Dec 24, 2010

Sex and the chimpanzee

I heard a story on All Things Considered the other night that set my hairs a-bristle. The story began, "For the parents who have been wondering, 'Why is it that my daughter plays with dolls and my son just wants a toy fire truck?' scientists reported this week that they've seen something similar among chimpanzees in a forest in Uganda." Immediately, alarm bells started going off in my head.

Whenever I hear parents say things like "but my daughter just likes pink" or "he just loves anything with wheels," my stomach tightens in fury. People seem to think that children are blank slates, unaffected by culture or social norms, so anything they do must be "natural" or "biological," especially as it pertains to gender. But researchers have uncovered minute ways that parents treat girl infants and boy infants differently from birth, and are looking into the effects of this disparate treatment on children's development. Other researchers have shown that toddlers recognize gender stereotypes by the time they're two years old. I would challenge anyone who thinks that their three-year-old boy "just loves trucks" to show me that he hasn't been given a huge number of trucks as presents, or been read tons of books about trucks, or heard lots of songs about trucks - many times more than the his three-year-old sister had. I find it amazing how many people shrug their shoulders about "how my kid got like that," but then continue to perpetuate this pattern by giving their daughters fairy books and their sons baseball books. This is not, of course, to say that everyone does this to the same extent or in the same way, just that I see it all the time and it makes me a little crazy.

So I'm always suspicious when I hear stories about scientific "experiments" that purport to show that gender differences are inborn or natural. I just find it impossible to believe that a penis or a bit of extra testosterone creates an innate love of fire trucks. To start with, that kind of inborn trait would require millenia of evolution, and fire trucks have only been around for about 150 years (although human-powered fire vehicles go back a few hundred years more). I'm not saying that there's no biological basis of gender, but I am saying that we have not yet found a way to separate the biological from the cultural and understand the contributions of each.


Anyway, back to the ATC story.

It reported on a Harvard professor of biological anthropology, Richard Wrangam, who decided that young chimps carried sticks the same way they would carry infants. Moreover, he "figured he would see young females doing it more often than males, because in many primate species, it's the girls who play with real babies." So he went through 14 years of data and looked at all the times that researchers had recorded "that a young chimp did something with a piece of wood." Lo and behold, the girl chimps "carried around sticks twice as often" as the boy chimps. Other studies have not replicated this finding, leading Wrangham to argue that this behavior is socially induced. He suggested that the particular chimp community he studied had a culture that promoted girls carrying sticks - er, babies.

You'd think that Wrangham's cultural analysis would salvage the study for me, and it did a little bit, but this is exactly the kind of study that gets my panties in a twist. There are so many points of human intervention here, as with all science, and at all of these points our own cultural biases come into play. In this instance they include: (1) Wrangham's decision that the original chimp playing with a stick looked like he was playing with a baby, (2) his assumption that therefore the chimps equated sticks with babies, (3) his hypothesis that when he looked at the data he would find girls playing with sticks/babies more often than boys, (4) the series of decisions by researchers over the 14 years about when to write down that a chimp was playing with a stick, (5) the decision by peer reviewers that Wrangham's study was noteworthy and deserved publication in the journal Current Biology, and (6) the decision by ATC journalists and editors that this journal report deserved national news coverage.

I just read a fascinating article by Jonah Lehrer in the New Yorker about the so-called "decline effect" in scientific research. This happens when a researcher finds a statistically significant experimental outcome but, when the experiment is replicated, the outcome is much less significant. For instance, researchers testing a group of antipsychotic drugs (the most well-known of which is Zyprexa) found a "dramatic decrease in the subjects' psychiatric symptoms" in the first studies of these drugs in the 1980s and 1990s. The drugs seemed to work extremely well. But more recently, when scientists have conducted studies of the same drugs, they've found a much smaller decrease in the patients' symptoms. Looking at the series of studies over time, Lehrer writes, "the therapeutic power of the drugs appeared to be steadily waning." The drugs hadn't changed, and neither had the patients (as far as anyone knew), but there was a decline in the statistical significance of the experimental outcomes - which is why it's called the "decline effect."

This poses a real problem for researchers because one of the premises of scientific experimentation is that any finding must be replicable to be considered valid. But why does this "decline effect" occur? No one knows, and it probably doesn't happen for the same reason all the time, but the most likely causes have to do with the publishing process and the researchers' intentions.


No one gets published for doing an experiment that finds an insignificant outcome. If you test a drug against a placebo and the drug's effects are no better than the placebo's, you don't get to publish an article in the Journal of the American Medical Association - but if your drug's outcomes are significantly better than the placebo, you do. That means that if 50 studies find no effect and 1 does, it's more likely that the outlier will be published than any of the other 50. There are, of course, exceptions to this generalization, but since the publication process prizes original discoveries, it stands to reason that statistical outliers often slip in among the actual discoveries, because both (the statistical flukes and the true discoveries) seem to be original. 

So, as ATC reported, other studies have not found any gendered differences in stick-carrying chimps. But Wrangham's, the outlier, is the one that got the news coverage.

The other issue Lehrer describes has to do with scientists finding what they expect to find. He describes a meta-study that did a statistical analysis of published findings on a particular topic, and found that the findings "skewed heavily toward positive results." There are lots of reasons for this, ranging from unconscious omissions to our desire to prove our hunches right, but they add up to the same thing: scientists intervene in their results all the time, and it's pretty hard to tell the objective data from the subjective stuff.

Our culture is science-obsessed. We love hearing about "studies," believing that information based on the scientific method is more valid that other information, and we like to attribute things we see in the world around us to scientific causes. Take, for instance, the search for a "gay gene," or a "smart gene," or an "obesity gene." But we must always remember that, just like history is the story told by historians, so science is the story told by scientists. They're people too, and they carry their own ideas into their research - as they're taught to do from the time they begin their first lab report with a hypothesis.

In a post-women's movement era, many people want to believe that we're post-sexism, just like people want to believe that we're post-racism. So they insist that there must be some "natural," "biological" reason that girls like pink and boys like blue, or boys like football and girls "throw like a girl." But these things are socially produced, and the more we hear that "chimps do it too," no matter the explanation, the more we'll think it's okay to reproduce those behaviors in the next generations of children.**

Recommended reading: An interview with Anne Fausto-Sterling, a biologist who studies the biological and cultural construction of sex and gender. Her book, Sexing the Body: Gender Politics and the Construction of Sexuality, is a must-read if you want to really learn about science and gender. You can also find out more about her ongoing research into how cultural and social behavior affects physical bodies, such as the development of bone structure and the brain - really fascinating stuff.


**except, of course, for those anti-evolutionists who refuse to believe that people are related to chimps. But I guess they have their own reasons for believing that men and women fulfill distinct social roles. You know, because of that whole "God created Eve out of Adam's rib" thing.

No comments:

Post a Comment