June 12, 2006

David Brooks, cognitive neuroscientist

In David Brooks' most recent column ("The Gender Gap at School" 6/11/2006), he observes that "reading rates are falling three times as fast among young men as among young women", and suggests that the problem is that "in most classrooms boys and girls are taught the same books in the same ways". His prescription is "more Hemingway, Tolstoy, Homer and Twain" for the boys. (I guess that the girls can make do with what they get now, though I suspect it is not, alas, mostly Atwood, Austen, Sappho and Alcott.)

I share Brooks' worries about educational gender gaps. I'm 100% in favor of more classic literature. And if there's an educational prejudice in favor of Jane Austen as opposed to Mark Twain, I'm 100% in favor of rectifying the balance. But there are two things about Brooks' column that bother me: he bolsters his argument with apparently misunderstood or made-up results of brain research on sex differences, and (if we fix those mistakes) he applies a black-and-white interpretation to observations of shades of gray.

Brooks starts his column by observing that men and women stereotypically like different kinds of books, relying on studies whose methodology magnifies group differences:

Researchers in Britain asked 400 accomplished women and 500 accomplished men to name their favorite novels. The men preferred novels written by men, often revolving around loneliness and alienation. Camus's "The Stranger," Salinger's "Catcher in the Rye" and Vonnegut's "Slaughterhouse-Five" topped the male list.

The women leaned toward books written by women. The women's books described relationships and are a lot better than the books the men chose. The top six women's books were "Jane Eyre," "Wuthering Heights," "The Handmaid's Tale," "Middlemarch," "Pride and Prejudice" and "Beloved."

Here's the first example of boosting the contrast to turn grays into blacks and whites.

Everyone knows that there are gender differences in reading preferences. I haven't been able to find any of the raw numbers from surveys like the one that Brooks cites, but I can guess what they would be like. We'd find that there are statistically significant group differences, sometimes strong ones, but we'd also find that individuals are complicated and groups are variable.

Speaking for myself, I dislike Camus and Salinger, and can take Vonnegut only in small doses. I like Austen and Atwood, but have to force myself through Brontë (any of them), Eliot and Morrison. I love Twain but find Tolstoy tedious. I'm hardly typical -- but then few people are, in the sense of conforming exactly to group patterns. I know a manly man with a secret habit of reading romance novels, and a womanly woman who loves military science fiction, even though the audiences for those genres are no doubt overwhelmingly gendered.

Adding up group preferences is a fine thing to do, but expressing the results by saying that "men preferred" one thing while "women leaned toward" something else is dangerous. It's like saying that "women preferred Kerry to Bush in 2004", when (according to exit polls) 48% voted for Bush as opposed to 51% for Kerry.

Brooks continues:

There are a couple of reasons why the two lists might diverge so starkly. It could be men are insensitive dolts who don't appreciate subtle human connections and good literature. Or, it could be that the part of the brain where men experience negative emotion, the amygdala, is not well connected to the part of the brain where verbal processing happens, whereas the part of the brain where women experience negative emotion, the cerebral cortex, is well connected. [emphasis added]

Or it could be that Brooks is deeply confused, not just about the meaning of "stark" differences in grouped preferences, but also about the literature on brain localization of emotional processing.

He warned us in his column of 5/25/2006 that

... I, a scientific imbecile, have spent several weeks trying to understand the amygdala and the orbitofrontal cortex.

Though I'm no expert in cognitive neuroscience, I think Brooks might need to go back to the books on this one. As far as I can tell from looking at (some of) the relevant literature, the amygdalas (paired subcortical structures at the inward tips of the medial temporal lobes) and the cerebral cortex are involved in the experience of negative emotion in a rather similar way in both sexes. While there both neural and behavioral group differences in emotional processing have been reported, the results are rather like those for book preferences: individually complex and variable within groups. And in any case, the pattern of sex differences is not, as per Brooks, men:amydala :: women:cortex.

To illustrate the kind of reading Brooks might (have one of his assistants) do, I hit Google Scholar with {amygdala sex difference}, and I'll summarize for you the relevant parts of one of the papers I found on the first page of results.

Turhan Canli et al. ( "Sex differences in the neural basis of emotional memories", PNAS vol. 99 no. 16, August 6, 2002) did a study that used functional MRI to study localization of brain activity in 12 men and 12 women "while they rated their experience of emotional arousal in response to neutral and emotionally negative pictures". Three weeks later, memory for the pictures was tested: "highly emotional pictures were remembered best, and remembered better by women than by men".

The study found that "activation in the left but not right amygdala was correlated with emotional arousal ratings for both sexes".

In addition to the left amygdala, men and women had other common (although not necessarily overlapping) areas in which activation correlated significantly with reported emotional experience. These areas included the bilateral superior frontal gyrus [Brodmann area (BA) 6], right middle (BA 46), and bilateral inferior frontal (BA 44, 45) gyri, left-lateralized anterior cingulate (BA 32), right precentral gyrus (BA 4), left thalamus, and left insula. This pattern of correlation loci suggests that both sexes share an extensive network of structures associated with attention, language, and motor control that are associated with emotional arousal. [emphasis added]

Against this general pattern of similarity across the sexes, there were also some (complicated) group differences:

Women but not men exhibited correlations in the postcentral gyrus and hippocampus. Men but not women exhibited a significant correlation in the putamen. Correlations in BA 37 of the fusiform gyrus were lateralized by gender: right-lateralized for men and left-lateralized for women. There was evidence for different patterns of hemispheric asymmetry between the sexes. Women had significantly more clusters in the left than in the right hemisphere ({chi}2 = 5.90, P < 0.05), whereas men showed no hemispheric asymmetry in the number of clusters in either hemisphere ({chi}2 = 0.12, P = not significant).

I found the last point interesting, and perhaps even relevant to Brooks' argument, since many studies have found that language-related phenomena are more lateralized in men than in women. But this is a sex difference in the degree of lateralization (i.e. left-right asymmetry) in activity of parts of the cerebral cortex, not a difference (that Brooks asserted to exist) between the amygdala(s) and the cerebral cortex as a whole.

As usual in such studies, the raw data from the Canli et al. study is not available. If it were -- and especially if a larger number of subjects were available -- we'd be able to see that there is a great deal of variation within the male and the female groups. It's quite possible (I would guess it's likely) that the within-group variation is quite a bit larger, both qualitatively and quantitatively, than the between-group differences.

In the memory test three weeks later, the researchers found that "highly emotional pictures were remembered best, and remembered better by women than by men". Here's the paper's detailed summary of the behavioral data, which most of you will not care to read:

There were sex differences in reported emotional experience and in memory for emotionally provocative pictures. ANOVA was performed with factors of "sex" (male and female), "arousal rating" (ratings 0–3), and "memory accuracy" (0, forgotten; 1, familiar; 2, remembered). Pictures that provoked high emotional arousal were more likely to be remembered than those that yielded little or no sense of arousal when collapsed across sex (arousal rating x memory accuracy, F(6,120) = 7.40, P < 0.0001). There was a significant interaction between sex and emotional arousal (sex x arousal rating, F(3,66) = 3.37, P < 0.025). Women rated significantly more pictures as highly arousing (rated 3) than did men [t(22) = 2.41, P < 0.025]. Women had better memory for emotional pictures than men (Fig. 1D); pictures rated as most highly arousing were recognized significantly more often by women than by men as familiar [t(20) = 2.40, P < 0.05] or remembered [t(20) = 2.38, P < 0.05]. There were no significant sex differences in memory for pictures rated less intense (0–2) or in false-positive rates (12 and 10% for women and men, respectively). Thus, women had superior memory for only the most intensely negative pictures even when subjective ratings of arousal were equated.

Easier to assimilate is the graphical representation of the results:

You can see the interactions easily in the plots: male subjects rated more of the pictures towards the low end of the scale of arousal ratings, while female subjects rated more of the pictures towards the high end; as arousal ratings increase, women showed an increasing advantage over men in percent recognition in the memory test. On the other hand, the differences were not enormous ones: the biggest (group) difference in percentage of arousal ratings is about 10%, and likewise the biggest group difference in picture recognition is about 10%. These differences are meaningful; it's worth exploring and explaining their causes and consequences; but we're talking about shades of gray here, not black and white. And individual differences among men and women, both in classification of emotionally-loaded pictures and in free recall of picture sets, will be much larger than these group differences.

There was another interesting sex difference found in this study. When they correlated the memory results with the (three-week earlier) fMRI data, they found (among a lot of other complicated stuff) "correlation clusters in the right amygdala for men (Tailarach coordinates, +16, –8, –17) and left amygdala for women (–25, –8, –17)". Here's a picture:

So we've got what looks like a meaningful sex difference (though I'd like to know more about individual differences, and I'd like to see the experiment replicated with different pictures, and so on), but it's not for to the experience of negative emotion, it's for the memory of pictures associated with negative emotions. And it's not men:amygdala :: women:cortex; rather, it's men:right-amygdala :: women:left-amygdala.

This reminds me of the old Soviet-era joke:

Question to Radio Yerevan: Is it correct that Grigori Grigorievich Grigoriev won a luxury car at the All-Union Championship in Moscow?

Answer: In principle, yes. But first of all it was not Grigori Grigorievich Grigoriev, but Vassili Vassilievich Vassiliev; second, it was not at the All-Union Championship in Moscow, but at a Collective Farm Sports Festival in Smolensk; third, it was not a car, but a bicycle; and fourth he didn't win it, but rather it was stolen from him.

In fairness to Brooks, I guess that because of the tendency for speech and language to be lateralized in the left cerebral hemisphere in most people (and somewhat more strongly in men than in women), you could imagine that the Canli et al. sex difference for memory of negative emotions (men:right-amygdala :: women:left-amygdala) might indeed have the result Brooks wants:

It could be that women are better at processing emotion through words.

Maybe, though there's a chain of about ten unsupported inferential steps between us and even a shades-of-gray version of that argument. Is there any conclusion whose empirical and logical foundations are adequately established by Brooks' column? In my opinion, there's at least one: we shouldn't accept any public policy recommendations on the basis of Brooks' understanding of cognitive neuroscience.

[I should add that there's another fallacy implicit in Brooks' use of neuroscience. He writes as if demonstrated group differences in brain activity, being "biological", must therefore be innate and essential characteristics of the groups, and not "socially constructed". But how else would socially constructed cognitive differences manifest themselves? In flows of pure spiritual energy, with no effect on neuronal activity, cerebral blood flow, and functional brain imaging techniques?

I ask this as someone who is quite prepared to believe in genetically-influenced cognitive differences. If such differences exist, let's understand what they are and decide what to do on the basis of the facts. But Brooks appears to believe that measured group differences in brain physiology are ipso facto evidence of innate cognitive differences, rather than different life experiences. If that were true, there would be no point in ever trying to teach anyone anything. ]

[Update -- Lameen Souag emails:

I wonder how Brooks would account for similar phenomena elsewhere, such as Qatar, where men's dropout rates are higher than women's even at primary school and more than twice as many women as men attend university, or Algeria, where 20%  more women than men make it to the baccalaureate, or Kuwait, where two-thirds of university students are women.  Learning styles yes - sitting down in one place and paying attention all day is a sore trial for most boys - but there's surely something broader going on here than choice of violence-filled vs. touchy-feely literature, never mind his further inferences about brains.

Qatar: http://lughat.blogspot.com/2006/04/more-from-qatar.html
Algeria: http://jazairana.blogspot.com/2006/06/60-of-bac-candidates-are-women.html
Kuwait: http://gender.pogar.org/countries/gender.asp?cid=8

Whatever is happening to schoolboys in Qatar, Algeria and Kuwait, I doubt that Brooks' explanation will work:

During the 1970's, it was believed that gender is a social construct and that gender differences could be eliminated via consciousness-raising. But it turns out gender is not a social construct. Consciousness-raising doesn't turn boys into sensitively poetic pacifists. It just turns many of them into high school and college dropouts who hate reading.

Seriously, I wonder what the global demographics of school attendence are -- how have things changed over the past few decades, with respect to sex and otherwise? If there is an overall trend for males to drop out or fall behind, to a greater extent than in earlier times, why is it? Could it be that boys are acting the way they always have, but girls are getting increased opportunities, and turn out to be better adapated (whether by nature or by nurture) to taking advantage of them? ]

[Update: more on the source of Brooks' misunderstandings, Leonard Sax.]

Posted by Mark Liberman at June 12, 2006 08:47 AM