Flynn, Ceci, and Turkheimer on Race and Intelligence: Opening Moves

William Saletan may be the first journalist to so directly acknowledge the scientific evidence suggesting a non-trivial genetic basis for racial differences in IQ (hereafter, the partly-genetic hypothesis) and to be allowed to publish his views. Publication of his 3-part article in Slate (Nov. 18-20, 2007) represents a stunning departure from usual journalist practice, which is to enforce the taboo against voicing such thoughts. But Saletan does something more important. He urges us to look the dreaded possibility straight in the eye to see whether it is really as fearsome as we imagine. He thinks not, and I agree.

Flynn, Ceci, and Turkheimer comment on both issues raised by Saletan: How strong is the scientific evidence favoring a partly-genetic hypothesis? And is it wise to speak openly and honestly about it? All three reject Saletan’s judgment favoring the genetic hypothesis, though with varying degrees of finality. On the second question, Turkheimer opposes “cool-headed evaluation of the empirical evidence” because the hypothesis is, in his view, “offensive” and “dangerous.” Flynn and Ceci side with Saletan in allowing such evaluation, but Flynn echoes Turkheimer’s moral distaste when he labels as “offensive” the IQ-based scholarship with which all three disagree.

The series of posts by these three authors illustrates, in microcosm, the melange of criticism commonly marshaled against mainstream science on intelligence in order to seem to discredit it without actually engaging its large interlocking body of evidence. Indeed, the criticism succeeds precisely by avoiding such engagement. There are two general strategies for avoiding the totality of relevant evidence: (1) create doubt about some small portion of it as if that isolated doubt nullified the totality of evidence, and (2) put unwelcome evidence off-limits by labeling it immoral or ill-motivated.

What Is the Debate About — Phenotypic or Genotypic Differences in Intelligence?

IQ tests measure only phenotypic (developed) differences in general intelligence (g), not what causes them. Ability differences among groups can be real without necessarily being genetic, either in whole or part. When scientists speak of race or sex differences, this is short-hand for average differences between the groups—just as when we refer to men being taller than women or Americans being fatter today than decades back. There is no implication that the observed differences are caused by genetic rather than non-genetic differences between the groups.

Public commentary on black-white IQ differences generally confuses these two questions: whether blacks and whites differ in average developed intelligence level, and, if so, why. The first question has been scientifically settled for decades. As a task force of the American Psychological Association wrote in its flagship journal in 1996, “The differential between the mean intelligence test scores of [Western] Blacks and Whites … does not result from any obvious biases in test construction and administration” (Neisser et al., 1996). The gap in abilities is real, and it has real-world, practical consequences for the individuals and societies involved (Gottfredson, 1997). Racial-ethnic differences in phenotypic intelligence are the rule, not the exception, worldwide.

In contrast, the “why” question—are these differences partly genetic?—remains a matter of much scientific debate. For instance, despite agreeing that the American black-white IQ gap reflects a gap in general learning and reasoning abilities, intelligence researchers differ on the extent to which the average difference is genetic in origin. Some have written that current evidence makes the black-white gap more plausibly 80% genetic than 0% genetic and others that 50% genetic is more plausible than 0% genetic, and yet others have argued that the available evidence is equally consistent with 0% genetic or is not yet sufficient to venture an opinion. I suspect that most experts on the topic now believe that the gap is at least somewhat genetic, because that was the plurality judgment when solicited confidentially twenty years ago (Snyderman & Rothman, 1988). My view is that the current weight of evidence favors a 50-80 percent rather than a zero percent genetic component (Gottfredson, 2005b).

Gambits for Seeming to Discredit the Genetic Hypothesis Without Actually Doing So

The question Saletan raises is thus whether the part-genetic hypothesis explains average racial differences in phenotypic intelligence better than does a no-genetic hypothesis. Flynn, Turkheimer, and Ceci seem to favor the latter. Their arguments for it, to the extent they provide any, rest on creating doubt about the evidence for g and its relation to race.

The part-genetic hypothesis becomes plausible only if intelligence differences among members of the same race are demonstrated to be real, important, and at least partly genetic. It gains further plausibility when the difference in average intelligence between two races is also real, important, and stubborn. The longer and stronger this chain of evidence for both blacks and whites, the more scientifically plausible a genetic component to their average IQ difference becomes.

Each link in the chain of evidence has been vigorously tested many times because all were once controversial hypotheses among the scientists testing them. Often contrary to the researchers’ initial expectations, each has proven robust. That is why many critics of the genetic hypothesis will concede one or more of the links, at least implicitly, but then assert that crucial other links have been broken.

Table 1 lists a series of seven “yes-but” gambits for such purposes. The first five deny the validity of basic facts about intelligence: the existence of intelligence (g), its fair measurement, practical importance, stability (lack of malleability), and high heritability. None of these facts relates to race, per se. The last two gambits concede all prior five facts but assert that racial differences in IQ are not genetic or must not be thought so. The first, Non-Existence gambit is to concede racial differences in scores on IQ tests but then to assert that IQ tests do not measure intelligence because, for instance, “intelligence” is only that which a particular culture chooses to value. In contrast, the Mismeasured (Test-Bias) rejoinder concedes what the Non-Existence rejoinder disputes (“Yes, IQ tests do measure intelligence”) but then targets a different link in the chain of plausibility (“but, they produce falsely low readings for blacks”). Both rejoinders, however, urge us to abolish IQ tests or question testers’ motives rather than take the IQ gaps seriously.

Table 1. “Yes But” Rejoinders Used to Ignore Scientific Findings of Intelligence and Race

(Click image for full table)

Flynn, Turkheimer, and Ceci make mutually inconsistent claims over which well-established facts we should doubt. Turkheimer deploys primarily the Nonexistence and Unthinkable gambits. Ceci rejects both and relies primarily on Malleability. Flynn also deploys Malleability, but has it both ways on heritability (secular changes but not individual differences are Just Environmental) and the Nonexistence of g (he “shatters” it across generations but not within them), hints at Unimportance, and explicitly rejects Unthinkable while providing reasons to accept it.

The “but” segment of the “yes-but” rejoinders in Table 1 have all been plausible scientific hypotheses at one time or another—for instance, that IQ tests are biased against non-whites or that intelligence is highly malleable. That is why there has been so much research on them in the last half century. A once plausible “but” becomes a misleading gambit, however, when it is falsely asserted to be proved or to enjoy a scientific consensus, especially once the weight of high-quality evidence has clearly tipped in the opposite direction. The three commentators sometimes accomplish this by simply asserting counterfactual claims as “clearly” true.

As the center of scientific inquiry has moved along the chain of plausibility, from newly-confirmed links to still-contested ones, those who would quash the part-genetic hypothesis have shifted their “yes-but” rejoinders accordingly. Where once they could seem to discredit the reports of racial differences in intelligence simply by asserting that intelligence (or fair measurement or genetic influence within race) does not exist, they now frequently assert that the existence of biological races has been disproved. If race is only a social construct, without any biological component, then any genetic component to average racial differences in intelligence is ruled out. Readers are thus invited to ignore or disparage all evidence in the entire chain of plausibility. Our commentators do not comment on the biological reality of race, perhaps because they all deny one or more prior links in the chain of plausibility.

Each “yes-but” gambit contradicts prior ones in the list but enlarges the menu of options for impugning the science. Their mutual inconsistency is muted by the “yes” concessions usually being left implicit. While a “but” criticism can draw public accolades, a “yes” to bedrock facts can draw fire. So, while our commentators sometimes disagree sharply about which links are weak, they join in pronouncing the chain broken.

Those making “but” statements generally justify doing so by providing an evidentiary “because.” The last column of Table 1 exemplifies such appeals to seeming evidence. Many are themselves once-plausible hypotheses, now disproved; others are straw men or emotional appeals designed to arouse fear and disgust. What catches public attention, however, is the great variety of ways in which the science seems to be rendered deficient, or worse.

The trump card is the Unthinkable gambit. It gains popularity when the empirical evidence starts to seem incontestable. This gambit moves the partly-genetic hypothesis out of the scientific realm and into the moral arena. There, we need not consider whether the hypothesis is true or false (that is, supported or not by the weight of evidence) but whether accepting it makes one good or evil (or, at least, feel good rather than guilty). Scientific truth is no defense in the moral realm and, indeed, an unwelcome idea may be attacked all the more fiercely for possibly having truth on its side. By this code, moral duty requires suppressing and censuring evil speech and mandating good speech, whether true or false. Turkheimer plays this card openly.

Critics sometimes apply moral standards in the guise of scientific rigor, for example, when they demand that unwelcome conclusions be proved beyond all possible doubt. Besides falsely suggesting that moral blame is the question, it shifts the scientific standard for adjudicating competing claims in science from the preponderance of the evidence (Which side made the best factual case?) to burdening the unwelcome side with impossible standards of proof, as Flynn himself notes in one post. The favored side is presumed true until the disfavored side answers an infinite regress of manufactured doubt. All our commentators explicitly reject that double-standard, but they all enforce it implicitly by manufacturing and celebrating doubt about some facts while ignoring the fuller, interleaved network of evidence.

In the court of moral judgment, defending the scientific validity of an unwelcome conclusion amounts to confessing moral guilt. Scientific strengths are rendered moral flaws (objectivity becomes insensitivity; integrity under fire, a taste for controversy). The innocuous is made suspicious (dedication is depicted as a fetish, collaboration as cabal). Scientific conclusions on what “is” (racial differences in IQ) are construed as personal preferences for what “ought to be.” Appeals to academic freedom are derogated as self-serving attempts to escape accountability or well-deserved opprobrium. Thus, while all our commentators defend rights to academic freedom and free speech, they impugn the characters of opponents to make them seem less credible.

Let us turn now to the specifics in our three commentators’ recent posts.

Avoiding the Totality of Evidence on g

Ceci is the most candid because he concedes there is “impressive data” for the first six bedrock facts in the chain of plausibility in Table 1. They “seem to accord with each prong of the [following] syllogism:”

  • An underlying ability (called g) is needed for all forms of cognitive performance
  • g is manifest in any broad cognitive battery such as IQ
  • g is related to many types of biological markers and is highly heritable
  • Large individual and group differences exist in g
  • Variation in g predicts differential life outcomes
  • Therefore, variation in life outcomes is at least partly rooted in biological differences in g

He is also candid about wishing to reject the syllogism’s conclusion that “g reflects an underlying biologically-based and stable intellectual ability, rather than a specific skill learned in school or taught by parents.” This is why he says “all of us owe Flynn a deep debt of gratitude for complicating what had started to seem like a closed case”—for “throwing a spanner” of doubt. Ceci has pinpointed here why Flynn is feted worldwide by the critics of mainstream research on intelligence: not for disproving any part of the syllogism, and not for solving the mystery of why certain scores on certain IQ subtests have risen (it remains unsolved), but for cultivating doubt about the syllogism.

In fact, the syllogism has already been empirically verified. Level of education, occupation, and income are themselves moderately heritable (60-70%, 50% and 40-50%, respectively), and these heritabilities overlap that for g by at least half (yielding 40%, 25%, and 20% of the phenotypic variation in the three life outcomes being jointly heritable with g: e.g., Rowe, Vesterdal, & Rodgers, 1998).

Ceci implies that Flynn has rendered Table 1’s bedrock facts 5 and 6 (demonstrated lack of malleability, high heritability) irrelevant when he says Flynn “has shown beyond doubt that general intelligence fluctuates systematically over time and this cannot be due to our having better genes than our grandparents.” Flynn has shown neither, of course. There is no evidence that g itself has risen much over time, if at all. Nor is there any evidence that rules out genetic change (e.g., heterosis) or that rules in environmental influences of any particular kind, if any at all, for the observed increases on certain subtests, the limited nature of which suggests that only a narrow band of skills has been affected.

For Ceci, Flynn provides a life-line of doubt to escape the decades of evidence showing that intelligence (g) is not malleable. Now freed to search for levers of malleability, Ceci proposes education. He acknowledges that the patterns of change in IQ subtest scores are opposite of what one would expect if school instruction were the cause, so he suggests that there is “clearly” something “indirect” about having more schooling that enhances intelligence. He also notes, however, that some children show greater cognitive growth in a year of schooling than do others. He interprets this, not as expected differences in children’s maturation rates, but as evidence that “clearly” the “everyday experiences that some individuals engage in” improve intelligence. For empirical support he must turn away from the large body of contrary evidence in the US from standard educational and work settings and toward bits of research from odd corners of the globe (rural Kharwar Indian and Brazilian children) or forms of “natural stimulation” during childhood (herding, running errands, playing videogames or Legos).

Turkheimer selects a different target for scientific doubt: g itself (bedrock fact 1). “Practically everything [is] wrong with the concept of g.” He chides Flynn for conceding too much to me by accepting g as an empirical phenomenon. His strategy is to assign an “intuition” to g’s advocates which he then demolishes with a thought experiment. The experiment is that he can construct a universe in which abilities are independent. He shows us that factor analysis will extract a g factor nonetheless. His conclusion is that g is therefore an “illusion.” His other-worldly demonstration does nothing, however, to invalidate the independent bodies of evidence showing that this “illusion” exists beyond factor analysis as a biological (Jensen, 1998) and cross-species phenomenon (Chabris, 2007), and that it has more predictive substance in the real world than does any other cognitive ability. In any case, Humphreys (1986) and others put his hypothesis to the test half a century ago when they earnestly tried, but failed, to build useful ability tests that did not also measure general intelligence.

Flynn seems less concerned with breaking any particular link in the empirical chain of evidence than with simply declaring the most troublesome ones irrelevant. The most troublesome are g and its high heritability, which constitute the core of the syllogism that Ceci would like to see nullified. So, while Flynn does not deny the existence of g or its high heritability, he attempts to “shatter” it with secular changes in IQ. Explaining the secular increases seems secondary, except that the proferred explanations must emphasize the cultural malleability of “intelligence.”

He accomplishes this “shattering” by reversing perhaps the most important advance in intelligence measurement of the 20th Century, which was to separate the latent trait, g, from the specific devices used to measure it, including IQ tests. This separation is what allows us to describe different mental tests according to their “g loading” (their correlation with the latent trait, g). All mental tests measure mostly g, whatever their original intent. They often differ in the narrower skills and abilities that also contribute somewhat to good performance on them, but their power to predict important life outcomes has been shown to reside almost exclusively with their common g component.

How does Flynn “shatter” this psychometrically unitary phenomenon? He ignores latent traits and focuses on the measurement device. First, he notes that raw scores on IQ tests have risen over time because people are doing dramatically better on a few of the various subtests that contribute points to the overall IQ score. He cannot say that g has increased, because scores on some g loaded subtests have increased a lot (e.g., Similarities) but others hardly at all (e.g., Vocabulary), so he merely says “intelligence” has increased. Because the score increase rests on only a few subtests, he pulls them out for special attention as the source of this increase in “intelligence.” This pulling apart of the measurement device as if it were pulling apart “intelligence” is presented as the “shattering” of g. The impression is solidified by presenting sports analogies in which the athletic “skills” in question are “functionally independent.”

Flynn can now examine these shattered pieces of the IQ battery as “interesting in their own right,” as if they no longer measured mostly g. He points to selected surface characteristics of their items to generate scenarios for very general cultural factors that could have increased performance on them. These mechanisms are but metaphors, such as the donning of “scientific spectacles” by the population at large owing to the industrial and scientific revolutions. By these means “intelligence” is not only made malleable, but made amenable to change via “cultural priorities” which we can presumably manipulate at will.

The problem with his explanation, however, is that IQ (g) remains highly heritable. Flynn “solves this paradox” by postulating, in effect, that environments mimic genetic influence. For this he turns to William Dickens’ proposed environmental “multipliers.” By this model, any one of our independent abilities will trigger environmental reinforcement of all our abilities. That is, environmental demands for exercising our abilities come bundled together—trigger one demand and you trigger all, which boosts all abilities in tandem. The result is to reconstitute g from the outside in. Instead of g being internally generated, it is now apparently the product of positive manifold among opportunities and demands for cognitive development in the external environment. The resulting g is highly heritable because of the single, genetically-influenced ability that happened to recruit a very powerful environment. Dickens’ is a valiant attempt to make g seem malleable by having genes themselves recruit environments to make it so.

Moral Denigration of Politically Unwelcome Research

Ceci argues that the current taboo against discussing one side of the argument on race and intelligence is not good for science. Indeed, he describes how the taboo can produce one-party science: all professional advantages and accolades flow toward scientists who explicitly disavow the politically incorrect side, as he has. I appreciate his candor.

I am fortunate that these are my views because they are politically correct and garner me praise, speaking and writing invitations, and book adoptions at the same time those who disagree with me are demeaned, ostracized, and in some cases threatened with tenure revocation even though their science is as reasonable as mine….I can tell my side far more easily than they are permitted to tell theirs.

Unlike Ceci, who believes that this double-standard is “unfair,” Turkheimer contends that those who disagree with their views “deserve the vigorous disapprobation they often receive.” The more eminent the disagreeing scientist, presumably the more vigorous the opprobrium ought to be—a principle exemplified by the recent vilification of Nobel laureate James Watson.

Turkheimer’s stance allows Flynn to claim the scientific high ground by turning Turkheimer’s peremptory question to “racial scientists” against him: is there any evidence that you, Turkheimer, would be willing to accept as disproving your view? Unlike Turkheimer, Flynn would allow “rational discussion of offensive ideas”—it is “okay,” he says. Flynn yields none of the moral high ground to Turkheimer, however, because he, too, labels the research conclusions with which they all disagree as being “offensive.” Thus, Turkheimer and Flynn seem to disagree only on how immoral the politically incorrect ideas are (“dangerous” or merely “offensive”) and thus whether the “offensive” may be allowed to compete in the marketplace of ideas. Flynn grants entry.

One supposes I should thank Flynn for his magnanimity in allowing others to rationally discuss my “offensive” research. My writings certainly meet his offensiveness criterion, if only because I review evidence and draw conclusions without regard to political correctness (e.g., Gottfredson, 2005a; Wainer & Robinson, in press; see www.udel.educ/educ/gottfredson/reprintsfor other publications). Perhaps he will show others who also believe I am mistaken how to dissect my evidence and argument to locate my errors and prove me wrong. I have waited decades for someone to do so.

In view of his stance on Watson, Turkheimer seems more likely to favor “vigorous [opprobrium],” which is much easier as well. Moral opprobrium requires no mucking about hip-deep in data or the complications of analyzing it properly. It is accomplished most easily by simply evoking the specter of genocide, as Turkheimer does in his first posting.

[H]istory has shown us only too clearly what can happen when simplistic views of human ability make poorly informed contact with biology and genetics….I am by training a behavior geneticist, and as such I am too well-acquainted with the ugly places oversimplified thinking about human ability and genetics can lead to let the phrase “good genes” pass without a shiver.

Lest readers miss his allusion to genocide, he refers to researchers who study group differences in (phenotypic) intelligence as “racial scientists” and “racialists” and, moreover, in the context of “200 years of [black] enslavement…, followed by 100 years of Jim Crow oppression.”

Turkheimer begins his post on race and IQ by moving the genetic question beyond the reach of empirical inquiry: “The important questions about the role of genetics in the explanation of racial differences in ability are not empirical, but theoretical and philosophical.” This he does by substituting the word “innate” for “genetic,” and then criticizing the substituted term. By this logic, we could shift his own field of study—behavior genetics—into his university’s philosophy department.

Turkheimer continues to invoke scientific expertise, however, when it serves to amplify Watson’s moral culpability or his own moral authority (as in the quotation above). Specifically, he chastises the great geneticist for speaking on scientific questions that are, according to Turkheimer, “ultimately outside [his] real expertise.” I do not know whether Turkheimer would grant me expertise for my decades investigating the nature and consequences of racial disparities in intelligence, or that he would claim any special expertise of his own, but I suspect he would avoid the question altogether by reverting to his opening argument, in which he renders scientific expertise irrelevant.

Criticisms of research into intelligence differences have become more strident and ad hominen as the research base has become, in Ceci’s words, ever more “impressive.” The postings by Flynn, Turkheimer, and even Ceci are laced with aspersions on the characters of the researchers with whom they disagree and with efforts, explicit or not, to encourage moral panic in the service of suppressing unwelcome truths.

Turkheimer describes the “possibility of genetic differences between the races for IQ” as “ugly,” “offensive,” “destructive,” and “dangerous.” He heightens our sense of outrage by listing slurs against various ethnic groups. He does not, however, explain why the genetic hypothesis is “dangerous”—because it is false, or because it might be true? Again, I await rational analyses by those who encourage moral panic.

Flynn is less direct in alleging that unwelcome conclusions about racial differences do harm, yet he specifies how he thinks they could do so. He prefaces his allegation by telling us why we are fortunate that political pressure did not dissuade Arthur Jensen from posing the part-genetic hypothesis in 1969. It prompted Flynn to greatness.

I would never have made any contribution to psychology. It was arguing versus Jensen that led me to investigate IQ trends over time, led me to formulate a more sophisticated defense of affirmative action, led me to my book What is Intelligence, which (I hope) has done much to unfreeze our thinking about g, or the general intelligence factor.

Flynn then plays the trump card for suppressing unwelcome inquiry: “It may be said that truth is not the only value.” For support, he refers to imaginings of the social harm possibly done by Jensen’s work, seemingly because his conclusions were true.

A colleague once told me that all that has been learned in the wake of Jensen was not worth hundreds of acts of discrimination, those that may have occurred from naïve racists feeding on his views.

For additional support, Flynn cites 18th century philosopher Jean-Jacques Rousseau on the dangers of atheism: “Is this single truth worth the danger that thousands of people of simple faith will become immoral, rather than acting responsibly?” In so doing, Flynn echoes Turkheimer’s implied fear that bad ideas about race will push average Americans down the slippery slope toward racial genocide. By branding such ideas as socially dangerous, both Flynn and Turkheimer encourage their suppression—despite Flynn’s ostensible admonition that we ought not do so.

All three commentators belittle as well as morally taint the unspecified scientists with whom they disagree. Flynn consistently describes the “g-men” and their ideas in derogatory terms (see especially his book). They have a “blinding obsession” for an “imperialistic” g, empirical challenges to whose “supremacy” even the best of them “strongly resist” and therefore must be irrationally wedded. One might conclude after reading Flynn’s book that Arthur Jensen, arguably the greatest intelligence researcher of the 20th Century (Detterman, 1998), is but a buffoon. Turkheimer also speaks of researchers who study the general intelligence factor as “g-men” who “revere” the positive correlations among ability tests and wish, thereby, to establish the “supremacy of g.” Ceci likewise derogates the “IQ mafia” as being “enamored with g” and “touting” its ability to tie together myriad seemingly unrelated phenomena—as if surely exaggerating g’s explanatory power.

Moral slander and fear mongering stifle scientific dialog, as is their purpose. They also preclude rational inquiry into the social costs of telling untruths about racial disparities in intelligence.

The Truth—What Possible Good? Lies—What Possible Harm?

Flynn implies that the public interest would be served by withholding the truth about race and IQ. But exactly which truth? It cannot be the part-genetic hypothesis because he rejects it as false. He uses Jensen’s research post-1969 to argue his point, but that research focused almost entirely on differences in phenotypes. So, Flynn is encouraging either the suppression of well-established facts about racial differences in phenotypic intelligence or else the common confusion between genotype and phenotype, which serves the same purpose.

The game here is not to suppress discussion of genetic differences but to suppress knowledge of phenotypic differences. The latter make the former more plausible, so the specter of genetic causation is used as a club to beat back scientific knowledge about racial disparities in developed abilities, whatever their origins.

None of our commentators mentions the large, persisting, and socially important black-white differences in phenotypic g, except to suggest that they are rapidly disappearing (something I’ve heard now for over thirty years). Instead, Flynn, Turkheimer, and Ceci focus almost entirely on cultivating doubt about the scientific edifice for g itself. If g is an illusion, then there can be no racial differences in g. If the scores on some IQ subtests rise over time, then perhaps we can declare “intelligence” malleable and disregard the high heritability of g. Nature’s constraints may not be so constraining after all, our commentators imply, so we have no lessons to learn from the many prior educational interventions and social policies that failed because they presumed ready malleability. If there is a lesson, says Flynn, it is that we should “impose” such interventions on individuals more vigorously and into their adult years. We begin to envision the state, guided by a firm hand and correct priorities, trying to reshape us to eliminate cognitive diversity for egalitarian ends.

Proponents of the taboo on discussing race and IQ assume that the taboo is all for the common good, but whose good, exactly, is served? It is most certainly not individuals of below-average intelligence, who face a tremendous uphill battle in modern, literate societies where life becomes increasingly complex by the day. General intelligence (g) is simply a general proficiency to learn and reason. Put another way, it is the ability to deal with complexity or avoid cognitive error. Virtually everything in life requires some learning or reasoning and thus confers an advantage on brighter individuals. Life is complex, and complexity operates like a headwind that impedes progress more strongly for individuals lower on the IQ continuum. Everyone makes cognitive mistakes, but lower intelligence increases the risk of error.

Take, for example, health care. Patients differ enormously in intelligence level, and these differences have life and death consequences for them. Individuals of lower health literacy, or IQ, are less likely to seek preventive care even when it is free, use curative care effectively when they get it, understand and adhere to treatment regimens, or avoid health-damaging behavior. They have worse health, more accidental injuries, higher health costs, and die sooner—regardless of income, insurance coverage, or quality of health care. Health care matters, as do material resources and motivation, but mental resources matter too. They are critical in the prevention and self-management of chronic illnesses such as diabetes and heart disease. Health self-care is an increasingly complex life-long job for all of us, which becomes even more complex as we age and experience more health problems.

It overstates only slightly to say that health care providers currently pay no attention to patient differences in the ability to learn and understand. As health literacy researchers have shown, however, a sizeable fraction of patients in urban hospital outpatient clinics are unable to understand an appointment slip (when to come back), a label indicating how to take four pills a day, or, among their insulin-dependent diabetic patients, the signs of low (or high) sugar and what action to take to bring their blood sugar back under control. Do proportionately more blacks have such problems? Yes, many more. Is that a reason to continue ignoring or disputing individual and group differences in g?


Chabris, C. F. (2007). Cognitive and neurobiological mechanisms of the law of general intelligence. In M. Roberts (Ed.), Integrating the mind: Domain general versus domain specific processes in higher cognition. Hove, UK: Psychology Press.

Detterman, D.K. (Ed.). (1998). A king among men: Arthur Jensen

[Special issue]. Intelligence, 26(3).

Gottfredson, L. S. (Ed.) (1997). Intelligence and social policy [Special issue]. Intelligence, 24(1).

Gottfredson, L.S. (2005a). Implications of cognitive differences for schooling within diverse societies. In C. L. Frisby & C.R. Reynolds (Eds.), Comprehensive handbook of multicultural school psychology (pp. 517–554). New York: Wiley.

Gottfredson, L. S. (2005b). What if the hereditarian hypothesis is true? Psychology, Public Policy, and Law, 11, 311-319.

Humphreys, L. G. (1986). Commentary. Journal of Vocational Behavior, 29(3), 421-437.

Jensen, A. R. (1998). The g factor: The science of mental ability. Westport, CT: Praeger.

Neisser, U., Boodoo, G., Bouchard, T.J., Jr., Boykin, A.W., Brody, N., Ceci, S.J., et al. (1996). Intelligence: Knowns and unknowns. American Psychologist, 51, 77–101.

Rowe, D. C., Vesterdal, W. J., & Rodgers, J. L. (1998). Herrnstein’s syllogism: Genetic and shared environmental influences on IQ, education, and income. Intelligence, 26, 405-423.

Snyderman, M. & Rothman, S. (1988). The IQ controversy, the media and public policy. New Brunswick, NJ: Transaction Books.

Wainer, H., & Robinson, D. (in press). Interview with Linda Gottfredson. Journal of Educational and Behavior Research. (see also )

Also from this issue

Lead Essay

  • In this month’s lead essay, famed intelligence researcher James R. Flynn draws on his new book, What Is Intelligence? Beyond the Flynn Effect, to help answer some of the puzzles surrounding the controversial issue of IQ. Flynn, who first discovered that IQ scores were rising generation after generation, notes that this improvement has been more dramatic on some parts of IQ tests than on others, challenging ideas of a unified general intelligence or “g.” Flynn argues that the fact cognitive abilities do not develop together has important implications for education. Additionally, Flynn contends that the evidence for the co-determining reciprocal influence of brain and environment should lead us to set aside simple ideas about the primacy of nurture and nature in intelligence. Once we grasp that “the brain is much more like our muscles than we had thought,” we can do more to improve cognitive performance by doing more to exercise the brain.

Response Essays

  • Linda Gottfredson, co-director of the Delaware-Johns Hopkins Project for the Study of Intelligence and Society, defends the unity of general intelligence, or g, against Flynn’s attempt to “unravel g into its component parts” by charging that “his core argument rests on logical fallacies that profoundly misinterpret the evidence.”

  • University of Virginia psychologist Eric Turkheimer contests even James Flynn’s modest accommodation with the construct of g, or general intelligence. Turkheimer argues that the “fundamental intuition” of g is that “universal positive relations among mental tests compels a single dominant explanatory construct,” but that “the fundamental intuition is wrong.” According to Turkheimer, g is not “discovered,” but is simply posited as a convenience, like the Prime Meridian. g may be useful for some purposes, but a multidimensional explanation of ability may still be correct. “The trick is not to get hooked on any particular way of dividing up the pie,” Turkheimer writes, “because it is a short step from there to trying to find the Greenwich Meridian at the bottom of the North Atlantic.”

  • Stephen J. Ceci, the Helen L. Carr Professor of Developmental Psychology at Cornell University, explains why Flynn’s work has been so important. According to Ceci, the received wisdom about the nature and effects of g, or general intelligence, led “some to argue that inequality in the distribution of wealth, prestige, and educational attainment is, in part, a consequence of unequal distribution of the intellectual capacity needed for high levels of functioning.” However, Ceci says, “[Flynn] has shown beyond doubt that general intelligence fluctuates systematically over time and this cannot be due to our having better genes than our grandparents,” and he goes on to explore the puzzles raised by this discovery.