I’ve mentioned the unusually high incidence of certain kinds of weirdness among prominent mathematicians and physicists. I wondered if there may be a few deleterious mutations that increase intelligence but make you squirrelly – and there may be, but there is an easier way. Imagine someone with very high intelligence who is much more interested in math than any normal person – in a way that would have reduced fitness in past environments, and likely in this one as well. If you take two mathletes with equal talents and unequal obsessions, the guy who has a life is going to be far less productive.
This is fairly obvious, but I think that there is an interesting implication. The sort of personality we’re talking about is something like autism. I hate to use that word, because I think autism is probably a not a good natural category, but it’s useful in this case. Anyhow, high-functioning autism is far more common in males than females: maybe as much as 15 times more common.
At the highest levels of achievement in math and physics, men far outnumber women, maybe as much as 100-fold. This may account for a fair bit of that difference.