The Sex Difference Evangelists
In analyzing the sex-difference claims of authors Susan Pinker and Louann Brizendine, let's start with language. Who hasn't heard that women are naturally more verbal than men—better at expressing themselves, better at reading and writing, chattier? These clichés crop up in various forms. In her book, for instance, Pinker emphasizes that girls speak earlier, outperform boys on various measures of verbal skill when they're young, and are less likely to be dyslexic. She notes that women have an advantage in verbal fluency. And in an interview, she told me that "huge differences in literacy" exist between college-age men and women. Meanwhile, Brizendine casts women as virtual talkaholics. The hardcover edition of her book asserts that "girls speak faster on average—250 words per minute versus 125 for typical males." It also claims that females use an average of 20,000 words per day compared to males' 7,000.
What is the scientific basis for these claims? Well-established literature suggests that girls tend to acquire language earlier than boys and are less likely to develop dyslexia (though the sex difference in dyslexia is less striking than some older research would suggest). But while adolescent girls may perform better on some tests of verbal ability, the gender gap is not large, according to meta-analyses assessed here. In the past couple of years, scores on the critical reading section of the SAT essentially show a dead heat for boys and girls: In 2007, they averaged 504 and 502, respectively. The new writing test on the SAT shows an advantage for girls, but it's small: In 2007, those male and female averages were 489 and 500. Sex differences on reading comprehension and vocabulary tests also appear to be small or close to zero, when all ages are taken into account. To some degree, differences in verbal ability in children or adolescents may reflect different paces of development that even out later on.
Some differences—for instance, on tests of verbal fluency—do appear in adults. (A typical verbal fluency test might ask people to list as many words as possible beginning, say, with the letter B.) But the differences between average men and women are small compared with the variation within each gender. For instance, if we take an average measure of verbal fluency for men, about 50 percent of men will score higher that that mark, and about 60 percent of women will. Which means that you'd do pretty badly if you tried to predict a person's gender from his or her verbal fluency score. What's more, these tests may have little to do with real-life communication. "When does any conversation call upon you to produce as many words as you can think of starting with B?" asks Deborah Cameron, a professor of language and communication at Oxford and author of The Myth of Mars and Venus. People may assume that "verbal fluency" means that women are more articulate or can find the words to express themselves better, she says, but that leap has not been substantiated.
Meanwhile, Brizendine's claim that women talk faster than men is unfounded, as linguist Mark Liberman has pointed out. Brizendine told me she omitted the two-to-one speed ratio from her paperback edition because she discovered that no primary sources verified it. Similarly, her assertion that women utter more words a day than men is bunk. Thanks in part to Liberman's provocation, last year University of Arizona psychologist Matthias Mehl conducted a new analysis of daily word budgets. He and his colleagues sampled speech from male and female college students, who wore recording devices that turned on every 12½ minutes throughout the day. The findings, published in Science, show that on average women use about 16,000 words per day. And so do men. (Brizendine says that this study convinced her to drop the 20,000-to-7,000-words-per-day claim. But her paperback still says that "on average girls speak two to three times more words per day than boys"—an assertion that is just as flimsy. Here's her explanation, and a critical response from the scientist she relies on.)
What makes the claims of a stark male-female split sexy, of course, is the appeal to neuroscience. Pinker, for instance, highlights a study of brain cell density, which suggests that female brains are more densely packed with neurons in an area called the posterior temporal cortex, which is associated with language. This "might explain the general female advantage in language fluency and spelling," among other things, she writes. Brizendine cites the same paper, asserting that in the "brain centers for language and hearing … women have eleven percent more neurons than men." But this paper looked at four men and five women—hardly a sample size to inspire grand claims. Neither Brizendine nor Pinker mentions this crucial caveat.And even if a difference in neuron density for that area were to be established, it's not at all clear what that would mean, if anything, given the complex circuitry that language involves.
These writers also home in on the structure that connects the brain's left and right hemispheres. Some research suggests that part of this structure, which is called the corpus callosum, is thicker in women than in men. This could mean that women have a "faster superhighway for neural messages" (Pinker) and therefore an advantage when it comes to language (as well as emotional processing). But that claim is tricky to make, and the significance of any purported size difference is deeply unclear.
Finally, Pinker argues that men "are simply less versatile when it comes to language." At first she seems to mean that they are more vulnerable to language-related problems like dyslexia, but in the paragraphs that follow, she slips from dyslexia to general measures like language fluency (and then back to dyslexia). So "less versatile" becomes a broader comment on ability, too. She suggests that men may rely more heavily on one brain hemisphere—the left—while women are more likely to use both. In particular, Pinker cites a 1995 study that asked men and women to answer questions about words and nonsense words, like whether they belonged to the same category or rhymed. Using fMRI images of the subjects' brains, the researchers found that men and women both relied on areas of the left hemisphere when answering questions. But women also used areas of the right hemisphere while men tended not to. According to Pinker, this means that if a problem occurred in the left hemisphere, "women would simply access the right hemisphere instead. Under normal circumstances sex differences would be subtle, but when things went wrong, sex differences would be extreme."
But since 1995 and the study that Pinker cites, a more complex picture has emerged, with some researchers finding that women are more likely to draw on both sides of the brain for certain language tasks, and others finding no sex difference. Maybe that's because of the types of tasks involved, but Pinker doesn't really discuss the controversy. Also, even if, on average, men and women used different neural strategies for playing certain word games, that doesn't necessarily mean one sex will perform better overall. This study, for instance, which asked subjects to put real and nonsense verbs in the past tense, found that women did tend to rely more on both brain hemispheres while men tended to rely more asymmetrically on the left hemisphere. But it made no difference in how quickly or accurately they performed.
Subtle differences could turn out to matter for men and women with specific clinical conditions. Pinker likes to say that the extremes tell us something about the rest of us. But the relevance is hard to see. (And, as it turns out, the connection between these gender-related brain findings and dyslexia is not well-established, either, however logical the connection seems.) All told, what's striking about the evidence on language is not so much a profound gap between the sexes, but the large gaps in our understanding of the brain.
Amanda Schaffer is a science and medical columnist for Slate.