So here we are again, debating the tired and discredited thesis propounded by Charles Murray and his late co-author Richard Herrnstein in their book The Bell Curve about the role of intelligence in socioeconomic success and as its pertains to race. These issues have resurfaced in a new book by David Reich and Murray has once again returned to the spotlight. Andrew Sullivan who, as then editor of The New Republic gave a huge amount of space to publicize that earlier book, has emerged to defend Murray’s ideas. Sam Harris has once again aligned himself with odious ideas and defends them as courageous truth-telling in the fight against political correctness.
Crip Dyke has done some excellent posts on this (see here and here) as has PZ Myers (see here and here). There is also a critique of Reich’s book signed by 67 scholars in a range of disciplines that touch on this issue. Rather than try to cover the ground well covered by them, I thought I would provide excerpts from a long article that I published way back in 1995 after The Bell Curve came out that analyzed their argument in great detail, and deal with some issues not as well-covered in the current debate.(Race and Intelligence: What are the Issues?, Phi Delta Kappan, December 1995, p. 271-278, citations omitted.) For some reason, WordPress is not allowing me to upload the pdf of the article, so I give below the main points.
Murray and Herrnstein make much of the 15-point gap in average IQ scores between black and white students. After dealing with some general background problems (the reification of IQ scores and their changing meaning and use over time, race being a social and not a biological construct, the problems involved with measuring heritability and inferring causation from correlations, heritability and determinism, the ontological prioritization of genes, and problematic data) I then turn to the book’s arguments.
Simply stated, [Murray and Herrnstein’s argument] goes like this: If the variation in a particular individual trait is caused by genes, then the difference in average values of the trait in populations must also be caused by the genes. In the I.Q. debate, this argument takes the following form: variation in I.Q. among individuals in a population is caused (to a large extent, at least) by each individual’s genes. For group A, the average I.Q. score is higher than for group B. Hence members of group A must, on average, have higher I.Q. genes than members of group B.
However plausible this sounds initially, the fallacy of this logic becomes immediately obvious with a little thought and the use of a popular analogy. If we randomly take some corn seed and plant it in uniform, rich, well-tended soil, we will get a distribution of plant heights whose variation is caused by their genes. If we take a sample of seed from the same source and plant in poor soil, we will again get a variation of heights that is caused by the genes. But the second group will have a lower average height than the first, even though the plants come from the same gene pool. This difference in average values is caused by the environment and not the genes, a fact known to every farmer.
So it is possible to have a variation that is purely genetic in some trait within a single group, while the difference in average values of the same trait between different groups is caused purely by environment. For example, the variation in individual heights has a substantial genetic component. But in Japan, which has been a relatively isolated country, average heights have risen considerably since World War II, a fact easily explainable by better nutrition.
Whenever one deals with experimental data that seem plausible, one must be especially careful that bias does not enter into obtaining the data and interpreting meaning. And it is here that research on I.Q. and intelligence has performed most shamefully. The historical record of the field is riddled with examples of influential people who have been biased in their collection of data, in their interpretation of results, or in both.
The bias has ranged from the unconscious to outright fraud. However, what has turned out to be remarkably consistent is that the bias has occurred in such a manner that the results have “proved” that whichever group was dominant in society at the time was dominant because of high intelligence. This was so even when the segments that composed the dominant and dominated populations underwent changes. In the U.S. in the 1920s, there were detailed studies that “showed” that Jews and Slavic groups (who happened to be recent immigrants) were of lower intelligence than the Nordic groups who were already established here. No such claims are made today about these two groups. All these formerly “inferior” groups are encompassed within the broad category of “white,” whose members are held to be superior to those who are “black.” In English society, where race was less an issue than in the U.S., class played a crucial role, and, sure enough, research in England “showed” that the upper classes were of higher intelligence than the poor.
It must be extremely comforting to those in positions of wealth, power, and privilege to feel that they got where they are by virtue of innate ability and not because of privileged family connections, wealth, race, education, and all the other factors so crucial to success. It should not be surprising then that these “results” from times past were welcomed so enthusiastically and uncritically by those whose positions they validated. Much the same is still true today. And the moral that can be drawn is that we should always examine most carefully those results that confirm the status quo, because it is then that our critical faculties are most likely to be dulled
This brings us to a consideration of The Bell Curve. This book does not contain any new data. Instead, it uses statistical analyses on old data to draw inferences about the relationship of race and socioeconomic status to intelligence. And the inferences the authors draw are breathtaking in their simplicity and sweeping in their scope. The authors are firmly in Spearman’s camp (despite Spearman’s own defection) in their belief in a single measure of general intelligence, and, when they mention them at all, they give short shrift to opposing views. Herrnstein and Murray seem to find that intelligence is equivalent to I.Q. score and that I.Q. is substantially genetic, heritable, and unchangeable. There’s more. They also seem to believe that low I.Q. causes such social problems as welfare dependency, crime, divorce, illegitimacy, poor parenting, and unemployment.
Having found that I.Q. is the key to understanding all of society’s problems, Herrnstein and Murray contend that, when comparing the conditions of different groups in the population, it only makes sense to compare groups with the same I.Q. scores. What do they find? They find that the chief victims in our society, those suffering the most injustice, are in the high-I.Q. group – the ones commonly referred to as gifted in intelligence. These are the very same people who, the authors had shown earlier, are congregated at the top of the socioeconomic ladder. The problem is that, even though this group is doing very well by any reasonable yardstick, they are not doing as well as they should be, given their intelligence.
Blacks, by contrast, are overrepresented in white-collar and professional occupations and in higher education.
Herrnstein and Murray ponder the consequences of their results, and they do not shrink from the implications for policy that naturally follow. If success in life is so determined by I.Q. and if I.Q. is largely immutable, then it is foolish to waste precious resources in a well-meaning but misguided effort to lift the poor out of their state by giving them assistance based on the assumption that they are as able as everyone else. It would be far better to accept the notion that the lot of these people is determined by natural factors and that we should simply make sure they are kept out of trouble and misery. We should focus our scarce resources instead on making sure that the “cognitively gifted” get all that they need to enable them to reach their full potential, because it is from this group that society can expect to reap real benefits. The authors also recommend abolishing affirmative action as we know it and discouraging low-I.Q. women from having children.
Herrnstein and Murray are not totally heartless about the disadvantaged. They do recommend taking children away from single mothers with low I.Q.s and having them adopted into “better” environments. They also recommend vitamin supplements for children, because there is some suggestion that this measure could raise cognitive ability. In addition, they recommend that all people who work full time should be entitled to a decent standard of living. Although they do not specify what they mean by “decent,” it is interesting that the mechanism they prefer for achieving this end is not raising the minimum wage (which they feel would be a bad move because it increases the cost of doing business for employers), but expanding the earned-income tax credit (which is essentially a federal subsidy to employers that indirectly pays part of the wages of their low-income employees out of federal revenues). Without any sense of irony or déjà vu, they go on to recommend limiting immigration so that low-I.Q. immigrants are kept out, thus mirroring the very arguments that were used to shut out Jews and Slavic groups in the 1920s.
But do the authors really believe all their conclusions about intelligence and I.Q.? Can they really be so unaware of the logical and methodological problems that bedevil any attempt to relate race to intelligence? Surely they could not have been completely unaware, because any reasonably well-informed layperson knows something about them. Actually, it is hard to know exactly what beliefs the authors held. One finds caveats scattered throughout their text that show that they were well aware of most of the problems and lacked adequate arguments to counter many of them. Any single one of these caveats, if taken seriously, would undermine their whole thesis, to say nothing of negating the value of any policy recommendations that follow from them. But the authors, having made the caveats in a pro forma way, blithely proceed to ignore them and to charge ahead. To put it mildly, this is a highly unusual approach for any book that purports to be scientific.
So why was this book written in the way it was? It does not offer any new data. It does not have any new theoretical constructs. It does not have any convincing new arguments to lay to rest the problems raised about the old data and constructs. Yet its publication caused a major stir, and the reasons for that must be examined. It is always dangerous to speculate on the motives of others, but the curiously convoluted nature of this book cries out for explanations, even if we can only guess at them.
We have passed the stage of history when people can simply assert unchallenged that the poor or blacks or women are of lower intelligence than the rest of the population. Anyone who said that would be written off as a crank or a racist or a misogynist. But it is possible to state that “science” says that the poor or blacks or women are of lower intelligence than the rest of the population. So powerful is the mystique of science and so widespread the ignorance of the limits of scientific inference that such a statement can be widely publicized and accepted – even if few understand the basis for it.
I believe that this book was written with two audiences in mind. To achieve any credibility at all, it had to be perceived as a work of science and had to engage the attention of and be debated by the academic community. In order to do that it had to be written in a technical way, with charts, graphs, and all the trappings popularly perceived as constituting science. It also had to contain all the caveats I alluded to above, so that the authors can point to them in response to challenges to their scientific rigor.
But the other – perhaps the main – audience consists of the nonscientific community, especially those journalists and politicians who can be expected to seize upon the statements that support their attempts to further marginalize the poor and minorities. These people can now claim that “science” justifies the gutting of all programs that are based on the belief that what separates the poor from the rest of us is that they were not fortunate enough to grow up with adequate housing, food, education, and security. Herrnstein and Murray must have realized that this second audience would ignore the reservations buried in the dense text and would seize on the broad, sweeping generalizations that are contained in the one-page chapter summaries that the authors wrote for those not willing to read the entire book.
A curious feature of the book and of subsequent statements by the authors is the close attention paid to Asian I.Q. scores. The authors make much of the fact that the I.Q. scores of Asians are even higher than those of whites, and hence Asians are “smarter.” But the average score they report for Asians is 105, compared to 100 for whites. Given the immense problems that exist with the data and the interpretation and the consequent uncertainties that arise from them, a five-point difference is not particularly striking. It could well lie in the range of “noise.”
So why emphasize this feature so? Could it be that the Asian scores provide a preemptive response to the expected charges of racism? The white authors could claim that they could not be racists, because they concede the superior intelligence of a darker-skinned group. The Asian scores are just high enough to deflect charges of racism but not so high as to strain the credulousness of even the sympathetic reader. (It is amusing to speculate as to what might have been the reaction had the Asian I.Q. scores been higher than white scores by the same margin that whites exceed the blacks.) I am not aware of any response from members of the Asian community to the notion that they are somehow smarter than the rest, but they should be wary of being used as pawns in a struggle that pits black against white and rich against poor.
For the reader who does not share the ideology propounded in the book, another curious (and highly irritating) feature is the authors’ ingratiating efforts to co-opt readers into believing their thesis. They appeal to the reader’s vanity with statements asserting that the fact that the reader has waded so far through the book is a sign that the reader has a superior I.Q. Like many of the assertions in the book, one could quite convincingly argue the opposite.
In the book itself and in interviews and in articles that have appeared since its publication, Murray has taken pains to paint himself and his late co-author as courageous scientists who are taking great personal risks in putting forth unpopular points of view. He seems to view himself and Herrnstein as crusaders for unpalatable truths, valiantly battling the overwhelming forces of political correctness that have for so long successfully suppressed the truth out of a misplaced sense of egalitarianism.
This is patently absurd. When all the dust settles, the authors of The Bell Curve are merely saying that the people and groups who dominate our society do so because of their intrinsic ability and merit, that this is the way things were meant to be, and that they and those like them should be benefited even more. When has it ever been an act of courage to assert that those in power have a natural right to that power? True, the authors were subjected to some scorn in the scientific community because of the flaws in their work. But facing those attacks does not imply courage. It takes courage to defend the powerless against the powerful; it merely takes gall to claim courage for defending the privileges of the powerful.
It is safe to say that, despite decades of effort by very determined people, we are not much closer to a definitive answer to the question of the roles of race and intelligence in the processes of social and economic stratification of society. All kinds of hypotheses can be invoked to explain the data. And this shouldn’t be too surprising. As I emphasized above, both race and intelligence are poorly defined and operationally ambiguous. When you have two variables that are ill-defined, it is asking too much to expect a simple relationship between them to emerge.
So what are we to make of the periodic bursts of interest in this topic in the absence of any real advances in research? Each of these epochs was contemporaneous with a period of social stress and increasing racial tension. The fuss over The Bell Curve says less about the merits of the book than about the times we live in. Unfortunately, the only thing we can be sure about is that the same ground will be covered the next time issues of race and immigration come to the boiling point. If we wish to disentangle the real science from the wishful thinking of those who seek to promote a particular ideology or social policy, it is essential that we understand clearly the rules (and limits) of scientific logic and inference and how to apply them.
My last paragraph, written in 1995, seems prescient about the resurgence of this discussion now but it is such an obvious prediction that I do not deserve any credit for it.