espite its largely technical nature, The Bell Curve has already secured a prominent place in American consciousness as a "big," "important," and "controversial" book. In a manner more befitting a chronicle of sex or spying, the publisher withheld it from potential critics until the date of publication. Since then it has grabbed front-page attention in influential publications, ridden the talk-show waves, and catalyzed academic conferences and dinner table controversies. With the untimely death of the senior author, psychologist Richard Herrnstein, attention has focused on his collaborator Charles Murray (described by the New York Times Magazine as "the most dangerous conservative in America"). But this volume clearly bears the mark of both men.
The Bell Curve is a strange work. Some of the analysis and a good deal of the tone are reasonable. Yet, the science in the book was questionable when it was proposed a century ago, and it has now been completely supplanted by the development of the cognitive sciences and neurosciences. The policy recommendations of the book are also exotic, neither following from the analyses nor justified on their own terms. The book relies heavily on innuendo, some of it quite frightening in its implications. The authors wrap themselves in a mantle of courage, while coyly disavowing the extreme conclusions that their own arguments invite. The tremendous attention lavished on the book probably comes less from the science or the policy proposals than from the subliminal messages and attitudes it conveys.
Taken at face value, The Bell Curve proceeds in straightforward fashion. Herrnstein and Murray summarize decades of work in psychometrics and policy studies and report the results of their own extensive analyses of the National Longitudinal Survey of Labor Market Experience of Youth, a survey that began in 1979 and has followed more than 12,000 Americans aged 14-22. They argue that studies of trends in American society have steadfastly ignored a smoking gun: the increasing influence of measured intelligence (IQ). As they see it, individuals have always differed in intelligence, at least partly because of heredity, but these differences have come to matter more because social status now depends more on individual achievement. The consequence of this trend is the bipolarization of the population, with high-IQ types achieving positions of power and prestige, low-IQ types being consigned to the ranks of the impoverished and the impotent. In the authors' view, the combined ranks of the poor, the criminal, the unemployed, the illegitimate (parents and offspring), and the uncivil harbor a preponderance of unintelligent individuals. Herrnstein and Murray are disturbed by these trends, particularly by the apparently increasing number of people who have babies but fail to become productive citizens. The authors foresee the emergence of a brutal society in which "the rich and the smart" (who are increasingly the same folks) band together to isolate and perhaps even reduce the ranks of those who besmirch the social fabric.
Scientifically, this is a curious work. If science is narrowly conceived as simply carrying out correlations and regression equations, the science in The Bell Curve seems, at least on a first reading, unexceptional. (My eyebrows were raised, though, by the authors' decision to introduce a new scoring system after they had completed an entire draft of the manuscript. They do not spell out the reasons for this switch, nor do they indicate whether the results were different using the earlier system.) But science goes far beyond the number-crunching stereotype; scientific inquiry involves the conceptualization of problems, decisions about the kinds of data to secure and analyze, the consideration of alternative explanations, and, above all, the chain of reasoning from assumptions to findings to inferences. In this sense, the science in The Bell Curve is more like special pleading, based on a biased reading of the data, than a carefully balanced assessment of current knowledge.
Moreover, there is never a direct road from research to policy. One could look at the evidence presented by Herrnstein and Murray, as many of a liberal persuasion have done, and recommend targeted policies of intervention to help the dispossessed. Herrnstein and Murray, of course, proceed in quite the opposite direction. They report that efforts to raise intelligence have been unsuccessful and they oppose, on both moral and pragmatic grounds, programs of affirmative action or other ameliorative measures at school or in the workplace. Their ultimate solution, such as it is, is the resurrection of a world they attribute to the Founding Fathers. These wise men acknowledged large differences in human abilities and did not try artificially to bring about equality of results; instead, Herrnstein and Murray tell us, they promoted a society in which each individual had his or her place in a local neighborhood and was accordingly valued as a human being with dignity.
The Bell Curve is well argued and admirably clear in its exposition. The authors are, for the most part, fair and thorough in laying out alternative arguments and interpretations. Presenting views that set a new standard for political incorrectness, they do so in a way that suggests their own overt discomfort--real or professed. Rush Limbaugh and Jesse Helms might like the implications, but they would hardly emulate the hedges and the "more in sorrow" statements. At least some of the authors' observations make sense. For example, their critique of the complex and often contradictory messages embodied in certain governmental social policies is excellent, and their recommendations for simpler rules are appropriate.
Yet I became increasingly disturbed as I read and reread this 800-page work. I gradually realized I was encountering a style of thought previously unknown to me: scholarly brinkmanship. Whether concerning an issue of science, policy, or rhetoric, the authors come dangerously close to embracing the most extreme positions, yet in the end shy away from doing so. Discussing scientific work on intelligence, they never quite say that intelligence is all-important and tied to one's genes; yet they signal that this is their belief and that readers ought to embrace the same conclusions. Discussing policy, they never quite say that affirmative action should be totally abandoned or that childbearing or immigration by those with low IQs should be curbed; yet they signal their sympathy for these options and intimate that readers ought to consider these possibilities. Finally, the rhetoric of the book encourages readers to identify with the IQ elite and to distance themselves from the dispossessed in what amounts to an invitation to class warfare. Scholarly brinkmanship encourages the reader to draw the strongest conclusions, while allowing the authors to disavow this intention.
In a textbook published in 1975, Herrnstein and his colleague Roger Brown argued that the measurement of intelligence has been the greatest achievement of twentieth-century scientific psychology. Psychometricians can make a numerical estimate of a person's intelligence that remains surprisingly stable after the age of five or so, and much convergent evidence suggests that the variations of this measure of intelligence in a population are determined significantly (at least 60 percent) by inheritable factors. As Herrnstein and Murray demonstrate at great length, measured intelligence correlates with success in school, ultimate job status, and the likelihood of becoming a member of the cognitively entitled establishment.
But correlation is not causation, and it is possible that staying in school causes IQ to go up (rather than vice versa) or that both IQ and schooling reflect some third causative factor, such as parental attention, nutrition, social class, or motivation. Indeed, nearly every one of Herrnstein and Murray's reported correlations can be challenged on such grounds. Yet, Herrnstein and Murray make a persuasive case that measured intelligence--or, more technically, "g," the central, general component of measured intelligence--does affect one's ultimate niche in society.
But the links between genetic inheritance and IQ, and then between IQ and social class, are much too weak to draw the inference that genes determine an individual's ultimate status in society. Nearly all of the reported correlations between measured intelligence and societal outcomes explain at most 20 percent of the variance. In other words, over 80 percent (and perhaps over 90 percent) of the factors contributing to socioeconomic status lie beyond measured intelligence. One's ultimate niche in society is overwhelmingly determined by non-IQ factors, ranging from initial social class to luck. And since close to half of one's IQ is due to factors unrelated to heredity, well over 90 percent of one's fate does not lie in one's genes. Inherited IQ is at most a paper airplane, not a smoking gun.
Indeed, even a sizeable portion of the data reported or alluded to in The Bell Curve runs directly counter to the story that the authors apparently wish to tell. They note that IQ has gone up consistently around the world during this century--15 points, as great as the current difference between blacks and whites. Certainly this spurt cannot be explained by genes! They note that when blacks move from rural southern to urban northern areas, their intelligence scores also rise; that black youngsters adopted in households of higher socioeconomic status demonstrate improved performance on aptitude and achievement tests; and that differences between the performances of black and white students have declined on tests ranging from the Scholastic Aptitute Test to the National Assessment of Educational Practice. In an extremely telling phrase, Herrnstein and Murray say that the kind of direct verbal interaction between white middle-class parents and their preschool children "amounts to excellent training for intelligence tests." On that basis, they might very well have argued for expanding Head Start, but instead they question the potential value of any effort to change what they regard as the immutable power of inherited IQ.
The psychometric faith in IQ testing and Herrnstein and Murray's analysis are based on assumptions that emerged a century ago, when Alfred Binet devised the first test of intelligence for children. Since 1900, biology, psychology, and anthropology have enormously advanced our understanding of the mind. But like biologists who ignore DNA or physicists who do not consider quantum mechanical effects, Herrnstein and Murray pay virtually no attention to these insights and, as a result, there is a decidedly anachronistic flavor to their entire discussion.
Intoxication with the IQ test is a professional hazard among psychometricians. I have known many psychometricians who feel that the science of testing will ultimately lay bare all the secrets of the mind. Some believe a difference of even a few points in an IQ or SAT score discloses something important about an individual's or group's intellectual merits. The world of intelligence testers is peculiarly self-contained. Like the chess player who thinks that all games (if not the world itself) are like chess, or the car salesman who speaks only of horsepower, the psychometrician may come to believe that all of importance in the mind can be captured by a small number of items in the Stanford-Binet test or by one's ability to react quickly and accurately to a pattern of lights displayed on a computer screen.
Though Herrnstein deviated sharply in many particulars from his mentor B.F. Skinner, the analysis in The Bell Curve is Skinnerian in a fundamental sense: It is a "black box analysis." Along with most psychometricians, Herrnstein and Murray convey the impression that one's intelligence simply exists as an innate fact of life--unanalyzed and unanalyzable--as if it were hidden in a black box. Inside the box there is a single number, IQ, which determines vast social consequences.
Outside the closed world of psychometricians, however, a more empirically sensitive and scientifically compelling understanding of human intelligence has emerged in the past hundred years. Many authorities have challenged the notion of a single intelligence or even the concept of intelligence altogether. Let me mention just a few examples. (The works by Stephen Ceci and Robert Sternberg, as well as my own, discuss many more.)
Sternberg and his colleagues have studied valued kinds of intellect not measured by IQ tests, such as practical intelligence--the kind of skills and capacities valued in the workplace. They have shown that effective managers are able to pick up various tacit messages at the workplace and that this crucial practical sensitivity is largely unrelated to psychometric intelligence. Ralph Rosnow and his colleagues have developed measures of social or personal intelligence--the capacities to figure out how to operate in complex human situations--and have again demonstrated that these are unrelated to the linguistic and logical skills tapped in IQ tests.
Important new work has been carried out on the role of training in the attainment of expertise. Anders Ericsson and his colleagues have demonstrated that training, not inborn talent, accounts for much of experts' performances; the ultimate achievement of chess players or musicians depends (as your mother told you) on regular practice over many years. Ceci and others have documented the extremely high degree of expertise that can be achieved by randomly chosen individuals; for example, despite low measured intelligence, handicappers at the racetrack successfully employ astonishingly complex multiplicative models. A growing number of researchers have argued that, while IQ tests may provide a reasonable measure of certain linguistic and mathematical forms of thinking, other equally important kinds of intelligence, such as spatial, musical, or personal, are ignored (this is the subject of much of my own work). In short, the closed world of intelligence is being opened up.
Accompanying this rethinking of the concept of intelligence(s), there is growing skepticism that short paper-and-pencil tests can get at important mental capacities. Just as "performance examinations" are coming to replace multiple-choice tests in schools, many scientists, among them Lauren Resnick and Jean Lave, have probed the capacities of individuals to solve problems "on the scene" rather than in a testing room, with pencil and paper. Such studies regularly confirm that one can perform at an expert level in a natural or simulated setting (such as bargaining in a market or simulating the role of a city manager) even with a low IQ, while a high IQ cannot in itself substitute for training, expertise, motivation, and creativity. Rather than the pointless exercise of attempting to raise psychometric IQ (on which Herrnstein and Murray perseverate), this research challenges us to try to promote the actual behavior and skills that we want our future citizens to have. After all, if we found that better athletes happen to have larger shoe sizes, we would hardly try to enlarge the feet of the less athletic.
Scientific understanding of biological and cultural aspects of cognition also grows astonishingly with every passing decade. Virtually no serious natural scientist speaks about genes and environment any longer as if they were opposed. Indeed, every serious investigator accepts the importance of both biological and cultural factors and the need to understand their interactions. Genes regulate all human behavior, but no form of behavior will emerge without the appropriate environmental triggers or supports. Learning alters the way in which genes are expressed.
The development of the individual brain and mind begins in utero, and pivotal alterations in capacity and behavior come about as the result of innumerable events following conception. Hormonal effects in utero, which certainly are environmental, can cause a different profile of cognitive strengths and limitations to emerge. The loss of certain sensory capacities causes the redeployment of brain tissue to new functions; a rich environment engenders the growth of additional cortical connections as well as timely pruning of excess synapses. Compare a child who has a dozen healthy experiences each day in utero and after birth to another child who has a daily diet of a dozen injurious episodes. The cumulative advantage of a healthy prenatal environment and a stimulating postnatal environment is enormous. In the study of IQ, much has been made of studies of identical and fraternal twins. But because of the influences on cognition in utero and during infancy, even such studies cannot decisively distinguish genetic from environmental influences.
Herrnstein and Murray note that measured intelligence is only stable after age five, without drawing the obvious conclusion that the events of the first years of life, not some phlogiston-like "g," are the principal culprit. Scores of important and fascinating new findings emerge in neuroscience every year, but scarcely a word of any of this penetrates the Herrnstein and Murray black-box approach.
Precisely the same kind of story can be told from the cultural perspective. Cultural beliefs and practices affect the child at least from the moment of birth and perhaps sooner. Even the parents' expectations of their unborn child and their reactions to the discovery of the child's sex have an impact. The family, teachers, and other sources of influence in the culture signal what is important to the growing child, and these messages have both short- and long-term impact. How one thinks about oneself, one's prospects in this world and beyond, and whether one regards intelligence as inborn or acquired--all these shape patterns of activity, attention, and personal investments in learning and self-improvement. Particularly for stigmatized minorities, these signals can wreck any potential for cognitive growth and achievement.
Consider Claude Steele's research on the effects of stereotyping on performance. African-American students perform worse than white students when they are led to believe that the test is an intellectual one and that their race matters, but these differences wash out completely when such "stereotype vulnerable" conditions are removed.
To understand the effects of culture, no study is more seminal than Harold Stevenson and James Stigler's book The Learning Gap: Why Our Schools Are Failing and What We Can Learn from Japanese and Chinese Education (1992). In an analysis that runs completely counter to The Bell Curve, Stevenson and Stigler show why Chinese and Japanese students achieve so much more in schools than do Americans. They begin by demonstrating that initial differences in IQ among the three populations are either nonexistent or trivial. But with each passing year, East Asian students raise their edge over Americans, so that by the middle school years, there is virtually no overlap in reading and mathematics performance between the two populations.
Genetics, heredity, and measured intelligence play no role here. East Asian students learn more and score better on just about every kind of measure because they attend school for more days, work harder in school and at home after school, and have better-prepared teachers and more deeply engaged parents who encourage and coach them each day and night. Put succinctly, Americans believe (like Herrnstein and Murray ) that if they do not do well, it is because they lack talent or ability; Asians believe it is because they do not work hard enough. As a Japanese aphorism has it, "Fail with five hours of sleep; pass with four." Both predictions tend to be self-fulfilling. As educator Derek Bok once quipped, Americans score near to last on almost all measures save one: When you ask Americans how they think they are doing, they profess more satisfaction than any other group. Like Herrnstein and Murray, most Americans have not understood that what distinguishes the cultures is the pattern of self-understanding and motivation, especially the demands that we make on ourselves (and on those we care about) and the lessons we draw from success and failure--not the structure of genes or the shape of the brain.
Like Murray's earlier book Losing Ground, The Bell Curve views most recent governmental attempts at intervention as doing more harm than good and questions the value of welfare payments, affirmative action programs, indeed, any kind of charitable disposition toward the poor. To improve education, Herrnstein and Murray recommend vouchers to encourage a private market and put forth the remarkable proposal that the government should shift funds from disadvantaged to gifted children. And while they do not openly endorse policies that will limit breeding among the poor or keep the dispossessed from our shores, they stimulate us to consider such possibilities.
Nowhere did I find the Herrnstein and Murray analysis less convincing than in their treatment of crime. Incarcerated offenders, they point out, have an average IQ of 92, eight points below the national mean. They go on to suggest that since lower cognitive aptitude is associated with higher criminal activity, there would be less crime if IQs were higher. But if intelligence levels have at worst been constant, why did crime increase so much between the 1960s and 1980s? Why have crime rates leveled off and declined in the last few years? Does low IQ also explain the embarrassing prevalence of white-collar crime in business and politics or the recent sudden rise in crime in Russia? Astonishingly, no other influences, such as the values promoted by the mass media, play any role in Herrnstein and Murray's analysis.
Considering how often they remind us that the poor and benighted at society's bottom are incapable through no fault of their own, Herrnstein and Murray's hostility to efforts to reduce poverty might seem, at the very least, ungenerous. But, at the book's end, the authors suddenly turn from their supposed unblinking realism to fanciful nostalgia. Having consigned the dispossessed to a world where they can achieve little because of their own meager intellectual gifts, Herrnstein and Murray call on the society as a whole to reconstitute itself: to become (once again?) a world of neighborhoods where each individual is made to feel important, valued, and dignified. They devote not a word to how this return to lost neighborhoods is to be brought about or how those with low IQs and no resources could suddenly come to feel worthwhile. It is as if we were watching scenes from Apocalypse Now or Natural Born Killers, only to blink for a minute and to find the movie concluding with images from a situation comedy or "Mr. Rogers' Neighborhood."
Perhaps the most troubling aspect of the book is its rhetorical stance. This is one of the most stylistically divisive books that I have ever read. Despite occasional avowals of regret and the few utopian pages at the end, Herrnstein and Murray set up an us/them dichotomy that eventually culminates in an us-against-them opposition.
Who are "we"? Well, we are the people who went to Harvard (as the jacket credits both of the authors) or attended similar colleges and read books like this. We are the smart, the rich, the powerful, the worriers. And who are "they"? They are the pathetic others, those who could not get into good schools and who don't cut it on IQ tests and SATs. While perhaps perfectly nice people, they are simply not going to make it in tomorrow's complex society and will probably end up cordoned off from the rest of us under the tutelage of a vicious custodial state. The hope for a civil society depends on a miraculous return of the spirit of the Founding Fathers to recreate the villages of Thomas Jefferson or George Bailey (as played by Jimmy Stewart) or Beaver Cleaver (as played by Jerry Mather).
How is this rhetorical polarization achieved? At literally dozens of points in the book, Herrnstein and Murray seek to stress the extent to which they and the readers resemble one another and differ from those unfortunate souls who cause our society's problems. Reviewing the bell curve of the title, Herrnstein and Murray declare, in a representative passage:
You--meaning the self-selected person who has read this far into this book--live in a world that probably looks nothing like the figure. In all likelihood, almost all of your friends and professional associates belong to that top Class l slice. Your friends and associates who you consider to be unusually slow are probably somewhere in Class II.
Why is this so singularly off-putting? I would have thought it unnecessary to say, but if people as psychometrically smart as Messrs. Herrnstein and Murray did not "get it," it is safer to be explicit. High IQ doesn't make a person one whit better than anybody else. And if we are to have any chance of a civil and humane society, we had better avoid the smug self-satisfaction of an elite that reeks of arrogance and condescension.
Though there are seven appendices, spanning over 100 pages, and nearly 200 pages of footnotes, bibliography, and index, one element is notably missing from this tome: a report on any program of social intervention that works. For example, Herrnstein and Murray never mention Lisbeth Schorr's Within Our Reach: Breaking the Cycle of Disadvantage, a book that was prompted in part by Losing Ground. Schorr chronicles a number of social programs that have made a genuine difference in education, child health service, family planning, and other lightning-rod areas of our society. And to the ranks of the programs chronicled in Schorr's book, many new names can now be added. Those who have launched Interfaith Educational Agencies, City Year, Teach for America, Jobs for the Future, and hundreds of other service agencies have not succumbed to the sense of futility and abandonment of the poor that the Herrnstein and Murray book promotes.
When I recently debated Murray on National Public Radio, he was reluctant to accept the possibility that programs of intervention might dissolve or significantly reduce differences in intelligence. If he did, the entire psychometric edifice that he and Herrnstein have constructed would collapse. While claiming to confront facts that others refuse to see, they are blind to both contradictory evidence and the human consequences of their work. Herrnstein and Murray, of course, have the right to their conclusions. But if they truly believe that blacks will not be deeply hurt by the hints that they are genetically inferior, they are even more benighted--dare I say, even more stupid--than I have suggested.
It is callous to write a work that casts earlier attempts to help the disadvantaged in the least-favorable light, strongly suggests that nothing positive can be done in the present climate, contributes to an us-against-them mentality, and then posits a miraculous cure. High intelligence and high creativity are desirable. But unless they are linked to some kind of a moral compass, their possessors might best be consigned to an island of glass-bead game players, with no access to the mainland.