The Bell Curve Redux

I decided to publish the entire review in one place. On Goodreads it was split up.

One would hope that decisions are made based on solid evidence and a modicum of rational thought. Often that is not the case, however Sometimes rehashed data and superficial analysis, particularly in the area of social policy, appeal to society because they reflect changes in society's perceptions of reality To some extent that explains the popularity of The Bell Curve by Richard Herrnstein and Charles Murray. There seems to be an unconscious desire to locate society's ills in our genes. Perhaps another misplaced wish is to allocate blame on something or someone else. The premise of The Bell Curve is that there are inherent genetic differences in intelligence between groups and races, e.g., whites, on the average, score lower than Asians; blacks, score lower than whites, etc. and that intelligent people are more successful, i.e. make more money. (Surely, mixed races score higher than everybody, so score one for interracial marriage.)

Charles Lane ("The Tainted Sources of The Bell Curve," in The New York Review of Books, December 1, 1994) and Stephen Jay Gould ("Curveball" in The New Yorker, November 28, 1994) have taken the trouble to actually look at the documentation Herrnstein and Murray used to support The Bell Curve, and they have found it wanting.

The Bell Curve does not purport to be a piece of original scholarship, but a review of the literature, so examination of the sources is certainly relevant. One source for the book was a publication entitled The Mankind Quarterly or, more specifically, articles written by contributors to that journal. Unfortunately, that magazine was founded for the sole purpose of selling the idea that whites are genetically superior to other races. Its founder and editor-in-chief was a vocal supporter of apartheid and segregation in the United States. Most reputable anthropologists have denounced the magazine. One of the major sources that Herrnstein and Murray use to show evidence of lower I.Q. scores of African blacks is an I.Q. test that had been declared invalid for non-Americans. (One of the questions, for example, showed a tennis court without a net and the test taker was supposed to sketch in the net to get credit for the answer) Lane also discovered that the source Herrnstein and Murray used to document the higher scores of Asians sampled the children of only wealthy Japanese, compared to a much broader sample of American children. A study done by a prominent social scientist in Minnesota that carefully matched socioeconomic and demographic factors found no difference in I.Q. at all between Japanese, Taiwanese and American children. (It is interesting to note that Herrnstein was the author of a 1971 Atlantic article that promoted paying well-educated mothers for higher birth rates.)

But it remains for that most lucid of commentators, Stephen Jay Gould, to put the whole issue of heritability of I.Q. into perspective; "Take a trait that is far more heritable than anyone has ever claimed I.Q. to be but is politically uncontroversial - body height. Suppose that I measure the heights of adult males in a poor Indian village beset with nutritional deprivation, and suppose the average height of adult males is five feet six inches. Heritability within the village is high, which is to say that tall fathers... tend to have tall sons while short fathers tend to have short sons. But this high heritability within the village does not mean that better nutrition might not raise average height to five feet ten inches in a few generations. Similarly, the well-documented fifteen-point average difference in I.Q. between blacks and whites in America, with substantial heritability of l.Q. in family lines within each group, permits no automatic conclusion that truly equal opportunity might not raise the black average enough to equal or surpass the white mean.

Herrnstein and Murray conveniently ignore documented high I.Q. scores of poor black children adopted into affluent, intellectual white families. They also overlook average I.Q. increases in some nations since the Second World War equal to the entire fifteen-point difference now separating blacks and whites in America. Gould has another gripe; the failure of lay readers to penetrate the authors' scientism. He quotes many reviewers who said in their reviews that they were unable to judge the adequacy of the arguments because of their lack of scientific training. Gould says, "The book is a rhetorical masterpiece of scientism, and it benefits from the particular kind of fear that numbers impose on nonprofessional commentators. It runs to eight hundred and forty-five pages, including more than a hundred pages of appendixes filled with figures. So the text looks complicated, and reviewers shy away with a knee-jerk claim, that while they suspect fallacies of argument, they really cannot judge." Yet the central premise of The Bell Curve rests entirely on two entirely unsupported assumptions; "(1) that there is a single, general measure of mental ability, and (2) that the I.Q. tests that purport to measure this ability... aren't culturally biased." Ironically, Herrnstein and Murray fail to document these assumptions in their book. According to Gould, "they simply declare that it has been decided."

Gould examined their statistical methodology and found it, too, lacking in precision and accuracy. But he finds their solutions completely abhorrent. They actually write in The Bell Curve that those with lower I.Q.s should be placed in a custodial state ... a high-tech and more lavish version of the Indian reservation for some substantial minority of the nation's population, while the rest of America tries to go about its business." Do you suppose they would let them have guns or TV's? Gould quotes John Stuart Mill; "The tendency has always been strong to believe that whatever received a name must be an entity or being, having an existence of its own. And if no real entity answering to the name could be found, men did not for that reason suppose that none existed, but imagined that it was something particularly abstruse and mysterious." And Gould ends his review; "How strange that we would let a single and false number [I.Q.:] divide us, when evolution has united all people in the recency of our common ancestry-thus undergirding with a shared humanity that infinite variety which custom can never state. E pluribus unum."

Interestingly, there is a very revealing piece of data contained in Appendix 5 of The Bell Curve (and yes, I have read the book) and that is the results of ACT, SAT, and GRE scores of whites and blacks between T 970 and 1990. Blacks score on average generally lower than whites, but what is interesting is that the difference has narrowed. "Overall the evidence seems clear beyond a reasonable doubt... the narrowing was achieved because black scores rose more than white scores, not because white scores were falling." That would seem to provide evidence that perhaps some of the social tinkering may have been working, contrary to Murray's thesis in Losing Ground, a book he published some years ago that was an indictment of the welfare system as a failure.

Murray and Herrnstein make some statements in The Bell Curve that made me wonder about their cognitive ability. For example, on page 201 they state; "Going on welfare really is a dumb idea, and that is why women who are low in cognitive ability end up there; but also such women have little to take to the job market, and welfare is one of their few appropriate recourses when they have a baby and no husband to help." So I guess it was pretty smart, huh.

A recent study that bears on the problems raised by Herrnstein and Murray reports that many children suffer permanent intellectual damage before they enter first grade. "Neuroscientists now believe that a child's future intellectual growth is shaped during these years by the kind of stimulation a child gets." The child's brain can only become organized and make associations if stimulated early in life, which makes the role of the parent crucial.

Studies done on kittens where one eye was sutured shut - we'll discuss cruelty in laboratory experiments in another issue - and then returned to a normal sensory world left the kittens now permanently blind.

"In 1991, 53 percent of all women with one-year old babies were in the workforce, up from 17 percent in 1965, and nearly half of the children under three were being looked after by someone other than their parents."

The report ["Starting Points; Meeting the Needs of Our Youngest Children." Carnegie Corporation, 1994:] cites studies that show the "care infants and toddlers get is often of such substandard quality that it adversely affects their development." The most discouraging aspect suggests that there may be little that educators and parents can do after age three. "There may be a permanent gap between youngsters who have had the stimulation necessary to their mental development and those who have not - no matter what schools and teachers do." And' of course, since they would never take I.Q. tests that early, the differences in I.Q. may stem from early deprivation of stimulation rather than an innate cognitive difference, yet the outcome may be depressingly similar.

One major thesis of Murray and Herrnstein's book is that during the last sixty to seventy years there has been a partitioning of society based on education and intelligence. During the 1930's, for example, there was little difference in the I.Q.s of students at various colleges throughout the country. A student at a small church related school in Idaho was likely to have an I.Q. not too far from the average I.Q. of a student at Harvard, whose average score on the SAT even in 1950 was only 528. Since W.W.II there has been an enormous shift. Society is much more efficient now at sending its brightest students on to college and success. Bright students have been going to the more elite schools, and the population in general that used to include a broader range of intellectual levels, is now sending more of the bright students to college, which is a great sorter - 20% of those in the lower 2 deciles entered college but only 2% got degrees, whereas 70% in the top decile got a BA's. - and this has resulted in the development of what they call a "cognitive elite," a group that gets better jobs, has more education, and associates with itself, a partitioning of society. To some extent, the subject of the book is how the social fabric has been changed by the development of this group that is different from the rest of society by virtue of its being selected out of society. "When people live in encapsulated worlds, it becomes difficult for them, even with the best of intentions, to grasp the realities of the worlds with which they have little experience but over which they have great influence, both public and private." Education is the first sorting mechanism, which leads to the second great sorter: occupation.

Murray has argued that their book should not be used to create policy, a disingenuous position at best. I suspect they did not think through the logical outcome of their proposition. Murray has argued elsewhere that the welfare system is a failure and that we need to eliminate programs as they exist, yet their book screams for more welfare, for if indeed there is a group of people inherently unable to care for themselves or achieve on a high level, then society has no other choice but to put them on welfare and act paternalistically toward them, a proposition, T, for one, don't find compelling evidence for.

Some of the more disingenuous quotes from The Bell Curve: "Measures of intelligence have reliable statistical relationships with important social phenomena, but they are a limited tool for deciding what to make of any given individual." [their italics!... This thing we know as I.Q. is important but not a synonym for human excellence." Now let me get this straight: We can use I.Q. for determining social policy that has enormous impact on individuals, but an individual's l.Q. has nothing to do with their performance in society So treat them as a group because of the group's I.Q., even though it may hinder the individual's performance. The book is filled with such non sequiturs.

There is one aspect of The Bell Curve that I found to be quite useful. Appendix I contains one of the most enlightening chapters on statistics I have read. The authors explain clearly and non-technically what the standard deviation is, how linear regression is used, and how statistics are used to measure and interpret data.