Medical Examiner

Mars, Venus, Babies, and Hormones

According to sex-difference evangelist Louann Brizendine, women are like emotion-seeking F-15’s, deciphering and responding to other people’s feelings and needs. By contrast, it’s “only when men actually see tears that they realize, viscerally, that something’s wrong,” she writes. Brizendine and Susan Pinker not only argue that women are more empathetic than men, a claim that is dicey to begin with. They also say that the gap springs from innate difference.

Amanda Schaffer and Emily Bazelon discuss the science of empathy.

To make that case, these authors look to studies of infants, children, and twins. Pinker writes that girls not only “show more empathy toward friends and family” but “remarkably, demonstrate signs of these skills from early infancy, well before any cultural expectations about women as nurturers can be absorbed.” Newborn girls “respond more to the cries of another baby—and to human faces—than do boys,” Brizendine says. Pinker also argues that newborn girls show greater interest in looking at faces. And she hangs her hat on twin research, writing: “Studies of twins show that the ability to understand social situations—which requires empathy—is largely inherited, and that there are large differences between boys and girls that are most noticeable when children are young.”

What’s at play here are leaps between early rudimentary behavior and complex behavior later on. It’s true, for instance, that girls tend to make more eye contact and cry more easily in response to another infant’s cry, according to some research. But why is it particularly clear that these measures are relevant to empathy, which emerges later in development and involves a far more sophisticated set of responses? If baby girls are, in fact, more strongly drawn to some displays of faces than to objects, is that because they have a categorical preference for people versus objects, asks Harvard psychologist Elizabeth Spelke? Or because they’re responding to some other contrast between the two presentations, like “their rate of motion or distribution or color or contrast”? These data can be hard to interpret. In an interview, even Simon Baron-Cohen, another doyen of sex-difference claims, offered up some caution.

Caveats are also in order for the length of time infant girls gaze at faces and what to conclude from that. In one of his more controversial studies, Baron-Cohen found that 1-day-old girls were more inclined to look at a human face, while 1-day-old boys were more inclined to look at a mechanical mobile. But that work has not been replicated. Brizendine cites psychologist Erin McClure, but this reference is also problematic, as McClure herself pointed out to me. Meanwhile, an older body of research suggests that “male and female infants are equally interested in people and objects,” as Spelke puts it.

Pinker tops the slippery charts when she buttresses the case for innate difference with twin research. She cites a study that looks at twins and suggests a difference between boys’ and girls’ social understandings—which includes, say, the ability to pick up on body language or not to interrupt when other people are talking. Contrary to her explicit claim of “large differences,” when I calculated the gap between average measures for boys and girls, it turned out to be small —comparable, again, to the difference in average height between 15- and 16-year-old girls. Also, while the twin study Pinker likes does find that social cognition, or the ability to infer what others are feeling, is largely inherited (as Pinker correctly claims), its authors conclude that the disparity they observe between boys and girls cannot be attributed to genetic difference. Pinker, amazingly, fails to mention that the authors on whom she’s relying for proof of relevant genetic difference disavow that explanation of their findings. This is precisely the sort of selective reporting that makes her book misleading.

To shore up their claims that sex difference is innate, Pinker and Brizendine also fetishize hormones like testosterone and oxytocin, which they say may underlie crucial sex differences in empathy. For instance, they rely on Baron-Cohen’s argument that higher levels of prenatal testosterone diminish boys’ drives to empathize later on. His team is tracking a group of children born around Cambridge, England, some from birth through early childhood, and writes that, in general, fetal testosterone “predicts how sociable a child will be,” with higher levels of the hormone linked to lower scores on social measures.

But the evidence for that should be qualified. Baron-Cohen finds that at age 1, boys with higher levels of fetal testosterone appear to make less eye contact with their parents (usually their mothers). The ranges for boys and girls, however, overlapped significantly. (In the course of 20 minutes, the boys looked at the parent’s face between 3.0 and 46.2 times, the girls between 3.8 and 55 times.) At age 4, children with higher testosterone tended to have lower “quality of social relationships,” according to questionnaires their parents filled out. But that was only true when data for boys and girls were pooled. No relationship between fetal testosterone and the quality of social relationships was found among boys as a separate group. And none was found among girls, either.

For children between the ages of 6 and 8, the links between fetal testosterone and two measures of empathy were somewhat more convincing. Children with higher testosterone tended to score lower on a questionnaire and on a test in which they tried to discern emotion from pictures of eyes. And this association held when boys were considered alone. But, confusingly, on the eye-reading test, there was no overall difference in how well the boys and the girls performed. This work has not been replicated, either. Since Baron-Cohen’s results come from a nonrepresentative sample from one geographic area, his findings should not be treated as the final word.

Then there is oxytocin, which Susan Pinker calls “the hormone that greases the wheels of attachment” and “a feel-good, nurturing drug that happens to be homegrown.” Brizendine describes it rhapsodically in her “cast of neuro-hormone characters” as “fluffy, purring kitty; cuddly, nurturing, earth mother; the good witch Glinda in The Wizard of Oz; finds pleasure in helping and serving.”

Here’s what we actually know about oxytocin: The hormone is important to childbirth and lactation. It may also contribute to mother-child bonding and possibly to feelings of calm in breast-feeding mothers. Yet it is also linked to feelings of social distress. One theory is that the body releases oxytocin to promote social connection, and if that connection is positive, the hormone may help to reduce stress. If it’s negative, however, oxytocin may actually make stress worse. In other words, the hormone’s effects are apparently paradoxical—it is not simply a “feel-good” drug.

At the moment, research that includes a control group (and is therefore more rigorous) doesn’t tell us much about empathy and gender. Pinker emphasizes two studies: One finds that subjects who received intranasal puffs of the hormone were more trusting of other players in an investment game; the other shows that those who got oxytocin were better able to discern emotions in photographs of faces. Crucially, though, both these studies were conducted in men, as Pinker acknowledges. So far, for the most part, women haven’t been in the research pool, according to social psychologist Jennifer Bartz of Mount Sinai. This is starting to change, but the bottom line for now, she says, is that “we can’t say oxytocin makes women more empathetic.”

Finally, Brizendine and Pinker lean on neuroimaging studies, which compare male and female responses to stimuli like pictures of sad and happy faces or other imagery. But this kind of data is notoriously hard to interpret. Consider this meta-anlysis by psychologist Tor Wager, who looked at 65 functional MRI and PET studies of gender and emotion. Wager found some differences in the brain activity patterns of men and women in response, say, to films or pictures meant to elicit emotion. The differences were subtle, however, compared to the similarities.

And the kicker is that these studies don’t tell us whether differences are innate. Brizendine moves seamlessly from references to fMRI studies to phrases like “distinct female and male brain operating systems.” (She also jumps off the deep end with a claim about male and female mirror neurons.) Pinker suggests that fMRI studies can show how women’s “neural hardware” gives them an edge in discerning emotion. But our brains change in response to how we use them—what we think, see, feel, and practice doing over a lifetime. This is the plasticity of the brain, demonstrated most colorfully in this famous study of London cabbies. With its potential connection to a person’s response to the culture he or she lives in, plasticity could explain much—or potentially all—of the difference between brain scans of men and women responding to emotional stimuli. Pinker knows this and says she does not suggest otherwise. “You can’t look at a brain scan and say therefore we know the cause,” she told me. But because she and Brizendine largely devote their books to excavating innate difference, they should write that caveat in red.