The Larry Summers Debate Continues

As everyone knows by now the President of Harvard University has come under serious criticism for his recent remarks wherein he suggested that there was significant research supporting a biological explanation for the dearth of women in jobs as professors of math and the sciences. This debate is ripe for a treatment on Smidgen of Truth, because many of Summers' defenders have depended on exactly that to come riding to his rescue. Nevertheless, I had avoided blogging about this topic since it seemed that everyone else had already said everything that could be said on the matter. But, today I read exactly the kind of misuse of science that gets the blood of this old empiricists boiling.

You can skip on over to Captain Ed's post on the Captain's Quarters blog: http://www.captainsquartersblog.com/mt/archives/003921.php to make sure that I am staying honest. But here's my summary of his statement with commentary.

Basically Captain Ed makes a few common mistakes 1) he assumes that the brain reflects only biological effects, ignoring the fact that socialization can affect brain functioning; 2) ignores missing pieces of data that would greatly affect our ability to draw conclusions 3) he dresses up an assumption as fact with no empirical support for it and 4) he uses this faulty reasoning to draw up a big conclusion and basically accuses the opposition of being a bunch of screaming idiots who confuse politics with science.

Captain Ed begins by offering a summary of an article that summarizes some research that finds:

The human brain--male or female--is composed of about 40 percent gray matter and 60 percent white matter. When given intelligence tests, men used 6.5 times more gray matter than women, while women used nine times as much white matter.

First of all, I am not even going to go into the problems posed by the fact that Captain Ed's argument is based on what someone says the researchers say. Of course, reporters never misinterpret research or overstate the results to make a "newsworthy" story (and yeah, that's sarcasm you are detecting). But, second all this quote says is that there is a difference in the part of the brain that men and women use to solve problems in an IQ test. Note that this finding does not show that men are more likely than women to get the answers right. In fact, it could be that men and women arrive at the same answers using different skills which make neither of them inherently better or worse at the task.

However, the article goes on to note that there is research that suggests that these different types of brain matter are adapted to doing different tasks. For instance, the reporter says gray matter is associated with skills like map-reading, mathematics and what I think of as a fairly vague phrase "intellectual thought." My problem with this statement has little to do with what Captain Ed does with it and more to do with the fact that I have a problem with how the reporter is defining "intellectual thought." Namely, it seems to me that the reporter is assuming that the kind of thinking associated with men is "intellectual" whereas the kind of thinking associated with women is not. Read the following quote with this idea in mind:

Gray is central to processing information and plays a vital role in aiding skills such as mathematics, map-reading, and intellectual thought. White matter connects the brain's processing centers and is central to emotional thinking, use of language, and the ability to do more than one thing at once. Because women use less gray matter--critical to map-reading--they tend to have more difficulty with this skill than men.

Since when did "use of language" not constitute an aspect of intellectual thought. Frankly, how do we know a person is thinking intellectually unless they are able to express those thoughts via language? Also note the final sentence provides the first time we've been given a concrete task that women have difficulty with...that is...map reading...Now if you are thinking like me you say this might explain why women are less likely than men to be cab drivers but has little to do with sex differences in the number of Calculus professors.

But, neither Captain Ed nor the reporter are solely responsible for trying to link this finding to a more general question about differences in mathematically ability. It is the study's own co-author who suggests that such a link might be supported. See the following quote:

"This may help explain why men tend to excel in tasks requiring more local processing, like mathematics and map-reading, while women tend to excel at integrating information from various brain regions, such as is required for language skills," co-study author and neuropsychologist Rex Jung of the University of New Mexico told the Daily Telegraph. "These two very different pathways and activity centers, however, result in equivalent overall performance on broad measures of cognitive ability, such as those found on intelligence tests."

However, note Professor Jung's qualification that this difference in use of each part of the brain may explain differences. Suggesting that in his professional opinion the research does not yet reach a level of robustness that allows him to posit a definite causal relationship.

But, here's where Captain Ed really offends me:

As the article says in preface to these remarks, the study doesn't say that women can't do math; in fact, the study concludes that "human evolution has created two different types of brains designed for equally intelligent behavior." It does indicate that men have a more natural ability to perform in this area. People tend to migrate towards disciplines in which their natural talents allow them to flourish, and so a gender gap in the sciences -- especially engineering and physics, which rely so heavily on mathematics -- should not come as a surprise. Summers pointed out the gap and wondered if the reason might not be organic so that a solution could be found to overcome it. It appears that the relevant research has been found to affirm his suspicions, and that a lot of people owe Summers an apology.

So let's think this through, does the study show that men have more of a natural ability to perform in this area? NO!!! It shows that in adult men this area is more often activated than in adult women. This means that socialization could well play a part. As Bad Cat Robot says

I am also curious that no one has speculated that the differences in abilities (or in brain matter used) are *learned*. The brain changes with environment and use. There was a study done that compared the brain activity of London cabbies to ordinary drivers, and the cabbies had markedly more neural connections. Are you arguing that only very well-connected brain people are deciding to become cabbies? That something in the water makes a London cabbie more brainy than a New York one? Brain usage also changes after a stroke and people learn to compensate to work around the damaged area.

Following BCR, I'd argue that we might compare women who excel at map reading with men who do the same and see if we still observe this difference in gray and white brain matter activation. If we observe the same difference then we can begin to argue that such differences do not on their own explain abilities to read maps or do math. If the difference disappears, we can do a longitudinal study of young people and see whether their activation patterns change over time or whether in fact, there is a strong gender-neutral relationship between this pattern of brain matter activation and abilities.

Another problem with Captain Ed's conclusion is his assertation that "People tend to migrate towards disciplines in which their natural talents allow them to flourish, and so a gender gap in the sciences -- especially engineering and physics, which rely so heavily on mathematics -- should not come as a surprise." Ummm...okay, well this might seem like commonsense but we don't take anything for granted in science we test it. In addition, as a social scientists I would add the caveat that people also go into fields which are more structurally available to me. Which explains why for a long time almost no women at all went into any of the disciplines, because SOCIETY DIDN'T ALLOW IT. Therefore, we should never ASSUME that career choices reflect only what the jobseekers desire, we should test to see how true this is.

Despite these weaknesses, Captain Ed seems to think that this is the definitive evidence that Summers is correct. That is, Captain Ed is more convinced of the implications of the study than even the co-author himself.

One wonders, how Captain Ed would deal with the fact that there is no math/science sex gap in high school or college and the fact that on average women receive better grades than men in these subjects? Despite the fact that women are less likely to be called on in classes. Hmmm....so much for natural propensities.

Further, in the end, what does any of this have to do with explaining differences in tenure? The distribution of math/science abilities (not to mention map-reading) has little impact on tenure decisions since this a process that occurs not among the total population but among those people who have self-selected to go into these professions and have excelled throughout their lives in these subjects. It is the very nature of the problem that suggests that social (or even sociobiological, e.g. the biological clock and the social pressure to reproduce) causes are likely to be the strongest causal factors in the sex difference in tenure.

I could spend another couple of hours going over all the good social science research that suggests that Summers is wrong to say that discrimination plays little part in sex differences in tenure in math and science, but I'll save that for another rant.