This paper originally appeared in Society, March-April 1994, v. 31 (7), no. 3, p. 53
Linda S. Gottfredson is professor of educational studies at the University of Delaware and co-director of the Project for the Study of Intelligence and Society. She has published widely on fairness in testing and racial inequality, focusing most recently on race-norming and the dilemmas in managing workforce diversity. Her current work examines social policy based on the egalitarian fiction.
Social science today condones and perpetuates a great falsehood one that undergirds much current social policy. This falsehood, or "egalitarian fiction," holds that racial-ethnic groups never differ in average developed intelligence (or, in technical terms, g, the general mental ability factor). While scientists have not yet determined their source, the existence of sometimes large group differences in intelligence is as well-established as any fact in the social sciences. How and why then is this falsehood perpetrated on the public? What part do social scientists themselves play, deliberately or inadvertently, in creating and maintaining it? Are some of them involved in what might be termed "collective fraud?" Intellectual dishonesty among scientists and scholars is, of course, nothing new. But watchdogs of scientific integrity have traditionally focused on dishonesty of individual scientists, while giving little attention to the ways in which collectivities of scientists, each knowingly shaving or shading the truth in small but similar ways, have perpetuated frauds on the scientific community and the public at large. Perhaps none of the individuals involved in the egalitarian fiction could be accused of fraud in the usual sense of the term. Indeed, I would be the first to say that, like other scientists, most of these scholars are generally honest. Yet, their seemingly minor distortions, untruths, evasions, and biases collectively produce and maintain a witting falsehood. Accordingly, my concern here is to explore the social process by which many otherwise honest scholars facilitate, or feel compelled to endorse, a scientific lie.
It is impossible here to review the voluminous evidence showing that racial-ethnic differences in intelligence are the rule rather than the exception (some groups performing better than whites and others worse), and that the well-documented black-white gap is especially striking. All groups span the continuum of intelligence, but some groups contain greater proportions of individuals that are either gifted or dull than others. Three facts regarding these group differences are of particular importance here for together they contradict the claim that there are no meaningful group differences. Racial-ethnic differences in intelligence are real. The large average group differences in mental test scores in the United States do not result from test bias, which is minuscule overall, as even a National Academy of Science panel concluded in 1982. Moreover, intelligence and aptitude tests measure general mental abilities, such as reasoning and problem solving, not merely accumulated bits of knowledge and thus tap what experts and laymen alike view as "intelligence."
Regardless of how we choose to construe them, differences in intelligence are of great practical importance. Overall they predict performance in school and on the job better than any other single attribute or condition we have been able to measure. Intelligence certainly is not the only factor that affects performance, but higher levels of intelligence greatly increase people's odds of success in many life settings. Group disparities in intelligence are stubborn. Although individuals fluctuate somewhat in intelligence during their lives, differences among groups seem quite stable. The average black-white difference, for example, which appears by age six, has remained at about 18 Stanford-Binet IQ points since it was first measured in large national samples over seventy years ago. It is not clear yet why the disparities among groups are so stubborn the reasons could be environmental, genetic, or a combination of both but so far they have resisted attempts to narrow them. Although these facts may seem surprising, most experts on intelligence believe them to be true but few will acknowledge their truth publicly.
The 1988 book The IQ Controversy: The Media and Public Policy, by psychologist-lawyer Mark Snyderman and political scientist Stanley Rothman provides strong evidence that the general public receives a highly distorted view of opinion among "IQ experts." In essence, say Snyderman and Rothman, accounts in major national newspapers, newsmagazines, and television reports have painted a portrait of expert opinion that leaves the impression that "the majority of experts in the field believe it is impossible to adequately define intelligence, that intelligence tests do not measure anything that is relevant to life performance, and that they are biased against minorities, primarily blacks and Hispanics, as well as against the poor." However, say the authors, the survey of experts revealed quite the opposite: On the whole, scholars with any expertise in the area of intelligence and intelligence testing ... share a common view of [what constitute] the most important components of intelligence, and are convinced that [intelligence] can be measured with some degree of accuracy. An overwhelming majority also believe that individual genetic inheritance contributes to variations in IQ within the white community, and a smaller majority express the same view about the black-white and SES [socioeconomic] differences in IQ.
Unfortunately, such wholesale misrepresentation of expert opinion is not limited to the field of intelligence, as Rothman has shown in parallel studies of other policy-related fields such as nuclear energy or environmental cancer research. However, the study of IQ experts revealed something quite surprising. Most experts' private opinions mirrored the conclusions of psychologist Arthur Jensen, whom the media have consistently painted as extreme and marginal for holding precisely those views. As Snyderman and Rothman point out, the experts disclosed their agreement with this "controversial" and putatively marginal position only under cover of anonymity. No one, not even Jensen himself, had any inkling that his views now defined the mainstream of expert belief. Although Jensen regularly received private expressions of agreement, he and others had been, as Snyderman and Rothman note, widely castigated by the expert community for expressing some of those views. Several decades ago, most experts, among them even Jensen, believed many of the views that the media now wrongly describe as mainstream for example, that cultural bias accounts for the large black-white differences in mental test scores. While the private consensus among IQ experts has shifted to meet Jensen's "controversial" views, the public impression of their views has not moved at all. Indeed, the now-refuted claim that tests are hopelessly biased is treated as a truism in public life today. The shift in private, if not public, beliefs among IQ experts is undoubtedly a response to the overwhelming weight of evidence which has accumulated in recent decades on the reality and practical importance of racial-ethnic differences in intelligence. This shift is by all indications a begrudging one, and certainly no flight into "racism."
Snyderman and Rothman found that as many IQ experts as journalists and science editors (two out of three) agreed with the statement that "strong affirmative action measures should be used in hiring to assure black representation." Fully 63 percent of the IQ experts described themselves as liberal politically, 17 percent as middle of the road, and 20 percent as conservative not much different than the results for journalists (respectively, 64, 21, and 16 percent). Moreover, as Snyderman and Rothman suggest (and as is consistent with personal accounts by Jensen and others), many of the surveyed experts, while agreeing with Jensen's conclusions, may disapprove of his expressing these conclusions openly. Consistent with this, when queried about their respect for the work of fourteen individuals who have written about intelligence or intelligence testing, the IQ experts rated Jensen only above the widely (but apparently unjustly) vilified Cyril Burt. Despite the fact that most agreed with Jensen, they rated him far lower than often like-minded psychometricians who had generally stayed clear of the fray. Jensen even received significantly lower ratings than his vocal critics, such as psychologist Leon Kamin, whose scientific views are marginal by the experts' own conclusions. By contrast, the experts in environmental cancer research behaved as one would expect; they gave higher reputational ratings to peers who are closer to the mainstream than to high-profile critics. Snyderman's and Rothman's findings therefore suggest that a high proportion of experts are misrepresenting their beliefs or are keeping silent in the face of a public falsehood. It is no wonder that the public remains misinformed on this issue.
IQ experts feel enormous pressure to "live within a lie," to use a phrase by Czech writer and leader Vaclav Havel characterizing daily life under communist rule in Eastern Europe. Havel argued, in The Power of the Powerless, that, by living a lie, ordinary citizens were complicit in their own tyranny. Every greengrocer, every clerk who agreed to display official slogans not reflecting his own beliefs, or who voted in elections known to be farcical, or who feigned agreement at political meetings, normalized falsification and tightened the regime's grip on thought. Each individual who lived the lie, who capitulated to "ideological pseudo-reality, " became a petty instrument of the regime. As many commentators have noted, Americans may not speak certain truths about racial matters today. To adapt a phrase, there is a "structured silence."
Social scientists had already begun subordinating scientific norms to political preferences and creating much of our current pseudo-reality on race by the mid-1960s. Sociologist Eleanor Wolf, in a 1972 article in Race, for example, detailed her distress at how fellow social scientists were misusing research data to support particular positions on civil rights policy: presenting inconclusive data as if it were decisive; lacking candor about "touchy" subjects (such as the undesirable behavior of lower-class students); blurring or shaping definitions (segregation, discrimination, racism) to suit "propagandistic" purposes; making exaggerated claims about the success of favored policies (compensatory education and school integration) while minimizing or ignoring contrary evidence. As a result, social science and social policy are now dominated by the theory that discrimination accounts for all racial disparities in achievements and well-being. This theory collapses, however, if deprived of the egalitarian fiction, as does the credibility of much current social policy. Neither could survive intact if their central premise were scrutinized.
No wonder, then, that IQ researchers find themselves under great professional and institutional pressure to avoid not only engaging in such scrutiny but even appearing to countenance it. The scrutiny itself must be discredited; the egalitarian fiction must be raised above serious scientific question. Scientists must at least appear to believe the dogma. As was the case in Havel's communist-dominated Eastern Europe, in American academe feigned belief in the official version of reality is maintained largely by routine obeisance of academics as they pursue their own ambitions.
Scholars realize their scholarly ambitions primarily through other scholars. Peer recognition is the currency of academic and scientific life. It is crucial to a scholarly reputation and all the steps toward status and success publications, professional invitations and awards, promotion, tenure, grants, fellowships, election to professional office, appointment to prestigious panels. One's ability even to carry out certain kinds of research, funded or not, may be contingent upon peer recognition and respect for instance, getting collaborators, subjects, or cooperation from potential research sites. Just as in personal life, a high professional reputation depends upon a sustained history of "appropriate" behavior, and it may be irreparably damaged by hints of scandal or impropriety. Similarly, the reputations of scientists and their organizations are enhanced or degraded by those for whom they show regard and approval. Associating oneself with highly regarded individuals or ideas enhances, even if slightly, one's own status.
Awarding an honor to a luminary can enhance the reputation of one's own organization, especially if the recipient accepts the honor with genuine appreciation. By the same token, one risks "staining" one's reputation by associating with, honoring, defending, or even failing to condemn the "wrong" sort of individual or idea. In short, how one gives or withholds one's regard is important for one's professional reputation because it affects the regard one receives. Such a social system enhances the integrity of science and is furthered by personal ambition when the members of the community base their regard on scholarly norms, such as competence, creativity, and intellectual rigor. However, such a system breeds intellectual corruption when members systematically subordinate scientific norms to other considerations money, politics, religion, fear. This is what appears to be happening today in the social sciences on matters of race and intelligence. As sociologist Robert Gordon argues, social science has become "one-party science."
Democrat or Republican, liberal or conservative, virtually all American intellectuals publicly adhere to, if not espouse, the egalitarian fiction. And many demonstrate their party loyalty by enforcing the fiction in myriad small ways in their academic routine, say, by off-handedly dismissing racial differences in intelligence as "a racist claim, of course," criticizing authors for "blaming the victim," or discouraging students and colleagues from doing "sensitive" research. One can feel the gradient of collective alarm and disapproval like a deepening chill as one approaches the forbidden area. Researchers who cross the line occasionally face overt censorship, or calls for it. For example, one prominent (neoconservative) editor rejected an author's paper, despite finding it scientifically sound, because there are social "considerations" which "overweigh the claims of social science." Another eminent editor, after asking an author to soften the discussion in his article, recently published the revised paper with an editorial postscript admonishing scientists in the field to find a "balance" between the need for free exchange of research results on intelligence and the (presumably comparable) "need" that "no segment of our society ... feel threatened" by it.
Whether motivated by a sincere concern over supposedly "dangerous" ideas or by a desire to distance themselves publicly from unpopular ideas, editors who use such non-academic standards discourage candor and stifle debate. They deaden social science by choking off one source of the genuine differences of opinion that are its lifeblood. Overt censorship of research is uncommon, probably because it is an obvious affront to academic norms. Less striking forms of censorship directly affect many more academics, however, and so may be more important. Easier to practice without detection and to disguise as "academic judgment," they serve to keep scholars from pursuing ideas that might undermine the egalitarian dogma.
A less obvious form of censorship, which has become somewhat common recently, is indirect censorship. It is accomplished when academic or scientific organizations approve some views but repudiate or burden others on ideological grounds. Sometimes the ideological grounds are explicit. Campus speech codes are a well-known example which, had they been upheld in the courts, would have made repudiation of the egalitarian fiction a punishable offense on some campuses. The earlier (unsuccessful) attempt to include possible "offense to minority communities" as grounds for refusing human subjects approval is another example.
Gordon reports yet others, including the National Institutes of Health's new extra layer of review for politically "sensitive" grant proposals and the University of Delaware's recent policy (reversed by a national arbitrator) of banning a particular funding source because, so the university claimed, it supports research on race which "conflicts with the university's mission to promote racial and cultural diversity." Gordon also outlines in detail as political scientist Ian Blits has done the covert application of ideological standards to facilitate expression of some views but burden others. This form of indirect censorship, also falling under the rubric of "political correctness," occurs when university administrators, faculty, or officers of professional associations disguise as "professional judgment" an ideological bias in their enforcing of organizational rules, extending faculty privileges, protecting faculty rights, and weighing evidence in faculty promotions and grievances.
Recently, some American universities have invoked "professional judgment" as a pretext for reclassifying "controversial" scholarly publications in their annual merit reviews as "non-research," to misrepresent outside peer reviews in evaluating controversial professionals up for promotion, and to limit student access to these professors. Such thinly veiled bias publicly demonstrates the officials' own adherence to the prescribed institutional attitudes and their willingness to enforce them, not only protecting those officials from protest but also encouraging fellow members of the institution to toe the line.
Covert censorship is far more common than overt or indirect censorship. It consists of bias in the application of scientific norms when reviewers evaluate their peers' work for funding, publication, presentation, or dissemination. Individual ideological biases are found in all fields, of course, but the hope is that such biases remain small and will cancel each other out over the long run hence the importance of a free and open exchange of data, theories, and results. What I have in mind is systematic bias and a pervasive double standard which impedes one line of research and accords another undeserved hegemony. In one-party science, the disfavored line of work is subjected to intense scrutiny and nearly impossible standards, while the favored line of work is held to lax standards in which flaws are overlooked (called "oversight bias" in the psychological literature). Similarly, the disfavored idea is rejected unless it is "balanced" by including proponents of the favoapplication of scientific norms when reviewers evaluate their peers' work for funding, publication, presentation, or dissemination.
The broader circle of critics in the social sciences often implicitly dismisses the legitimacy of research on intelligence itself by arguing that "intelligence" is undefinable or unmeasurable as if the critics' own favored constructs (social class, culture, self-concept, anxiety, and so on) were as well validated and operationalized. Others now also seek to deny IQ researchers (but not themselves) use of the concept "race" because, they assert, race is not a biological condition, but is socially constructed. The double standards can even ricochet back and forth, depending on the particular question being considered. Gordon recalls how sociologists failed to criticize sociologist James Coleman for omitting student ability from his analyses of school integration (which led to overstating the impact of integrated schools on black achievement-for sociologists a favorable outcome), but how they criticized him roundly for the very same omission in analyses of private versus public schools, (which led to overstating the impact of private schools on black achievement an unfavorable outcome). In short, in one-party science, scientific regard flows like political patronage to loyal and active party members, who can demonstrate their loyalty by being alert to hints of dissidence. Like all one-party political systems, one-party science becomes intellectually corrupt and arrogant as it gains confidence in its power.
The most insidious corruption to which one-party science leads is pervasive self-censorship, what involved researchers generally prefer to regard as "prudence" or "avoiding unnecessary trouble." Coleman has drawn particular attention to the problem of "self-suppression" "the impulse not to ask the crucial question" in research on race. In an example from his own research for the influential "Coleman Report," he describes his failure to conduct important analyses that might have produced embarrassing findings about the abilities of black teachers. Another way of avoiding unwanted results is to ignore certain data, subjects, or variables. Or unwanted results can be omitted, buried in footnotes, explained away, or simply ignored in one's conclusions. The most subtle form of self-censorship is deliberate avoidance of making crucial connections, or denying them. Psychologist Richard Herrnstein has noted that it was his drawing out the implications of one such connection, namely, that some portion of (white) social class differences in intelligence is genetic, that sparked his public excoriation in the 1970s.
Normally, scholars are eager to explicate illuminating connections between subspecialties. They are reluctant to do so, however, when these connections put in question the egalitarian dogma on race. Virtually all sociologists and economists ignore the literature on intelligence despite its central importance to core issues in their disciplines, such as inequalities in occupation and income.
Many psychometricians, especially those working for large testing organizations, avoid referring to "intelligence" and often seem reluctant to say much about the practical or theoretical meaning of the racial differences they observe on unbiased tests. But even remaining within one's subfield is often not enough, for the field of intelligence itself is widely suspect. Hence some scholars explicitly disavow unpopular connections that critics might attribute to them. For example, they will argue in favor of the importance of intelligence for scholastic performance but then assure their readers, over-optimistically, that the racial gap "seems to be closing rapidly." The tenor of these preemptive disclaimers is clear. While researchers in any field may lightly dismiss the credibility of key connections regarding race and intelligence, no one ever lightly endorses their credibility with impunity. Even those of us committed to candor are exceedingly cautious when expressing informed opinions on certain topics, especially the genetics of race. Thus, publicly stated opinions of researchers about matters outside their subfields tend in one direction -to dispute or undercut the facts necessary for toppling the egalitarian fiction. What may be tolerable behavior at the individual level becomes intolerable bias at the aggregate level. Censorship even self-censorship requires justification, or at least apparent justification. On the whole, those who would squelch open inquiry of the egalitarian fiction base their justification on two assertions: 1) Research on racial differences in intelligence has already been scientifically "discredited." 2) Inquiry into racial differences is immoral.
Point one asserts that the egalitarian premise is absolute truth and hence beyond scientific scrutiny. Point two is indifferent to its truth. Both counsel outrage at the very thought of the research. The claim that the research has been discredited rests largely on extensive misrepresentation that is often embarrassingly crude or casual for example, contradicting arguments an author never made, while ignoring what was actually stated; attributing policy preferences to an author which are opposite of what the author actually expressed; or simply alleging fraud or gross incompetence without any substantiation whatsoever. The claim that the research is immoral rests squarely on the view that, regardless of the truth, the study itself can only be harmful. In fact, some critics assert (mostly privately) that the greater the truth, the greater the danger it poses to lower-scoring groups, and thus the greater the need to suppress it.
Despite their differences, both justifications for censorship often take the form of demonizing open inquiry by labeling it (and the people who practice it) as "dangerous," "fascist," "ideological," or "racist." The study of race and intelligence is something, they tell us, that no decent person let alone a serious scientist would ever do and that every decent person and serious researcher would oppose. Thus, in a kind of Orwellian inversion, marked by what Gordon calls "high talk and low blows," the suppression of science presents itself as science itself. Intellectual dishonesty becomes the handmaiden of social conscience, and ideology is declared knowledge while knowledge is dismissed as mere ideology. Neither social policy, nor science, nor society itself is served well by scientific silence on racial differences in intelligence.
Enforcement of the egalitarian fiction has tragic consequences, especially for blacks. The outcomes are even worse than researchers of intelligence predicted two decades ago. The falsehood, because it tries to defy a reality that has conspicuous repercussions in daily life, is doing precisely what it was meant to avoid: producing pejorative racial stereotypes, fostering racial tensions, stripping members of lower-scoring groups of their dignity and incentives to achieve, and creating permanent social inequalities between the races. Enforcement of the lie is gradually distorting and degrading all institutions and processes where intelligence is at least somewhat important (which is practically everywhere) but especially where it is most important (in public schools, higher education, the professions, and high-level executive work). The falsehood requires that there be racial preferences and that their use be disguised, wherever intelligence has at least moderate importance. Society is thus being shaped to meet the dictates of a collective fraud. The fiction is aiding and abetting bigots to a far greater degree than any truth ever could, because its specific side-effects racial preferences, official mendacity, free-wielding accusations of racism, and falling standards are creating deep cynicism and broad resentment against minorities, blacks in particular, among the citizenry.
Enforcement of the egalitarian fiction is not a moral or scientific imperative; it is merely political. It is terribly short-sighted, for it corrupts both science and society. However, just as the fiction is sustained by small untruths, so can it be broken down by many small acts of scientific integrity. This requires no particular heroism. All that is required is for scientists to act like scientists to demand, clearly and consistently, respect for truth and for free inquiry in their own settings, and to resist the temptation to win easy approval by endorsing a comfortable lie.
Jan H. Blits and Linda S. Gottfredson, "Equality or Lasting Inequality?", Society, 27 (3) March/April 1990.
Robert A. Gordon, The Battle to Establish a Sociology of Intelligence: A Case Study in the Sociology of Politicized Disciplines, Baltimore, Md.: The Johns Hopkin University, Department of Sociology, 1993.
Linda S. Gottfredson, "Dilemmas in Developing Diversity Programs." In Diversity in the Workplace: Human Resources Initiatives, Susan Jackson (ed.). New York: The Guilford Press, 1992.
Linda S. Gottfredson and James C. Sharf (eds.), "Fairness in Employment Testing." Journal of Vocational Behavior, 33, December 1988.
Richard J. Herrnstein, "A True Tale from the Annals of Orthodoxy", Preface to IQ in the Meritocracy, Boston, Mass.: Little, Brown and Company, 1973.
Daniel Seligman, A Question of intelligence, New York: Birch Lane Press, 1992.