1. Milton Diamond, Elizabeth Adkins-Regan, William Byne, Donald Dewesbury, Marc Breedlove, and, indirectly, Kim Wallen—all of whom study the role of hormones in behavior and/or study the comparative psychol­ogy of animals—took time to comment on an earlier draft of this chapter. Their critiques were generous and of enormous help. I am very grateful for their efforts. Their commitment to helping me ‘‘get it right,’’ even when I occasionally stepped on their toes, represents the spirit of open, scientific inquiry at its best. Of course, I am solely responsible for the final outcome.

2. Aberle and Corner i933. Borell ^987) lists the transfer date as i93i. See also the discussion in Clarke i998.

3. Borell Q987) quotes a BSH memorandum explaining the transfer: ‘‘It had been felt by the Bureau for some time that this would be an advantageous move, inasmuch as the Foundation, through its biological experts, could fur­nish an advisory control which the Bureau could not supply; and the Founda­tion also inclined to the view that the administration of this program and the evaluation of the results of the researches conducted are more clearly in the field of the present research programs of the natural sciences and medical sciences of the Foundation than in the program of the Bureau’’ (p. 79).

4. Quoted in Borell i987,p. 79. Borell notes that this new independence of scientific researchers resulted in the abandonment of the search for an easy – to-use spermicide contraceptive, which had anyway ‘‘never summoned the interest of scientists as the contraceptive pill was to do’’ (p. 85). The contra­ceptive pill was in the end developed (with Sanger’s support and funding) at a private foundation founded by Gregory Pincus after he was denied tenure at Harvard following intense controversy over his early work on artificial par­thenogenesis in mammals. See also Clarke 1990a and b.

5. Quoted in Kohler 1976^.291.

6. For the story of how these events led to the modern-day fields of molec­ular biology, see Kohler 1976; Kay 1993; and Abir-Am 1982.

7. Aberle and Corner i953,p. 100.

8. Aberle and Corner (1953 ) list Terman’s last CRPS grant for the prepa­ration of a ‘‘report on the marital adjustment of intellectually superior sub­jects’’ (p. 129). For the path from Yerkes to Carpenter to modern primatology as a model for human sex behavior and social organization, see Haraway 1989.

9. For a brief history of the latter two fields in the United States, see Dewsbury 1989.

10. Many received funds from the Rockefeller Foundation both indepen­dently and via the Committee for Research in Problems of Sex. Before 1938, 25 percent of grants from CRPS funded behavioral research, with most of the rest addressing the basic physiology of sex and reproduction. From i938 to 1947, however, 45 percent of CRPS’s grants went for research into sex-related behavior, with a major focus on the role of hormones. For a complete list for this period, see Aberle and Corner 1953.

ii. There is a large parallel literature on primates, work that hormone researchers have always felt to be particularly applicable to humans. Some of the concepts developed with rodents do not hold up well in primates. But the primate work is expensive and difficult, because of the long lives of the ani­mals, the need for breeding colonies, and the growing recognition that pri­mate behavior, even more than rodent behavior, needs naturalistic settings if one wants to draw conclusions about ‘‘normal’’ development. There is also an influential literature on birds, one of the few groups for which the relationship between hormones and certain aspects of brain development is fairly clear. See Schlinger 1998. For a current and extensive review of the work on verte­brates, see Cooke etal. 1998.

12. Squier i999,p. 14 of electronic printout.

13. Quoted in May i988,p. 93.

14. Quoted in D’Emilio 1983^.41. For a deeper discussion of the inter­twining of anticommunism, homosexual repression, emphasis on a narrowly defined family structure, and the staking out of clear cultural definitions of masculinity and femininity, see May 1988 and 1995, Breines 1992, andEhren – reich 1983. See D’Emilio 1983, Ehrenreich 1983, and Reumann 1998 for discussion of the enormous secondary literature on homosexuality and gender in the postwar period.

15. Schlesinger 1958, p. 63. Women, wrote Arthur Schlesinger, Jr., ‘‘seem an expanding, aggressive force, seizing new domains like a conquering army, while men, more and more on the defensive, are hardly able to hold their own and gratefully accept assignments from their new rulers. A recent book bears the stark and melancholy title The Decline of the American Male” (p. 63).

16. Quoted in May 1988^.140.

17. Quoted in May i988,p. 66.

18. In the 1930s, masculinity did not require special attention. The field of andrology emerged as an independent discipline in the 1970s. See, for ex­ample, Bain et al. 1978. Niemi (1987) notes that the idea of andrology ap­peared as early as 1891, but that the first societies and journals didn’t coalesce until the i970s.

19. Quoted in May 1988^.147.

20. This idea resurfaces from time to time. In response to the increase in single mothers, Robert Bly described his vision of the ‘‘deep masculine’’— something sons imbibed bodily from their fathers and that single mothers, no matter how well meaning, could never give (Bly 1992).

21. SeeD’Emilio 1983. For a thorough and illuminating discussion of the Kinsey reports and the national discussion of sex and sexuality, see Reumann I998.

22. Quoted in Elger et al. 1974, p. 66, from remarks made at a 1969 workshop conference on ‘‘Integration of Endocrine and Non-endocrine Mechanisms in the Hypothalamus.’’

I have limited this entire discussion to mammals; there are significant differences found in other vertebrates.

23. Jost 1947, 1946a, 1946b, and 1946c.

24. Wiesner i935,p. 32; emphasis in original.

25. Greene etal. 1940b, pp. 328,450.

26. These experiments all addressed the question of secondary sex deter­mination—i. e., the development of the gonadal duct system and external genitalia. Jost did not examine primary sex determination— i. e., the differ­entiation of the gonad as either a testis or an ovary.

From his first publication through the 1970s, Jost also actively promoted his work—publishing many times, often in review articles or in symposium proceedings, so that his original data, while supplemented steadily by new results, also received ongoing attention.

27. Jost 1946c, p. 301; his emphasis, my translation. Later experimenters identified two culprits. Embryonic testosterone encouraged differentiation of the male duct system and masculine external genitalia, while a new hor – mone—a proteinlike structure dubbed Mullerian Inhibiting Substance (MIS for short)—induced degeneration of the embryonic female duct system. The embryonic male testis makes both fetal testosterone and MIS. Jost tried re­moving only one testis; under those circumstances male development contin­ued apace, while the female duct system degenerated as it normally would in an unoperated male fetus. From this and other experiments he concluded that the testis secreted one or more factors that caused male duct differentiation and female duct degeneration. Jost also grafted testes onto female embryos or ovaries into males, but the grafted tissue did not affect the embryos’ develop­ment, a failure he attributed to the fact that he had to use older embryos, which he presumed had already passed through the stage during which devel­opment was plastic. However, the androgen-supplemented embryos still differed from normal males because they had at least some uterine develop­ment, although the vaginal region was ‘‘more or less inhibited.’’ Jost never reported testing the possible actions of estrogen on the development of either castrated male or castrated female embryos, although it is possible that he tried such tests but that the estrogen caused the embryos to abort.

28. MIS is currently an object of great research interest because it has been identified as an important and ubiquitous growth factor—transforming growth factor-^. Gustafson and Donahoe (1994) review the work on MIS molecular biology (pp. 509—16).

29. Jost 1946c, p. 307; my translation. Jost soon extended his studies ex­amining male and female duct tissue grown in in vitro tissue culture. This work, however, did not eliminate the possibility of hormonal effects on female development. As he, himself, pointed out, his culture system was not ‘‘anhor – monal.’’ In 1951, Jost wrote that the action of trace estrogens contained in the serum used as his culture medium ‘‘cannot be neglected a priori. We must ultimately return to the use of a synthetic hormone-free medium’’ (Jost and Bozic 1951, p. 650); see also Jost and Bergerard 1949. But by 1953 his inter­pretation had begun to change. Acknowledging that female development might be affected by maternal hormones produced in the placenta or the mother’s gonads, or by nonovarian fetal hormones (e. g., the adrenal glands), and reminding his readers that he had provided evidence of some ovarian ac­tivity, he nevertheless felt ‘‘that maternal or extragonadal gynogenic sub­stances can hardly account for the feminization of the gonadectomized fetus’’ (Jost 1953, p. 387). This, he concluded, despite again acknowledging earlier reports that estrogens could feminize male fetal development (Greene et al. 1940a, 1940b; Raynaud 1947. Jost (1953) wrote that ‘‘the interpretation of this experiment was not evident’’ (p. 417).

30. Jost’s rhetoric changed with time. In 1954, he wrote ‘‘the foetal testis plays the principle role’’ in normal sexual development (implication: females become female because they lack a testis) (Jost 1954, p. 246). In i960, he wrote: ‘‘the anhormonal (sex in mammals) is feminine and the testis prevents males from differentiating as females’’ (Jost i960, p. 59). By 1965 he was saying thatfemale mammals are ‘‘the neutral sex type’’ (Jost i960, p. 59). In 1969, ‘‘becoming a male is a prolonged, uneasy, and risky adventure; it is a kind of struggle against inherent trends toward femaleness’’ (Jost 1965, p. 612). Finally, in 1973, Jost wrote: ‘‘masculine characteristics. . . have to be imposed in males by the fetal testicular hormones against a basic feminine trend of the mammalian body. Female organogenesis results from the mere absence of testes, the presence or absence of ovaries being unimportant’’ (Jost etal. i973,p. 41).

When computer terminology entered the language in the 1980s, research­ers updated Jost’s description of an inherent trend toward femaleness into a metaphor of female development as a ‘‘default pathway.’’ The earliest use I can find of the phrase ‘‘default sex’’ to describe female development is 1978. The editors of the journal Trends in Neuroscience use the term in the introduction to Dohler 1978.

31. Jost et al. 1973. Jost was French, and I have not looked at the specifics of such discussions in France after World War II. But his ideas were known and discussed internationally and gained rapid acceptance in the United States. The production of scientific knowledge not only involves doing experi­ments and interpreting results, but being in the right place at the right time for a particular result and interpretation to be culturally intelligible. For more on this issue, see Latour 1987.

The mono-hormonic theories also echo nineteenth-century views of women, children, and nonwhites as being closer to nature. All races and sexes developed identically up to a point, but only white males continued devel­oping into true adulthood. For a full treatment of these nineteenth-century views, see Russett 1989 and also Herschberger 1948.

32. Aristotle wrote: ‘‘The female is a female by virtue of a lack of certain qualities. We should regard the female nature as afflicted with a natural defec­tiveness.’’ St. Thomas thought that women were imperfect men, incidental beings. In the oedipal drama of becoming (a la Freud), the female psyche must accommodate to the absence of a penis, while the male psyche must adjust to the fear of its loss and thus a return to some basal female state (quoted in de Beauvoir 1949, p. xxii).

Additional explanations for the acceptance of the female = absence, male = presence theory may include the difficulty of the necessary experi­ments, the time needed to fill in difficult-to-get details, which could be ob­tained only by diverting attention from easier and more immediately produc­tive (in terms of publications) experiments. One component of scientific success is the ability to balance a forward-moving program against the impor­tance of digging into a recalcitrant problem.

Some of the unresolved experimental results included: (1) the possibility that Jost’s castrations were not done early enough to detect an effect of remov­ing the fetal ovary; (2) that injected estrogens could feminize male develop­ment and stimulate the growth of female organs; (3) while Jost tried substitut­ing injected testosterone for the removed testis, he never performed parallel experiments for the removed ovary; (4) he did not work to identify possible nonovarian sources of estrogen or other nonestrogenic differentiative factors that might govern female differentiation; (3) Jost knew that the fetal ovary made estrogen from an early point, but did not seem to worry about the func­tion of the estrogen.

The possibility that estrogen, from either the mother or the fetal ovary, might play a role in secondary sex determination is still not completely re­solved. Certainly in some vertebrates it ‘‘is thought to play a major role in the gonadal differentiation,’’ (di Clemente et al. 1992, p. 726); see also (Reyes et al. 1974. George et al. (1978) found that the embryonic rabbit ovary begins to make large quantities of estrogen at exactly the same time as the fetal testis begins to make testosterone. They suggest further studies to clarify the func­tion of this fetal estrogen (Ammini et al. 1994; Kalloo et al. 1993). These latter authors find that: ‘‘The presence of estrogen receptors suggests that maternal estrogen may play a direct role in female external genital develop­ment, challenging the widely held view that female external genital develop­ment is passive because it can occur in the absence of fetal gonadal hormones’’

(p. 692).

33. Greene’s results showing estrogen’s potential to actively feminize male embryos rankled. Jost continued to note the need for further experi­mentation to resolve such contradictory results. Slowly, however, references to Greene’s work and calls for further experimentation disappeared from Jost’s writing. By 1963, the mono-hormonic theory appeared in Jost’s writing as fully proven fact, rather than tentative theory requiring further experimen­tal verification. Although continuing to note that estrogens could feminize male embryos, he suggested that injected estrogen did not actively cause differentiation. Rather, it damaged the testes’ ability to make testosterone, thus allowing the ‘‘natural’’ femininity of the embryo to emerge. In 1963, Jost still considered the presence/absence theory of male and female develop­ment to be ‘‘speculative.’’ Although he could present a tidy-looking story, such a presentation ‘‘should not conceal the necessity ofnew crucial experiments” (Jost 1963, p. 614). but he never did perform all of the critical experiments suggested in is 1947 paper.

34. See notes 43 and 46 for evidence of this claim. In a 1999 debate about the concept of default development, one Loveweb member wrote: ‘‘Maybe the female program is also dependent on some hormone—all we know is that gonadal hormone isn’t required. How about the 20—30 hormones that probably exist that we haven’t discovered yet? The more I grow old and cur­mudgeonly, the less sense this default stuff makes. I don’t think it means a da**ed thing, but merely poses as a phrase that means something.’’ Evidence continues to emerge that suggests the importance of events in the ovary for regulating sexual differentiation (Vainio et al. 1999). It does seem likely, how­ever, that in mice neither progesterone nor estrogen is a major actor in initial events (Smith, Boyd, etal. 1994; Lydonetal. 1995; Korach 1994).

35. This process is called primary sex determination.

36. Once a fetal gonad appears, it can produce hormones that induce sec­ondary sexual development—the problem that researchers in the 1930s through the 1950s addressed, and to which I will return later in this chapter.

37. Schafer etal. i995,p. 271; emphasis added.

38. Wolf 1995, p. 325; emphasis added.

39. Capel I998, p. 499.

40. Angier 1999, p. 38.

41. See, for example, Mittwoch 1996.

42 . The metaphor that promotes feminist glee can also fuel masculine oppression. ‘‘Western culture,’’ writes the psychologist Helen Haste, ‘‘has a strong tradition of rationality overcoming the forces of chaos that is closely interwoven with masculine versus feminine. . . . One pole is not only anti­thetical to, it triumphs over, the other pole. Dark forces must be challenged and conquered’’ (Haste 1994, p. 12). In a similar vein, the feminist historian Ludmilla Jordanova notes how the Enlightenment brought us word pairs such as nature/culture, woman/man, physical/mental, mothering/thinking, feeling and superstition/abstract knowledge and thought, darkness/light, na – ture/science and civilization (Jordanova 1980 and 1989).

43. Wolf 1995, p. 325. At least one scientist I have corresponded with disputes this claim, but I believe it is justifiable. Many embryology texts have a section entitled ‘‘sex determination’’ that discusses only male development. For example, Carlson considers the topic of ‘‘the genetic determination of gender.’’ He first notes that females develop in the absence of a Y chromosome, then spends the entire section discussing male development. Figure 15—22 in his book illustrates a complex and detailed account of the mechanisms of male development, but there is no analogous illustration of the mechanisms of fe­male development (Carlson 1999, pp. 375—76). The only modern textbook that treats male and female development in an evenhanded manner is by Scott Gilbert (Gilbert 1997). And it is no accident that one of Gilbert’s public hats is that of a feminist historian of science. See also Swain et al. 1998; Haqq et al. 1994; and McElreavey et al. 1993.

44. Fausto-Sterling 1989.

45. Eicher and Washburn i986,pp. 328—29.

46. Wolf acknowledges that ‘‘female development is undoubtedly not spontaneous’’ (p. 325), but does not otherwise discuss female development. Two articles by Sinclair discuss the testis-determining pathway, and although acknowledging that there is complexity underlying both ovarian and testicu­lar determination, never hypothesizes an ovarian determining pathway (Sin­clair i995, i998). Capel writes that the default terminology ‘‘may be mis­leading because it suggests that the female pathway is not an active, genetically controlled process’’ (1998, p. 499). Hunter gives a paragraph to Eicher and Washburn’s hypothesis, but then spends the rest of a sixty-six-page chapter (entitled ‘‘Mechanisms of Sex Determination’’) discussing genes for testis de­termination (Hunter 1995). Swain et al. write: ‘‘ovary differentiation is un­likely to be passive as there are changes in gene expression that occur very early in XX genital ridge development’’ (1998,p. 761).

Only three current papers picture genes active in female development. Such accounts of ‘‘sex differentiation’’ (as opposed to male differentiation) are still in the minority (Werner et al. 1996; Jimenez and Burgos 1998; Schafer and Goodfellow 1996).

47. The ‘‘master’’ gene hypothesis weighs heavily in this story. Most cur­rent work on primary sex determination considers that the Y chromosome contains a ‘‘master gene,’’ a switch that starts the development ball rolling. It takes, according to this model, only one gene to determine male develop­ment. Others argue that development is a process for which many genes are critical, as is getting the timing of gene action right. On the latter viewpoint, see Mittwoch 1989, 1992, and 1996.

48. Milton Diamond writes: ‘‘Asa graduate student my first thesis attempt in this vein was to see if estrogens could feminize male fetuses as androgens masculinized females. My injections of estrogens into pregnant guinea pigs invariably resulted in fetal death. This was a great disappointment to me since it’s hard to study behavior that way’’ (Diamond 1997a, p. 100). Another re­searcher wrote me that the effects of estrogen on animal behavior were small and hard to measure. ‘‘That doesn’t mean that they are unimportant, of course, but if you were an Assistant Professor and you wanted to be produc­tive, I hope you’d choose to study robust effects and not subtle ones’’ (anony­mous, personal communication).

49. For Jost’s description of meeting Wilkins, see Jost i972,pp. 38—39.

50. Frank Beach writes: ‘‘The importance of support distributed by this committee to development of hormone behavior research has never been ade­quately recognized. . . the decision of the Committee for Research in Prob­lems of Sex to encourage investigations of copulation in rats . . . , or fre­quency of orgasm in married women. . . , was a courageous one that eventually opened the way for general expansion of research on effects of hor­mones on a very important category ofbehavior’’ (Beach 1981 ,p. 354).

The work on hormones and animal behavior that evolved from the late 1930s through the 1960s built directly and consciously on the issues addressed by early hormone researchers. Beach cites Lillie, Moore, Marshall, Heape, and many others as early contributors to the field (Beach 1981).

Beach’s Ph. D. mentor at the University of Chicago was Karl S. Lashley (1890—1958). Lashley’s work on brain mechanisms and intelligence empha­sized a holistic view of brain function, and his views are clearly reflected in

Beach’s work and manner of thinking. For more on Lashley, see Weidman, 1999-

Beach discussed his results on brain-injured rats with an endocrinologist who suggested that brain injury could disturb pituitary secretion and thus affect gonadal hormone secretion. Beach wrote of this encounter: ‘‘I hadn’t the faintest understanding of what he was talking about; but after reading a bit of endocrinology I decided to inject some of my brain-operated, de-sexed males with testosterone just to see what would happen. . . lo and behold! The injected rats regained their libido; and I thought I was on the way to a Nobel Prize’’ (Beach 1985, p. 7.).

In the United States, the field of animal psychology was known as compar­ative psychology. In Europe, a related but distinct tradition was known as ethology. Only during the 1950s did European ethology strongly influence American comparative psychologists. For a historical treatment of compara­tive psychology, see Dewsbury 1989 and 1984.

51. As one researcher wrote me, ‘‘For the behaviorist, that is the beauty of it because there is so much that can be easily measured.’’ Even closely re­lated species differ in the details. Male guinea pigs, for example, resemble primates in the use of repetitive thrusts but a single intromission (anonymous, personal communication).

5 2. Behaviors such as nest building, maternal care, and aggression in ter­ritorial defense also defined masculinity and femininity in rats, but in this period, Beach focused primarily on figuring out the components of mating behaviors. For recent theories on hormones, experience, and parenting be­havior, see Krasnegor and Bridges 1990.

53. For an overwrought account of the necessity of psychoanalysis for daily life, see Lundberg and Farnham 1947.

54. See, for example, Watson 1914 and Dewsbury 1984.

55. Beach was not enamored of Watson and the behaviorists. In 1961 he wrote: ‘‘It seems to me the time has come for a re-examination ofthese prob­lems with great attention being given to genetically-influenced biologic fac­tors which may contribute to some of these differences’’ between the sexes and between racial groups (Beach 1961, p. 160).

56. In a retrospective moment William C. Young wrote: ‘‘Research on the relationships between the hormones and sexual behavior has not been pur­sued with the vigor justified by the biological, medical, and sociological im­portance of the subject. Explanation may lie in the stigma any activity associ­ated with sexual behavior has long borne. In our experience, restraint has been requested in the use of the word ‘sex’ in institutional records and in the title of research proposals. We vividly recollect that the propriety of present­ing certain data at scientific meetings and seminars was questioned’’ (Young

19^ p. 2I2).

57. Beach і942іТ p. 173.

58. In 1947 he wrote: ‘‘Importance of the holistic approach: Physiological experiments designed to identify the nervous pathways involved in a particu­lar genital reflex, or to measure the importance of secretions from a single gland to the occurrence of copulatory reactions, have contributed a great deal to our understanding of sexual behavior. It should be obvious, however, that the full significance of such findings becomes apparent only when they are viewed against the broader background of the total sexual pattern as it appears in the normal animal’’ (Beach 1947, p. 240).

59. ‘‘Individual differences in the ease with which various inexperienced males become sexually aroused constitute an important factor which must be taken into consideration in any attempt to define the adequate stimulus for mating behavior. A stimulus situation eliciting copulation in one male may fail to call forth the mating reactions of a less excitable individual of the same species’’ (Beach 1942c, p. 174).

60. ‘‘The appearance of the overt copulatory pattern depends jointly upon the male’s sexual excitability, and the intensity of the stimulation afforded by the incentive animal. A highly excitable male may attempt copula­tion with an incentive animal of relatively low stimulus value. … A less ex­citable male fails to show mating reactions in response to all incentive animals other than the receptive female with which he will copulate. A male of low excitability may not be aroused to the point of copulation even when offered the receptive female’’ (Beach i942e, p. 246). Beach and other researchers commented on the fact that there were always males and females in a colony that seemed to have no interest in mating. Eventually, it became common practice to eliminate such animals from tests of mating activity.

61. Beach 1942c.

62. Apparently animals could still mate, even with their cortex removed. See Beach i942b, c, pp. 179—181; and Beach 1943.

63. Beach 1941.

64. Beach 1942a. Normal females did not require outside testosterone to show a male mating pattern. Beach and Priscilla Rasquin raised females in sexually segregated quarters and then tested them daily through four mating cycles. During the test, they allowed the female to adapt to the test cage, placed her with a receptive female for five minutes, and then with a sexually active male. They divided the female’s masculine mating behaviors into three types: (i ) mounting and embracing the mounted animal with her forepaws; (2 ) mounting, touching the mounted animal with her forepaws, and pelvic thrusting; and (3) mounting, touching, and ‘‘giving a final forceful thrust and dismounting with a pronounced backward lunge.’’ Of 20 females, 18 exhib­ited the sexual clasp, i8 showed mounting, touching, and pelvic thrusting, and 5 engaged in the complete ‘‘male’’ copulation pattern. The masculine behaviors occurred whether or not the females were in heat.

Beach and Rasquin drew some startling conclusions: First, they noted that the majority of the female rats in their colony had the brain and muscular anatomy needed for a male mating pattern. Second, they concluded that the same stimulus—a female in heat—elicited this male pattern in both sexes. Finally, they noted that ovarian hormones did not control masculine mating in female rats (Beach and Rasquin 1942); see also Beach 1942a and f. Beach first reported on these cross-gendered behaviors in 1938. He quotes from his own lab journal, dated 1937, on Male No. 136 interacting with Female No. 192:

10:03 AM: Female dropped into observation cage containing male. . . . 10:13 AM: . . . Both animals display all signs of intense sexual excitement but the male never actually mounts and palpates the female. 10:16: Fe­male whirls about, approaches male from the rear and mounts and pal­pates actively. The forepaws of the female clasp and palpate the male. . . and the female’s pelvic region is moved in and out with the piston-like action characteristic of the copulating male. After this brief display of mas­culine activity the female dismounts, without the typical masculine lunge, and does not clean the genital region. 10:17: Female responds to male’s investigatory activity by crouching, arching the back, and vibrating the ears rapidly.

Beach notes that this particular female mounted and palpated the male seven times in a fifteen-minute observation period. He emphasizes that she exhibits both masculine and feminine responses (Beach 1938 ,p. 332).

63. Beach 1942b, p. 183. To reinforce his point, Beach also cites Carl Moore’s earlier debates with Steinach, especially Moore’s insistence that indi­vidual rats varied too much to be used as an indicator of hormone presence or absence.

66. This mass effect is compatible with Lashley’s approach to brain function.

67. Later he reported on experiments confirming this hunch. He contin­ued to emphasize the holistic approach: ‘‘Evidence makes it plain that the effects of androgen are mediated by a complex combination of mechanisms, of which the supposed tactile functions of the glans are only one’’ (Beach and Levinson i930,p. 168).

68. Beach 1947—i948,p. 276.

69. Beach i963,p. vii.

70. Beach writes that certain statements in his text ‘‘are based upon data generously made available by Dr. A. C. Kinsey of Indiana University, whose extensive interview study of sexual behavior in more than i0,000 humans is to be published in the future (Beach 1947, p. 301). Kinsey and Beach were both funded by CRPS, a fact that Kinsey notes in the introduction to his 1948 study. They met and talked about their common interests.

71. Marc Breedlove, personal communication (May 1999). Kinsey used personal interviews to gather his data. He personally recruited and trained in­terviewers.

72. In Jones (1997) and Gasthorne-Hardy (1998) Beach discusses his friendship with Kinsey. Kinsey got his first CRPS grant in 1941 and received new funding yearly, in increasing amounts, through 1947 (Aberle and Cor­ner 1953).

73. Kinsey et al. 1948, 1933. In the 1933 volume, Kinsey specifically thanks Beach for contributing information on animal behavior (p. ix).

74. See, for example, Berube 1990; Katz 1993.

73. For the many strands and complexities of this discussion, see Reu – mann 1998.

76. The architect of this new work had done many studies on individual variation and concluded with others that individuality emerged because each body—or, as the scientists call it, ‘‘substrate’’—differed. He wrote: ‘‘it was clear that the central problem for the investigator interested in accounting for the great variability in patterns of mating behavior was to identify the factors which determine the character of the substrate on which the gonadal hor­mones act’’ (Young i960, p. 202). This article brought the latest in rat re­search into the psychiatric community. The theory of organization and activa­tion is only infrequently used to explain differences among individuals of the same species, even though this question originally stimulated experiments that led to the O/A theory.

77. For a complete bibliography of Young’s publications and a brief biog­raphy, see Goy 1967. He was supported through money Lillie obtained from CRPS (Dempsey 1968); see also Roofe 1968. Although he worked on other animals, especially rats and monkeys, and some of his students focused espe­cially on primates, the bulk of Young’s publications were devoted to guinea pig behavior.

78. Quoted in Goy i967,p. 7.

79. Young 1941, p. 141.

80. Young and Rundlett i939,p. 449.

81. Youngetal. 1939. They wrote: ‘‘in any measurement of sexual drive, mounting activity and receptivity should be regarded as separable compo­nents of a sexual behavior complex and measured directly by whatever means are considered most appropriate’’ (p. 63).

82. Young and Rundlett 1939.

83. In this cyclical dependence guinea pig females differ from rats. Young indicates the lingering confusion discussed in the previous chapter engen­dered by expectations for so-called male and female hormones: ‘‘Early in the work it was anticipated that estrogen-androgen rather than estrogen – progesterone action would stimulate mounting activity. The relative ineffec­tiveness of the androgens which have been employed is surprising, but their ability to substitute for progesterone more efficiently in the induction of heat than in the induction of the male-like mounting activity is even more puz­zling’’ (Young and Rundlett 1939, p. 439).

84. Young 1941, p. 311. Here we see the culture of scientific practice at work. To do science at all, some measurable starting point was needed. Young, like the others, needed steady results in order to obtain funding, train stu­dents, and continue his work. Successful scientific practice, in other words, does not necessarily lead to a balanced overview of organismal function. It does lead to carefully designed experiments of the sort that give specific re­sults and pave the way for more carefully designed experiments.

83. They also included another category of mating response called ‘‘other’’ (Young and Grunt 1931).

86. ‘‘It is postulated. . . that much of the difference between individuals is attributable to the reactivity of the tissues rather than to differences in the amount of hormone’’ (Grunt and Young 1932, p. 247). See also Grunt and Young 1933 and Riss and Young 1934.

87. Valenstein et al. 1933, p. 402. The additional papers detailing the importance of genetic background and experience in males are Valenstein et al. 1934; Riss et al. 1933; Valenstein and Young 1933; and Valenstein and Goy 1937. Young’s group began in this period to flirt more seriously with the distinction between early organization of neural patterns and their activation at a separate time by circulating hormones. In one paper they write: ‘‘The data suggest the role of t. p. [testosterone propionate] to be that of an activator rather than a direct organizer of sexual behavior. The organization is depen­dent on variables associated with the strains and upon opportunity to learn the techniques of mounting and maneuvering a female’’ (Riss et al. 1933, p. 144). At the time Young also felt that the organization of sexual behaviors in males ‘‘is not as sharply restricted to an early critical period as is’’ imprinting in birds (Young 1937, p. 88). After 1939, Young and others began to insist on the importance of a critical period, and once they had demonstrated a prenatal organizing effect of testosterone, they no longer wrote about the social isola­tion effects as ‘‘organizing.’’ Robert Goy, whom Young trained, also found strain differences and experience important for the organization of female mating responses. These findings assume importance in view ofthe later focus on the role (or lack thereof) of prenatal estrogen in organizing female mating patterns. See Goy and Young 1936—37; Goy and Young 1937; Goy and Jak-

way 1939.

88. Phoenix etal. 1939.

89. IbidM p. 370.

90. FordandBeach 1931 ,p. 123; HampsonandHampson i96i, p. 1,423. Although this paper appeared two years after that of Phoenix et al., Young edited the volume in which it was published. Thus he and his co-workers had read it and could refer to it ‘‘in press.’’

91. The dispute was complex. Hampson and Hampson, for example, wrote that their study ‘‘of human hermaphroditism points strongly to the tre­mendous influence of rearing and social learning in the establishment of nor­mal gender role. . . and by analogy, disordered psychologic sex.’’ At the same time they did not entirely rule out genetic or constitutional contributions. But they thought the ‘‘evidence militates too strongly against a theory of innate, preformed and inherited behavioral imperatives, hormonal or otherwise’’ (1961 ,p. і,428). There were also debates within Young’s lab about the mean­ing of the findings: ‘‘The younger members of the team were more convinced [than Young] that this was a direct brain effect. . . . This was a hotly debated issue in the lab while the paper was being written and the somewhat contra­dictory views finally presented reflect a balance between what they suspected had occurred and what they could actually demonstrate’’ (Kim Wallen, per­sonal communication, July іі, 1997).

92. I discussed the technical details of this paper at some length in Fausto – Sterling 1995. I have been convinced by critics that some aspects of this earlier treatment were in error, especially my failure to give Young his full historical due and my assertion that Phoenix et al. claimed a brain effect, when in fact they were more cautious, claiming a central nervous system effect. But the paper is useful for showing how the O/A theory has been modified in fairly fundamental ways since its original publication, and I stand by my critique that the model leaves out experience and genetic and individual difference. It is not the sort of holistic model that Beach wanted, nor that I develop in this chapter and the next.

93. Phoenix al. 1959, p. 372. They performed four basic experiments: (і) they injected prenatally exposed females in adulthood with estradiol and progesterone and measured aspects of their mating responses, concluding that prenatal androgen exposure suppressed the lordosis response, but not male­like mounting; (2) they tested for ‘‘permanence’’ of the effects of prenatal androgen and found them present at 6—9 and then again at 11-12 months of age (guinea pigs live for 10—12 years), concluding that ‘‘the suppression of the capacity for displaying the feminine components of the sexual behavior pat­tern. . . appears to have been permanent’’ (p. 377); (3) they studied the effects of injecting adults who had been prenatally exposed to androgen with testosterone, finding such females more responsive (i. e., more likely to ex­hibit a masculine mating pattern) to testosterone than were untreated females and concluding that ‘‘the earlier appearance and greater strength ofmasculine behavior by the hermaphrodites given testosterone propionate are believed to be effects of the prenatally administered testosterone propionate on the tissues mediating masculine behavior and therefore to be expressions of its orga­nizing action’’; (4) they examined the behavior of adult male siblings—also exposed to androgen prenatally; here they found no apparent effect of prenatal testosterone treatment.

Here I discuss only mating behavior. The authors were well aware of other sex-differentiated behaviors (e. g., maternal behaviors, nest building, territo­rial aggression), but Young and Beach had spent decades defining mating be­haviors in a manner that could be quantitatively measured and evaluated.

94. Grady and Phoenix 1963, p. 483. Rats began to be used for these studies because there is a biological difference of practical importance be­tween rats and guinea pigs. The important anatomical and organizational events in guinea pigs take place in utero because guinea pigs are long-gestation animals. Rats, however, gestate for a shorter time and are born far more sexu­ally undifferentiated. Young and colleagues never succeeded in doing prenatal castrations (in utero) in guinea pigs, but in rats they could work on individual newborns rather than do surgery on a pregnant female. Furthermore, they could directly treat test individuals with hormones rather than inject pregnant females (Grady et al. 1963).

93. Beach 1981. Beach discusses both Young’s and his early work in Beach (1981). In an autobiographical piece Beach lists the organizational effects of hormones during early development under the topic ‘‘Discoveries I almost made.’’ He also discusses his dog experiments in this context (Beach 1978, p. 3°).

96. Phoenix et al. 1939 ,p. 381. The central nervous system refers to the brain and spinal cord. Although they suspected brain involvement, the au­thors were cautiously agnostic, since they did not have evidence to this effect.

97. Phoenix etal. i939,p.379.

98. Ibid., p. 380. It took less than a decade for Young to adopt the pres – ence/absence language introduced by Jost. In 1967,he wrote: ‘‘Manyofthose traits which are sexually dimorphic. . . appear to be influenced in the mascu­line direction by appropriate treatment with androgen and in the feminine direction by the absence of early steroid hormones’’ (Young 1967, p. 180).

99. Phoenix etal. i939,p.38o.

100. Young continued to debate this question with both Beach and the Hampsons during the early 1960s. In 1961 and 1962, CRPS hosted two con­ferences, the organization’s last actions before going out of the sex-study busi­ness, work by then so fully supported by the National Science Foundation and the National Institute of Mental Health as to render CRPS obsolete. Following the conferences, CRPS ‘‘recommended to the Chairman of the Division of Medical Sciences that the Committee for Research in Problems of Sex be dis­charged when the book resulting from the Conference on Sex and Behavior has been prepared for publication’’ (Beach 1963, p. ix). Beach edited a volume that summarized the two meetings, and it is in this volume that we find Young and Hampson talking to each other, with Beach’s editorial hand clearly egging on the debate. For example, Young addressed John Hampson’s upcoming pa­per: ‘‘By ‘bisexuality’ I do not mean. . . that an individual can move equally well in one direction or the other’’ (here an asterix refers the reader to Hamp – son’s account of neutrality in the upcoming chapter). ‘‘I believe,’’ Young con­tinued, ‘‘that. . . the evidence in the clinical literature’’ and primates ‘‘will reveal a predominance of masculine characteristics in the genetic male, and a predominance of feminine characteristics in the female. . . . Even in human beings before the individual is born the stage’’ may be ‘‘set for selective re­sponsiveness to experiential and psychologic factors’’ (Young 1965, p. 103). Young reiterates this point in Young 1967. Beach twice flagged John Hamp – son’s rejection of the idea of ‘‘sex hormones as a single causal agent in the establishment of an individual’s gender role and psychosexual orientation’’ (p. 115), referring the reader back to Young’s discussion. Hampson concludes ‘‘that an individual’s gender role and orientation as boy or girl, man or woman, does not have an innate, preformed instinctive basis. . . .Instead. . . psychologic sex is undifferentiated at birth—a sexual neutrality one might say—and. . . the individual becomes psychologically differentiated as mas­culine or feminine in the course of the many experiences of growing up’’ (Hampson 1965, p. 119).

101. ‘‘The possibility must be considered that the masculinity or feminin­ity of an animal’s behavior beyond that which is purely sexual has developed in response to certain hormonal substances within the embryo and fetus’’ (Phoe­nix et al. 1959, p. 381; emphasis added).

102. Ibid., p. 381. The possibility that fetal or perinatal estrogen plays a role in the developing female brain remains in dispute to this day. See Fitch and Denenberg 1998; Fitch etal. 1998; Etgen et al. 1990; Fadem 1995; and Ogawa et al. 1997.

103. Van den Wijngaard 1991b.

104. Beatty 1992.

105. By the late 1960s, John Money and Anke Ehrhardt had applied the paradigm to the study of CAH girls (chapter 3). In a popular account of their work, they introduced the idea that prenatal androgen exposure masculinized the brains of XX kids exposed to high levels of testosterone in utero. Just as with the guinea pigs and rats, Money and his colleagues argued that prenatal hormones induced such girls to engage in a more masculine style of play (Money and Ehrhardt 1972). Also in this period, the German endocrinologist Gunther Dorner suggested that the new understandings proffered by the O/ A theory might offer a cure for homosexuality. Citing experiments showing that perinatal castration seemed to prevent masculinization of a rat’s brain, Dorner hoped that the same might be true for humans. ‘‘These results,’’ he wrote, ‘‘suggest. . . that male homosexuality may be prevented by androgen administration during the critical period’’ (Dorner and Hinz 1968, p. 388).

106. Young 1961, p. 1,223. Young wrote of the role of genes on female behavior: ‘‘As in the male, differences were seen in every measure of behavior studied: responsiveness to [hormone] treatment. . . , duration of induced heat. . . duration of maximal lordosis, and amount of male-like mounting’’ (Young 1961, p. I,2I^).

In order to obtain usable data, scientists often made their experimental animals more uniform. In one sense, then, working scientists produced a typi­cal account of sexual behaviors by systematically eliminating genetic diversity from their studies. A brief recent article on contemporary, commercially pro­duced laboratory rats notes that commercial companies have selected them to breed as rapidly as possible (thus increasing profit margins). As a result, they now average almost double what they used to weigh twenty years ago, and they die much younger. There is little doubt that this selective breeding has changed the physiology of our ‘‘standard’’ laboratory rat to meet both com­mercial and experimental needs. Thus scientific theories based on these rats—especially, I suspect, ones having to do with energy metabolism—are peculiarly structured to the laboratory. In this sense, we have ‘‘created’’ biol – ogy—i. e., the facts from which we will generalize attempts to devise medi­cines, diet regimes, and theories of biology will come from a peculiar creature that is subject only to human selections, not natural selection (see Wassersug 1996; Clause 1993).

Young cited especially his experiments on social isolation. These showed that for one genetic strain the development of mounting, intromission, and ejaculatory behaviors depended ‘‘almost completely… on the contact [the animals] had with other young animals’’ (Young 1961 ,p. i, 218).

107. Young 1964, p. 217. Of course, some researchers continued to ac­knowledge the importance ofsocial interactions and experience and to design experiments based on such acknowledgment. This was not, however, the reigning paradigm, and to many inside and outside the field, and to the general public, this other, more complex approach was indeed invisible.

108. Phoenix I978,p. 30.

109. During the 1960s, Beach continued to challenge the O/A and insist on adult bisexuality. He explained the evidence on lordosis by calling on neu­romuscular units developed before birth in both sexes: ‘‘They are present in both sexes and their organization during development is not dependent upon gonadal hormones’’ (Beach 1966, p. 532). As the male matures, the reflexes come under inhibitory influences that a variety of environmental circum­stances can release.

110. Money and Ehrhardt (1972) were anxious about the judgment of women’s liberationists, who, they noted, were not going to like all they had to say. Their book also has an odd index entry. Under ‘‘Women’s Liberation: Quotable Material,’’ they list the page locations of items that they apparently felt would bolster the feminist viewpoint (see p. 310). The psychologist Rich­ard Doty wrote a paper in which he called on researchers to extend more ‘‘equal opportunity’’ to female rodents (Doty 1974, p.169), while the psy­

chologist Richard Whalen expressed concern about whether his theories of gender formation in rodents were ‘‘sexist.’’ See, e. g., Whalen 1974, p. 468. In a symposium held in 1976 in honor of Beach’s sixty-fifth birthday, his student Leonore Tiefer infuriated Beach with a talk offering a feminist perspective on contemporary research. Later, when Beach read the piece, he apologized and suggested that her viewpoint was indeed worth listening to. See Tiefer 1978 and van den Wijngaard 1991.

in. During its first ten years of publication, Hormones and Behavior de­voted fully 80 percent of its research articles to hormones and gendered be­havior.

Supporters of the organizational theory, Beach suggested (1971), had got­ten carried away with the embryological metaphor, but Young and his follow­ers could not specify what, exactly, was being organized. He also found the notion that androgen organizes a brain (a male one, at that) problematic, sug­gesting that castration would thus disorganize the brain (and the female brain, at that). What, he wondered, could a disorganized brain imply? He pressed home his point with a doctored photo purporting to show disorganized neural pathways. He expressed concern about the loss of ‘‘hard-earned knowledge regarding relationships between gonadal hormones and behaviors. Many the­orists are so sadly and seriously affected with neurophilia (which in its termi­nal phases inevitably develops into cerebromania) that they are able seriously to entertain only those interpretations of behavior couched in the vocabulary of the neurologist’’ (Beach i97i, p. 286).

The published work offered a concentrated dose of his famous acid wit. But rather than burning a hole into the heart of the organizational theory, his words fell, I gather (from corresponding with some who were there), on somewhat embarrassed ears. Beach understood what the reception would be. He wrote: ‘‘No one is more fully aware than I that many readers will feel that I am tilting windmills’’ (Beach i97i, p. 291).

112. In his own history of the field, Beach carefully works his way around his earlier objections without citing the Beach i97i paper (Beach i98i). The remarkable nature of Beach’s silence can be seen in McGill et al. i978. This 436-page volume celebrating Beach’s sixty-fifth birthday contained articles on the current research of at least seventeen of his former students. Only one even referred to Beach’s critical article, and then only to mention a particular fact, not the critique itself.

There are, of course, microexplanations: (i ) a new breed of biochemists was taking over, and Beach knew little of the molecular approach, so he was off base, but his respectful junior colleagues were too kind to tell him so in public; (2) his article was so intemperate that it exceeded acceptable norms of behavior, and people chose to turn the other cheek rather than return the insults.

In addition to attacking linguistic ambiguities, Beach considered alternate explanations for results of experiments on early hormone treatment. He looked particularly at the claim that testosterone organized male and female copulatory behavior. He noted that androgen strongly affected the postnatal growth of the penis. Thus males castrated in infancy might later fail to achieve intromission and ejaculation because their penises were too small, not because their brains had failed to be masculinized. For more on the penis-size debate, see Beach and Nucci 1970; Phoenix et al. 1976; and Grady et al. 1965. In general, he argued that many of the experimental results achieved could have resulted from effects on the peripheral nervous system or genitalia rather than on the central nervous system. See, for example, Beach and Nucci 1970. In 1968, the first evidence suggesting that the brain was at least one component of the central nervous system involved in organizing behavior was published by Nadler. During the 1970s and early 1980s, additional evidence accumu­lated on this point. See Christensen and Gorski 1978; Hamilton et al. 1981; and Arendash and Gorski 1982. (Thanks to Elizabeth Adkins-Regan for this chronology.)

Beach further insisted that whatever the effects of early hormones in males, they did not permanently wipe out the neural connections needed to express lordosis. Perhaps, as he had suggested earlier, prenatal hormones changed the level of sensitivity of nerve cells to later hormone stimulation. But the meta­phor of mutually exclusive permanent electronic circuits (male or female) seemed untenable. Beach cited a study of male rats castrated as adults. Ac­cording to the O/A theory, these males should not exhibit lordosis even when stimulated by estrus-inducing hormones because their brains had been prop­erly masculinized in and around the time of birth. Indeed, normal quantities of estrogen did not elicit lordosis. However, a more prolonged series of injec­tions induced these castrated males to exhibit lordosis almost as frequently as normal females in estrus. He wrote: ‘‘It becomes increasingly apparent that neural mechanisms capable of mediating lordosis and possibly ancillary re­ceptive responses as well are organized in the central nervous system of male rats despite the presence of testis hormone during prenatal and early postnatal periods’’ (Beach i97i, p. 267).

113. Ibid., p. 270.

114. Beach i976,p. 261.

115. Beach and Orndoff 1974; Beach 1976.

116. Hart (1972) concluded that manipulation of neonatal androgen affected both penile development and the central nervous system.

117. Raisman and Field 1973.

118. Goy and McEwen 1980 ,p. 18. The conference that led to this book took place in i977.

119. Beach 1975.

120. Federi98i, p. 141.

121. Evaluating Beach’s critique in 1990, Michael Baum wrote:

Ironically, Beach’s warning that we should resist the temptation to attri­bute all steroid-induced changes in behavioral potential to structural changes in the central nervous system still has some merit in the beginning of the 1990’s. . . . While at present most workers would agree that the developmental effects of androgen on masculine coital responsiveness to adult steroids probably reflect a change in the nervous system, such behav­ioral changes cannot be localized in any of the rather limited current in­ventory of sexually dimorphic brain structures of the various mammalian species studied to date. Furthermore, some aspects of steroid-induced changes in mating potential may, as Beach predicted, result from the indi­rect perinatal action of androgens on the developing masculine genital organs.’’ (Baum 1990, pp. 204—5)

Balthazart et al. echo Baum’s point, writing, ‘‘in all model species… it is still impossible to identify satisfactorily brain characteristics that differentiate under early steroid action and explain the sex differences in behavioral activat­ing effects of steroids’’ (1996, p. 627). Cooke et al. (1998) and Schlinger (1998) make similar points.

122. For a good overview of these changes, see Chafe 1991. For specific information on the history of the U. S. gay liberation movement, see D’E – milio 1983.

123. Money and Ehrhardt i97 2,p. xi.

124. Doty 1974. Doty also noted that the sense of smell might be a key aspect of mating behavior totally unobserved by studies relying on visual com­ponents of behavior. One implication: some hormone effects might be medi­ated by changes in odor or odor responsiveness, rather than changes in the brain or central nervous system. This concern paralleled Beach’s interest in hormonal effects on peripheral sensory systems.

125. Doty was not the first to develop such a critique. Whalen and Nadler, for example, had called for better experimental definition of female receptiv­ity: ‘‘If receptivity is defined by the presence of spermatozoa in the vagina, some estrogen-treated females are receptive. If receptivity is defined by the rapid and easy elicitation of the lordosis response, spontaneous and hormone – induced receptivity is suppressed’’ (1965, p. 152). Whalen continued his methodological critiques during the 1970s. See, for example, Whalen 1976.

126. De Jonge 1995, p. 2. If a female is not in estrus, not even a much larger male can succeed in mating with her. Several researchers have empha­sized to me that a rodent male cannot succeed in mating with an unwilling female and that in some species a female may attack and even kill an unwel­come suitor.

127. Clark 1993b, p. 37. In the wild, an unwilling female hides in her burrow, while the interested male tries to entice her to emerge. In her test cage, with no possible escape, a female may respond aggressively, screaming and biting the male (Calhoun 1962; deJonge 1993).

128. Speaking at a symposium in honor of Young, held shortly after Young’s death, Beach noted the difficulty of proving the absence of a particular neural representation (Beach 1968). This seems to be a good example of his point. Under some circumstances the lordosis response was absent and pre­sumed missing because the neural substrate needed for it had been suppressed in early testosterone treatment. But under some experimental circumstances a positive result—frequent lordosis—appeared, thus suggesting that the neu­ral substrate was there after all.

129. Gorski i97i, p. 231.

130. ‘‘The rapidly increasing precision and sophistication in endocrino­logical techniques,’’ he wrote, ‘‘have not been accompanied by comparable advances in the definition and measurement of behavioral variables’’ (Beach I976, p. I03).

131. More recent work shows quite clearly that proceptive and receptive behaviors respond to different activating hormones in adulthood (de Jonge 1986; Clark 1993).

132. In a later (1977) paper, Madlafousek and Hlinak offered a thick de­scription—to borrow an anthropological term—of the various aspects of a female rat’s behavior as she proceeded through estrus. (A ‘‘thick description’’ offers a lot of detail out of which a nuanced interpretation is thought to emerge.)

133. Whalen 1974; Davis et al. 1979.

134. Whalen and Johnson 1990.

133. Bem 1974. The parallel between the 1974 Bem and Whalen publica­tions is striking: each noted the independence of masculinity and femininity. Whalen writes: ‘‘Bem and I had no contact about the ideas that we put forth at the time. The time must have been right’’ (personal communication, Sep­tember 19, 1996). SandraL. Bem writes: ‘‘I think the Zeitgeist. . . is probably another hypothesis that must be considered in addition to direct contact. . . . I’m quite sure that I had never met or talked to Whalen in the time period you’re asking about’’ (personal communication, September 28, 1996).

136. Goy and McEwen 1980 ,pp. 3, 6. They noted the new respectability afforded hormone research: ‘‘While there is still reasonable and serious dis­pute regarding the biological cause of different organizations of sexuality. . . hormonal hypotheses have earned a respectability that allows their inspection even for problems of human sexual behavior, a permission that was not readily granted by clinical workers a few decades ago.’’

137. Brain cells contain an enzyme called aromatase, which transforms testosterone into estrogen. Recent studies show that the hypothalamus of de­veloping male mice contains higher activities of aromatase than does that of developing female brains. This implies that some masculine behaviors may result from higher concentrations of estrogen in male than in female brains! The aromatase enzyme system is not distributed uniformly throughout the brain, and the multiple and complex roles of the sex steroids in their various molecular incarnations, as well as the enzymes that transform them and the various brain regions that contribute to their synthesis, still awaits some uni­form understanding or unifying hypothesis. See, for example, Naftolin et al. 1971; Naftolin et al. 1972; Naftolin and Ryan 1975; Naftolin and Brawer 1978; Naftolin and MacLusky 1984; and Hutchison et al. 1994.

While the conversion hypothesis produced a small tidal wave of research on estrogen production by various organs in the male, only a very small num­ber ofresearchers seem to have noticed that the results ought also to call for a reevaluation of the presence/absence hypothesis of male and female develop­ment. In 1978 one researcher raised the question ‘‘Is female sexual differenti­ation hormone-mediated?’’ and again in 1984, another pointed out that ‘‘sex­ual differentiation in males and females is hormone dependent (Dohler 1976, 1978; Dohler etal. 1984; Toran-Allerand 1984; emphasis in original).

138. Bell et al. 1981.

139. See, for example, the mixture of articles in Young and Corner 1961 or de Vries etal. 1984.

140. Beach emphasized the normal status of female mounting and urged it be studied as a typical female behavior. He also reasoned that humans had the neural mechanisms needed for same-sex attraction, although he thought that exclusively homosexual attractions resulted from the complexities ofcul – ture and experience (Beach 1968).

141. Kinsey et al. write: ‘‘Several investigators (Ball, Beach, Stone, Young et al.) have shown that the injection of gonadal hormones may modify the frequency with which an animal shows an inversion of behavior. . . . Among many clinicians this work has been taken to mean that the sex hormones con­trol the heterosexuality or homosexuality of an individual’s behavior. This, of course, is a totally unwarranted interpretation’’ (Kinsey etal. 1948, p. 613).

142. This is a culturally specific attitude. In many Latin American cul­tures, for example, only the receptive male is understood to be homosexual.

143. Nothing stirred up this debate more than Simon LeVay’s 1991 publi­cation. See also LeVay 1991; Byne and Parsons 1993; Byne 1993.

144. Adkins-Regan 1988. She noted that this distinction was often lost on medical researchers applying animal results to humans, despite many animal researchers having clearly articulated it in the past. See esp. p. 336 for this dis­cussion.

143. In one study, researchers removed the ovaries of adult females and then injected them with testosterone that had been chemically altered to pre­vent its conversion to estrogen, or progesterone. Female rats treated with the altered testosterone preferred to mate with males, but had no lordosis response, while progesterone facilitated both receptive (lordosis) and precep­tive (hopping and darting) behaviors, but did not induce male sexual prefer­ence. Thus in female rats, the mechanisms for sexual preference and actual mating behaviors differ. Furthermore, prenatal androgens seem to have no effect on the sexual orientation of female rats. Rather, the adult hormonal environment interacts with the rats’ prior experience (de Jonge et al. 1986; de Jonge et al. 1988; Brand et al. 1991; Brand and Slob 1991a and 1991b).

146. Francien de Jonge and her co-workers removed the ovaries of adult female rats, some ofwhich had had prior sexual experience and some ofwhich had not. They then induced sexual behaviors by injecting testosterone (or, for controls, plain oil). Inexperienced females preferred the company of males when they got testosterone but showed no preference without it, while fe­males with prior mount experience with other females continued to prefer females regardless of whether they received oil or testosterone. If, instead, their prior experience had been with males, they subsequently showed no particular sexual preference (de Jonge et al. 1986). Although adult hormones and prior experience seem to be the keys to female laboratory rat sexual pref­erence, in male lab rats, prenatal hormones assume a greater importance. Julie Bakker completed a series of experiments showing that male rats for whom the conversion of testosterone into estrogen is blocked at birth later develop strongly bisexual or asexual potentials. If left intact and put on the right kind of light/dark cycle, they will run back and forth between test males and test females, exhibiting both altered mating behaviors and altered prefer­ences. In adulthood, estrogen induces homosexual preferences in such males, while testosterone seems to permit greater bisexuality (Bakker 1996). Bakker also showed that, for males, social isolation from the moment of weaning to adulthood had no effect on sexual preference, although such isolation drasti­cally impaired sexual performance. Adult social interactions, however, did affect male sexual preference. Aromatase inhibitor-treated rats required phys­ical interactions with their potential partners in order to differentiate them­selves from control males. Although I have primarily used Bakker’s Ph. D. the­sis to write this section, much of her work also appears in the following publications: Brand and Slob 1991a and 1991b; Brand et al. 1991; Bakker et al. 1995a; Bakker, Brand, et al. 1993; Bakker, van Ophemert, et al. 1993; Bakker, 1995; and Bakker et al. 1994.

147. See, for example, LeVay 1996.

148. Schlinger 1998.

149. Wallen 1996.

150. The psychologist Gilbert Gottlieb (1997) summarizes his lifetime of experiments on the development of bird behaviors such as imprinting and applies the tradition of systems theory to his results. It’s a good read!

151. Ward 1992.

152. See, for example, Houtsmuller et al. 1994. There is a fairly large literature on the effects of location in the uterus on future behavior.

153. Gottlieb 1997.

134. Laviola and Alleva 1995.

155. Harris and Levine 1965.

156. DeJongeetal. 1988.

157. Harris and Levine 1965.

158. Feder1981.

159. Gerall et al. 1967; Valenstein and Young 1955; Hard and Larsson 1968; Thor and Holloway 1984; and Birke 1989.

160. For example, when nonovulating female rats were housed with sex­ually experienced males, they would not mate. But after 3 months of continu­ous cohabitation, 18 out of 60 of these females responded to male mounting (Segal and Johnson, cited in Harris and Levine 1965).

161. Ward 1992.

162. Moore et al. 1992. Moore describes the effects of early testosterone treatment as either a web or a cascade. Her model has no linear connections. The number of affected organs grows as hormones influence the scent glands and the brain early in the process and subsequently alter liver physiology, geni­tal anatomy, and muscle development. Finally maternal licking, overall body size, play, exploration, and self-grooming behaviors all interact with hor­monal effects. Thus, behavior results from the intersection of links among physiology, anatomy, and behavior. For example, maternal licking causes and is caused by the interrelationships between pup odor, pup urine production and retention, pup leg-extension behavior, maternal water and salt balance related to lactation and attraction to pup odor. The relationships are complex and decentralized. Hormones become part of a web that includes—among other things—experience, the brain, peripheral muscles, and general physi­ology (Moore and Rogers 1984; Moore 1990).

163. Drickamer 1992.

164. Moore and Rogers 1984; Moore 1990.

165. Arnold and Breedlove 1985.

166. Breedlove 1997, p.801. There are other hormone effects as well. Prenatal or perinatal testosterone treatment lowers thyroid function, affects the liver, and causes a wide variety of reproductive system abnormalities (Moore and Rogers 1984; Moore 1990; Harris and Levine 1965; de Jonge et al. 1988; de Jonge 1986).

167. Sodersten describes a strain of rats in which intact males exhibit sig­nificant levels of lordosis, often considered to be an exclusively female behav­ior, while van de Poll and colleagues report on one showing no hormonally induced alterations in aggressive behavior. Finally, Luttge and Hall and McGill and Haynes discuss strain differences in how mice respond to testosterone treatment (van de Poll et al. 1981; Sodersten 1976; McGill and Haynes 1973; Luttge and Hall 1973).

168. See, for example, Calhoun 1962; Berry and Bronson 1992; Smith, Hurst etal. 1994.

169. Geralletal. 1973.

170. Sodersten 1976.

171. Adkins-Regan et al. 1989.

172. De Jonge et al. 1988. This result is consistent with the report that an ovary present around puberty in either male or female rats facilitated the appearance of female behavior when the animals were examined as adults (Gerall et al. 1973).

173. Tobet and Fox 1992.

174. Toran-Allerand 1984, p. 63; emphasis added.

173. One correspondent who read this comment scoffed, suggesting it would be a waste of time to do long-term studies, since he was certain the outcome wouldn’t change. Given the current explosion of information on neural plasticity, I believe that long-term studies that manipulate environ­mental variables are quite appropriate.

176. Brown-Grant 1974.

177. Beach 1971.

178. Feder i98i, p. 143.

179. Arnold and Breedlove 1983.

180. Thor and Holloway 1984 review work on social play in juvenile rats.

181. The pituitary of adult female rats, for example, controls the repro­ductive cycle with periodic or cyclical secretions. In contrast, the male rat pituitary controls reproduction with a constant flow of hormones. Perinatal testosterone seems to permanently suppresses cyclicity in treated females, while castration of newborn males results in adults with a cyclically function­ing pituitary (Harris and Levine 1963). In primates, however, prenatal hor­monal effects on pituitary function are not permanent. Thus the development of sex differences in pituitary physiology differs in rats and primates. In the latter group, functional modulation in adulthood is possible (Baum 1979).

182. Feder 1981; Adkins-Regan 1988.