Legacy of racism.

HUMAN intelligence is an eel-like subject: slippery, difficult to grasp, and almost impossible to get straight. Charles Murray and the late Richard Herrnstein make a heroic attempt to lay before the public a topic of writhing complexity: the interaction of intelligence, class, and ethnicity in America. The authors have not succeeded wholly, either in presenting the information or in convincing this reader of their conclusions, but I must applaud them for the clarity and honesty of their attempt. Who else has had the audacity to try to teach a nation raised on factoids and ten-second sound bites to think in subtle terms of probabilities, correlations, and standard deviations?

The authors' conclusions are so unwelcome that many readers will find themselves, as I did, slogging slowly and carefully through each paragraph, poring over every footnote, making irritated notes to themselves to seek out this or that study from the original literature to satisfy their skepticism. The research that Herrnstein and Murray summarize is exquisitely sensitive to the way a question is framed, so that the thinking reader cannot coast for even a paragraph without paying attention. But in the end, it all comes down to three questions: What do they say? Is it true? What should we do now?

Through summaries of myriad studies, the authors paint a vivid picture of the dark side of the American dream. The United States is the country of immigrants, the country where (at least in theory) name and family mean nothing and personal accomplishment is all. You can come to America with nothing, work hard, and rise to the top. This, Murray and Herrnstein show convincingly, is true if you are smart (and if, not incidentally, you are white). The consolidation of this meritocracy throughout this century has produced a class of smart, powerful, and wealthy individuals--the "cognitive elite"--who enjoy life at the high end of the bell curve.

But the shadowy inverse, rarely seen clearly, is also true: if you are not smart, you will fall to the bottom. The book tolls funereally, in chapter after careful chapter, ringing out the stunning relationship between low IQ and the tendencies to perform poorly in school; drop out of school; live in poverty; become dependent on welfare; bear children out of wedlock; go to jail; hold, perform badly at, and often lose menial jobs; achieve only a low socio-economic status; earn little money; maintain households that score poorly in factors important in nurturing children; and even suffer disabilities that prevent working altogether. The land of golden opportunity inevitably offers the chance to fail abjectly as well.

In the latter part of the book, Herrnstein and Murray present the fearsome possibility that cognitive class and race are now coincident. They report data that, as a population, African-Americans have a bell curve of IQ scores that is shifted to the lower side of the white mean. So do Africans, while East Asians have a bell curve of IQ scores shifted slightly to the right of the white curve. The authors are quick to observe that this does not mean that all blacks are stupider than all whites; there are many highly intelligent African-Americans who perform as well as or better than their white counterparts on the various measures of achievement. Indeed, one of the brightest points in the book is the demonstration that the average annual incomes of blacks, Latinos, and whites of the same IQ fall within a few hundred dollars of each other. But there seem to be disproportionately more blacks at the lower end of the bell curve and thus disproportionately more caught in poverty, ignorance, helplessness, and depression. Furthermore, considerable data indicate that those at the lower end of the IQ scale, regardless of race, are breeding faster than those at the top. We seem trapped in a downward spiral of ever-increasing stupidity.

Having sounded the death knell, Herrnstein and Murray do not abandon their readers to this vision of doom. Since they find eugenics an abhorrent policy, they suggest we revise the affirmative-action laws to reap the economic benefits of a more intelligent and more productive work force; find a "valued place" and useful occupations for those who are not very smart; strengthen the bonds of community responsibility and interdependence by reintegrating the cognitive elite into the rest of society; and encourage breeding among the cognitive elite so that the intelligence of our nation as a whole is not swamped by the fertility of the less intelligent.

But is it true? Do the data Herrnstein and Murray report about black IQ support their conclusions about black intelligence?

Underlying their thesis are two crucial issues. First is the premise that intelligence--of whatever it may consist--can be measured accurately and reliably by various tests, including the familiar IQ test. Herrnstein and Murray discuss the debates over psychometric testing fairly and clearly, and conclude that IQ and other such measures do reflect the elusive quality or qualities we label "intelligence." This point is the basis for the authors' compelling argument for the existence of a cognitive elite and its dark twin. The second issue is the heritability of intelligence. Heritability does not mean the extent to which a particular trait, such as intelligence, is genetically determined. Rather, heritability is the faithfulness with which a trait's measured expression (or phenotype, like IQ) mirrors the underlying genetic basis (or genotype). Heritability is always time- and population-specific, which is why the heritability of intelligence in studies ranges from .4 to .8 (out of a maximum possible of 1.0). Some populations have a genuinely higher heritability for intelligence than others, which renders cross-population comparisons of IQ and its correlates problematic (as the authors know).

The point is that the value assigned to heritability indicates the amount of the variation in measured intelligence that can be explained by genetic factors; heritabilities of .4 to .8 thus explain from a modest 40 per cent to a robust 80 per cent of the observed variations in IQ within the samples studied. Statistically, both values may be highly significant, if the sample sizes are large enough. Yet even a heritability of .8 leaves a substantial portion of the variance in intelligence to be attributed to something non-genetic.

If African-Americans have a lower heritability for intelligence, then their IQ scores are more heavily influenced by non-genetic factors. The authors reject the hypothesis that the mean IQ differences between blacks and whites are caused solely by environmental differences, commenting: "The average environment of balcks would have to be at the 6th percentile of the distribution of environments among whites, and the average environment of East Asians would have to be at the 63rd percentile of environments among whites, for the racial differences to be entirely environmental." This is the crux of the issue: Could the prejudicial treatment of blacks in America during the last two hundred years have been so crippling as to produce a downshift of the mean IQ by some 15 points? I find this thesis more plausible than do the authors, especially since most of the legislation outlawing racila discrimination is only about thirty years old. Little more than one generation ago, life was deeply different for blacks and whites in America, and social changes follow legislation slowly.

The issue cannot be resolved yet, but it deserves to be grappled with thoughtfully. I think Herrnstein and Murray missed an opportunity to examine the potential effects of prejudice on IQ and the other measures discussed here. They might have taken the circumstance of women and IQ as one way of establishing the pattern of changes that can be wrought by socialization. The advantage of looking at male--female differences (rather than racial or ethnic ones) is that men and women of the same racial background share the same gene pool; because it takes one of each sex to make a child, it is difficult to imagine how genes for high intelligence could become segregated in one sex or the other.

Today, it is acceptable and even admirable for white girls to be smart in school--at least until the age of 12 or 13, when school performance and girls' self-confidence plummet. In some segments of contemporary American society, females are encouraged to perform well all the way through college, with the critical drop-off point coming when the educated young woman wishes to enter graduate or professional school or obtain a job. The potency of the effect of discrimination over time can be seen by taking as an example the Johns Hopkins Medical School, a highly competitive professional school that trains many of the leaders of academic medicine in America. At its founding in 1892, the institution reluctantly agreed to admit women into every class on an equal footing with men--on pain of repaying a large financial contribution (plus accumulated interest) to a group of Baltimore woemn if the school ever failed in this undertaking. Yet not until 1994 did Hopkins Medical School have an entering class that was 50 per cent (indeed slightly over 50 per cent) female. Of the full professors at Hopkins who will teach this incoming class, only 11 per cent are female: an indication of the environment for bright women who entered the system twenty to thirty years ago. Progress was very slow for the first ninety-odd years of Hopkins's history, accelerating rapidly in the last decade.

Data in The Bell Curve trace the outlines of a pattern of discrimination. The mean IQ score for females is about 2 points lower than for males (versus the 15 points that recurs in black--white comparisons), and the variance of scores--the scatter around the mean--is more restricted in females than in males. This suggests that part of the bell curve of female scores has been truncated; given the lower mean score, a reasonable hypothesis is that the upper end of the bell curve has been cut off. The correlates of IQ are also distorted by discrimination. Although IQ correlates highly with job status, job performance, and income among white males and working women, the correlation becomes meaningless with the inclusion of women throughout the range of IQ who are unemployed or denied employment.

Discrimination against blacks in America, until very recently, has been far stronger and more pervasive than that leveled at women, and its effects can be expected to be more dramatic. While the brightest and most determined blacks have succeeded--as have the brightest and most determined women--there has surely been a cost to everyone else. It will take further insightful analysis to determine just how great that cost has been. By opening the discussion and daring to educate their readers, Herrnstein and Murray have set the stage for such work.

As The Bell Curve suggests, whether low intelligence is fostered by genetic inheritance or nurtured by a culture of poverty, it is nonetheless passed from generation to generation. Herrnstein and Murray make some brave and radical suggestions, presented out of a genuine sense of responsibility and a fervent hope for a better future than the one our current policies will produce. Their prognosis is one we must take seriously, whether or not we accept their interpretations of the IQ data; and their prescription for social change, though daunting, is one we must listen to carefully. Ask not for whom The Bell Curve tolls; it tolls for all of us.