Daring to Discuss

Today the New York Times ran an article by John Tierney titled "Daring to Discuss Women in Science." Sounds brave!

The subject of Tierney's article is a piece of legislation just passed by the House called "Fulfilling the potential of women in academic science and engineering." According to Tierney, who's dubbed it "Larry's Law," this legislation is mainly a reaction to the controversy over former Harvard president Larry Summers. You may remember his comments in 2005 about women in science and math. He suggested that a lack of female professors in these areas might not be due to biased hiring, but to an innate difference in women's abilities. Larry Summers doesn't work for Harvard anymore. He does work for Barack Obama, though.

Tierney says:

This proposed law, if passed by the Senate, would require the White House science adviser to oversee regular “workshops to enhance gender equity.” At the workshops, to be attended by researchers who receive federal money and by the heads of science and engineering departments at universities, participants would be given before-and-after “attitudinal surveys” and would take part in “interactive discussions or other activities that increase the awareness of the existence of gender bias.”

I agree that "attitudinal surveys" sound a little goofy. But the legislation also calls for, at these workshops, "research presentations" (that's actually right before "interactive discussions" in the quote Tierney chose; I guess he wanted it to sound sillier) and "information on best practices and the value of mentoring undergraduate and graduate women students as well as outreach to girls." Mentoring children doesn't sound so silly to me.

"At the risk of being shipped off to one of these workshops," Tierney says (good one, John!), he'd like to discuss "new evidence supporting Dr. Summers's controversial hypothesis."

Here's how Tierney describes Summers's theory:

Yet even if all these social factors were eliminated, he hypothesized, the science faculty composition at an elite school like Harvard might still be skewed by a biological factor: the greater variability observed among men in intelligence test scores and various traits. Men and women might, on average, have equal mathematical ability, but there could still be disproportionately more men with very low or very high scores.

What Summers said, and Tierney repeats, is that assuming men have "greater variability" in intelligence, there will be more outliers who are men. So if you look at the top .01 percent of the population, the super geniuses, most of them will be men. And by extension, I guess, if you are the president of Harvard and you only want to hire super geniuses, it's not your fault if all of them happen to be male.

The "new evidence" that Tierney cites to support Summers's theory is a study done at Duke on SAT scores of seventh graders. (Yes, 13-year-olds.) In the super genius portion of the bell curve--the top .01 percent of kids--there are four times as many boys as girls.

The 4-to-1 ratio was reached in 1991. Yet only a decade earlier, it was 13-to-1. This crazy increase in super genius girls happened, the researchers assume, because of "sociocultural factors." But since the ratio has stayed at 4-to-1 since then, they think this difference is innate: "Our data clearly show that there are sex differences in cognitive abilities in the extreme right tail."

Here's the first obvious problem I have with this: Did I miss some sort of huge breakthrough where we discovered how to measure intelligence, or "cognitive abilities," completely removed from cultural influences? Did we fix the problem where African American students do worse on tests when they have to check a box for their race? Or where women do worse on math tests when men are in the room? Suddenly we can look at middle schoolers' test scores and say that they "clearly show" something about their innate intelligence as adults?

Secondly, I thought we were talking about biases against women in academia. Tierney claims that maybe there is no such bias, and promises to address the question in his next column. In the meantime, why on earth are we talking about possible cognitive differences in one ten-thousandth of the population? What does that have to do with biased hiring or funding? Even if a cognitive difference exists, does that mean we shouldn't bother assessing people based on their actual academic work?

Is it possible that there's an innate cognitive difference between women and men? Sure. I'd bet there are lots of them. We have different genes and different brain structures. But right now there's no way to tease out a cognitive difference from every other difference in how a person has been treated and formed by their family and society since birth. So I don't see the point in talking about it. All it accomplishes is to give a prominent Times writer a platform to describe workshops promoting women in science as "re-education."

If you feel as discouraged as I do, please check out this girl. She's eight, she loves science, and she has a business card with robots drawn on it.