A Comparative Study of Male Transvestites, Male to Female Transsexuals, and Male Homosexuals VERN BULLOUGH, BONNIE BULLOUGH, AND RICHARD SMITH One of the problems in the field of sex research has been a lack of empirical testing of definitions. In the past, the term homosexual, for example, has been used for individuals who are totally oriented towards the same sex, as well as bisexual persons (a 6, a 5, or even a 3 on the Kinsey scale, Kinsey, Pomeroy, & Martin, 1948). What is true for defining homosexuality is even more true for the phenomena of transsexualism and transvestism. The current definitions of who belongs to which group were constructed by an accretion of clinical wisdom, based mostly upon observations made by therapists who used the case study mode as their approach to a search for understanding. Usually, the individuals who were observed were being treated, so the treatment process may have entered into and even contaminated the data gathering process. The purpose of this research was to test some of the current assumptions about the characteristics of transvestites and male to female transsexuals by divorcing the research from the treatment process. Instead, the data were gathered from the points of view of the members of these groups themselves. It was hypothesized that this subjective viewpoint might reveal some commonalities which would distinguish transsexuals from transvestites and both groups from homosexuals or heterosexuals. These two groups were selected because the categories have been established comparatively recently, and there is not always agreement about who belongs to which group; there is still also considerable confusion of those two categories with homosexuality. If they arc to be separate categories,there ought to be a clear distinction among the three groups. Transvestism (TV), as a phenomenon, has a long history, ranging from mythical figures to medieval saints who cross dressed; from the many instances of berdache in anthropological literature to historical figures such as the Chevalier d'Eon (Bullough, 1974; Bullough, 1976). Still, the diagnostic category was so uncertain that Kinsey did not include it in his examples of sexual behavior (Kinsey et al., 1948), in spite of the fact that the term transvestism appeared in the writings of Hirschfeld in two different books entitled Die Tranvestiten (1925). Hirschfeld's term emphasizes the cross dressing aspect of transvestism. Later writers, mostly post-Kinsey, have added other characteristics, particularly after organized transvestite groups appeared in this country and in Europe. Stoller (1871), for example, held that transvestites have a fetish for the clothing of the opposite sex. Prince has attempted to refine the definition further by arguing that "true" transvestites are heterosexuals (Prince, 1978; Prince & Bentler, 1972), and, by implication, she indicates that homosexuals who cross dress are not transvestites. Prince, who classifies herself as a transvestite, has, at times, used other terms such as femmephile to distinguish her definitions from the more ambiguous term transvestism. The transsexual phenomenon (TS), historically, is more difficult to document (Bullough, 1976 Hoyer, 1933) since the surgical technology required to meet the current usage of the term was not possible until well into the 20th century. The term was introduced in the literature in the article "Psychopathia Transsexualism," by Cauldwell (1950). Benjamin popularized the term transsexualism in a lecture at the New York Academy of Medicine in 195 3 (Benjamin, 1954). In that same year, the term reached public attention when Christine Jorgensen, who had undergone sex change surgery, was featured in the public press (Bullough & Bullough, 1977). Interestingly, the first scientific description of the case labeled her a transvestite (Hamburger, Sturup, & Dahl- Iversen, 1953). Obviously, as in the case of transvestism, the term "transsexualism" is loaded with implications that go beyond the literal meaning of a desire to change sex. Benjamin, for example, felt that transsexuals considered their sex organs sources of disgust (Benjamin, 1954; Buhrich & McConaghy, 1977). On the basis of our extensive personal contact with transsexuals it is known that some who call themselves transsexuals and have had considerable surgery to achieve that end, however, stop short at their genitalia. This has been true of many female to male transsexuals. In our experience a genitalia change does not necessarily change sex preference, and there are numerous individuals who have completed surgery only to find out that their sex preference was the same as before they began surgery. There is, for example, a growing community of male to female transsexuals who are lesbians. Others change both their sex and their sex preference. The issue of just who is a transsexual has been complicated by the need, if only to protect the surgeon from charges of mutilation or malpractice, for some kind of counseling for those seeking surgery. Whether the gender dysphoria clinics accept multiple definitions of what constitutes the condition of transsexualism is undoubtedly debatable; among transsexuals and would-be transsexuals, however, there is a widespread belief that only certain responses are acceptable. The key element seems to be a statement by the patient that he (in the case of males) feels like a woman trapped in a man's body or vice versa (MacKenzie, 1978; Prince, 1978; Prince & Bentler, 1972). Is there any agreement among the groups themselves as to who belongs to which group? Though numerous descriptive studies have been conducted of various sexual groups, very few have been done of transvestites or transsexuals, perhaps because they are not so numerous and are not as well organized. The earliest large sample study of transvestite males was that by Prince and Bentler (1972). Over 500 subscribers to the magazine Transvestia were studied by means of a questionnaire focusing on their social and demographic characteristics. The readership and the responses, in retrospect, seem to have been deeply influenced by Prince 's own definition of the phenomenon. As editor of the journal Transvestia and as a missionary for transvestism, Prince had continually emphasized that transvestism was a heterosexual phenomenon which differed radically from transsexualism (Prince, 1978). Consequently, the magazine and the organized groups which grew out of it at first insisted on the Prince definition, and, in the groups at least, those who did not conform were re moved from membership. Still, in spite of this limitation, 11% of the sample members in the Prince and Bentler survey did not classify themselves as heterosexual. Overall, the subjects in this pioneering study tended to be well educated with substantial incomes and responsible jobs. Since the Prince and Bentler study, others have entered the field, but the number of studies re mains limited. Buhrich and McConaghy (1977, 1978, 1979) studied 34 transvestite and 29 transsexual men, along with 29 male homosexuals gathered together from three different sources. The transsexual group members had presented them selves to the Prince Henry Hospital Transsexual Clinic seeking sex change surgery; the transvestites were club members, and the gay males had sought psychiatric help because of their homosexuality. In a series of articles, they reported demographic and behavioral differences among their sample. The transsexuals were younger than the transvestites and they (the TS's) were more likely to dress fully and to report homosexual interests than the TV's. The TV's were older, and more likely to be heterosexual and to cross dress only partially. They con cluded that these were two clinical entities. MCauley and Ehrhardt (1977) studied 15 females requesting TS surgery who ware presented for treatment at the psychoendrocrinology clinic, and 15 lesbian volunteers from local gay groups. Tests of cognitive function did not discriminate between the two groups, but on the personality scales the transsexual women were more stereotypically male, whereas the lesbians were more androgynous. More female transsexuals held jobs that were within the male dominated domain, whereas lesbians filled more neutral work roles. The authors speculated that the more stereotypic responses may reflect a desire to be convincing as males. Studies Related to Childhood A part of the scenarios often used to describe individuals from one of the sexual minority groups has been to indicate that it was their childhood experiences which formed them. This has been particularly true of the psychoanalytic approach. Though only a few such studies have paid any detailed attention to transvestites and transsexuals, those dealing with homosexuality have sometimes made passing references to the topic. The so-called "Bieber Mom," the dominant, overprotective, and closebinding woman who superintended a family situation sufficiently pathological to induce sexual variation in her child, has been utilized to explain transvestism and transsexualism (Bieber, 1962, 1968). Socarides' (1968, 1978) belief, that a key element in sexual variation was a fixated wish for mother-child unity which resulted in behavior designed to forestall such a powerful affective state threatening to destroy the individual, has been used similarly. Although almost all recent studies have challenged the picture for homosexuals (Bell, Weinberg, & Hammersmith, 1982; Whitam, 1980), no such challenge has been made for transvestism and transsexualism. Stoller (1978) differentiated between two types of feminine boys. The first type resulted from a mother-infant symbiosis in which the father was totally absent. He saw this pattern as predisposing to child hood transsexualism. The second and more common picture in his conceptual scheme involved some separation from the mother and a situation of conflict marked by hatred and blackmail of the boy by the mother so that his sense of self as a male was threatened. Stoller visualized this as the precursor to castration anxiety which, in turn, led to transvestism, effeminate homosexuality, and transsexual patterns that are more conflict ridden than he first described. Stoller held that in the case of the transsexual in particular, the father would be physically absent but psychologically present in that he, the father, is constantly cited by the mother as a model of failure. The totally absent father is seen by Stoller as a less consistent but common feature of the childhood of both homosexual and transvestite males. Stoller (1976, 1978), however, questions whether the analysis of adults is a sufficient source of information about what went on in infancy. The psychoanalytic case study method has come under attack on the grounds that it favors the sought for diagnosis, making it possible to construct a pathological case for anyone, regardless of that individual's well-being (Karlen, 1971; Mavahedi, 1975; Weinberg & Williams, 1974). To avoid this charge Green (1974) approached the problem longitudinally in his study of feminine boys. Although his review of the literature indicated that there were a number of sociological and psychological variables for adult sexual behavior, he believed that child hood socialization was a key process. In his study, he identified such elements as a close relationship with the mother and conscious or unconscious encouragement of feminine behavior patterns as present in his case histories. In his study of 50 feminine boys, he reported that 75% cross dressed before their 4th birthday and only 3% after their 6th birthday, and that 93% of his boys began doll playing, which all children probably exhibit and emphasized that he was describing marked and persistent cross gender behavior. Green had originally studied five feminine boys; a follow-up study of these was reported by Money and Russo (1979), who found all five to be homosexual. In a follow-up on his own study, Green (1979) reported that in the nine cases where genital contact had occurred there was a high incidence of homo sexual and bisexual experience. Since most of the boys in his sample were still in their early or middle adolescence, it is probably too early to fully assess their adult sexual identity. One of the difficulties with a study such as Green's is that we are not yet certain what childhood factors to look for in any longitudinal study of transvestites and transsexuals. It might be that feminine behavior is only one factor or might not exist at all. Buhrich and McConaghy (1977, 1978, 1979) found the only statistically significant family-related variable was that more of the transvestite and transsexual subjects reported, that their mothers hoped for a girl prior to their birth. No such report was made by the control groups. In a Swedish study of 12 male and 12 female transsexuals, which compared their subject with male army conscripts and female nulliparous pregnant women, Uddenberg, Walinder, and Hojerback (1979) found significant differences between transsexuals and other young adults. The transsexuals were much more likely to report unsatisfactory relationships with parents than other young adults. The pattern described by Stoller (1978) of intense relationship with the opposite parent was found in only half of the male transsexuals and none of the females. They concluded that probably both parents were important in defining gender and sexual identity. The implications of these studies seem rather limited. They support the early genesis of these problems, as outlined by Money and Ehrhardt (1972), and suggest that this period is not touched by their type of retrospective studies. They suggest that gender misidentification may impair the parents' efforts to establish satisfactory relationships with the child. If this is the case, it would challenge some of the psychoanalytic concepts which put the burden on the parents. Perhaps even the psychoanalytic explanation is the result of, rather than a cause of, the behavior being described. The major recent study of homosexuality is that of Bell, Weinberg, and Hammersmith (1982). Their report also does much to dispel myths related to some of the early psychoanalytic assumptions relative to homosexuals. Focus of This Study From this literature review, there are some assumptions and finding which emerge that can be subjected to further testing. The first is the theme, pervasive throughout the psychoanalytic literature, of the absent father, as a basis not only for homosexuality, but also for transvestism and transsexualism. Other assumptions focus on feminine behavior of boys and childhood adjustment. These assumptions can be converted to the following hypotheses: a. Transvestites and transsexuals will be more likely to have grown up in a family headed by a female than will members of the sexually unidentifies comparison group but will be very similar to the homosexual comparison group. b. Transvestites and transsexuals will be less likely to have engaged in typical boy activities than will the sexually unidentified comparison group and will have a pattern very similar to that of the homosexual group. c. Transsexuals, because they are most rejecting of their childhood sex, will report a less happy childhood than any of the other groups. d. Transsexuals will report more academic difficulties as children than members of other groups. Several other foci, which also emerge from the literature review, were examined but not as major hypotheses: demographic and social variable, life histories, and current sex patterns of male to female transsexuals and transvestites as compared to comparison groups of gay males and the group of men whose sexual orientation was not identified. This article concentrates on sex patterns and childhood differences. Method Subjects Subjects were 65 transvestites (TV's), 33 male to female transsexuals (TS's), 57 gay males, and an undifferentiated control group of 61 men. Subjects, except for the undifferentiated control group, were identified by their membership in organized groups. These sexual subcultures have been described in several studies (Sagarin, 1969;Siegel & Zitrin, 1978; Warren, 1974). They are a well-known phenomenon in large cities and have been studied at least since the time of Hirschfeld (1910) in Berlin. In Los Angeles, a few of the more scholarly members of these groups are also members of, or friends of, the local chapter of the Society for the Scientific Study of Sex. These ties were used for locating members. The cooperating organizations included three transvestite sororities (including two groups whose members had split from the original Prince-sponsored TV group over definitional matters). The TS's were members of a Southern California transsexual support group which included both preoperative and post-operative persons who met for purposes of mutual support and socializing. The gay comparison group came from a random sample of an established city-wide gay organization and a nearby campus gay group. The second comparison group included men from occupations represented in the two samples, as well as a group of randomly selected college students whose age was in the range of the transsexual sample. The occupational population was obtained by searching through the library for local membership lists of professional groups or trade unions similar to the groups to which many of our transvestite and transsexual members belonged. This group was then randomly sampled to obtain a mailing list. This group did not receive the sexually oriented questions. They were used only for some baseline data comparisons on the non-sexually oriented questions. Questionnaire The questionnaire was a nine-page instrument in four parts. The first part dealt with background information about age, height, weight, occupation, living arrangements, marital status, education, in come, parents, political views, sports, etc. There were 28 items in this part. The second part was a standardized Spence-Helmreich Scale (1978) and consisted of 24 questions. The third part was the Texas Social Behavior Inventory (Helmreich, Stapp, & Ervin, 1974), a 16-question instrument. The fourth part consisted of a series of questions about sex, cross dressing, and related matters. These questions were pretested and revised, as necessary, prior to their use in this investigation. Procedure Each person was mailed a questionnaire. Mailing was done by the organizations, except in the case of the unidentified comparison group. The researchers furnished the postage. This procedure was followed to protect the identity of the person who did not wish to respond. The questionnaire was accompanied by a cover letter from us and the organization of which they were a member. It indicated that the respondent had been selected for participation in this research be cause of his membership in that organization. It explained that the purpose of the research was to clarify misconceptions concerning various erotic minority groups. All our credentials were described and our telephone numbers indicated for those who wanted further information. The overall response rate was 60%. The response rate for the three erotic groups was roughly comparable, but the response rate for the undifferentiated group was somewhat lower. A follow-up letter to the latter group produced additional responses, making the total percentage almost comparable to the other three groups. One obvious defect in this method of sample selection is that we reached no truly isolated or nongroup members. A second limitation in the study related to the sampling technique is the fact that studies have indicated that members of organized groups tend to be better integrated into society, are more affluent, and are better educated than the non-organized (Durkheim, 1951; Nisbet, 1953). In defense, however, the sample is more representative of the universe of transsexuals and transvestites than previous ones drawn from patients who have presented themselves for psychotherapy. Interestingly, how ever, 50% of the total sexually differentiated sample (including homosexuals) reported that they had sought some type of counseling or psychotherapy, with transsexuals being the most overrepresented in patient populations (transsexuals, 93%; transvestites, 54%; homosexuals, 29%), perhaps because of the requirements of reputable medical centers that there be psychiatric evaluation before hormone therapy or surgery. Results The first hypothesis, which dealt with family patterns, was tested in several ways. A summary of the responses to the question about who raised the subjects is shown in Table 1. No statistically significant relationship between group membership and parenting figures was found, x2 (6) = 8.56, p > .05. In fact, comparisons of national data for female heads of household (Table 2) suggest that absent fathers are no more common among the sexual minority groups than they are in the general population. Table 1 Summary of Responses to the Question: "By Whom Were You Raised?" Both Parents Mother Othera Transvestite (n = 65) 85% 8% 8% Transsexual (n = 33) 79% 6% 15% Homosexual (n = 57) 81% 16% 4% Undifferentiated (n = 61) 89% 7% 2% Note. X2(6) = 8.56. p > .05. aIncluded father, grandparents, other relatives, and adoptive parents. Table 2 National Statistics on the Percentage of Female Heads of Households 1978 25% 1970 21% 1960 18% 1950 15% 1940 15% 1930 19% 1890 14% Note. From Historical Statistics of the United States: Colonial Times to 1970, United States Bureau of the Census. Washington, D.C.: U.S. Government Printing Office 1975; Statistical Abstract of the U.S., 1979, United States Bureau of the Census. Washington, D.C.: U.S. Government Printing Office, 1979 p. 47. A summary of the data pertaining to mother's work status is contained in Table 3. Whether the subjects' mothers worked when they were children emerged as a statistically significant relationship. The transvestite group presented the most extreme pattern: Members of this group were much more likely to have had mothers whose occupation was listed as "housewife." Whether this is an important finding or simply a function of the age and economic standing of the TV group as compared to the others is not clear. Other data, not shown in tabular form, were also sought to determine whether family constellation and placement were important. Data relating subjects' birth order (older, middle, younger, or only children) to the four groups did not result in a statistically significant association. So, if the mother tended to smother or favor a child, she apparently did so in a way that was not related to birth order. Number of male and female siblings also proved to be nonproductive. Thus, the psychoanalytic picture of the overwhelming mother and absent father was not supported by the data. The second cluster of hypotheses dealt with the masculine vs. feminine behavior because of the emphasis on this behavior as a marker for later gender and erotic development. The question was asked if sports were an important part of the subject's adolescent life, and whether their involvement in sports was a participant or an observer. Here, the TV's and the sexually unidentified group stand out as most interested in sports, both as participants and as observers. These data are summarized in Table 4. Table 3 Subjects' Mothers' Work Status Transvestites Transsexual Homosexuals Undifferentiated (n = 58) (n = 30) (n = 52) (n = 60) Housewife 81% 63% 56% 53% Occupation indicated 19% 37% 44% 47% Note.: x2(3) = 11.71, p < .01. Table 4 Summary of Responses to the Question: "Were Sports an Important Part of Your Adolescent Life?" Responses Transvestites Transexuals Homosexuals Undifferentiated (n = 65) (n = 33) (n = 57) (n = 61) No 23% 64% 51% 26% Somewhat, as an observer 6% 6% 11% 3% Somewhat, as a participant 26% 12% 26% 26% Yes, as an observer 14% 6% Ñ 3% Yes, as a participant 31% 12% 12% 41% Note.: x2(12) = 31.90, p < .0001. Table 5 Age First Cross Dressed Age Transvestites Transsexals Homosexuals (n = 63 (n = 33) (n = 51) Up to age 10 73% 21% 7% 11-15 19% - - 16-20 2% 3% 2% 20-29 5% 3% - 50-59 2% - - Never cross dressed - 73% 91% x2(10) = 105.56, p < .001. Table 6 Responses to an Open Ended Question Which Asked How Happy the Respondents Were as Children. Responses Transvestites Transsexuals Homosexuals Undifferentiated (n = 64) (n = 32) (n = 55) {n = 61) Happy 38% 16% 64% 60% Mixed 39% 34% 16% 21% Unhappy 23% 50% 20% 19% Note. x2(6) = 29.75, p < .0001. One of the marks of feminine boys is youthful cross dressing and female role modeling. Table 5 shows patterns of cross dressing. Here, the major difference appeared between the transvestites and the other groups. (The unidentified group was not asked this question.) Cross dressing, where it existed started at an early age in all three groups, but as an index of later behavior it would seem to be more predictive of transvestism than either transsexualism or homosexuality. The transsexuals seem to be less focused on clothing than the transvestites, and comparatively few of the homosexuals crossdressed as children. This seems to be contrary to the Green (1979) findings, but it could be that the TV's learned early to compensate for their cross dressing by playing the masculine role fully enough to avoid questions in the minds of their parents. There was a cluster of variables related to child hood adjustment which emerged as significant and tended to differentiate the groups from each other. Respondents were asked how they would describe their childhood. In order to get the subjects' impressions, each was asked to indicate whether they were happy, lonely, full of guilt, etc. Responses were categorized as happy; as mixed or neutral, which included both negative and positive responses; and as unhappy, with only negative adjectives. These retrospective views of childhood appear in Table 6. The transsexual group stands out as the most unhappy, with the TV's intermediate, and the gay group the most happy. Although not at such an extreme level, the transsexual group also perceived themselves as poorer students. Table 7 shows responses to the question. "How strong were you academically in high school?' Three choices were supplied: (l) Excellent (A to B), (2) Fair (B to C), Table 7 Summary of Responses to the Question: "How Strong Were You Academically in High School?" Academic Strength Transvestites Transsexual Homosexuals Undifferentiated (n = 65) (n = 33) (n = 57) (n = 61) Excellent student (A to B) 35% 30% 47% 59% Fair student (B to C) 54% 49% 53% 36% Poor Student 11% 21% - 5% Note. x2(6) = 21.57, p < .001. Table 8 Sexual Orientation: Responses to the Question: "Are You More Sexually Attracted to Males, Females, Both, or Neither?" Attraction to Transvestites Transexuals Homosexuals (n = 65) (n = 33) (n = 55) Females 82% 24% 2% Males 6% 52% 80% Botha 12% 17% 17% Neither 2% 6% 2% Note. x2(6) = 98.65, p < .0001. a"Both" included people who were attracted to males at one time in their life and females at another time as well as currently active bisexuals. Sexual Orientation Since the inclusion of the subjects in the study was determined by their ties to a subcultural grouping and these cultures carried a sexual identity with them, we anticipated that there would be considerable uniformity within the sample groups. This did not turn out to be the case. Table 8 indicates the responses to the question, "Are you more sexually attracted to males, females, both, or neither?" This question was not asked of the sexually unidentified control group. Most noteworthy is the degree of blurring of the lines between the groups. Although the majority of the homosexual sample was, in fact, homosexual, and the majority of transvestites was heterosexual, an ambisexual minority was evident in all the groups. This suggests that a definition of transvestism limiting the phenomenon to heterosexual males is much too confining. Even among groups who proclaim their heterosexuality there is considerable ambivalence. Some of the transsexuals had serial orientations: having been attracted to females at one point in their lives and then, having decided they were transsexuals, they switched over. Not all changed their preference in sexual partners, and one of the post-operative transsexuals included in the sample is a leader in the lesbian movement. Though such a situation is decried by Raymond in The Transsexual Empire (1979), it should be emphasized that a desire to change one's sexual identity may or may not involve a change in sexual orientation. Table 9 Gender Identity & Self-descriptions Descriptions Transvestite Transsexuals Homosexuals (n = 65) (n = 33) (n = 57) Woman trapped in male body 11% 64% 4% Man trapped in female body 2% - - Man with feminine side 46% 6% 14% Woman with masculine side 3% 3% 4% Person who enjoys opposite clothing 37% - 4% Other: a. Gay - - 75% b. Woman - 27% - c. Chromosomal Disorder 2% - - Note. x2(14) = 135, p = .0001. Sexual Identity In an effort to see how some of the phrases from the sex and gender identity literature matched with the self-perception of the subjects, a question was posed asking respondents to describe gender identity and interest in cross dressing in some of the terms used by the various sexual subcultures. Those terms and the responses are shown in Table 9. As can be noted, 75% of the members of the gay sample rejected these alternative. Their self- perception was in terms of their homosexual sex preference. Only a few endorsed statements suggesting any blurring of gender identity. Members of the transvestite sample chose all possible gender identities, although the two most popular phrases were that of a man with a feminine side and a person who enjoys wearing the clothing of the opposite sex. Eleven per cent, however, chose the statement that they were women trapped in a man's body, and one person explained that he was born with a chromosomal disorder. The majority of the transsexual sample accepted the traditional view of themselves as having been born women but trapped in male bodies. Unfortunately, however, this term is known to have evaluation which clears the way for surgery (MacKenzie, 1978). The sample members were part of the transsexual subculture, and one of the important functions of that subculture is to orient its members to the appropriate answers which one gives to the psychiatric or psychological evaluator. Table 10 Responses to the Question: "To What Extent Are You Interested In Sex Change?" Interest Transvestites Transsexuals Homosexuals (n = 65) (n = 33) (n = 51) Not at all 54% 3% 96% Looked into the possibility and rejected it 20% - - Am favorably inclined 2% 15% 2% Have taken or am now taking sex hormones 25% 49% 2% Have had surgery - 33% - Note. x2(8) = 126, p < .0001. Though transsexuals are obviously the groups who have dealt with the sex change surgery in the most direct way, consideration of the possibility of sex change or actual sex change was also found in all the groups, as noted in Table 10. Clearly the emphasis is different in the three groups but the blurring of lines is also seen. One "transsexual" indicated he was not at all interested in the complete sex change, although he considered himself a transsexual. Nearly half of the transvestites have either considered sex change or taken sex hormones, and two members of the gay sample had seriously thought of sex change. Note, also, that only 33% of the TS sample members are completely post-operative. This suggests that the transsexual subculture may be most important pre-operatively. Young men who feel that they somehow do not fit into established traditional gender identities apparently seek out the group for support and advice during the decision making process. Only a small minority of the post operative TS's known to the authors actually participate in the TS groups after surgery. Table 11 Detailed List of Occupations Occupations Rating Transvestites Transsexuals Homosexuals Undifferentiated Waiter/waitress 49 1 1 Refinery worker 50 1 Security guard 50 1 Dance instructor 53 1 Cook 55 1 Warehouseman 55 1 Salesperson/clerk 56 6 1 4 7 Medical assistant 57 1 Truck driver 59 1 Mechanic 62 3 1 Foreman-laborer 62 1 Piano repairman 63 1 Beauty operator/hairdresser 63 2 5 1 Machine operator 63 2 Fireman 63 1 Recreation assistant 63 Student 64 1 3 13 10 Electrician 65 1 1 Carpenter 65 1 Small businessman (manager) 67 3 2 Telephone operator 67 1 Housewife 68 1 Real estate agent 69 2 Insurance agent 69 1 Office clerical, data processing, teller 70 1 6 5 1 Dispatcher 71 1 Survey researcher 71 1 1 Retired 71 1 1 1 2 Child care (handicapped) 71 1 Technician 75 4 1 3 Small businessman (owner) 76 1 Music engraver 78 1 Artist 78 4 1 1 Advertising 78 1 1 Writer/author 78 5 1 1 Actor/musician 78 2 2 2 Owner/manager 80 1 1 Contractor 80 1 2 Safety director 80 1 Librarian 81 1 1 Teacher 81 1 1 Stockbroker/investor 81 3 1 3 Accountant 81 3 Administrator 81 5 1 3 Counselor 81 3 1 Health professional 83 1 Banker 85 1 Engineer 86 10 1 1 13 Pilot 86 1 Lawyer 89 1 2 3 College professor 90 1 1 1 Physician 93 1 In sociological terms, one of the most important factors relating to an individual's self-identification as a TV, as distinguished from a TS, was income and occupation. A standard scale of occupational prestige developed by Hodge, Siegel, and Rossi (1964) was used to rank the subjects. This scale utilizes a range between 94 for Supreme Court justices and 34 for workers at the lowest level of occupational prestige (shoe-shining). An average occupation ranks at 71. A detailed list of the jobs held by members of the four groups is shown in Table 11 and the summarization is shown in Table 12. Table 12 Occupation Level of the Four Subsamples Split at the Median of the Total Sample Transvetites Transsexuals Homosexuals Undifferentiated Occupational Level (n = 63) (n = 32) (n = 53) (n = 60) Above median 64% 31% 30% 48% Below median 37% 69% 70% 52% Note x2(3) = 16.03, p < .001. Though the median of the total sample was 75, which places it somewhat above the national aver age, the transvestite sample was heavily represented in the mainstream white collar world, including engineering, accounting, teaching, sales, and clerical work. These men clearly are high prestige workers, and, except for the two hairdressers and the per forming artists, they were not represented in stereotypically feminine jobs. On the other hand, the transsexuals were much more likely to choose the feminine image jobs, with five hairdressers among their ranks. The gay group showed greater occupational variation, although again transvestites as a group had the highest median. Discussion Although the data suggest that there is considerable blurring of categories, certain things, not always anticipated by the TS's in their preparation to pass the psychiatric examination for surgery, appear important. TS's are more likely to be unhappy in their childhood and to seek employment in low status occupations, many of which are usually held by women. One of the differences between the transsexuals and transvestites has been the willingness of the transsexual to assume publicly the role of the opposite sex to which they were born. The transvestites, with Prince as a major exception, have carefully protected their male identities. They presented a feminine persona only on occasions when they cross dressed in seclusion. To do more was to risk exposure, and with this, a significant status loss. The history of an unhappy childhood and poor academic performance found in the transsexual sample has implications in light of the differences both with gays and transvestites, as well as with the sexually undifferentiated samples. It might well be that the existence of a category of transsexual has allowed many of the people with the most troubled childhoods and most marginal adult occupations and happiness to classify themselves as transsexuals rather than homosexuals. This suggests that many of the early case studies on homosexuality are no longer valid, since those so classified might today have been classified as transsexual. The absent father as a variable simply did not hold up in this study. If sports and cross dressing can be thought of as clues as to what is regarded as feminine behavior in young boys, the evidence is contradictory. TV's were most interested in sports, and they also participated most actively in cross dressing. Gays and TS's were not so sports-minded, and although some TS subjects also cross dressed, hardly any of the gays did. From what the individuals can reconstruct of their childhood on a questionnaire, few would have acted in an overly feminine manner. The family constellation of the groups also lacked significance, although the transsexuals again seem to have been the most marginal. The sample is still too small to make any firm generalizations, but it would seem that there might well be an overemphasis on childhood in the psychoanalytic literature, at least as it relates to relationships with parents. It might well be that, ultimately, whether a person classifies himself as a transvestite or as a transsexual relates to his success in the world at large. Transvestites were eminently successful, were heavily into male identified occupations, and, outwardly, seemed secure in their role as males. Perhaps the very security of success in the male world carries a burden which the adoption of a feminine persona allows them to escape. If this speculation is correct, transvestites might well be distinguished from transsexuals in their life goal orientation and their motivation towards material satisfaction and occupational prestige. Success in these lines makes them unwilling to risk more than periodic cross dressing. Failure to be successful in these areas might well encourage them to adopt more radical solutions such as surgery. This, however, is only a speculation. There are obvious differences between the various groups, but there is also a considerable overlap among them. Perhaps, as the definitions become further refined and the self-selection process develops, there will be less overlap.