For another take on the Summers story, click here.
When an uproar erupted last month over Harvard President Lawrence Summers' suggestion that genetic differences played a role in the lack of women in the sciences, his defenders claimed he'd been made a martyr to politically correct alarmists who were neither intellectually rigorous nor curious. Now that Summers' transcript has finally been released, it's clear that it was Summers' grasp of issues that was less than rigorous. Given the hornet's nest that Summers knew he'd be walking into by making such a presentation, what's surprising is that he would attend this conference prepared only to demonstrate a college sophomore's grasp of social science; as my colleague Will Saletan pointed out, Summers "grossly overreached the evidence." If you're hunting big game, though, you have to go armed with the right ammunition.
First of all, Summers relied on a naive notion of "choice." According to him, the leading cause for women's underrepresentation at top levels of the hard sciences is that women are far less willing than men to put in the 80-hour weeks that Summers believes top-level jobs require. As he put it, "It is a fact about our society that that is a level of commitment that a much higher fraction of married men have been historically prepared to make than of married women." To Summers' credit, he did ask whether contemporary workplace expectations are rational. But his discussion of the disparity between men and women's desires to work obsessively was framed by the implicit assumption that, given the status quo, men and women make free and clear choices. The central question, he said, is "who wants to do high-powered intense work?" (emphasis added).
The problem is that when it comes to choice, Summers gives very little credit to the notion that a woman who "wants" to do "high-powered intense work" faces a slew of obstacles that men don't. Some of these are obvious, others probably aren't. A woman is going to find it much harder than a man to find a spouse who is ready to tolerate her 80-hour work weeks and obsessive relationship to her job. She'll find it far more difficult to raise a family, and far harder to move her family from city to city if her job demands it. Men still do a lot less than 50 percent of the domestic chores and housework, according to recent studies. And working mothers don't feel they have the same kind of built-in professional peer group, to judge by the continuous arrival of books about the confusions and hostility that accompany modern working motherhood. These are all ways in which institutional and social factors shape both people's desires and their ability to act on them. But when a questioner asked if socialization might not have some bearing on what personal sacrifices women are willing to make for their careers, Summers responded skeptically.
In part, this is because Summers seemed to have an impoverished definition of socialization, or the cultural biases that affect individual performance. For Summers, "socialization" basically comes down to how much parents shape a child's identity, which is why he cited identical twin studies (which appear to show that parenting has less influence on factors like intelligence than we once thought it did) and invoked the example of his twin daughters "who found themselves saying to each other, look, daddy truck is carrying the baby truck" despite Summers' (self-described) gender-neutral parenting.
But the form of socialization that shapes performance in math and science has less to do with parenting than it does with the impact of collective beliefs on our identities—that is, how implicit and unconscious cultural attitudes mold an individual's sense of self. As a host of studies have shown, this shaping is powerful and begins remarkably early in our lives. To offer a simple but striking example: In a study about how gender perceptions affect our response to young children, participants were shown a videotape of a 9-month-old baby startled by a Jack-in-the-Box; when the participants thought the baby was a girl, they were more likely to think she was afraid; when it was a boy, to think he was angry. (Consider what a difference it makes if, from the age of 9 months, you treat one group of people as fearful and another as aggressive.) Perceptions have quite practical consequences: In math class, teachers are more likely to call on boys than on girls. More generally, although matters have improved, we still treat the hard sciences and math as essentially masculine domains. The phenomenon of the girl math geek who frets that she can't get any dates continues to be a stereotype for a reason. Even if there were no overt discrimination, these factors undoubtedly contribute to the complicated interplay of socialization, bias, and genetics that lead to high-level disparities later on.
But Summers' presentation didn't acknowledge the reality of this kind of culture-wide conditioning. There is little room in his view for the idea that the factors that lead people to decide to work in higher-level math—or, in the case of women, to opt out—might have social rather than biological origins. His presentation of the genetic "variances" that supposedly lead to higher-up discrepancies similarly relied on the notion that test scores are measures of pure intellectual aptitude, uncontaminated by social factors.
Summers' failure to take culture seriously was strikingly demonstrated in the Q &A session that followed his talk. One of the most interesting things about gender discrepancies in the hard sciences in the United States is that women in other countries—most notably France and southern Europe—are noticeably better represented in top-level departments. On the face of it, this seems to suggest that biology doesn't trump culture. (Even in today's climate, no one thinks the French are genetically different.) But one someone asked Summers about the difference between the United States and France, he responded, "Good question. I don't know much about it."
Summers' "provocative" mish-mash of biological essentialism and assertions about what women "want" to do was far less rigorous than the most of the critiques leveled at him. But, to Summers' credit, the debate of the past month seems not to have passed him by. In his most recent letter to the faculty, he wrote: "My January remarks substantially understated the impact of socialization and discrimination, including implicit attitudes—patterns of thought to which all of us are unconsciously subject," pinpointing the deepest problem with his presentation. Maybe Summers just said this to keep his job. (And needless to say, he shouldn't be fired for this misstep alone.) But let's give him the benefit of the doubt. Summers said his goal was to provoke thought. Now it looks like it did, for at least one person.