.0 65.1 0.07 ns Hy 58.5 62.5 62.9 1.21 ns Pd 61.1 70.4 68.3 2.52 0.05 Pa 55.5 62.8 63.3 2.59 0.05 Pt 58.8 62.8 62.9 0.66 ns Sc 57.7 70.1 68.3 2.84 0.05 Ma 52.4 59.2 55.6 1.65 ns Si 55.4 52.3 54.3 0.43 ns Rorschach Responses 42.8 44.6 43.9 0.87 ns F + % 70.5 57.8 51.1 9.60 0.001 Popular responses 7.4 6.6 7.2 0.63 ns Sex responses 0.7 6.4 5.7 4.87 0.01 Gender scores from the DAP and the MF scale of the MMPI are presented in Table III. At least 60% of each group drew the female first indicating a high degree of femininity in all three groups. Differences among groups were not statistically significant. In a similar way, their mean T scores of over 80 on the MF scale of the MMPI is consistent with their identification with the female sex. The differences among groups were again nonsignificant. A series of one-way analyses of variance were performed on MMPI scale scores transformed into K-corrected, T-score equivalents. Significant differences were observed among groups on five MMPI scales (see Table V). Post hoc testing using Duncan Multiple Range Test revealed the following significant (p < 0.05) between-group differences, with higher scores reflecting greater psychological impairment. The Pleasure and Inactive groups both scored higher than the Avoidant group on scales F, Hs, Pd, Pa, and Sc. The differences between the Pleasure and Inactive groups were nonsignificant on these five scales. The Avoidant transsexual sample produced a mean MMPI profile indicative at best of a modest level of psychological impairment. Only scores on Scales D and Pd were more than 1 standard deviation (T score of 60) above the normative mean score of 50. On the other hand, the peak scale scores on the Sc and Pd scale in both the Pleasure and Inactive transsexual groups are often obtained by males who show significant psychopathology in areas of general functioning. Rorschach variables are presented in Table V. Inspection revealed that the F + % and number of sexual content statistically distinguished the three groups. A control for number of responses was not needed as all groups produced approximately the same mean number of responses. Chi-square analyses with 2 degrees of freedom were used to compare frequencies of the three groups on variables listed in Table II. Chi-square tests for two independent samples were used for pairwise comparisons. To minimize overestimating chi-square values, Yates Correction for Continuity was applied to each analysis involving cell sizes of less than 5. The groups were significantly different on six of the seven variables listed. The Avoidant group differed from the two other groups on six of the seven variables listed. All 15 patients in this group were single and all denied a history of sexual attraction to females. By comparison, 10 of the 32 patients in the Pleasure group reported a history of sexual attraction; 8 of these had tried married life. In the Inactive group, 21 of the 30 patients reported a history of sexual attraction and 17 of these had married. Differences between the Pleasure and Inactive groups of these two variables were also significant. There was less attraction to females and fewer marriages in the Pleasure group. The Avoidant group also showed a much smaller incidence of erotic arousal to cross-dressing. Only one patient in the Avoidant group reported a history of fetishistic arousal. This compares with 10 of 30 in the Pleasure group and 18 of 36 in the Inactive group. The difference between the Pleasure and Inactive group was nonsignificant, X2(1) = 1.86, p > 0.17. The Avoidant group also differed from the other groups in the amount of feminine sex-typed behavior in the developmental history; 86% of the group reported a developmental history involving cross-sex preference for both play partners and play toys. The three groups also differed relative to attitudes regarding the penis. In the Avoidant group, 87% stated that the penis was repulsive, compared to 36.2% in the Inactive group, and 16.7% in the Pleasure group. The difference between the Pleasure and Inactive group was nonsignificant, X2(1) = 2.21, p > 0.14. Data relating to cross-dressing are presented at the bottom of Table I. While the groups did not significantly differ on mean age of onset, patients in the Avoidant group had lived significantly longer in the feminine role than patients in either of the other two groups. Differences in full-time living between the Pleasure and Inactive group were also significant (p < 0.05). DISCUSSION Level of gender disturbance is not a clinically discriminating variable. Whether measured by the Draw-A-Person Test, the MF Scale of the MMPI, or the Bem Sex Role Inventory, comparable levels of gender disturbance are found among the three transsexual groups, with a higher incidence of femininity and a lower incidence of masculinity noted. All three male groups are clearly more feminine than heterosexual males when comparisons are made to normative data (Gravitz, 1966; Bem, 1981). Clinically, however, transsexuals in the Avoidant group are different from transsexuals in the other two groups along a number of important clinical dimensions. On measures of feminine functioning, they appear to follow a different developmental pattern. They show greater amounts of cross-gender behavior early in life and more consolidation of feminine identity later in life. Feminine toy play and feminine playmates are almost universal in their histories; whereas only one in two report this in the Inactive group, and only slightly more in the Pleasure group. The Avoidant group also seems to have experienced less difficulty adopting the cross-gender role as a full time way of living. On average, they have been living full-time as women for 5 years; this contrasts with 3 years in the Pleasure group and less than a year in the Inactive group. Transsexuals in the Avoidant group are also distinguishable on the basis of heterosexual history. This group is remarkable for the absence of heterosexual behavior. As a group, they totally deny sexual attraction to females, and none report marriage. This compares with an attraction and marriage rate of approximately 50% in the Inactive group and 33% in the Pleasure group. Similar patterns are found for the variables measuring fetishistic arousal and attitudes towards the penis. Transsexuals in the Avoidant group share a common aversion to their penis and rarely experience fetish arousal. By contrast, at least 33% of the transsexuals in the other two groups report a history of fetish arousal and at least 60% deny being repulsed by their genitals. The Avoidant group also differs from the Inactive and Pleasure group on the basis of psychopathology. Transsexuals in this group appear psychologically healthy by scores on both the MMPI and the Rorschach. There is no evidence that their disorder is based on psychopathology as suggested by some (Roberto, 1983). Transsexuals in both the Inactive and Pleasure group show significant psychiatric impairment and interestingly emphasize sex content in responding to Rorschach stimuli. Prominence of sex content in protocols is generally interpreted as an indicator of disruptive problems in sexual adjustment (Philips and Smith, 1953). Transsexuals who are sexually active with males but do not allow their penis to be involved in sexual activity share a constellation of traits and experiences that generally conform to characteristics of the nuclear transsexual (Buhrich and McConaghy, 1977). The picture of the nuclear transsexual conveyed by the literature is that of sustained, nonfluctuating femininity, developmental patterns involving a preference for girls' games and company, aversion to rough-and-tumble activity in early childhood, later patterns involving a desire to posses a women's body, to live in society as a woman, and to attract heterosexual male partners. The nuclear transsexuals experience intense disgust and aversion for their penis and deny a history of heterosexual orientation or fetish arousal. If patients in the Avoidant group are accepted as nuclear transsexuals, then only 18.5% of our sample meet those criteria. This implies that most of our sample seeking reassignment are not nuclear transsexuals (Newman and Stoller, 1974). This supports Lothstein's claim (1982) that most "patients seeking surgery are secondary transsexuals, i.e., transvestite or effeminate homosexuals." Transsexuals who derive pleasure from the penis are clinically more puzzling. They interact with males in ways that are more classically viewed as homosexual, yet they share more in common with transsexual males who have been the least sexually active with males and the most heterosexually oriented. They are particularly similar to this group on variables involving early developmental patterns and psychopathology. Both display more masculinity in their developmental patterns and more emotional disturbance probably rooted in the sexual sphere. They are somewhat less alike in respect to repugnance for their genitals and for fetish arousal, but these differences are not statistically significant. Transsexuals who derive pleasure from the penis however display substantially less in the way of heterosexual patterns than do transsexuals in the Inactive group. They were less attracted to women, fewer had married, and most have lived longer in the feminine role on a full-time basis. The level of psychopathology in this group is similar to that of other transsexual groups who have involved their penis to a considerable degree in sexual relations. Langevin et al. (1977) reported significantly more suicide preoccupation and more abnormal MMPI patterns for this subset of transsexuals. The prominence of fetishism in this group is unusual since it is considerably higher than typically reported for transsexual groups. A fetish rate among homosexuals of only 8% was reported by Hellman et al. (1981). It was noted that the homosexuals who exhibited arousal to cross dressing also scored higher on femininity. This may suggest that one feminine form of homosexuality is fetishistic. If individuals of this nature come to look upon themselves as transsexual, this could explain the puzzling as sociation of penile pleasure and fetishism in this group. Other studies have found similar subgroups of transsexuals who show sexual patterns that conform to those of nontranssexual homosexuals. Bentler (1976) in a study of 42 postoperative male-to-female transsexuals classified 36% of this sample as homosexual on the basis of a five-item sex questionnaire. All members of this group admitted to viewing themselves as homosexual prior to surgery; following surgery, they were much more sexually active than either the heterosexual or asexual comparison groups. Sexual activity involving fellatio and anal intercourse was prominent, with pleasure reported by 73 and 33%, respectively. A 23% incidence of fetish arousal was observed among these patients. Their postoperative adjustment was poor. One-quarter reported that life as a female was not up to expectations in comparison to none in the two contrasting groups. Transsexuals in the Inactive group report characteristics that most depart from the nuclear transsexual pattern. Strong heterosexual orientations and fetish histories are prominent. They show the most difficulty in making the transition to full-time feminine living. Despite reporting erotic interest in males, the pattern exhibited generally conforms to that exhibited by patients referred to as heterosexual transsexuals (Buhrich and McConaghy, 1978). Studies by Blanchard et al. (1985) suggested that some heterosexual transsexuals adjust their histories of erotic preference to bolster their chances of receiving a positive decision on sex reassignment surgery. Since this group accounts for 40% of the sample and shares little in common with homosexuals, except a stated erotic interest in males, the use of the homosexual label for patients with a stated erotic attraction to males seems to have little merit; the label with its pejorative baggage may guide clinical thinking in a misleading direction. Adoption of more neutral descriptive terms such as androphilia (Freund et al., 1982) to indicate erotic preference for physically mature males may be desirable. The present study contains a potential methodological flaw that needs to be considered in evaluating the heterogeneity of results concerning fetish arousal. The study selected transsexuals with strong sexual attraction to males using a cutoff score of 7.49 on the Androphilia scale. It is possible that some transsexuals with a strong sexual attraction to males are bisexual rather than homosexual. Sexual attraction to females was not directly measured in this study. Instead, the 7.49 cutting score was calculated to eliminate 77% of bisexuals in our study using Blanchard's normative data (1985). His bisexual transsexuals had a mean score of 4.15 (SD 4.18) on the Androphilia scale. Based on probability statistics of the normal curve, it is estimated that only five patients in our final sample are bisexual. While fetish arousal is common among bisexuals, an n of 5 cannot account for the fact that 18 of 36 subjects in the Inactive group, 10 of 30 in the Pleasure group, and 1 of 15 in the Avoidant group report fetish arousal (see Table II). It remains possible that the cutting score selected was not as successful as planned in excluding bisexual transsexuals. The data suggest several kinds of clinical patterns among candidates for sex reassignment surgery who are erotically attracted to males that need to be differentiated in the development of topologies. These types cannot be understood on the basis of different levels of gender disturbances, since high levels of femininity are experienced by all groups. It appears clinically important to carefully document history of sexual activity with males since this variable seems to separate this part of the "presenting transsexual" population into three forms that are sufficiently distinct to raise the possibility that clinicians may be dealing with separate and distinct entities. Our data point to the centrality of sexual disturbance and psychopathology in two of these groups and raise the question whether clinicians should be more cautious when history places candidates in either the Pleasure or Inactive groups. Beyond that lies the question of whether these are genuine transsexuals or pseudo-transsexuals. Patients label their gender states in terms of the limited cognitions available and/or acceptable to them. Should clinicians accept the same? There may be entirely different paths for shaping feminine gender in males, and these paths may reflect entirely different etiologies and disorders. The data point to no specific etiology for those transsexuals who seem to fit the nuclear transsexual pattern best. Though speculative, they show the consolidation of feminine behavior and the stability of personality functioning that one might expect to see in a condition with an organic cause. They may represent the most appropriate group for surgical intervention. REFERENCES American Psychiatric Association. (1980). Diagnostic and Statistical Manual of Mental Disorders, 3rd ed. APA, Washington, DC. Beck, S. J., Beck, A. G., Levitt, E. E., and Molish, H. B. (1961). Rorschach Test: 1. Basic Processes, 3rd ed., Grune and Stratton, New York. Bem, S. L. (1977). On the utility of alternate procedures for assessing psychological androgyny. J. Consult. Clin. Psychol. 45: 196-205. Bem, S. L. (1981). Bem Sex-Role Inventory: A Professional Manual, Consulting Psychologist Press, Palo Alto, CA. Bentler, P. M. (1976). A typology of transsexualism: Gender identity theory and data. Arch. Sex. Behav. 5: 567-584. Blanchard, R. (1985). Typology of male-to-female transsexualism. Arch. Sex. Behav. 14: 247-261. Blanchard, R., Clemmensen, L. H., and Steiner, B. W. (1985). Social desirability, response set and systematic distortion in the self-report of adult male gender patients. Arch. Sex. Behav. 14: 505-516. Buhrich, H., and McConaghy, N. (1977). The discrete syndrome of transvestism and transsexualism. Arch. Sex. Behav. 6: 483-495. Buhrich, H., and McConaghy, N. (1978). Two clinically discrete syndromes of transsexualism. Br. J. Psychiat. 133: 73-76. Fleming, M., Koocher, F., and Nathans, J. (1979). Draw-A-Person Test: Implications for gender identification. Arch. Sex. Behav. 8: 55-61. Freund, K., Langevin, R., Satterberg, J., and Steiner, B. (1977). Extension of the Gender Identity Scale for Males. Arch. Sex. Behav. 6: 507-513. Freund, K., Steiner, B. W., and Chan, S. (1982). Two types of cross-gender identity, Arch. Sex. Behav. 11: 49-63. Gravitz, M. (1966). Normal adult differentiation patterns on the figure drawing test. J. Proj. Tech. Pers. Assess. 30: 272-273. Hellman, R. E., Green, R., Gray, J. L., and Williams, K. (1981). Childhood sexual identity, childhood religiousity, and 'homophobia' as influences in the development of transsexualism, homosexuality, and heterosexuality. Arch. Cen. Psychiat. 38: 910-915. Langevin, R., Paitich, D., and Steiner, B. (1977). The clinical profile of male transsexuals living as females vs. those living as males. Arch. Sex. Behav. 6: 143-153. Leavitt, F., and Garron, D. C. (1982). Rorschach and pain characteristics of patients with low back pain and "conversion V" MMPI profiles. J. Pers. Assess. 46: 18-25. Lothstein, L. N. (1982). Sex reassignment surgery: Historical, bioethical and theoretical issues. Am, J. Psychiat. 139: 417-426. Machover, K. (1949). Personality Projection in the Drawing of Human Figures, Charles C. Thomas, Springfield. Ill. Morgan, A. J. (1978). Psychotherapy for transsexual candidates screened out of surgery. Arch. Sex. Behav. 7: 273-282. Newman, L. E., and Stoller, R. J. (1974). Non-transsexual men who seek sex reassignment. Am. J. Psychiat. 131: 437-441. Phillips, L., and Smith, J. G. (1953). Rorschach Interpretation: Advanced Techniques, Grune and Stratton, New York. Roberto, L. G. (1983). Issues in diagnosis and treatment of transsexualism. Arch. Sex. Behav. 12: 445-473. Stoller, R. J. (1973). Male transsexualism: Uneasiness. Am. J. Psychiat. 130: 536-539.