Read the introduction 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.        
The term gender dyphoria refers to discontent with one's biological sex, the desire to possess the body of the opposite sex, and the desire to be regarded by others as a member of the opposite sex. Transsexualism may be defined as extreme gender dysphoria that has persisted without fluctuations for a considerable time - 2 years, according to the DSM-III-R (American Psychiatric Association, 1987). The term transvestism will be used here in the contemporary clinical sense of recurrent and persistent cross-dressing that, at least in puberty or adolescence, is accompanied by genital excitement.
The terms homosexual and hetereosexual are applied to gender dysphorics (including pre- and postoperative transsexuals) exactly as they are to other individuals, to refer to erotic attraction to members of the same or the opposite chromosomal sex, respectively. Similarly, the label bisexual refers to gender dysphorics attracted to both sexes. The term analloerotic - from Greek borrowings an-, "lacking," and alloerotic, "sexual feeling or activity finding its object in another person" (Webster's Third International Dictionary, 1981) - will be used in this paper to designate gender dysphorics who report no sexual attraction to either sex. This term, unlike asexual, does not necessarily imply a lack of all erotic interests, only erotic interest in other persons. The term nonhomosexual will be understood, in this paper, to include hetereosexual, bisexual, and analloerotic gender dysphorics.
Hirschfeld (1918) first noted that one class of cross-dressing men are sexually aroused by the thought or image of themselves as women. He referred to this group as automonoseuxals. [Footnote: A term borrowed from Rohleder (1901). Rohleder, however, used the term to denote a kind of pathological narcissism in which the individual is excited by his own body in its real (i.e., male) form, whereas Hirschfeld's automonosexual cross-dresser is aroused by the fantasy that his body is that of a woman.] "They feel attracted not by the women outside them, but by the woman inside them" (Hirschfeld, 1948, p. 167). Many later observers have alluded to this same phenomenon (e.g., Buckner, 1970; Freund et al., 1982; Ovesey and Person, 1976), but the basic concept has never been elaborated beyond Hirschfeld's epigrammatic formulation of it.
Blanchard (1989) proposed that a male's propensity to be sexually aroused by the thought of himself as a female be labeled autogynephilic ("love of oneself as a woman"). In Blanchard's view, autogynephilia, like heterosexuality, homosexuality, and pedophilia, is conveniently indexed by an individual's tendency to respond with penile erection to the erotic stimulus, but also includes the capacity for pair-bond formation (or something like it) with that stimulus.
The concept of autogynephilia underlies Blanchard's (1985b, 1988b, 1989) hypothesis that there are only two fundamentally different types of gender identity disturbance in males: homosexual (aroused by men) and autogynephilic (aroused by the idea of being a women). Autogynephilia, according to this hypothesis, may be manifested in a variety of ways, fetishistic cross-dressing being the most striking. Transvestites on this view, would be understood as autogynephiles whose only - or most prominent - symptom is sexual arousal in association with cross-dressing and who have not (or not yet) become gender dysphoric.
The hypothesis asserts that the various discriminable syndromes of nonhomosexual gender dysphoria are the results of autogynephilia interactions with additional, constitutional, or experiential factors. Hetereosexual gender dysphorics represent those cases in which the autogynephilic disorder interferes the least with normal erotic attraction to other persons. It should be noted, however, that many hetereosexual gender dysphorics are able to maintain potency with their wives only by means of cross-gender fantasy during intercourse (Fruend, 19745; Levine and Lothstein, 1981). In many cases, the individual prefers to have intercourse with his wife in the female superior position. He then fantasizes that his wife, imagined as a man, is penetrating him, a woman (Benjamin, 1966; Lukianowicz, 1959). Others fantasize during heterosexual intercourse that they and their wives are two women having lesbian relations (Newman and Stoller, 1974).
Bisexual gender dysphorics represent those cases in which the autogynephilic disorder gives rise to some secondary interest in men that coexists with the individual's basic attraction to women. Autogynephilia, as indicated above, may find expression in the fantasy of having intercourse, as a woman, with a man. In bisexual gender dysphoria these fantasies are especially strong; they are therefore more likely to be actualizes - or rather, approximated - with anal or oral intercourse substituting for vaginal - particularly with the bisexual gender dysphoric in partial or complete cross-dress (Benjamin, 1967; Person and Ovesey, 1974). The effective erotic stimulus in these interactions however, is not the male physique of the partner, as it in in true homosexual attraction, but rather the thought of being a women, which is symbolized in the fantasy of being penetrated by a man. For these persons, the male sexual partner serves te same function as women's apparel or make-up, namely, to aid and intensify the fantasy of being a woman.
Analloerotic gender dysphorics represent those cases in which the autogynephilic disorder nullified or over-shadows any eortic attraction to women; those cases, in Hirschfeld's metaphor, in which "the women within" completely supplants her fleshy rivals. Some analloerotics are most aroused by tangible symbols of their femininity, for example, chanfing into women's attire or putting on make-up. Others are most aroused by transsexual ideas, such as the thought of having women's breasts or a vagina. The feature common to all members of this group is their erotic self-sufficiency.
The forgoing hypothesis does not deny that homosexual gender dysphorics also have erotic fantasies or that their fantasies also include cross-gender content. The occurrence of such fantasy should not, however obscure the essential feature of the homosexual type: Homosexual gender dysphorics are directly aroused by the objective features of the male physique, especially the sight and feel of the male genitalia, and this arousal in not dependent on the mediation of cross-gender fantasy.
The notion that the nonhomosexual gender dysphorias are variant forms of the same disturbance predicts that the heterosexual, bisexual, and analloerotic groups will report developmental histories similar to one another and dissimiliar from the homosexual group. Such predicition have been confirmed in two previous studies. Blanchard (1985b) found that similar majorities of analloerotic, bisexual, and hetereosexual transsexuals acknowledged some history of erotic arousal in association with cross-dressing, whereas only a small minority of homosexual subjects did so. Blanchard (1988b) found no different in the average age at which analloerotic, bisexual, and heterosexual transsexuals first presented for clinical assessment. All three groups were significantly (9 to 15 years) older at initial presentation than the homosexual transsexuals. Blanchard (1988b) also found no difference in the average degree of childhood femininity reported by analloerotic, bisexual, and heterosexual transsexuals. All three groups reported significantly less feminine identification than did the homosexual group.
The above-mentioned findings all tend to support the hypothesis that the main types of nonhomosexual gender dysphoria are variant forms of one underlying disturbance. Neither of the previous studies, however, directly tested the notion that the characteristic feature of this disturbance is the tendency to be sexually aroused by the thought of being a women. The present study, therefore, tested the prediction that all three types of nonhomosexual gender dysphoric will be more likely than the homosexual type to report symptoms of autogynephilia. This comparison necessitated the prior development of an objective, questionnaire measure of autogynephilia, which thus became a secondary goal of the study.
This is page 1 (above links) & 8 (below links) of 9.