Steve Pinker Responds to Ron Unz

Professor Richard Lynn has sent us a June 3 note from Professor Steven Pinker of Harvard to Ron Unz. Professor Pinker told Prof. Lynn that if he wished to circulate the note, it should include the following statement by Professor Pinker:

You’re welcome to circulate it on the condition that you add that I do NOT endorse your theory that cross-national differences in IQ have a genetic cause. It’s not that I would argue that these differences are necessarily 100% environmental, but I remain skeptical on this issue (while maintaining that the debate be fair, open, and accurate – hence my objections to Unz’s characterization).

Hi, Ron,

Interesting piece, and it’s important to engage with these issues. I think your argument has some (fixable) flaws, though. You characterize Lynn as claiming that 100% of the variance in intelligence (within and across groups) is genetic. For example you often describe his position as “determinism,” and the technical sense of a deterministic process is one with probability = 1. That allows you to cite evidence of changes of between-group rankings over time as a refutation of Lynn. But he didn’t make that claim, and so your argument doesn’t work–indeed, is just the kind of straw-manning that Gould was repeatedly guilty of. Say that he claims that half of the variation among groups is genetic (that’s not a bad ballpark for the variation among individuals within a group). Then all of your findings are completely compatible with Lynn–they would pertain to the 50 percent of the variance that is not genetic–and he would still be right that in a world in which all economic, political, and geographic differences were somehow annulled, there would still be substantial differences in the economic and technological fortunes of different countries. Which of course is still an incendiary statement in today’s intellectual climate.

Now, you could make your argument more respectable by reformulating it as follows. IF Lynn’s argument for genetic variance across countries is based SOLELY on the differences in mean tested IQ between countries (together with an extrapolation of the genetic contribution of within-group differences), you could use the existence of non-genetic variation over time to question that inference–viz., by showing that a genetic explanation of the difference in IQ scores is otiose; we already have an environmental cause that is sufficient to account for his data. But saying that historical variation makes the genetic contribution to the crossnational differneces otiose or unnecessary or unparsimonious is different from saying that it refutes such a hypothesis. It only refutes the hypothesis that genetic determination is perfect, which no one except for Gould’s straw enemies believes.

And of course it would have to be true that Lynn had no data supporting a genetic hypothesis (even a partial one) other than the national differences. I haven’t read his book so I don’t know whether that’s fair. In the case of racial differences within the United States, Jensen and Rushton do have additional data, such as that when socioeconomic status, income, education, and the like are all thrown into a regression, the black-white gap doesn’t go away; the fact that the children of black and white couples matched in IQ regress to different means; and others. This is not to endorse their arguments, just to say that if Lynn has similar ancillary data (other than the existence of variation) then your arguments are not enough to prove that the genetic contribution to group differences is 0; all you’ve shown is that it’s less than 100 percent, which Lynn has always acknowledged.

A few other points:

-You may want to give Tom Sowell credit for making this argument first, in several books, including Race and Culture, and the second edition of Intellectuals and Society.

-You write as if Lynn were a well-respected psychologist whose findings have been widely accepted. This is very far from the case. Outside the circle of a handful of bloggers and behavioral geneticists he is somewhere between obscure and radioactive. (I believe several of his books are either self-published or put out by fringe publishers.) This is not to say that such a reputation is deserved or not, but it would be a mistake to imply that you’re arguing against a widely accepted hypothesis — Lynn’s hypothesis is anathema to 99.99% of psychologists and, for that matter, academics.

-You suggest that the evidence for a genetic contribution to variance within groups (as opposed to between groups) is based solely on a small sample of monozygotic twins reared apart. This is not true. It is based on a vast amount of data, including comparisons of identical and fraternal twins reared together, comparisons of biological versus adopted siblings, studies of full- versus half-siblings, studies of inbreeding depression, and other data. Many of these samples are enormous, coming from countries that administer IQ tests to every schoolchild and draftee and make the data available to researchers.

Despite these caveats in your argument, I don’t disagree with the conclusion. But I think you need to firm up the argument and attributions. Thanks for sharing it,

Steve