Mark Kleiman writes:
For excellent evolutionary reasons, human males display higher variance than human females on many important traits, including measures of mental capacity. That means that they are likely to predominate among the top one hundredth of one percent of almost any cognitive talent, unless women are on average much better endowed in that particular department.I'm not sure what excellent evolutionary reasons Kleiman has in mind here.
Human beings evolved. Any theory of human nature therefore has the following constraint on it: it has to be consistent with evolutionary theory. But it is very important to see just how weak a constraint that is. An awful lot of theories about human beings are consistent with the theory that human beings evolved. This, in a nutshell, is why extrapolative evolutionary psychology gets intellectually dicey so quickly. Since so many different and incompatible theories of human nature are consistent with the fact of human evolution, the fact that we've evolved gives us no reason by itself to prefer any of the various theories on offer. In popular discussions, that usually means that the most cynical or crude theories get pushed on us by people who say that, sadly, they're just following the science where it leads them. Which is a pile of steaming crap.
Of course, claiming that evolutionary theory gives us reason to believe some theory or other of human nature is often just loose talk. Extrapolative evolutionary psychology typically involves a story about the environment of evolutionary adaptation (the EEA): say, that the early human environment favoured bias towards kin, so that that tendency was gradually selected for. But then we're not just talking about what evolutionary theory gives us reason to believe, but what evolutionary theory combined with (an often tendentious) story about early human life was like. More than that: we're usually asked to believe in addition (an often extremely tendentious) characterization of the relevant psychological states which would have produced the relevant advantageous behaviours in the EEA. And finally, we're often asked to believe that the relevant psychological states are hard-wired in some fairly inflexible way.
The point is not that the theories of human nature promoted in this way are false. It's that there are an awful lot of very thorny methodological issues involved in extrapolative evolutionary psychology, and so it's hardly surprising that we end up with an awful lot of scientific sounding just-so stories. Nor is it - or at least nor should it be - terribly surprising that we end up with an awful lot crappy scientific literature in the field that tends to basically reproduce established prejudices. The prejudices can make it very hard to even notice crucial methodological issues. Again, the prejudices might all be right in the end. But the evolutionary reasons cited for them will still be stupid, for all that.
Which is why I wonder what exactly Kleiman means when we says that we have excellent evolutionary reasons for "human males display higher variance than human females on many important traits, including measures of mental capacity." It might be that Kleiman has something else in mind. He's a smart guy, and I like his blog - I really hope he hasn't been taken in by crude, unreflective, warmed-over sociobiology.
Sorry to jump on this. The whole subject just really pisses me off.
I've been meaning to write about this subject for a long time, and to write about it at greater length and more coherently. But Larry Summers - the bastard - has everyone talking about it, and I couldn't resist chiming in with something a bit half-baked. If you want a serious discussion of the methodological problems in extrapolative evolutionary psychology, the best place I know to start is an obscure paper by a philosopher of science at Cornell, Richard Boyd (for whom I TAed a course in this subject). But, damn it, I've recently lent the paper to a friend, and can't remember the citation. If you're really curious, send me an email, and I'll try harder to track it down.
UPDATE: I noticed on re-reading this that I say a few times that the prejudices might for all I say be right. I hope it's clear that I say that in order to make a logical point - that I don't take the methodological worries I mention to decisively refute any particular theory by themselves. But for what it's worth I do think that the prejudices typically manifested in the scientific just-so stories are false.
Posted by Chris at January 19, 2005 08:29 PMGreat post, Chris.
And finally, we're often asked to believe that the relevant psychological states are hard-wired in some fairly inflexible way.
That really is an important point to make. Even if there's some true evolutionary tale about some specific trait or set of traits, a whole new argument is needed to demonstrate that the trait's manifestation is a foregone conclusion. Human beings are rational creatures, and for all of psychology's power as a natural, motive force, reason and reflection seem pretty darn powerful too. That isn't to say some Kantian-libertarian kazzam theory is true, whereby all natural psychological impulses can be overcome at any moment by the effort of our metaphysically sui generis Will. But it is to say that we have the intellectual capacity to diagnose our more base psychological compulsions and proclivities, and then to set ourselves on courses that are likely to change those proclivities and cultivate new dispositions.
The relation is tenuous, but this also gets me thinking about the sort of "cognitive psychology" that people are talking about vis-a-vis George Lakoff. There is all this talk about what psychological buttons we need to push, given that a person is either of the nurturant parent bent, or the strict father bent. The way to combat the more crude and dogmatic uses of this theory is to ask, "So, what psychological model explains Lakoff's theory? Was he driven by a nurturant nature or a strict nature when he sat down at his desk to think about cognitive Psychology?" This will produce a blank stare, but it should also, in time, get people to see that those two models can't be all there is to the explanation of the peculiarities of our mental lives.
True, we might not be able to explain everything that we want to explain. But you don't have to be religious in order to believe that there just are some mysteries.
Posted by: Paul at January 19, 2005 10:24 PMIn the environment in which humans evolved, the variance among males in reproductive opportunties was much greater than the variance among females, with the leading males siring most of the children and the bottom half of the male status distribution getting largely shut out. In such circumstances, characteristics leading to high variance will be reproductively rewarded. (In measures of risk-taking, men score consistently higher than women.)
Whether that particular explicans is right or not, the explicandum is indisputable: measure whatever mental ability you like, and men have a higher variance than women. That, in turn, surves as an explicans for the observation that there are more men than women at the very top of the distribution, even though women score as high, or higher, on average.
Posted by: Mark Kleimanq at January 20, 2005 02:58 PM"with the leading males siring most of the children and the bottom half of the male status distribution getting largely shut out."
When and where did this happen? What was *the* environment in which we evolved, and what (and how) do we know about the familial-social-political arrangements that characterized human life at this time (again, what time period are we talking about?) In other words, on the basis of what historical evidence do you advance this claim? And what has "leadership" in "the evolutionary environment in which humans evolved" to do with the kinds of traits/characteristics/abilities that are currently subjected to various attempts at measurement? (eg, what qualified the evolutionary ancestral male for leadership -- was it brains or brawn, say?)
"(In measures of risk-taking, men score consistently higher than women.)"
What kind(s) of risk(s), and how is it/are they measured?
