The prenatal-testosterone studies are, however, just one source of evidence for the fetal fork hypothesis. The period shortly after the baby is born supposedly provides another:
One of the first things your daughter’s female brain will compel her to do is study faces. Whereas child developmental specialists originally thought all infants came wired for mutual gazing, your daughter may be more interested in staring at a human face than the newborn male.25
This quote from the Gurian Institute’s book It’s a Baby Girl! is a typical popular take on a study conducted several years ago by Simon Baron-Cohen, together with graduate student Jennifer Connellan and other colleagues. They looked for gender differences in newborns who were on average just a day-and-a-half old. The logic was simple: any differences between the sexes seen at this tender age can’t be chalked up to socialisation. One hundred and two babies were offered, one at a time, Connellan’s own face and a mobile to look at. The idea was to measure the babies’ interest in the face versus interest in the mobile: empathising versus systemising. Each baby’s eye gaze was filmed, and this recording was later used to time how long each baby spent looking at the face and the mobile. Male and female babies spent equal amounts of time looking at the face: both sexes, on average, spent just under half the total looking time (which was about a minute) looking at Connellan’s face. However, males looked longer at the mobile than did females (51 percent of looking time versus 41 percent for females) and females, as a group, looked longer at the face than the mobile (49 percent versus 41 percent of looking time).26
Much has been made of the significance of this study. ‘The results of this experiment suggest that girls are born prewired to be interested in faces while boys are prewired to be more interested in moving objects’, writes Leonard Sax in his book Why Gender Matters,27 a conclusion echoed in the popular media around the world. The implications for career choices are clear. Cambridge academic Peter Lawrence, citing the newborn study, argues that men and women are ‘constitutionally different’ and thus unlikely to ever become professors of physics and literature in equal numbers.28 And in his contribution to the book Why Aren’t More Women in Science? Baron-Cohen suggests from the newborn study that the ‘“bias” in attention to things rather than emotions (in boys) and vice versa (in girls)’ reflects ‘partly innate differences’ that culture then amplifies. Sex differences in the empathising versus systemising bias, Baron-Cohen argues, ‘suggests that we should not expect the sex ratio in occupations such as math or physics to ever be 50-50 if we leave the workplace to simply reflect the numbers of applicants of each sex who are drawn to such fields.’29 In other words, short of some very heavy-handed social engineering, gender equality in the workplace is an impossible ideal.
But unfortunately, as some researchers have pointed out, the study was simply not done well.30 When you are claiming nothing less than evidence of the biological origins of a gender-stratified society, it helps to have a methodology that stands up to scrutiny. No study is perfect, of course, but this one was flawed in ways it simply need not have been, as psychologists Alison Nash and Giordana Grossi have pointed out. Some of these problems concern the sort of detail that may provoke a small yawn in the non-specialist, but a severe case of eyebrow-in-the-hairline for experts. First of all, there are standard procedures when it comes to testing newborns for their visual preferences. A baby’s attention span is not at its peak in the first few days of life, waxing and waning over short periods of time. For this reason, when infant researchers want to find out which of two stimuli a newborn finds most interesting, they usually present them simultaneously. If you don’t, and instead present them one after another, then you don’t really know whether the baby looked at stimulus A more because she genuinely found it more interesting, or whether she was irritated by some internal rumblings, about to fall asleep, or simply a little tired of life when stimulus B was on show.
In Connellan’s study, the face and the mobile were presented separately.
Another important thing to know about very little babies is that they can’t see very well. They actually aren’t even drawn to faces per se but to visual stimuli that, like the face, have a top-heavy pattern. In fact, before the age of three months, babies actually prefer top-heavy, facelike patterns over real faces. It’s important, therefore, to ensure that babies all view the stimuli from the same angle, otherwise the same stimulus can appear to be different, including its degree of top-heaviness.
In Connellan’s study, some babies were tested on their backs in a cot, and other babies were tested in a parent’s lap. (If more girls than boys, say, were on their backs, then on average your groups of boys and girls are not seeing the same stimuli.)
But the most major problem with the study, described by Nash and Grossi as a ‘striking design flaw’, was its potential for experimenter expectancy effects.31 If you have ever visited a new mother in a maternity ward, there is a good chance that you will have seen one or more of the following items: a baby wearing a pink or blue (or otherwise gendered) outfit; a pink or blue balloon; a pink or blue blanket; an arrangement of predominantly pink or blue flowers; pink or blue congratulation cards; or even (as was the case in the hospital in which I gave birth) a pink or blue name card on the baby’s bassinet. Clues, in short, as to the baby’s sex. Now if you are an experimenter and stimulus rolled into one neat package with a particular hypothesis in mind (not to mention a head full of cultural assumptions), you have to make sure that this information doesn’t unconsciously affect your behaviour towards the baby. This, of course, is impossible. As we saw in the first part of this book, even information that doesn’t register with consciousness can subtly change behaviour. Researchers therefore usually take this problem very seriously and go to some effort to eliminate experimenter expectancy effects. Here, for example, are the precautions taken by another recent study that also looked for gender differences in newborn eye gaze:
We instructed all participants that the infant must be dressed in a gender-neutral outfit and that the interacters in the study room must remain unaware of the baby’s sex throughout the interaction, as well as after the interaction was complete. Because parents often had either pink or blue outfits for their newborn, many opted to dress their baby in the white outfits provided by the hospital …
We decided that the study should take place in a room other than the mother’s room in order to decrease the likelihood that something in the room would provide clues to the interacters as to the sex of the infant.…
To keep the interacters blind to the sex of the infant all identifying information on the infant’s bassinet was covered or removed upon arrival to the study room.
Researchers do not go to such lengths merely to make life awkward for themselves and the parents of newly born babies. (In this carefully designed study, no gender differences in eye gaze were found in newborns although, interestingly, they did find gender differences in eye gaze in a follow-up three to four months later. This, they point out, suggests the possibility ‘that the gender-typed behaviour pattern is not innate but, instead, learned in early infancy.’)32
No such precautions were taken in Connellan’s study.
She knew the sex of at least some of the newborns she tested, and it’s not beyond the realm of possibility that, on other occasions, clues as to the baby’s sex unconsciously undetected could have swayed her behaviour in a direction consistent with gender stereotypes.33 Unfortunately, this was a study in which even slight differences in the experimenter’s behaviour could well create experimenter expectancy effects. Motion, open eyes and mutual eye gaze are all visual stimuli that newborns especially like and are sensitive to.34 It is, I imagine, rather hard to hold up a mobile, and look at a newborn, in exactly the same way 102 times. What if Connellan inadvertently moved the mobile more when she held it up for boys, or looked more directly, or with wider eyes, for the girls?
But even if a redoing of the study, performed with a less cavalier approach to normal poli
cy and procedure in infant testing, got the same result, what would it actually signify? Nash and Grossi have argued that if the sex differences in the newborn study reflect differences in brain organisation then we should see increasing divergence between girls and boys as these skills develop. Yet boys’ greater interest in the mobile doesn’t seem to serve them much advantage. As Nash and Grossi have pointed out, as has Harvard University developmental psychologist Elizabeth Spelke, there is little evidence for a systemising advantage in young boys: a large body of research exploring infants’ understanding of objects and mechanical motion finds no advantage for males.35 As for the development of empathy, evidence of divergence is modest. Boys and girls develop an understanding of the mental states of others at a similar rate. But girls do have a small advantage, on average, in facial expression processing and, overall, studies find signs of greater affective empathy in girls. However, as is the case in adults, this difference is much smaller when based on observations rather than self-report or report by another (such as a parent).36 But also, these psychologists have pointed out, why think that what a newborn prefers to look at provides any kind of window, however grimy, into their future abilities and interests? It might come down to something as boring as girls responding more or less to some other difference between the two kinds of stimuli – visual, auditory or olfactory – that has nothing to do with faces versus objects per se. We have no idea whether newborn preferences reflect what their later abilities will be – such an assumption is, as Neil Levy puts it, ‘essentially unargued for’ and ‘questionable at best’.37
Many studies have methodological flaws. Many studies are overinterpreted. But not many studies inspire in their authors and others the conclusion that innate differences in part lie behind our gender-stratified society.38 This is a study that really needs to be repeated before it is taken too seriously, and with closer attention to what the results might actually mean, as well as those little details that make all the difference between the study the expert feels she can trust and the study that leaves her eyebrow muscles aching and exhausted.
So what does go on in the darkness of the womb? Consider the boldness of the statements made in the popular media about the effect of foetal testosterone on the brain. Now consider the inadequacy of the data showing links between exposure of the foetal brain to testosterone (which, you will recall, these studies might not even be tapping) and brain ‘type’. Contrast, for a moment, the confidence of claims that boys and girls arrive with differently prewired interests, against the flimsiness of the evidence. There’s something a little shocking about the discrepancy between the weakness of the scientific data on the one hand and the strength of the popular claims on the other. As Simon Baron-Cohen himself has written, ‘the field of sex differences in mind needs to proceed in a fashion that is sensitive … by cautiously looking at the evidence and being careful not to overstate what can be concluded.’39
At last, something on which we can all agree.
It’s a good life. If I die tomorrow, I’ll die a happy woman, because I’ll feel like I’ve done a lot of good work.
—Kerin Fielding, orthopaedic surgeon1
Today, women are strongly represented in fields such as biology, psychology, medicine, and forensic and veterinary science. Some think this reflects ‘the feminine propensity to protect and nurture – and the desire to work with living things’, as Christina Hoff Sommers suggested by way of explaining the recent influx of women into the once male-dominated domain of veterinary medicine.2
Maybe. But there is something a little unsatisfying about this reframing of the life sciences as: Now with added empathising for extra feminine appeal! Is the supposed female drive to work with living things, or to engage with mental states, really likely to be satisfied by looking at cells under microscopes or de-sexing cats? Even academic psychology, most of which is at least about people, is devoted to the pursuit of understanding the laws and principles – one might even say systems – that underlie cognition and behaviour. Apart from the lab teamwork common to science in general, the core work of an academic psychologist – making sense of the literature, designing experiments and analysing and interpreting data – puts few demands on empathising abilities. And what about forensic science, which draws in more than three times as many women as men?3 On the one hand, it does indeed sometimes have people as its subject of study. But, on the other hand, when it does, often they are dead.
As journalist Amanda Schaffer has pointed out:
[I]f history is any guide, today’s gender breakdowns are likely to keep changing. What’s so magical, after all, about the current numbers? A few decades ago, most biology and math majors were men. So were most doctors. Now maths undergraduate majors split close to 50/50. In 1976, only 8 percent of Ph.D.s in biology went to women; by 2004, 44 percent did. Today, half of M.D.s go to women. Even in engineering, physics, chemistry, and math, the number of women receiving doctorates tripled or quadrupled between 1976 and 2001. Why assume that we have just now reached some natural limit?4
It’s a good point. Perhaps in a few decades we will be redefining women’s new levels of participation in the physical sciences, politics and business as reflecting their innate drive to nurture. After all, is there any more powerful way to help others than to develop sustainable technologies, set tough emissions targets or, like Bill Gates, write big fat cheques to charitable causes?
As some psychologists have pointed out, such historical shifts – including the movement away from male dominance in teaching and secretarial work – don’t lend themselves especially well to explanations in terms of hormones and genes.5 So with this malleability of sex segregation in mind, let’s turn to the next two ways of investigating the link between foetal testosterone and later sex-typed behaviour: females whose in utero living conditions were, hormonally speaking, wrong for their chromosomal sex; and monkeys.
In a condition called congenital adrenal hyperplasia (CAH), the child’s genetic state results in the foetus’s being exposed to unusually high levels of testosterone. In girls with CAH, this triggers development of male external genitalia. (The female internal reproductive organs, however, develop normally.) Girls with CAH are born with genital virilisation – that is, they look more-or-less like a boy at birth, depending on the severity of the condition. Usually the condition is detected at birth. The child is then given ongoing hormonal treatment, some time later undergoes surgery to feminise her genitalia, and is raised as a girl. This offers an opportunity for researchers to explore the effects of high foetal testosterone, disentangled from what normally comes with that experience, namely, also being reared as a boy. However, it’s important to point out that girls with CAH are not simply girls plus extra foetal testosterone. Not only are other hormone levels also awry (and are therefore potential candidates for being behind any differences in behaviour), but also these girls are born with ambiguous genitalia, and receive continuous hormonal treatment as well as, most likely, extensive surgery on the genitalia. (When this happens seems to be quite variable.) It’s not impossible to imagine that this could create a certain ambivalence around the child’s gender in the mind of a parent, and perhaps in the child herself, for which there is a little evidence.6
But, nonetheless, are girls with CAH more likely to be systemisers than empathisers? So far, we can’t say. Older girls and adults with CAH do report less tender-mindedness, interest in infants, and social skills than their non-CAH relatives. But on the other hand, they report equal communication ability (assessed with questions like I am good at social chit-chat, and I find it easy to ‘read between the lines’ when someone is talking to me) and no greater dominance (which includes masculine qualities like being aggressive, authoritative and competitive).7 So the evidence is a little mixed and, as we learned in Chapter 2, self-report scales may tell us little about people’s actual empathic tendencies and skills. As for systemising, in the absence of an actual test of this ability it’s impossible to know. One study found that
girls with CAH report less attention to detail than control girls (a skill that Baron-Cohen considers especially important for systemising).8 And there’s no evidence that the high prenatal-testosterone levels of CAH serve to improve mathematical performance – it’s even been suggested that it impairs it.9 Researchers have also tested girls with CAH on the ubiquitous mental rotation tasks, and the evidence currently points towards an advantage for them over unaffected girls.10 But, as has been pointed out, this could be the result of their more boyish play experiences, rather than prenatal testosterone per se.
And girls with CAH definitely do differ from their non-CAH sisters and relatives in their play. In as much as we can take at face value their caregivers’ reports and behaviour when under observation in the lab, this seems to be despite the best efforts of their parents.11 Girls with CAH play much more at boyish activities and toys than do control girls (although not quite as much as boys do), and they are also less interested in girlish toys and pastimes.12 This boyishness seems to continue into adolescence. For example, adolescent girls with CAH are intermediate between boys and girls in their interest in sex-typical activities (football versus needlepoint, embroidery or macramé) and future occupations (like engineer versus professional ice skater).13
These tomboyish interests seem to provide a compelling case for the idea that foetal testosterone organises the brain to be drawn to certain kinds of stimuli that lie behind sex differences in play behaviour and, by implication, occupational segregation.14 But what is a little odd is that no attempt seems to have been made to work out whether girls with CAH are drawn to some particular quality in boyish toys and activities or whether they are drawn to them simply by virtue of the fact that they are associated with males.15 Take, for instance, the Pre-School Activities Inventory, on which girls with CAH score more like males than unaffected girls. The inventory includes questions about playing with cars and dolls, and so on.16 But girls with CAH can also get a higher score than unaffected girls by, for example, showing little interest in jewellery, pretty things, dressing up in girlish clothes and pretending to be a female character.17 Another study (drawing on a different clinical group) found that greater prenatal androgen exposure led to less interest in activities like ballet, dressing up as a fairy, dressing up as a witch, dressing up as a woman, gymnastics, playing hairdresser and working with clay, but more interest in basketball, dressing up as an alien, dressing up as a cowboy, dressing up as a man, dressing up as a pirate and playing spaceman.18 Likewise, women with CAH asked to recall their childhood activities score significantly differently from controls on a questionnaire that, among other questions, asks about use of cosmetics and jewellery, hating feminine clothes, the gender of admired or imitated characters on TV or in movies and whether they dressed up more as male or female characters.19
Delusions of Gender Page 14