Read the discussion to Ray Blanchard's (1989) paper "The concept of autogynephilia and the typology of male gender dysphoria."
This is page 1 (above links) & 8 (below links) of 9.        
Two limitations of this study should be noted: a) The analloerotic and heterosexual groups were considerably smaller than the two others. Findings involving these groups might therefore prove less reliable than findings involving the bisexuals and homosexuals only. b) The questionnaire measures of autogynpehilia and alloeroticisms had small numbers of items, and the items themselves were rather crude and blunt. It is possible that more sophisticated instruments would yield a somewhat different pattern of results. The present findings, therefore, need replication; their provisional status should be kept in mind in the following discussion of their taxonomic implications.
The result obtained with the Core Autogynephilia Scale confirmed the prediction that heterosexual, bisexual, and alloerotic gender dysphorics all report sexual stimulation by cross-gender fantasy more frequently than homosexual gender dysphorics. This result parallels the previous findings that all three nonhomosexual groups report histories of fetishistic cross-dressing more frequently than the homosexual group (Blanchard, 1985b); that all three nonhomosexual groups report less childhood femininity than the homosexual group (Blanchard, 1988b); and that all three nonhomosexual groups present for clinical assessment at a greater age than the homosexual group (Blanchard, 1988b). The present finding that the analloerotics (like heterosexuals and bisexuals) also report more heterosexual experience than the homosexual group was not predicted in advance; this finding also, however falls into the same pattern.
The cumulative evidence is consistent with the view that the three nonhomosexual groups are subtypes of one underlying disorder, and that this disorder is different from that found in homosexual gender dysphorics. If this hypothesis is not accepted, then some other explanation is required for the repeated occurrence of the same pattern of empirical results. It should be noted, however, that the present hypothesis not only explains the empirical findings but also predicted most of them beforehand. This establishes its competitiveness in relation to alternative explanations that might be advanced after the fact.
The central finding of this study also supports the view that the nonhomosexual type of gender identity disorder is characterized by an abnormal tendency to be sexually aroused by the thought of being a woman. This finding suggests that nonhomosexual gender dysphoria can, at least provisionally, be labeled "autogynpehilic" gender dysphoria, in line with the tradition of classifying gender identity disorders according to the accompanying erotic anomaly (e.g, Money and Gaskin, 1970-1971).
The foregoing argumentation is based entirely on the evidence of formal, systematic studies. There are, however, other lines of evidence that point toward the same conclusions. These include the epidemiological observation that fetishistic cross-dressing, seen here as a symptom of autogynephilia, is extremely rare in biological females, as are all three types of nonhomosexual gender dysphoria. This is the prevalence pattern one would expect if the nonhomosexual types were related to one another and to autogynephelia. There is also a variety of clinical observations that can be seen as supporting the present views; the clinical material and the epidemiological evidence have been reviewed at length elsewhere (Blanchard, 1989).
The field of gender identity disorders is already overburdened with terminology, and the reader may well wonder what is gained by introducing a new label - autogynephilia - for a phenomenon that is tolerably well described by the familiar term transvestism. One might point out that sexual arousal by women's garments is not truly pathognomic of the disorder in question; a colleague of the writer's, for example, had a patient who was sexually aroused by contemplating his shaved legs in the mirror [Footnote: K. Freund, personal communication, 1979.] Minor problems of definition or diagnosis, however, were not the writer's motivation for introducing (or reintroducing) this concept.
It seems likely that, at some level in the development of nonhomosexual gender dysphoria, the distinction between wearing women's clothes and being a woman - a distinction so apparent to the adult, conscious mind - simply does not exist (cf. Katcher, 1955; Levin et al., 1972; McConaghy, 1979). It may therefore be ultimately unimportant whether one labels nonhomosexual gender dysphoria as "transvestic" or "autogynephilic." The narrow focus of many modern investigators on fetishistic cross-dressing, however, has not been especially productive of new ideas or new findings. It has, for example, failed to suggest any plausible link between an individual's masturbating in women's garments at one stage of life and earnestly pursuing vaginoplasty at another. The finding of this study suggest that returning to a broader conception of the underlying erotic phenomenon might have greater heuristic value at this time.
The heuristic potential of the autogynephilia concept is exemplified by the present results concerning the bisexual group. The hypothesis that there are only two fundamentally different types of male gender dysphoria - homosexual and autogynpehilic - asserts that the discriminably different forms of nonhomosexual gender dysphoria result from the interaction of autogynephilia with additional, constitutional, or experiential factors. This hypothesis, in its present state of development, remains silent on the identity of these differentiating factors. The present study yielded an unexpected clue to one such factor. This was the finding that bisexual subjects are more likely than all others to report sexual stimulation from the fantasy of being admired, in the female persona, by another person. This suggests that bisexual gender dysphorics' interest in male sexual partners is mediated by a particularly strong desire to have their physical attractiveness as women validated by others. This "bisexual" behavior need not reflect an equal erotic attraction to the male and female physiques (and would perhaps be better characterized as pseudobisexuality). [Footnote: It is interesting, in this connection, to note the remark of a married cross-dresser concerning one of his homosexual encounters: "It was all from the point of vanity of being a woman. I have absolutely no taste for homosexuality itself" (Henry, 1948, p. 495).]
One must consider at least one alternative explanation of the bisexuals' high scores on the Autogynephilic Interpersonal Fantasy Scale. This is the possibility that the bisexual group had a general tendency to endorse any and all questionnaire items acknowledging sexual behavior - perhaps the same tendency that got them classified as bisexual in the first place. This possibility is not borne out by the study's other findings: The bisexuals did not produce the highest scores on the Alloeroticism, Heterosexual Experience, or Cross-Gender Fetishism Scales. It thus appears conceivable that the desire four outside validation as attractive "women" is a specific factor in the differentiation of the bisexuals from the other nonhomosexual groups.
The analloerotic group in the present study probably included a handful of subjects who were not truly analloerotic but rather were consciously motivated by the desire for sex reassignment to deny all sexual behavior, whether or not this was accurate. This possibility is raised both by formal studies demonstrating the general unreliablility of self-report in gender patients (Blanchard et al., 1985, 1986) and by casual examination of the present data. The questionnaire data of five subjects in the analloerotic group suggested near total asexuality. Information on the clinical charts of four of these indicated that the patient's self-report regarding sexual behavior could not be considered reliable. In two of these cases, the patient's wife had directly advised the author's clinic of sadomasochistic and fetishistic activities concealed by the patient. In the other two, aspects of the patient's sexual history were contradictory or highly improbable. It should be stressed, however, that the occurrence of such "false alloerotics" does not preclude the existence of a genuine type, which differentiates out of the larger autogynephilic group for reasons unknown at this time.
The closing remarks of this paper concern the relationships between autogynephilia and sexual interest in others. The results of this study's factor analysis and the low correlations of the Alloeroticism Scale with the Core Autogynephila and Interpersonal Fantasy Scales indicate that autogynephilia is not strongly associated with a lack of erotic interest in other persons. This finding appears to depart from the picture of "automonosexualism" painted by Hirschfeld (1918). Hirschfeld (1925, pp. 199-200) modified the epigram, previously quoted in this paper, to the effect that certain cross-dressers love the woman inside them in addition to the women outside them; remarked that all his cases had a pronounced need to feel close to others, which was heightened by cross-dressing; and concluded that these individuals lack the erotic self-sufficiency that defines Rohleder's (101) automonosexualism. It is possible that the net impact of autogynephilia on erotic interest in others is small because autogynpehilia produces new types of erotic need for others at the same time that it diminishes an individual's direct interest in other people's bodies. The two effects, which are not contradictory, might predominate in different groups or coexist in the same person.
This is page 1 (above links) & 8 (below links) of 9.