+ _y E I'lllfllfllfllll”ll‘fllfi‘l 9-‘IFQ Ff..4i ;-__'.-4_.l...’3 e. ..~..".;.v This is to certify that the dissertation entitled PATTERNS IN MASCULINE GENDER ROLE-IDENTIFICATION, BODY SATISFACTION AND SELF-IMAGE IN HOMOSEXUAL AND HETEROSEXUAL MALES presented by WILLIAM REID LEDFORD has been accepted towards fulfillment of the requirements for Ph . D . degree ingsychology Major professor Bertram P. Karon Date 7/ 10/85 MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771 ‘ MSU LIBRARIES “ RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if‘book is returned after the date stamped below. :33 W-Ww‘was. :W L.‘ U. 9,1. 112 A133 “1?: O4 ‘87 a APR} 0 :9 WIN 1 22009 llflllllllI“lllllllllllllllllllllllllllllllilllmllllllllll ~- ~ a E 31293 10643 6664 If8_,-;O_ “‘5 _/. I 3’...» ——.n---—u¢-).w-j grin-u“ This is to certify that the dissertation entitled PATTERNS IN MASCULINE GENDER ROLE-IDENTIFICATION, BODY SATISFACTION AND SELF-IMAGE IN HOMOSEXUAL AND HETEROSEXUAL MALES presented by WILLIAM REID LEDFORD has been accepted towards fulfillment of the requirements for Ph. D . degree inpsychology / {,1 r. . r M Major professor Bertram P. Karon Date 7/10/85 MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771 ‘ MSU LIBRARIES RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if‘book is returned after the date stamped below. 57 KIF7511§09 ' APR ‘30 .999 {IAN 1 2 2009 PATTERNS IN MASCULINE GENDER ROLE-IDENTIFICATION, BODY SATISFACTION AND SELF-IMAGE IN HOMOSEXUAL AND HETEROSEXUAL MALES BY William Reid Ledford A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1985 ABSTRACT PATTERNS IN MASCULINE GENDER-ROLE IDENTIFICATION, BODY SATISFACTION AND SELF-IMAGE IN HOMOSEXUAL AND HETEROSEXUAL MALES BY William Reid Ledford It is frequently implied in clinical psychoanalytic literature that homosexuality in males is motivated by the ego's need to secure masculine components of the ego-ideal. Aspects of this theory were evaluated for a sample of 53 gay and heterosexual men ranging in age from 19 to 45 years. It was found that for both gay and heterosexual nonpa- tient males, the semantic differential concepts MYSELF AS I MOST WOULD LIKE TO BE and THE PERFECT LOVER were virtually identical in certain respects for almost every subject. The differences between the two concepts for heterosexual sub- jects seems to be accounted for only by the differences in gender-role traits between heterosexual men and their lovers. Regardless of sexual orientation, it seems that people de- sire lovers who possess traits which correspond to their ideal selves. It was also found that for gay men, the stronger the identification with the father, the more an individual's actual self-image approximated his ideal self; a finding not obtained for heterosexual subjects. It was also found that for gay men, one's perceived masculinity was significantly and positively related to the degree of similarity between William Reid Ledford perceived actual and ideal self. Interestingly, body satis- faction was found to be related to certain qualities pos- sessed by fathers of subjects in both groups, qualities most likely related to kindness and nurturance. There was substantial evidence that any differences in characteristics of parents that may exist between the two groups involve fathers rather than mothers. Homosexual and heterosexual men seem to conceptualize their mothers in very similar terms. This finding contradicts traditional theor- ies which attribute maladaptive qualities to mothers of homosexual men and imply that such characteristics play a causal role in the development of homosexual orientation in men. Contrary to assertions frequently made in clinical literature, it was concluded that there is no evidence that homosexual men have a greater discrepancy between actual and ideal self than do heterosexual men, or that such a discrepancy plays any role in the development of one's sexu- al orientation. To the fifty-three incredibly dedicated subjects who made this possible--0ne million thanks. ii ACKNOWLEDGMENTS With affection and gratitude I acknowledge the contri— butions made by the following people in the preparation of this dissertation: My chairperson, Dr. Bertram Karon, dedicated a great deal of time and energy to reading numerous preliminary drafts, willingly helping me to clarify and organize the material presented--frequently until the wee hours. My committee members, Drs. Robert Caldwell, Elaine Donelson and Terry Stein, along with Dr. Karon, offered ex— tremely helpful advice and criticism in regard to measures, reference materials, statistical procedures, theoretical issues, and final textual presentation. Their involvement in my education over the course of what hardly seems to be so many years has led to very much personal and professional growth on my part. I wish earnestly to thank them for their patience, their time and consideration. In addition, per- haps unknowingly, they have given me tools,too intangible to describe, which will help me continue my development as a psychologist, perfecting skills according to principles they have taught me. These four people are fondly remembered. I also wish to thank those who helped me with many difficult details in connection with research procedure. I cannot imagine having completed this project without the help of William Bathie, Larry Little, Eddie Medina, and Dianne Roycraft. I am also very grateful to Dr. A.I. Rabin for being of great help to me in the earliest stages of this study. Many additional thanks to my parents, Margaret and Robert Ledford,for their unremitting interest and sup- port during my education process. Finally, many warm, affectionate thanks to Deborah Berry, a source of support to me for many years, and a still very constant friend and colleague. She will know to what I refer in saying, "This is the little white card--I was glad that we were in line together." iii TABLE OF CONTENTS LIST OF TABLES Vii LIST OF FIGURES ix CHAPTER l--INTRODUCTION 1 Historical Review of Psychodynamic Theories of Male Homosexuality 4 Overview of Major Research on Developmental Origins of Homosexuality 14 The Completion Hypothesis 24 Empirical Literature on Self-Concept and Sexual Preference 34 Summary 45 Statement of Hypotheses 49 CHAPTER 2--METHOD 56 Subjects 56 Procedure 63 Instruments 63 Measures 64 CHAPTER 3--RESULTS 77 Results of Hypothesis Tests 77 Additional Findings for Gay Subjects 93 Relationships Among Semantic Differential Concepts 93 Correlations Among "Self" and "Parent" Semantic Differential Concepts and the Psychometric Tests 100 iv Intercorrelations Among the Psychometric Instruments 101 Additional Findings for Heterosexual Subjects 102 Relationships Among Semantic Differential Concepts 102 Correlations Among "Self" and "Parent" Semantic Differential Concepts and the Psychometric Tests 110 Intercorrelations Among the Psychometric Instruments 112 Additional Similarities and Differences Between Groups 113 The Semantic Differential Concepts 113 Body Satisfaction and Personal Sense of Power 113 Emphasis on Physical Attractiveness of Sexual Object and Gender Role Identifi- cation 115 Gender-Role Identification and Actual- Ideal Self Discrepancy 117 CHAPTER 4--DISCUSSION 120 Discussion of Hypothesis Tests 120 Discussion of Additional Findings for Gay Subjects 131 ' Relationships Among Semantic Differential Concepts 131 Correlations Among "Self" and "Parent" Semantic Differential Concepts and the Psychometric Tests 136 Discussion of Additional Findings for Heterosexual Subjects 138 Relationships Among Semantic Differential Concepts 138 Correlations Among "Self" and ”Parent" Semantic Differential Concepts and the Psychometric Tests 141 V Discussion of Additional Similarities and Differences Between Groups The Semantic Differential Concepts Body Satisfaction and Personal Sense of Power Emphasis on Physical Attractiveness of Sexual Object and Gender-Role Identifica- tion Gender-Role Identification and Actual- Ideal Self Discrepancy CHAPTER 5--CONCLUSION LIST OF REFERENCES APPENDIX vi 143 143 144 145 146 148 155 160 Table 1. Table 2. Table 3. Table 4. Table 5. Table 6. Table 7. Table 8. Table 9. Table 10. Table 11. LIST OF TABLES Frequency Counts for Characteristics of Gay Men in Sample Frequency Counts for Characteristics of Heterosexual Men in Sample Descriptive Statistics for Questionnaire Items Assessing Physical Emphasis for Gay Subjects Descriptive Statistics for Questionnaire Items Assessing Physical Emphasis for Heterosexual Subjects Tests for Between-Group Differences in Means for Physical Emphasis Measures Correlation Coefficients Among Body Satisfaction and Physical Emphasis Variables for Sample of Gay Males Correlation Coefficients Among Body Satisfaction and Physical Emphasis Variables for Sample of Heterosexual Males Correlation Coefficients Among Body Satis- faction and Gender Self-Perception/ Role Identification Variables for Gay Subjects Correlation Coefficients Among Body Satis- faction and Gender Self-Perception/ Role Identification Variables for Heterosexual Subjects Relationships Among Concepts for the Semantic Differential by Dimension for Gay Subjects Intercorrelations Among Self-Concept, Body Satisfaction and Gender-Role Scales for Gay Subjects vii 59 61 80 81 82 84 85 90 92 96 103 Table Table Table Table 12. 13. 14. 15. Relationships Among Concepts for the Semantic Differential by Dimension for Gay Subjects Intercorrelations Among Self-Concept, Body Satisfaction and Gender-Role Scales for Heterosexual Subjects Correlations Among Body Satisfaction Measures and Measures of Personal Sense of Powerfulness Correlations Between Personal Attributes Questionnaire and Actual Self--Ideal Self Discrepancy viii 104 114 116 119 Figure 1. Figure 2. LIST OF FIGURES Relationships Among Semantic Differential Concepts in Two Dimensions (Evaluation and Potency) for Gay Subjects 98 Relationships Among Semantic Differential Concepts in Two Dimensions (Evaluation and Potency) for Heterosexual Subjects 107 ix CHAPTER 1 INTRODUCTION There is probably more nonsense written about homosexuality, more unwarranted fear of it, and less understanding of it than any other area of human sexuality. (Pomeroy, 1969, p. 1) There are at least two prominent factors which, in P0- meroy's view, account for the heap of nonsensical litera- ture devoted to the subject of homosexuality. The first of these is the acceptance of the phenomenon as a single, uni- tary entity. He points out that whereas some homosexuals develop long-term emotional relationships with another per- son of the same sex and live "monogamously" for as much as the remainder of their lives, others devote themselves to pluralistic (i.e., nonmonogamous) styles of sexual expres- sion. He also observes, in relation to male homosexuals, that some engage in homosexual encounters for money, others may begin homosexual activity when confined to an exclusive- ly male environment. Pomeroy concludes that ...it is obvious that to lump all homosexuals together is as grossly misleading as to lump all heterosexuals together. Homosexuality is no re- specter of age, religion, or social level. It occurs as frequently among physicians, psychia- trists, clergymen, judges and politicians as among truck drivers and ditch diggers. (p. 10) 2 The second factor is a persistent disinclination to ap- proach homosexuality as a sexual variation independent of stereotypes. More specifically, regarding the controversy as to whether homosexuality should be considered a disorder, Pomeroy writes, If my concept of homosexuality were developed from my practice, I would probably concur in thinking of it as an illness. I have seen no homosexual man or woman in that practice who was not troubled, emo- tionally upset, or neurotic. On the other hand, if my concept of marriage in the United States were based on my practice, I would have to conclude that marriages are all fraught with strife and conflict, and that heterosexuality is an illness. In my twen- ty years of research in the field of sex, I have seen many homosexuals who were happy, who were par- ticipating and conscientious members of their com- munity, and who were stable, productive, warm, re- laxed, and efficient. Except for the fact that they were homosexual, they would be considered normal by any definition. (p. 10) In the Final Report and Background Papers of the NIMH Task Force on Homosexuality (Livingood, 1972), top priority is assigned to refinement of sampling methods to comprise the entire range of homosexual phenomena. It is urged, fur- thermore, that investigations be conducted in a way that in- cludes homosexual individuals who "do not come into contact with medical, legal, or other social control or treatment resources and who therefore have been least studied" (p. 3). In spite of a generally more liberal approach to the issue of homosexuality , however, strong emphasis is still placed by the Task Force on prevention and treatment. On the other hand, in the third and most recent edition of the 3 Diagnostic and Statistical Manual of Mental Disorders (DSM- III), no general classification exists for homosexuality, but rather for "ego dystonic homosexuality" (American Psy- chiatric Association, 1980, p. 281). This is a change from earliest printings of DSM-II which has elicited opposition from psychiatrists who regard homosexuality in itself as a clinical entity (Bayer, 1981). DSM-II describes this change and the rationale for its implementation as follows: In December 1973, the Board of Trustees of the American Psychiatric Association voted to eliminate homosexuality per se as a mental disorder and to substitute a new category, Sexual Orientation Dis- turbance, reserved for those homosexuals who are "disturbed by, in conflict with, or wish to change their sexual orientation." This change appeared in the seventh and subsequent printings of DSM-II. (p. 380) The removal of homosexuality per se from DSM-II was supported by the following rationale: The crucial issue in determining whether or not homosexuality per se should be regarded as a mental disorder is not the etiology of the condition, but its consequences and the definition of mental disorder. A significant proportion of homosexuals are apparently satisfied with their sexual orientation, show no significant signs of manifest psychopathology (unless homosexual- ity, by itself, is considered psychopathology), and are able to function socially and occupationally with unimpairment. If one uses the criteria of distress or disability, homosexuality per se is not a mental disorder. If one uses the criterion of inherent dis- advantage, it is not at all clear that homosexuality is a disadvantage in all cultures or subcultures. In DSM-III, the category of Ego-dystonic Homo- sexuality is a modification of the DSM—II category of Sexual Orientation Disturbance. The change in ter- minology was made to make it clear that the category is limited to individuals with a homosexual arousal pattern. (p. 380) 4 Historical Review of Psychodynamic Theories of Male Homo- sexuality Among earlyiflmnrists who devoted formal study to the subject of homosexuality, it was widely believed that the basis for such same-sex preferences was biological. Richard von Krafft-Ebing, for example, became distinguished for his work in the field of human sexuality in the late nineteenth century. Krafft-Ebing originally believed homosexuality, or "inversion," as he called it, to be a functional sign of neuropathic and psychopathic degeneration which in most cases was the result of unnamed hereditary factors. He ultimately took the position, however, that homosexuality represented not so much a state of degeneration, but was more likely a simple variation or anomaly. In commenting upon Krafft- Ebing's contribution to our understanding of homosexuality, another early student of human sexuality, Havelock Ellis, writes: At the time of his death, Krafft-Ebing, who had begun by accepting the view, at that time prevalent among alienists, that homosexuality is a sign of degeneration, thus fully adopted and set the seal of his authority on the view, already expressed by some scientific investi- gators as well as by inverts themselves, that sexual inversion is to be regarded simply as an anomaly.... The way was even opened for such a view as that of Freud and most of the psy- choanalysts today who regard a strain of homo— sexuality as normal and almost constant, with a profound significance for the psychonervous life. (Ellis, 1942, pp. 70,71) 5 Interestingly, although Freud saw homosexuality as an "arrest of psychosexual development, he regarded it as nei- ther illness nor vice and suggested a dismal outlook for those who might undertake to "reverse" it. In 1935, Freud wrote a now well-known letter to an American woman whose son was homosexual: Homosexuality is assuredly no advantage, but it is nothing to be ashamed of, no vice, no degra- dation, it cannot be classified as an illness; we consider it to be a variation of the sexual func- tion produced by a certain arrest of sexual devel- opment. Many highly respectable individuals of ancient and modern times have been homosexuals, several of the greatest men among them (Plato, Michelangelo, Leonardo da Vinci, etc.). It is a great injustice to persecute homosexuality as a crime, and cruelty too... By asking me if I can help, you mean, I sup- pose, if I can abolish homosexuality and make normal heterosexuality take its place. The answer is, in a general way, we cannot promise to achieve it. In a certain number of cases we succeed in developing the blighted germs of heterosexual ten- dencies which are present in every homosexual, in the majority of cases it is no more possible... What analysis can do for your son runs in a different line. If he is unhappy, neurotic, torn by conflicts, inhibited in his social life, ana- lysis may bring him harmony, peace of mind, full efficiency, whether he remains a homosexual or gets changed. (Freud, 1935/1963) Basic to the psychoanalytic approach to the origins of homosexuality is what is known as the "biological bisexual- ity of man," a concept which holds the ability of everyone initially to develop sexual feelings without regard to the object's gender to be a phylogenetic given. Seeming to sup- port this hypothesis is the apparently indiscriminant nature 6 of infantile sexuality. Similarly, homosexual behavior has been described by Blos as "part and parcel" of teenage life (Blos, cited in Fraiberg, 1961, p. 78). In addition, ob- servations of frequent homosexual experimentation in adoles- cence, at least among males, are well documented. Just as these phenomena may be considered behaviorally manifested vestiges of the original homosexual component of basically bisexual man or woman, what is often referred to as "situa- tional homosexuality" is seen as a consequence of an innate flexibility of object choice. This type of homosexuality is the homosexual activity of men or women in situations where sexual partners of the other sex are unavailable, such as in prisons or at sea. At the heart of the psychoanalytic ap- proach to homosexuality, then, is the question of what events take place in later development to cause an individual's sexual preference to be limited to object of his or her own sex. According to Fenichel (1945), and true to the psycho- analytic tradition, a readiness to develop the homosexual orientation is in part determined by constitutional factors among which the hormonal components are crucial (p. 330). Under the pressure of certain conflicts, then, which are dis- cussed below, these constitutionally predisposing factors may facilitate the development of a homosexual orientation. At this point, it is necessary to turn our attention specifically to male homosexuality given that the divergen- cies between females and males in development necessitate a separate discussion for each group, and that the focus of 7 this investigation is homosexuality in males. What is not applicable to this group is not here considered. The classical psychoanalytic position holds the rejec- tion of the heterosexual object in male homosexuality to be distinctly genital. The homosexual man may engage in social relationships with women, come to admire them and develop se— cure platonic relationships, but is repulsedcnrfrightened by the idea of genital contact with them due to the fact that the homosexual man is dominated by a strong castration com- plex. For such an individual, the idea of being without a penis is so terrifying that he would not consider engaging in sexual intercourse with a partner who did not have one (Fenichel, 1945). According to this perspective, the sight of feminine genitals may arouse anxiety in the male child in two ways: first, once recognizing that there does exist a class of hu- man beings who have no penis, the boy comes to fear that he might lose his as well. Second, certain oral fears may be aroused in which the female genitals are seen as a device for castration, viz., a "vagina dentata." The first sight of the female genitals and the sudden anxiety which it may arouse in a boy is referred to as "castration shock," and may be found in the histories of both homosexual and heterosexual males. The decisive factor in terms of object choice is the type of reaction to the shock which an individual undergoes. Homo- sexual men are thought to have reacted by refusing hetero- sexual contact from that point onward (Freud 1909/1963; Fenichel, 1945). According to Fenichel (1945), the homosexual man typi- cally exhibits an oedipal attachment to his mother which Fe- nichel describes as an intense "mother fixation." Central to the phenomenon of homosexuality in males, for this per- spective, is the fact that all object loss or disappointment entails a tendency to regress from the level of object love to the level of identification with the object. Taking the form of castration shock, this disappointment in the mother's genitals precipitates just such a regression. What decides whether the boy becomes homosexual is how and in what respect the regressive identification takes place. It takes place in the boy who later becomes homosexual when he identifies with his mother, whom he cannot possess, and like her, he loves men. Psychodynamic theories advanced subsequent to classical psychoanalysis have characteristically de-emphasized the role of biology in personality development in favor of greater attentipn to the impact of social forces on the psychological histories and ongoing mental lives of individuals. Typically less complex than more orthodox psychoanalytic formulations, the corresponding conceptualizations of homosexuality natu- rally have assumed the flavors of the theories which spawned them. In general, these later psychodynamic theories will not be elaborated here except as they have some bearing on the hypotheses examined in the present investigation. The theories of Sullivan (1953) and Kohut (1971) are among those 9 which bear such a relevancy. Sullivan (1953) places particular emphasis on the ne- cessity for intense and intimate (not necessarily sexual) preadolescent relationships between boys as a prerequisite to heterosexual development. In this connection he describes a group of boys who had attended high school in a small Kan- sas town. As adults, the two men who had not participated in the group's homosexual experimentation were discovered by Sullivan to be overt homosexuals. He adds that "those who had participated in mutual sexuality were married, with children, divorces and what not, in the best tradition of American society" (p. 256). Although he does not say that preadolescent homosexual experimentation is necessary to la— ter heterosexual development, he does insist that the need for intimacy manifested toward members of one's own sex nor- mally predates the maturation of the "lust dynamism." This maturation, according to Sullivan, ideally accompanies a shift in the intimacy need to the other sex. According to this point of view, homosexual behavior, whether transient or enduring, is largely given rise to by "accidents" in which such a shift fails to occur at or near the time of puberty. In other cases, according to Sullivan, the homosexual beha- vior of an individual may be motivated by a pathological need to separate interpersonal relations based on lust from those based on a need for intimacy. In cases in which this need to separate lust from intimacy leads to the dissociation of lust, any of a variety of anomalies in personality may be 10 engendered, depending upon the aspect of lust dissociated. Sullivan considers male homosexuality in these cases to de- velop in response to the "uncanny feeling" associated with the female genitals and anticipation of the "physical inter- genital situation" (p. 275). Kohut (1971) writes, in reference to homosexuality and the other psychoanalytically so-called "perversions," "It is...my impression that specific circumscribed disturban- ces in the narcissistic realm are usually the nucleus of these widespread disorders." According to Kohut's psycho- analysis, the basis of the narcissistic personality disor— ders is a disturbance in the "narcissistic configurations" which evolves in response to normal disruptions to the equi- librium of primary narcissism. This comfortable state of oneness with omnipotent objects is, of course, necessarily disturbed by the shortcomings of maternal caregiving. The child, then, "replaces the previous perfection by (a) estab- lishing a grandiose and exhibitionistic image of the self: the grandiose self; and (b) by giving over the previous per— fection to an admired, omnipotent...self—object: the ideal- ized parent imago (p. 25). The term self—object here refers to those object representations which are not experienced as separate and independent from the self. Normally, the gran- diosity and exhibitionism of the former are gradually modi- fied and along with the latter are integrated into the adult personality. These two major "configurations," then, are the precursors, respectively, to normal mature forms of 11 (a) positive self-esteem and self-confidence and (b) the ability for enthusiasm and admiration for others. Moreover, the grandiose self, once integrated into the adult personal- ity, provides the motive force for ego-syntonic goals and ambitions while the idealized internal representation of the parent (idealized parent imago) is introjected as the idealized superego. An individual with a narcissistnzpersonality disorder, however, has remained fixated on archaic grandiose self-con- figurations and/or archaic idealized "narcissistically ca- thected" objects (self-objects) and has not had the benefit of integration of these with the rest of his personality. Because they remain unaltered in their unintegrated form, they threaten the mature self with intrusion of archaic nar- cissistic aims; and the ego remains deprived of the corre- sponding narcissistic investments. Particularly in the realm of object relations, which concerns us in the present study, the regression from normalcy and the corresponding continuance of the narcissistic strivings in the narcissis- tic personality disorders involves a "compelling need for merger with [the] powerful object" (p. 9). Kohut contrasts the process obtaining under favorable circumstances with its failure as follows: Under optimal circumstances the child experiences gradual disappointment in the idealized object-- or, expressed differently: the child's evaluation of the idealized object becomes increasingly real- istic--which leads to a withdrawal of the narcis- sistic cathexes from the imago of the idealized 12 self object to their gradua1...internalization, i.e., to the acquisition of permanent psychologi— cal structures which continue, endopsychically, the functions which the idealized self-object had previously fulfilled. If the child suffers trauma- tic...disappointment in it, then optimal inter- nalization does not take place. The child does not acquire the needed internal structure, his psyche remains fixated on an archaic self-object in what seems to be an intense form of object hun- ger. (p. 45) Since all bliss and power now reside in the ideal- ized object, the child feels empty and powerless when he is separated from it and he attempts, therefore, to maintain continuous union with it. (p. 37) In addition, these archaic, regressive psychic struc- tures (e.g., grandiose self, idealized parent imago) may be- come sexualized in narcissistic personality disorders; one manifestation of which, in Kohut's opinion, may be homosex- uality. To illustrate this principle, Kohut describes the case of a man, Mr. A., who although not overtly homosexual, reported homosexual attractions of such strength as to lead him to seek analysis. In Kohut's opinion this patient's ho- mosexual preoccupations were subordinate to his overall per- sonality configuration which originated largely from a trau- matic disappointment in the idealized father imago in early latency. Consequently, according to Kohut, he was able to obtain a sense of heightened self-esteem only by attaching himself to strong and admired male figures. This tendency was nonsexual in nature, however, insofar as the sexualization of the narcissistic configurations was only a part of the total narcissistic personality organization. 13 In regard to more specifically sexual acts, Mr. A. never engaged in homosexual activities and—-apart from some sexually tinged, playful wrestling in adolescence and the buying of "physical culture" magazines which contained photographs of athletic men--his homosexual preoccupations were consummated only in fantasy, with or without masturbation. The objects of his homosexual fantasies were always men of great bodily strength and perfect physique.... Occasionally he achieved orgasm and a feeling of triumph at the thought of masturbating a strong and physically perfect man and draining him of his pow- er. (PP- 69, 70) Kohut regards this sexualization of the narcissistic confi- gurations as having come about via failure in the ego's drive neutralizing capacity prior to the traumatic loss of the patient's idealized parent imago. Kohut considers his fantasies of pursuing physically powerful men and the orgas- tic experience of draining power from "fantasied imagoes of external perfection" as a means to the vicarious acquisi— tion of the strength and perfection which characterized them. Existing psychodynamic formulations, then, generally regard homosexuality in males as a function of developmen- tal arrest or regression and/or avoidance of heterosexuali- ty. These factors are often posited as interactive with unspecified constitutional predisposers. In particular, psychodynamic theories of male homosexuality, the theories of Sullivan and Kohut among them, often suggest that male homosexuality represents an unconscious striving to 14 complete an identification with a male which normally is accomplished during childhood. Such strivings, according to these theories, which in childhood are pregenital for heterosexual males, in homosexual males continue post-puber- tally and are attached to adult masculine objects. Conse- quently, they assume a manifestly adult sexual character. The mechanisms by which these strivings are satisfied are often referred to in terms of incorporation or introjec- tion of qualities of masculinity which have become highly admired, or which exist in a highly admired object-~usually the father. Although Sullivan's emphasis is upon inter- personal processes rather than libidinal strivings mediated by the processes of incorporation and identification, he suggests that the absence of preadolescent homosexual ex- ploration and intimacy in males is conducive to adult homo— sexuality. At this point, we turn to a brief survey of some major empirical findings in connection with the psychodyna- mics of sexual object choice. Overview of Major Research on Developmental Origins of Homosexuality In 1952 Bieber and his collaborators (Bieber et al., 1962) undertook an intensive study of male homosexuality em- ploying 106 male homosexual and 100 male heterosexual sub- jects as controls, all of whom were involved in psychoanaly- tic treatment with members of the Society of Medical Psycho- analysts. This investigation is of importance since it 15 represents the first successful attempt at (a) compiling such detailed data from individual psychoanalyses for such a large sample of homosexual men, and (b) subjecting such data to statistical and clinical analysis. In general, Bieber is critical of Freud's emphasis on biological determinants in the etiology of male homosexual- ity, and considers the emphasis to be more suitably placed on family dynamic patterns. What Bieber refers to as the "classical" situation is one in which the mother is close- binding and intimate, dominant, and takes a deprecatory at- titude toward her husband. The father is described as de- tached and often hostile toward the son in question. Bie- ber and his collaborators conclude that from their statis- tical analysis, the chances seem high that any son exposed to such a parental combination will either become homosexual or develop homosexual conflicts. With regard to what constitutes a close-binding mother, the Bieber group observed that such mothers were sexually overstimulating toward their sons by means of excessive in- timacy or outright seductiveness. Secondly, such mothers sexually inhibited their sons. Although they were sexually overstimulating, they nevertheless suppressed overt manifes- tations of heterosexual responsiveness on the part of the child. Bieber also observes that such suppression seems to have served as a defensive means of concealing from them- selves as well as from others their own sexual feelings to- ward their sons. Most such mothers also held antisexual 16 attitudes which were reflected in a tendency to portray sex- uality as unacceptably distasteful and brutish. Close-binding mothers also had the general tendency to discourage masculine attitudes and behavior patterns on their sons' part and interfered with their peer group parti- cipation, minimizing opportunities for masculine identifica- tion with other boys. These mothers also typically interfered with the fa- ther-son relationship in a number of ways. First, they en- couraged the child's wish for exclusive maternal possession by openly expressing a preference for the son over the fa— ther. Similarly, they fostered father-son competitiveness by finding ways to pit each against the other for maternal favor. The mothers behaved romantically toward the sons in ways that seemed to compensate for deficiencies in the mari— tal relationship, and permitted or encouraged the sons' par- ticipation in situations in which their involvement was not appropriate. Such sons, for example, were sometimes in- volved in parents' arguments or were allowed to sleep with their parents in the same bed. By selecting a particular child for preferential treatment, usually the son who later became homosexual, they fostered competitive sibling relationships. In addi- tion, they interfered with the development of independence by preempting the decision-making process for the child. They discouraged self-assertiveness and typically infantil- ized their sons by their oversolicitous treatment. 17 According to Bieber, the most striking aspect of the father-son relationships in both homosexual and heterosexual groups was the consistency with which psychopathological phenomena tended to appear. He adds that "profound inter- personal disturbance is unremitting in the father-son rela- tionships [of the homosexual subjectsl" (p. 114). The fa- thers of controls, in addition, presented a far more whole- some picture than fathers of the homosexual subjects. The vast majority of fathers of homosexual men were classified as "detached," of which most were distant and indifferent, hostile, or dominating-exploitative. Bieber indicates that homosexual development in the sons of such fathers can be largely traced to the fact that paternal detachment, as a traumatic circumstance, is compen- sated for by reparative relationships with other males. The seeking of need fulfillment from other men, according to Bieber, has a clear point of origin in fathers who were de- tached. The Bieber study suggests that sons of such fa- thers sought in homosexual partners the qualities which were absent in their own fathers, such as warmth, friendli— ness, closeness, and the reassurance of physical presence. Bieber suggests that because the detached fathers spent little time with their sons,they contributed to the devel- opment of homosexuality in that they failed to provide the sons with adequate male models for identification. Although subsequent investigations indicate that this family constellation occurs with marked frequency in the 18 histories of homosexual men (White & Watt, 1973), this is by no means always the case. There has been ample criticism of Bieber's position that homosexuality per se is a patholo- gical entity even among those who do not dispute his find- ings. Before turning to a discussion of these issues, it might be noted that these findings seem to bear some consis- tency with the psychodynamic hypotheses previously described which suggest that motivations underlying homosexuality in males consist, at least in part, of adult sexual versions what occurs in all males sooner or later--strivings toward appropriation of idealized masculine qualities via physical and/or emotional closeness. Bieber's statement that sons of the detached fathers sought in their homosexual partners qualities which were absent in their own fathers is not at all divergent from Kohut's position. This point is made much more strikingly, however, in later treatises by Kaplan (1963) and Tripp (1975), discussed below. Finally, in this connection it is apparent that Bie- ber's findings regarding inhibited peer-group participation among homosexual males as children are in harmony with Sullivan's observations. It might be noted, however, that rather than engendering homosexuality, disruptions to such peer-group experiences might as well have been brought about by variables sometimes associated with homosexual de- velopment such as variant gender-role preferences and lack of traditionally "masculine" interests, or merely the sense of being different (Bieber et al., 1962; Saghir & Robins, 19 1973). Precursors to homosexuality, then, would have exist- ed prior to the observed among-peer interactions and could not be said to have arisen from them. Whether the charac- teristic nature of the peer-group interaction bolsters con- stitutionally determined homosexual proclivities is subject to question. Gonsiorek (1982a) observes that Bieber's study is fraught with sampling problems characteristic of those stu- dies involving patients in treatment for psychological prob- lems. Aside from this issue, which is discussed at length below, Gonsiorek describes the Bieber study as noteworthy in regard toresearcher bias. He observes that the same psy- choanalysts with whom the subjects were in treatment were those who developed the theory of homosexuality propounded by Bieber et a1. These investigators, he notes, addition- ally developed the questionnaire used to test their theory, served as raters in the study, and interpreted the results concluding that their theory had been verified. These facts make it unclear, according to Gonsiorek, as to whe- ther the findings were a function of built-in researcher biases, adding that "it would be difficult to imagine how to build more potential for research bias into experimental procedures than the Bieber group did" (p. 69). Hooker (1972), in reference to various studies appar- ently supporting familial pathology theories of homosexual- ity remarks: 20 The evidence from these and many similar studies does not support the assumption that pathological parent-child relations are either necessary or sufficient antecedents or determinants of adult homosexuality. The evidence does indicate, however, that some forms of familial pathology appear to be associated with increased vulnerability of some in- dividuals to homosexual development, and it sug- gests that psychopathology is more frequently asso- ciated with homosexuality in these individuals. (p. 13) Hooker has been a leading figure in pointing out the unfa- vorable tendency of researchers to treat homosexuality as a unitary and clinical entity. According to her, the lines of investigation pursued by researchers are largely directed by the prevailing climate of professional opinion. Research on the development of homosexuality, therefore, has general- ly been conducted with its focus on clinical rather than so- cial and cultural phenomena. Furthermore, Hooker points out that phenomena judged as psychopathological which appear es- pecially characteristic of homosexual groups are often at- tributed to the variable "homosexuality." Instead, these may typically represent "ego-defenses" against victimiza- tion, which are characteristic not only of homosexuals, but of other oppressed minority groups as well (Hooker, 1965). In this connection, however, Schur (1972) has comment- ed: Notwithstanding evidence from the Hooker studies indicating that there may be--even under present circumstances in the United States--some con- firmed homosexuals who appear reasonably "well adjusted" psychologically, it is hard to see how 21 any homosexual in our society can completely avoid feeling the psychological impact of strong social disapproval and legal condemnation and proscription. (p. 37) A further consequence of concealment is the fre- quent need to maintain silence in the face of ex- pressions of contempt for homosexuals. In all this, the homosexual cannot remain unaffected by the pressures.... (p. 37) Hooker is particularly critical of conclusions drawn by Bieber that homosexuality is a specific form of psycho- sexual disorder. She argues that none of the evidence used to support such an assumption was specific to his homosexual group. Furthermore, she observes that in a number of stu- dies conducted outside of treatment or correctional set- tings, "results obtained by the use of the MMPI, TAT, Ror- schach, and other psychological measures did not justify the conclusion that homosexuality is necessarily and invariably a concomitant or symptom of psychopathology. In many indi- viduals no evidence of psychopathology was found" (1972, p. 15). Hooker writes: It comes as no surprise that some homosexuals are severely disturbed...but what is difficult to accept (for most clinicians) is that some homo- sexuals may be very ordinary individuals, indis- tinguishable, except in sexual pattern from indi- viduals who are heterosexual. (1963, p. 159) Gonsiorek (1977, 1982b) has reviewed the literature on homosexuality and psychological adjustment, and emphasizes that a consistent and clear pattern emerging from studies 22 on homosexuality and psychological testing is that homosexu- ality in and of itself is unrelated to psychological distur- bance. He points out that differences which are obtained between homosexual and heterosexual and heterosexual groups lie within normal ranges and that attempts to differentiate homosexuals from heterosexuals on the basis of psychological testing have been generally unsuccessful. Gonsiorek, then, cautions that although significant differences between groups may be of theoretical interest, these are not indica- tive of greater disturbance of one group over another unless the former has scores falling in a range which has been val- idated as psychopathological. Secondly, he advises that findings of difference between groups in regard to family constellation are not valid as a basis for inference about difference in psychological adjustment of individuals from such families. Hooker tentatively concludes that homosexuality does not exist as a clinical entity, its forms being as varied as are those of heterosexuality; and that homosexuality may be a deviation in sexual pattern which psychologically lies within the normal range. Saghir and Robins (1971), in a much cited study, un- dertook a detailed investigation of groups of 89 male homo- sexuals, 57 lesbians, and corresponding groups of hetero- sexual controls, from a developmental point of view. Their procedure involved semistructured interviews which yielded the following findings: 23 Homosexual males and females for the most part show during childhood preferences in terms of roles and identifi- cations which are most typical of the opposite sex. Their findings indicated that the childhood and adolescence of most homosexual men are characterized by a lack of contact with male companions and by a preference for female play- mates as opposed to what was found to be typical for male heterosexuals. In addition, the homosexual males, as boys, generally did not participate in team sports and rough play. The majority of the lesbians reported being tomboys during childhood. They typically bad boys as playmates and en- joyed sports rather than dolls and domestic activity. In a recent effort to explore the origins of homosexu- ality, Bell, Weinberg and Hammersmith (1981) compiled inter- view data for 979 homosexual and 477 heterosexual men and women. The homosexual men were found generally to have been less stereotypically masculine as boys than their he- terosexual counterparts. More homosexual than heterosexual men recalled some dislike for typical boys' activities and enjoyment of those which they considered to be for girls. On the basis of their path analysis, the researchers conclude that such gender nonconformity is directly related to adult homosexual preference. They also report that the homosexual men identified less with their fathers than did heterosexual men, and less with their fathers than with their mothers. Nevertheless, they add that "our causal analysis convinces us that the tendency for homosexual 24 males to perceive their fathers in a relatively negative fa- shion has little eventual influence on their sexual orienta- tion" (pp. 61, 62). They point out that influences are not necessarily unidirectional from father to son, but recipro- cal in which "the prehomosexual boy may simply be recipro- cating his father's disinclination to identify with him" (p. 60). It was also found that as children the homosexual men did not differ significantly from heterosexual men in the degree to which they felt similar to and wanted to be like their mothers. The authors conclude that as children, identification with the opposite-sex parent appears to have had no signi- ficant impact on whether male as well as female respondents became homosexual or heterosexual. Similarly, they argue that identification with same-sex parents seems to have ex- erted no decisive influence on the development of adult sexual orientation. Nevertheless, they report that there is a powerful link between gender nonconformity and the de- velopment of homosexuality and that "the homosexual men's generally negative relationships with their fathers...dis- played a very modest but direct connection to their gender nonconformity..." (p. 190). The authors conclude that on the basis of their overall findings, no single phenomenon of family relationships can be singled out as especially consequential in the development of adult sexual prefer- ence. The ways in which much of this earlier literature may be brought specifically to bear on the present 25 investigation is the issue to which we now turn. The Completion Hypothesis According to Tripp (1975), "Homosexuality in all its variations always means that same-sex attributes have become eroticized, that is, have taken on erotic significance. No matter how or when this takes place, each individual per- ceives a disparity between his own qualities as they pre- sently are, and as they might be with certain additions-- thus his struggle to bridge the gap. In all their essen— tials, the sought-after rewards of homosexual and hetero- sexual complementations are identical: the symbolic posses- sion of those attributes which, when added to one's own, fill out the illusion of completeness" (p. 93). This idea, previously described by Freud (1922), Reik (1944, 1957) and others, has been called the "Completion Hypothesis" (Cen- ters, 1971). Implications of this hypothesis on the study sexual motivations for homosexual males is the subject of this investigation. In this connection, Tripp (1975) has developed a thorough formulation of the hypothesis specifi- cally in reference to homosexual men. Although he describes several possible theoretical pathways which may eventuate in the development of homosexu- al proclivities in males, he notes that in every case some aspects of maleness have been invested with erotic signifi- cance to the extent that sexual arousal is evoked by act or fantasy in which the stimulus is masculine. He argues that 26 this erotization, by its very nature, appreciates its tar- get. By such a raising of its target's value, the erotiza- tion of male attributes "alerts a boy to a hierarchy of male qualities and invites him to make comparisons in which his own assets may seem outpaced and outdistanced by those of a particularly admired partner" (p. 78). Erotization always tends to raise the value of the items it touches, not only by exalting them, but by keeping a person's aspiration level soaring ahead of his own attainments. Often the result is to make a person feel a sharp disparity between what he has and what he would like to have. Even the...utterly secure male who has eroticized male attributes is ready to improve what he has by sexually importing refinements and additions from an admired partner. Thus, in a sense, it hardly matters what a person thinks of himself; an exalted ideal is never fully satis- fied... (p. 78) What Tripp implies here is that strong urges toward fulfilling an elusive masculine ideal in many cases is an inevitable component of the homosexual orientation. Tripp takes a critical view, however, of the simplistic assump- tion that homosexual inclinations are given rise to by feelings of inferiority. He concedes that a correct theory of male homosexuality may legitimately imply inferiority feelings at some level, but only as they arise as a result of eroticizing masculine attributes, not as if the inferi- ority feelings themselves were responsible for the homosex— uality. He does not describe whether erotization of female attributes for heterosexual males can lead to a sense of 27 comparable inferiority. The act of sexual importation to which Tripp refers is quite thoroughly elucidated elsewhere in the same work. The import-export model of complementation is offered by Tripp as a description of the process underlying both homo- sexual and heterosexual motivations. Whereas a person of either orientation may "import" qualities admired in a sexu- al partner and lacking in himself, in the homosexual case it would seem that the advantages of complementarity are seri- ously threatened by direct comparability. Especially on the anatomical level is it the case that men are more alike than they are different. If Tripp's assumptions are correct, the question then arises as to what extent it is possible for a male to consider other men so attractive as to become sexu- ally aroused by them without also being beset by feelings of self-dissatisfaction, and/or persistent strivings toward self-improvement in the area of perceived masculinity. Coming to conclusions similar to those of Tripp, but by a different route, Kaplan (1967) examined same-sex attrac- tions with the psychoanalytic concept of the ego-ideal as a point of departure. He is careful at the outset, however, to point out that homosexual behavior, for both female and male individuals, may be seen as the culmination of a series of experiences and relationships unique to each individual for whom it is chosen as a mode of sexual expression. Thus, with no single, unitary causality implied, the search for the ego-ideal is considered by Kaplan to be a possible 28 factor which plays a part in the development of a homosexual orientation in individuals whose self-images are "devalued or impaired." Freud considered the neurotic to be "impover- ished in his ego and incapable of fulfilling his ego-ideal" (Freud, cited in Kaplan, 1967). In the narcissistic quest for something to replenish these failing resources, a sexu- al ideal may be chosen "which possesses excellencies to which he cannot attain" (Kaplan, 1967). Kaplan fails to ex- plain his introduction of Freud's observations concerning neurotics to support a theory of homosexuality, but it is clear that a mechanism Freud considered to underlie neuro- sis, Kaplan proposes as a motivational force for homosexual behavior in some men. Kaplan states that individuals whose sexual orienta- tion is primarily or exclusively homosexual typically place major emphasis on the physical or personal attractiveness of the sexual object chosen; and to an extent considerably greater than do heterosexuals, they choose as sexual part— ners or fantasied sexual objects persons who possess char- acteristics in which they themselves feel deficient. Thus, dissatisfaction with the self, with the way one is, measured against internalized standards about how one would like to be, may be one of the major roots of some homosexual feeling and behavior. (p. 356) Kaplan's emphasis here, like Tripp's, is that of the perceived disparity between what one is and introjections 29 of what one would like to be; i.e., between the self-image and ego-ideal. For both male homosexuals and lesbians, then, these feelings may take the form of how one would like to be as a man, and how one would like to be as a woman; re- spectively. In this way, Kaplan suggests that the homosexu- al object choice may be directed more toward the acquisition of an idealized object with whom to identify or introject than toward sexual gratification per se. As possible sup- port for this argument, he offers Bieber's (1962) study of male homosexuals in which the investigators found that al- most half of their subjects waw themselves as physically frail, inadequate or effeminate. He further speculates that under conditions where a suitable appealing model for masculine identification is lacking, the need for such iden— tification combined with a sense of personal inadequacy and a potent although undifferentiated sexual drive may even- tuate in a homosexual adaptation. In a partial response to why the mode of obtaining the desired identification must be seuxal, Kaplan makes the observation that the sexual ex- perience may stimulate identification fantasies, thus pro- viding partial motivation for relationships with other men. Orgasm is likened, by Kaplan, to drug-induced euphoria in which the wish to be like one's partner may be intensified and transformed into a sensation of such union with the partner that one actually feels he is the partner. Al- though Kaplan does not deny that such an experience may characterize orgasm in heterosexual intercourse, he says 30 it may be one additional source of motivation for homosexual men to engage in intercourse with other men. He does not address the significance of such an experience for hetero- sexual relationships. At this point, it is useful to return to Kohut's re- marks concerning the psychic fixation on archaic self-ob- jects that appears to manifest itself in what he refers to as an "intense form of object hunger" and "the compelling need for merger with the powerful object" (see page 9). The similarity of Kaplan's observations to Kohut's later re- marks in connection with children are striking: Since all bliss and power now reside in the ideal- ized object, the child feels empty and powerless when he is separated from it and he attempts, there- fore, to maintain continuous union with it. (Kohut, 1971, p. 37) Recall as well his comments about the patient, Mr. A., the objects of whose homosexual fantasies were always men of great bodily strength and perfect physique.... Occasionally he achieved orgasm and a feeling of triumph at the thought of masturbating a strong and physically perfect man and draining him of his power. (pp. 69, 70). The theme of power as a component of masculinity and its fantasied acquisition via the sexual experience appears with such frequency in this literature that it is one point of emphasis in the present study. 31 Kaplan reports the case of a 21 year old homosexual man who described feelings of physical inadequacy, who had of- ten daydreamed of being taller, stronger and more virile. These daydreams became transformed, according to Kaplan, in- to fantasies about sex with physically powerful men. Fur- thermore, the patient reported a wish to have his body be more like those of his partners and that sex became a means by which temporarily to accomplish this goal in fantasy. Kaplan argues that the discrepancy between self—image and ego ideal need not center around physical or anatomical details but may involve deficiencies in intellectual endow- ments, socioeconomic status, and interpersonal adroitness as well. In summary, Kaplan concludes that because of such self-dissatisfaction the homosexual man cannot love himself as he is, so he loves his ego-ideal in the person of his ho- mosexual partner. From this perspective, homosexuality is considered a narcissistic form of gratification. Kaplan makes this sweeping generalization in spite of his more ten- tative introductory statement in which he limits his specu- lations to men whose self-images are impaired. The dyna- mics implied here are virtually identical to those described in theories set forth by Tripp and Kohut,as previously out- lined. Kaplan, for example, asserts that "Covert admiration for the individual who possesses these highly valued charac- teristics may become sexualized, and 'instant identifica— tion' may be achieved...in the homosexual relationship" (p. 358). Tripp, in fact, goes even beyond this in 32 declaring that once these highly prized attributes are ero- ticized, they are raised in value all the more. This pro— duces a self-perpetuating cycle in which it would seem that one's personal and sexual goals never find their realiza- tion. It may be argued that if a person feels especially lacking in certain qualities, then this dissatisfaction it- self requires a prior attribution of a high degree of impor- tance to them. These features, then, will be greatly ad- mired when they exist in others, owing to the considerable importance placed on them. Such a "high-intensity admira- tion," according to 2253 Tripp and Kaplan, may lead to the erotization of certain features by homosexual persons in connection with men. These same-sex features, then, repre- sent for homosexuals characteristics which, if appropriated, would make them more like the persons they wish to be. Note that although Tripp argues that the goal of both heterosexu- al and homosexual attachments is symbolically to "fill out the illusion of completeness," for homosexuals actual pos- session of specifically same-sex attributes desired in lovers is an aspiration. For homosexual persons, then, the objects of erotization are at once stimuli for specifically and identifiably sexual arousal and a part of the conscious ideal self. There is probably no definitive way to ascer— tain if ever homosexual behavior is motivated by a narcis- sistic search for the ego-ideal, although the argument may 33 be rendered more plausible if the discrepancy between self- image and ego-ideal is found to be significantly greater for homosexuals than for heterosexuals. Centers (1971) undertook to test such a hypothesis for heterosexual attachments, and obtained findings which failed to support the principle that heterosexual attachments are based upon a motivation to "complete" the ego-ideal. Never- theless, if Kaplan's hypotheses are taken seriously, one would expect self-image--ego-ideal discrepancies not neces- sarily existing for heterosexuals, in homosexual persons. Furthermore, one would expect to find compensatory strivings for objects of love or sex who possess the highly desired qualities. For homosexuals, according to Kaplan and Tripp, qualities sought in a prospective lover are also those to- ward which aspirations are directed, including those quali- ties which are anatomical and isosexual in nature. Although Kaplan states that the envy and idealization of qualities in other men may center around any set of qual- ities and not only physical characteristics, it would seem that this latter aspect of his theory would be the one most likely to prove at least to some extent valid since in this are men most clearly distinguishable from women. It is not likely a narcissistic quest for the vicarious acquisition of social facility or status that makes a man homosexual since these qualities may be easily "imported" from women. There- fore, the body image of subjects is a point of emphasis in the present investigation. As previouSly pointed out, the 34 alleged assignment of such considerable importance to the variable power, especially on the anatomical level, by homo- sexual men is salient in the clinical literature. Hypothe- ses associated with this idea, therefore, are also tested in the present investigation. Empirical Literature on Self-Concept and Sexual Preference Chang and Block (1960) attempted to test the hypothesis that homosexual males are more strongly identified with their mothers and less identified with their fathers than are nonhomosexual males. The investigators measured strength of identification in terms of the degree of cor- respondence between a subject's description of the parent under consideration and that of his ideal self. This mea— sure was obtained through the use of a list of 79 adjec— tives to which the subject was to respond as either charac- teristic or uncharacteristic of the given parent and of his ideal self. There were 20 homosexual and 20 heterosexual subjects, none of whom were receiving psychotherapy. The hypotheses that homosexual men show a significantly greater degree of identification with their mothers and a signifi- cantly lesser degree with their fathers were both supported. The homosexual and control groups did not differ signifi- cantly in their degree of self-acceptance as measured by the degree of correspondence between the perceived self and ideal self of subjects. In addition, it was found that the two groups did not differ significantly in regard to the 35 kind of ego ideal to which they aspired. Similarly, Greenberg (1973) found that although male homosexual subjects tended to evince greater feelings of alienation from society, they exhibited self-esteem scores comparable to those of heterosexuals on the Rosenberg Self- Esteem Scale (Rosenberg, 1965). These findings may be ques- tioned to some extent, however, on the basis that data for his "control group" were derived from other studies conduct- ed by other researchers as far back as 1955. Sallee (1976) attempted to assess self—concept of male homosexuals as classified according to the variable of sex- role identification. This approach was based upon the sug— gestion that male homosexuals are as variable in their sex- role identification as heterosexuals. Subjects were classi- fied as masculine, feminine, androgynous and undifferentiat- ed according to the Bem Sex-Role Inventory (BSRI)(Bem, 1974) and were then administered the Tennessee Self-Concept Scale (TSCS)(Fitts, 1965). Sallee found that no significant dif- ferences obtained in overall self-concept or frequency of sex-role identification classification between homosexual and heterosexual groups, nor did self-concept vary signifi— cantly according to sex-role identification for either group. Sallee notes that significant trends were more likely to be obtained in analyses of individual TSCS scales. Interest- ingly, within the male homosexual group, subjects identified as androgynous and feminine scored higher than masculine and undifferentiated subjects on several TSCS subscales. 36 Using the TSCS as a measure of psychological adjust- ment, Hart (1978) undertook a similar investigation. Gender characteristics were assessed by multiple methods including the BSRI as well as a questionnaire developed by the inves- tigator. In direct contrast to Sallee's findings, Hart re- ports that homosexual men with "norm-violating" gender char- acteristics obtain lower self-concept scores than homosexual men with more typical gender characteristics. She observes that the homosexual subjects who displayed few masculine- typed attributes and many gender traits typically associated with femininity tended to come from disturbed families and suffered from symptoms associated with neurosis, personality disorders, and hypochondriasis as adults. Hart speculates that men with a fragile sense of their own masculinity do not benefit from the feminine typed attributes in their personalities because they experience them as threats to their masculine gender identities. Ra- ther than enhancing adjustment, Hart says, for a man who is less sure of his masculinity these feminine components dimin- ish self-esteem. Somewhat along these lines, Peretti, Bell and Jordan (1976) define a typology of homosexual men on the basis of the nature and outcome of the childhood Oedipal situation. The so-called Oedipal male homosexual is described by these authors as maintaining a great attachment to his mother which 37 should have been resolved roughly between the ages of three and four. The child who does not resolve the Oedipal cri— sis so successfully, then, goes on to internalize feminine characteristics as a dominant part of himself,according to Peretti, et a1. These authors maintain that the Oedipal male homosexual is often harassed by feelings of guilt and shame, appears more withdrawn and lonely, and is intent upon concealment of his homosexuality. They add that he does not like what he is and attempts to change himself with or without resort- ing to professional counseling. Finally, they suggest that the Oedipal homosexual man tended to be overprotected by his mother as a child, more dependent upon her, and openly pre- ferred by her to his father. They add that these boys were also more "likely to establish a coalition with mothers a- gainst fathers than non-Oedipal male homosexuals." Although the specific criteria used for classifying their subjects are not reported by the authors, their re- sults indicate several tendencies in Oedipal subjects which are associated with a negative self-image. Oedipal male homosexuals were found to have less self-worth, self-confi- dence and self-acceptance than non-Oedipal subjects. Self- concepts, self-attitudes and self-motivation were additional variables which, consistent with the hypotheses outlined above, were consistently less favorable for the Oedipal group. In spite of the investigators' failure to specify the bases for classification of subjects, this investigation 38 raises questions concerning the universal applicability of Bieber's findings in regard to homosexuality and suggests a need to reconsider the tendency to approach research as if homosexuality were a developmental and phenomenological uni- ty. Dickey (1961) examined feelings of adequacy in homosex- ual males as defined by two measures: (a) a measure of self- image--ideal-self discrepancy by which traits were rated on a seven-point scale and (b) a direct measure consisting of statements pertaining to adequacy and self-concept to which subjects either agreed or disagreed. Consistent with the findings of Hart, previously cited, Dickey reports that (a) homosexual males who perceive more desirable characteristics in the role of the typical heterosexual male tend to feel more adequate and (b) feelings of adequacy are probably greatest in homosexual males who see themselves as more like the typical heterosexual male than like the typical homosex- ual male as the subjects themselves defined them. It seems, then, on this basis, that a likely candidate among sources of low self-esteem and feelings of inadequacy among homosex- ual men who experience them is a large self-ideal perceptual discrepancy in connection with masculinity. Although it has been suggested that such a discrepancy is typically given rise to by insalubrious family dynamics or cross-sex iden- tification patterns, recent research argues that systematic distortions in perceptions of the ideal man and what consti- tutes the ideal male role account for such discrepancies 39 rather than the actual quantity or salience of feminine at- tributes (See Skrapec and MacKenzie, 1981, below). In an effort to assay the components of low self-esteem among homosexual men based on research done by Rosenberg (1965), Sobel (1976) remarks that homosexual men tend to have both low body satisfaction as well as poorly modeled sex typed behavior. In a study designed to assess such body self-image characteristics of homosexual men, Prytula, Well- ford and DeMonbreun (1979) sought to determine whether and to what extent differences existed between heterosexual and homosexual men during adolescence relative to actual body characteristics. They also studied how subjects' body char- acteristics were perceived by their peers and families, and examined their perception of how others perceived the sub- jects' body characteristics. The study investigates the in- teraction of the homosexual man's recalled body image and his overall self—concept including interpersonal and famili- al factors which related to his self-image during adole- scence. On the basis of prior research, Prytula, et a1. point out that many male homosexuals report low self-esteem and ty- pically have stronger feelings of physical and social inade- quacy than heterosexuals. On the basis of findings reported by Saghir and Robins (1973), these investigators posit that repeated negative feedback prompted by effeminacy in child- hood has contributed to the development of a negative body image and overall self-concept during adolescence: 40 Saghir and Robins (1973) suggest that the dif- ference between the childhoods of most male homosexuals and heterosexuals was not only in the particular behavior patterns of the homo- sexuals during childhood, but also in their physical appearance, the perception of their physical appearance by others, and their per- ception of how their physical appearance was perceived by others. (p. 567) Using their retrospective self-report inventory with scales developed to assess six areas of general adjustment during adolescence, the researchers report that over all scales homosexual males scored significantly lower (in- dicating poorer adjustment) than heterosexual males. They account for this difference as being a result of greater recalled dissatisfaction with general physical characteris- tics and body image as compared to heterosexuals. Further- more, the homosexuals characterized themselves as having significantly different body charactristics, receiving nega- tive feedback because of their body characteristics from peers and/or family, and as having generally less positive body self-images and overall self-concepts during adolescence than did heterosexuals. Prytula et al. suggest further research via longitudinal studies to identify the possible role of body image varia- bleS'in development of. homosexual preference. Further- more, they point out the necessity for taking such variables into account in connection with therapeutic and prophylactic intervention where indicated. In spite of the fact that the 41 authors do not specify the indications for such interven- tion, it might be added that where such body image concerns are in evidence, both therapeutic and preventive interven- tion might be more appropriately directed toward negative self-concept than toward sexual orientation. These authors do recognize that inferior self-concept may in fact be a variable frequently associated with homosexuality in males though not a causal factor underlying it. Alpert (1978) used a semantic differential rating scale to assess the degree to which homosexual men have internal- ized popular negative attitudes regarding homosexuality. It was hypothesized by Alpert that there is essentially no dif- ference between the way homosexuals and heterosexuals char- acterize the ideal man, but that homosexual men rate them- selves significantly further from their characterizations of this ideal. Alpert found that homosexual and heterosexual men rated the ideal man in essentially the same way. The remaining hypothesis was not confirmed, however, since on several adjective pairs the homosexual subjects rated them- selves as significantly closer to the ideal man than did heterosexuals. Based on these data, Alpert concludes that internalized stigma appears largely confined to feelings of being less adequate than heterosexual men in terms of attri- butes associated with the stereotype of masculinity. In their examination of gender related components of self-perception, Skrapec and MacKenzie (1981) compiled test data for three matched groups of eight homosexual, eight 42 heterosexual and eight preoperative transsexual anatomical males. Self-concept is regarded by these investigators as a composite of interacting subsystems including core self- esteem, core gender identity, etc. They assume that a clear gender identification is the outcome of a developmental pro- cess which involves drawing referents "from an organized system of beliefs as to the psychosexual meaning of being a male or female" (p. 358). In other words, being a male or a female involves having a set of beliefs about physical appearance, gender roles, sexual preference and psychologi- cal makeup. As a process of development, then, individuals incorporate the resultant composite into their sense of self; which makes it inevitable that the individual will as- sess the degree of congruence between his or her gender identity and anatomical sexual identification. Based on results from the Repertory Grid Technique, a procedure derived from Kelly's (1955) theory of personal constructs, the researchers found that transsexual subjects described themselves as more like females. Homosexual sub- jects described themselves as more like males, and the he- terosexuals described themselves as equally similar to males and females. These results were based upon each subject's own descriptions of males and females. Additionally, Skrapec and MacKenzie found that scores on the Rosenberg Self-Esteem Scale (Rosenberg, 1965) indi- cated that transsexuals had the most negative global self— esteem and homosexuals the most positive, with heterosexuals 43 scoring in the middle. All comparisons were statistically significant. One heterosexual subject scored at almost max— imum self-dissatisfaction, however, and the authors observe that when this score is removed the heterosexual and homo- sexual groups are not significantly different from each other. Regarding gender identification, the Derogatis Sexual Functioning Inventory (Derogatis, 1976) revealed that homo- sexual men more strongly endorsed masculine traits for them- selves in contrast to transsexuals who revealed the most femininity with respect to gender role definition. Inter- estingly, on eight of the ten subscales homosexual subjects scored higher than either group. The only subscale in which the homosexual subjects scored lower than heterosexuals were Affect Balance and Body Image. The latter category had a mean standard score lower than that of any other group on any measure except two (i.e., transsexuals had lower scores On Body image and Gender role definition). Body image was also lowest among mean standard scores for the ten subscales in the homosexual group. Results on the BSRI were equivocal when a comparison was made between t—ratio and median-split procedures for analyzing the data, although according to both methods the homosexual group was found generally to describe themselves in masculine terms. Scores for heterosexual subjects large- ly suggested masculine or undifferentiated gender identities while transsexuals appeared either feminine or androgynous. 44 Since over all measures used to assess gender identity homosexuals generally described themselves in a more rigidly stereotypic fashion with reference to masculinity, the au- thors suggest that One might suspect that the gender identification component of self-concept was most fragile in this group of individuals. The data suggests that they introduced a systematic distortion in- to their perceptions of maleness and male roles. One possible explanation could be the systematic use of denial and reaction formation. The re- sults could be explained by using the idea of com- pensatory masculine responding, where masculini- ty is defensively exaggerated in the face of gen— der role "threat"... Such a response would be un- derstandable in situations where the homosexual's wish for enduring interpersonal relationships is met by a reality of brief, more superficial en- counters. In a sense then, his global self-eval- uation is "over-determined," with some kind of compensation operating at a cognitive level. (p. 368) Finally, it might be noted that homosexual subjects re- ported a higher global self-regard than self-ratings on the Repertory Grid would suggest. While the correlation between scores on the Rosenberg Scale and the Repertory Grid was significant for the heterosexual subjects (r=0.88, p<.01), no such relationship was found for homosexual (r=0.06) or transsexual (r=0.10) groups. These discrepancies suggest that homosexual subjects, who reported highest global self- esteem, used different referents in describing themselves on the Repertory Grid. This information, along with their aty- pical scores on the Body Image subscale of the Derogatis 45 Sexual Functioning Inventory further suggests that continued exploration of body self-perceptions among homosexual men is an area of potentially revealing investigation. Summary Investigations to date which address self-esteem in global terms generally indicate that no significant differ- ences exist between homosexual and heterosexual men on this dimension. Data on sex role identification are equivocal, with a significant proportion of findings indicating that there are no differences in overall frequency of cross-sex gender identity between groups and that in some cases homo- sexual men describe themselves in more masculine terms than heterosexual men. When subjects are further classified ac- cording to gender role identification there is considerable agreement that male homosexual subjects who see themselves in more feminine terms, as well as those who as children have identified with or developed atypically intense attach- ments to their mothers, are more likely to have more nega- tive self-concepts. Comparable data for heterosexuals are less available and indicate a need for further research. The data support that homosexual men in general, while ex- hibiting more negative self-concepts in some circumscribed areas, have more positive self-perceptions in others. On the basis of a number of investigations it appears that less positive self-concepts of the homosexual men for whom this is a problem are characteristically in the area of perceived 46 masculinity, often in connection with body image in particu- lar. To determine to what extent this is indeed the case and the particular aspects of the masculine self-concept which are most problematic is one aim of this study. A final consideration in this connection is whether discrepancies in self-concept between homosexual and hetero- sexual groups of men result from systematic distortions in concepts of what is appropriately and desirably masculine. In regard to the important implications of this research to psychotherapy, Hart (1978), for example, urges closer atten- tion on the part of the therapist to gender characteristics of clients and to helping them become more secure in their sense of their own masculinity. Kaplan suggests that treat- ment of homosexual men, such as he describes in his afore- mentioned case studies, consist of exploring the origins of the negative self-image which may be found in family dyna- mics, peer attitudes, etc. (Kaplan, 1967). Of perhaps equal importance, however, are the origins of the ”ego-ideal" which is itself a "composite of early identifications, in- trojections and wishes" (p. 356). He concludes that "the self-image, the ego-ideal, or both, may be unrealistic con- structs in the patient's mind, he may devalue himself out of proportion to his real life situation" (p. 358). Before turning to a statement of the hypotheses with which this study is concerned, a word about sampling consi- derations is in order. Gonsiorek (1982a) describes the problem of defining and obtaining a suitable sample as the 47 largest single methodological problem in the scientific study of homosexuality. The most frequently used sources of subjects for such research have consisted of psychiatric and legally involved populations, patrons of gay or lesbian bars, and individuals obtained through lesbian and gay or- ganizations. He calls attention to the fact that the phe- nomenon of homosexuality traverses the entire range of so- cial class, age, ethnicity, etc. As a research entity, therefore, it is among those most consistently uncorrelated with demographic groupings typically considered of import- ance to social scientists. Even research which draws homosexual and heterosexual subjects from apparently comparable groups may fail to de- tect subtle interactions between sexuality and other factors such as psychiatric diagnosis. For example, Gonsiorek ob- serves that for reasons less than straightforward, it might be more difficult to be homosexual and schizophrenic than heterosexual and schizophrenic. For these reasons, then, disparities between homosexual and heterosexual patients or legally involved groups may be considerably exaggerated from those which may exist for groups not characterized by such specific problems. The foregoing is not to suggest that sound research on human sexual orientation can be accomplished only with Egg representative sample, this being a hypothetical construct the approximation of which is our goal. It is less useful to make the degree of this approximation a criterion for 48 "goodness" of our sample than it is to precisely specify limitations on the generalizability of findings. Research using legally or psychiatrically involved samples, then, is useful if the questions asked by the investigator have specifically to do with these populations. Investigators who pose questions about the population at large, however, will find such samples highly limited in their ability to provide reliable or correct answers. These considerations apply equally to samples drawn from gay organizations, which involve subjects likely to be more open about their sexuali- ty and probably more politically conscious; as well as to samples recruited via friendship networks, since they tend to be demographically homogeneous. As Gonsiorek advises, even though it is impossible to obtain a completely representative sample, any sample of homosexual subjects should attempt to mimic the major demo- graphic characteristics of the locality from which the sam- ple is derived and should be as diverse as possible. In cases in which skewness of sample is apparent-~a detailed description of sampling procedure will make limitations on generalizability clear. 49 Statement of Hypotheses Psychoanalytic theories frequently argue that homosexu- ality in males symbolically expresses the ego's need to ac- quire masculine components of the ego-ideal. Based on clin- ical observations, the assumption is made that the discre- pancy between actual and ideal self is greater for homosexu- al than for heterosexual men. It is the purpose of this study to test this and related assumptions for nonclinical populations. The following hypotheses, therefore, are eval- uated: Hypothesis 1 There is a significant difference in global self-esteem between homosexual and heterosexual males. The findings are not consistent across studies for this hypothesis, a fact which may be a function of the variety of measures and research designs employed. Hypothesis 2 There is a significant difference between homosexual and heterosexual subjects in the overall discrepancy between self-image and perceived ideal self. That this hypothesis will be supported is predicted by Kaplan (1967) since, like Tripp (1975), he believes homo- sexual motivations to consist of one's quest to incorporate, through the sexual encounter, qualities pertaining to the ideal self, and of which he or she feels destitute. It should be noted that this hypothesis is different from Hypo- thesis 1 since global self-esteem does not necessarily pre- dict in absolute terms the quality of either actual or ideal 50 self, and may only imperfectly reflect the distance between them (Skrapec and MacKenzie, 1981). Hypothesis 3 There is a significant difference between homosexual and heterosexual males in their degree of overall body satisfaction. This hypothesis is based upon clinical reports which frequently emphasize the impaired body image characteristic of the patients studied (Bieber et al., 1962; Kaplan, 1967; Kohut, 1971). Hypothesis 4 There is a significant difference between homosexual and heterosexual men in the relative degree of impor- tance placed on the physical attractiveness of the sexual object. In support of his argument that homosexual men are mo- tivated to have sexual experiences with men as a way to ap- propriate vicariously the characteristics in which they feel deficient, Kaplan (1967) states that People whose sexual orientation is predominantly or exclusively homosexual usually place major emphasis on the attractiveness of the homosexual object- choice. At times the personal or physical charac- teristics of the potential homosexual partner seem to be of considerably greater importance to the ac- tive homosexual...than to the person seeking he- terosexual intercourse. (PP. 355, 356) This hypothesis refers to sexual encounters regardless of whether in the context of more stable relationships. Be— cause what he refers to as "personal characteristics" is not explicit, the hypothesis will be tested in terms of physical 51 attributes alone. Stated in these terms, the following hy- potheses logically follow: Hypothesis 5 For homosexual males there is a positive correlation between degree of body dissatisfaction and the rela- tive degree of importance placed on the physical at- tractiveness of the sexual object. Hypothesis 6 For heterosexual males there is a positive correlation between degree of body dissatisfaction and the rela- tive degree of importance placed on the physical at- tractiveness of the sexual object. On the basis of Kaplan's argument summarized under the previous hypothesis as well as Tripp's (1975) observations, also previously cited, it would be expected that for homo- sexual males there would be a significant positive correla- tion between degree of body dissatisfaction and the degree to which physical attractiveness in the sexual object is em- phasized. Such a correlation, on this basis, would not be expected for heterosexual males. The purpose of these hypo- theses, then, is to assess whether a "search for the ego- ideal" in Kaplan's (1967) terms is more generally true of homosexual men than of heterosexual men. They examine, how- ever, only one possible form which the hypothetical search might take and tests whether it is specific to or more generally characteristic of one of the two groups. Kaplan has prOposed, it will be recalled, that for a homosexual man, this search may comprise an unconscious attempt to ap- propriate not only physical, but social or intellectual characteristics as well. This hypothesis is tested here in 52 regard to the physical dimension alone. Hypothesis 7 There is a significant difference between male homosex— ual and male heterosexual subjects in their degree of present identification with the primary male caregiver in childhood (father, stepfather, etc.). According to findings by Chang and Block (1960), homo- sexual men identified more strongly with their mothers and less with their fathers. Bieber et a1. (1962) report that as children homosexual men had mothers who behaved toward them in a way described by him as "close-binding" and inti- mate and fathers who were described as detached or hostile. Bell, Weinberg,and Hammersmith (1981) report that homosexual men identified less with their fathers than heterosexual men, but add that this tendency appeared to have little eventual influence on their adult sexual orientation. The analogous hypothesis with reference to mothers is considered not so germane to the theories under consideration since it is with the relationship between homosexual motivations and masculine gender identification that this investigation is primarily concerned. The following two hypotheses are fur- ther elaborations of Hypothesis 7. Hypothesis 8 There is a negative correlation between the degree of present identification with the primary male caregiver in childhood and the degree of discrepancy between ac— tual self-image and ideal self for male homosexuals. 53 Hypothesis 9 There is a negative correlation between the degree of present identification with the primary male caregiver in childhood and the degree of discrepancy between ac- tual self-image and ideal self for male heterosexuals. According to Kaplan (1967), "One might speculate that the search for an ego-ideal via homosexual relationships is often a substitute for the more usual ascription of this role to the father" (p. 356). This predicts a negative cor- relation between the degree of childhood identification with a masculine object and the degree of discrepancy between self-image and ideal self for homosexual subjects. With in- formation acquired from tests of Hypotheses 7, 8, and 9, the correspondences among variables such as gender-role identi- fication, identification with the primary masculine child- hood object, and aspects of self-image may be examined. It may then be determined to what extent, if any, these corres- pondences differ between groups. For example, findings which indicate that positive self—regard is more highly de- pendent upon masculine self-perception for homosexual men than for heterosexual men suggests an interesting line of further investigation (Hart, 1978; Skrapec and MacKenzie, 1981). Hypothesis 10 There is a significant difference between homosexual and heterosexual males in the degree to which they feel a personal sense of powerfulness. This hypothesis is based in part upon the observations of Tripp (1975) and the clinical data adduced by Kaplan 54 (1967) and Kohut (1971); each in support of his own version of what has been called the "Completion Hypothesis" (Cen- ters, 1971). Fantasied sexual partners were found by these writers to be men whose most salient feature was great phy- sical powerfulness. They, like Bieber et al. (1962), add that these homosexual men saw themselves as physically weak. This hypothesis is intended as the first step in the effort to determine if the motivation toward a vicarious acquisi- tion of a general sense of powerfulness is more characteris- tic of male homosexual than of male heterosexual experien- ces. Hypothesis 11 There is a positive correlation between masculine self- perception/masculine gender role identification and overall body satisfaction for homosexual males. Hypothesis 12 There is a positive correlation between masculine self- perception/masculine gender role identification and overall body satisfaction for heterosexual males. If data support Hypothesis 11, then plausibility is en- hanced for the position that body dissatisfaction in homo- sexual men for whom it is salient involves deficiency in masculine self-concept and is engendered by weak masculine gender role identification (see Bieber, 1962; Hart, 1978; and Skrapec and MacKenzie, 1981). Hypothesis 12 is tested to assess whether the same is true for heterosexual men. Hypothesis 13 There is a significant difference between homosexual and heterosexual males in their degree of masculine self-perception/masculine gender role identification. 55 As noted above, prior research (Hart, 1978; Skrapec and MacKenzie, 1981) has suggested that differences may exist between groups in terms of masculine self-perception and/or masculine gender role identification. Findings obtained in connection with this hypothesis are of interest in deter- mining to what extent masculine self-perception matches ac- tual gender role identification for the two groups (Hart, 1978). CHAPTER 2 METHOD Subjects Subjects were 53 homosexual and heterosexual males ranging in age from 19 to 45 years. There were 24 gay sub- jects ranging in age from 19 to 45 years, with a mean age of 28.1 The remaining 29 subjects were heterosexual men rang- ing in age from 19 to 37 years, having a mean age of 26. One of the gay subjects was legally heterosexually married at the time of his participation in the study, but along with the other 23 unmarried gay subjects affirmed that he had had sex only with men for a period of at least one year. Average highest level of education completed for gay sub- jects in this study was 3 years of college; for heterosexual subjects, 4 years of college. Occupations represented with- in each of the two samples varied widely, including blue and white collar workers as well as full-time college students. Major demographic features of the two samples are summarized in Tables 1 and 2. One variable pertaining to gay subjects, labeled In what follows, the term "gay" is used as well as the term "homosexual." The former term has distinctive psychologi- cal, social and political connotations and implies self-ac- ceptance in regard to sexual orientation. For that reason, individuals subsequently described as gay are those who choose to apply the term to themselves. The term "homosex- ual"is much broader in scope and applies to all individuals characterized exclusively or almost exclusively by same-sex sexual attraction and behavior. 56 57 Self-Disclosure Regarding Sexual Orientation, could not be classified objectively. The values listed under this vari- able inlhbhsl.are intended as a roughly descriptive cate- gorization. Subjects were assigned to one of these cate- gories by the investigator as follows: Subjects described as "not out at all" were two sub- jects who said, without elaboration, that none of their fam- ily, friends, or acquaintances know that they are gay; and one who said he presumes others know but that he never dis- cusses the issue with anyone. Subjects described as "entirely out" were the five sub- jects who said that they were completely open about their sexuality, all of the family and friends knowing they are gay. If they were employed, they also added that they are known as gay by their co-workers. Subjects in the remaining category were described as "partially out." Clearly, there was room for considerable variability in this category. In general, these subjects specified that only some family members and some close friends know that they are gay or that they have not been open about their sexuality with their family and/or their co-workers, but are known by their friends as gay. The subjects were recruited by the investigator and two social science students at Michigan State University. Sources through which gay subjects were obtained were vari- ous, including gay organizations and friendship networks. Heterosexual subjects were recruited once the gay sample 58 had been composed so that the two groups could be made to approximate each other in terms of age, education, occupa- tion, and relationship status (has or does not have a lover/ romantic involvement). Like the gay subjects, heterosexual participants were from Southcentral Michigan and were re- cruited from a variety of sources such as factories, banks, real estate agencies, other local businesses, and Michigan State University. All subjects stated that they had never been hospitalized for a mental or emotional disorder and had not had psychotherapy or psychotropic medication for at least six months prior to their participation in this study. The homosexual men who participated in this investiga- tion described themselves as "gay" and as having had exclu- sively same-sex sexual/romantic involvements for a period of at least one year. The heterosexual subjects described themselves as "heterosexual" and as having had sexual/ro- mantic involvements only with women for a period of at least one year. Insert Table 1 about here Insert Table 2 about here 59 Table 1. Frequency Counts for Characteristics of Gay Men in Sample Variable Classification Range (in years) Age 19-21 22-25 26-29 30-33 34-37 38-41 42-45 5 6 4 2 3 3 1 Years High School and Above Education 4 5 6 7 8 9 10 4 4 4 l 7 2 2 Legal Marital Status Never Married Divorced Married 20 3 1 Occupation Laborer/Hourly Manager/ College/ Worker Clerk Graduate Std. Other [4] Factory [6] [9] [5] Worker (2) Computer Hairdres- Systems ser Bricklayer Mgr Physical Waiter Banker Therapist Real Est Store- Agent keeper Clerk - Librarian AccountantIZ) [No data] 60 Table 1 (cont'd.). Describes Self as "Out" Entirely Partially Not at All Self-Disclosure 5 16 3 Regarding Sexual Orientation Degree of Activity (1-3) (4-7) (8-10) in Gay Organizations No Activity Medium High (lo-point Likert Activity Activity Scale Ratings) 11 5 8 Binary Classification Yes N2 Relationship Status 15 9 (Has a lover/romantic involvement) Church Membership 10 14 (Is a member of an organized religion) Belief in God 13 8 (Believes in a single, personal God) (Missing values--3) 61 Table 2. Frequency Counts for Characteristics of Heterosexual Men in Sample Variable Classification Range (in years) Age 19-21 22-25 26—29 30-33 34-37 38-41 42-45 6 ll 4 5 3 0 0 Years High School and Above Education 4 5 6 7 8 9 10 12 13 14 2 2 4 l 8 4 4 l 2 1 Legal Marital Status Never Married Divorced Separated 25 3 1 Occupation Laborer/Hourly Manager/ College/ Worker Clerk Graduate Std. Other [6] Factory [6] I8] [9] Worker (4) Accounting Store Mgr Assistant Restaurant Barber Cook/Waiter (2) Energy Con- Univ. Prof servation Programs Librarian Mgr Mental Environ— Health Wkr mental ' Quality Sw1m / Specialist Sports (4) Trainer(3) Engineer 62 Table 2 (cont'd.). Yes N2 Relationship Status 18 11 (Has lover/romantic involvement) Church Membership 1? 11 (Is a member of an organized religion) (Missing values--1) Belief in God 22 6 (Believes in a single, personal God) (Missing values-—1) 63 Procedure Measures were administered by the investigator and the two research assistants. The same instructions for the com- pletion of the measures were given in writing to all sub- jects (see Appendix) so that they were able to complete the questionnaire/test packet on their own in a period of sixty to ninety minutes. Subjects identified themselves on mea- sures only by means of a numerical code, were assured that all materials would be kept confidential, and were told that in no case would names be attached to any test protocol or questionnaire. Subjects were paid $5.00 for completing the packet as agreed prior to their participating in the study. Instruments The following is a complete list of instruments used in this investigation appearing in the order in which they were administered. A more thorough description of these measures is provided in the subsequent section, "Assessment of Varia- bles." Semantic Differential Technique (Osgood, Suci, and Tannenbaum, 1957) Tennessee Self-Concept Scale, Physical Self Subscale (TSCS-PS) (Fitts, 1965) Personal Attributes Questionnaire (PAQ)(Spence and Helmreich, 1978) Body Cathexis Scale (BC)(Secord & Jourard, 1953) Rosenberg Self-Esteem Scale (RSE)(Rosenberg, 1965) 64 A questionnaire on personal background and current functioning composed of miscellaneous items, con- structed specifically for this investigation Measures Variables and concepts referred to in the preceding statement of hypotheses are listed below in connection with the instruments used to measure them and the hypotheses to which they pertain (hypothesis numbers in parentheses): Global self-esteem (1). For this variable, the Rosen- berg Self—Esteem Scale (RSE)(Rosenberg, 1965) is a conveni- ently administered measure which has the advantage not only of brevity, but of direct comparability with the recent findings of Skrapec and MacKenzie (1981) as well as those of Greenberg (1973), previously cited. In support of construct validity for the RSE, Wylie (1974) cites findings of Silber and Tippett (1965) in which monotrait-heteromethod conver- gent validity coefficients ranged among the highest found for such intermeasure correlations. These coefficients ex- ceeded both heterotrait-monomethod and heterotrait-hetero- method correlations, thus suggesting both convergent and discriminant construct validity for the instrument. Rosen- berg (1965) found a reproducibility index for the RSE of .92 and Silber and Tippett (1965) report a test-retest re- liability of .85. In addition, it is pointed out by Wylie that this method, constructed as a lO-item Guttman scale, has to some extent a built-in control for acquiescence re- sponse set since there are an equal number of alternately 65 presented items in which "agree" and "disagree" responses indicate high self-esteem. Also available in support of the RSE's validity as a measure of self-esteem is a large quantity of published data based on "known-group" comparisons (assumed validity). These involve variables such as depressive affect, interper— sonal insecurity and psychosomatic symptoms. For a more complete review of these, the reader is referred to Wylie (1974) and Burns (1979). In view of Wylie's (1974) critique of measures which consist of heterogeneous items whose relation to a unitary self-esteem dimension is questionable, and since the RSE was constructed according to a Guttman scaling procedure ex- pressly for the purpose of obtaining unidimensionality, per- formance of the RSE is here considered the best criterion for testing Hypothesis 1. Of course, performances on relat- ed measures and their possible relationships to this hypo- thesis are noted where indicated (e.g., where scores are either highly consistent or highly inconsistent with per- formance on the RSE). Self—image--ideal self discrepancy (2,8,9). A straightforward device which has a considerable range of applicability and which is less vulnerable than most stan- dard scales to response distortion through the influence of social desirability factors, the semantic differential tech- nique (Osgood, Suci, & Tannenbaum, 1957) has been chosen to asses the self-image--ideal self discrepancy. This 66 technique is a means by which the connotative or affective meaning of a concept for a particular individual may be measured. The semantic differential offers a way of ap- praising the dissimilarities in meaning among concepts as they exist according to the individual's personal semantic organization as well. In this case, the semantic differen- tial not only provides qualitative descriptions of the con- cepts MYSELF AS I AM and MYSELF AS I MOST WOULD LIKE TO BE; but permits quantification of the semantic "distance" be- tween them. For this technique, as it is used in this study, the investigator chooses the concepts to be measured, each of which is presented on a separate page. Each concept is rated on a series of seven-point bipolar scales with an ad- jective at one end and its opposite at the other. Subjects are instructed to evaluate each of the scales in connection with the concept under consideration and to put a check mark along the continuum at a point most descriptive of the rela- tive applicability of the two adjectives. The number of scales used is typically determined by the investigator as is the particular set of descriptors used for each. Factor analysis of Osgood's original set of 50 scales yielded the three principal orthogonal factors: Evaluative, Potency, and Activity, with the first of these accounting for the largest percentage of total variance. The Evalua- tion dimension is composed of scales primarily associated with "good" and "bad." The Potency dimension, on the other 67 hand, involves traits related to "strong" and "weak." Fi- nally, the third dimension, Activity, is composed of scales associated with "active" and "passive." Osgood recommends that the investigator choose an equal number of scales to represent each factor, in order to provide the subject with a "balanced semantic space which he may actually use as he sees fit..." (Osgood, et al., 1957, p. 78). Among earlier applications of the semantic differen- tial technique to certain attitudes (self-concept to a minor extent) in relation to sexual orientation, is Kendrick and Clarke (1967) to which the interested reader is referred. Relative degree of emphasis on physical attractiveness of sexual object (4,5,6). Subjects were presented with a series of blank lines on which they were instructed to fill in the characteristics or attributes most desired in a lover or romantic partner. It was further specified that these might include any details of personality, social, economic, physical or anatomical characteristics, as specifically stated as possible. Subjects were asked to try to list the characteristics as they occurred to them, and to do so as honestly as possible. The following measures were constructed to evaluate emphasis on physical attractiveness of the sexual object: Physical-General Frequency Ratio (PGFRQRTO)--Subjects were asked to list the characteristics or attributes they 68 most look for in a lover or romantic partner. The number of features listed which pertained to physical characteristics was compared to the number of general (nonphysical) features as an arithmetic ratio. Physical-General Difference in Rank (PGDIFRNK)--Sub- jects were instructed to rank the characteristics listed (as described above) on ten-point Likert scales. This vari- able was computed for each subject by subtracting the mean ranking of general characteristics from the mean ranking of physical characteristics. Frequency-Rank Product (FRQXRNK)--This variable is de- fined as the product of the values for the preceding vari- ables. To the extent that both PGDIFRNK and PGFRQRTO are valid measures of the Physical Emphasis dimension, the com- posite variable is even more sensitive to variations in Phy- sical Emphasis. Overall bogy satisfaction (3,5,6,ll,12). The measure- ment of this construct in connection with Hypotheses 5 and 6 has already been briefly discussed. The evaluation of Hypotheses 11 and 12 similarly depends upon correlations ob- tained in connection with the Body Cathexis Scale (Secord & Jourard, 1953). Body Cathexis is defined by the authors as the degree to which an individual is satisfied or dis— satisfied with various parts or processes of the body. The original test consists of a 46-item list of body parts and functions, although some studies have employed a 40-item modification (Jourard & Secord, 1954). These listed 69 functions and parts are rated by subjects on a five-point scale ranging from (1) Have strong feelings and wish change could be made somehow (strong negative) to (5) Consider my— self fortunate (strong positive). Eleven of the items most negatively cathected by a standardization group define the "body anxiety" subscales, one each for male and female sub— jects. The authors report split-half reliability coefficients for the 46-item scale respectively as .83 and .78 for 45 female and 45 male subjects. For the 40-item version, Wein- berg (1960) reports coefficients of .75 for females and .84 for males. There have been various approaches to the determination of the Body Cathexis Scale's construct validity as a measure of an individual's overall attitude toward body parts and functions. Secord and Jourard (1953) report a correlation of .58 for women and .66 for men between Body Cathexis and a global self-concept measure based on the same rating scale as used for the Body Cathexis Scale. This suggests a degree of construct validity for the test in that the way persons feel about their bodies is not unlike their self-referent feelings in many other areas. Similarly, a number of hypothesis tests have used theory-based expected correlations between the Body Cathex- is Scale and variables as wide ranging as nudist group af- filiation (Sugarman & Roosa, 1968), mental illness (Cardone & Olson, 1969), security-insecurity (Weinberg, 1960) and 70 size of body parts (Jourard & Secord, 1954). For a brief overview of these studies, the reader is referred to Wylie (1974). It may be mentioned here, however, that predicted correlations between the Body Cathexis Scale and these vari- ables have been obtained to an extent consistent with expec- tations for construct validity. Because of the general lack of measuring devices which purport to assess the same domain as the Body Cathexis Scale, convergent validity has not been adequately deter- mined. What information is available on its discriminant validity is considerably less persuasive than studies based on assumed validity, although considerably more work along these lines is needed (Wylie, 1974). It might also be added that the authors of the Body Ca- thexis Scale deliberately omitted bodily parts and functions explicitly referring to sexual and excretory functions "be- cause it was feared that their presence in the scale might give rise to an evasive attitude which would transfer to other items..." (Secord & Jourard, 1953, p. 344). Almost three decades later such an attitude seems considerably less likely, so the form of the Body Cathexis Scale used in the present study includes the additional items "penis" and "buttocks" added at the end of the form with no discontinui- ty in format or typeface. Finally, although it is not in- tended to affect actual evaluation of the hypotheses here listed, these results will be compared with those found on the Physical Self subscale of the Tennessee Self Concept 71 Scale (Fitts, 1965). Identification with the primary male caregiver in childhood (7,8,9). This and related variables have been previously evaluated by a variety of methods. As described above, Chang and Block (1960) used a list of adjectives to 60 of which a subject was to respond as either particularly characteristic or particularly uncharacteristic first of ideal self, then of father, mother, and actual self. For each of these, subjects were allowed to choose the 60 adjec- tives about which they felt strongly enough to respond, but were required to respond with only and exactly 30 X's (par- ticularly characteristic) and 30 O's (particularly uncharac- teristic). For each subject, a mother identification score and a father identification score vereobtained as follows: a score of "l" was assigned each adjective in which the rating given (X, 0, or unmarked) was the same for both 2222; self and the parent under consideration, and the total of these scores was taken as the "identification score." Skrapec and MacKenzie (1981), it will be recalled, used the Role Construct Repertory Test (Repertory Grid Technique) (Kelly, 1955) as a measure of gender identification. This procedure permits the subject to devise a set of descriptors according to his or her own system of "personal constructs" rather than respond to a set of predetermined attributes. Skrapec and MacKenzie assessed gender identification in terms of "grid distances" (i.e., degree of similarity or dissimilarity) between a subject's description of himself 72 and his descriptions of both males and females. Identification is conceptualized in the present study in broad terms and it is understood to comprise elements of imitation, affiliation, admiration, idealization, etc. Be— cause identification in this study has as its emphasis out- come rather than process, it is assessed as perceived simi- larity between self and parent under consideration, much in the way as discussed in connection with the Chang and Block study described above. Whereas in that investigation iden- tification was determined as the degree of correspondence between descriptions of ideal self and parent, such a cor- respondence between perceived actual self and parent is con- sidered a more accurate index of this variable as defined in this study. The semantic differential technique was used to assess these correspondences. The method by which identification is measured is identical to that described by Endler (1961) involving use of Osgood's 2 index (Osgood, et al, 1957) as a measure of semantic "distance" between concepts. Where necessary, instructions for completion of the se- mantic differential provided that the concepts MY FATHER and MY MOTHER might be replaced by the role name of the primary male and female caregiver in childhood, respectively. Such substitutions were permitted only if (a) the biological par- ent was absent, and (b) the parental substitute functioned in the role typically assumed by the parent. There was only one instance of such a substitution, and the subject 73 reported that the male caregiving substitute was not known to him during childhood, but was an employer for a short time during adolescence. Factor scores, in this instance, were not used in the analysis of the data. Personal sense of powerfulness (10). The reasons for choosing the semantic differential technique as the measure to assess several of the variables in this investigation have been specified. In order to test Hypothesis 10, the additional concept A POWERFUL MAN was added to the collec- tion. The wording of this concept was intentionally devised to be ambiguous to allow subjects to "project" into it their own meanings. In other words, it was hoped that each sub- ject's responses to the scales listed under this heading would be reflective of the aspects of power which to them were most personally salient. Subjects were advised that in responding to the scales they might have in mind a real per- son, a composite of several individuals, or someone purely imaginary. This concept was placed last in the sequence and at the end of the test, on the page following, subjects were asked to describe the mental picture they used in responding to this final concept. The semantic "distance" between the concepts MYSELF AS I AM and A POWERFUL MAN was computed for each subject in the usual manner and the means of these scores were tested for significant difference (Hypothesis 10). Finally, although it did not influence evaluation of Hypothesis 10, the degree of similarity between semantic differential ratings of 74 MYSELF AS I MOST WOULD LIKE TO BE and A POWERFUL MAN for subjects in the two groups was considered a matter of com- parable relevance to this investigation as were the semantic correspondences among MYSELF AS I AM, A POWERFUL MAN, MYSELF AS I MOST WOULD LIKE TO BE, and THE PERFECT LOVER. Masculine self-perception/masculine gender role iden- tification (ll, 12, 13). These two general constructs are listed as a single variable since they are more difficult to separate in terms of theoretical definition than in terms of operationalization. Masculine self-perception (in the nonphenomenal sense) is yet another variable the assessment of which involves the semantic differential technique, with correspondences between the concepts MYSELF AS I AM and MAN determining its evaluation. In addition, ratings of MYSELF AS I AM vis—a-vis WOMAN were examined for the two groups. Masculine gender role identification was determined by the Personal Attributes Questionnaire (PAQ)(Spence & Helm- reich, 1978). This is a self-report inventory which con- sists of 24 trait descriptions each applied in the form of a bipolar scale. These are distributed among three general scales labeled Masculinity (M), Femininity (F), and Mascu- linity-Femininity (M-F). The first two of these consist of items which Spence and Helmreich found to be socially desir- able in both sexes, but which were judged to be more charac- teristic of the sex to which each pertains. The third scale, however, Masculinity-Femininity, consists of items the social desirability of which tends to be considered 75 gender-specific. The scale itself is bipolar, scored in a masculine direction. The investigators found significant differences between means for the two sexes on each item in two independent sam- ples of college students. For each of the three scales dif- ferences were found in the expected directions (Spence, Helmreich, and Stapp, 1975). In terms of intercorrelations among the scales, they found, contrary to conceptualizations which demand a single bipolar masculine-feminine dimension, that correlations between M and F were low positive in both sexes. The authors consider this finding supportive of a "dualistic" conceptualization of gender identity. On the other hand, correlations between the M-F scale and the M scale were moderately high positive whereas the correspond— ing correlations involving the F scale were less striking yet substantially negative. The view of the authors, based on subsequent analyses as well, is that the data support a simultaneously dualistic and bipolar model of masculinity and femininity and that the M-F scale has the potential of yielding information not available from the M and F scales alone. Correlations reported between PAQ and the Bem Sex-Role Inventory (BSRI)(Bem, 1974) on comparable scales, for males and females have ranged from .57 to .75 (Stapp and Kanner, cited in Spence and Helmreich, 1978). Spence and Helmreich speculate that the differences between the two instruments are accounted for by factors such as the lack of a bipolar 76 masculinity-femininity scale in the BSRI. They point out, in addition, that a number of the trait descriptions per- taining to the M-F scale of the PAQ were found on both the M and F scales of the BSRI. For Hypotheses 11 and 12, appropriate product-moment correlations were computed between scores on the Body Ca- thexis Scale and each of the two PAQ subscales M and M-F. Similar correlations were computed between scores on the Body Cathexis Scale and the discrepancy scores (Osgood's 2) between the concepts MYSELF AS I AM and MAN. Since Osgood's 2 is used in this study to calculate all "distances" between semantic differential concepts, fre- quent references are made to distances between concepts as follows: 2 (concept, concept). For example, 2 (MYSELF AS I AM, MAN) is the actual geometric distance between the con- cepts MYSELF AS I AM and MAN when these concepts are located in three-dimensional "semantic space." CHAPTER 3 RESULTS Results of Hypothesis Tests Hypothesis 1 There is a significant difference in global self-esteem between homosexual and heterosexual males. A two-tailed test of significance was applied to scores obtained by the two groups on the Rosenberg Self-Esteem Scale. The mean RSE score for gay subjects (Mé35.04, §2é 4.33) was significantly higher than for heterosexual sub- jects (Mé3l.07, §2?4.46), E (51)=3.27, p<:.002. Hypothesis 2 There is a significant difference between homosexual and heterosexual subjects in the overall discrepancy between self-image and perceived ideal self. For each subject, the semantic "distance" between con- cepts MYSELF AS I AM and MYSELF AS I MOST WOULD LIKE TO BE was computed using the 2 formula (Osgood et al., 1957).1 No significant difference between homosexual and heterosexu- al groups was found for 2 (MYSELF AS I AM, MYSELF AS I MOST WOULD LIKE TO BE) (g;8.71, gp;4.7o and §;9.14, sp;4.22; 1The index here referred to as D is actually a modified version of Osgood's measure. To avoid cumbersome mani- pulation of very small decimal values, raw, rather than averaged, semantic differential factor scores were used. The relationships among factor scores for the concepts, and hence among the 2, remain unchanged (see Osgood et al., 1957). 77 78 respectively), 2 (51)=-0.35, (NS).2 Hypothesis 3 There is a significant difference between homosexual and heterosexual males in their degree of overall body satisfaction. Means were compared for homosexual (Mél76.17, §2=25.96) and heterosexual (Mé168.4l, §2é19.66) scores on the Body Ca- thexis Scale. No significant difference was found between the two groups, 2 (51)=l.24, NS. Similarly, no significant difference was found on the Physical Self subscale of the Tennessee Self-Concept Scale (TSCS—PS)(Mé51.88, §2?8.49 for homosexual subjects and Mé48.79, §Q?7.62 for heterosexual subjects), 2 (51)=l.39, NS. Hypothesis 4 There is a significant difference between homosexual and heterosexual men in the relative degree of impor- tance placed on the physical attractiveness of the sexual object. Descriptive statistics (means, variances, standard deviations, minimum and maximum scores) for questionnaire items assessing emphasis on physical attractiveness of the sexual object (subsequently referred to as "Physical Empha- sis") for homosexual and heterosexual groups are listed in 2Osgood et al. (1957) discuss the possible appropriateness of nonparametric tests for differences in central tendency of 2 scores. Mann-Whitney U tests and Wilcoxon's matched- pairs signed-ranks tests were computed as appropriate, and in every case the outcomes of hypothesis tests in this study were identical to those in which 2_was used. The latter results are reported here given that there is no reason to expect violations of assumptions underlying para- metric tests. 79 Tables 3 and 4. Insert Table 3 about here Insert Table 4 about here No significant differences were found between the two groups on any of the Physical Emphasis measures. Although gay subjects listed significantly more physical characteris- tics when asked to specify attributs most desired in a lover or romantic partner, 2 (37.31)=2.l6, p< .05, they also list- ed more nonphysical characteristics (NS). Results of sta— tistical analyses applied to Physical Emphasis measures in connection with Hypothesis 4 are listed in Table 5. Insert Table 5 about here Hypothesis 5 For homosexual males there is a positive correlation between degree of body dissatisfaction and the rela- tive degree of importance placed on the physical at- tractiveness of the sexual object. Results of two-tailed tests of significance revealed no significant correlations among Body Satisfaction and Phy- sical Emphasis variables for the gay subjects. The only OH NH XME N mm.H :flE om em.m mm.o mm.m om.v .um> mm.h ma.m m~.h mm.~ SI moaumfiuouomumnu Homosow mo mcflxoom mowumfluouomuonu Aboammnm mo moflxnmm mowumwuouomumno Hmuocou wo mocosvoum mowumwuouooumnu Havananm mo mocosvoum mne puma mlw uumm file “Hum ale puma 80 umuoe OQ‘ N m Nocmowoum onEmm emumuu mowumfiuouomumso amowmmsmcoc oznm ooumfla mowumfiuouomumno Hmowmmsm mo HH03 mm Hmuocwwua omumuu mowumfluouomumco Hmowmmnm ozuo «um puma gowumofiwwmmmau uncommom cowuoom ouamccowumoso muoonosm mow uow mfimozmfim aboammom mcammommd mEouH ouflmccoflumoso Hem monumwumum o>wumfluomoo .m OHQMB 81 m v mw.a ma.~ om.h mowumfluouoouosu Hmuocow mo usaxcmm mnv puma OH O hm.m mm.HH mm.m moaumfluouomuosu HmOHmwnm mo mcflxcom mla uumm ca m mm.H mm.m mm.m mowumfluouomumno Hmuocou mo accosvoum «1v uumm m o mm.a om.H mh.a monumfluouomumcu Hmowmwom mo accosvmum «1v puma me CHE mm .Hm> m Aooam> mcammfls He mm Hmuoe coumfla o moflumfluouomumno HMOfimhnmcoc ozum ooumwa moaumfluouomumco om Hmoflmanm mm Hams mm Houocouua m cwumwa moflumfluouomumno HMOfim>Am ozuo 41m uumm mocmnmmum mHmEmm coauMUAMHmmmau sowuoom uncommom muwmccofiumooo muoonozm Hmsxowoumuom How mammcmEm Hm0flm>om mcfimmommd mEouH ouwmccoflumoso uom mofiumwumum o>flumwuomoo .a magma 82 Table 5. Tests for Between-Group Differences in Means for Physical Emphasis Measures Physical Gay Subjects Heterosexual Subjects Emphasis Measure M 22 M §2 22 2 PGFRQRTO 0.46 0.41 0.30 0.27 50 1.68* PGDIFRNK -1.71 2.51 -1.64 3.36 50 0.08 FRQXRNK -0.44 2.21 0.03 0.17 50 -1.72* *p_< .10 (NS) , 83 correlation approaching significance was between the Physi- cal Emphasis measure FRQXRNK and Body Cathexis (2;.37, p< .10), in direct opposition to Hypothesis 1. The other correlations were consistently positive, ranging from .12 to .29. These results are summarized in Table 6. Insert Table 6 about here Hypothesis 6 For heterosexual males there is a positive correlation between degree of body dissatisfaction and the rela- tive degree of importance placed on the physical at- tractiveness of the sexual object. Results of two-tailed tests revealed no significant correlations among Body Satisfaction and Physical Emphasis variables for heterosexual subjects. Correlations were near zero or moderately positive. No probabilities exceeded the .2 level. These results are summarized in Table 7. Insert Table 7 about here Hypothesis 7 There is a significant difference between male homosex- ual and male heterosexual subjects in their degree of present identification with the primary male caregiver in childhood (father, stepfather, etc.) The semantic distances between concepts MYSELF AS I AM and MY FATHER were computed for each subject. Although this 84 Table 6. Correlation Coefficients Among Body Satisfaction and Physical Emphasis Variables for Sample of Gay Males Body Satisfaction Variable Physical Emphasis Variablea PGFRQRTO PGDIFRNK FRQXRNK Body Cathexis Scale .29 .15 .37 Tennessee Self-Concept .23 .12 .24 Scale (Physical Self) a"Physical Emphasis"refers to the degree of emphasis an indi- vidual places on the physical attractiveness of the object. 1: p< .10. 85 Table 7. Correlation Coefficients Among Body Satisfaction and Physical Emphasis Variables for Sample of Heterosexual Males Body Satisfaction Variable Physical Emphasis Variablea PGFRQRTO PGDIFRNK FRQXRNK Body Cathexis Scale -.05 .00 .21 Tennessee Self-Concept .14 -.05 .01 Scale (Physical Self) Note. None of the above correlations were significant. a"Physical Emphasis" refers to the degree of emphasis an individual places on the physical attractiveness of the object. 86 value was smaller for heterosexual (2é13.87, £2?6.81) than for homosexual subjects (2é17.66, 22=9.62), this difference did not quite reach significance, 2 (50)=-l.66, p< .10. On the other hand, a significant difference was ob- tained between the two groups on the semantic distance mea- sure 2 (MYSELF AS I MOST WOULD LIKE TO BE, MY FATHER). The difference between means of homosexual (2é18.80, §2é9.80) and heterosexual subjects (2;13.20, 22é7.26) is evidence that a closer correspondence exists between the two concepts for heterosexual than for homosexual subjects, 2 (50)=2.36, p< .05. No significant differences between groups were found for the corresponding obtained 2 values involving the concept MY MOTHER. Hypothesis 8 There is a negative correlation between the degree of present identification with the primary male caregiver in childhood and the degree of discrepancy between ac- tual self-image and ideal self for male homosexuals. For each of the gay subjects, Osgood's 2 index was cal- culated to determine the "distance" between (a) MYSELF AS I AM and MYSELF AS I MOST WOULD LIKE TO BE and (b) MYSELF AS I AM and MY FATHER. A moderately high positive correlation was found between the two 2 measures (2;.51, pwuoommou .asowuma>oo oumpcoum can mosam> some ho .momonucoumm ca .om30HH0m mum w>onm ooumua mummocou .ouoz Aua.ao lam.ao Aa~.~u. .au.a. um on uaouo uuuu loo.av Aoa.ao Am~.ao Aaa.ao ...zuum Ama.ao ouooz .Ha.oav .aa.oao Aaa.a. .am.ao ...u>a: Aum.av Aaa.au. eao: au>ou :< H :ous Aam.anxaa.av :4: uoa H a: aoua auue u a: ueuz auua Ana.m1. uuuzoa a auua»: naua may :4: uaa »: aqua»: 2a: 4 lo: »: z<:os »Osouoa Aa~.ao Ian.av Aaa.u~. Aa~.ea. um on uaouo uauu Aao.ev ...zuuu ouooz Aem.o~. Ama.ev Ama.av ...u>a: lea.mo Aum.av eao: au>ou Asa.au. lav.ao lam.ouo u :ouz Au~.mu. :aa.a. Ama.ouo H a: eoua :4 H a: Aaa.au. mums aeus 2a: qua mums Aao.a. aqua»: naua uue aqua»: 2::03 lo: »: za: a uauzoa a uaa »: za: o>uumoum>u a a a a a a m N a scamsoeflo o9~m> coo: co ooawm umooooo mo xsmm Hmcuouo MUOOnosm how How . cowacoEHo mo HmwucouOMMHo vaucmfiom mop u0m aumoosou vcoad amwnasoHumHom .OH.OHQMB -_‘n. A 97 .»Ho>Huoomaou .acoflumu>oo oumocnua can mosam> some wo .aoaooucoumm CH .oosoHHOM mum o>onm ooumHH mumoozoo .ouoz Ham.mo “Hm.mo .ae.a. Aaa.m. Aaa.mo um OH uaoqo Ama.a. loo.a. uqu ...zuuu au>oq Amm.ao oqooz ...u>a: aoua :u eao: H :ous uaua H a« H a: ueHz uue aqua»: aqua»: 2a: a Immzom d z<203 Z¢E Ama.ao Hm~.oo lam.ao .em.e. lam.eo Aaq.ao Ama.m. .om.. :4: qua .mm.eo .am.e. auue mama no: »: 1aa »: »uH>Huo¢ . A.U.ucoov OH OHQMB 98 Poten means 22 Cy I ) 20 18 . A POWERFUL MAN 16 14 1 MYSELF AS I MOST WOULD LIKE 2 TO BE 10 . MAN THE PERFECT LOVER . 8 . MY FATHER . MYSELF AS I AM 6 0 ~ A MAN WITH WHOM I HAVE RECENTLY BEEN 4 ROMANTICALLY CLOSE 2 Evaluation (means) 0 2 4 6 8 10 12 14 16 18 20 22 _ 2 MY MOTHER . - 4 WOMAN . Figure 1. Relationships Among Semantic Differential Concepts in Two Dimensions (Evaluation and Potency) for Gay Subjects 99 study in connection with the first of these, Lovers: (1) On all factors, Evaluation, Activity, and Potency, MYSELF AS I MOST WOULD LIKE TO BE and THE PERFECT LOVER were almost identical in their associated mean values. These two concepts, in fact, were closer in the three-dimensional se- mantic space than either alone was to any other concept. Concepts included in the second category, 22222, clus- tered in groups defined by (a) whether the male specified by the concept was gay, and (b) whether the concept specified and actual or ideal entity. (1) MYSELF AS I AM, A MAN WITH WHOM I HAVE RECENTLY BEEN ROMANTICALLY CLOSE, THE PERFECT LOVER, and MYSELF AS I MOST WOULD LIKE TO BE (the four "gay” concepts) each had significantly higher values on the Evaluative dimension than MAN and MY FATHER (p< .05 for the comparisons involving A MAN WITH WHOM I HAVE RECENTLY BEEN ROMANTICALLY CLOSE, otherwise, p< .001). (2) On the Potency dimension, however, the following ordering of variables appeared: A MAN WITH WHOM I HAVE RE- CENTLY BEEN ROMANTICALLY CLOSE (gay and actual) < MYSELF AS I AM (gay and actual)<:MY FATHER<=MAN‘:THE PERFECT LOVER (gay and ideal) < MYSELF AS I MOST WOULD LIKE TO BE (gay and ideal) (no two adjacent variables were significantly differ- ent from each other). For MYSELF AS I AM in relation to Mothers and Fathers, the following was found: 100 (l) The semantic distance between MYSELF AS I AM and MY MOTHER was not significantly different from that between MYSELF AS I AM and MY FATHER (2:15.49, 22:7.23 and 5517.66, §2é9.62; respectively), 2 (22)=-.90, NS. (2) Similarly, 2 (MYSELF AS I MOST WOULD LIKE TO BE, MY MOTHER) was not significantly different from 2 (MYSELF AS I MOST WOULD LIKE TO BE, MY FATHER). Mean values were 18.82 (§Q?8.28) and 18.80 (§2=9.80), respectively; p (22): 0.08, NS. Finally, for Men and Women: (1) The distance between MYSELF AS I AM and WOMAN (2% 18.15, §2é10.07) was not significantly different from MY- SELF AS I AM and either MAN (2%16.00, 22;9.96) or A POWERFUL MAN (§é17.01, §2?9-03): E (22)=0.81 and E (22)=o.24, respec- tively; NS. (2) The distance between MYSELF AS I MOST WOULD LIKE TO BE and WOMAN (2%20.49, §2é11.07) was not significantly dif- ferent from MYSELF AS I MOST WOULD LIKE TO BE and either MAN (§é17.53, 22:11.29) or A POWERFUL MAN (2%14.48, 22:9.61). 2 (22)=0.94 and 2 (22)=1.81, respectively; NS. B. Correlations Among "Self" 2nd "Parent" Semantic Differential Concepts and the Psychometric Tests The variables most relevant to the present investiga- tion include those which assess the quality of an indivi- dual's relationship with his parents and aspects of self- image including gender-role and body satisfaction as well as overall self-esteem. Further examination of the 101 relationships among these variables, for gay men, yielded several interesting findings. First, one's sense of per- sonal powerfulness, as defined by Potency ratings of MYSELF AS I AM, was found to be significantly related to both gen- der-role variables and body satisfaction variables, as well as overall self-esteem. The following measures were significantly correlated with Potency measures of MYSELF AS I AM: (A) TSCS-PS (2;.40, p< .05); (B) BC (£=.59, p< .002); (C) RSE (£=.46, p<.05); (D) PAQ-M (£é.69, p< .001); and (E) PAQ-MF (2?.45, p< .05). The negative correlation between MY FATHER and BC was also significant (Es-.42, p< .05). Corresponding correlations were not significant for heterosexual subjects except for the correlation between MYSELF AS I AM and PAQ—MF (noted below). The remaining corresponding correlations for he- terosexual men were as follows: (A) MYSELF AS I AM and TSCS (5?.23, NS); BC (5;.13, NS); RSE (2?.32, NS); and PAQ-M (r=.34, NS); and (B) MY FATHER and BC (Es-.13, NS). Of the between-group differences in these correlations, only that between MYSELF AS I AM and BC approached significance (2é1.88, p< .10). C. Intercorrelations Among the Psychometric Instru- ments Correlation coefficients obtained for all pairs of psy- chometric scales and subscales for gay subjects are listed in Table 11. As expected, high positive intercorrelations were found among self-esteem and body satisfaction measures. 102 Correlations between these measures and PAQ subscales were all positive, unquestionably highest of which were those in- volving the Masculinity subscale. Insert Table 11 about here Additional Findings for Heterosexual Subjects A. Relationships Among Semantic Differential Concepts Consistent patterns among the concepts, when examined by dimension, were extremely rare for the heterosexual group. The only finding of relevance replicated a finding obtained for gay subjects: The concept MY FATHER, on the Evaluative dimension, was usually less than MY MOTHER (21 cases), 2 (28)=3.89, p<1.001. This was also found to be true for gay subjects in 19 cases and the difference between the concepts was significant for that group as well, 2 (22)= 3.21, p< .005. The between-group difference in regard to this discrepancy between mean values of the two concepts was not significant, 2 (50)=0.62, NS. Semantic differential concepts ordered according to mean rank are presented in Table 12 for each of the three dimensions. Insert Table 12 about here 103 Table 11. Intercorrelations Among Self-Concept, Body Satisfaction and Gender-Role Scales for Gay Subjects Psychometric Scale BC RSE PAQ -M PAQ-MP PAQ-F Tennessee Self-Concept .70*** .56‘7 .62** .42* .18 Scale--Physical Self Body Cathexis Scale .61** .80*** .34 .36 Rosenberg Self-Esteem Scale .63*** .53** .32 Personal Attributes Questionnaire--M .52** .21 Personal Attributes Questionnaire--M-F -.22 *p< .05. **p< .01. ***p< .001. 4 o l .wHo>Huoomaou .mcoHHMH>oo osHm> one: :0 women umoocou mo acmm Hmchuo oumccmua com mosHm> some >3 .aoaocucoumm CH .oo3oaaow mum o>oom oouaHH auooocou .ouoz Hoo.ao AmH.aV Rho.MHv AHm.H1v uu OH uaoqo uqu Aaa.mo ...zuuu Aoa.mv oqooz Aaa.ao :Ha.mo la~.ao Ham.-. ...u><: Aa~.aqo Hao: AHm.HHV Hoa.ao .ma.ao Aaa.1o uu>oq H :o:: Asa.ev zu: qua H a« mums Aoa.oqc :u H aa auue Houa :eHz Avq.auo uauzoa a aqua»: uua »: zu: aqua»: no: »: naua uuu zazoz < zazoz »Osouoa Aaa.eo Aaa.av Ama.aqo Amm.mqo um 09 uaoqo xqv.av uqu ...zuuu :Hm.aqo oqoo: Asa.ao Aqm.av ...u>¢: Ha~.ae Hao.mqv uu>oq eao: Ama.mqo Aaq.mqv H :ouz AHo.ao qu.oqo “ma.ao Ham.mv Houa H a: auue :< H a: :HHz AHa.HHo mums AmH.mv za: qua uuua uue aqua»: no: »: aqua»: z<:oz a z<:oz 1Huoon>u a a a a m a m m H :oncoEHo auommosm Hmsxomououom How :oHacoEHo >o HmHucouowmwa 0Hucm5mm on» How mumoocou mGOE¢ mmHnmcoHumHom .NH OHQMB .»Ho>Huoommou .maowumw>oo oumocmua can mosHm> some an .aomooucoumm CH .oo3oHH0m mum o>onm oouaHH aumoocou .ouoz 5 0 l Aaa.vv Am~.ao laa.av Aqm.ae uu OH uaoqo “ma.eo uqu Hma.mv ...zuuu laa.av oqooz Aaa.ao an.mo ...u><: Haa.va Aaa.ac au>oq Hao: :u Aaa.mv H :ous Hma.ma Ham.ea Hma.~o Aa~.V 90mm H m< H m< mmmB mBHB Amm.¢v Amm.NV 2&2 ADE mmmfi Immm WEB mammwz mqmmwz :02 M2 Z4203 d 24203 2&2 Immzom ¢ IUH>HUO¢ .A.U.ucoov NH OHQMB 106 Relationships among semantic differential concepts for heterosexual subjects are spatially represented in the two primary dimensions in Figure 2. Insert Figure 2 about here In a way similar to the treatment of data for gay sub— jects, MYSELF AS I AM and MYSELF AS I MOST WOULD LIKE TO BE were examined in connection with the following classes of concepts: (A) Lovers--A WOMAN WITH WHOM I HAVE RECENTLY BEEN RO- MANTICALLY CLOSE and THE PERFECT LOVER; (B) Females--WOMAN, MY MOTHER, THE PERFECT LOVER, and A WOMAN WITH WHOM I HAVE RECENTLY BEEN ROMANTICALLY CLOSE; (C) Mothers and Fathers—~MY MOTHER and MY FATHER; and (D) Men and Women--MAN and WOMAN. In connection with Lovers, the following is of inter- est (see Figure 2 and Table 12): (1) As it can be seen in Figure 2, the "actual" con- cepts MYSELF AS I AM and A WOMAN WITH WHOM I HAVE RECENTLY BEEN ROMANTICALLY CLOSE were virtually identical to each other in mean value on Evaluation as were the "ideal" con- cepts MYSELF AS I MOST WOULD LIKE TO BE and THE PERFECT LOVER. Concepts representing lovers actual and ideal, how- ever, were significantly different from each other on 107 20 Potency (means) 18 . A POWERFUL MAN 16 14 MYSELF AS 12 I MOST . . WOULD LIKE TO BE MAN 10 MY FATHER 8 ‘I’ MYSELF AS I AM 6 4 2 0 2 4 6 8 10 12 14 16 18 20 22 Evaluation (means) i A WOMAN THE WITH WHOM MY PERFECT ‘2 I HAVE RECENTLY MOTHER LOVER BEEN ROMANTICALLY CLOSE -4 -6 ‘l) WOMAN -8 Figure 2. Relationships Among Semantic Differential Concepts in Two Dimensions (Evaluation and Potency) for Heterosexual Subjects 108 Evaluation, (ideal lover was rated significantly higher than actual lover), 2 (28)=-3.39, p<< .005; as were the two con- cepts, actual and ideal, representing the 2222 (ideal self was rated significantly higher than actual self), 2 (28): -5.87, p< .001. (2) The "actual" concepts were significantly different from each other on Potency (actual lover was rated lower than actual self) as were the "ideal" concepts (ideal lover was rated lower than ideal self), 2 (28)=6.38, p< .001 and 2 (28)==10.04, p< .001, respectively. These differences were substantially large. (3) The two concepts representing lovers actual (A WO- MAN WITH WHOM I HAVE RECENTLY BEEN ROMANTICALLY CLOSE) and ideal (THE PERFECT LOVER) were almost identical to each other on Potency. Concepts actual and ideal representing the self (MYSELF AS I AM, MYSELF AS I MOST WOULD LIKE TO BE), however, were significantly different on this dimen- sion, 2 (28)=-5.45, p< .001. Two observations about concepts representing Females and those representing the self are: (1) A WOMAN WITH WHOM I HAVE RECENTLY BEEN ROMANTICALLY CLOSE, MY MOTHER, and THE PERFECT LOVER form a clear group- ing on Potency, these values being almost equivalent. WOMAN is lower than all three, significantly different from each at the .001 probability level (see Table 12). (2) The ordering of "female" concepts on Evaluation and Activity are identical. On Evaluation, MY MOTHER lies 109 between (a) MYSELF AS I AM and A WOMAN WITH WHOM I HAVE RE- CENTLY BEEN ROMANTICALLY CLOSE toward the lower extreme and (b) MYSELF AS I MOST WOULD LIKE TO BE and THE PERFECT LOVER at the upper extreme. Among all concepts representing "self" and "females," WOMAN is lowest on both Potency and Evaluation. Mothers and Fathers of heterosexual subjects appeared in relation to the "self" concepts as follows: (1) MYSELF AS I AM was found to lie almost exactly mid- way between MY FATHER and MY MOTHER on Evaluation and was significantly different from each in mean value (see Table 12). MYSELF AS I AM was also midway between MY MOTHER and MY FATHER on Potency. The difference between MYSELF AS I AM and MY MOTHER was significant, the difference between MYSELF AS I AM and MY FATHER was not (Table 12). In three-dimen- sional space, the distance between MYSELF AS I AM and MY FATHER was not significantly different from the distance be- tween MYSELF AS I AM and MY MOTHER, (2%13.87, 13.54; respec- tively), t (28)=0.17, NS. (2) 2 (MY MOTHER, MYSELF AS I MOST WOULD LIKE TO BE) was not significantly different from 2 (MY FATHER, MYSELF AS I MOST WOULD LIKE TO BE) (§é16.45, 22:6.73 and §é13.2o, §2=7.26; respectively), 2 (28)=l.60, NS. (3) On Potency, MYSELF AS I MOST WOULD LIKE TO BE was significantly larger in mean value than MY MOTHER, 2 (28)= 9.69, p<'.001, but not significantly different from MY FA- THER, t(28)=1.25, NS. 110 (4) On Evaluation, MYSELF AS I MOST WOULD LIKE TO BE was significantly greater than both MY MOTHER and MY FATHER, 2 (28)=2.17, p< .05 and 2 (28)=7.10, p< .001; respectively. Finally, for Men and Women: (1) The distance between MYSELF AS I AM and MAN (2: 14.05, §Q?Q.52) was not significantly different from the distance between MYSELF AS I AM and WOMAN i§é17.77, 22s7.85), t (28)=-1.95, p<:.10. Similarly, the distance between MY- SELF AS I AM and A POWERFUL MAN (2%18.30, 22:9.97) was not significantly different from the distance between MYSELF AS I AM and WOMAN, 2 (28)=0.22, NS. B. Correlations Among "Self" and "Parent" Semantic DIfferential Concepts and the ngchometric Tests As noted in connection with gay subjects, variables re- lated to quality of an individual's relationship with his parents, gender-role identity, body satisfaction and other aspects of self-concept are those with which this study is most directly concerned. As with gay subjects, therefore, correlations were computed between the Evaluation and Poten- cy ratings of the "self" and "parent" semantic differential concepts, and the four psychometric tests. Whereas all of the significant correlations between the concept MYSELF AS I AM and a given psychometric test involved the Potency dimen- sion for gay subjects, the majority of such significant cor- relations for heterosexual subjects involved the Evaluative dimension. The following results were obtained: Evaluative dimension: Correlations were significant 111 between ratings of MYSELF AS I AM on Evaluation and (A) the Tennessee Self Concept Scale, Physical Self Subscale (TSCS- PS) (2é.64, p<:.001), (B) the Body Cathexis Scale (BC) (2? .44, p<=.02), (C) the Rosenberg Self-Esteem Inventory (RSE) (2é.47, p .01) and (D) the Personal Attributes Question- naire,Femininity Subscale (PAQ-F) (2;.66, p< .001). This pattern of findings suggests that quality of self-concept varies directly with semantic differential ratings of self associated with "goodness." Correlations were significant between Evaluative ratings of MY FATHER and TSCS-PS (2s.57, p<3.01), BC (2?.59, p< .001), RSE (2?.45, p< .01) and PAQ-F (2;.42, p< .05). This pattern of results suggests that qua- lity of self-concept also varies directly with semantic dif- ferential ratings of MY FATHER,which are associated with traits related to goodness. Evaluative ratings of MY MOTHER correlated significantly only with PAQ-F (2é.47, p<:.01). Corresponding correlations for gay subjects were not signi- ficant, except for the one between MYSELF AS I AM and PAQ-F (noted above). The remaining corresponding correlations (nonsignificant) for gay subjects were as follows: (A) MY- SELF As I AM and TSCS-PS (3;.28, NS), BC (rs-21: NS), RSE (2=.11, NS); (B) MYSELF AS I MOST WOULD LIKE TO BE and TSCS- PS (2;.12, NS), BC (5;.19, NS), PAQ-F (ET-49: NS); (C) MY FATHER and BC (2?.07, NS), RSE (2s.14, NS), PAQ-F (r=.26, NS); and MY MOTHER and PAQ-F (2;.12, NS). The between-group difference was significant for the correlations between MY FATHER and BC only (2?2.08, p,< .05). 112 Potency dimension: The only two significant correla— tions involved PAQ-MF. PAQ-MF was found to correlate sig— nificantly with Potency ratings of MYSELF AS I AM (26.47, p< .01) and MYSELF AS I MOST WOULD LIKE TO BE (2?.46, p< .01). One of the corresponding correlations was signifi- cant for gay subjects, between MYSELF AS I AM and PAQ-MP (noted above). The remaining corresponding correlation, be- tween MYSELF AS I MOST WOULD LIKE TO BE and PAQ-MF was not significant (2?.25, NS). No between-group difference in these correlations was significant. C. Intercorrelations Among the Psychometric Instru- ments Correlations obtained for all pairs of psychometric scales and subscales for heterosexual subjects are listed in Table 13. As with heterosexual subjects, high positive intercorrelations were found among self-concept and body- satisfaction measures, a predictable finding. Correlations among self-concept measures and PAQ-M were moderately posi- tive. As noted above, the correlation between Body Cathex- is and PAQ-M was Significantly stronger for heterosexual subjects. Correlations between PAQ-MF and the self-concept scales were virtually nonexistent for heterosexual subjects. Also, heterosexual subjects' scores for PAQ-F were strongly and positively correlated with TSCS-PS, a finding discussed in connection with Hypothesis 12. 113 Insert Table 13 about here Additional Similarities and Differences Between Groups In addition to the preceding findings, there were cer- tain similarities and differences of interest between the homosexual and heterosexual groups for which no specific hy- potheses were stated. A. The Semantic Differential Concepts The five concepts which were significantly different for homosexual and heterosexual groups were, for Evaluation, A POWERFUL MAN, WOMAN, MYSELF AS I AM, and MYSELF AS I MOST WOULD LIKE TO BE. For Potency, THE PERFECT LOVER was signi- ficantly different for the two groups. In every case, the mean value for heterosexuals was lower (p<=.01). No mean value on Activity was significantly different for the two groups. Of between-group differences in 2 values between the pairs of concepts, only 2 (MYSELF As I MOST WOULD LIKE TO BE, MY FATHER) was significant (2 =13.2o, 22s heterosexuals 7.26, and fi'gays=18.ao, 22:9.80), 2 (50)=-2.36, p< .05. B. Body Satisfaction and Personal Sense of Power Correlations among Body Satisfaction variables and the two variables 2 (MYSELF AS I AM, A POWERFUL MAN) and 114 Table 13. Intercorrelations Among Self-Concept, Body Satisfaction and Gender—Role Scales for Heterosexual Subjects Psychometric Scale BC RSE PAQ-M PAQ -MF PAQ-F Tennessee Self-Concept .70*** .63*** .49** -.10 .50** Scale--Physica1 Self Body Cathexis Scale .52** .47** -.01 .36 Rosenberg Self-Esteem Scale .64*** .19 .30 Personal Attributes Questionnaire--M .43* .29 Personal Attributes Questionnaire--M-F -.27 *p< .05 **pg .01 ***23 .001 115 AMP(MYSELF AS I AM, MYSELF AS I MOST WOULD LIKE TO BE)1 were uniformly negative for both groups. Three of the four cor- relations were significant for gay subjects, suggesting that body satisfaction is positively correlated with personal sense of powerfulness for gay men. None of the correlations reached singificance for heterosexual subjects. On the other hand, none of the corresponding correlations between the two groups were significantly different. These results are shown in Table 14. Insert Table 14 about here The correlations between _D_' (MYSELF AS I AM, MYSELF AS I MOST WOULD LIKE TO BE) and the Body Satisfaction variables were also significant only for gay subjects (r=-.40, p‘=.05 for BC and £?-.53, p< .01 for TSCS-PS). Corresponding nonsig- nificant correlations for heterosexual subjects were negli- gible (EI'°O3' NS for BC and 2s-.14 for TSCS-PS, NS). The between-group differences in these correlations were not significant. C. Emphasis on Ppysical Attractiveness of Sexual Ob- ject and Gender Role Identification The overall pattern of correlations among these varia- bles indicates that emphasis on physical attractiveness of 1A2 Prefers to the difference in means between concepts on Potency. 116 Table 14. Correlations Among Body Satisfaction Measures and Measures of Personal Sense of Powerfulness Body Satisfaction Measure Gay Subjects Heterosexual Subjects 0) 3%} 0) a BC TSCS-PS 22 TSCS-PS .394 D -.44* -.25 -.26 -.23 AM. (MYSELF AS I AM, MYSELF AS I MOST WOULD LIKE TO BE) - Potency -.60** -.48* -.36 -.26 *p< .05 **p< .01 117 the sexual object (Physical Emphasis) is most strongly (and positively) related to degree of masculine gender-role iden- tification in both groups. Correlations among Physical Emphasis and Gender-Role Identification variables were significant for the Physical Emphasis variable FRQXRNK and PAQ-M for gay subjects and for FRQXRNK and PAQ—MF for heterosexual subjects (2;.49, p< .01 and £F°51r p< .01, respectively). None of the correlations between the Physical Emphasis variable PGFRQRTO and gender— role variables (PAQ subscales M, M—F, and F) were signifi- cant. None of the correlations between the Physical Empha- sis variable PGDIFRNK were significant except for the cor- relation between PGDIFRNK and PAQ-MF for heterosexual sub- jects (2?.37, pf<.05). A nonsignificant trend was found for PGDIFRNK and PAQ-M for gay subjects (2?.38, p< .10). Of similar interest is the correlation obtained between PGFRQRTO and 2 (MYSELF AS I AM, WOMAN) (2;.54, p<:.01 for gay subjects and EI'44' p< .05 for heterosexual subjects). Corresponding correlations involving 2 (MYSELF AS I AM, MAN) were not significant for either group (2é.37, pg .10 for gay subjects and 2§-.28, p<:.20 for heterosexual subjects). No between-group differences in the correlations were signifi- cant. D. Gender-Role Identification and Actual-Ideal Self DIScrepangy Correlations were negative and significant among PAQ-M and PAQ-MF scales and 2 (MYSELF AS I AM, MYSELF AS I MOST 118 WOULD LIKE TO BE) only for gay subjects, indicating that higher "masculinity" self-ratings on PAQ-M and PAQ-MF were associated with actual self-images which were closer to gay subjects' ideal selves. Correlations involving PAQ—F were nonsignificant for both groups. These findings are listed in Table 15. It might also be noted here that one other as- pect of actual-ideal self discrepancy was significantly re- lated to PAQ-M only for gay subjects. As described previ- ously, this variable is related to one's "personal sense of power," and is defined as the difference between factor scores on Potency for the concepts MYSELF AS I AM and MYSELF AS I MOST WOULD LIKE TO BE. The correlation obtained for gay subjects was —.49 (p< .01). A negative correlation be- tween these two variables implies a positive relationship between sense of powerfulness and masculinity as measured by the Personal Attributes Questionnaire. The correlation for heterosexual subjects was -.31, and approached significance at the .10 level. None of the correlations reported in Table 15 were significantly different between groups. Insert Table 15 about here 119 Table 15. Correlations Between Personal Attributes Questionnaire and Actual Self--Ideal Self Discrepancy 2 (MYSELF AS I AM, MYSELF AS I MOST WOULD LIKE TO BE) :figscale Gay Subjects Heterosexual Subjects PAQ-M -.50* -.27 PAQ-MP -.41* .01 PAQ—F -.31 -.21 *p< .05 CHAPTER 4 DISCUSSION Discussion of Hypothesis Tests Hypothesis 1 was supported by findings obtained. There was a modest but highly significant difference between groups on the Rosenberg Self-Esteem Scale. While both groups displayed generally positive overall self-concepts, gay men Showed better overall self-esteem. As has been pointed out, various findings have been reported in exist- ing literature on this issue. It is probably most often found that there are no differences in global self-esteem between homosexual and heterosexual men. It might be point- ed out that all of the gay men involved in this study were to some extent open about their homosexuality to others. Some, in fact, were completely open with family and friends about their sexual orientation and were actively involved in gay organizations. It is possible that on the way to de- veloping this degree of self-acceptance, in spite of keen Opposition from a nonaccepting society, many gay men emerge with self-concepts significantly more positive than average. Hypothesis 2, on the other hand, was not supported by the data. There was no significant difference between groups in the overall discrepancy between self-image and perceived ideal self. Findings in connection with Hypothe- ses 1 and 2 precisely replicate Skrapec and MacKenzie's (1981) findings, cited above. In that study, although gay 120 121 men rated themselves more positively on direct measures of self-regard, on indirect measures consisting of adjective rating scales for ideal vs. actual self, this difference was 222 obtained. It seems that although gay men conceptualize themselves as no more or less distant from their ideal selves that do heterosexual men, they describe themselves in more positive terms. Given backgrounds of discrimination and varying degrees of overt oppression, gay men may find themselves more frequently compelled to reaffirm their posi- tive self-valuation and pride in sexual identity. This may account for gay males' more positive self-ratings on self— esteem scales. This phenomenon may not be as discernible on measures which do not directly invite evaluations of self-worth. This may be the case since, as it has been noted, heterosexual men rated themselves significantly lower on the Evaluative dimension than homosexual men for both MYSELF AS I AM and MYSELF AS I MOST WOULD LIKE TO BE. Per- haps as self-concept improves, aspirations correspondingly rise so that the actual-ideal self discrepancy tends to re- main constant regardless of sexual orientation. Similarly, there was no significant between-group dif- ference on the Body Cathexis Scale or on the Physical Self Subscale of the Tennessee Self-Concept Scale, contrary to Hypothesis 3. The data yield no evidence, therefore, that homosexual and heterosexual men differ in their degree of overall body satisfaction. Hypothesis 4 predicts a significant difference between 122 homosexual and heterosexual groups in "Physical Emphasis," defined as the degree of importance placed on physical at— tractiveness of sexual object in relation to the degree of emphasis on nonphysical features of the sexual object. The two groups were found to emphasize physical features to a similar extent in comparison to nonphysical features. Hypo- thesis 4 was not confirmed. Hypothesis 5 was not supported by obtained correlations among Physical Emphasis variables (emphasis on physical at- tractiveness of sexual object) and measures of Body Satis- faction (satisfaction with one's own body). These correla- tions were without exception positive for gay subjects, con- trary to the idea that emphasis on physical features of the sexual object varies in direct proportion to a gay man's own body dissatisfaction. This idea has been expounded by proponents of the Completion Hypothesis of male homosexuali- ty. Kaplan (1967), for example, describes a gay man's em- phasis on the physical attractiveness of the sexual object as often a manifestation of the gay man's "search for the ego-ideal." Given the present findings, it is not unlikely that the more an individual gay male felt satisfied with his own bo- dy, the more emphasis he was likely to place on physical features in a romantic partner or lover. It must be empha- sized, however, since the correlations did not quite reach significance, that support for the hypothesis that Body Sa- tisfaction and Physical Emphasis are positively correlated 123 for gay men is equivocal. Hypothesis 6 states that for heterosexual subjects there is a positive correlation between degree of body dis— satisfaction and degree of emphasis placed on the physical attractiveness of the sexual object. This hypothesis was not supported. The obtained correlations were even weaker for heterosexual subjects than for gay subjects, were non- significant and generally near zero in both negative and positive directions. Hypothesis 7 predicts a significant difference between homosexual and heterosexual men in their degree of identifi- cation with the primary male caregiver in childhood. With identification measured as the degree of similarity between one's father and actual self, this hypothesis was not con- firmed. Interestingly, it was found that heterosexual men conceptualized their fathers as significantly closer to their 22222 selves than homosexual men. These results strongly suggest that although homosexual men see themselves as no more or less like their fathers than heterosexual men, they aspire less to be like their fathers than do their heterosexual counterparts. Neither subjects' perceived similarity to their mothers, however, nor aspirations to be like their mothers had any relation- ship to sexual orientation. These findings are consistent with those reported in prior investigations (Chang and Block, 1960; and Bell, Weinberg and Hammersmith, 1981, cited above). Chang and 124 Block found a greater similarity between descriptions of father and 22222 self for heterosexual than for homosexual men. Furthermore, Bell et a1. emphasize that identification with the opposite-sex parent appeared to have had 22 signi- ficant impact on development of adult sexual orientation, whereas they imply an indirect connection between sexual orientation and identification with the 2222-sex parent. Hypothesis 8 was clearly supported by the data in this study. It was found that the stronger the identification with the primary male caregiver in childhood, the smaller the discrepancy between self-image and ideal self for gay males. It is frequently argued, in connection with the Com- pletion Hypothesis of male homosexuality, that a stronger identification with one's father leads to a smaller discre- pancy between actual and ideal self (for both gay and he- terosexual men). For gay men, this was supported by data obtained in the present study. It seems, therefore, that identification with one's father contributes to greater self-satisfaction among gay men. It has already been noted, however, that certain beha- vioral manifestations frequently assumed to follow from the Completion Hypothesis were 222 shown in this study to hold. For example, it was 222 found that gay men overemphasize the physical attractiveness of the sexual object as a function of dissatisfaction with their own bodies. The Completion Hypothesis is also the theoretical basis for predictions that gay men have a greater self-image—-ideal self 125 discrepancy than heterosexual men. This was found 222 to be the case. Although the finding obtained in connection with Hypothesis 8 is not inconsistent with the Completion Hypothesis, it is not in and of itself a sufficient condi- tion of the Completion Hypothesis. Hypothesis 9 states that there is a negative correla— tion between the degree of identification with the primary male caregiver in childhood and the degree of discrepancy between actual and ideal self for male heterosexuals. This is the hypothesis for heterosexual subjectS'which: corre- sponds to Hypothesis 8. No correlation was found between identification with the primary male caregiver in childhood and the degree of discrepancy between self-image and ideal self for male heterosexuals. Hypothesis 9, therefore, was not supported. Although the two groups differ in magnitude of the obtained correlations, this difference was not sig- nificant. It can only be concluded, therefore, that the degree to which a 22y man's actual self-image approximates his ideal self is related to perceived similarity to his father. This relationship has not been shown to hold for heterosexual men. It will be recalled that the correlation between male gender identity and the discrepancy between actual and ideal self was found to be substantially negative for gay men (Table 15). Actual self-image, then, is closer to ideal self in gay men who see themselves as similar to their fa- thers and who describe themselves in terms considered 126 masculine (as measured by the Personal Attributes Question- naire). As reported in a preceding section, one other as— pect of self-concept, satisfaction with one's body, was found to be significantly related to paternal traits for both groups. For gay men, satsifaction with their bodies was associated with low Potency ratings of MY FATHER. For heterosexual subjects, body satisfaction was associated with high Evaluation ratings of MY FATHER. As discussed in a subsequent section, traits associated with nurturance or expressiveness may characterize fathers rated relatively high in Evaluation and relatively low in Potency. Such fa- thers, therefore, are likely to contribute to body satisfac- tion in their sons, although the process by which this hap- pens is unknown. The overall pattern of findings for gay men strongly indicates that body satisfaction and similari- ty of actual and ideal self are associated with similarity to one's father and more "masculine" self-ratings. Fathers who are nurturant are more likely objects of their sons' identification. Such an identification with a positive male role model, therefore, appears more conducive to an overall sense of self-satisfaction and body satisfaction. These re— lationships are most demonstrable in connection with gay men. The correlations between measures of masculine gender identity and discrepancy between actual and ideal self were not significant for heterosexual men. The differences be- tween the two groups in magnitude of these correlations, however, did not reach significance. 127 Hypothesis 10 was not supported. There was no Signifi- cant difference between homosexual and heterosexual groups in the degree to which subjects felt a personal sense of powerfulness. This finding is in direct contrast to clini- cal literature which emphasizes feelings of impotence or ef- feminacy among male homosexual patients. This outcome fails to support the assumptions made in psychoanalytic theories in which the Completion Hypothesis in some form is implicit. As has been discussed, one assumption underlying these theories is that homosexuality in males is motivated by a sense of deficiency in regard to qualities of strength and potency. These theories view homosexual acts as the means by which to acquire vicariously the experience of powerful- ness. Hypothesis 11 was strongly confirmed by the data in this study. For gay men, correlations among Body Satisfac- tion and Masculine Self-Perception/Gender Role Identifica- tion variables were uniformly positive. Correlations be- tween the PAQ-M subscale and each of the two Body Satisfac- tion variables, however, were very much higher than corre- sponding correlations for the PAQ-MF subscale. One finding pertaining to this hypothesis has yet to be explained. It is not clear why the correlations between 2 (MYSELF AS I AM, MAN) and the two Body Satisfaction varia- bles were negligible if Masculine Self-Perception and Body Satisfaction are positively correlated (Hypothesis 11). Given the relationship of MAN to the other male concepts, 128 A POWERFUL MAN in particular, it is suggested that this con- cept was translated by subjects as "the average man" rather than as "the most masculine man." The concept MAN, there- fore, would not serve as an accurate index of masculinity. The overall pattern of results shown in Table 8 is highly consistent with Hypothesis 11. It is interesting that the correlations between the Body Satisfaction variables and the M—F scale were so much lower than the corresponding correlations for the M scale. Rating scales composing the M and F scales, it will be re- called, have at their upper extremes descriptors which were found to be socially desirable for both sexes; and at their lower extremes, descriptors relatively undesirable for both sexes. In contrast, the rating scales contributing to M-F have traits at their upper extremes which were considered desirable for males only. At their lower extremes the M—F rating scales have qualities judged as desirable for females only. It is hard to determine exactly what set of variables differentiate the M and M-F scales since both comprise at- tributes most typical of or socially acceptable for males. The M-F scale consists of traits such as "never cries," "very little need for security," "not easily hurt," "ag- gressive," and "very dominant." The M scale is composed of traits such as "independent," "active," "competitive," "self-confident," and "stands up well under pressure." The M-F scale is composed, therefore, of traits which 129 collectively might be described as self-serving and impas— sive; whereas the traits which contribute to the M scale might best be described as self-reliant and instrumental. The M-F rating scales may be considered typically masculine, but actual social desirability is probably less than for the M scale. Although the upper extremes of both scales are composed of traits regarded as most typical of males, Body Cathexis seems most related to traits which are both masculine 222 unquestionably positive. The correlation coefficients are clearly suppressed in magnitude when involving masculine characteristics which have to some extent a negative ele- ment. Nevertheless, the results Clearly demonstrate that a positive correlation exists between masculine self-percep- tion and body satisfaction for gay males. The corresponding prediction for heterosexual males is stated in Hypothesis 12. As was the case for gay males, coefficients between Body Satsifaction variables and M-F were much lower than corresponding coefficients for the M scale. Since all corresponding correlations were much lower for heterosexual subjects, those involving the M-F scale were virtually nonexistent. In addition, for the TSCS-PS, correlations with the F scale were comparable to those for the M scale. The data do support Hypothesis 12, but suggest that, for heterosexual men, Body Satisfaction is positively correlated with socially desirable "feminine" traits as well. This finding is not self-contradictory 130 since, as noted above, Spence et a1. (1975) found a low pos- itive correlation between M and F for both sexes. F was found to be negatively correlated with M-F only. In summary, the data indicate that, for males, Body Ca— thexis is related to socially desirable masculine traits (as measured by the PAQ-M) regardless of sexual orientation. On the other hand, this relationship was found to be signifi- cantly stronger for gay males. This fact implies that body satisfaction is much more strongly associated with, or per- haps even dependent on, masculine gender identification for homosexual than for heterosexual men. The implications of such a finding for psychotherapy have been discussed by Hart (1978). As has been noted in a preceding section, any discrepancy which might be found in self-concept between homosexual and heterosexual groups might result from distortions in personal construct systems defining what is appropriately or desirably masculine. Tra- ditional stereotypic assumptions which associate effeminacy and homosexuality are likely one source of such distortions. It will be recalled that Skrapec and MacKenzie (1981) found in their study that homosexual men described themselves in the most rigidly stereotypic fashion with reference to mas- culinity. They attempted to explain their findings as a manifestation of "compensatory masculine responding, where masculinity is defensively exaggerated in the face of gender role 'threat'..." (p. 368). As described above, Hart's sug- gestion is that in treating a gay male in psychotherapy, the 131 therapist be sensitive to such implicit distortions. Results offer partial confirmation of Hypothesis 13, which predicts a difference between groups in degree of mas- culine gender-role identification. Although statistically significant, the between-group difference on PAQ-MP was not very large, and no significant difference was found on PAQ- M. The between-group difference found on the F scale was comparable to that found on the M-F scale, but in the op- posite direction. That is, gay males were significantly more characterized by socially desirable traits defined by the F scale. The overall pattern of results indicates that while the two groups of men are comparable in the extent to which they ara characterized by clearly desirable masculine traits, heterosexual men have fewer desirable traits typi- cally associated with women. On the other hand, gay men possess fewer traits judged as specific to males but which are probably less desirably socially. Discussion of Additional Findings for Gay Subjects A. Relationships Among Semantic Differential Conquts On the semantic differential Evaluative factor, it was found that for gay subjects the concept MYSELF AS I AM was significantly greater than MY FATHER and not significantly different from MY MOTHER. Heterosexual subjects differed in that they rated MYSELF AS I AM significantly lower than MY MOTHER, discussed subsequently. Gay men also rated them- selves as significantly less like their fathers on the 132 Evaluative dimension than did heterosexual men. The scales which contributed to the Evaluative factor were loyal-dis— loyal, good-bad, generous-stingy, beautiful-ugly, compas- sionate-—hard-hearted, nurturing-depriving, dirty-clean, na- tural-pretentious, and harsh—gentle. With respect to these traits, then, gay subjects were more like their mothers. As noted below, overall, gay subjects were found to be as much like their fathers as like their mothers. It might also be noted that on Evaluation, MAN was significantly less than WOMAN and MY FATHER was significant- ly less than MY MOTHER for both groups; so the dimension was composed of scales also associated with masculinity and femininity. Gay men, then, who rated themselves as higher on the Femininity subscale of the Personal Attributes Ques- tionnaire, would expectably rate themselves as more like their mothers than like their fathers on the Evaluative factor. On both dimensions, THE PERFECT LOVER lies between MY— SELF AS I MOST WOULD LIKE TO BE and MYSELF AS I AM, and the first two of these are almost identical in value. The dif- ference between MYSELF AS I AM and THE PERFECT LOVER is sig- nificant at the .01 level on Evaluation, but nonsignificant on Potency. This latter finding seems not to support psy- choanalytic theories which incorporate to any extent the Completion Hypothesis, since many of these theories would predict a discrepancy especially on the Potency dimension. Several of the subjects, in fact, in describing the ideal 133 lover or romantic partner, wrote that they did not want a lover much stronger, taller or better looking than they, in direct contrast to the Completion Hypothesis. It is possi- ble, however, that the ideal lover of erotic fantasies may be different from the ideal lover hoped for in real life. The latter, according to the hypothesis, might not be some- one whose obvious superiority on any dimension would pose a threat to an individual's own positive self-evaluation. The concept THE PERFECT LOVER might, in fact, be a combination of the two, emerging higher in Potency than MYSELF AS I AM, but not significantly so. These speculations are consistent with findings previously cited of Skrapec and MacKenzie (1981) and Hart (1978). In other words, traits in others typically more associated with femininity and/or considered desirable for everyone may be less threatening to one's mas- culine self-concept than those associated with stereotypic masculinity. Perhaps it is in this context that we can also under- stand findings obtained in connection with Hypothesis 5. In connection with that hypothesis, a trend toward a positive relationship between Body Satisfaction and Emphasis on Phy- sical Attractiveness of the sexual object was found for gay men, contrary to expectations following from the Completion Hypothesis. Since it was found in the present study that body satisfaction is related to sense of masculinity, par- ticularly for gay men, then gay men who see themselves as more masculine and are more comfortable with their bodies 134 may also be more comfortable in expressing an emphatic sexu- al interest and enjoyment in connection with physically at- tractive male others. It was also found that for all of the semantic dif- ferential dimensions combined into a three-dimensional se- amntic Space, MYSELF AS I MOST WOULD LIKE TO BE and THE PERFECT LOVER were significantly closer spatially to each other than either was to any other concept. The application of this finding to the preceding theoretical discussions is straightforward. What appears to be the case is that the hypothetical ideal lover is very close to the ideal self. What these men desired in partners was not a set of traits highly disparate from what they would value in themselves. The gay man's ideal lover, therefore, possesses features not experienced as similar to what he actually 22, but, rather, what he would like to be. Similar, although not identical, findings were obtained for heterosexual subjects, and are discussed below. Gay subjects also saw themselves as no more different from one parent than from the other. The actual three-di- mensional relationships may be inferred readily from Figure l where MYSELF AS I AM lies roughly equidistant from MY MO- THER and MY FATHER. They were more like MY MOTHER on the Evaluative dimension, however, and more like MY FATHER on the Potency dimension. What is indicated, then, is not that subjects saw themselves as representing an ”average" of features defining male and female parents, but that they 135 saw themselves as having the most desired features of each parent. Finally, these findings are precisely analogous to those obtained in connection with MAN and WOMAN. MYSELF AS I AM was not significantly more or less distant conceptual- ly from MAN than from WOMAN, but was significantly closer to both WOMAN and MAN than was MAN to WOMAN. In addition, MYSELF AS I AM more closely approximated WOMAN on Evalua- tion, but was more similar to the concept MAN on Potency. Gay subjects, then, did not just conceptually locate them- selves between the average man and woman, but saw themselves as possessing the most desired features typical of each. The ordering of concepts on the Activity dimension is less readily interpretable. There is only one significant difference between adjacent concepts on this dimension, be- tween A MAN WITH WHOM I HAVE RECENTLY BEEN ROMANTICALLY CLOSE and MYSELF AS I MOST WOULD LIKE TO BE. There is no immediately apparent way to interpret the clusterings list- ed in the bottom panel of Table 10. The Activity dimension does not straightforwardly translate "active-passive," but also seems to incorporate a heavy component of emotional expressiveness (e.g., "emotional-unemotional"). Because there are so many ways to be active or passive, emotional or unemotional, there are no clear grouping by sex or sexual orientation for the concepts. The similarity between MYSELF AS I MOST WOULD LIKE TO BE and THE PERFECT LOVER, however, is further emphasized by superimposing the Activity 136 dimension on Figure l. Across dimensions, these two con- cepts form the most consistent pairing. B. Correlations Among "Self" and "Parent" Semantic Differential Concepts and the Psychometric Tests On the Potency dimension, significant correlations were found between the concept MYSELF AS I AM on the one hand, and self-esteem, body satisfaction, and masculine-scored PAQ subscales on the other. Findings for gay men clearly indi- cate that body satisfaction, overall self-esteem, and mascu- line gender identity are positively related to personal sense of potency. Only one of these correlations, the cor- relation between MYSELF AS I AM and PAQ-MF, was significant for heterosexual subects. The difference between groups for one of these correlations approached significance, viz., the correlation between Body Cathexis and Potency ratings of MYSELF AS I AM. For gay subjects, there was a moderately negative cor- relation between Potency ratings of MY FATHER and the Body Cathexis Scale. This unexpected finding suggests a differ- ent way'to understand Bieber's controversial descriptions of gay men and their fathers. As noted above, Peretti et al. (1976) found that only a subgroup of male homosexuals had particularly negative self-concepts. These men were also observed to have had seriously defective relationships with their fathers. Bell et al. (1981) discuss the relatively negative perceptions their gay respondents had of their fa- thers as developing in reaction to the fathers' negative 137 attitudes toward them. Because Bieber's subjects were all in psychoanalytic treatment and therefore were possibly more likely to have negative self concepts, it is not surprising that these men were found to have had defective relation- ships with their fathers. Further research is needed to determine what paternal characteristics associated with potency may contribute to diminished body satisfaction in gay sons. Given the likely relationship of the Potency dimension to masculinity, and given that some traits generally characterized as masculine are also self-serving and emotionally impassive, one might speculate that fathers possessing these stereotypically male traits are more derogatory in their attitudes toward gay sons. It is of the utmost importance to keep in mind, how- ever, that there is no suggestion that body satisfaction of gay men is less than for heterosexual men, on the basis of these findings. It must also be pointed out, conversely, that the data equally imply that for gay men, the lack of these paternal traits appears to be associated with positive body images. Furthermore, as noted in the corresponding section for heterosexual subjects, similar findings were ob- tained in connection with that group. What is important about what has been found is that body concept was signifi- cantly related to different, but probably related, paternal traits for the two groups; and that only the characteristics of fathers seem to have such importance. 138 Discussion of Additional Findings for Heterosexual Subjects A. Relationships Among Semantic Differential Concepts For heterosexual subjects, MY FATHER was usually less on Evaluation than MY MOTHER, which was also true of an equal proportion (three-fourths) of gay subjects, and is not surprising. The more interesting phenomenon is that this pattern did not hold for exactly one-fourth of subjects in each group. Future research may reveal significant implica— tions of this "deviant" pattern, but there is no reason cur- rently to consider it any more or less adaptive than the more typical situation. An interesting pattern of relationships was obtained among concepts pertaining to "self" and those representing "lovers" (Figure 2). On the Evaluative dimension, the "ac- tual" concepts MYSELF AS I AM and A WOMAN WITH WHOM I HAVE RECENTLY BEEN ROMANTICALLY CLOSE were nearly identical and were rated much lower than the corresponding "ideal" con- cepts MYSELF AS I MOST WOULD LIKE TO BE and THE PERFECT LOVER, which were virtually equivalent. The latter pair of concepts were similarly close in value on Evaluation for gay subjects. Both groups, therefore, have aspirations for themselves which match what they imagine in the ideal lover. The heterosexual subject, on the other hand, differentiated the "self" concepts sharply from "lovers" on Potency, the latter being rated much lower. This differentiation is most likely due to heterosexual males' preference for character- istics in lovers typically associated with femininity. 139 Interestingly, whereas A WOMAN WITH WHOM I HAVE RECENT- LY BEEN ROMANTICALLY CLOSE, MY MOTHER and A PERFECT LOVER are roughly equivalent on Potency, WOMAN is significantly and considerably less than these other three "female" con- cepts. Heterosexual men consistently seem to regard their lovers (actual and ideal) and their mothers as much stronger than the average woman. On all three dimensions, MY MOTHER and A WOMAN WITH WHOM I HAVE RECENTLY BEEN ROMANTICALLY CLOSE were not significantly different. The lovers that he- terosexual men actually choose appear to be very similar to their own mothers. The "perfect" lover, however, is rated significantly higher in Evaluation. The ideal lover, then, is differentiated from other female concepts in her posses- sion of a great many qualities associated with "goodness," almost identical to what is characteristic of MYSELF AS I MOST WOULD LIKE TO BE. MYSELF AS I AM was roughly midway between MY MOTHER and MY FATHER on the two primary dimensions although, as with gay males, the difference between MYSELF AS I AM and MY FA- THER was not significant of Potency. Heterosexual subjects were significantly less than their mothers in Evaluation (although this difference appears small), whereas gay sub- jects were roughly equivalent to their mothers on this di- mension. Like gay subjects, in three-dimensional space, the heterosexual subjects were no more like their fathers than their mothers. On the whole, heterosexual men desired to be slightly 140 more like their fathers than like their mothers. Although it did not quite reach significance, this discrepancy was much greater than the near zero discrepancy found for gay men. The implications of the significant between-group dif- ference for the semantic distance between MYSELF AS I MOST WOULD LIKE TO BE and MY FATHER was discussed in connection with Hypothesis 8. This again emphasizes the fact that if there are between-group differences for parents of homosexu- al and heterosexual men, they probably have more to do with fathers than with mothers. This is a finding, it has been noted, which is supported by similar observations of Bell, Weinberg and Hammersmith (1981). This finding does not jus- tify the attribution of maladaptive qualties to mothers of homosexual men as has been traditionally made (e.g., Bieber et al., 1962). It was also found that overall semantic dif- ferential ratings of MY MOTHER were not significantly dif— ferent between groups on any of the three dimensions. Homo- sexual and heterosexual men, therefore, see their mothers in very similar terms. Finally, like gay subjects, heterosexual men saw them- selves as more like WOMAN on Evaluation and more like MAN on Potency. Overall, as it has been pointed out, heterosex- ual men saw themselves as no more distant overall from WOMAN than from MAN. Like gay subjects, they seem to see them- selves as possessing the most desirable qualities of each SEX . 141 B. Correlations Among "Self" and "Parent" Semantic Differential Concepts and the Psychometric Tests For heterosexual subjects, correlations were signifi- cant between each of the self-concept scales TSCS-PS, RSE, and BC on the one hand, and the Evaluation score for MYSELF AS I AM on the other. These findings were significant only for heterosexual subjects. As with gay subjects, MYSELF AS I AM on Evaluation was also positively correlated with PAQ- F. Findings were similar in regard to MYSELF AS I MOST WOULD LIKE TO BE. The correlations which were significant for the heterosexual group suggest that self-concept for this group (global as well as physical self) is positively related to qualities perceived in the self which are asso- ciated with "goodness." These are qualities which, in turn, have been shown to be associated with femininity. No such significant correlations were found for gay subjects, buttme magnitude of obtained correlations was not significantly different between groups. Overall, this pattern was reversed for the two groups on the Potency dimension. MYSELF AS I AM and MYSELF AS I MOST WOULD LIKE TO BB were not significantly related to self-concept measures for heterosexual subjects on Potency. For gay subjects, on the other hand, correlations between the self-concept scales and Potency scores for MYSELF AS I AM were moderately high positive. In addition, the between- group difference for the correlation between Body Cathexis and MYSELF AS I AM on Potency approached significance at the 142 .10 level for gay subjects. Overall, then, it appears that comfort and satisfaction with one's own body as well as with one's overall self is related to possession of positive "masculine" as well as "feminine" gender characteristics in oneself. Given the gender-role differences found between the two groups, however, there does seem to be some evidence that self-concept in heterosexual men is enhanced by gender role traits possessed to a greater extent by gay men. In gay men, on the other hand, it may be enhanced by gender role traits more characteristic of heterosexual men. Overall self-esteem and body satisfaction were also positively related to Evaluation ratings of MY FATHER for heterosexual subjects. As pointed out in connection with gay subjects, it appears that certain traits in their fa- thers are related to variations in body satisfaction for heterosexual men. These paternal characteristics appear to be different for the two groups, but not unrelated. Body satisfaction was positively related to Evaluation scores of MY FATHER for heterosexual subjects and negatively rela- ted to Potency scores of MY FATHER for gay subjects. The difference between groups in magnitude of the former corre- lation was significant, the difference between groups for the latter approached significance at the .10 level. For both groups it is likely that qualities in fathers associated with kindness or nurturance are related to posi- tive body images, although the specific qualities may be different between groups. 143 Discussion of Additional Similarities and Differences Between Groups A. The Semantic Differential Concepts Highly significant differences were obtained between groups for five semantic differential concepts. As expect- ed, THE PERFECT LOVER was significantly lower for hetero- sexual subjects on Potency. This concept was not signifi- cantly different between groups on Evaluation, however. It is not clear why heterosexuals rate A POWERFUL MAN, WOMAN, MYSELF AS I AM and MYSELF AS I MOST WOULD LIKE TO BE as lower on Evaluation. Overall, the instances of such differences among con- cepts on the three factors are few. Furthermore, the dif- ference between groups in the relative positioning of con- cepts on the two primary dimensions was not especially great. For example, excluding the concept A POWERFUL MAN, there was no difference in the relative positioning of male concepts on Evaluation. The two most apparent between- group differences were (a) the difference between mean scores for MYSELF AS I AM and MY FATHER was much greater for gay subjects, and (b) the difference between mean scores for MAN and MY FATHER was considerably greater for hetero- sexual subjects. The first finding is probably due to the fact that gay men characterize themselves significantly more by adjectives associated with "goodness," as suggested by their self-ratings on Evaluation. The second finding is primarily due to the exceptionally low mean rating of MAN 144 on Evaluation for heterosexual subjects. There were likewise many more similarities than differ- ences in relative positioning of concepts on Potency; and where there were differences, they minimal. For heterosexu- al subjects, MYSELF AS I AM is lowest in Potency among the male concepts, and second to lowest for gay subjects. The differences among MYSELF AS I AM, MAN and MY FATHER were not significant on this dimension for either group, supporting the conclusion that overall there is no difference between the two groups in the degree to which they feel a sense of powerfulness. B. Body Satisfaction and Personal Sense of Power It was found that the correlations between measures of body satisfaction and personal sense of power were signifi- cant for the gay subjects. The absolute magnitudes of cor- relations, however, were not found to be significantly dif— ferent for gay and heterosexual subjects. One's personal sense of power was measured by computing (a) the semantic distance between MYSELF AS I AM and A POWERFUL MAN, and (b) the difference in mean Potency scores for the concepts MY- SELF AS I AM and MYSELF AS I MOST WOULD LIKE TO BE. The moderate to high negative correlations found for gay sub- jects support the hypothesis that a gay man's body satisfac- tion increases with his sense of powerfulness. The Poten- cy dimension is composed of traits more likely to be judged as masculine, and a positive relationship was found, for gay men, between the PAQ Masculinity subscale and Body 145 Satisfaction. It is not surprising, therefore, that a gay man's body satisfaction increases with his sense of power- fulness. C. Emphasis on Physical Attractiveness of Sexual Object and Gender-Role Identification The obtained correlations found among emphasis on the physical attractiveness of the sexual object (Physical Em— phasis) and Masculine Gender-Role Identification variables were strongly positive for the two groups. Given the pro- bable crudeness of the former measure, the relatively strong correlations are even higher than expected. Physical Empha- sis was found to be positively associated with PAQ-M for gay subjects, and to an equal degree with PAQ-MF for heterosexu- al subjects. The discrepancy between MYSELF AS I AM and WOMAN was similarly related to Physical Emphasis for both groups. Together, these findings make it clear that empha- sis on the physical attractiveness of the sexual object is related to masculinity, regardless of sexual orientation. It has been suggested by this investigation's findings that differences do not exist between groups primarily in absolute degree of masculinity, but more in the way it is expressed. The M—F scale of the PAQ was judged as sex- specific because the masculine extreme of this bipolar scale consists of traits considered clearly undesirable for women. The upper extremes of scales comprising the M scale on the other hand, are composed of traits socially desirable for both sexes, yet considered more charactristic 146 of men. Items characterizing the M-F scale, it has been mentioned, might best be described as "self-serving and im- passive," in contrast to the "self-reliant and instrumental" quality of traits composing the M scale. It must be emphasized that overall between-group dif- ferences are not great in terms of gender-role identifica- tion. Insofar as they do exist, however, the masculinity of heterosexual men is more likely characterized by "M-F" traits than is that of heterosexual men. This may be one reason that emphasis on physical attractiveness of sexual object may be more strongly related to M for gay males and to M-F for their heterosexual peers. D. Gender-Role Identification and Actual-Ideal Self DiScrepancy Masculine-scored PAQ subscales (M and M-F) were found to correlate significantly and negatively with the discre- pancy between actual and ideal self for gay subjects. These correlations were moderately high, suggesting that mascu- line gender-role identification is positively related to the degree to which actual self approximates ideal self for gay males. This was found to be true for the aspect of self-concept described as "personal sense of powerfulness" as well. That is, for gay subjects, masculine gender iden- tity appeared to be particularly associated with one's ac- tual sense of potency relative to his ideal self. The ex- tent to which these findings are true for heterosexual men is less clear since neither the corresponding correlations 147 for heterosexual men, nor the difference between the two groups in magnitude of these correlations was found to be significant. CHAPTER 5 CONCLUSION Psychoanalytic theories frequently assert or imply that homosexuality in males is motivated by a sense of deficiency created by a faulty or nonexistent identification with a male object. It is assumed that homosexual activity (fanta- Sied and actual) is an unconscious attempt to achieve this identification and vicariously to fill the void experienced. These theories are typically based on clinical cases and frequently emphasize aspects of the self in which these ho- mosexual patients feel especially deficient. These observa- tions, in turn, are generalized to include nonpatient homo- sexual men. It was found that for both gay and heterosexual nonpa- tient males, their idea of the ideal lover was in many ways virtually identical to their descriptions of themselves as they most would like to be. This is consistent with a theory which predicts that, regardless of sexual orienta- tion, people desire lovers who possess traits which corre- spond to their ideal selves. Some psychodynamic theories imply that emphasis on the physical attractiveness of the sexual object, for homosexual men, is related to a sense of dissatisfaction with their own bodies and a desire to ac- quire symbolically such features via the relationship with a male lover. Contrary to this prediction, the data suggest that that as body satisfaction for gay men increased, so 148 149 did emphasis on the sexual object's physical features. It was also found that emphasis on physical attractiveness of the sexual object was, regardless of sexual orientation, po- sitively related to one's sense of masculinity. Consistent with these findings, then, is the fact that for gay men, body satisfaction was found to be positively related to mas- culinity. The relationships among other aspects of self-concept were examined for this sample of heterosexual and gay men, and similarities and differences between groups were noted. In addition, these findings were discussed in connection with perceived similarity to each parent or primary care- giver in childhood to assess the possible relationship of these identifications to relevant aspects of self-concept. It was found that gay men described themselves in some- what more favorable terms than heterosexual men on the Rosenberg Self-Esteem Scale, although the discrepancy be- tween semantic differential ratings of actual and ideal self was comparable for the two groups. Both findings con- tradict theories which assume that self-concepts of homo- sexual men are less positive than self-concepts of hetero— sexual men. These findings also indicate that discrepancy between actual and ideal self is not directly predictable from self- esteem ratings. Gay men rated both actual and ideal self higher on the semantic differential Evaluative factor, so that the actual discrepancy between the two was similar to 150 that for heterosexual men. Similarly, there was no between- group difference in body satisfaction, a finding which fails to support theories which emphasize the especially negative body images alleged to be characteristic of homosexual men. There was no significant differences between groups in overall perceived similarity to one's father, but gay men were found to aspire less to be like their fathers. Gay men also rated themselves as significantly less like their fa- thers on the semantic differential Evaluative factor than did heterosexual men. The two groups were not different in the degree to which they saw themselves as like their mo- thers or in the degree to which they wanted to be like their mothers. It was found that for gay men, the stronger the identification with the father, the more an individual's actual self-image approximated his ideal self; a finding not obtained for heterosexual subjects. This finding may be re- lated to the finding that, for gay men, one's perceived mas- culinity was significantly and positively related to the degree of similarity between perceived actual and ideal self. Interestingly, body satisfaction was found to be re- lated to certain qualities possessed by fathers of subjects in both groups, qualities most likely related to kindness and nurturance. The specific qualities were different for the two groups, however, although why these difference oc- curred is not clear. There was substantial evidence, therefore, that any differences in characteristics of parents that may exist 151 between groups involve fathers rather than mothers. Homo— sexual and heterosexual men seem to conceptualize their mo— thers in very similar terms. This finding contradicts tra- ditional theories which attribute maladaptive qualities to mothers of homosexual men and imply that such characteris- tics play a causal role in the development of homosexual orientation in men. The findings regarding fathers in the study notwithstanding, there were also striking between- group similarities in the overall characteristics of parent- child relationships. For example, on the semantic differen- tial Evaluative dimension, a dimension which is character- ized by traits associated with "goodness," exactly three- fourths of the subjects in each group rated their mothers higher than their fathers. In each group, conversely, exactly one-fourth of the subjects did not show this pat- tern. In addition, the actual mean rating scores for MY FATHER and MY MOTHER were not significantly different be- tween groups on any of the three semantic differential factors. The two groups were comparable in the absolute quan- tity of masculine traits, but gay men were found to express their masculinity somewhat less in ways which were self- serving and/or emotionally impassive. They were also more characterized by traits characteristic of the Femininity subscale of the Personal Attributes Questionnaire. Ironi— cally, there was evidence that aspects of self-concept such as body image are actually favorably influenced by the 152 presence of feminine traits for heterosexual males. In contrast, there was also evidence that masculine gender identity is more important to body satisfaction for gay men than for heterosexual men. Interestingly, overall it ap- pears that self-concept in heterosexual men is enhanced by gender role traits possessed to a greater extent by gay men, and that in gay men it is most enhanced by gender role traits more characteristic of heterosexual men. It might be suggested, based on this finding, that comfort and satisfac- tion with one's body, as well as with his overall self, is related to degree of comfort with possession of positive "masculine" as well as "feminine" gender characteristics. The relationships among psychological androgyny and body sa- tisfaction as well as overall self-esteem and adjustment is an area for which continued research is recommended. In summary, contrary to theories which conclude that negative self-concept is intrinsic to a homosexual orienta- tion, there was no evidence that homosexual men have lower overall self-esteem, poorer body image, or a greater discre- pancy between actual and ideal self. Although findings from this study are not necessarily inconsistent with cli- nical observations frequently reported in the literature, they emphasize the fact that the observations reported in that literature can reliably be generalized to other clini- cal populations only. Certainly clinical observations which emphasize differences between groups contribute to the un- derstanding of difficulties uniquely confronting homosexual 153 men, and should not be dismissed. On the other hand, the usefulness of such data is greatly diminished when under- stood as intrinsic to homosexuality or as causal in its de- velopment. The findings obtained in this study, then, are not inconsistent with the view that homosexual patients are different in certain ways from heterosexual patients, be- cause this study did not employ a patient sample. In fact, knowledge of the areas of self-concept which are differen- tially central in nonpatients for the two groups may be of clinical relevance as well. Each type of data may shed light on the other, but care must be exercised to avoid making direct inferences about one population based on data from another. The findings of this study, then, do not necessarily diminish credibility of clinical observations typically reported, but strongly suggest that these clinical data at least have been misapplied. When confined to clinical ap- plications, the exploration of similarities and differences between clinical groups is enlightening. The misapplica- tions of such findings, on the other hand, lead to conclu- sions the destructiveness of which is hard to overestimate. One useful approach to future research on human sexu- ality would recognize that neither heterosexuality nor ho- mosexuality is a necessarily "either-or" phenomenon. Fur- thermore, a sexual orientation may be different both cau- sally and phenomenologically for different individuals. The complexity of sexual variations probably demands a more 154 idiographic approach to understanding all of the similari- ties and differences among humans in the way our sexuality is experienced and expressed. LIST OF REFERENCES LI ST OF REFERENCES Alpert, P. (1979). Internalized stigma among homosexual men. Dissertation Abstracts International, 39, 3580B. University Microfilms No. 7901729 American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. Bayer, R. (1981). Homosexuality and American psychiatry. New York: Basic Books. Bell, A., Weinberg, M., & Hammersmith, S. (1981). Sexual preference, its development in men and women. Blooming- ton: Indiana University Press. Bem, S. (1974). The measurement of psychological andro- gyny. Journal of Consulting and Clinical Psychology, 42, 155-162. Beyer, W. (1966). Handbook of tables for probability and statistics. Cleveland, OH: The Chemical Rubber Co. Bieber, I., Dain, H.J., Dince, P.R., Drellich, M.G., Grand, H.G., Gundlach, R.H., Kremer, M.W., Rifkin, A.H., Wil- bur, C.B., & Bieber, T.B. (1962). Homosexuality: A psychoanalytic study of male homosexuals. New York: Basic Books. Blank, L., Sugarman, A., & Roosa, L. (1968). Body concern, body image, and nudity. Psychological Reports, 23, 963-968. Burns, R. (1979). The self concept in theory, measurement, development and behavi'or. New York: Longman, Inc. Cardone, S., & Olson, R. (1969). Chlorpromazine and body image: Effects on chronic schizophrenics. Archives of General Psychiatry, 29, 576-582. Centers, R. (1972). The completion hypothesis and the com- pensatory dynamic in intersexual attraction and love. The Journal of Psychology, £2, 111-126. 155 156 Chang, J., & Block, J. (1960). A study of identification in male homosexuals. Journal of Consulting Psychology, 24 (4), 307-310. Derogatis, L. (1976). Psychological assessment of the sexu- al disabilities. In J.K. Meyer, (Ed.), Clinical manage- ment of sexual disorders. Baltimore: Williams & Wilkins. Dickey, B. (1961). Attitudes toward sex roles and feelings of adequacy in homosexual males. Journal of Consulting Psychology, 25 (2), 116-122. Ellis, H. (1928). Studies in the psychology of sex (Vol. 2, 3rd ed.). Philadelphia: F.A. Davis Co. Endler, N. (1961). Changes in meaning during psychotherapy as measured by the semantic differential. Journal of Counseling Psychology, 8, 105-111. Fenichel, O. (1945). The psychoanalytic theory of neurosis. New York: W.W. Norton. Fitts, W. (1965). Tennessee self concept scale (manual). Nashville: Counselor Recordings and Tests. Fraiberg, S. (1961). Homosexual conflicts. In S. Lorand & H. Schneer (Eds.), Adolescents: Psychoanalytic approach to problems and therapy. New York: Paul B. Hoeber, Inc. Freud, S. (1963). Three essays on the theory of sexuality (J. Strachey, Edi). New York: Basic Books. (Original work published 1905) Freud, S. (1963). Analysis of a phobia in a five year-old boy. In S. Freud, The sexual enlightenment of children (P. Rieff, Ed.). New York: Collier Books. (Original work published 1909) Freud, S. (1922). Group psychology and the analysis of the ego. London: Hogarth. Freud, S. (1963). Letter to an American mother. In H.M. Ruitenbeek (Ed.), The problem of homosexuality in modern society. New York: E.P. Dutton. (Original work puB? lished 1935) Gonsiorek, J. (1977). Psychological adjustment and homo- sexuality. JSAS Catalog of Selected Documents in Psy- chology, l (2), 45. (Ms. No. 1478) Gonsiorek, J. (1982). Introduction. In W. Paul, J. Wein— rich, J. Gonsiorek, and M. Hotvedt (Eds.), Homosexuality: Social, psychological and biological issues. Beverly Hills, CA: Sage Publications. 157 Gonsiorek, J. (1982). Results of psychological testing on homosexual populations. In W. Paul, J. Weinrich, J. Gonsiorek, and M. Hotvedt (Eds.), Homosexuality: Social, psychological and biological issues. BeVerly Hills, CA: Sage Publications. Greenberg, J. (1973). A sutdy of the self-esteem and alien- ation of male homosexuals. Journal of Psychology, 83, 137-143. Hart, M. (1979). Gender attributes and adjustment in homo- sexual men. Dissertation Abstracts International, 32, 40313-4032B. (University Microfilms No. 7902500) Hooker, E. (1963). The adjustment of the male overt homo— sexual. In H.M. Ruitenbeek (Ed.), The pgoblem of homo- sexuality in modern society. New York: E.P. Dutton. Hooker, E. (1965). Male homosexuals and their "worlds." In J. Marmor (Ed.), Sexual inversion: The multiple roots of homosexuality. New York: E.P. Dutton. Hooker, E. (1972). Homosexuality. In J. Livingood (Ed.), NIMH Task Force_9n Homosexuality: Final Repgrt and Back- ground Papers. Rockville: NatiOnal Institute of Mental Health. Jourard, S. & Secord, P. (1954). Body-size and body cathex- 13. Journal of Consulting Psychology, 18, 184. Kaplan, E. (1967). Homosexuality: A search for the ego- ideal. Archives of General Psychiatry, 16, 355-358. Kelly, G. (1955). The psychology of personal constructs. New York: W.W. Norton. Kohut, H. (1971). The analysis of the self. New York: International Universities Press. Livingood, J. (Ed.) (1972). NIMH Task Force on Homosexuali- ty: Final Report and Background Papers. Rockville: Na— tional Institute of Mental Health. MacDonald, A. & Games, R. (1974). Some characteristics of those who hold positive and negative attitudes toward homosexuals. Journal of Homosexuality, 1 (1), 9-27. Osgood, C., Suci, J., & Tannenbaum, P. (1957). The mea- surement of meaning. Urbana: University of Illinois Press. ‘* 158 Peretti, P., Bell, R., & Jordan, J. (1976). Self-percep- tions of oedipal and non-oedipal homosexuals. Indian Journal of Psychology, 51 (2), 98-105. Pomeroy, W. (1969). Homosexuality. In R.W. Weltge (Ed.), The same sex. Philadelphia: Pilgrim Press. Prytula, R., Wellford, C. & DeMonbreun, B. (1979). Body self-image and homosexuality. Journal of Clinical Psy- cholog ,‘35 (3), 567-572. Reik, T. (1944). A psychologist looks at love. New York: Farrar & Reinhart. Reik, T. (1957). Of love and lust. New York: Farrar, Strauss, & Cuhady. Rosenberg, M. (1965). Society and the adolescent self image. Princeton: Princeton University Press. Saghir, M., & Robins, E. (1971). Male and female homo- sexuality: Natural history. Comprehensive Psychiatry, 1971, 12 (6), 503-510. Saghir, M., & Robins, E. (1973). Male and female homosexu— ality: A comprehensive investigation. Baltimore: Williams & WiIkins. Sallee, D. (1979). Relationship between sex—role identi- fication and self-concept in a comparison of homosexual and heterosexual males. Dissertation Abstracts Inter- national, 39, 5583B. (University Microfilms No. 7909616) Schur, E. (1972). Sociocultural factors in homosexual be- havior. In J. Livingood (Ed.), NIMH Task Force on Homo- sexuality: FinallReport and Background Papers. Rock- ville: NatiOnaI Institute of Mental Health. Secord, P., & Jourard, S. (1953). The appraisal of body cathexis: Body cathexis and the self. Journal of Con- sulting Psychology, 11, 343-347. Silber, E., & Trippett, J. (1965). Self-esteem: clinical assessment and measurement validation. Psychological Reports, 16, 1017-1071. Skrapec, C.,& MacKenzie, K. (1981). Psychological self- perception in male transsexuals, homosexuals, and hetero- sexuals. Archives of Sexual Behavior, l9 (4), 357-370. Snider, J.G.,& Osgood, C.E. (1969). Semantic differential technique. Chicago: Aldine. 159 Sobel, H. (1976). Adolescent attitudes toward homosexuality in relation to self concept and body satisfaction. Ado- lescence, 11 (43), 443-453. Spence, J., Helmreich, R., & Stapp, J. (1975). Ratings of self and peers on sex-role attributes and their rela- tion to self-esteem and conceptions of masculinity and femininity. Journal of Personal and Social Psychology, 32, 29-39. Spence, J., Helmreich, R. (1978). Masculinity and feminin- i y. Austin: University of Texas Press. Sullivan, H. (1953). The interpersonal theory of psychia- try. New York: W.W. Norton. Tripp, C. (1975). The homosexual matrix. New York: New American Library. Weinberg, J. (1960). A further investigation of body ca- thexis and the self. Journal of Consulting Psychology, 24, 277. White, R., & Watt, N. (1973). The abnormal personality (4th ed.). New York: The Ronald Press. Wylie, R. (1974). The self-concept (Vol. 1). Lincoln: University of Neraska Press. APPENDIX APPENDIX INSTRUMENTS Semantic differential Tennessee Self-Concept Scale Physical Self Subscale Personal Attributes Questionnaire Body Cathexis Scale Rosenberg Self-Esteem Scale Background questionnaire 160 161 TO THE SUBJECT: In order to participate in this investiga- tion, it is necessary that your response be "yes" to all of the following: A. I describe myself as "gay" and have had exclusively same-sex sexual and/or romantic involvement(s) for a period of at least one year. I am not currently receiving psychological counseling, psychotherapy or medication for a mental or emotional or nervous disorder and have not received such for a period of at least six months. I have never been hospitalized for a mental or emotional disorder. I have consented freely to participate in this investi- gation, understand that all materials are to be kept confidential, and that I may discontinue my participa— tion at any time without recrimination. I am also aware that results of the investigation will be provided to me upon request . PLEASE DO NOT ATTACH YOUR NAME OR OTHER IDENTIFYING INFORMA- TION TO THIS SHEET. YOUR ANSWERS ARE TO BE SPOKEN, NOT WRITTEN. (Form administered to gay subjects) 162 TO THE SUBJECT: In order to participate in this investiga- tion, it is necessary that your response be "yes" to all of the following: A. I describe myself as "heterosexual" and have had sexual and/or romantic involvement(s) only with women for a period of at least one year. I am not currently receiving psychological counseling, psychotherapy or medication for a mental or emotional or nervous disorder and have not received such for a period of at least six months. I have never been hospitalized for a mental or emotional disorder. I have consented freely to participate in this investi- gation, understand that all materials are to be kept confidential, and that I may discontinue my participa— tion at any time without recrimination. I am also aware that results of the investigation will be provided to me upon request. PLEASE DO NOT ATTACH YOUR NAME OR OTHER IDENTIFYING INFORMA- TION TO THIS SHEET. YOUR ANSWERS ARE TO BE SPOKEN, NOT WRITTEN. (Form administered to heterosexual subjects) 163 Instructions: The purpose of this part of this study is to measure the meanings of certain things to various people. On each of the pages immediately following, you will find a word or phrase in capital letters. Beneath the word or phrase there will be a set of scales each of which has a de- scriptive word at one end and its opposite at the other. For example: POLITICIANS fair : : : : : : unfair honest : : : : : : dishonest generous : : : : : : stingy For each page of this test, think carefully about the word at the top of the page and what the word (or concept) means to you. Next, look at the first scale beneath the word or concept. If you feel that the word or concept at the top of the page is very closely related to one end of the scale, you should place a check mark as follows: fair X unfair or fair : - X unfair If you feel that the word at the top of the page is quite closely related to one or the other end of the scale (but not extremely), you should place your check mark as follows: honest x dishonest or dishonest X honest If the concept seems only slightly related to one side as opposed to the other side (but is not really neutral), then you should check as follows: 164 N generous : stingy or generous 3 3 : X : : stingy If you consider the concept to be neutral on the scale or if the scale is completely irrelevant to the concept, then you should place your check mark in the middle space: generous : : : X : ° stingy IMPORTANT: (1) Place check marks in the middle of spaces, not on the borders: (Like this) X X (Not like this) (2) Be sure you check every scale for every concept-- do not omit any. (3) Never put more than one check mark on a single scale. Finally, you sometimes may feel as though you've had the same item before on the test. This will not be the case, so do not look back and forth through the items. Do not try to re- member how you checked similar items earlier in the test. Work at a fairly high speed through this test. Do not worry or puzzle over individual items. It is your first impres- sions, the immediate "feelings" about the items and what they mean to you that we want. On the other hand, please do not be careless, because we want your true impressions. IF YOU HAVE ANY QUESTIONS ABOUT THESE INSTRUCTIONS, PLEASE ASK THE EXAMINER. 165 N933. Nine copies of the following set of scales were ad- ministered to each subject with a different concept heading each page. The concepts used in the present study were the following: MYSELF AS I AM, MY MOTHER, MAN, THE PERFECT LOVER, WOMAN, MYSELF AS I MOST WOULD LIKE TO BE, MY FATHER, A MAN WITH WHOM I HAVE RECENTLY BEEN ROMANTICALLY CLOSE (gay subjects), A WOMAN WITH WHOM I HAVE RECENTLY BEEN RO- MANTICALLY CLOSE (heterosexual subjects),and A POWERFUL MAN. The concepts were administered in the sequence listed above. fast brave emotional possessive accessible weak impetuous individualistic needy compassionate bold playful hard beautiful large nurturing compliant warm rational accepting restrained independent dirty competitive calm harsh generous rugged good peaceable distant passive dominant feminine natural loyal rigid 166 O. .0 O. O. O. O. O. O. O. O. O. O. O. O. O. O. O. O. O. C. O. O. O. O. O. O. .0 O. .0 .0 .0 .0 O. O. .0 O. O. O. O. O. O. O. O. O. O. .0 O. O. .0 slow cowardly unemotional nonpossessive aloof strong controlled conforming providing heard-hearted meek serious soft ugly small depriving stubborn cold intuitive rejecting expressive dependent clean cooperative agitated gentle stingy delicate bad hostile close active submissive masculine pretentious disloyal flexible 167 Instructions: Please respond to these items as if you were describing yourself to yourself. Read each item carefully, then select one of the five alternative responses. Do not omit any item. Place a check mark in the appropriate box to the right of each item to indicate whether you think the statement as applied to you is (1) Completely false, (2) Mostly false, (3) Partly false and partly true, (4) Mostly true, or (5) Completely true. Entirely False Mostly False Part False and Part True Mostly Entirely True True 1. I have a healthy body. 2. I am an attractive person. 3. I consider myself a sloppy person. 4. I like to look nice and neat all the time. 5. I am full of aches and pains. 6. I am a sick person. 7. I am neither too fat nor too thin. 8. I like my looks just the way they are. 9. I would like to change some parts of my body. 10. I am neither too tall nor too short. 11. I don't feel as well as I should. 12. l3. 14. 15. 16. 17. 18. I should have more sex appeal. I take good care of myself physi- cally. I try to be careful about my appearance. I often act like I am "all thumbs." I feel good most of the time. I do poorly in sports and games. I am a poor sleeper. (.1 ox CD Entirely False Mostly False Part False and Part True Mostly True Entirely True 169 The items below inquire about what kind of a person you think you are. Each item consists of a pair of character- istics, with the letters A-E in between. For example: Not at all Artistic A....B....C....D....E Very Artistic Each pair describes contradictory characteristics--that is, you cannot be both at the same time, such as very artistic and not at all artistic. The letters form a scale between the two extremes. You are to choose a letter which describes where ypp fall on the scale. For example, if you think you have no artistic abil- ity, you would choose A. If you think you are pretty good, you might choose D. If you are only medium, you might choose C, and so forth. Circle the letter that you choose. 1. Not at all ag- A....B....C....D....E Very aggressive gressive 2. Not at all in- A....B....C....D....E Very independent dependent 3. Not at all emo- A....B....C....D....E Very emotional tional 4. Very submissive A....B....C....D....E Very dominant 5. Not at all ex- A....B....C....D....E Very excitable citable in a in a major cri- major crisis sis 6. Very passive A....B....C....D....E Very active 7. Not at all able A....B....C....D....E Able to devote to devote self self completely completely to to others others 8. Very rough A....B....C....D....E Very gentle 9. Not at all A....B....C....D....E Very helpful to helpful to others others 10. Not at all A....B....C....D....E Very competitive competitive 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. Very home oriented Not at all kind Indifferent to others' approval Feelings not easily hurt Not at all aware of feelings of others Can make de- cisions easily Gives up very easily Never cries Not at all self-confi- dent Feels very inferior Not at all understand- ing of others Very cold in relations with others Very little need for security Goes to pieces under pressure 170 ACOOOBOOOOCOOOODOOOOE AOOOOBOOOOCOOOODOOOOE AOOOOBOOOOCOOOODOOOOE A....B....C....D....E AOOOOBOOOOCOOOODOOOOE AOOOOBOOOOCOOOODOOOOE A....B... .C....D....E AOOOOBOOOOCOOOODOOOOE AOOOOBOOOOCOOOODOOOIE AOOOOBOOOOCOOOODOOOOE AOOOOBOOOOCOOOODOOOOE A....B....C....D....E A....B....C....D....E AOOOOBOOOOCOOOODOOOOE Very worldly Very kind Highly needful of others' approval Feelings easily hurt Very aware of feelings of others Has difficulty making decisions Never gives up easily Cries very easily Very self-confi— dent Feels very supe- rior Very understanding of others Very warm in re- lations with others Very strong need for security Stands up will under pressure 171 Instructions: On the following pages are listed a number of things characteristic of yourself or related to you. You are asked to indicate which things you are satisfied with exactly as they are, which things you worry about and would like to change if it were possible, and which things you have no feelings about one way or the other. Consider each item listed below and encircle the number which best represents your feelings according to the following scale: 1. Have strong feelings and wish change could somehow be made. 2. Don't like, but can put up with. 3. Have no particular feelings one way or the other. 4. Am satisfied. 5. Consider myself fortunate. 1. hair . . . . . .1 2 3 4 5 2. facial complexion. .1 2 3 4 5 3. appetite. . . . .1 2 3 4 5 4. hands. . . . . .l 2 3 4 5 5. distribution of hair over body. . .1 2 3 4 5 6. nose . . . . . .1 2 3 4 5 7. fingers . . . . .1 2 3 4 5 8. elimination. . . .1 2 3 4 5 9. wrists . . . . .1 2 3 4 5 10. breathing . . . .1 11. waist. . . . .1 12. energy level . . .1 13. back . . . . . .1 14. 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. ears . . chin . exercise. ankles . neck . . shape of head body build . profile . height age . . width of shoulders arms . . chest. . eyes . . digestion hips . . skin texture lips . . legs . . teeth. . forehead. feet . . sleep. . voice. . health . sex activities. knees. . 172 www wwwww U101U1U1UIU'IU1U'IU1UI U'IU'IU1 41. 42. 43. 44. 45. 46. 47. 48. posture . . . . face . . . . . weight . . . . sex (male or female).1 back view of head. trunk . . . . penis . . . . buttocks. . . . .1 174 Instructions: Place a check mark in the appropriate box to show how you feel about yourself. 45 g; 4} é”0 0 <8 69 O O m lb 0 0 J o? o? w” #4” @‘T 2' <3 Q‘D 1. I feel that I am a person of worth, at least on an equal plane with others. 2. All in all, I am inclined to feel that I am a fail- ure. I feel that I have a num- ber of good qualities. I am able to do things as well as most other people. 5. I feel I do not have much to be proud of. 6. I take a positive atti- tude toward myself. 7. On the whole I am satis- fied with myself. 8. I wish I could have more respect for myself. 9. I certainly feel useless at times. 10. At times I think I am no good at all. 175 QUESTIONNAIRE PART I About you: Use the space below to describe yourself in as much detail as you think will give a fairly complete picture. Please do not use your name. As usual, con- fidentiality is promised. This sheet will be kept in a place different from where the other test materials are stored so that there will be nothing to identify you on the other tests and questionnaires. 176 PART II SOME BACKGROUND INFORMATION: Instructions: Your answers to the questions below will provide information essential to this study. Please com- plete all items. As usual, confidentiality is promised. Responses will be kept separate from other test materials and under no circumstances will your name be attached. A. Your birthdate B. Racial identification: Black (non-Hispanic) (circle one) White (non-Hispanic) Hispanic Native American Asian American Other C. Highest education level Specify number of years elementary school, high school, college, etc.: Degrees you hold: Degree Major D. Occupation E. Marital Status (circle one) 1. legally divorced 2. never legally married 3. separated 4. other If you were ever legally married, how many times? How long were you married each time? (In years and months. Continue on reverse side if necessary.) 177 PART II F. G. Marital history of parents: Please describe in the space below the marital history of your parents (natural as well as adoptive). Include all information you have about divorces, remarriages, etc. In addition, you should provide to the best of your knowledge information about all of the people you have lived with until the age of 21 (whether parents, stepparents, spouses, lov- ers, etc.). Try to be specific; for example, if you were adopted or if your parents divorced and remarried, tell what age you were when each such event occurred. Current living situation: Are you living with anyone now? Yes No If yes, with whom are you living? with parents with relative(s) other than parent with friend(s) (not a lover) with lover alone other (please specify) Relationships: Do you have a current lover or romantic involvement? (yes or no) If yes, how long have you been involved in this way? If you have a lover or romantic involvement,.rate on the following scale how satisfied overall you have been with the relationship during the past six months (circle one number on the scale even if you have been involved for less than six months): 178 Very Very dissatisfied satisfied /1/2/3/4/5/6/7/8/9/10/ I. If you are not currently in a relationship such as de- scribed on the preceding page, rate on the following scale how much you would like to have such an involve- ment: /1/2/314/5/6/7/8]9/10/ Would not like to Would like very much have such a rela— to have such a rela- tionship--am satis- tionship. fied as I am. Look at the five sentences below: 1. My lover and I are sexually and romantically involved only with each other and do not have sexual and roman- tic involvements with others. 2. My lover and I are sexually and romantically involved with each other, but sometimes have sexual and romantic involvements with others. 3. I am not currently involved with any one person, but have relatively brief sexual and romantic involvements from time to time. 4. I am not currently involved with any one person, but tend to have relatively long-term sexual and romantic involvements. 5. I am not currently involved with any one person, but tend to have few or no sexual involvements of any kind. Which of the sentences above best describes your life now? (Circle one of the following numbers): 1 2 3 4 5 Which of the sentences above best describes your life as you most would like it to be? 1 2 3 4 5 If you have anything to add to the above two questions, please do so in the space below, continuing on the other side if necessary: 179 PART II How long have you considered yourself gay? years/months (Explain below if necessary) Who are the persons to whom you are known as gay--that is, to what individuals have you come out and to what persons are you to any extent open about being gay? About how many of your friends/acquaintances at your place of work know that you are gay? Is this most of the people you work with or only a small part? (Explain in a sentence or two) OR IF YOU ARE A STUDENT: About how many of your friends/acquaintances at the school you attend know that you are gay? Is this most of your friends/acquaintances or only a small part? (Explain in a sentence or two) (This form administered to gay subjects only) 180 PART III A. In the space below, describe an imaginary person that is your idea of the perfect lover, companion or romantic partner. Use as much space as you think it will take to give a fairly complete picture. Your identity is not relevant to this study, so keep in mind that no informa- tion other than the code number listed above will be used to identify you. 181 PART IV A. 10. Instructions: List on the lines below the characteris- tics or attributes that you most look for or would like to have in a lover or romantic partner. These may in- clude any details of personality, social, economic, phy- sical or anatomical characteristics; but please be spe- cific. The characteristics that you list do not have to be written in order of importance, but try to list them as you think of them, being as honest as possible. You should list only as many characteristics as are of interest or importance to you. If you need more space, please write additional characteristics below the lines provided. Regardless of the number of characteristics that you list, however, it is important that you be specific. Remember, confidentiality is guaranteed and under no circumstances will your name be attached to this sheet. 182 PART IV B. 10. Now look at the characteristics that you just listed on the preceding page. Each of the characteristics you listed there corresponds to the scale of the same number below. You are now asked to rate these characteristics on the corresponding lO-point scale below by placing an "X" in the space above the number on the scale which best illustrates how much interest you have in having that characteristic in a lover or sexual partner--that is, how important that characteristic in a lover is to you. If you listed more than 10 characteristics on the preceding page, please rate the remaining characteris- tics in the same fashion by adding enough scales to correspond to the additional characteristics listed. Least Important Most Important / / / / / / / / / / / 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 l 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 3‘ 6 7 8 9 10 183 PART IV C. Now go back two pages to your list of characteristics in Section A and place a check mark next to the three characteristics you listed on that page that you think you tend to emphasize most or that most interests you in connection with a lover, sexual or romantic partner. D. Look at each of the three items you just selected, and of these, place a "1" next to the most important, a "2" next to the second most important. and a "3" next to the least important of the three in terms of how much interest you have in finding a lover, sexual or romantic partner with these characteristics. PART V A. Are you a member of an organized religion? Yes No If yes, what denomination? B. In the space below, please describe briefly your reli- gious orientation. If you believe in God or a speci- fic religion, you should describe your beliefs. Also, note whether you feel some conflict between your reli- gious values and beliefs and your sexuality, and briefly describe the nature of any such conflict that may exist. Continue on the reverse side if necessary. 184 PART VI A. Please describe below what involvement you have in gay organizations, if any, and rate the extent of your in- volvement on the ten-point scale below: (Circle one number) / 1 / 2 / 3 / 4 / 5 / 6 / 7 / 8 / 9 / 10 / Not involved Very at all involved Describe type of involvement and type of organization(s). B. About how often do you visit gay bars? (Specify, for ex- ample, daily, once per week, three times per week, once per month, etc.) C. About how often do you visit straight bars? (Specify, for example, daily, once per week, three times per week, once per month, etc.) (This form administered to gay subjects only) 185 PART VI A. About how often do you go to bars? (Specify, for example, daily, once per week, three times per week, once per month, etc.) (This form administered to heterosexual subjects only)