William Saletan's ongoing Maoist-style self-criticism for the crimethink of pointing out that James Watson knows more about the genetics of IQ than Watson's critics continues in Slate:

Not Black and White: Rethinking race and genes. By William Saletan

Five months ago, I wrote a series on race, genes, and intelligence. Everything about it hurt: the research, the writing, the reactions, the regrets. Not a day has gone by that I haven't thought about it. I've been struggling to reconcile two feelings that won't go away: that what I wrote was socially harmful and that I can't honestly renounce the evidence I presented. That evidence, which involved the proposed role of heredity in trait differences by race, is by no means complete or conclusive. But it's not dismissible, either. My colleague Stephen Metcalf summarized the debate better than I did: "It's a conflict between science and science."

When you find yourself in a dilemma this difficult, sometimes the best thing to do is let it sit in your head until you find a way to make sense of it within your value system. I think I'm beginning to find the answer that works for me: I was asking the wrong question.

In last fall's series, I asked myself why I was writing about such an ugly topic. "Because the truth isn't as bad as our ignorant, half-formed fears and suspicions about it," I concluded. "And because you can't solve a problem till you understand it." I wrote my commitment on a piece of paper and leaned it against my computer monitor: The truth doesn't care what you want.

Sometimes, with time and perspective, it's the small, overlooked things that turn out to be big. In retrospect, I was consumed by the wrong word. The flaw in my approach wasn't truth. It was the. Even if hereditary inequality among racial averages is a truth, it's less true, more unjust, and more pernicious than framing the same difference in nonracial terms. "The truth," as I accepted and framed it, was itself half-formed. It was, in that sense, a half-truth. And it flunked the practical test I had assigned it: To the extent that a social problem is genetic, you can't ultimately solve it by understanding it in racial terms.

A study published two weeks ago in Nature Medicine illustrates the point. Gina Kolata of the New York Times explains what happened:

Doctors who treat patients with heart failure have long been puzzled by a peculiar observation. Many black patients seem to do just as well if they take a mainstay of therapy, a class of drugs called beta blockers, as if they do not. [Now researchers] have discovered why: these nonresponsive patients have a slightly altered version of a gene that muscles use to control responses to nerve signals. … As many as 40 percent of blacks and 2 percent of whites have the gene variant, the researchers report. The findings, heart failure specialists say, mean that people with the altered gene might be spared taking what may be, for them, a useless therapy.

In other words, racial observation turned out to be a temporary step toward a deeper genetic explanation. Most blacks don't have the altered gene, and some whites do. Given these findings, prescribing or not prescribing beta blockers based on race rather than genes would be malpractice.

In a similar way, policy prescriptions based on race are social malpractice. Not because you can't find patterns on tests, but because any biological theory that starts with observed racial patterns has to end with genetic differences that cross racial lines. Race is the stone age of genetics. If you're a researcher looking for effects of heredity on medical or educational outcomes, race is the closest thing you presently have to genetic information about most people. And as a proxy measure, it sucks.

Okay, but the reason people get so irrationally upset when talk turns to race is because, much of the time, it's not a proxy measure: "Watch what you say, mister—you're talking about family here." People care about what you say about their races for the same reasons they care about what you say about their families. And that's not a metaphor.

To say that somebody is, say, white is not just a crude way of saying that they are unlikely to have the gene variant that makes beta blockers ineffective. It's actually much more of a way of saying that on, average, they are more likely to be genealogically related to another white person than to a non-white person. In other words, a white person has more family ties to white people than to nonwhite people. And who you are related to matters, in all sorts of ways, genetic, sociological, political, and personal, ways both subtle and bleedingly obvious.

It's irritating that after a full decade of my yammering away over and over again about a single insight that can clear up a remarkable amount of confusion in public discourse —that a racial group is an extended family that's partly inbred—confusion reigns unabated.