1. Hanley 1983.

2. My description of these events is based on the following reports: de la Chapelle 1986; Simpson 1986; Carlson 1991; Anderson 1992; Grady 1992; Le Fanu 1992; Vines 1992; Wavell and Alderson 1992.

3. Quoted in Carlson 1991 p. 27.

4. Ibid.

The technical name for Patino’s condition is Androgen Insensitivity Syn­drome. It is one of a number of conditions that leads to bodies having mixtures of male and female parts. Today we call such bodies intersexes.

3. Quoted in Vines i99 2,p. 41.

6. Ibid., p. 42.

7. The contradiction plagued women’s athletics at all levels. See, for ex­ample, Verbrugge і997.

8. The Olympics specifically, and women’s sport in general, have built all sorts of gender difference into the heart of its practice. Barring women from certain events or having different rules for the men’s and women’s games pro­vide obvious examples. For a detailed discussion of gender and sport, see Cahn 1994. For other examples of how gender itself contributes to the con­struction of different male and female bodies in sports, see Lorber 1993 and Zita 1992.

9. Money and Ehrhardt define ‘‘gender role’’ as ‘‘everything that a person says and does to indicate to others or to the self the degree that one is either male, or female, or ambivalent.’’

They define ‘‘gender identity’’ as ‘‘the sameness, unity, and persistence of one’s individuality as male, female, or ambivalent. . . . Gender identity is the private experience of gender role, and gender role is the public experience of gender identity’’ (Money and Ehrhardt 1972^.4. For a discussion of Money’s separation of‘‘sex’’ from ‘‘gender,’’ see Hausman 1993.

Money and Ehrhardt distinguish between chromosomal sex, fetal gonadal sex, fetal hormonal sex, genital dimorphism, brain dimorphism, the response of adults to the infant’s gender, body image, juvenile gender identity, pubertal hormonal sex, pubertal eroticism, pubertal morphology, and adult gender

identity. All of these factors, they believe, work together to define a person’s adult gender identity.

10. See, for example, Rubin 1975. Rubin also questions the biological basis of homosexuality and heterosexuality. Note that feminist definitions of gender applied to institutions as well as personal or psychological differences.

11. The sex/gender dichotomy often became a synonym for debates about nature versus nurture, or mind versus body. For a discussion of how to use debated dichotomies as an aid to understanding the intertwining of social and scientific belief systems see, Figlio 1976.

12. Many scientists and their popularizers claim that men are more com­petitive, more aggressive or assertive, are more sexual, more prone to infi­delity and more. See, for example Pool 1994 and Wright 1994. For a critique of such claims, see Fausto-Sterling 1992, 1997 a, b.

13. For feminists this debate is very problematic because it pits the au­thority of science, especially biology, against the authority of social science— and in any battle of this sort, social science is bound to lose. Science in our culture brings with it all the trappings of special access to the truth: the claim of objectivity.

14. Spelman labeled feminist fear of the body‘‘somatophobia.’’ See, Spel – man 1988. Recently a colleague commented to me that I seem scared of bio­logical theories of behavior. This puzzled him because at the same time he could see that I am devoted to biological studies as one way of gaining interest­ing and useful information about the world. He was right. Like many femi­nists, I have good reason to be scared of bringing biology into the picture. It is not only my knowledge of centuries of arguments in which the body has been used to justify power inequities. I have also encountered such arguments at a personal level throughout my life. In grade school, a teacher told me that women could be nurses but not doctors (after I had announced my intention to become the latter). When, as a young Assistant Professor, I joined the fac­ulty at Brown, a Full Professor in the History Department told me kindly, but with great authority, that history showed that there had never been any women geniuses in either the sciences or the field of letters. We were, it seemed, born to be mediocre. To cap it off, when I returned from scientific meetings, emotionally shaken by my inability to break into the all-male conclaves, where the true scientific exchanges occurred (chatting at the socials and at meals), I read that ‘‘men in groups’’ was a natural outcome of male bonding that had evolved from prehistoric hunting behaviors. Nothing, really, was to be done about it.

I now understand that I experienced the political power of science. This ‘‘power is exercised less visibly, less conspicuously (than overt state or institu­tional power), and not on but through the dominant institutional structures, priorities, practices and languages of the sciences’’ (Harding 1992, p. 567, emphasis in the original). Thus it is no wonder that I and other feminists were (and are) suspicious of grounding the development of the psyche in some bodily essence. We responded to what came to be called ‘‘essentialism.’’ A century ago and today, feminist essentialists argue that women are naturally different—and that such difference forms the basis for either social equality or superiority. For entree into the extensive feminist debates about essen­tialism, seeJ. R. Martin 1994 and Bohan 1997.

15. For a discussion of this recalcitrance in terms of gender schema in adulthood, see Valian 1998a, 1998b.

16. See chapters 1—4 herein; also Feinberg 1996; Kessler and McKenna 1978; Haraway 1989, 1997; Hausman 1995; Rothblatt 1995; Burke 1996; and Dreger 1998b.

One recent sociological account of problems of embodiment considers that ‘‘‘the cutting edge’ of contemporary social theorising around the body may in fact be located within feminism itself’’ (Williams and Bendelow 1998, p. i3°).

17. Moore i994,pp. 2—3.

18. My social activism has included participation in organizations work­ing for civil rights for all people, regardless of race, gender, or sexual orienta­tion. I have also worked on traditionally feminist issues such as shelters for battered women, reproductive rights, and equal access for women in the academy.

19. I am actually willing to broaden this claim to include all scientific knowledge, but in this book I make the argument only for biology—the scien­tific endeavor I best understand. For extended arguments on the topic, see Latour 1987 and Shapin 1994.

20. Some would point to the fact that people express very unpopular sex­ualities despite strong contrary social pressure, even the threat of bodily harm. Clearly, they say, nothing in the environment encouraged the develop­ment of such behavior, but the body will out. Others argue that there must be some prenatally determined disposition that, in interaction with unknown environmental factors, leads to a strongly held, often immutable adult sexual­ity. Members of this latter group, probably the majority ofLoveweb members, call themselves interactionists. But their version of interactionism (meaning that the body and the environment interact to produce behavior patterns) calls for a large dose of body and only a little sprinkling of environment. ‘‘The real issue,’’ one of the staunchest and most articulate interactionists writes, ‘‘is how the body generates behavior’’ (‘‘Lovenet’’ discussion).

2i. Scholarship is not the sole agent of change; it combines with other agents, including traditional means such as voting and forming consumer preference blocks.

22. Haraway 1997, p. 217. See also Foucault 1970; Gould 1981; Schie – binger 1993 a, b.

23. See, for example, Stocking 1987, 1988; Russett 1989 ; Poovey 1995.

24. The historian Lorraine Daston notes that the idea of nature or the natural invoked in debates about the body changed between the eighteenth and nineteenth centuries: ‘‘Early modern nature was incapable of ‘hard facts.’

. . . Modern nature abounded in bitter revelations about the illusions of ethics and social reform, for nature was ruthlessly amoral’’ (Daston i99 2,p. 222).

25. During this time, Foucault maintains, the change from Feudalism to Capitalism required a new concept of the body. Feudal lords applied their power directly. Peasants and serfs obeyed because God and their sovereign told them to (except, of course, when they revolted, as they did from time to time). The punishment for disobedience was, to the modern eye, violent and brutal: drawing and quartering. For a stunning description of this brutality, see the opening chapters of Foucault 1979.

26. Foucault i978,p. 141.

27. These efforts created ‘‘an anatomo-politics of the human body’’ (Foucault 1978 p. 139; emphasis in the original).

28. Because some of the arguments about sex and gender represent the old nature/nurture arguments in modern drag, their resolution (or, as I argue for, their dissolution) is relevant to debates about racial difference. For a dis­cussion of race in terms of modern biological knowledge, see Marks 1994.

29. Foucault 1978 ,p. 139; emphasis in the original.

30. Ibid. In chapter 5 I discuss how the rise of statistics enables twentieth – century scientists to make claims about sex differences in the human brain.

31. Sawicki 1991, p. 67; see also McNay 1993 for specific discussions of Foucault in a feminist context.

32. Foucault i98o, p. 107.

33. Quoted in Moore and Clarke i995,p. 271.

34. Illustrating the anatomo-politics of the human body.

35. Exemplifying the biopolitics of the population.

36. Harding 1992, 1995; Haraway 1997; Longino 1990; Rose 1994; Nel­son and Nelson i996.

37. See also Strock 1998.

38 . Furthermore, the theories derived from such research deeply affect how people live their lives. Recently, for example, a movement to turn homo­sexuals into ‘‘straight’’ people has garnered a lot of publicity. It matters a lot to individual homosexuals if they and others think they can change or if they believe their homosexual desire is permanent and unchangeable (Leland and Miller 1998; Duberman 1991).

For further discussion on this point, see Zita 1992.

For a detailed analysis of bisexuality, see Garber 1995 and Epstein 1991.

The sociologist Bruno Latour argues that once a scientific finding becomes so thoroughly accepted that we dignify it by calling it a fact, placing it without question in textbooks and scientific dictionaries, it moves out of view, behind

a veil that he refers to as a blackbox (Latour 1987). Place a fact in a Latourian black box and people stop looking at it. Nobody asks whether, at the time of its origin, it functioned ideologically in the social or political arena or whether it embodied particular cultural practices or ways of seeing the world.

39. Kinsey etal., 1948; Kinsey et al., 1933.

Kinsey’s Eight Categories. o: ‘‘all psychologic responses and all overt sexual activities directed towards persons of the opposite sex.’’ 1: ‘‘psychosexual responses and/or overt experience are almost entirely toward individuals of the opposite sex.’’ 2: ‘‘the preponderance of their psychosexual responses and/or overt experiences are heterosexual, although they respond rather differently to homosexual stimuli.’’ 3: Individuals who ‘‘stand midway on the heterosexual-homosexual scale.’’ 4: Individuals whose ‘‘psychologic re­sponses are more often directed toward other individuals of their own sex.’’ 3: ‘‘almost entirely homosexual in their psychologic responses and/or their overt activities.’’ 6: ‘‘exclusively homosexual.’’ X: ‘‘do not respond erotically to either heterosexual or homosexual stimuli and do not have overt physical contacts’’ (Kinseyetal. i933,pp. 471—72).

40. When they looked at accumulated homosexual encounters, from ado­lescence through age forty, they reported that homosexual responses had reached 28 percent for women and almost 30 percent for men. When they asked about interactions that led to orgasm, the numbers were still high: 13 percent for women and 37 percent for men (ibid., p. 471). Kinsey did not endorse the notion of homosexuality as a natural category. His system, em­phatically, did not carve nature at the joints.

4i. He did, of course, study these other aspects of human sexual exis­tence, but they were not explicitly part of the 0—6 scale and Kinsey’s com­plexity and subtlety of analysis were often lost in subsequent discussions. As recently as 1989, some researchers complained about the adequacy of the Kinsey scale and proposed more complex grid-like models. One created a grid with seven variables down (sexual attraction, sexual behavior, sexual fan­tasies, emotional preference, social preference, self-identification, hetero/ homo lifestyle) and a time scale (past, present, future) across (Klein 1990).

42. See, for example, Bailey et al. 1993; Whitam et al. 1993; Hamer et al. 1993; and Pattatucci and Hamer 1993.

From the very beginning Kinsey fell under both political and scientific attack. He lost his funding after certain members of Congress became out­raged. Scientists, especially statisticians, attacked his methodology. Kinsey had obtained data from an impressively large number of men and women, but he had collected his overwhelmingly middle class, white, Midwestern popula­tion using what sociologists now call a snowball sample. Starting with stu­dents as one source, he had branched out to their friends and family, their friends’ friends and family, and so on. As word of the study spread (for exam­ple, through his public speaking engagements), he picked up more people, some volunteering after hearing him speak. Although he actively sought out people from different environments, there seems little doubt that he selected a segment of the population who was especially willing, and in some cases even eager, to talk about sex. Might this have accounted for the high frequen­cies of homosexual encounters in his reports?

On the positive side, Kinsey and a small number ofhighly trained co-work­ers (in a fashion true to the racism and sexism of the period, Kinsey’s inter­viewers had to be male, white, and WASP) conducted all of the interviews. Rather than use preset questionnaires, they followed a memorized procedure and had the leeway to pursue lines of questioning in order to be sure they had gotten complete answers. More modern survey approaches have exchanged this more flexible, but also more idiosyncratic, interview process for a level of standardization that permits using less highly trained interview personnel. It is very hard to know whether important data are lost as a result. I owe this point to James Weinrich (personal communication) (Brecher and Brecher 1986; Irvine 1990a, b).

43. This is a necessary feature of doing molecular linkage studies (for any multifactorial trait) because the power of resolution is so low. (See Larder and Scherk 1994.) If the trait is not narrowed enormously, it is impossible to find statistically significant association. But narrowing the trait makes it inappro­priate to generalize a finding to the general population (Pattatucci 1998).

44. For the grid model, see Klein 1990. For one version of an orthogonal model, see Weinrich 1987.

43. Chung and Katayama 1996.

In the most important recent survey of human sexual practices in the United States, Edward O. Laumann, John H. Gagnon, Robert T. Michael, and Stuart Michaels categorized their results along three axes: same-sex sexual behavior, desire, and identity (Laumann, Gagnow, etal. 1994). For example, Laumann and colleagues found that 39 percent of women with at least some homosexual interest expressed same-sex desire but no other behaviors, and 13 percent reported that they had same-sex desire, behaviors, and self-identity as lesbian. Thirteen percent reported same-sex behaviors (sexual interac­tions) but without strong homosexual desire and without identifying as les­bian. Although the precise distributions for men differed, the same general conclusion held. There is a ‘‘high degree of variability in the way that differing elements of homosexuality are distributed in the population. This variability relates to the way that homosexuality is organized as a set of behaviors and practices and experience subjectively, and it raises provocative questions about the definition of homosexuality’ (Laumann Gagnow et al. i994,p. 300). The sample size for these studies was 3,432, age range 18 to 39. There were discrepancies in the data, which the authors note and discuss. Among them: 22 percent of women report being forced into some sexual act, but only 3 percent of men say they have force women into sex. Men say they have more sex partners than women do, so who are the men having all that sex with? See Cotton 1994; see also Reiss 1993.

46. I often hear from my biology colleagues that our compatriots in other fields have an easier time than we because scientific knowledge changes con­tinuously while other fields are static. Hence we need constantly to revise our courses, while a historian or Shakespearean scholar can legitimately give the same old lecture, year after year. In fact, nothing could be further from the truth. The field of literature changes all the time as new theories of analysis and new philosophies of language become part of the academic’s tools. And an English professor who does not regularly update her lectures or develop new courses to reflect the changing field receives just as much criticism as the biochemistry professor who reads his lectures directly from the textbook. My colleagues’ attitudes represent an attempt at boundary maintenance—a method of trying to make scientific work special. The entire thrust of current analyses of science, however, suggests that it is not so different after all. For an overview of work in the social studies of science, see Hess 1997.

47. Halperin 1990, pp. 28—29.

48. Scott i993,p. 408.

49. Duden 1991 ,pp. v, vi.

3°. Katz i993.

31. Trumbach 1991a.

32. McIntosh 1968.

33. In philosophy the question of how to categorize human sexuality is usually discussed in terms of ‘‘natural kinds.’’ The philosopher John Dupre writes more generally about the difficulties of biological classification of any sort: ‘‘There is no God-given, unique way to classify the innumerable and diverse products of the evolutionary process. There are many plausible and defensible ways of doing so, and the best way of doing so will depend on both the purposes of the classification and the peculiarities of the organisms in question’’ (Dupre 1993, p. 37). For other discussions of natural kinds with regard to classifying human sexuality, see Stein i999 and Hacking i992 and

i993.

Even now many of us spend idle moments speculating about whether So and so is ‘‘really’’ straight or ‘‘really’’ a queer just as we ‘‘might question whether a certain pain indicated cancer’’ (McIntosh 1968 ,p. 182).

34. Only through time travel, Latour argues, can one understand the so­cial construction ofa particular scientific fact. Interested parties must journey back to a period just before the fact in question appeared on Earth and follow along as citizens of an earlier time participated in its ‘‘discovery,’’ argued about its reality, and finally agreed to place it in the dark box of facticity (see note 38). Thus we cannot understand modern scientific formulations of the structure of human sexuality without traveling back in time to their point of origin.

33. There is now a rich and growing literature on the history of sexuality. For an overview of ideas about masculinity and femininity, see Foucault 1990 and Laqueur 1990. For studies of sexuality in Rome and in early Chris­tendom, see Boswell 1990 and Brooten 1996. For up-to-date scholarship on the Middle Ages and the Renaissance, see Trumbach 1998 and 1987; Bray 1982; Huussen 1987; and Rey 1987). For changing expressions of sexuality in the eighteenth and nineteenth centuries, see Park 1990; Jones and Stallybrass 1991; Trumbach 1991a, b; Faderman 1982; and Vicinus 1989. For additional historical accounts, see Boswell 1993; Bray 1982; Bullough and Brundage 1996; Cadden 1993; Culianu 1991; Dubois and Gordon 1983; Gallagher and Laqueur 1987; Groneman 1994; Jordanova 1980 and 1989; Kinsman 1987; Laqueur 1992; andMort 1987. For looks at how ideas about health and disease have been linked to our definitions ofsex, gender, and morality, see Moscucci 1990; Murray 1991; Padgug 1979; Payer 1993; Porter and Mikulas 1994; Porter and Hall 1993; Rosario 1997; Smart 1992; and Trumbach 1987 and :989.

36. Katz 1976 and Faderman 1982.

37. Halwani 1998 provides one example of the ongoing nature of this debate.

38. Sometimes touted as the seat of modern democracy, Athens was, in fact, ruled by a small group of elite male citizens. Others—slaves, women, foreigners, and children—had subordinate status. This political structure provided the scaffolding for sex and gender. There were, for example, no specific prohibitions against men having sex with one another. What really mattered was what kind of sex one had. A citizen could have sex with a boy or a male slave so long as he actively penetrated and the other passively received. This sort of sex did not violate the political structure or bring into question the masculinity of the active partner. On the other hand, penetrative sex be­tween citizens of equal status ‘‘was virtually inconceivable’’ (Halperin 1990, p. 31). The sex act declared one’s social and political standing. ‘‘Sex between social superior and social inferior was a miniature drama of polarization which served to measure and define the social distance between them’’ (idem, p. 32). Position mattered. In the pattern that emerges from analyzing the variety of sex acts depicted in drawings on Greek vases, male citizens always pene­trated women or male slaves from the rear. (No, the missionary position is neither universal nor ‘‘natural’’!) But in the much-touted relationships be­tween older men and their younger male citizen proteges, sex (without pene­tration) happened face to face (Keller 1983). Weinrich 1987 distinguishes among three forms of homosexuality identified either in different cultures or in previous historical eras: inversion homosexuality, age-structured homo­sexuality, and role-playing homosexuality. See alsoHerdt i99oaand 1994a, b.

59. Katz 1990 and 1995. Other authors (Kinsman 1987) note the written use of the word in 1869 by the Hungarian K. M. Benkert. Something must have been in the air.

60. Hansen 1992 and 1989. French, Italian, and American accounts fol­lowed soon after.

61. Ellis 1913. A number of historians point out that the medical profes­sion’s involvement in defining types of human sexuality was only part of the story. For a variety of more nuanced accounts see Krafft-Ebing 1892; Chauncey 1985 and 1994; Hansen 1989 and 1992; D’Emilio 1983 and 1993; D’Emilio and Freedman 1988; and Minton 1996. Duggan writes: ‘‘turn-of – the-century sexologists, far from creating or producing new lesbian identi­ties, drew their ‘cases’ from women’s own stories and newspaper retellings of them as well as from French fiction and pornography as ‘empirical’ bases for their theories’’ (Duggan i993,p. 809).

62 . In earlier periods male and female sexuality was understood to lie along a continuum from hot to cold (Laqueur 1990).

63. The true invert of this period cross-dressed and, when possible, took up appropriately masculine work. Ellis, writing in 1928, described the in­verted lesbian: ‘‘The brusque, energetic movements, the attitude of the arms, the direct speech, . . . the masculine straightforwardness and sense of honor. . . will all suggest the underlying psychic abnormality to a keen observer. . . there frequently a pronounced taste for smoking cigarettes. . . but also a decided tolerance for cigars. There is also a dislike and sometimes incapacity for needlework and other domestic occupations, while there is often some capacity for athletics.’’ Ellis 1928, p. 250. No single book made this point more clearly while affecting the lives of thousands of lesbians well into the 1970s than Hall 1928. See also chapter 8 of Silverman 1992.

64. Although the notion of the invert strongly influenced turn-of-the cen­tury sex experts (who became known as sexologists), the idea was unstable, changing as strict sex roles weakened and men and women began more often to appear in the same public spaces. Ellis and then Freud began to note that in men one might separate masculine behaviors and roles from same-sex desire. Thus object choice (or what we today often call sexual preference) grew in importance as a category for classifying sexuality. A similar division came more slowly to women, perhaps not fully emerging until the feminist revolu­tion of the 1970s smashed rigid sex roles into bits. For more on the history of sexology, see Birken 1988; Irvine i990a, b; Bullough 1994; Robinson 1976; and Milletti 1994.

For a fascinating description of this transformation from the point of view oflesbians themselves, see Kennedy and Davis 1993.

65. Although male-male sex did not bother them, the Greeks recognized the existence of molles, unmasculine men who wanted to be penetrated, and tribades, women who, although engaging in sex with men, preferred other women. They considered both groups mentally troubled. But the abnormality lay not in same-sex desire. Rather, what worried Greek physicians was that molles and tribades were gender deviants. They either mysteriously wished to surrender male power by becoming a passive sex partner, or, intolerably, they tried, by becoming the active partner, to assume male political status. Both the molle and the tribade differed from normal folk by having too much of a good thing. They were understood to be oversexed. (Molles apparently devel­oped the desire to be penetrated because taking the active role did not offer sufficient sexual release.) David Halperin writes: ‘‘these gender-deviants de­sire sexual pleasure just as most people do, but they have such strong and intense desires that they are driven to devise some unusual and disreputable. . . means of gratifying them’’ (Halperin i99o, p. 23).

66. The historian Bert Hansen writes: ‘‘A tentative sense ofidentity facili­tated further interaction. . . which then facilitated the formation of a homo­sexual identity for more individuals’’ (Hansen i992,p. 109).

67. Ibid., p. 125. See also Minton 1996.

The historian George Chauncey provides impressive evidence for a large and fairly open and accepted social world for urban gay men during the first third of this century. He argues that, in contrast to that period, gay culture encountered a great period of repression from the 1930s through the 1950s (Chauncey 1994). Allan Berube (1990) documents the participation of gay men and women in World War II. He suggests that the modern gay movement forms one of the ultimate legacies of their struggles in the armed services. For a fascinating oral history of the postwar gay rights movement, see Marcus

1992. Additional essays on the postwar period maybe found in Escoffier et al. 1995. For discussions of historiographical problems in writing histories of sexuality, see Weeks 1981a, b and Duggan 1990.

68. Its English language entree occurred in 1889 with the English transla­tion of Krafft-Ebing’s Psychopathia Sexualis.

69. Katz 1990 p. 16.

Today the concept of heterosexual appears to us as inexorably natural. But the first 30 years of the twentieth century had passed before it solidified on American shores. In 1901 neither the terms heterosexual nor homosexual ap­peared in the Oxford English Dictionary. During the teens and 20s novelists, playwrights, and sex educators fought censorship and public disapproval to make a public space for the erotic heterosexual. Only in 1939 did the word heterosexual finally emerge from the medical demi-monde to achieve that honor of all honors, publication in the New York Times. From there to Broadway, as a lyric in the musical Pal Joey, took another decade.

Katz 1990. The full Pal Joey lyric is quoted on p. 20; for a more detailed account of the history of the modern concept of heterosexuality, see Katz 1995. In 1929 the sex educator Mary Ware Dennett was convicted of sending obscene material—a sex education pamphlet for children—through the mails. Her criminal writings declared the joys of sexual passion (of course, within the confines of love and marriage). The author Margaret Jackson ar­gues that the development of the field of sexology undermined feminists of the period ‘‘by declaring that those aspects of male sexuality and heterosexu­ality were in fact natural, and by constructing a ‘scientific’ model of sexuality on that basis’’ (Jackson 1987^. 55). Forfurther discussion of feminism, sex­ology, and sexuality in this period, see Jeffreys 1985.

70. Nye 1998, p. 4.

71. Boswell i99o, pp. 22, 26.

72. Nye 1998, p. 4.

73. As, for example, James Weinrich suggests (Weinrich 1987).

74. Not all anthropologists agree on the exact number of patterns; some cite as many as six patterns. As with many of the ideas discussed in this chap­ter, the academy is still in flux as new data pour in and new approaches to analyzing old data proliferate.

75. McIntosh 1968.

76. In the years since McIntosh’s essay, books’ worth of scholarship on the topic have been published. See, for example, Dynes and Donaldson 1992a, 1992b, and Murray 1992.

77. For reviews of cross-cultural studies of human sexuality, see Davis and Whitten 1987; Weston 1993; and Morris 1995.

78. See, for example, how Weinrich uses the notion of human universals to infer the biological basis of behavioral traits (Weinrich 1987).

79. Vancei99i, p.878.

80. Note that such a definition permits Boswell to be a mild social con­structionist while still believing that homosexual desire is inborn, transhistor­ical, and cross-cultural. Indeed, the phrase social construction does not refer to a unified body of thought. The meaning of the phrase has changed with time; more modern ‘‘constructionists’’ are generally more sophisticated than early ones. For a detailed discussion of the different forms of constructionism and essentialism, see Halley 1994.

81. Vance 1991, p. 878. Halperin certainly falls into this more radical constructionist category.

82. Herdt 1990a, p. 222.

83. A careful reading of Herdt’s account of Melanesian societies reveals three underlying (Western) assumptions: that homosexuality is a lifelong practice, that it is an ‘‘identity,’’ and that these definitions of homosexuality may be found worldwide.

84. Elliston 1995, p. 849.

Ibid., p. 852 .

Anthropologists have similar disagreements about the implications of Na­tive American practices that scholars refer to as ‘‘Berdache’’—a variety of practices involving culturally sanctioned cross-gender roles and behaviors. Some argue that the existence of Berdache proves that the assumption of cross-gender roles and behaviors is a universal expression of inborn sexuality, but others find this to be an ahistorical, simplistic view of practices that have varied greatly across North American cultures and history. Carolyn Epple, who has been studying contemporary Navajo nadleehi (the Navajo word for ‘‘Berdache’’) observes, for instance, that Navajo definitions of nadleehi vary from case to case. Such variation makes sense because the Navajo worldview she studies ‘‘seems to place more emphasis on situation-based definitions than on fixed categories.’’ Epple is very careful to qualify phrases such as the ‘‘Na­vajo Worldview’’ by indicating that she is talking about the one her informants discuss. There is no singular worldview, because it changes historically and regionally, and is better understood as a complex of overlapping belief sys­tems. This contrasts with Euro-American assumptions that homosexuality is a fixed or natural kind.

(For discussions of natural kinds, see Dupre 1993; Koertge 1990; and Hacking 1992 and 1993.) Moreover, Epple points out, the Navajos don’tnec – essarily regard nadleehi as gender transgression. The Navajo Epple studies con­ceptualize everyone as both male and female. Thus they would not describe a man with a woman’s mannerisms as feminine. ‘‘Given that both male and female are ever-present,’’ Epple observes, ‘‘a gender valuation of ‘masculine’ versus ‘feminine’ will generally reflect the perspective of the observer, and not some absolute value’’ (Epple 1998, p. 32). For additional critiques of the ‘‘Berdache’’ concept, see Jacobs etal. 1997.

83. See, for example, Goldberg 1973 and Wilson 1978.

86. Ortner 1996.

87. Although they didn’t invent the concept, Kessler and McKenna use the idea to excellent effect in their analysis of cross-cultural studies of gender systems (Kessler and McKenna 1978).

88. Ortner i996,p. 146.

89. Ortner writes: ‘‘Hegemonies are powerful, and our first job is to un­derstand how they work. But hegemonies are not eternal. There will always be (for both better and worse) arenas of power and authority that lie outside the hegemony and that may serve as both images of and points of leverage for alternative arrangements’’ (ibid., p. 172).

90. Oyewumi i998,p. 1033. See also Oyewumi 1997.

91. Oyewumi i998,p. 1061.

92. Oyewumi 1997, p. xv. Oyewumi notes that gender divisions are espe­cially visible in African state institutions, which were derived originally from colonial formations—that is, they represent the transformed impositions of colonialism, including the gender beliefs of the colonizers.

93. Stein 1998. For a full treatment of Stein’s ideas, see Stein 1999.

Much of contemporary biological, psychological, and anthropological re­search uses homosexuality as real or natural categories. Some examples in­clude Whitam et al. 1993; Bailey and Pillard 1991; Bailey et al. 1993; and Buhrich et al. 1991.

94. One other feminist biologist, Lynda Birke, has moved in the same direction, but because her book is forthcoming, and I have only read an early outline and the advanced publicity, I cannot cite it more specifically (Birke I999).

93. Halperin 1993, p. 416.

96. Plumwood I993,p. 43.

Plumwood also argues that dualisms ‘‘result from a certain kind of denied dependency on a subordinated other.’’ The denial, combined with a relation­ship of domination and subordination, shape the identity of each side of the dualism’’ (ibid., p. 41). Bruno Latour uses a different framework to make a similar point—that nature and culture have been artificially divided in order to create modern scientific practice. See Latour 1993.

97. Wilson I998, p. 33.

98 . In her words, she ‘‘wants to ask how and why ‘materiality’ has become a sign of irreducibility, that is, how is it that the materiality of sex is under­stood as that which only bears cultural constructions and, therefore, cannot be a construction’’ (Butler I993,p. 28).

99. Ibid., p. 29.

100. Ibid., p. 31.

ioi. For other examples of sedimented meanings in science, see Schie – binger 1993a, on Linnaeus’s choice of the breast as the word to designate the class mammalia, and Jordanova 1989 on Durkheim’s account of women in his 1897 book Suicide.

102. Butler I993,p. 66.

103. Hausman I993,p. 69.

104. Grosz I994,p. 33.

103. Singh 1942; Gesell and Singh 1941; Candland 1993; andMalson and Itard 1972.

106. ‘‘The body image cannot be simply and unequivocally identified with the sensations provided by a purely anatomical body. The body image is as much a function of the subject’s psychology and sociohistorical context as its anatomy’’ (Grosz I994,p. 79). See also Bordo 1993.

107. The philosopher Iris Young considers a similar set of problems in her book and essay of the same title (Young 1990).

108. Phenomenology is a field that studies the body as an active partici­pant in the creation of self. Young writes: ‘‘Merleau-Ponty reorients the entire tradition of that questioning by locating subjectivity not in mind or conscious­ness but in the body. Merleau-Ponty gives to the lived body the ontological status that Sartre. . . attribute(s) to consciousness alone’’ (Young 1990), p. 147-

Grosz relies heavily on a rereading of Freud, on the neurophysiologist Paul Schilder (Schilder 1950), and on the phenomenologist Maurice Merleau – Ponty (Merleau-Ponty 1962).

109. Grosz 1994.P. 116.

110. Ibid., p. 117. The scholars to whom Grosz turns to understand the processes of external inscription and subject formation include Michel Fou­cault, Friedrich Nietzsche, Alphonso Lingis, Gilles Deleuze, and Felix Gu – attari.

in. For continuing discussion of the positions Grosz develops, see Grosz 1995; Young 1990; and Williams and Bendelow 1998.

112. I suspect that Grosz understands this, but has chosen the ill-defined starting point of a ‘‘drive’’ (hunger, thirst, etc.) because she needed to begin her analysis somewhere. In fact, she mentored Elisabeth Wilson, whose work provides part of the theoretical basis needed to dissect the notion of drive itself.

113. In discussing developmental systems theory, I do a lot of ‘‘lumping.’’ I have found new ways of thinking about organismic (including human) devel­opment among thinkers working in a number of different disciplines. They have not always read each other, but I discern common threads that link them. At the risk of doing one or more of them an injustice, I will refer to them under the rubric of developmental systems theorists. The disciplinary back­grounds out of which this work comes include: Philosophy: Dupre 1993; Hack­ing 1992 and 1995; Oyama 1985, 1989, 1992a, 1992b, 1993; andPlumwood

1993. Biology: Ho et al. 1987; Ho and Fox 1988; Rose 1998; Habib et al. 1991; Gray 1992; Griffiths and Gray 1994a, 1994b; Gray 1997; Goodwin and Saunders 1989; Held 1994; Levins andLewontin 1985; Lewontin et al. 1984; Lewontin 1992; Keller and Ahouse 1997; Ingber 1998; Johnstone and Gott­lieb 1990; and Cohen and Stewart 1994. Feminist Theory: Butler 1993; Grosz 1994; Wilson 1998; and Haraway 1997. Psychology and Sociology: Fogel and Thelen 1987; Fogel etal. 1997; Lorber 1993 and 1994; Thorne 1993; Garcia – Coll et al. 1997); Johnston 1987; and Hendriks-Jansen 1996. Law: Halley

1994. Science Studies: Taylor 1995, 1997, 1998a, and 1998b; Barad 1996.

114. Many social scientists and some geneticists view organisms as re­sulting from the addition ofgenes and environment. They study organisms by looking at their variability and ask what proportion of the variability can be attributed to genes and what proportion to environment. A third term, which they designate as a gene-environment interaction, may be added to the equa­tion of the simple sums if genetic and environmental cause don’t account for all of the variance. This approach has been roundly criticized on more than one occasion. Sometimes such scientists call themselves interactionists, be­cause they accept that both genes and environment are involved. Their critics note that this approach to the analysis of variance portrays genes and environ­ment as separately measurable entities. Some of these critics also refer to themselves as interactionists because they consider it impossible to separate the genetic from the environmental. I prefer to use the idea of a develop­mental system because of this confusion of terminology and because the idea of a system entails the concept of mutual interdependence of its parts. For critiques of the partitioning of variance, see Lewontin 1974; Roubertoux and Carlier 1978; and Wahlsten 1990 and 1994.

115. Oyama 1985, p. 9. The revised and expanded edition of Oyama’s book is due out in the year 2000 (Duke University Press).

116. Taylor 1998a, p. 24.

117. For references on this point, see Alberch 1989, p. 44. As another example, an embryo needs to move in the womb to integrate nerve, muscle, and skeletal development. Mallard ducklings still in the shell must hear them­selves quack in order to respond to maternal quacks. (Wood ducks need to hear their siblings quack in order to develop the ability to recognize Mom.) (Gottlieb 1997).

118. Ho i989,p. 34. Alberch makes a similar point, writing, ‘‘itisimpos – sible to state that form determines function or vice versa since they are inter­connected at the level of the generative process’’ (Alberch i989,p. 44).

119. LeVay’s results still await confirmation and in the meantime have been subject to intense scrutiny (LeVay 1991). See Fausto-Sterling i992aand 1992b; Byne and Parsons 1993; and Byne 1995. I do not read anything into the current lack of confirmation other than that it is a difficult study to do because of the relative scarcity of autopsy material from individuals with a known sexual history. A confirmation of his results will not help us under­stand very much about the development or maintenance of homosexuality unless we place the information into a developmental system. Standing alone, his findings can prove neither nature nor nurture.

120. I was horrified to start getting mailings and phone calls from right­wing Christian organizations that assumed my public argument with LeVay meant I was sympathetic to their homophobic agenda.

121. Bailey and Pillard 1991; Bailey et al. 1993; Hamer et al. 1993.

i22 . In a detailed and brilliant analysis of the problems posed by the na – ture/nurture, essential/constructed, biology/environment dichotomies, the lawyer Janet Halley calls for the development of common ground from which to struggle for personal, political, and social equality (Halley 1994).

123. Oyama 1985.

124. LeVay 1996.

125. Extraordinary, because it is not customary to use a strictly scientific report to discuss the potential social implications of one’s work. Hamer et al. I993,p.326.

126. Wilson is more interested in the philosophical nature of the attacks on LeVay’s work than in the technical critiques. She willingly grants the valid­ity of many of these, as, indeed, does LeVay himself (see LeVay 1996). For the technical critiques see Fausto-Sterling 1992a and 1992b; and Byne and Parsons 1993.

127. Wilson includes me in the list of feminists who had a knee-jerk anti­biology response to LeVay. While I don’t think that I have ever thought of human sexuality in terms that discard the body, it is true that I have been wary of putting many such thoughts into print because I was caught in the grip of the essentialist/antiessentialist dualism. The history of essentialist ideology in the oppression of women, homosexuals, and people of color has been an enormous counterweight in my thinking. Only now that I see how systems theory provides a way out of this dilemma am I willing to commit myself to discussing these questions on the printed page.

128. Wilson i998,p. 203.

129. I will discuss here some of those connectionists who apply their ideas to brain function or who model brain function using computer models of neural networks.

130. The psychologist Esther Thelen writes: ‘‘A view now is that multimodal information is bound together frequently and in multiple sites along the processing stream and that there is no single localized area in the brain where perceptual binding occurs’’ (Thelen i995,p. 89).

Connectionists postulate processing elements called nodes, or units (which might, for example, be nerve cells). The nodes have many connections that enable them both to receive and send signals to other nodes. Different connec­tions have different weights or strengths. Some nodes receive signals while others send them. Between these two types of nodes lie one or more layers that transform signals as they are sent. The transformations happen according to basic rules. One type is a i: i (i. e., linear) transmission, another is a thresh­old (i. e., above a certain level of input, a new response is activated). It is the nonlinear responses of neural network models that most resemble actual human behavior and that have excited the imagination of cognitive psychol­ogists.

131. I have cobbled together this primitive account of a complex field from three sources: Wilson 1998; Pinker 1997; and Elman et al. 1996.

132. This has recently been shown to be the case for studies of mouse behavior. Three groups of researchers on different parts of the North Ameri­can continent took genetically identical strains of mice and attempted to get them to exhibit the same behavior. To do this, they standardized the experi­ments in every way they could think of—same time of day, same apparatus, same testing protocol, etc.—but they got markedly different results. There were clearly laboratory-specific environmental effects on behavior in these mice, but the experimentalists cannot figure out what environmental cues are important. They urge caution and multiple-site testing before concluding that a genetic defect affects a behavior (Crabbe et al. 1999).

133. When researchers ask identical twins to solve puzzles, the twins come up with answers that are more alike than those of paired strangers. But if monitored using PET scans, while working on the puzzles, the twins’ brains do not show identical function. ‘‘Those identical twins with their identical genes never have identical brains. Every measure differs.’’ This result is expli­cable with a developmental systems account of behavior, but less so with an accountthat suggests that genes ‘‘program’’ behavior (Sapolsky 1997, p. 42).

134. Elman et al. i996,p. 339. See also Fischer 1990.

133. Joan Fujimura writes: ‘‘Just because something is constructed does not mean that it is not real’’(Fujimura 1997, p. 4). Haraway writes: ‘‘The bodies are perfectly ‘real.’ Nothing about corporealization is ‘merely fiction.’ But corporealization is tropic and historically specific at every layer of its tis­sues’’ (Haraway 1997, p. 142).

136. Haraway envisions objects such as the corpus callosum as nodes out of which grow ‘‘sticky threads’’ that ‘‘lead to every nook and cranny of the world’’ (see the last two chapters of this book for concrete examples). Biolo­gists, doctors, psychologists, and sociologists all employ a ‘‘knot of knowledge-making practices,’’ including ‘‘commerce, popular culture, social struggles. . . bodily histories. . . inherited narratives, new stories,’’ neurobi­ology, genetics, and the theory of evolution to construct beliefs about human sexuality (Haraway 1997, p. 129). She refers to the construction process as material-semiotic practice and the objects themselves as material-semiotic objects. She uses this complex phrase very specifically to bypass the real/ constructed divide. Human bodies are real (i. e., material) but they interact only via language—the use of signs (verbal and otherwise). Hence the word semiotic.

137. This is a good example of Dupre’s argument that there is no fixed way to divide up nature (Dupre 1993) and of Latour’s plea to look at science in action (Latour 1987).

138. Connectionists, of course, do not believe that behaviors and motiva­tions have a permanent location in the brain; instead, they view behavior as the result of a dynamic process.