Top Quotes: “Brain Storm: The Flaws in the Science of Sex Differences” — Rebecca Jordan-Young

“Most books set out to answer questions. This book sets out to question answers. The answers I question have to do with the nature and causes of differences between men and women, and between straight people and gay people. Specifically, I question what we ‘know’ about male and female brains, or gay and straight brains.

When Simon LeVay reported in 1991 that he’d found a difference in brain structure between gay and straight men, which was trumpeted as the discovery of ‘The Gay Brain,’ I found it interesting but also puzzling. How could gayness take on a single identifiable form in the brain when it takes such varied forms in people’s lives?

At the time, I ran a project on injection drug users and first met a lot of gay men. These men were not the poster children of the gay rights movement, but were poor, struggling with addiction and recovery, and trying to avoid or outlive AIDS. And so were their straight brothers, with whom I also worked. In fact, these men were so similar in demeanor, dress, and daily struggles that without our detailed interviews, it was impossible to tell the difference between the gay and straight men. In LeVay’s study, the gay men were a singular ‘type’ unlike the (also homogeneous) straight men, and they were also somehow similar to (presumably straight) women. Those are common enough ideas both inside and outside of science, but our research challenged these notions.

The reason for this was our research methods. Since our main question didn’t have to do with sexual orientation, our preconceptions about sexual orientation didn’t shape the way we gathered our info. We asked people to join the study based on their drug use or their connection with drug users, not their sexuality — nor their sexual practices or identities, not the gay or straight reputation of the bars or neighborhoods they hung out in, not even how we perceived them as straight or gay. We asked everyone extremely detailed questions about their sexual practices with both same-sex and other-sex partners. Scientifically, this approach is a much more reliable way to get info about the nature of sexuality in a population than to go out looking for gay and straight people to compare. The number and variety of people who talked about same-sex relationships surprised all of us — including the gay and lesbian staff members who thought we had finely tuned ‘gaydar.’ The bias in epidemiology at the time was to see same-sex behavior among people outside of well-defined gay communities as being ‘instrumental’ — meaning that it was due to drug use, incarceration, or sex work. But as we spent hours and weeks and eventually years with our participants, growing to know and love many of them, it was clear that they had same-sex relationships for the same reasons they had straight relationships: desire, affection, and love. The thousand or so sexual histories we gathered fed into an enormous pool of research that eventually included info on the sexual and drug-using behavior of tens of thousands of people from 50+ cities. And on the basis of this research, it would be very hard to suggest that either men who desire and have sex with men, or women who desire and have sex with women, are a distinct type of person, or are somehow ‘like’ heterosexuals of the other sex.”

“In a text that deserves much wider audience than it’s received, Oudshoorn (1994) advanced the somewhat astonishing claim that ‘sex hormones’ were ‘invented’ by 20th-century endocrinologists. Oudshoorn doesn’t dispute that steroid hormones like estradiol, testosterone, or progesterone are actual material substances, nor does she claim that these substances were not already being produced and used by bodies long before scientists had any notions about them. Her argument specifically concerns the idea that certain steroid hormones are fundamentally about and for sex. She documents how preexisting ideas of masculinity and femininity caused scientists to look for, create tests for, classify, and perceive steroid hormones in a way that fit them into a dualistic system of sex. This might not be particularly compelling if she didn’t also document how this commitment to a fundamentally sexual classification of hormones systematically blocked some kinds of info. For example, scientists had repeated difficulty assimilating the info that both males and females produce and use both androgens (‘male sex hormones’) and estrogens (‘female sex hormones’). Likewise, biologist Marianne van der Wijngaard (1997) has documented the great difficulty mid-20th century experimental psychologists had with the info that estrogen — the quintessential ‘female’ hormone — was even more important than testosterone for the development of certain ‘masculine’ characteristics. Van den Wijngaard and Anne Fausto-Sterling (2000) have both shown how a dualistic notion of ‘sex hormones’ has been a repeated stumbling block in recognizing the great variety of functions that steroid hormones accomplish, apart from those related to development and function of reproductive capabilities. In this vein, a colleague and I recently confirmed that attachment to the term and concept sex hormones is associated with systematic misinformation in widely-used high school biology texts (Nehm and Young 2008).”

“Kessler’s groundbreaking analysis showed how ideas about gender and sexuality are called upon by doctors and psychologists, both to determine the male or female status of bodily structures and to reshape those structures so that they can be seen as fitting with ‘normal’ male vs. female bodies. For example, Kessler explores the centrality of the idea that ‘boys urinate standing up’ in determining whether a phallic structure can, or cannot, count as male. She and, more recenlty, Katrina Karkazis (2008) have also demonstrated how doctors who conduct ‘feminizing’ surgeries on intersex infants deploy a definition of ‘normal’ female sexual function that’s built around vaginal penetrability rather than around the capacity for sexual pleasure and orgasm. Together with analyses of biology and endocrinology, these studies add up to a take-home message that gender causes us to perceive the natural world (the body) in a particular way, and thereby to impose upon it the dichotomous category ‘sex.’ Sex, then, is no longer the raw material from which culture produces gender. Indeed, sex is in some important sense an effect of gender.

In some sense, the entire foundation of brain organization theory takes the idea that gender and sex proceed from biological processes as a starting point, and researchers devise studies that are intended to demonstrate a causal trajectory from the sexed body to the gendered experience.”

“As the 1950s drew to a close, much of the data regarding hormone effects on behavior seemed disjointed and even contradictory. In humans, a great deal of work has gone into investigating hormones as an explanation for homosexuality. The first formal studies in this vein were conducted by Viennese physiologist Eugen Steinach, whose innovative experiments with rats and guinea pigs ‘marked the beginning of modern experiments on the role of hormones in sexual differentiation.’ In the prevailing understanding at the time, homosexuality was an ‘inversion,’ meaning that the mind and soul of gay men was feminine and that of lesbians was masculine. Although others before him had proposed a biological basis for homosexuality, Steinach suggested that male homosexuality could be traced to aberrant cells in the testes that produced ‘female’ hormones. Steinach’s role as a physician as well as an experimental physiologist made it fairly simple to test his theory via ‘therapeutic’ interventions with humans. Under his guidance, 7 gay men each had one of his own testicles removed and replaced with a testicle from a heterosexual man (the latter having been removed because it was undescended). Initial enthusiasm for the transplants was short-lived, as the ‘heterosexual’ testes failed to stimulate heterosexual desire in the recipients.

Steinach’s transplants didn’t work, but the idea that gay men had more ‘feminine’ hormone profiles continued to be quite popular with scientists. Proponents of the hormonal inversion idea were again thwarted by the dismal failure of gay men to become straight in response to testosterone injections, and by the gradual accretion of evidence that studies seeming to support a ‘feminine’ hormonal profile among gay men were badly flawed.”

“The contemporary idea that men have a much greater sexual appetite than women is the exact reverse of the idea in Renaissance Europe that women were generally sexually insatiable and men were much more capable of controlling their ‘base desires’ with their naturally greater rationality. Another example concerns the type of conclusions one might draw about the sexual habits of a man who’s extremely concerned with his appearance and dress. According to current American stereotype, the man is likely to be gay; but according to the common wisdom of early modern Europe, he was likely to be a womanizer. Interestingly, being a womanizer in this earlier period was considered effeminate, because it signaled a lack of manly self-control. So much for the popular idea that a desire for lots of sex and multiple partners has always and everywhere been considered masculine!”

“Young men’s descriptions, from the 1950s and 1960s, of the emotional content of their sexual desires and behavior are a particularly compelling indicator of just how much changed in the second half of the century. The current popular and scientific view holds that male sexual desire is largely, if not entirely, independent of concerns for love and intimacy, but this ‘masculine’ separation of lust and emotional connection was demonstrably less pronounced just a few decades ago. In interviews about ‘premarital’ sex, some young men justified their sexual activities outside of marriage with reference to emotions, especially love: ‘We didn’t believe in petting because of the sex alone, but because we were very much in love and this was a means of expressing our love to each other.’ Others felt that it was disrespectful to make sexual advances to ‘nice’ girls, even if that meant settling for little or no sex themselves. As one college male explained, ‘the fact that I was never sexual with anyone is because I never dated anyone I didn’t care for.’”

“What, specifically, is wrong with the fact that brain organization researchers came to expect relatively high libido, frequent sex, multiple partners, masturbation, sexual night dreams, and initiation of sexual activity from sexually normal women? Updating their assumptions about feminine sexuality might seem to be a fairly straightforward matter of scientific progress — especially when the later model seems to fit better with contemporary women’s sexual behavior, and when many early studies of brain organization have been criticized as sexist due to their outmoded assumptions. So why the critique? Aren’t the changes relayed in the preceding discussion a good thing? Unfortunately, it’s not that simple. Newer and more complex conceptualizations of feminine sexuality may be an improvement in abstract theory, but unfortunately they make it difficult to grasp the meaning of research underlying brain organization theory. We might read that 2 different studies have shown that prenatal androgens ‘masculinize’ women’s sexuality, but we can’t be confident that those 2 studies are talking about the same thing — in fact, there’s a good chance the studies directly contradict each other. My point is not that updating the definitions is bad, but that updating the definitions has serious consequences for how brain organization studies can be compared to see if the theory is supported, on the whole, by the evidence.

I don’t believe that scientists shifted definitions on purpose in order to make their favored theory look better. I’m convinced that they’re totally unaware of the transformation this key concept has undergone over time in their studies. These scientists — like most people, probably — tend to think that masculine and feminine sexuality is a no-brainer. My sense is that over time, as sexual-revolution ideas about women’s sexuality became increasingly mainstream, scientists’ definitions changed without their even realizing it. This was an especially comfortable and ‘natural’-feeling change because of the ways that data on DES-exposed women seemed to fit hypotheses with the newer definition of feminine sexuality. But regardless of why the definition of feminine sexuality changes or whether scientists are aware of it, one we notice the change, we have to acknowledge that studies with different definitions of feminine sexuality generate irreconcilable evidence about the theory.

According to symmetry principles, studies with definitions that are starkly different or even contradictory can’t be lumped together as evidence of how hormones affect feminine sexuality. The general term feminine sexuality is used to gloss divergent — and in some cases diametrically opposed — concepts in the research. This level of departure from earlier definitions veers dangerously close to making brain organization literally unfalsifiable. It begins to look as though any difference from a comparison group could be interpreted as in keeping with the idea that CAH-affected women are masculinized, or DES-exposed women are defeminized, and so on. Further, because scientists have not noticed the changes and confronted them directly, they’ve both reinforced the notion that ‘masculine’ and ‘feminine’ sexuality are universal, timeless constructs and created the illusion of a seamless line of evidence supporting human sexuality as hardwired by hormones.”

“The point is not that some scientists choose the ‘wrong’ measures for sexual orientation. The point, instead, is that the measures they choose often put them at odds with one another. As a result, their studies just don't add up to the consistent findings about brain organization that are commonly claimed. In fact, a good many of the studies supposedly showing that prenatal hormones influence sexual orientation contradict other studies that supposedly show the same thing. Between slippery measures and other problems, such as illogical dose-response inferences about hormone effects, the overall evidence that early hormones affect sexual orientation is murky, at best.

Measurement problems amplify some breaches in evidence that appear when comparing the major subsets of studies. That is, some studies start with hormones, and other studies start with endpoints (sexual orientation). Somehow the evidence from all these studies should meet up, but it doesn’t. In part because researchers have very different notions of what counts as homosexual (usually related to what measures will give ‘better’ results in the individual studies), it’s impossible to get the findings to align in a way that supports brain organization theory.”

“A 3-part test for ‘male orientation’ among rams consists of the following:

Only rams that would not mount estrous females in the preliminary tests satisfied the criterion for entrance into the Sexual Preference Paradigm…Briefly, in November and December of the second year of life, rams were exposed simultaneously to 2 restrained estrous females and two males that were chosen at random for use. Rams that courted and mounted males in preference to females during a 30-min test that was repeated at least 3 times were classified as male-oriented. Male-oriented rams were given an additional preference test 5 days before they were killed. Using these procedures, we identified 6 males that wouldn’t mount females in the preliminary sexual tests, mounted males in a group setting, and mounted males in preference to females in the sexual preference tests.

It turns out, though, that most, if not all, of these ‘male-oriented’ rams are having sex with ewes when the scientists aren’t looking. In a study that was designed to test the relationship between successful reproduction (siring of lambs) and behavioral measures of ‘sexual performance,’ 5 male-oriented rams impregnated at least 330 ewes over 21 days, resulting in 480+ offspring. The number of ewes they impregnated and the number of lambs sired fell roughly midway between the groups of rams whose behavioral tests had classified them as ‘high-performing female-oriented’ and ‘low-performing female-oriented.’ In behavioral tests after breeding, only 3 of the 5 ‘male-oriented’ rams exhibited any sexual activity, and 2 of the 3 continued to show a strong (though not exclusive) preference for mounting other rams. Offering the understated conclusion that ‘individual sexual partner preference tests didn’t absolutely reflect the breeding performance of the male-oriented rams,’ the scientists speculated that aspects of the testing environment differed sufficiently from the ‘competitive breeding environment’ to make the preference test an imperfect predictor of rams’ mating behavior.

It’s not an easy task to come up with good animal models for partner preferences, and much research focuses on animals’ sexual positions rather than on preferred category of partners. Second, even when sexual preference tests in animals focus on identifying partner preferences, the preferences that look like a ‘stable, dispositional trait’ in 1 context look unstable and fluid in other contexts.”

“Subject groups were split into homosexuals vs. heterosexuals by different scientists at literally every point along the Kinsey Scale between 0 and 6.”

“Researchers who study people with unusual prenatal hormone exposures (intersex individuals and the offspring of hormone-treated pregnancies) measure adult sexual orientation in very finely graded ways that can pick up the slightest difference between hormone-exposed and unexposed groups. But studies that look for evidence that gay men or lesbians have had different prenatal hormone exposures than their straight counterparts seek to create comparison groups that are as starkly different as possible in terms of sexual orientation, so they choose ‘extreme’ homosexuals and ‘extreme’ heterosexuals. It’s easier in the former group (the cohort studies) to find hormone effects with a very loose criterion for homosexuality, while the latter (case-control) approach will benefit from a very restrictive criterion. It’s impossible not to see this as scientific gerrymandering — that is, moving the boundaries of sexual orientation categories around in a way that favors the scientist’s own hypothesis.”

“Of all the people who have any same-sex experiences as adults, less than 20% of men and just 10% of women have been exclusively homosexual in behavior since age 18. Among those who report same-sex attractions and/or behavior, 85% of women and 72% of men identify as heterosexual. Brain organization researchers are unaware of this issue, at least in terms of their female samples. In the lit on biology and sexual orientation, as well as in the interviews I conducted with prominent scientists, there was general agreement that women are even less likely than men to be totally consistent among the various dimensions of orientation.

One result of filtering out ‘inconsistent’ responses is that many brain organization studies exclude the majority of self-described lesbians and many gay men from studies on sexual orientation. This most certainly hampers the utility of brain organization research for elucidating the processes that underlie sexual orientation development. But it’s equally important to note that the process of ‘cleansing’ samples to remove apparent sexual inconsistency results in artificially homogeneous ‘heterosexual’ groups, as well. This creates a serious problem of bias. In an artificially homogeneous group, other traits besides sexual orientation will be present to a degree that one would not see in a truly random group of ‘heterosexuals’ or ‘gay men’ or ‘lesbians. When we analyze the data, there will be irrelevant variables present that can make it seem like there are group differences related to sexual orientation, when in fact there’s another hidden variable at play. A historical example will help clarify the problem. Prior to the 1980s, it was common to recruit samples of gay men and lesbains in gay bars, since there were few other public meeting places for sexual minorities in most communities. Unsurprisingly, these samples showed higher rates of problem drinking than heterosexual people who were recruited from a wide range of community settings.”

“Sensibly, in the precise passages where [cohort study] scientists discuss subtle differences in sexual orientation, they tend to avoid attaching labels such as heterosexual or homosexual to their subjects. Yet this careful approach drops away quickly — sometimes even in the article abstracts and titles that scientists themselves choose.

Thus, in the sections where most readers look for the ‘bottom line,’ it seems that the exhaustive assessments work on a sort of sexual ‘one-drop rule’: any hint, in any one of the elements, of same-sex attraction or response is glossed as evidence of homosexuality. It’s little wonder, then, that by the time studies are reported in the popular press, terms like homosexual and lesbian are used freely.

An interesting example is found in Meyer-Bahlburg and colleagues’ report on women exposed to DES: the sections on methods and results carefully refer to the women in this study by their dichotomized Kinsey scores of 0–1 and 2–6, while the title and intro section address the development of ‘homosexual orientation.’ In a striking understatement, the authors note that ‘the extent to which bisexuality and homosexuality were increased in DES women was rather modest.’ In fact, only 1 of 90 women was rated as being ‘largely homosexual but incidentally heterosexual’ and none of the 90 was rated as ‘entirely homosexual.’ Does it make sense to talk about the ‘development of homosexual orientation’ based on a sample that many people — including most brain organization researchers — would agree does not include any homosexuals? If we count the 1 woman who was rated as ‘largely homosexual,’ that’s still a rate of lesbianism that’s just 1%, which is lower than most estimates for the general population.

Even if we were to agree that researchers can use whatever definitions they like for ‘homosexual’ or ‘heterosexual,’ as long as they’re clear and consistent in their own research, the yawning gap between the approaches in the 2 main sets of studies creates a problem. Although scientists’ choices of different measures make sense in terms of studies’ internal logic (it gives each kind of study a better chance of finding that there is some kind of hormone effect on sexuality), it’s very difficult to reconcile findings across the studies. Bluntly, you can’t have ‘homosexual’ groups that are mutually exclusive in 2 sets of studies, but then claim that the studies complement or build upon each other. Still, as serious as the measurement discrepancies are between cohort and case-control studies of brain organization, the findings from research on prenatal hormones and human sexual orientation are just as problematic. It’s not just that they don’t add up ‘yet’ because of incomplete data. Rather, the pattern of findings in the hundreds of studies done to date, and especially between the cohort and the case-control studies, suggests that brain organization theory isn’t a good explanation for human sexual orientation.”

CAH

“A very interesting example of how abstract models about boys’ and girls’ toys can be misleading is found in a study of girls with CAH in Sweden. Anna Servin is among the few scientists who used actual observation of play behaviors, rather than interviews or questionnaires, to determine children’s play preferences. They arranged toys that had previously been defined as masculine, feminine, or neutral in a semicircle on the floor, and invited children, one at a time, to play with the toys however they wished. Although girls with CAH spent significantly more time playing with masculine toys, and less time playing with feminine toys, than did the control girls, the most popular toy with all of the girls was a toy coded as masculine: the Lincoln Logs. Noting that ‘twice as much time was sepnt with that toy as with any other toy,’ the scientists addressed the discrepancy by speculating that the character of this particular toy, as well as ‘its novelty to Swedish children,’ makes it inherently time consuming: ‘This toy contains many different parts…and the means of putting the logs together must be figured out.’ This fits the nature of tasks that Baron-Cohen has designated ‘systemizing’ — the sine qua non of ‘boy play.’ But it was the most popular activity among girls.

Somehow the scientists seemed to overlook the fact that the second most popular toy among both sets of girls was also a so-called masculine toy: a garage with 4 cars. (Certainly Swedish children are familiar with cars and garages, so the draw there couldn’t have been the sheer novelty of it all.) This time it isn’t necessarily the idea of ‘systemizing’ that takes a hit. Instead, it’s the idea that ‘motion’ and objects with moving parts are intrinsically interesting to boys rather than girls.

It’s even more interesting to compare the amount of time the children spent playing with these ‘boys’ toys’ compared to the quintessential ‘girls toy’ they were offered: a baby doll with a blanket. On average, the normal control girls spent 3x as long playing with the garage and toy cars as they did playing with the baby doll (The Lincoln Logs-to-doll ratio was more than 6:1). The only ‘girls’ toy’ that was in the ballpark (if you’ll excuse the boyish metaphor) with these boys’ toys was a pair of Barbie and Ken dolls. (I suspect that Barbie and Ken were riding around in some of those cars.) A glance at another result further confounds the designation of toys as ‘masculine’ and ‘feminine.’ Each child was offered the choice of a masculine toy (car), feminine toy (doll), or neutral toy (ball) to keep as a gift. This alone is somewhat remarkable, because other scientists routinely code ball play as masculine. But the real news here is the sad fate, again, of the baby doll: even among the control girls, the boy was chosen least frequently (control girls were about 36% more likely to choose the ball than the doll.)

How should we interpret these results? Servin and colleagues are perfectly correct in asserting that the girls with CAH did spend more time with the so-called boys’ toys, on average (though, as they note, the differences aren’t significant for all of the toys). So there’s a group difference. But what about the idea that ‘boys favor construction and transportation toys, whereas girls favor toys such as dolls’? Obviously, the larger story about sex-typed preferences is damaged by the nitty-gritty details of studies like this.”

“It’s important to recognize that [intersex] surgeries often don’t have the intended effect: somewhere between 1/3 and 2/3 of adult women with CAH have a vaginal opening that’s not adequately large or flexible to permit heterosexual intercourse (should they desire this), and the vast majority of women have either impaired clitoral sensation or no sensation at all. One of the few long-term outcome studies to report on multiple components of sexual function strongly indicates that, for most women with CAH, vaginal penetration is painful. Gastaud and colleagues (2007) found, first, that 37% (13 of 35) said they’d never had heterosexual intercourse with vaginal penetration’: moreover, 81% of the women who did have heterosexual intercourse experienced pain during vaginal penetration. This report also included data on lubrication, orgasm, and overall satisfaction with sex acts. They found severe impairment in all aspects of sexual function in women who’d been born with the most severely affected genitalia; notably, even though these researchers had expected to find lower sexual functioning in this group, they reported that the ‘extent of these differences…was unanticipated.’ Given the extent and the number of ‘feminizing’ surgeries in this group, which is even substantially higher than for the other women with CAH, it’s curious that this finding should come as a surprise. Moreover, the women with CAH who had less severely affected genitalia also reported significantly higher rates of vaginal pain, and less lubrication and orgasm, than the healthy women to whom they were compared. Another study found similar results, even when CAH women were compared, not to healthy controls, but to women with early-diagnosed diabetes mellitus (which may also impair sexual function): CAH women were found to be less sexually experienced in all areas, and reported higher levels of penetration difficulties pain, and lubrication problems, than the diabetic women.

While the idea that women with CAH are more likely to be bi or lesbian is perhaps the most widely popularized finding from studies of sexuality in this condition, the evidence has long indicated that the most profound difference in sexuality between women with CAH and unaffected women is that the former have very low rates of sexual activity and partnerships across the board, as well as lower levels of sexual arousability, genital sensation, and orgasmic capacity. In a very large study that’s now 20+ years old, Mulaikal and colleagues found that the sexual orientation and activity women reported was more closely related to their vaginal condition than to the degree of prenatal androgen. This isn’t surprising, given that heterosexual eroticism typically centers on penile-vaginal intercourse. Whether this is simply taken for granted, as the normal and natural state of affairs (as is the case in virtually all brain-organization research), or bemoaned as a male-centered definition of sexuality that women can and should challenge, there’s virtually no one who questions that intercourse is, in this time and place, the very definition of ‘sex’ for heterosexuals. Is it so surprising, then, that women who, as a group, have extremely low rates of ‘typical’ vaginal function would also have somewhat increased rates of at least fantasizing about female partners, where sexual intimacy wouldn’t necessarily hinge on the depth and flexibility of the vagina? And that’s only to consider the physiological effect on genitals that medical treatment of CAH has.

What about the psychological effects? In 1 follow-up study, women with CAH ‘used a language of rape, invasion, and violation when talking about vaginal exams and other procedures carried out during visits to pediatric and adult clinics.’ This is consistent with the stories that intersex adults as well as the parents of intersex children told to anthropologist Katrina Karkazis. Medical visits every 3–4 months are often considered necessary to monitor children’s hormone levels as well as their response to treatment. As Karkazis documents, girls with CAH, as well as their parents, often experience the genital scrutiny as ‘intrusive and dehumanizing,’ as well as of questionable medical value. The words of an informant make this point:

I constantly ask [the endocrinologist]: ‘I know you have to monitor her clitoris, but at what point can this stop? At what point can my daughter say, ‘No, I’m not comfortable with you looking at my bottom like that?’ He hasn’t given an answer yet, but I want to raise my daughter to be able to tell someone ‘No’ when someone is doing something that doesn’t feel comfortable. I’m cringing to think about this happening when she’s 13. This is an awful enough period for a young girl, how does that affect her confidence that every time she goes to the doctor, someone’s pulling her pants down and looking at her?

It’s not only heterosexual intercourse, or even sexual partnerships, that are affected by the physical and psychological interventions meant to ‘normalize’ sexual function for women with CAH. Masturbation, too, is less frequent and less pleasurable for women with CAH. The report by May, Boyle, and Grant (1996) offers particularly poignant data on the relationship between masturbation and the medical interventions tat women with CAH have endured. Less than 1/2 of the women with CAH reported ever masturbating, compared with 3/4 of the women with diabetes, but women with CAH ‘commonly spoke about masturbation as a necessary medical procedure rather than primarily as a sexual activity.’ That is, they associated masturbation with dilation required to keep the vagina open, or as a way to understand the medical interventions they’d received.

Most women masturbate by stimulating the clitoris, which is the most efficient way to achieve orgasm, but these comments seem to suggest that women with CAH may more often think of masturbation as vaginal penetration, perhaps because this has been so explicitly prioritized in the eway they’ve been medically ‘prepared’ for sexual interaction.”

“In a study, a short film is shown of an infant baby in an infant seat who’s presented with 4 different stimuli — a teddy bear, a jack-in-the-box, a doll, and a buzzer — each of which is presented several times in succession. 200+ observers were shown the film and were randomly assigned the info that the child was a boy vs. a girl, without making an obvious point of the infant’s sex. There were remarkable differences in the way observers saw the same infant when they thought they were watching ‘David’ vs. when they thought they were watching ‘Dana,’ and these differences emerged most strongly when the child’s response was somewhat ambiguous. That is, in situations where the infant’s response was unambiguous, as in the clear pleasure s/he showed with the bear and the doll, or the clear distress s/he showed with the buzzer, the ratings didn’t differ much, except that the ‘boy’ was seen as displaying more pleasure and less fear across all situations than the ‘girl’ was. But 1 stimulus, the jack-in-the-box, elicited a much more complex response: ‘At first the infant stares at the box and shows a slightly startled reaction when it’s first opened. Upon successive presentations the infant becomes more and more agitated and after the third presentation the infant cries when the box is pushed forward (even before it’s opened) and screams when it jumps up.’ Given the sequence of responses, those observers who thought they were watching ‘David’ were more likely to describe the child as ‘angry,’ while those who thought they were watching ‘Dana’ described the child as ‘afraid.’ The Condrys pose an interesting question about how this differential attribution process might affect real-world responses to ambiguous emotional displays: ‘If you think a child is angry do you treat ‘him’ differently than if you think ‘she’ is afraid?’

Many subsequent studies have extended these observations, including by controlling for actual variations in infant characteristics. There’s even evidence that modern mothers, who typically know their child’s gender before birth, begin this labeling process to interpret the child’s ‘behavior’ or personality while still in the womb.”

“Girls with CAH are in some small respects ‘masculinized,’ and this is likely to be at least in part because they’re expected and believed to be masculinized. It may also be because potential masculinization in these girls is a source of fear and dread both for their parents, and for many of the girls and women themselves. Another factor might be a function of how ‘masculinization’ is conceptualized: if it’s a lower degree of certain behaviors and traits that are considered ‘feminine’ — such as fertility, heterosexual activity, and desire, or libido — then it’s important to recognize that genital intervention and long-term outcomes may be all the explanation that’s necessary. Finally, broader physiological differences — including higher weight-height ratios, shorter stature, hirsutism, disruptions in mood-regulating hormones, and more — may affect behavior through a variety of pathways, none of which are well-understood. And last, but not least, it’s possible that early androgen effects on the brain could explain some of these effects. But all of these possible effects must be seen in context. Importantly, the effect sizes for behavior differences between girls with CAH and other girls are quire small in most domains — so whatever influences are driving the slight difference in psychosexual behavior relative to unaffected women, the cumulative effects of all influences are relatively weak. It’s important to recognize that even if girls with CAH are to some degree masculinized (from whatever factors), most reports likely overestimate this masculinization because they are a) retrospective rather than concurrent; b) based on general assessments of behavior patterns rather than on specific observations; and c) collected under circumstances that make the condition of CAH salient, thus ‘priming’ research subjects to assign more masculinity to subjects with CAH.

Given the extensive factors that are demonstrably and significantly different in the development of girls with CAH compared with unaffected girls, the fact that researchers interested in brain organization focus single-mindedly on the potential effects of early hormone exposures on the brain is troubling. In fact, I think the persistent ignorance of broad physical and psychological health issues for women with CAH is downright stunning. Historians and philosophers of science are giving increased attention to the way that gaps in knowledge, as well as knowledge itself, are actively produced and maintained. The study of this phenomenon reveals that specific ideologies, cultural schema, and political interests systematically block certain forms of into and cause people to ‘forget’ or fail to incorporate certain facts into the overall thinking on a subject. Together with other patterns, the systematic disregard for the context in which girls and women with CAH develop constitutes more than an oversight. The attentive reader may have noticed the relative recency of the outcome studies that begin to systematically document pain, lack of sensation, and other difficulties with sexual function. These studies have been done in part because of brilliant organizing on the part of some intersex adults, who’ve begun to insist on having a voice in evaluating and reshaping the medical process that many believe constitutes serious harm. But it’s important to recognize that clinicians must surely have heard reports of pain, discomfort, and lack of sensation for decades, since the first children who underwent early surgery reached childhood in the 80s. This is a classic case study for agnotology: the ‘not-known’ regarding CAH as not merely info that’s not yet known, or not yet investigated. 40 years into the serious study of this disorder, these lacunae persist because they help to keep brain organization theory alive.”

“The interesting thing, from my perspective, is that women with CAH end up being so very close to typical. As a feminist, I don’t assume that ‘typical’ outcomes in terms of gender and sexuality are necessarily optimal. If heterosexuality is not the ‘goal’ of development, for example, then same-sex desires and relationships look like a perfectly good, and possibly better, outcome for many women with CAH. Because even if penile-vaginal intercourse can be ‘decentered’ in a heterosexual relationship, both partners are likely to be frustrated if they cannot, at least sometimes, participate in this activity. It’s conceptually important to the heterosexual script and is also likely to be an activity that male partners have found satisfying in other sexual relationships. So contrary to the usual presumption, from the perspective of maximizing opportunity for mutually satisfying sexual partnerships, it might be better if women with CAH were less, rather than more, typical.”

Conclusion

“People’s willingness and ability to ‘pass’ as heterosexual because of gender conformity affects sexual identification, visibility, and disclosure of oneself as gay or lesbian, and these also affect participation in research on sexual orientation. Thus, both men on the DL and femme lesbians are probably underrepresented, which are but 2 examples of why the association of gender conformity and heterosexuality, or nonconformity and homosexuality, should not be taken at face value.

Context matters. In our ongoing development, context matters from the earliest stages of gestation, through infancy, childhood, adolescence, and on into adulthood, especially when we’re talking about domains of personality and behavior that are as complex and irreducibly social as gender and sexuality. It also matters in science, affecting the way we recruit our research subjects and interpret the info they offer us, the way subjects perceive the goals of the research and unconsciously emphasize or filter info about themselves and important others (especially children) on whom we ask them to report. But there has been next to no attention to context in brain organization research. And context is demonstably different for the very groups that scientists compare when they do brain organization studies, so the effect of contextual variables is hidden within the differences they may observe.

Can we conclude, then, that there are no meaningful differences in the initial predispositions of male and female infants, at the group level? No, because we can’t remove children from the socialization process in order to test this — this would require intervening to an unacceptable degree with children’s initial contact with parents and others, which is critical for healthy emotional attachment. But we also can’t conclude that there are such differences, and the evidence from brain organization research adds very little reason to suspect that differences in initial predispositions make a meaningful contribution to gender.

Why do I place persistent emphasis on similarity? Is it because I think ‘difference is bad’ and everyone should be the same (as critics accuse feminists of thinking)? No. I think difference and variety are good, and my problem with the current story of ‘sex in the brain’ is chiefly that it underplays and mislocates the creative sources of human difference. Gender schemas are powerful, permanent shapers of both our materially based, lived selves, and the way we perceive ourselves and others. In this way, gender is like a polarized lens, channeling the ‘unruly’ vibrations of sunlight into a single direction. It’s not that light only ‘looks’ less glaring after passing through the lens: it is, in fact, transformed, and more homogeneous. If you view sunlight only after it’s been polarized, you’ll have a much narrower view of how light can move. Of course, people don’t become ‘gender automations’ — there is variety and flexibility, and I’m of the opinion that these are good things.”

“Metaphorically speaking, whatever is ‘written in our genes’ must be a very open-ended story, because gene expression is a dynamic, contingent process that’s responsive both to specific conditions during development and to random events. In Balaban’s words, ‘Development processes inseparably fuse experience-dependent and experience-independent components, have important stochastic contributions, and exhibit a greater degree of mechanistic continuity between developing and adult nervous systems than previously thought.’

This is unfamiliar territory to most biologists, so what exactly does that mean? The 3 key concepts are the inseparability of experience and heredity, the importance of random events, and the fact that development is a lifelong process. Outcomes in the cognitive domain, in particular, are always contingent, rather than ultimate.”

“They began by gathering samples of naturally occurring variants of a number of plant species, an ecotype, from climatically distinct regions in CA. Individual plants from each of these ecotypes were then cloned by propagating cuttings from the wild-gathered parent plant. The researchers tested the effect of the growing environment by following this procedure: ‘Clone-members of different ecotypes of perennial species were brought into gardens at contrasting altitudes in CA, one in the mild coastal climate near sea level at Stanford, one halfway up the western Sierra Nevada slope at 4,600 ft, and a third at timberline near the crest of the Sierras at 10,000 ft. Notably, ‘the environment of the plants of different genetic/geographic origin within each experimental field was as small as possible.’

Clausen’s team followed a painstaking process of charting the growth of successive generations of each of these transplants, measuring and plotting the characteristics of tens of thousands of individual plants over the years. Obviously this sort of experiment couldn’t be done for people, nor for most other animals. There’s a second reason that plants provide a good example: it removes the issue of socialization and learning. Plants are simpler. Yet these experiments dramatically disrupt any simple sense of genetics as a ‘blueprint’ for development, because the appearance of genetically identical clones was strikingly different in different climates. On the one hand, among ‘clone-members’ from one region transplanted to another, ‘details of structure and general habit of growth are essentially unaltered by transplanting.’ In other words, basic features common to the species remained relatively constant. On the other hand, ‘striking modifications in size of vegetative parts, especially of leaves and stems, and in the extent of branching flowering stems’ were observed among genetically identical plants grown at different altitutdes.

This point is so important, and so antithetical to the popular notion of genetic ‘blueprints.’

Note two interesting thing about these plants. There’s no way to arrange the plants by height that’s ‘environment-neutral’: the plant that’s tallest at high elevation and low elevation is one of the shortest plants at medium elevation. And there’s no single environment that’s uniformly favorable for producing taller plants. Plant 4, for example, grows tallest at medium elevation, while plant 3 grows highest at low elevation, and plant 6 grows highest at high elevation.

The environmental dependence of the relationship between the genotype and the phenotype is not anything special in Achillea, by the way, nor is it restricted to plants. Instead, it’s a general feature of development that is well recognized by developmental and evolutionary biologists.”

“There are a number of examples of differences in the way that specific environmental conditions affect sex-typed behavior in m ales vs. females in other species, such as that there’s no environment-neutral way to specify the sex difference. For example, territorial exploration is a trait that’s considered male-typical in rhesus monkeys, but in this species males seem to be more vulnerable to ‘aberrant’ rearing conditions such as isolation. Thus, ‘isolation-reared males [show] much more frequent and intense fear, and less exploration, than isolation-reared females.’ Likewise, both male and female rats have a capacity for ‘maternal’ behavior (which raises the question of why grooming and caring for rat pups isn’t just called ‘paternal’ behavior). Leboucher has shown that previous exposure to pups has a more dramatic effect on stimulating ‘maternal’ behavior among male vs. female rats. Another interesting example concerns the way that males vs. females navigate through space. The use of ‘spatial location” (such as orientation in relation to compass points) vs. ‘local markers’ is considered to be a reliable M-F difference in a number of species, with males preferring (or excelling at) the former and females the latter. But recent data complicate this perspective. Herman and Wallen studied spatial navigation among rhesus monkeys, manipulating both hormones and the info that was available to the animals for using the 2 spatial strategies. The interesting finding was among the animals that weren’t hormonally manipulated. Here, in a novel environment in which local markers weren’t available, the researchers found that the females, not the females, were able to make better use of the ‘male typical’ navigational strategy. They also found that blocking the males’ prenatal androgen exposures improved the use of local markers (not that blocking androgens prevented development of the ‘male-typical trait, which in their model would’ve meant impairing the ability to use ‘spatial consistency.’)

These examples suggest that although these traits — maternalism, exploration, spatial relations — are staples of the brain organization lit, it isn’t accurate to simply talk about these traits as representing ‘stable differences’ between the sexes, even though the differences may in fact be quite real and even substantial in any given time and place. Given what researchers themselves have shown about about how sex differences in these traits can and do change in different environments, it’s teleological to pronounce such environment-dependent status as ‘sex-typed.’ That is, because the nature of the sex difference depends on specific environments, the environments that produce more pronounced sex differences are defined in a circular manner as ‘ideal’ or ‘natural.’ But — especially when the variation in outcomes produced by environmental change isn’t pathological — one must ask on what basis is that environment considered ideal? Ideal for what? For emphasizing a particular pattern of difference.”

“The current basic theory behind sex difference in height is this:

The growth plates are generally more sensitive to the effects of estrogen than to those of testosterone. During puberty in the female, the rising levels of estrogen seal the epiphysical plate earlier than testosterone does males. The effects of testosterone are felt at a later stage. Thus, females stop growing earlier than males do.

The scientific actors and relationships — the narrative — should be very familiar by now. An observable sex difference in phenotype (height) and a physiological process that involves the so-called sex hormones. The story implies that estrogen would be responsible for hardening of growth plates only in females, while the actor in males would be testosterone (in fact, both hormones are important for stopping bone growth in both sexes). The story doesn’t help account for a number of interesting problems we could observe about men, women, and height across populations and time. For example, if ‘sex hormones’ are the primary explanatory factors, how do we explain differences in groups that aren’t defined by sex? For instance, why are Dutch people so tall, even when compared with other W. Europeans with similar health and nutrition metrics? How can we account for the fact that the average height among Dutch women is now somewhat higher than the average height among Spanish men was just about 60 years ago? And why has height in Dutch people (and the rest of the industrialized world) been significantly rising during the precise period when age at menarche (the time when estrogen levels dramatically rise in girls) has been significantly dropping?

The problematic habit of classifying hormones as fundamentally related to, and even for, sex (rather than about broader functions, such as growth regulation) and as ‘sex-specific’ (that is, properly belonging to 1 sex or the other, rather than being related to normal function in healthy individuals of both sexes) is part of the difficulty here. But it’s only 1 part. The other part is that when height is conceptualized as a sex (read: biological) difference, height is then theorized as fundamentally flowing from sex in a deterministic way. This model has the unhappy effect of suggesting that the general shape of sex differences is stagnant over time and across populations. Notice, for example, that Lippa (2009) pronounces that sex differences in height are fully biological, even though he’s gathered info about only 1 single time period! But that’s a conclusion one cannot draw without data over time, as well as across locations. In fact, we have enough data to show that the NOR map between sex and height is not flat, because there have been highly significant changes in height for both men and women in the past 100 years. Consider the fact that the average male height has been increasing faster than the average female height in most European countries over the last 70 years. This shows us that the NOR is not only flat but nonparallel: different environments affect height differently for women vs. men. Once we know all this, all bets are off for predicting how sex differences might look in a future, novel environment.

The ‘fundamental sex difference’ story makes it difficult to notice how sex differences, themselves, change shape in different environments. It also makes it hard to absorb info on important influences from social structures.”

“Although early development has a major impact on later characteristics, bone width and density continue to change and develop over time, following the same basic processes that guide early development. Fausto-Sterling recently used bone density and osteoporosis to push against the sex vs. gender distinction itself, asking ‘what it might mean to claim that our bodies physically imbibe culture.’ Because bone development is ‘an area often accepted as an irrefutable site of sex difference,’ showing how sex differences in bone development are responsive to culture not only improves our understanding of bone health and diseases, it destabilizes biological sex. This work is part of a larger feminist project that involves documenting the ways in which apparently ‘natural’ properties of sexed bodies are in fact dependent on gendered activities that actively shape bodies, as well as our perception of bodies. In the case of bone density and osteoporosis, some of the factors that underlie putatively innate biological sex differences include physical activity levels, especially weight-bearing work and exercise; diet, especially calcium intake, exposure to sunlight for Vitamin D; and hormones — of those, only hormones have a plausible ‘biological’ basis that’s easy to conceptualize apart from gender, socioeconomic position, and culture. And while biological factors like hormones are involved in translating specific inputs like weight-bearing exercise into a sex difference in bone density, we can immediately see that this will be affected by systematic gender differences in weight-bearing exercise. It’s harder to remember, though, that steroid hormones are also affected by the social organization of gender. Steroid hormones, including gonadal hormones, are responsive to a wide range of behaviors, mood states, and pharmacological interventions (such as hormonal birth control) that are profoundly affected by gender relations. Gender, in interaction with socioeconomic position and culture, heavily influences the kind of work and play that people engage in, the food they eat (including both type and quantity), the likelihood that they will diet, and so on.”

“It’s not just that there is some constant amount of bone density that’s attributable to sex and some other variable amount that you add on for gender. Nor is it accurate to think of sex as controlling some specific aspects of the processes related to bone density, while gender controls other processes. If we think of this interaction in Balaban’s terms as the ‘inseparable fusing’ of ‘experience-dependent’ and ‘experience-independent’ aspects of development, it suggests an alternative working definition for sex. If sex is the experience-independent part of maleness or femaleness, and gender is the experience-dependent part, what would that mean? For one thing, it would mean that the question of what counts as sex would become a matter for empirical investigation instead of being a matter of definition or philosophy. For another thing, it would mean the immediate constriction of the term sex to exclude all those aspects of biology that we already know are dependent on experience (where experience includes exposures that depend on personal biography as well as culture and social position in structures like gender, class, and race). Third, even once the term is restricted in this way, the meaning of this term for developmental endpoints must also change from current usage. That is because the way in which ‘sex’ works still depends on gender, as well as other aspects of culture. This is most especially true in the realm of behavior, particularly that portion of behavior that we understand as gender and sexuality.”

“As early as 1969 it was known that many of the ‘organizing’ effects of hormones aren’t permanent, but are easily modifiable by experience. In a little-cited study by UCLA researchers, for example, scientists found that allowing an androgenized female rat to have just 2 hours to adapt to a stud male completely eliminated the behavioral effects of prenatal testosterone injections. Money and Ehrhardt (1972) suggested that ‘neonatal androgen may have rendered the females more sensitive to the copulatory environment, possibly to olfactory cues, in the manner that’s usually typical of males. Once adapted to the environment, they became disinhibited. The behavior that was then released wasn’t masculine in type, but the feminine response of lordosis.’ Subsequent experiments have shown that a great many of the sex-typed behaviors that are supposedly permanently organized by prenatal hormones can be dramatically modified or even reversed by simple and relatively short-term behavioral interventions such as neonatal handling, early exposure to pups (in rats), and sexual experience, to cite just a few examples.

A recent very exciting example of plasticity in humans concerns dyslexia, a cognitive trait that has been theoretically linked to early testosterone and has even been examined in some brain organization studies as a marker of ‘masculinization.’ Simos and colleagues (2002) studied children with dyslexia before and after 80 hours of intensive remedial reading instruction. At the beginning of the study, magnetic source imaging showed that the children with dyslexia had a different pattern of brain activation compared with normal children with no reading problems. In particular, they showed very low signals associated with an area that’s normally involved in phonological processing. Remarkably, after the intensive intervention the children not only made substantial improvement in their reading skills but also showed much larger signals associated with the phonological processing area that formerly showed low signals.

Another powerful example concerns spatial cognition — one of the hallmarks of psychosexual differences. Feng, Spence, and Pratt (2007) identified a basic info-processing capacity that underlies spatial cognition and showed that differences in this capacity (the distribution of spatial attention) are related to differences in the higher-level process of mental rotation ability. They then showed that a remarkably brief intervention — just 10 hours of practice with an action video game — caused ‘substantial gains in both spatial attention and mental rotation, with women benefiting more than men. The 10-hour training didn’t completely eliminate the sex difference, but it came extraordinarily close — the mean scores after training were no longer statistically distinguishable between males and females.

Thus, even though early hormones affect neural development , the language of ‘hardwiring,’ ‘blueprints,’ ‘latency,’ ‘permanent organization,’ and so on clearly conveys an inaccurate picture of the nature of early hormone effects on behavior. As Doell and Longino noted 2 decades ago, these metaphors fail to accurately capture how development really works. Even in rats, early hormone exposures don’t create a solid foundation on which behavior must forever stand. At first glance the true process might seem to be captured by the notion of developmental ‘cascades,’ which several organization theory researchers raised in their interviews with me. The notion of developmental cascades suggests that hormones don’t directly determine behavior, but create a small push in one direction, which is then amplified by experiences and other inputs that in turn trigger additional inputs, such that a tiny push at the front end can end up in a sizable difference in outcome. But this is only half the story — one in which the small initial differences almost inevitably grow larger as additional effects accumulate. But an early push in a certain direction can be either enhanced or entirely eliminated by subsequent experience, such that development from that point forward would proceed as though the early hormone exposure had never happened.

Again, the point is not that hormone effects aren’t ‘real.’ Hormones are important growth mediators, and they do figure into development, including neural development, in a variety of important ways. Nor is the point that males and females aren’t ‘really’ different. There are demonstrable differences, on average, in a variety of characteristics, including some limited cognitive abilities, personality traits, and interests, including sexual interests. The problem is the way that brain organization theory brings together ideas about hormones with observations of M-F differences. The story attributes an unrealistic specificity and permanence to early hormone effects, as well as a demonstably false inevitability and uniformity to sex differences, which are inaccurate even for those animals whose sexual and other behaviors may turn out to be mechanistically less complicated than ours.”

“Let’s consider how animal experiments on brain organization could better incorporate the NOR principle, while remaining attentive to plasticity. Such experiments would require testing hormones-during-critical-periods as the first important variation in environment, and the phenotype that results from early hormones as an interim state in the organism’s development. Next, it would require systematically testing how early hormone exposures affect the behavior of animals who are reared in different ways, sexually socialized in different ways, exposed to different subsequent hormone environments, and have different diet and exercise regimens, and so on. A fundamental error in most research to date is that researchers have assumed that the ‘organizing effects’ of hormones are permanent, without testing similarly treated animals across a range of developmental conditions and genomes to test the stability of those effects. This is to proceed as though all important aspects of development are over, or as though development of the brain and behavior has a fundamentally different character after the ‘organization’ phase. But it doesn’t.

Thus, the notion of a blueprint or an unchangeable behavioral pattern, set early on and then lying ‘latent’ and waiting to be ‘released’ doesn’t really work. Add to the model the understanding of gender as a multilevel environment that impinges on development in ways that are both pervasive and only dimly recognized at present. What is the bottom line? It’s reasonable to expect that in the context of extremely gender-dichotomous socialization around sex, within-sex variability will be enhanced. This suggestion fits the empirical evidence regarding domains of development that were reviewed above (bones, spatial relations, educational interests and attainment).”

“The influence of biological variables that we think of as ‘sex-linked’ almost certainly plays a role in the overall iterative-and-looping process of development. But in terms of a scientific program for understanding development, we’ve reached the end of that road — in fact, we’ve gone way off the road into the woods and are now stuck in the deep mud of ‘innate sex differences.’ The data aren’t compelling when placed together in the ‘network of associations,’ and models of development that are current in biology give us little reason to continue pursuing this line of reasoning. So here are a few closing observations that might help us climb out of the mud:

• Steroid hormones are important, but they aren’t best conceptualized as ‘sex hormones.’ They do lots of things; ‘sex hormones’ was the original conceptualization that drove the research and classification on hormones, but it doesn’t fit the data on what hormones do any better than other possible schemes. And the ‘sex hormone’ framework demonstably blocks recognition of complex and accurate info.
• Personality traits and dispositions aren’t identical in individuals, but they also aren’t well captured by the binary system of gender — even in spite of pervasive cognitive schemas that exert pressures toward this pattern. We aren’t blank slates, but we also aren’t pink and blue notepads.
• Brains develop only in interaction; input from the external world, as well as from one’s own sensory apparatus, is as critical to development of the brain as food and water are to the entire organism.
• Brains change and develop over the lifetime. Few inputs are irreversible. Even the animal experiments on brain organization showed that the ‘permanent’ effects of steroid hormone exposures could be eliminated or even reversed by fairly brief interventions in the physical and/or social environment.
• Gender relations change, and these are demonstrably related to changes in psychosexual outcomes. For example, structural-level shifts in education have quickly reshaped the landscape in terms of the proportion of college grads who are female, as well as the sex composition of particular programs of study.”

“When research starts to look too much like an ‘infomercial’ for cherished beliefs, it’s no longer science. Brain organization theory is little more than an elaboration of long-standing folk tales about antagonistic male and female essences and how they connect to antagonistic male and female natures. The data don’t fit into the tidy M-F brain patterns, anyway. Why keep trying to fit the data into a story about sex?

A wise person once told me: ‘Life is full of choices that aren’t always clear. Err on the side of good narrative.’ The theory of sexual organization of the brain is getting in the way of the science of human development. And it’s also getting in the way of good narrative. What we need now is a way to cultivate and reinvigorate curiosity about how the body really matters in the development of human personality and behavior, because curiosity and skepticism are the real engines of scientific discovery. What good is science that doesn’t tell us anything new?”