Bell Curve:

No bell, no curve, no...

Note:  The following is the full text of a letter to the editor

published in abbreviated form in the Nation, 268(8), September 21, 1998.

 

To the Editor:

If I were king, I might well order my scientists to find the genes for kindness, as Prof. Williams suggests in her column ("Invictus", The Nation, June 15/22, 1998).

But I'm not.  I'm only the editor of Psychological Science, and so I only get to sift and winnow the best, most interesting material from what is submitted to me.

So when a paper by M.J. Chorney, Robert Plomin, and their associates crossed my desk, reporting the discovery of a gene implicated in IQ, the decision to publish it was, as we say in cognitive neuroscience, a no-brainer.  It is now pretty clear to almost everyone that genetic factors count for a nontrivial proportion of population variance in IQ -- perhaps as much as half, with the rest accounted for by the environment.  If IQ is to some extent heritable, then there must be genes for it; and it
is of some theoretical interest to know what and where they might be.  The research in question was methodologically sophisticated, and the investigators entered appropriate cautions about the practical, as opposed to the theoretical, significance of their work.  I was happy to publish the paper.

Contrary to Prof. Williams's implications, my confident prediction about the rise of a new branch of the genetic testing industry wasn't based on any sense of scientific hubris -- much less an ignorance of the subjunctive mood.  Rather, it was based on a progressive analysis of American culture -- one largely informed, I might add, by an adult lifetime's reading of The Nation.  We're already well along the road to designer babies, after all.  Some parents, in search of the "perfect" child off the shelf, will want to know whether their kids have this gene, and some biotechnology entrepreneurs will be only too happy to charge them a fee for this information. 

But even from a strict scientific point of view, such testing would be a crummy idea. 

By Chorney et al.'s estimate, the gene in question, IGF2R, accounts for at most 2% of the population variance in IQ.  But IQ accounts for only about 10% of the variance in important social outcomes.  That means that IGF2R accounts for only about 2/10 of 1% of the variance in educational achievement, lifetime income, socioeconomic status, and the like.  Claude Fischer and his
colleagues, in Inequality for Design (Princeton, 1996), show clearly that the major sources of social inequality lie not in IQ, much less in IQ genes, but in the social environment.  Family funds would be better invested in higher school budgets than on genetic testing.

Never mind the pressures placed on children who test positive for the gene, and the adverse consequences for those who don't, when there is no reason to believe that an IQ of 160 buys the bearer much more success, or any more happiness, than one of 140 or 120 -- or, for that matter, 100. 

If I were a behavior geneticist interested in intelligence, I would worry about IQ as my measure of general intelligence ("g"), and indeed about the whole concept of g.  There are good reasons for thinking that IQ tests mostly measure acquired knowledge, not raw intellectual ability, that g is, at best, only a feeble trait, and that the real action in intelligence may lie elsewhere, in abstract reasoning ability and in specific cognitive abilities that are only weakly linked together.  Robert
Sternberg made these points powerfully in his review of The Bell Curve (PS, September 1995).  I was also happy to publish this paper: it was one of the few commentaries on TBC to offer technical, as opposed to ideological, criticisms of its assumptions, methods, and conclusions. 

And if I were an intelligence researcher, I would be more interested in environmental than in genetic contributions to intelligence.  And I would be especially interested in research by Joel Myerson and his colleagues, which re-analyzed TBC's data and found, despite Herrnstein and Murray's denials, that there are returns to schooling after all.  Specifically, while college-educated blacks, on average, don't get much out of their high schools compared to their white counterparts, they get much more out of college, so that a college education cuts in half the black-white disparity in IQ scores apparent at the end of junior high.  I was also happy to publish this paper (PS, March 1998): a better scientific justification for efforts to improve inner-city  high schools, and for affirmative action in college admissions, would  be hard to find. 
    
It is perhaps not surprising that the right-wing press would ignore these papers.  And it is perhaps not surprising that the New York Times, which seems obsessed with the genetic, evolutionary, neural, and hormonal substrates of personality and social behavior, would ignore them as well.

But if I were a columnist for this country's most important progressive journal of politics and culture, I'd certainly be interested.