« Brass Crescent | Gene Expression Front Page | Spandrels, be damned »
January 17, 2005

Common sense breaks out at Harvard

...and naturally all right-thinking people are aghast.

As detailed by Michelle Malkin Lawrence Summers, president of Harvard, is into HBD with regards to gender differences.

Mr. Summers has enough cojones to stand up for HBD in public in what is one of the most politically correct environments in the US. I can't help but wonder what his real thoughs are on other, somewhat touchier admissions-related subjects...

(Nod to the Corner on NRO for linking Michelle)

Update from Razib: PZ Meyers responds in a rather hysterical manner. Starting off by stating that discrimination is the cause of "100%" of the male-female imbalances at elite universities, he seems to calmed down in the message boards and admits that he doesn't "think presence of an imbalance is necessarily a sign of discrimination. " He seems to be granting that it's a complicated topic, though he won't back off on his contention that Summers is biased.

Update II from Razib: Also, I note that PZ says the following: "I am certainly not assuming there are no genetic differences between males and females — that would be rather silly, don’t you think? I also expect that there are slight physiological and cognitive differences between them." I would observe that the physiological differences between females and males would likely not be characterized as "slight" by many, but, in any case, on the cogitive point, what exactly would a "slight" difference imply? The primary cognitive measure most people think of is some sort of IQ distribution, which ideally, though never in actuality, is conceived of as a normal distribution (a "Bell Curve"). One must remember that slight differences in the mean of two different distributions (imagine male and female mathematical aptitude) might result in very different numbers at the tails. For example, assume a "mathematical IQ test," where the mean for the population is 100, with a standard deviation of 15 points. Assume that the sex ratio is 50:50, if women had a mathematical IQ of 99, and men one of 101, a slight 2 point difference, while 0.47% of males would have scores above 140, 0.32% of females would have such scores. If a mathematical IQ of 140 is the minimum needed for someone to make enough of an impact to attain tenure, then you already have the die loaded toward males by a factor of 1.5. The SAT subtests have a standard deviation of 110, and in 2003 females scored a 503 on the math section, while males scored 537. Despite the modest mean difference, there tends to a rather large skew toward males at the very high end of the score distribution.

With all that said, I will not discount rational discrimination, stereotyping and societal pressures and expectations as factors in shaping the presence of women in any given field. But, that is not where it stops, the variables are confounded together and very difficult to tease apart, and many of the elastic social factors might emerge from slight differences in the means of variables which are less elastic. For every female discouraged from the sciences, there might be others who feel unduly pressured by teachers and parents who assume that because they have mathematical aptitudes they should become an engineer or physicist rather than a doctor, lawyer or musician (I have know of people in this situation). More females than males take the SAT and go on to college and earn degrees. More women attend law schools than males, and in medical schools the gender balance is nearing parity. Does this mean that all these fields are simply far less sexist than science and engineering? Or, is primatology far less hostile toward women than solid state physics? I've heard of some rather blatant sexism in science (usually in the form of so-and-so will not allow women into their lab, or so-and-so tends to hit on his female RAs). But, I don't see that the sexism is any worse than the stuff I hear about going on at law firms or at hospitals, where women are far more numerous and have made far greater strides after the formal and informal bars to their entry in these professions were removed.

Update III from Razib: Universal Acid observes that "There is also covert discrimination, such as...subtle forms of sociability like the guys in the department hanging around chatting about football." Agreed, men and women often have different focii in terms of what they like to talk about. But what would the solution to this be? A ban on non-gender neutral conversation in the office? Explicit regulations and laws can curtail overt discrimination, but the proposed solutions for the more subtle forms of bias (which often emerge out of preferences which might not be socially conditioned!) seem rather clunky and inappropriate in addressing the nuanced complexity of the situation at hand.

Update IV from Razib: Harvard Chief Defends His Talk on Women.

Update V from Razib: Not surprisingly, it seems Summers is aware that male cognitive phenotypes might display greater variance than those of females:


Summers referred repeatedly to the work of University of Michigan sociologist Yu Xie and his University of California-Davis colleague Kimberlee A. Shauman, who have found that women make up 35 percent of faculty at universities across the country, but only 20 percent of professors in science and engineering.

Their analysis of achievement test results shows a higher degree of variance in scores among men than among women. According to Ascherman Professor of Economics Richard Freeman, an organizer of the conference, the research found that “there are more men who are at the top and more men who are utter failures.”


Also, note how the assertion that there might be average differences in male and female aptitudes that translate into different representation in particular fields are transformed into value-based judgements:

Andrew G. Barr ’05, a government concentrator in Dunster House, said that “obviously my instinct is not to buy into any theory that there’s any sort of genetic flaw in women that prevents them from being good professors.

The vast majority of human likely doesn't have an innate mathematical aptitude that is a necessary condition to become a professor of science and engineering. That's not a "genetic flaw."

Update VI from Razib: I went to University of Minnesota at Morris's website and calculated that 28.5% of the science & mathematics faculty is female. If PZ resigned from the biology department, there would be a perfectly balanced ratio! (currently 5 males and 4 females)

Update VII from Razib: See the thread over at Yglesias' place. Another slash under the column "Why I am not a 'liberal'." One thing that is irritating, Summers rather qualified comments (they were embedded with a bunch of other factors and Summers admitted that he hoped he was wrong!) have been caricatured as bigoted anecdotal value judgements. Also, do English and Psychology graduate students and post-docs not have to put in "80 hour weeks?" Just curious, because one thing that is left out of the conversation seems to be the reality that in some fields women are very well, or over, represented. I can't believe math & the mathematical sciences (remember that in many fields of life science parity is almost there) are just so much more biased than other realms of study. Also, one thing I observed in high school, my calculus class had almost a 50:50 gender ratio. In physics there were 4 females out of a class of 30.

Posted by dobeln at 01:02 PM