Granted, the study of racial and sex differences in intelligence has not exactly covered itself in glory. There was that unfortunate incident in the mid-20th century, when British psychologist Cyril Burt apparently made up data to "prove" that genes make blacks and the poor innately less intelligent than whites and the wealthy. Later studies reaching similar conclusions were based on statistics that would have done Mark Twain ("lies, damned lies …") proud. But does this sorry record warrant the scientific equivalent of the death penalty for such research? That's what some scientists are arguing. In a heated debate that began in the journal Nature and spread online, they are calling for an end to research on possible links between race, gender and intelligence. "Stupid science" and "evil science" are the more polite terms being hurled. But the arguments for and against the research are not what you'd expect.

Political correctness—as in, it's offensive and destructive to even ask if women as a group, say, are less intelligent than men—doesn't merit more than a brief nod, thank goodness. Instead, argues neuroscientist Steven Rose of Britain's Open University, the problem is that both race and IQ are slippery concepts. Standard measures of intelligence are ridiculously flexible. In the 1930s and 1940s, for instance, when girls kept outscoring boys, IQ tests were repeatedly adjusted to make the results turn out "right." That calls into question what studies of intelligence actually measure, and whether it's too easy to choose and massage data to produce desired results. Worse, "race" in the sense of Caucasian, Asian and African is too broad to capture anything biological, including genetic differences. Only smaller groupings based on geographic ancestry (Basque, !Kung, Inuit …) do. Since each "race" is a hodgepodge of ancestries, it's as hard to draw meaningful conclusions about how it relates to intelligence as it is to draw meaningful conclusions about food and allergies by studying a stew with 27 ingredients.

As for sex, there are indeed structural and biochemical differences between male and female brains. But since boys and girls, and men and women, live very different lives and are treated differently first by parents and then by society, it's impossible to attribute those differences to innate biology rather than experience. That is especially true now that discoveries in neuroplasticity have shown that brains of any age can change their structure and function in response to experience. Even the visual cortex, which you'd think is pretty hard-wired, can switch from processing sight to processing touch if you are blindfolded for just five days.

Kudos to defenders of studies of how intelligence varies by race or sex for basing their case on something other than the obvious grounds of academic freedom. Instead, they argue, the studies must continue because of the wealth of important knowledge they produce. In the 1960s, for instance, psychologist Arthur Jensen presented evidence that African-Americans are inferior in intellect due to inherited genes. That prompted psychologist James Flynn of the University of Otago, New Zealand, to examine decades of IQ data from dozens of countries, something he never would have done without Jensen's work to goad him. He discovered what is now called the Flynn effect. One of the most fascinating phenomena in psychology, the Flynn effect is the increase in IQ scores over the last 70 or so years. The increase, of about .3 points per year and as much as 25 points in some countries, reflects generational improvements in abstract problem solving, a product of a more complex, mentally stimulating modern world. The Flynn effect "shows that substantial increases in IQ can and have occurred over a short period of time," says psychologist Wendy Williams of Cornell University. "Genetics cannot explain such changes. Thus we look to environment … As experiences [for blacks] improve, so can and does IQ." That has already happened: one quarter of the IQ gap between black and white Americans has been erased in 30 years (it's now 10 to 15 points). Cultural effects are more powerful than we thought, says Williams, a conclusion that would have remained undiscovered if race and IQ were off limits.

There has been a parallel increase in understanding sex differences in IQ. Boys outnumbered girls 13 to 1 in the top .01 percent of U.S. math scores 30 years ago; now that's down to 2.8 to 1, providing more evidence for culture's effect on intelligence, in this case evolving beliefs about what girls can be good at. The fact that experience shapes the brain, and that girls' and boys' experiences are different so their brain differences might be the result of different experiences, seems less like an argument against studying sex and IQ than a fascinating research project: how do sex-specific experiences leave a footprint in the folds of the cortex?

If race-IQ studies are viewed as the third rail of psychology, many scientists who might answer questions like that will stay away from them. That will leave the field to those whose agenda is to prove women and blacks intellectually inferior. If that happens, warns Flynn, they will win the debate "because the rest of us have all adopted a policy of unilateral disarmament." On this fraught issue, science must not give up without a fight.

© 2009