Wednesday, April 09, 2014

Larry on La Différence

Larry the Barefoot Bum takes me to task for my suggestion that some disparities in male-female payscales might be "justified."  He says:
When we are talking about establishing differences between classes of human beings (which seem, quelle surprise, to usually be construed as inferiority), our null hypothesis should be that there are no differences, until evidence compels us to reject the null. I do not believe that we have anywhere nearly enough evidence to conclude that that women are substantively inferior... oops... different in capabilities than men.
I'd agree 95 percent, maybe 97, maybe 94, whatever.  Assessing differences is a perilous business at best, and turns invidious at the flick of an eyebrow.  It's  very like (but perhaps not quite like) the Hegelian insight that we can't know what "man (sic) in the abstract" looks like because nobody has ever seen man in the abstract, nor women neither, for what it is worth.

But there is a dangerous slippage underfoot here.  Back in the 60s (say) we all learned (or were taught) that we shouldn't assert differences between men and women.  We subtly tramsmogrified that mandate into the proposition that there are no differences between men and women.  Narrowly interpreted, this little two-step is incoherent: if we cannot know that any categorization of men versus women is empirically based, how can we know that it is not?

Actually (one reason Hegel doesn't apply here) my take is that there are a few--perhaps very few--differences that we simply cannot explain away as matters of culture.  My pet is the fetal damage through drug use.  So far as I know, there is no dispute on the proposition that the male fetus is more vulnerable to such the risk of such damage than the female. I first ran across that one about 20 years ago.  I haven't yet run across any possible basis on which this variation could be cultural.   It's small potatoes I suppose; I suspect there are others but for the moment I only need one to make my point.

Larry is quite right that we have a long history of using this kind of stereotyping in ways that are adverse to women (Does Senator Dianne Feinstein fail to understand torture because she is "too emotional?"  No, I think not.)  But Larry might be overlooking an important cultural shift: these days, it's at least as likely that the stereotyping is used in ways that are invidious to men.    On that latter point, FWIW, I'd have to confess that I am a culprit.  I do tend to think that men are on the whole idiots. their minds clouded by sexuality and a tendency towards violence.  I suspect (though  certainly can't prove) that traits like this are hard-wired.  I think we'd make a mistake to say that of course it can't be so when the fact is we don't know at all (nb, I think I have just said that I'm not sold on Larry's "null hypothesis," supra).

By the way, does anybody remember Ashley Montagu's Natural Superiority of Women?  It was published in 1952, i.e., when I was just starting my senior  year in high school; when I was, in other words, obsessively alert to the question of the truth or falsity of just that proposition.  I see there's an Amazon review saying that "Most men will not care for [the book] at all."  I'm not at all sure that that is true: I suspect that most men know they are oafs and are not pleased with themselves for being so.  Stereotyping again.

1 comment:

Larry Hamelin said...

But Larry might be overlooking an important cultural shift: these days, it's at least as likely that the stereotyping is used in ways that are invidious to men. On that latter point, FWIW, I'd have to confess that I am a culprit. I do tend to think that men are on the whole idiots. their minds clouded by sexuality and a tendency towards violence. I suspect (though certainly can't prove) that traits like this are hard-wired.

I have not overlooked that cultural shift, although I don't think it's as important as you seem to think it is.

I do not share your suspicion, though. On general principles, absent evidence to the contrary, I hold the null that apparent differences are enculturated. Of course, as a scientist, every null hypothesis calls for a test: I want to actually look at the evidence.

There's another point, though. We have tons of evidence that human beings are incredibly cognitively labile. If natural selection has done anything to our brains, it is to select not for hard-wired traits (other than perhaps the hard-wired propensity to acquire language) but for flexibility. Even if there were statistically significant differences among various groups, our general propensity for cognitive flexibility swamps those differences on an individual level.

Which brings me to the last point. Small differences in means* between groups give almost no information whatsoever about individuals within those groups, especially if the standard deviations of both groups are large relative to the difference in means.

*Or other measures of central tendency

I don't think there's any scientific question that is wrong to ask by virtue of its content (although there are certainly unethical ways of asking), but I don't see the research program of trying to find inherent or intrinsic differences between groups to be politically important if interpreted correctly, and are easy to interpret incorrectly.