...but the 'you' was intended to be the impersonal and general 'anyone' you; I didn't mean you specifically. My bad.
Oh. Indeed. I'm sorry, I guess my english still lacks polishing.
My fault; in hindsight I wrote that in a way that was very easy to misread. Sigh. Must use more emoticons. Or less sarcasm.
I stand by my basic point, however - I think you're leaping from the facts you have to a conclusion that's completely unsupported. The data certainly suggests that there exists a reason why women do not pursue careers in physics research. I see no evidence that this reason is 'women are inherently less likely to be talented mathematicians'.
Rereading my post, I plead guilty on non sequitur. However, one thing is still right: there's no correlation (in my data) between the period that women usually have children and the rate of dropouts. And, anecdotally, I seem to recall my female teachers were all married with kids (except the lesbian ones). So, family-making is not a plausible explanation.
Cultural baggage is not either. It would have been about 30 years ago, but nowadays we're far more advanced.
I'd love to believe that, but consider it strongly unproven. The biases are certainly less visible; this isn't the same as 'gone away for good'. Perhaps older heads than mine can weigh in... have the career path requirements significantly changed since, say, the seventies? Because if not, then the old biases are still built in, and will remain so. (Certainly, all female mathematicians of my acquaintance claim that the career path is endemically biased in favour of men and of people who don't want families. But that's not a statistically significant sample.)
And if it were, we'd see the number of applications slowly climbing as we evolve, but they're mostly constant in the recent years.
In my country the number of applications was and is climbing, but that may be statistically meaningless due to a general expansion of higher education and several other external changes. So no hard data there.
So, there must be an alternative explanation. I have no serious data to support my hypothesis. But from my experience: the mean grade of the females was always lower than the male one (any chance of finding public records on this?).
As the group of females was always small, this is quite sensible to fluctuations. And, now I'm gonna sound like a real misogynist, none of the women I worked with was actually brilliant.
I don't find this misogynistic at all - it's a reasonable description of people you actually met. (It's anecdotal, but you knew that.)
Equally anecdotal: Grades for female undergrads in my department averaged slightly, but not significantly, higher. And the women I knew in my brief stint as a postgrad included several of the most brilliant minds I've ever worked with. The best of these had to endure an enormous amount of unjustified crap from her (male) supervisor, which held back her career, and still does... and which her successor as his student did not have any trouble with. But then, her successor was a lesbian.
If I wanted to stereotipise the groups (which I do), I'd say the females were hard working, while the males were slackers.
Hmmm... I'd agree, actually. But I'm not sure whether that supports your argument or mine. If any.
I could even imagine that, back to hunter-gatherer society, males had more necessity of understanding velocity, position and rates of change. Very useful in hunting. But I'm not a biologist, and this intuition would hardly do any good to someone studying quantum information.
I believe there is evidence that men rate more highly in spatial awareness... but I'm no biologist either.
Isolated, these aren't strong data, but collectively, and in absence of a better hypothesis, were enough to make up my mind.
This is where we part company again; if the history of gender/ethnic group/national/whatever relations tells us anything, it's that '[group X] are innately less talented at [activity Y]' is a natural, but extremely dangerous, default assumption. There too many easily concealed social biases for this one ever to be safe without strong, direct evidence. We know for certain that there were very strong social factors preventing equal opportunity for women in sciences, until (at best) recently. So any claim that these factors are now safely gone requires a strong burden of proof... it's certainly easy to name other areas where the biases have definitely not gone _anywhere_.