Is intelligence fixed?

THAT people differ in intelligence, that the more intelligent will do better at many things, that IQ tests (and a good number of other similar tests) will give us a pretty good picture of how people differ in intelligence--there is not much to object to in all of this. Elaboration on these matters of common judgment makes up half or more of The Bell Curve. Even this much has been found objectionable in the past, and will be found objectionable today, and by people of, one assumes, high intelligence. Thus one federal judge in California has asserted, as a matter of law, that intelligence tests cannot be used to place duller children in classes designed to help them, and intelligence tests have been subjected to massive, book-length assault by distinguished scientists because some of those who devised and used them believed there were inherited differences in intelligence among races.

On these matters Herrnstein and Murray are to me completely convincing. On more controversial matters, such as the notion that there are substantial differences in intelligence among different groups defined by a common inheritance and culture, they are also convincing to me. We take these differences in intelligence for granted in the case of families, and it stands to reason that such differences might also characterize larger groups that share some common features in genetic inheritance and culture. Otherwise we would have no explanation for the disproportionate presence of Jews and Asians in selective high schools and selective colleges, or the disproportionate presence of these groups in such occupations as medicine, law, and college teaching.

We could give a variety of explanations for these phenomena short of any reference to genes--for example, the disproportionate presence of persons of high education among immigrants from a number of Asian countries (some of these groups have a higher proportion of college graduates than do native Americans), or the high socio-economic position they obtain on the basis of this education, and the background they are thus able to offer to their children; or the urban background of Jewish immigrants as contrasted with the predominantly peasant background of other immigrant groups that arrived here around the turn of the century. I would have preferred that Herrnstein and Murray had paid more attention to these differences as explanations of group differences in intelligence, owing to what has been made in the past of differences based on race. But on reflection, it hardly matters whether the differences are genetic or environmental; people can be brutal to those different from themselves regardless of the ultimate reasons for the difference.

On three further extensions of the argument, however, I would take issue with Herrnstein and Murray. First, how fixed are these differences? Second, are there interventions that could raise the test performance of persons and groups who score below average? Third, what about policies--that is, affirmative action--that set aside differences in performance in favor of group representation?

Herrnstein and Murray give some surprising data (surprising in the light of their argument that intelligence is fixed early and can't be changed appreciably through environmental intervention) on the degree to which differences between whites and blacks in performance on educational tests have been reduced in the past twenty years. The National Assessment of Educational Progress has been giving tests in science, mathematics, and reading to groups of schoolchildren of different ages since 1969. The reductions in the differences between white and black performance in standard deviations (a crucial statistical measure of difference that will not be further explained here, but that Herrnstein and Murray explain admirably in the book) range from .12 to .44. As they write: "The overall average gap of .92 standard deviation in the 1969--1973 tests had shrunk to .64 standard deviation in 1990. The gap narrowed because black scores rose, not because white scores fell." There has also been a narrowing of the gap in SAT scores, they report.

I wonder why more is not made of this. After all, while we have tried to do a good deal, through various programs, to put more money into the education of lowachieving, mostly black inner-city children, this has not been an overwhelming national effort. Simultaneously there has been a drastic decline in the environment of many of these children--more drugs, more crime, more illegitimacy. If so much has been achieved with relatively little, why can we not expect further progress? Among all the wonderful charts in this book, it would have been interesting to see one mapping this reduction and extending it into the future.

One of the more intriguing data comes from a study of the children of black American servicemen and German women in Germany. There is no difference in IQ between these children and the children of white American servicemen and German women. This reminds one of the older literature, inspired by the desire to counter racism in the 1920s and 1930s, on the change possible in what were considered fixed racial characteristics. Fifty years ago, one read Otto Klineberg, who seemed to have given a fatal blow to theories of the fixity of intelligence by showing that blacks in the North scored higher on general intelligence tests than whites (or perhaps it was certain groups of whites) in the South. I would like to have seen what later research has done to this argument, but there is nothing on regional differences in intelligence in the book, and Klineberg is not in the massive bibliography.

Finally, on affirmative action Herrnstein and Murray tell us much that is not generally known but has been available for a long time--ever since Justice Powell, in his opinion in the Bakke case, contrasted Allan Bakke's scores on the Medical College Admission Test with the remarkably lower scores of those admitted under the affirmative-action program; ever since Thomas Sowell began making his powerful arguments on the too-large gap between black and non-black students in colleges that aggressively recruit the former; ever since Robert Klitgaard, in his important book Choosing Elites, demonstrated how far down in the pool highly selective college and graduate programs have to reach to get substantial numbers of black students.

We know the story, but what is to be done? Once again, white students who feel they have been discriminated against by an affirmative-action program are suing an institution of higher education (the University of Texas Law School), and the Supreme Court will have to consider the matter. The documents in the case--no surprise--show that without the program of special preference, blacks would constitute only 1 to 2 per cent of the class, a fraction of the number now enrolled. The degree of preference could be less, the amount of perceived unfairness reduced. But I do not see how a country that has struggled so long, and still struggles, to make blacks full and equal participants can take a purely meritocratic position on such matters. If higher education served only to qualify students to become theoretical physicists or Sanskritists, we could remain indifferent to group consequences of purely meritocratic selection. But it does considerably more than that. Group representation must be a consideration, and all we can do is argue about the details.