Father of 'g' tackles 'whys'

Michael Rutter delves into the psyche of a controversial scientist.

This is a thoroughly modern book in that it comprises the transcript of a series of lengthy interviews conducted by email rather than face-to-face interviews. The subject is Jensenism, the interviewee is Arthur Jensen himself and the interviewer is Frank Miele, the editor of Skeptic magazine. As the book illustrates well, Jensen is the subject of both extreme adulation (described in a special issue of the journal Intelligence as "a king among men") and extreme derision and hostility (Stephen Jay Gould claimed that Jensenism rested on "a rotten edifice").

The book is certainly a good read, and Miele asks some appropriately searching questions (with some cross-examination) about the science. He also uses the interviews as a means of providing a good picture of what Jensen is like as a person. Miele is reasonably even-handed in his approach, although he allows Jensen to get off the hook on several key issues, and presents a very positive view of his qualities as both a scientist and a responsible human being.

I doubt whether the book will persuade either the disciples or the critics to change their minds. Those who admire Jensen are likely to consider that vigorous empirical research has amply confirmed Jensen's view of the biological reality of the g factor - representing intrinsic general mental ability - and of the importance of genetic influences on individual differences in such mental ability. Critics, by contrast, are likely to respond by arguing that neither of these claims constitutes the main source of contention; rather it is all the other features that accompany them. So far as these features are concerned, critics will probably regard Jensen's vigorous defence as Blairite in style - denying the evidence as far as he can, retreating to the position that his words have been misunderstood, side-stepping the ethical issues and casting the blame on others. So, what are the issues and what does the evidence show?

Jensen would clearly like to be known most of all for his research into g.

The basic notion is that, on the whole, people who do well (or badly) on one sort of mental test tend to do well (or badly) on other mental tests.

The reason put forward is that there is a basic general mental ability that underlies the variety of special cognitive skills. It is considered to reflect the intrinsic biology of the brain (Jensen draws attention to the modest correlation between g and brain size, and to some physiological correlates of IQ), and that genetic factors predominate in bringing about individual differences in g. He estimates that they account for 75 per cent of population variance in IQ.

Jensen goes on to claim that Robert Plomin has identified at least four genes for intelligence, and that these findings have held up in repeated studies. Jensen further claims that the family environment has a near-zero effect on individual differences in IQ by the time individuals reach adult life. He argues that insofar as non-genetic influences have any effect, they derive from random biological effects in development, rather than from any kind of social-psychological influence. The problem with these claims is that they constitute a complicated mixture of good science and misleading assertions.

A Novartis Foundation symposium published in 2000 concluded in support of Jensen that g is real and does have a biological basis. It is not, as some critics have asserted, just the result of statistical jiggery-pokery. On the other hand, the experts considered it unlikely that the biological basis would prove to involve a unitary neural function and also that in addition to g there would be important special cognitive skills (as Jensen admits). Most people estimate the heritability of IQ as 50 to 60 per cent, rather than the 75 per cent claimed by Jensen. Nevertheless, it must be said that it does not matter very much which figure is more valid. Either way, genetic influences play a substantial role in individual differences, but so do non-genetic factors.

The claim that there are replicated findings on at least four genes for intelligence is simply wrong. Plomin is upbeat about the potential for molecular genetic research on intelligence, but he has been honest in noting that there are, as yet, no confirmed genes and that, indeed, for reasons that remain puzzling, he has been unable to replicate his own findings. It is also necessary to note that the heritability figure (whether 50 per cent or 75 per cent) does not provide a pure measure of genetic effects because it incorporates the effects of gene-environment correlations and interactions. In other words, nature often operates via nurture. Genes and environment are not as separate as they were once thought to be. As for the supposed importance of obstetric influences on IQ (as claimed by Jensen), the evidence is not particularly supportive of Jensen. Miele challenges Jensen on the implications of the Flynn effect - that is, the major rise in IQ that has taken place over time. This would seem to suggest some important environmental effect. Jensen accepts the findings but provides only a very cursory comment on their implications.

Surprisingly, Miele does not challenge Jensen on the finding from Michel Duyme's adoption study that the quality of the adoptive family environment is significantly associated with the extent to which late-adopted children's IQs rise after adoption.

Jensen is very selective in the evidence to which he pays attention, and he rarely even notes studies designed to test for environmental mediation effects. The consequence is a rather one-sided picture; although he is more right than wrong on the biological reality of g, he is more dismissive than he should be of environmental influences.

The situation with respect to Jensen's claims on the genetic basis of the lower average IQ of African-Americans as compared with whites is much more murky. To begin with, he relies on weasel words in noting that he never claimed that the racial difference in IQ was genetic in origin but only that it was a very reasonable hypothesis that genetic factors were implicated, together with possible environmental factors. That is correct and, of course, it is entirely possible that such genetic differences as there are among ethnic groups could play a role in general mental functioning.

However, that is not the main concern, at least not of the more reasonable critics. First, although expressed as a hypothesis, the whole tenor of Jensen's writings over the years leads to the conclusion that he thinks it is much more than that. Second, the quantitative genetic findings on the heritability of IQ largely derive from studies of white people. In 1969, there was virtually no evidence from studies of African-Americans, and the situation is not a lot better now. Third, Jensen's arguments ignore the possibility that the heritability of IQ may be lower in socially disadvantaged populations (evidence from a study by David Rowe provides some indication that this may be the case). Fourth, and most crucially, it is not possible to generalise from findings on the influences on individual variability within populations to the causes of between -group differences in level. The late Jack Tizard used the example of height to illustrate why Jensen's extrapolation was unjustified. He noted that height is one of the most strongly genetically influenced of all human characteristics - with a heritability of about 80 to 90 per cent. Yet, despite this, there has been a massive increase in the average height of the population over the past century - almost certainly due to improvements in nutrition. Most scientists, while accepting the need for the genetic hypothesis on race differences in IQ, are critical of the evidence and of the logic.

That might not matter very much were it not for the social and ethical issues involved. Regarding the social issues, it is Jensen's language that causes offence as much as the content of his claims. Thus, both in 1969 and in this book he refers to races as "breeding populations" and equates them with subspecies. Although he accepts that races do not constitute distinct categories, and although he notes the very high rates of inter-racial matings (giving rise to continua of gene frequencies and patterns), he continues to discuss the issues as if the ethnic groupings were much more distinct than they are. Also, although he accepts the reality of racial discrimination, he pays little attention to its effects.

Miele challenges Jensen on his repeated acceptance of funding from the Pioneer Fund - a charity widely regarded as racist in its aims. Jensen refuses to accept that this involves any ethical issue, arguing that the fund has never censored his findings and that, if there is no censorship, it is acceptable to take grants from any organisation.

It is also relevant to note whom Jensen particularly admires. Among researchers, he picks out Hans Eysenck and Cyril Burt, both of whom were regarded as pretty slippery characters. Having earlier considered that Burt's data were fraudulent, Jensen now recants this view and sees the errors as no more than careless mistakes, which led to his being framed.

That constitutes a very generous view of Burt's honesty.

Finally, there is the topic of Jensen's dismissal of the value of compensatory education. The book makes explicit that Jensen continues to believe that psychological or educational interventions can have little enduring effects on g. He considers that there is a decline in the average IQ of the population.He favours eugenics (albeit by family choice) and genetic engineering together with population control, and he recommends that educational programmes should be specifically tailored to individual abilities (which he sees as likely to be best measured by precise response-time measures of cognitive processes). Certainly, society lacks good answers on how educational performance should be improved, but I doubt that policy-makers will be much impressed by Jensen's musings on policy.

What should be the verdict on Jensen and Jensenism? Readers will need to make up their own minds. Is he the scrupulously rigorous scientist providing only dispassionate evidence, or has he an agenda that he seeks to defend? What should be the verdict on the book? That is easier. It does not try to draw conclusions on the controversies, but it does succeed in outlining many of the key issues in ways that are both interesting and thought provoking.

Sir Michael Rutter is professor of developmental psychopathology, King's College London.