12/02/2007 05:36 pm ET | Updated May 25, 2011

William Saletan, Race, Genes and Intelligence

If you want to make people -- just about everyone -- mad, write something called "Created Equal: Race, Genes and Intelligence." That's what William Saletan did last week in Slate and, hoo boy, did people pop veins in their apoplexy (just put "race, genes, and intelligence" into Google to see the articles and the comments). The New York Times noted that Saletan's three articles drew "particular attention -- and particular scorn" (oddly, the NYT put its coverage of the controversy on Page A17 -- the Saturday Arts Page. In a fourth piece, Saletan said he should have been more circumspect and done less judging. Slate's editor, Jacob Weisberg said that as a senior writer, Saletan's articles are not edited but that these should have been -- carefully ("IQ Debate Add a Chapter Online, December 1). Critics spanned the full range from its-all-nature zealots to its-mostly-nurture advocates.

Saletan attempts to make the argument that we can reconcile racial differences in intelligence with our commitment that "all men are created equal" (he assumes that IQ tests are adequate measures of intelligence, something that will lose him a lot of votes at the outset). It's not easy and I'm not sure Saletan is up to it even though he presents much interesting data. I'm not familiar with all of the research. Conceputal and moral issues aside, I am not at all sure he's got the analytical skills to handle that data.

For instance, he writes "White populations in Australia, Canada, Europe, New Zealand, South Africa, and the United States score closer to one another [on IQ tests] than to the worldwide black average. It's been that way for at least a century." Well, no. A century ago, IQ tests did not exist. Alfred Binet had developed something he called "mental age" in 1905 creating tests that would screen out those French children who could not benefit from a regular school education. But Wilhelm Stern had not yet pointed out that the mental age of a child had very different implications depending on the chronological age of a child.

A child who is 5 with a mental age of 8 is way ahead of his peers, while a child who is 11 and has a mental age of 8 is way behind. Stern suggested dividing the mental age by the chronological age and calling the results quotient the Intelligence Quotient or, IQ. Lewis Terman at Stanford revised Binet's tests in 1916 and called it the Stanford-Binet. The first revision with appropriate norms appeared only in 1937.

At a different level, Saletan asks, "How could genes cause an IQ advantage? The simplest pathway is head size. New studies using MRIs show that Asian-American kids have larger brains that white kids who in turn have larger brains black American kids. This is true even though the order of body size and weight runs in the other direction."

These are interesting data, but then Saletan goes on to write, "The new science of MRI finds at least a 40 percent correlation of brain size with IQ." No, it doesn't. It doesn't because it makes no sense to write about a "40 percent correlation" of anything with anything. The correlation might be .40. This .40 is not a percent, just an indicator of the strength of the relationship (no relationship would be 0.0, a perfect positive relationship would be +1.0. It would mean that brain size accounts for 16 percent of the variation in IQ, not a huge amount (the amount of variation accounted for by one variable in another is given by the square of the correlation coefficient).

Or, he might mean that brain size accounts for 40 percent of the variation in IQ. In this case the correlation would have to be around .63 in order for its square to reach 40 percent and this seems to me improbably large. Whichever, his sentence as written makes no sense.

Saletan argues that genetic differences in IQ are reasonable because there are other reliably found differences among ethnic groups in other characteristics: "On average, compared with whites, blacks mature more quickly in the womb, are born earlier, and develop teeth, strength, and dexterity earlier. They sit, crawl, walk, and dress themselves earlier. They reach sexual maturity faster, and they have better eyesight." He suggests that in differing environments, "survival of the fittest" might have led to the favoring of different skills and capacities.

The most important phrase in the preceding paragraph, though, is "on average." All of these differences reflect average differences in distributions where the groups overlap a great deal. As Saletan says, "Remember, these are averages, and all groups overlap. You can't deduce an individual's intelligence from her ethnicity. The only thing you can reasonably infer is that anyone who presumes to rate your IQ based on the color of your skin is probably dumber than you are."