For Whom the Bell Curves

Posted by WiredSisters on March 19th, 2014 filed in Uncategorized

Intelligence, Merit, and Rank

Some years back, I taught a college course in Humanities, based on Plato’s Republic, The Bell Curve, and Gould’s Mismeasure of Man. It was a lot of fun. But reading The Bell Curve was not part of the fun. I tried to piece out its argument for my students early on, and came up with something like this:

1) There exists such a thing as “g” (or “general intelligence”), which is heritable in considerable part
2) g can be accurately measured by several tests, including some of the ones to which children and adults in this country are most commonly exposed
3) To a considerable degree, g is immutable
4) Which is to say that Head Start and similar programs designed to increase g are a waste of money
5) g correlates, not only with academic and occupational success, but with numerous character traits essential to the maintenance of a civil society
6) People with low cognitive ability are out-breeding people with high cognitive ability in the US
7) Both the “cognitive elite” and the “cognitive underclass” are isolating themselves from the rest of society

 Conclusion: What society should do in response to these phenomena is to put the underclass on a relatively comfortable reservation, give them useful non-taxing work to do, and subject them to the discipline of a simplified legal system

 I think I’ve got all of that right. But I’m not altogether sure, because my cognitive ability may have been seriously eroded by reading the book. It can’t be good for the reasoning faculties to spend hours absorbing a book that repeatedly says things like “The evidence shows intelligence is most likely 40 – 60% heritable,” and then draws political conclusions based on a presumed 100% heritability. Again and again, the authors lay out the scientific data, explain it, and then ignore it. Again and again, they tell us that statistical data about the behavior and characteristics of populations tells us nothing about individuals–and then they draw political conclusions based entirely on the presumption of individual incompetence and depravity. They caution us repeatedly that the available data are partial and unclear–and then draw political conclusions that would be only barely acceptable based on 100% certainty. And they do all this in a writing style which (speaking of cognitive ability) would earn them (and their editors and proofreaders, if any) no better than a C+ from a competent English composition teacher. For instance, on p. 145, the authors state “…the first decades of the [twentieth] century saw American high school education mushroom in size without having to dip much deeper into the intellectual pool….”, and, on p. 157 “…the long-term employment trend of [young men’s] employment has been downhill….”

 Arguably, this is really two books–a scientific treatise which (some scientists tell us) has serious faults in its collection and treatment of data, and a political treatise which purports to be based on the science, but in fact bears only the most tenuous relationship to it.

 Aside from its flimsy scientific basis, the political treatise has serious logical problems of its own, mainly resulting from the authors’ unwillingness to follow their own argument to its ultimate conclusions. For instance: they admit that, whatever g is, East Asians have more of it than white Americans. Granting the premise that g correlates with various socially necessary character traits, would a Japanese reader not be justified in attributing various Western character traits of which the Japanese have always disapproved–our individualism, sloppiness, imprecision, poor manners, and tendency to violence–to our deficiency in g, and proposing to put us on a reservation where we can stay out of the way of the people most qualified to do the world’s brainwork, the East Asians? Obviously, despite their repeated references to East Asian cognitive superiority, the authors don’t really believe in it, or they would have raised such questions. On the contrary, the whole book is suffused with a “we happy few” tone of self-congratulation, extending from authors to readers, and presuming that both, being most likely white American, are and deserve to be, the “cognitive elite.” The East Asians are merely a statistical blip located in an interesting but minimally consequential place.

 Secondly, if g is fundamentally genetic, why do the authors not confront the problem of cognitive sexual differentiation within ethnic groups? That is to say, the genetic endowment of the average woman of any given ethnic group is in most part identical to that of the average man in the same group. Black men are no blacker than black women. It would follow from that that the spread of measured g should be identical in men and women of the same ethnic group. In fact, it is not. The range is the same, but women cluster much more strongly in the center of that range. Genetics could not possibly account for that. The only genetic differences between men and women lie in the material on the sex chromosomes. We don’t, of course, know exactly what genes are to be found on those chromosomes, nor can we fully explain sexual differentiation in areas other than g. But if intelligence were mostly genetic, it would have to be sex-linked to account for the gender differences in g within ethnic groups. And those differences would then be much greater than in fact they are. If intelligence were sex-linked (assuming as usual that the male is the norm) women would barely be able to drool and breathe at the same time. Obviously this is not the case. The only hard data suggesting a major gender difference in intelligence is the well-established fact that women, generally speaking, marry men, whereas men, generally speaking, marry women. This can easily be accounted for by environmental factors.

 But seriously, folks–looking at the book from a purely textual/editorial viewpoint, the fact that it is the acknowledged work of two authors is significant, particularly since one of the authors is now deceased. It is tempting to hypothesize that Herrnstein (may he rest in peace) did all the science and Murray did all the politics, and that the latter never seriously read the former’s work. It would certainly explain the non-sequiturs with which the book is littered.

 But what concerns me most about the book is not the exiguous link between the scientific and political material, but the even scantier connection between the premises of the political argument and the concluding proposal supposedly drawn from them. If I seriously believed that the increasing isolation of the elite and the underclass from each other and the rest of society were a problem, I would not propose to remedy it by putting the underclass on a reservation. And Murray’s proposal that members of the underclass should be given some “valued place” in such a reservation means nothing at all without some concrete link to the real job market, which Murray never draws. This general shoddiness in the conclusory section suggests strongly that these are not Murray’s real conclusions at all, but only the ones he feels pressured to write to avoid being completely discredited by the sinister forces of political correctness. If Murray had the courage of his convictions, he would be advocating forced sterilization and genocide, which are in fact the only social policies that follow logically from his premises. Murray may intend his overt proposals to be taken seriously long enough to be tried and proved useless (the way General Motors first tried installing cumbersome and unworkable seatbelts after being mandated to by law), or he may have meant them as an in-group joke shared with the other members of the cognitive elite. But anyone as smart as Murray thinks he and his readers are should have no trouble figuring it out.

 Finally, Murray is actually missing a much more factually substantial bet, if he is really serious about wanting to find a variable that correlates with occupational and academic success, high socio-economic status, and most of the pro-social character traits he talks about. There actually is one, and we can measure it with absolute validity, with a lot less controversy and at negligible cost–namely physical height. Which is heritable up to a point, but can be strongly influenced by well-known environmental factors. Of course, using it for the purposes for which Murray proposes to use intelligence testing would eliminate a lucrative service industry, whose lobbyists may have influenced Murray to leave them alone.

 But, assuming it is possible to get past the literary and editorial drawbacks of the book, what happens when we look more closely at the thinking behind it? Murray and Herrnstein talk not only about “g” but about a closely related phenomenon, merit. They, and most of us, have taken for granted, without seriously examining it, the notion that there is some real quality called “merit”, which we can and should usably define, accurately measure, and appropriately reward. Only by doing so, we believe, can we encourage the achievements necessary to the survival of our society.

 This principle goes back at least as far as Plato’s Republic. Plato was willing to grant that merit might not be directly inherited and might even turn up among the children of the less meritorious. Nonetheless, once found out, the meritocrats should be encouraged to reproduce, preferably in conjunction with each other. Which suggests that Plato had already figured out most of what we know today about merit: that the children, and the parents, of meritorious people are more likely to possess similar merit, than those whose parents and children lack it. We still don’t know why. Which means we still don’t know how to increase the proportion of meritorious people in the population, should we choose to do so.

 Let’s begin with definition–is merit equivalent to intelligence? Neither term is especially precise, but merit generally includes attributes of character as well as ability. On the other hand, many authorities, including the authors of The Bell Curve, presume that intelligence usually correlates with positive character traits such as sexual morality and law-abidingness. They also presume that “intelligence”, whatever it is, is something our society needs more of. If they defined intelligence only in terms of intellectual competence, the ability to perform certain cognitive tasks, then they might pay more attention to making better use of the intelligent people we know we already have. But they generally presume that society would be better off if everyone were “intelligent”–which suggests again that they are really concerned with the characterological dimensions of “intelligence.” So, for the moment, let’s stick with “merit”, and presume that it includes both cognitive and characterological dimensions, which are visibly related to each other in some as-yet-unclear way.

 To the extent that we have elevated the testing of human abilities to a pseudo-science, we are able to link certain test behavior in certain populations with certain other behavior patterns, including academic success and, in some cases, occupational competence. In larger populations, the same kinds of test performance may correlate with better socio-economic status and more conventional social behavior. As mentioned earlier, so does adult physical height, especially in males. (By the way, test performance in females is less closely linked to almost all other outcomes than it is in males.) If what we are looking for is a predictor of socio-economic success and lawful behavior, could we not save a lot of money and time (and eliminate a major service industry) by simply measuring and rewarding height, and doing everything we can to increase the height of the next generation? Since height correlates with good childhood nutrition and especially with high prenatal and childhood protein intake, we actually know how to achieve that end. Which gives us the opportunity to think about whether we want to. As Gilbert and Sullivan’s Duke of Barataria pointed out, when everybody is somebody, nobody is anybody. A society of people equally endowed with what we now define as merit would have no way to decide who takes out the garbage and who directs the fate of major corporations. In all likelihood, it would waste no time coming up with some other criterion–amount of melanin in the skin, for instance.

 Face it, what Murray and Herrnstein actually want is not a society in which everyone is more or less equally intelligent, but one in which there are slightly fewer unintelligent people and slightly more intelligent people than we now have, and in which that ratio remains constant over the generations. They are aghast at what they seem to consider a recent development–that the underclass is out-breeding the intelligentsia. So was Caesar Augustus. So were the eugenicists and their predecessors beginning in the 17th century. All of them are dealing with a phenomenon only secondarily related to intelligence: people who have enough access to resources to have some control over their own lives are likely to exercise that control, inter alia, in the area of family life, and specifically how many children they will rear. If they perceive children, or more than a few children, as a liability for any reason, they will therefore have fewer of them. While the people with less access to resources, and less control over their own lives, will have all the children they can, and raise all of those that survive. The vagaries of history and geography may make poor people and peasants more fertile (this also correlates with diet, to some extent) and infant mortality lower in some places than in others, while making the birth and rearing of upper-class children more or less expensive. An additional factor in some cultures has been the amount of personal involvement in child-rearing required of the upper-class mother. In most pre-20th-century cultures, upper-class women could and usually did have their children reared by lower-class women, which considerably decreased the disincentives of large families for them. But once lower-class women found ready employment in factory and service work, they were less likely to be willing to spend their lives raising other women’s children. Upper-class women are now required to invest a lot more time and energy rearing their own children than their grandmothers ever did. Which makes them far less inclined to do so more than once or twice. Apparently, Murray and Herrnstein, without ever talking about it explicitly, have already accepted that fact (unlike the earlier eugenicists, who wasted a good deal of energy trying to persuade upper-class women to have more children.) Murray and Herrnstein are concentrating on the other side of the equation exclusively: there will never be any more of Us, so we must do something to reduce the number of Them.

 On the other hand, how many fewer of Them do We really want? I suspect strongly that what Murray and Herrnstein really want isn’t a world full of Einsteins, but one in which it is possible to sign and send out one’s secretary’s letter without having to proofread it first, or in which one can hop in a cab and give the driver the address of one’s destination without having to instruct him on how to get there. And that, it seems to me, is a function not of the general level of intelligence in a society, but of where the market directs that intelligence. Just about all the really competent legal secretaries I have ever known were born before 1935. Younger women with the same interests and aptitudes went to law school instead–not because they were “too smart” to do secretarial work, but because they were smart enough to refuse to work for a secretary’s salary, and lucky enough to live in an era when they had other choices. Similarly, the omniscient and omnicompetent cabbie who could find any address in New York and deliver babies, was also either a full-time Yellow Cab employee with benefits, or an entrepreneur with a medallion of his own. Today, most cabbies are lessees who have to work the first 8 hours of their shift just to cover lease fees and insurance, and then make their own living in the next 5 hours or so. Anybody with the intelligence, the knowledge of English, and the citizenship or immigrant status to do anything else is doing it.

 In short, before we complain about the lack of intelligent people in our society, we should pay closer attention to what the intelligent people we do have are doing. In far greater proportions than in other industrialized countries, they are practicing law; if they are doing scientific research or engineering, they are very likely to be doing it for the military. On the other hand, they may even be high school dropouts if they are African-American or Hispanic or live in poor countries/ counties/ neighborhoods (yes, Virginia, smart people drop out too. They also commit crimes, engender children out of wedlock, and in general engage in antisocial behavior.) If they are female and their parents never went to college, they may be cashiers or waitresses. In general, if their parents never went to college, smart people may be virtually indistinguishable from the people they grew up with, except perhaps that they have an unusual grasp of sports statistics, or the Civil War, or model boat-building or some other quirky autodidactic fascination.

 Murray and Herrnstein presume that current social realities make it possible for more of the highly intelligent to achieve the rank they “deserve”–but accepting at face value their caveat that data about populations tells us nothing about individuals, we have to presume that there is still a lot of cream on the bottom, where it is either totally wasted or, worse still, put to antisocial uses. And if we really want a greater role for intelligence in our society, shaking up the bottle is still a faster and less expensive way to do it than either the solutions The Bell Curve overtly proposes, or the ones the authors probably really had in mind.

 Wired Sisters




2 Responses to “For Whom the Bell Curves”

  1. Sappho Says:

    Interesting combination of reading for that class. I like the idea of connecting Plato’s Republic with more modern debate over similar ideas.

    Re: “But if intelligence were mostly genetic, it would have to be sex-linked to account for the gender differences in g within ethnic groups.” The way for this to work would have to be that a lot of the IQ genes would be on the X chromosome. Since women get two of these, their high and low IQ genes moderate each other (although, maybe not so much, what with assortative mating by intelligence), while men get only one copy, and thus less moderated genes.

    But this has weird results, if you combine this theory with one in which the same genes are supposed to vary significantly by race. In that case, men, but only men, should be smarter if they have a white mother and black father than if they have a black mother and white father (because all their X DNA comes from their mother, but women’s X DNA comes from both sides). Meanwhile, women, but only women, should get the biggest IQ boost from a white paternal grandmother (since our father’s get X DNA only from their mothers, we get our largest share of X DNA from our paternal grandmothers). I am, for the sake of this example, assuming only white and black grandparents, and ignoring the Asians who should perhaps, if I followed the argument fully, be the best candidates for our philosopher kings. Anyway, I’m not aware that IQ tracks race in quite this way. So, whatever the genetic component, it doesn’t appear that it involves a sex-linked set of genes that also vary by race.

    Murray’s and Herrnstein’s argument that we should make laws so simple that even dummies could follow them was in some ways the most appealing of their policy prescriptions; laws do work better if they’re well understood and reliably enforced, and other things equal we’re better off with simple laws than difficult ones. But then I thought of which laws Murray probably most wanted cleared away, and concluded that he probably had in mind the regulation of corporations (affirmative action, workplace safety, environmental regulations), and concluded that, after all, corporations are generally capable of arranging to employ at least some people intelligent enough to understand the laws that apply to them, and that weaker labor regulations probably wouldn’t improve the lives of people with low IQ any more than they would improve the lives of people with high IQ.

  2. Wired Sisters Says:

    Thanks, I really was kind of winging it on the gender-and-genetics and I appreciate the clarification. Really, I just couldn’t resist the cheap shots.