There’s a simple story about sex differences in cognition, which traces these back to sex differences in early brain development, which are in turn due to hormone differences. Diagrammatically, it looks something like this:
Cordelia Fine’s “Delusions of Gender” (2010) accuses both scientists and popularisers of science with being too ready to believe overly simple, and biologically fixed, accounts of sex differences in cognition.
There is an undeniable sex difference in foetal testosterone in humans at around 6-8 weeks after conception. In Chapter 9 of her book, Fine introduces Simon Baron-Cohen, who seems to claim that this surge in male hormones is the reason why men are more likely to be Autistic, and why no woman had ever won the Fields Medal. So, diagrammatically:
This account may appear, at first, compelling, perhaps because of its simplicity. But Fine presents us with an antidote for this initial intuition, in the form of the neurodevelopmental story of a the spinal nucleus of the bulbocavernosus (SNB), a subcortical brain area which controls muscles at the base of the penis.
Even here, the route between hormone, brain difference and behaviour is not so simple, as shown by neat experiments with rats by Celia Moore, described by Fine (p.105 in my edition). Moore showed that male rat pups are licked more by their mothers, and that this licking is due to excess testosterone in their urine. Mothers which couldn’t smell, licked male and female pups equally, and female pups injected with testosterone were licked as much as male pups. This licking had an extra developmental effect on the SNB, which could be mimicked by manual brushing of a pup’s perineum. Separate work showed that testosterone doesn’t act directly on the neurons of the SNB, but instead prevents cell death in the SNB by preserving the muscles which it connects to (in males). So, diagrammatically:
One review, summarising what is known about the development of the SNB, writes ‘[There is] a life-long plasticity in even this simple system [and] evidence that adult androgens interact with social experience in order to affect the SNB system’. Not so simple!
What I love about this story is the complexity of developmental causes. Even in the rat, not the human! Even in the subcortex, not the cortex! Even in a brain area which direct controls a penis reflex. Fine’s implicit question for Baron-Cohen seems to be: If evolution creates this level of complexity for something as important for reproductive function, what is likely for the brain areas responsible for something as selectively-irrelevant as winning prizes at Mathematics?
Notice also the variety of interactions, not just the number : hormones -> body, body -> sensation in mother’s brain, brain -> behaviour, mother’s behaviour -> pup’s sensation, sensation -> cell growth. This is a developmental story which happens across hormones, brain, body, behaviour and individuals.
Against this example, sex differences in cognition due to early hormone differences look far from inevitable, and the simple hormone-brain-behaviour looks like a crude summary at best. Whether you take it to mean that sex differences in hormones have multiple routes to generate sex differences in cognition (a ‘small differences add up’ model) or that sex differences in hormones will cancel each other out, may depend on your other assumptions about development. At a minimum, the story of the SNB shows that those assumptions are worth checking.
Previously: gender brain blogging
Paper: Moore, C. L., Dou, H., & Juraska, J. M. (1992). Maternal stimulation affects the number of motor neurons in a sexually dimorphic nucleus of the lumbar spinal cord. Brain research, 572(1), 52-56.
Source for the 2009 claim by Baron-Cohen claim that foetal hormones explain why no woman has won the Fields medal: Autism test ‘could hit maths skills’.
In 2014 Maryam Mirzakhabi won the Fields medal.
Diagrams made with draw.io
Mind Hacks 
Great story! The bulbocavernosus is very useful for both males and females, at least in humans, in sex and to get the last drop out. It’s part of the muscle group targeted by Kegel exercises.
2 typos:
testostore -> testosterone
Mother’s which -> Mothers which
thanks Leo & I corrected those typos
I’m curious how exactly you think the argument structure goes here.
Hormones work in relatively direct ways in most cases; eg thyroid stimulating hormone stimulates the thyroid directly through biochemical pathways; it doesn’t motivate other people to engage in behavior that stimulates your thyroid. Let’s charitably say that 95% of hormone effects are like this.
So Fine reminds us that, in one case specifically selected for its weirdness, hormones can have an effect through a surprising pathway different from the strictly-biochemical ones they usually have..
Doesn’t that mean this should only change our minds if we were previously 100% sure that male mathematical ability was biological – and then only shift it down to being 95% sure? And then only if we have no other evidence either way?
Or to put this another way, the argument seems to be “Here is one bizarre case that works in a certain way, selected specifically for being unusual. Therefore, we should believe that this other unrelated case also works this way.”
Or to give an example of another argument that works this way which we can hopefully both agree is false, suppose we’re debating whether Barack Obama is lying about being born outside the United States. I search far and wide, and in the entire country, I manage to find one person who claimed to be born in the US but was actually lying and came from somewhere else (and hopefully you’ll grant that, in the whole US, there is at least one such person). Then I claim “See, Obama could be lying about being born outside the United States.”
It just seems like a strange way to support a position.
First, yes, I think the argument isn’t meant to be conclusive in any way – it is more of the type you suggest: it is meant to shift your 100% certainty to 95% certainty, to dent your confidence that the cognitive effects of hormones are mechanistically transparent.
Second, the context for this example (which is only half a page in the second section of Fine’s book) is to contrast the confidence with which some people interpret observed sex differences as due to the immutable effects of foetal testosterone, with the range of influences (hormonal, behaviour, social (simple and complex) that exist on those observed sex differences (sex differences like studying mathematics at Cambridge, winning the Fields Medal etc).
Third, in the book (again not quoted by me here), Fine suggests that this isn’t “one bizarre case”, but a model system which researchers have worked on for many years in an attempt to understand developmental neuroendoctrinological principles. Obviously, we may not accept this suggestion that the SNB not just one weird case, but there are reasons to think that it represents a simple case, and so one for which any discovered level of complexity is likely to be *less* than in other cases: it is subcortical, a reflex, in the rat, controlling a function which is directly selected upon / of high reproductive relevance.
This last gloss – that the SNB can guide our intuitions about a lower bound on the complexity of hormone-brain-behaviour developmental forces – is mine. I’m not a neuroendoctrinologist, so happy to hear otherwise, but it seems to be supported by the reviews cited in Fine’s book. Please send contrary evidence/references my way.
I guess the reason this annoys me so much is that I feel like the true situation is the opposite – most people seem a hundred percent sure that the differences are social, and any hint that they might be biological gets shot down with arguments like this. But that could just be my bias or my filter bubble speaking.
I am also not a neuroendocrinologist, but most hormones I know about work through pathways which are biologically very complex, but never veer off into the social realm (eg changing the smell of urine such that somebody else performs a behavior which then causes the desired effect).
“Mothers which couldn’t smell” (I assume “which” is a typo for “who”, saving the typo-ist’s feelings?). Anyone else wonder how these mothers got into that predicament? Especially when we learn that, surprise, surprise “even rats” are complex creatures whose maternal and developmental behaviours can shed light on ours? Doe no-one else find a tinge of sadness tainting their enjoyment of this interesting article?
@slatestarcodex
Yes, maybe different assumptions are common for the different “most people” around us. Is there a phrase for the situation where an argument is defined as much by the positions it is trying to exclude, and those are left implicit? It feels like lots of arguing / positioning on human nature, broadly conceived, is of that type. Non-foolish people don’t claim certainty on what is, just want to diminish other people’s certainty in positions that they believe are implausible.
Fine is very good in being explicit about the people/positions she is arguing against.
If I had more time I’d turn this into an empirical question: what do most people believe about the social and biological nature of sex differences?
Relevant: Misrepresentations of Evolutionary Psychology in Sex and Gender Textbooks http://journals.sagepub.com/doi/pdf/10.1177/147470491401200301