In mid-February, a couple of weeks before I began to write this piece, I downloaded a copy of the transcript of Harvard President Lawrence Summers' remarks to the Conference on Diversifying the Science and Engineering Workforce, held in Cambridge, Massachusetts, in January. Everyone had been talking about what he was purported to have said. Figuring out what was accurate, without the transcript, turned out to be as difficult as the questions raised by the conference itself.
So what did he say? Well, he was certainly hypothesizing, and he was liberal in his use of qualifiers, but he did propose three explanations for the small numbers of women in high-end scientific careers: first, the possibility that women are less attracted than men to "high-powered jobs"; second, the hypothesis of differential "availability of aptitude at the high end"; and third, the effect of "different socialization and patterns of discrimination." One can't really object to such speculation, though perhaps Summers crossed the line when he announced that he believed "their importance probably ranks in exactly the order that I just described." That's a conclusion that cries out for evidence, which he seems not to have had at hand.
Summers was more than a month tardy in releasing his remarks, perhaps because he was speaking "unofficially" and seeking to be "provocative" (he certainly succeeded!) and perhaps because he was hoping the whole thing would blow over. If he was hoping for the latter, he'd simply been spending too much time in Washington in the recent past and not enough time on the nation's campuses. The news cycle in the nation's capital might be lightning fast, and yesterday's news may be cold by breakfast time, but some issues in academe have real endurance. And equity issues are rightly among them.
Finally, Summers asked, "What to do?" And he did offer some sensible ideas, including child-care subsidies for university faculty members (though why he sees that as a male/female issue may reflect more bias than he's aware of). But simple solutions are not likely. …