It’s plausible but, as some researchers have pointed out, there are dangers in extrapolating from rats and birds to humans. Working from an implicit we’re-all-God’s-creatures framework that we do not apply when it comes to the right to not be killed and eaten, enjoy access to education or drive a car, there’s a tendency (especially among some popular writers) to assume that what goes for the rat can be readily applied to humans.11 Often, of course, this is the case. But while there are important similarities between all mammals great and small, there are also critical differences. As Melissa Hines points out (although she puts it rather less crudely), a penis is a penis, whether tucked between the legs of a rat or a man. Suitably scaled for size, it serves much the same function in both species, and the mechanism by which it’s produced may be much the same in the two species. But a rodent brain, even expanded to suitably grand proportions, would serve a human extremely poorly indeed. Whereas in the human brain the so-called association cortices, devoted to complex and clever higher-order thinking, have taken over much of the available space, in the rat brain the association cortex has to squeeze in where it can among the neurons devoted to smell, sight, sound, touch and movement. It’s for this reason that Hines cautions that ‘one cannot assume that early hormonal influences on neural development in other mammals, particularly those involving the cerebral cortex, are preserved in humans.’12 Likewise, the very point of the slur ‘birdbrain’ is to indicate that the thinking skills of the person in receipt of the insult are, in some important way worth commenting on, inadequate.
There are several other important dissimilarities, too, between how early hormones affect rats and humans.13 All in all, some researchers think that rat data may not be very helpful in illuminating what goes on in humans.14 That’s not to say that the same principle doesn’t apply – that foetal testosterone has some important effect on the brain. But it’s wise not to extrapolate too enthusiastically from rats. So what about primates? Unlike rats, female rhesus monkey infants treated prenatally with testosterone are no more aggressive than untreated females. In fact, even normal female infants are no less aggressive than males when they are reared in a normal social group.15 However, female infants experimentally treated prenatally with testosterone are keener than untreated females on rough-and-tumble play.16 And when prenatal testosterone is blocked in males, early in gestation, these males are a bit less interested in rough-and-tumble play.17
Researchers hypothesise that the changes they see in behaviour as a result of their hormonal manipulations are brought about by testosterone-induced changes in the foetal brain (or, in the case of the rat, the neonatal brain). But I say hypothesise because it has proved harder than you might think, even in the relatively humble rat, to connect the dots between prenatal hormones, brain changes and behavioural change. For example, more than twenty-five years ago it was discovered that a certain region of the rat brain (part of the preoptic nucleus) is much larger in male rats than in female rats. Treating female rats with androgens early in life makes this region bigger, and depriving male rats of androgens prevents the normal male supersize appearance of the preoptic nucleus.18 So far – hormone to brain – so good. But getting from brain to behaviour has proved a challenge. In 1995, the pioneer in this research, Roger Gorski, lamented, ‘We’ve been studying this nucleus for 15 years, and we still don’t know what it does.’19 Nearly a decade later, neuroendocrinologist Geert De Vries pointed out again that scientists have ‘not gotten an inch closer’ to working out how this sex difference in the brain translates into behaviour. And not for want of trying.20 Demand a story that includes a clear hormonal beginning, a neat neural middle, and a convincing behavioural end and the best that researchers have to offer involves a small area of the brain stem that innervates the penis. Without wishing in any way to denigrate the painstaking work of neuroendocrinologists (or, for that matter, the glory of the male machinery), so far they are falling way behind in the schedule of scientific discovery that Brizendine and others blithely attribute to them.21
And even here in the brain stem the story turns out to be much more complex than it first seems.22 Celia Moore is a developmental psychobiologist at the University of Massachusetts who has put a lot of effort into trying to understand how early hormones bring about sex-typical behaviour in postnatal life. Is it really by way of some direct enduring effect on the brain, or is it possible that ‘early hormones set all manner of processes into motion that could converge on behavioral differences days, weeks, months, or years down the road. What about those canines developing in young male rhesus monkeys? What about size differences resulting from early hormones? What about the genitalia? or odours, or other socially important cues?’23
Moore set out to investigate this very idea in the rat. Rat mothers lick the anus and genitals of their newly born pups, and Moore noticed that male pups are licked more than females. The reason for this, Moore discovered, is that mothers are attracted by the higher levels of testosterone in the urine of male pups. When Moore blocked the mothers’ noses, they licked male and female pups equally; and female pups injected with testosterone were licked as often as their brothers. But most remarkable of all was the effect of this anogenital licking on the young rats’ brains. When Moore stimulated the anogenital region of untampered-with female rats, using a paintbrush, the penis innervating nucleus in the brain stem got bigger (although not as big as the nucleus of a male rat). In other words, the sex difference in the nucleus size was not just due to neonatal testosterone, but was also influenced by the different maternal treatment of male and female pups.24 Even our simple hormone-to-brain-stem storyline has a social subplot.
This should make us concerned that social experiences might also be involved somewhere along the path between hormones and behaviour, and this flags the danger of leapfrogging directly from one to the other. As Moore puts it, this approach leaves ‘lots of unexplored territory and many possible pathways, perhaps convoluted ones, from the early hormones and end points of interest.’25 We should bear this in mind when, in the next chapter, we look at this kind of research with humans (and other primates). Moore’s work gives us a glimpse into the ‘amazingly complex interaction of brain, hormones, and environment in creating behaviour. And if the process is complicated in rats, imagine how much more so it is in humans’, as Rosalind Barnett and Caryl Rivers point out in their book Same Difference.26
But scientists are stout of heart. In the 1980s, Norman Geschwind and his colleagues suggested a very complex theory, part of which involved the idea that the high level of foetal testosterone experienced by males slows the growth of the brain’s left hemisphere.27 Geschwind went on to suggest that this leaves males with a greater potential for ‘superior right hemisphere talents, such as artistic, musical, or mathematical talent.’28 The Geschwind theory is the Teflon pan of the scientific literature. While other, lesser, theories become dirty and unusable when pelted with disconfirming data, these simply slide off the Geschwind theory, which continues to survive and inspire despite important critiques all pointing to the conclusion that the current status of the theory should be an-ambitious-idea-that-didn’t-work-out.29 For example, as the neurophysiologist Ruth Bleier pointed out more than two decades ago, the very starting point of the theory – the idea that the foetal male’s higher level of testosterone brings about a more cramped left hemisphere – was inconsistent with a large postmortem study of foetal brains.30 More recently, a neuroimaging study of seventy-four newborns also found no evidence of a relatively smaller left hemisphere in males.31
But still, the idea that higher foetal testosterone somehow creates a ‘male’ brain that is superior in masculine things like science and maths, while lower foetal testosterone leads to a touchy-feely, ‘female’ brain, has tremendous appeal. Baron-Cohen’s hypothesis is an elaboration of the Geschwind theory. His idea is that low levels of foetal testosterone result in a female, E-type brain; medium levels yield a balanced brain; and high levels of foetal testosterone make for
a male, S-type brain. (And really high levels of foetal testosterone create an ‘extreme male brain’ that is good at systemising, really bad at empathising, and is also known as autistic.)32 Since there is overlap between the sexes in foetal-testosterone levels in the second trimester – some girls have higher levels than some boys – this would explain why some females are systemisers and some males are empathisers. But because, on average, males have higher testosterone levels, they will be more likely to have S-type brains. That’s the idea: how do we test it? It’s not that easy. Higher levels of foetal testosterone are strongly correlated with having a penis. That means that a correlation between foetal-testosterone levels and later sex-typed behaviour, or differences between boys and girls, could have nothing to do with foetal testosterone and everything to do with the different socialisation of boys and girls. But as we’ll see in the next two chapters, there are several ways around this problem.
What will they tell us about the biological basis of gender inequality?
Without testosterone interfering, your daughter developed not only female genitalia but a decidedly female brain … it is your daughter’s girl brain that will direct her female approach to the world.
—the Gurian Institute, It’s a Baby Girl! (2009)1
At this point in the book, you may have begun to be a bit suspicious of phrases like ‘female approach to the world.’ As we discovered earlier, a person’s approach to the world can depend on what kind of social identity is in place or the social expectations that are salient. The girl brain directs not so much a female approach to the world as a flexible, context-sensitive one. But that’s not to say that foetal testosterone isn’t doing something in the brain. And perhaps the most obvious strategy for working out what that might be is to compare the empathising and systemising skills of children and adults who were exposed to different levels of foetal testosterone. If girls with higher foetal testosterone are more masculine than girls with lower levels (and ditto for boys), then this could mean that children with higher foetal testosterone have brains that have been more ‘masculinised’ in utero. (Then again, it might not.)2
One technical difficulty with this approach, however, is that only extremely rarely is blood sampled from an unborn baby. This means that researchers can’t directly measure the amount of testosterone circulating in the baby’s blood. So what do they do instead? Some researchers measure maternal testosterone, the testosterone level in the blood of the pregnant mother. Other researchers measure the amniotic testosterone in the amniotic fluid (which is taken from the sac surrounding the foetus for the purposes of prenatal testing). Yet other researchers study adults and use digit ratio as a proxy for the foetal testosterone levels. The 2D:4D digit ratio is the ratio of the length of the second (index) finger and the fourth (ring) finger. This ratio is, on average, different in men and women. (Men tend to have longer ring fingers relative to their index fingers, while women’s index fingers are about the same length as, or slightly longer than, their ring fingers.) The idea is that prenatal testosterone levels influence digit ratio. These very different approaches all have something very important in common: researchers don’t actually know for sure whether what they are measuring correlates well, or even at all, with the level of testosterone acting on the foetal brain.3 We won’t let this hold us back. (After all, we’re only trying to find the biological roots to gender inequality, so why be fussy, right?) But it’s worth bearing in mind.
With all the nitpicking done, we’re ready to look at the evidence that the ‘female approach to the world’ begins in the womb.4 In a series of articles, Simon Baron-Cohen and his colleagues have described a large group of children whose mothers had amniocentesis in the second trimester of pregnancy. According to his hypothesis, higher amniotic testosterone should bring about worse empathising skills. So, does amniotic testosterone negatively correlate, in boys and girls separately,5 with frequency of eye contact at twelve months old with a parent during play, quality of social relationships at four years old (as assessed by the mother), propensity to use mental-state terms, scores on the child version of the Empathy Quotient (EQ; as assessed by the mother), and performance on a child’s version of the Reading the Mind in the Eyes test? The answers are, respectively: no;6 not really;7 not really;8 no;9 and yes.10 And before you get too excited about this last yes for the Reading the Mind in the Eyes test, even though performance correlated with amniotic testosterone, girls scored no better than boys.11 Expanding the scope of the search to include digit-ratio studies also yields little in the way of support.12
What about prenatal testosterone and systemising? Systemising, you will recall, is ‘the drive to analyze or construct systems’, and ‘[a] system is defined as something that takes inputs, which can then be operated on in variable ways, to deliver different outputs in a rule-governed way.’13 As the observant reader might have noticed, we have yet to encounter an actual test of systemising ability. Nor can we even assume that a strongly systemising brain is the best kind to have to become a top-notch scientist. Philosopher Neil Levy has suggested that ‘[i]ntelligence, even in the hard sciences, and even in innovation, is as much an “empathizing” power as it is systemizing.’ Albert Einstein, for example, described his breakthroughs as being the result of ‘intuition, supported by being sympathetically in touch with experience’ rather than the end point of a ‘logical path’.14 Nobel Prize winners agree. An analysis of the transcripts of interviews with these illustrious men and women of science found that the majority accept that there is such a thing as scientific intuition that is distinct from conscious, logical reasoning and that can take place in the absence of all the information necessary for logical reasoning. In fact, their descriptions of scientific intuition bear a striking resemblance to Baron-Cohen’s characterisation of empathising as ‘an imaginative leap in the dark in the absence of complete data’.15 As one Nobel Prize winner in chemistry put it, ‘Intuition, I always feel, is when we don’t have enough components and yet we have to construct a picture.’ And while of course logical reasoning is vital, this intuitive scientific process that many laureates described as helpful to them can be undermined if this is the only approach taken, as a laureate of medicine describes:
This apparatus … which intuits has to have an enormous basis of known facts at its disposal with which to play. And it plays in a very mysterious manner, because … it sort of keeps all known facts afloat, waiting for them to fall in place, like a jigsaw puzzle. And if you press … if you try to permutate your knowledge, nothing comes out of it. You must give a sort of mysterious pressure, and then rest, and suddenly BING … the solution comes.16
This is another point to bear in mind when we consider the strength of the evidence for prenatal origins to gender inequality in science. In truth, ‘[n]o perfect set of cognitive abilities that makes one a successful scientist has been identified’.17 (Needless to say, this makes the task of finding the prenatal origins of such success that much harder.)
But let’s just accept the assumption that systemising is an important key to success in science, and return to the data. A study from Simon Baron-Cohen’s lab looked for, and found, correlations between amniotic testosterone and something promisingly named the Systemizing Quotient (SQ) for children (filled in by the mother).18 Yet while some of the items on this questionnaire have a systemising-y feel to them (asking, for example, whether the child can ‘easily figure out the controls of the video or DVD player’ or ‘knows how to mix paints to produce different colours’), for many other questions one struggles to understand how they tap into a desire to understand input-operation outputs. In what way does minding ‘if things in the house are not in their proper place’, becoming ‘annoyed when things aren’t done on time’, or noticing ‘if something in the house had been moved or changed’ reflect a mind driven to understand the rules of the law-bound universe?19 I’m not the expert here, but I can’t help wondering if some of the items from the Fusspot Quotient accidentally found their way into the SQ
.
Slightly more on target is a study of the toy choices of thirteen-month-old children. Boys spent more time than did girls playing with the boyish toys, which were a trailer with four cars, a garbage truck, and what was somewhat unhelpfully described as ‘a set of three plastic pieces of equipment’. Are these systemising toys? I suppose you could make a case for it. You push a car or a trailer, it moves. And we’ll give the ‘plastic pieces of equipment’ the benefit of the doubt. Certainly, these toys are probably better candidates than the tea set, dolls, baby bottle and cradle with which girls spent more time than boys. But then again, the three gender-neutral toys (a plastic friction dog, a wooden puzzle and a stacking pole with rings), with which boys and girls spent equal time, seem at least as systemising as the boyish toys, if not more so. Not that it matters, since neither amniotic testosterone nor maternal testosterone turned out to be related to play behaviour anyway.20 (Disclaimer: When I say ‘boyish’ toys, I am referring to toys traditionally marketed to boys; likewise for ‘girlish’ toys.)
Nor do studies of correlations between amniotic testosterone and cognitive performance lend much support to the idea that higher prenatal testosterone is associated with greater skill on visuospatial tasks, mathematics, or other vaguely scientific-like skills. Does accuracy on a mental rotation test at age seven correlate with amniotic testosterone? No.21 Does a four-year-old’s skill at copying a block structure, understanding number facts and concepts, and counting and sorting increase with higher levels of amniotic testosterone? No, it decreases in girls, and has no relationship in boys. Puzzle solving? No. Classification skills (for example, ‘find all the small ones’?) No.22 A test of spatial ability? No.23 And again, while some digit-ratio studies do provide a spattering of support, others have failed to find correlations between digit ratio and SQ score, and mental rotation ability. One study even found that physical scientists have more-feminine digit ratios than do social scientists.24 There are a few more prenatal testosterone studies, which we’ll come to in a later chapter. But there is, I think, something a little underwhelming about the evidence so far.
Delusions of Gender Page 13