Cambridge, Mass. -- AT A MEETING OF social scientists at the Harvard Business School last month, Richard Herrnstein's and Charles Murray's controversial book "The Bell Curve" came up.
One group reported that in an earlier conversation they had "trashed" it.
Heads around the room nodded in approval.
I asked the room at large -- about 20 people -- how many had actually read the book. Two raised their hands.
The condemnation of "The Bell Curve" in the media has been equally definitive.
Most of the analysis has focused on the question raised in the book of whether IQ is hereditary and whether racial differences in IQ are predominantly due to environmental or genetic factors.
The consensus appears to be that the book's argument is inherently racist and that Herrnstein (who died in September) and Mr. Murray are academic charlatans.
Yet, while their treatment of these issues has been justly criticized, much of "The Bell Curve" is not about race at all, and parts of it have been misrepresented.
For example, a frequent assertion about "The Bell Curve" is that it argues that intelligence is essentially inherited. In fact, the authors make the weaker claim that, according to existing research, between 40 percent and 80 percent of intelligence is in the genes. They adopt the middle of this range, 60 percent, as reasonable.
If you think this amounts to arguing that intelligence is "essentially" inherited, ask yourself whether you would be "essentially" receiving the same pay if you received a 40 percent cut in salary.
Herrnstein and Mr. Murray have been rightfully attacked for their shoddy and sometimes contradictory analysis of the relationship between race and intelligence. They acknowledge, for example, that there is no scientific way to determine even within broad ranges what proportion of the difference is due to environment and what proportion due to genes.
After offering this critical warning, however, the authors conclude that the racial gap is more likely genetic than environmental -- a divisive and irresponsible line of argument.
Yet, in spite of its serious flaws, "The Bell Curve" offers three potentially valuable insights that should not easily be dismissed. The first is that as a society we are becoming increasingly socially and economically stratified by level of cognitive ability.
This is an observation that has been made by others from widely different political perspectives, including Secretary of Labor Robert Reich.
The dramatic increase over the last two decades in the difference in incomes between high school and college graduates is strong evidence of this trend.
The second important assertion is that limited cognitive skills are strongly associated with myriad social problems. The authors find that among the poor, the unemployed, high school dropouts, those in prison, women on welfare and unwed mothers, 40 percent to 65 percent fall in the bottom 20 percent of measured IQ.
Most of these groups, by the way, contain more whites than blacks. Indeed, seeking to sidestep the race question altogether, the authors restricted a large part of their analysis to whites.
They find, as other social scientists have using the same data, that cognitive ability is a strong predictor of various social problems even when other factors such as family background are taken into account.
Given the strong suggestion of a link between intelligence and behavior, isn't further study of a possible causal relationship needed?
The third important claim in "The Bell Curve" is that cognitive ability is largely immutable.
Although the authors may well be overly pessimistic about the possibility of improving intellectual ability, surely we would be naive to think that simply increasing federal funding for early-childhood education, say, or for job-training programs would be sufficient to compensate for the increasing gap between the highly educated and the barely literate in U.S. society.
What are the consequences of ignoring such controversial but potentially important observations about our society?
Twenty-nine years ago, Daniel Patrick Moynihan, then an aide in the Labor Department, wrote a report that argued for an aggressive social policy to address the rising number of out-of-wedlock births in the African-American community, then about 30 percent of the total.
Today, nearly 70 percent of African-American children are born out of wedlock (as are 30 percent of white children, compared to about 12 percent in 1965).
However valid the warning, after the report was published Mr. Moynihan and his defenders were denounced as racists and the African-American family became a taboo subject for scholars for the next 20 years.
As we now try to grapple with the desperate situation of many black families in this country, we are missing two decades of research that could have informed current policy.
The furor about "The Bell Curve" risks the same perils. Many scholars are likely to back away from research on cognitive skills and social outcomes. Others will be inclined to present only findings consistent with the thesis that IQ and race differences of any kind are largely environmentally determined. This is hardly an atmosphere conducive to objective, rigorous scientific study.
Few of the most controversial assertions in "The Bell Curve" can be shown with any certainty to be either true or false. Only better, more unbiased and more sophisticated research can help us do this.
We need to ensure that neither the irresponsible statements in "The Bell Curve" -- nor the media's vitriolic response to the book as a whole -- prevents this research from being done.
In an era of increasing stratification by level of ability and income, it is critical that we understand what the relationship is, if any, between intelligence and entrenched social problems if we are to develop sensible public policy.
Christopher Winship is a professor of sociology at Harvard University.