‘The Bell Curve’: A Book Much Read About, But Rarely Actually Read

‘The Bell Curve: Intelligence and Class Structure in American Life’ by Richard Herrnstein and Charles Murray (New York: Free Press, 1994).

There’s no such thing as bad publicity’ – or so contends a famous adage of the marketing industry.

‘The Bell Curve: Intelligence and Class Structure in America’ by Richard Herrnstein and Charles Murray is perhaps a case in point.

This dry, technical, academic social science treatise, full of statistical analyses, graphs, tables, endnotes and appendices, and totalling almost 900 pages, became an unlikely nonfiction bestseller in the mid-1990s on a wave of almost universally bad publicity in which the work was variously denounced as racist, pseudoscientific, fascist, social Darwinist, eugenicist and sometimes even just plain wrong.

Readers who hurried to the local bookstore eagerly anticipating an incendiary racialist polemic were, however, in for a disappointment.

Indeed, one suspects that, along with ‘The Bible’ and Stephen Hawkins’ ‘A Brief History of Time’, ‘The Bell Curve’ became one of those bestsellers that many people bought, but few managed to finish.

‘The Bell Curve’ thus became, like another book that I have recently reviewed, a book much read about, but rarely actually read – at least in full.

As a result, as with that other book, many myths have emerged regarded the content of ‘The Bell Curve’ that are quite contradicted when one actually takes the time and trouble to read it for oneself.

Subject Matter

The first myth of ‘The Bell Curve’ is that it was a book about race differences, or, more specifically, about race differences in intelligence. In fact, however, this is not true.

Thus, ‘The Bell Curve’ is a book so controversial that the controversy begins with the very identification of its subject-matter.

On the one hand, the book’s critics focused almost exclusively on subject of race. This led to the common perception that ‘The Bell Curve’ was a book about race and race differences in intelligence.^[1]

Ironically, many racialists seem to have taken these leftist critics at their word, enthusiastically citing the work as support for their own views regarding race differences in intelligence.

On the other hand, however, surviving co-author Charles Murray insisted from the outset that the issue of race, and race differences in intelligence, was always peripheral to he and co-author Richard Herrnstein’s primary interest and focus, which was, he claimed, on the supposed emergence of a ‘Cognitive Elite’ in modern America.

Actually, however, both these views seem to be incorrect. While the first section of the book does indeed focus on the supposed emergence of a ‘Cognitive Elite’ in modern America, the overall theme of the book seems to be rather broader.

Thus, the second section of the book focuses on the association between intelligence and various perceived social pathologies, such as unemployment, welfare dependency, illegitimacy, crime and single-parenthood.

To the extent the book has a single overarching theme, one might say that it is a book about the social and economic correlates of intelligence, as measured by IQ tests, in modern America.

Its overall conclusion is that intelligence is indeed a strong predictor of social and economic outcomes for modern Americans – high intelligence with socially desirable outcomes and low intelligence with socially undesirable ones.

On the other hand, however, the topic of race is not quite as peripheral to the book’s themes as sometimes implied by Murray and others.

Thus, it is sometimes claimed only a single chapter dealt with race. Actually, however, two chapters focus on race differences, namely chapters 13 and 14, respectively titled ‘Ethnic Differences in Cognitive Ability’ and ‘Ethnic Inequalities in Relation to IQ’.

In addition, a further two chapters, namely chapters 19 and 20, entitled respectively ‘Affirmative Action in Higher Education’ and ‘Affirmative Action in the Workplace’, deal with the topic of affirmative action, as does the final appendix, entitled ‘The Evolution of Affirmative Action in the Workplace’ – and, although affirmative action has been employed to favour women as well as racial minorities, it is with racial preferences that Herrnstein and Murray are primarily concerned.

However, these chapters represent only 142 of the book’s nearly 900 pages.

Moreover, in much of the remainder of the book, the authors actually explicitly restrict their analysis to white Americans exclusively. They do so precisely because the well documented differences between the races in IQ as well as in many of the social outcomes whose correlation with IQ the book discusses would mean that race would have represented a potential confounding factor that they would otherwise have to take steps to control for.

Herrnstein and Murray therefore took to decision to extend their analysis to race differences near the end of their book, in order to address the question of the extent to which differences in intelligence, which they have already demonstrated to be an important correlate of social and economic outcomes among whites, are also capable of explaining differences in achievement as between races.

Without these chapters, the book would have been incomplete, and the authors would have laid themselves open to the charge of political-correctness and of ignoring the ‘elephant in the room’.

Race and Intelligence

If the first controversy of ‘The Bell Curve’ concerns whether it is a book primarily about race and race differences in intelligence, the second controversy is over what exactly the authors concluded with respect to this vexed and contentious issue.

Thus, the same leftist critics who claimed that ‘The Bell Curve’ was primarily a book about race and race differences in intelligence, also accused the authors of concluding that black people are innately less intelligent than whites.

Some racists, as I have already noted, evidently took the leftists at their word, and enthusiastically cite the book as support and authority for this view.

However, in subsequent interviews, Murray always insisted he and Herrnstein had actually remained “resolutely agnostic” on the extent to which genetic factors underlay the IQ gap.

In the text itself, Herrnstein and Murray do indeed declare themselves “resolutely agnostic” with regard to the extent of the genetic contribution to the test score gap (p311).

However, just couple of sentences before they use this very phrase, they also appear to conclude that genes are indeed at least part of the explanation, writing:

“It seems highly likely to us that both genes and the environment have something to do with racial differences [in IQ]” (p311).

This paragraph, buried near the end of chapter 13, during an extended discussion of evidence relating to the causes of race differences in intelligence, is the closest the authors come to actually declaring any definitive conclusion regarding the causes of the black-white test score gap.^[2]

This conclusion, though phrased in sober and restrained terms, is, of course, itself sufficient to place its authors outside the bounds of acceptable opinion in the early-twenty-first century, or indeed in the late-twentieth century when the book was first published, and is sufficient to explain, and, for some, justify, the opprobrium heaped upon the book’s surviving co-author from that day forth.

Intelligence and Social Class

It seems likely that races which evolved on separate continents, in sufficient reproductive isolation from one another to have evolved the obvious (and not so obvious) physiological differences between races that we all observe when we look at the faces, or bodily statures, of people of different races (and that we indirectly observe when we look at the results of different athletic events at the Olympic Games), would also have evolved to differ in psychological traits, including intelligence.

Indeed, it is surely unlikely, on a priori grounds alone, that all different human races have evolved, purely by chance, the exact same level of intelligence.

However, if races differ in intelligence are therefore probable, the case for differences in intelligence as between social classes is positively compelling.

Indeed, on a priori grounds alone, it is inevitable that social classes will come to differ in IQ, if one accepts two premises, namely:

1) Increased intelligence is associated with upward social mobility; and
2) Intelligence is passed down in families.

In other words, if more intelligent people tend, on average, to get higher-paying jobs than those of lower intelligence, and the intelligence of parents is passed on to their offspring, then it is inevitable that the offspring of people with higher-paying jobs will, on average, themselves be of higher intelligence than are the offspring of people with lower paying jobs.

This, of course, follows naturally from the infamous syllogism formulated by ‘Bell Curve’ co-author Richard Herrnstein way back in the 1970s (p10; p105).

Incidentally, this second premise, namely that intelligence is passed down in families, does not depend on the heritability of IQ in the strict biological sense. After all, even if heritability of intelligence were zero, intelligence could still be passed down in families by environmental factors (e.g. the ‘better’ parenting techniques of high IQ parents, or the superior material conditions in wealthy homes).

The existence of an association between social class and IQ ought, then, to be entirely uncontroversial to anyone who takes any time whatsoever to think about the issue.

If there remains any room for reasoned disagreement, it is only over the direction of causation – namely the question of whether:

1) High intelligence causes upward social mobility; or
2) A privileged upbringing causes higher intelligence.

These two processes are, of course, not mutually exclusive. Indeed, it would seem intuitively probable that both factors would be at work.

Interestingly, however, evidence demonstrates the occurrence only of the former.

Thus, even among siblings from the same family, the sibling with the higher childhood IQ will, on average, achieve higher socioeconomic status as an adult. Likewise, the socioeconomic status a person achieves as an adult correlates more strongly with their own IQ score than it does with the socioeconomic status of their parents or of the household they grew up in (see Straight Talk About Mental Tests: p195).

In contrast, family, twin and adoption studies and of the sort conducted by behavioural geneticists have concurred in suggesting that the so-called shared family environment (i.e. those aspects of the family environment shared by siblings from the same household, including social class) has but little effect on adult IQ.

In other words, children raised in the same home, whether full- or half-siblings or adoptees, are, by the time they reach adulthood, no more similar to one another in IQ than are children of the same degree of biological relatedness brought up in entirely different family homes (see The Nurture Assumption: reviewed here).

However, while the direction of causation may still be disputed by intelligent (if uninformed) laypeople, the existence of an association between intelligence and social class ought not, one might think, be in dispute.

However, in Britain today, in discussions of social mobility, if children from deprived backgrounds are underrepresented, say, at elite universities, then this is almost invariably taken as incontrovertible proof that the system is rigged against them. The fact that children from different socio-economic backgrounds differ in intelligence is almost invariably ignored.

When mention is made of this incontrovertible fact, leftist hysteria typically ensues. Thus, in 2008, psychiatrist Bruce Charlton rightly observed that, in discussion of social mobility:

“A simple fact has been missed: higher social classes have a significantly higher average IQ than lower social classes” (Clark 2008).

For his trouble, Charlton found himself condemned by the National Union of Students and assorted rent-a-quote academics and professional damned fools, while even the ostensibly ‘right-wing’ Daily Mail newspaper saw fit to publish a headline “Higher social classes have significantly HIGHER IQs than working class, claims academic”, as if this were in some way a controversial or contentious claim (Clark 2008).

Meanwhile, when, in the same year, a professor at University College a similar point with regard the admission of working-class students to medical schools, even the then government Health Minister, Ben Bradshaw, saw fit to offer his two cents worth (which were not worth even that), declaring:

“It is extraordinary to equate intellectual ability with social class” (Beckford 2008).

Actually, however, what is truly extraordinary is that any intelligent person, least of all a government minister, would dispute the existence of such a link.

Cognitive Stratification

Herrnstein’s syllogism leads to a related paradox – namely that, as environmental conditions are equalized, heritability increases.

Thus, as large differences in the sorts of environmental factors known to affect IQ (e.g. malnutrition) are eliminated, so differences in income have come to increasingly reflect differences in innate ability.

Moreover, the more gifted children from deprived backgrounds who escape their humble origins, then, given the substantial heritability of IQ, the fewer such children will remain among the working-class in subsequent generations.

The result is what Herrnstein and Murray call the ‘Cognitive Stratification’ of society and the emergence of what they call a ‘Cognitive Elite’.

Thus, in feudal society, a man’s social status was determined largely by ‘accident of birth’ (i.e. he inherited the social station of his father).

Women’s status, meanwhile, was determined, in addition, by what we might call ‘accident of marriage’ – and, to a large extent, it still is.

However, today, a person’s social status, at least according to Herrnstein and Murray, is determined primarily, and increasingly, by their level of intelligence.

Of course, people are not allocated to a particular social class by IQ testing itself. Indeed, the use of IQ tests by employers and educators has been largely outlawed on account of its ‘disparate impact‘ (or ‘indirect discrimination’, to use the equivalent British phrase) with regard to race (see below).

However, the skills and abilities increasingly valued at a premium in western society (and, increasingly, many non-western societies as well), mean that, through the operation of the education system and labour market, individuals are effectively sorted by IQ, even without anyone ever actually sitting an IQ test.

In other words, society is becoming increasingly meritocratic – and the form of ostensible ‘merit’ upon which attainment is based is intelligence.

For Herrnstein and Murray, this is a mixed blessing:

“That the brightest are identified has its benefits. That they become so isolated and inbred has its costs” (p25).

However, the correlation between socioeconomic status and intelligence remains imperfect.

For one thing, there are still a few highly remunerated, and very high-status, occupations that rely on skills that are not especially, if at all, related to intelligence. I think here, in particular, of professional sports and the entertainment industry. Thus, leadings actors, pop stars and sports stars are sometimes extremely well-remunerated, and very high-status, but may not be especially intelligent.

More importantly, while highly intelligent people might be, by very definition, the only ones capable of performing cognitively-demanding, and hence highly remunerated, occupations, this is not to say all highly intelligent people are necessarily employed in such occupations.

Thus, whereas all people employed in cognitively-demanding occupations are, almost by definition, of high intelligence, people of all intelligence levels are capable of doing cognitively-undemanding jobs.

Thus, a few people of high intellectual ability remain in low-paid work, whether on account of personality factors (e.g. laziness), mental illness, lack of opportunity or sometimes even by choice (which choice is, of course, itself a reflection of personality factors).

Therefore, the correlation between IQ and occupation is far from perfect.

Job Performance

The sorting of people with respect to their intelligence begins in the education system. However, it continues in the workplace.

Thus, general intelligence, as measured by IQ testing, is, the authors claim, the strongest predictor of occupational performance in virtually every occupation. Moreover, in general, the higher paid and higher status the occupation in question, the stronger the correlation between performance and IQ.

However, Herrnstein and Murray are at pains to emphasize, intelligence is a strong predictor of occupational performance even in apparently cognitively undemanding occupations, and indeed almost always a better predictor of performance than tests of the specific abilities the job involves on a daily basis.

However, in the USA, employers are barred from using testing to select among candidates for a job or for promotion unless they can show the test has ‘manifest relationship’ to the work, and the burden of proof is on the employer to show such a relationship. Otherwise, given their ‘disparate impact’ with regard to race (i.e. the fact that some groups perform worse), the tests in question are deemed indirectly discriminatory and hence unlawful.

Therefore, employers are compelled to test, not general ability, but rather the specific skills required in the job in question, where a ‘manifest relationship’ is easier to demonstrate in court.

However, since even tests of specific abilities almost invariably still tap into the general factor of intelligence, races inevitably score differently even on these tests.

Indeed, because of the ubiquity and predictive power of the g factor, it is almost impossible to design any type of standardized test, whether of specific or general ability or knowledge, in which different racial groups do not perform differently.

However, if some groups outperform others, the American legal system presumes a priori that this reflects test bias rather than differences in ability.

Therefore, although the words “all men are created equal” are not, contrary to popular opinion, part of the US constitution, the Supreme Court has effectively decided, by legal fiat, to decide cases as if they were.

However, just as a law passed by Congress cannot repeal the law of gravity, so a legal presumption that groups are equal in ability cannot make it so.

Thus, the bar on the use of IQ testing by employers has not prevented society in general from being increasingly stratified by intelligence, the precise thing measured by the outlawed tests.

Nevertheless, Herrnstein and Murray estimate that the effective bar on the use of IQ testing makes this process less efficient, and cost the economy somewhere between 80 billion to 13 billion dollars in 1980 alone (p85).

Conscientiousness and Career Success

I am skeptical of Herrnstein and Murray’s conclusion that IQ is the best predictor of academic and career success. I suspect hard work, not to mention a willingness to toady, toe the line, and obey orders, is at least as important in even the most cognitively-demanding careers, as well as in schoolwork and academic advancement.

Perhaps the reason these factors have not (yet) been found to be as highly correlated with earnings as is IQ is that we have not yet developed a way of measuring these aspects of personality as accurately as we can measure a person’s intelligence through an IQ test.

For example, the closest psychometricians have come to measuring capacity for hard work is the personality factor known as ‘conscientiousness’, one of the Big Five factors of personality revealed by psychometric testing.

Conscientiousness does indeed correlate with success in education and work (e.g. Barrick & Mount 1991). However, the correlation is weaker than that between IQ and success in education and at work.

However, this may be because personality is less easily measured by current psychometric methods than is intelligence – not least because personality tests generally rely on self-report, rather than measuring actual behaviour.

Thus, to assess conscientiousness, questionnaires ask respondents whether they ‘see themselves as organized’, ‘as able to follow an objective through to completion’, ‘as a reliable worker’, etc.

This would be the equivalent of an IQ test that, instead of directly testing a person’s ability to recognize patterns or manipulate shapes by having them do just this, simply asked respondents how good they perceived themselves as being at recognizing patterns, or manipulating shapes.

Obviously, this would be a less accurate measure of intelligence than a normal IQ test. After all, some people lie, some are falsely modest and some are genuinely deluded.

Indeed, according to the Dunning Kruger effect, it is those most lacking in ability who most overestimate their abilities – precisely because they lack the ability to accurately assess their ability (Kruger & Dunning 1999).

In an IQ test, on the other hand, one can sometimes pretend to be dumber than one is, by deliberately getting questions wrong that one knows the answer to.^[3]

However, it is not usually possible to pretend to be smarter than one is by getting more questions right simply because one would not know what are the right answers.

‘Affirmative Action’ and Test Bias

In chapters nineteen and twenty, respectively entitled ‘Affirmative Action in Higher Education’ and ‘Affirmative Action in the Workplace’, the authors discuss so-called ‘affirmative action’, an American euphemism for systematic and overt discrimination against white males.

It is well-documented that, in the United States, blacks, on average, earn less than white Americans. On the other hand, it is less well-documented that whites, on average, earn less than people of Indian, Chinese and Jewish ancestry.

With the possible exception of Indian-Americans, these differences, of course, broadly mirror those in average IQ scores.

Indeed, according to Herrnstein and Murray, the difference in earnings between whites and blacks, not only disappears after controlling for differences in IQ, but is actually partially reversed. Thus, blacks are actually somewhat overrepresented in professional and white-collar occupations as compared to whites of equivalent IQ.

This remarkable finding Herrnstein and Murray attribute to the effects of affirmative action programmes, as black Americans are appointed and promoted beyond what their ability merits because through discrimination.

Interestingly, however, this contradicts what the authors wrote in an earlier chapter, where they addressed the question of test bias (pp280-286).

There, they concluded that testing was not biased against African-Americans, because, among other reasons, IQ tests were equally predictive of real-world outcomes (e.g. in education and employment) for both blacks and whites, and blacks do not perform any better in the workplace or in education than their IQ scores predict.

This is, one might argue, not wholly convincing evidence that IQ tests are not biased against blacks. It might simply suggest that society at large, including the education system and the workplace, is just as biased against blacks as are the hated IQ tests. This is, of course, precisely what we are often told by the television, media and political commentators who insist that America is a racist society, in which such mysterious forces as ‘systemic racism’ and ‘white privilege’ are pervasive.

In fact, the authors acknowledge this objection, conceding:

“The tests may be biased against disadvantaged groups, but the traces of bias are invisible because the bias permeates all areas of the group’s performance. Accordingly, it would be as useless to look for evidence of test bias as it would be for Einstein’s imaginary person traveling near the speed of light to try to determine whether time has slowed. Einstein’s traveler has no clock that exists independent of his space-time context. In assessing test bias, we would have no test or criterion measure that exists independent of this culture and its history. This form of bias would pervade everything” (p285).

Herrnstein and Murray ultimately reject this conclusion on the grounds that it is simply implausible to assume that:

“[So] many of the performance yardsticks in the society at large are not only biased, they are all so similar in the degree to which they distort the truth-in every occupation, every type of educational institution, every achievement measure, every performance measure-that no differential distortion is picked up by the data” (p285).

In fact, however, Nicholas Mackintosh identifies one area where IQ tests do indeed under-predict black performance, namely with regard to so-called adaptive behaviours – i.e. the ability to cope with day-to-day life (e.g. feed, dress, clean, interact with others in a ‘normal’ manner).

Blacks with low IQs are generally much more functional in these respects than whites or Asians with equivalent low IQs (see IQ and Human Intelligence: p356-7).^[4]

Yet Herrnstein and Murray seem to have inadvertently, and evidently without realizing it, identified yet another sphere where standardized testing does indeed under-predict real-world outcomes for blacks.

Thus, if indeed, as Herrnstein and Murray claim, blacks are somewhat overrepresented among professional and white-collar occupations relative to their IQs, this suggests that blacks do indeed do better in real-world outcomes than their test results would predict and, while Herrnstein and Murray attribute this to the effect of discrimination against whites, it could instead surely be interpreted as evidence that the tests are biased against blacks.

Policy Implications?

What, then, are the policy implications that Herrnstein and Murray draw from the findings that they report?

In The Blank Slate: The Modern Denial of Human Nature, cognitive science, linguist and popular science writer Steven Pinker popularizes the notion that recognizing the existence of innate differences between individuals and groups in traits such as intelligence does not necessarily lead to ‘right-wing’ political implications.

Thus, a leftist might accept the existence of innate differences in ability, but conclude that, far from justifying inequality, this is all the more reason to compensate the, if you like, ‘cognitively disadvantaged’ for their innate deficiencies, differences which are, being innate, hardly something for which they can legitimately be blamed.

Herrnstein and Murray reject this conclusion, but acknowledge it is compatible with their data. Thus, in an afterword to later editions, Murray writes:

“If intelligence plays an important role in determining how well one does in life, and intelligence is conferred on a person through a combination of genetic and environmental factors over which that person has no control, the most obvious political implication is that we need a Rawlsian egalitarian state, compensating the less advantaged for the unfair allocation of intellectual gifts” (p554).^[5]

Interestingly, Pinker’s notion of a ‘hereditarian left’, and the related concept of ‘Bell Curve liberals’, is not entirely imaginary. On the contrary, it used to be quite mainstream.

Thus, it was the radical leftist post-war Labour government that imposed the tripartite system on schools in the UK in 1945, which involved allocating pupils to different schools on the basis of their performance in what was then called the 11-plus exam, conducted at with children at age eleven, which tested both ability and acquired knowledge. This was thought by leftists to be a fair system that would enable bright, able youngsters from deprived and disadvantaged working-class backgrounds to achieve their full potential.^[6]

Indeed, while contemporary ‘Cultural Marxists’ emphatically deny the existence of innate differences in ability as between individuals and groups, Marx himself, laboured under no such delusion.

On the contrary, in advocating, in his famous (plagiarized) aphorism “From each according to his ability; to each according to his need”, Marx implicitly recognized that individuals differ in “ability”, and, given that, in the unrealistic communist utopia he envisaged, environmental conditions were ostensibly to be equalized, these differences he presumably conceived of as innate in origin.

However, a distinction must be made here. While it is possible to justify economic redistributive policies on Rawlsian grounds, it is not possible to justify affirmative action.

Thus, one might well reasonably contend that the ‘cognitively disadvantaged’ should be compensated for their innate deficiencies through economic redistribution. Indeed, to some extent, most Western polities already do this, by providing welfare payments and state-funded, or state-subsidized, care to those whose cognitive impairment is such as to qualify as a disability and hence render them incapable of looking after or providing for themselves.

However, we are unlikely to believe that such persons should be given entry to medical school such that they are one day liable to be responsible for performing heart surgery on us or diagnosing our medical conditions.

In short, socialist redistribution is defensible – but affirmative action is definitely not!

Reception and Readability

The reception accorded ‘The Bell Curve’ in 1994 echoed that accorded another book that I have also recently reviewed, but that was published some two decades earlier, namely Edward O. Wilson’s Sociobiology: The New Synthesis.

Both were greeted with similar indignant moralistic outrage by many social scientists, who even employed similar pejorative soundbites (‘genetic determinism’, ‘reductionism’, ‘biology as destiny’), in condemning the two books. Moreover, in both cases, the academic uproar even spilled over into a mainstream media ‘moral panic’, with pieces appearing the popular press attacking the two books.

Yet, in both cases, the controversy focused almost exclusively on just a small part of each book – the single chapter in ‘Sociobiology: The New Synthesis’ focusing on humans and the few chapters in ‘The Bell Curve’ discussing race.

In truth, however, both books were massive tomes of which these sections represented only a small part.

Indeed, due to their size, one suspects most critics never actually read the books in full for themselves, including, it seemed, most of those nevertheless taking it upon themselves to write critiques. This is what led to the massive disconnect between what most people thought the books said, and their actual content.

However, there is a crucial difference.

‘Sociobiology: The New Synthesis’ was a long book of necessity, given the scale of the project Wilson set himself.

As I have written in my review of that latter work, the scale of Wilson’s ambition can hardly be exaggerated. He sought to provide a new foundation for the whole field of animal behaviour, then, almost as an afterthought, sought to extend this ‘New Synthesis’ to human behaviour as well, which meant providing a new foundation, not for a single subfield within biology, but for several whole disciplines (psychology, sociology, economics and cultural anthropology) that were formerly almost unconnected to biology. Then, in a few provocative sentences, he even sought to provide a new foundation for moral philosophy, and perhaps epistemology too.

‘Sociobiology: The New Synthesis’ was, then, inevitably and of necessity, a long book. Indeed, given that his musings regarding the human species were largely (but not wholly) restricted to a single chapter, one could even make a case that it was too short – and it is no accident that Wilson subsequently extended his writings with regard to the human species to a book length manuscript.

Yet, while Sociobiology was of necessity a long book, ‘The Bell Curve: Intelligence and Class Structure in America’ is, for me, unnecessarily overlong.

After all, Herrnstein and Murray’s thesis was actually quite simple – namely that cognitive ability, as captured by IQ testing, is a major correlate of many important social outcomes in modern America.

Yet they reiterate this point, for different social outcomes, again and again, chapter after chapter, repeatedly.

In my view, Herrnstein and Murray’s conclusion would have been more effectively transmitted to the audience they presumably sought to reach had they been more succinct in their writing style and presentation of their data.

Had that been the case then perhaps rather more of the many people who bought the book, and helped make it into an unlikely nonfiction bestseller in 1994, might actually have managed to read it – and perhaps even been persuaded by its thesis.

For casual readers interested in this topic, I would recommend instead Intelligence, Race, And Genetics: Conversations With Arthur R. Jensen (which I have reviewed here, here and here).

Endnotes

[1] For example, Francis Wheen, a professional damned fool and columnist for the Guardian newspaper (which two occupations seem to be largely interchangeable) claimed that:

“The Bell Curve (1994), runs to more than 800 pages but can be summarised in a few sentences. Black people are more stupid than white people: always have been, always will be. This is why they have less economic and social success. Since the fault lies in their genes, they are doomed to be at the bottom of the heap now and forever” (Wheen 2000).

In making this claim, Wheen clearly demonstrates that he has read few if any of those 800 pages to which he refers.

[2] Although their discussion of the evidence relating to the causes, genetic or environmental, of the black-white test score gap is extensive, it is not exhaustive. For example, Phillipe Rushton, the author of Race Evolution and Behavior (reviewed here and here) argues that, despite the controversy their book provoked, Herrnstein and Murray actually “didn’t go far enough on race”, omitting, for example, any real discussion, save a passing mention in Appendix 5, of race differences in brain size (Rushton 1997). On the other hand, Herrnstein and Murray also did not mention studies that failed to establish any correlation between IQ and blood groups among African-Americans, studies interpreted as supporting an environmentalist interpretation of race differences in intelligence (Loehlin et al 1973; Scarr et al 1977). For readers interested in a more complete discussion of the evidence regarding the relative contributions of environment and heredity to the differences in IQ scores of different races, see my review of Richard Lynn’s Race Differences in Intelligence: An Evolutionary Analysis, available here.

[3] For example, some of those accused of serious crimes have been accused of deliberately getting questions wrong on IQ tests in order to qualify as mentally subnormal when before the courts for sentencing in order to be granted mitigation of sentence on this ground, or, more specifically, in order to evade the death penalty.

[4] This may be because whites or Asians with such low IQs are more likely to have such impaired cognitive abilities because of underlying conditions (e.g chromosomal abnormalities, brain damage) that handicap them over and above the deficit reflected in IQ score alone. On the other hand, blacks with similarly low IQs are still within the normal range for their own race. Therefore, rather than suffering from, say, a chromosomal abnormality or brain damage, they are relatively more likely to simply be at the tail-end of the normal range of IQs within their group, and hence normal in other respects.

[5] The term Rawlsian is a reference to political theorist John Rawles version of social contract theory, whereby he poses the hypothetical question as to what arrangement of political, social and economic affairs humans would favour if placed in what he called ‘the original position’, where they would be unaware of, not only their own race, sex and position in to the socio-economic hierarchy, but also, most important for our purposes, their own level of innate ability. This Rawles referred to as ‘veil of ignorance’.

[6] The tripartite system did indeed enable many working-class children to achieve a much higher economic status than their parents, although this was partly due to the expansion of the middle-class sector of the economy over the same time-period. It was also later Labour administrations who largely abolished the 11-plus system, not least because, unsurprisingly given the heritability of intelligence and personality, children from middle-class backgrounds tended to do better on it than did children from working-class backgrounds.

References

Barrick & Mount 1991 The big five personality dimensions and job performance: a meta-analysis. Personnel Psychology 44(1):1–26
Beckford (2008) Working classes ‘lack intelligence to be doctors’, claims academic, Daily Telegraph, 04 Jun 2008.
Clark 2008 Higher social classes have significantly HIGHER IQs than working class, claims academic Daily Mail, 22 May 2008.
Kruger & Dunning (1999) Unskilled and Unaware of It: How Difficulties in Recognizing One’s Own Incompetence Lead to Inflated Self-Assessments. Journal of Personality and Social Psychology 77(6):1121-34
Loehlin et al (1973) Blood group genes and negro-white ability differences. Behavior Genetics 3(3): 263-270
Rushton, J. P. (1997). Why The Bell Curve didn’t go far enough on race. In E. White (Ed.), Intelligence, political inequality, and public policy (pp. 119-140). Westport, CT: Praeger.
Scarr et al (1977) Absence of a relationship between degree of white ancestry and intellectual skills within a black population. Human Genetics 39(1):69-86 .
Wheen (2000) The ‘science’ behind racism. Guardian, 10 May 2000.

Judith Harris’s ‘The Nurture Assumption’: By Parent or Peers

Judith Harris, The Nurture Assumption: Why Children Turn Out the Way They Do. Free Press, 1998.

Almost all psychological traits on which individual humans differ, from personality and intelligence to mental illness, are now known to be substantially heritable. In other words, individual differences in these traits are, at least in part, a consequence of genetic differences between individuals.

This finding is so robust that it has even been termed by Eric Turkenheimer the ‘First Law of Behviour Genetics’ and, although once anathema to most psychologists save a marginal fringe of behavioural geneticists, it has now, under the sheer weight of evidence produced by the latter, belatedly become the new orthodoxy.

On reflection, however, this transformation is not entirely a revelation.

After all, it was only in the mid-twentieth century that the curious notion that individual differences were entirely the product of environmental differences first arose, and, even then, this delusion was largely restricted to psychologists, sociologists, feminists and other such ‘professional damned fools’, along with those among the semi-educated public who seek to cultivate an air of intellectualism by aping the former’s affections.

Before then, poets, peasants and laypeople alike had long recognized that ability, insanity, temperament and personality all tended to run in families, just as physical traits like stature, complexion, hair and eye colour also do.^[1]

However, while the discovery of a heritable component to character and ability merely confirms the conventional wisdom of an earlier age, another behavioural genetic finding, far more surprising and counterintuitive, has passed relatively unreported.

This is the discovery that the so-called shared family environment (i.e. the environment shared by siblings, or non-siblings, raised in the same family home) actually has next to no effect on adult personality and behaviour.

This we know from such classic study designs in behavioural genetics as twin studies, adoption studies and family studies.

In short, individuals of a given degree of relatedness, whether identical twins, fraternal twins, siblings, half-siblings or unrelated adoptees, are, by the time they reach adulthood, no more similar to one another in personality or IQ when they are raised in the same household than when they are raised in entirely different households.

The Myth of Parental Influence

Yet parental influence has long loomed large in virtually every psychological theory of child development, from the Freudian Oedipus complex and Bowby’s attachment theory to the whole literary genre of books aimed at instructing anxious parents on how best to raise their children so as to ensure that the latter develop into healthy, functional, successful adults.

Indeed, not only is the conventional wisdom among psychologists overturned, but so is the conventional wisdom among sociologists – for one aspect of the shared family environment is, of course, household income and social class.

Thus, if the family that a person is brought up in has next to no impact on their psychological outcomes as an adult, then this means that the socioeconomic status of the family home in which they are raised also has no effect.

Poverty, or a deprived upbringing, then, has no effect on IQ, personality or the prevalence of mental illness, at least by the time a person has reached adulthood.^[2]

Neither is it only leftist sociologists who have proved mistaken.

Thus, just as leftists use economic deprivation as an indiscriminate, catch-all excuse for all manner of social pathology (e.g. crime, unemployment, educational underperformance) so conservatives are apt to place the blame on divorce, family breakdown, having children out of wedlock and the consequential increase in the prevalence of single-parent households.

However, all these factors are, once again, part of the shared family environment – and according to the findings of behavioural genetics, they have next to no influence on adult personality or intelligence.

Of course, chaotic or abusive family environments do indeed tend to produce offspring with negative life outcomes.

However, none of this proves that it was the chaotic or abusive family environment that caused the negative outcomes.

Rather, another explanation is at hand – perhaps the offspring simply biologically inherit the personality traits of their parents, the very personality traits that caused their family environment to be so chaotic and abusive in the first place.^[3]

For example, parents who divorce or bear offspring out-of-wedlock likely differ in personality from those who first get married then stick together, perhaps being more impulsive or less self-disciplined and conscientious (e.g. less able refrain from having children from a relationship that was destined to be fleeting, or less able to persevere and make the relationship last).

Their offspring may, then, simply biologically inherit these undesirable personality attributes, which then themselves lead to the negative social outcomes associated with being raised in single-parent households or broken homes. The association between family breakdown and negative outcomes for offspring might, then, reflect simply the biological inheritance of personality.

Similarly, as leftists are fond of reminding us, children from economically-deprived backgrounds do indeed have lower recorded IQs and educational attainment than those from more privileged family backgrounds, as well as other negative outcomes as adults (e.g. lower earnings, higher rates of unemployment).

However, this does not prove that coming from a deprived family background necessarily itself depresses your IQ, educational attainment or future salary.

Rather, an equally plausible possibility is simply that offspring simply biologically inherit the low intelligence of their parents – the very low intelligence which was likely a factor causing the low socioeconomic status of their parents, since intelligence is known to correlate strongly with educational and occupational advancement.^[4]

In short, the problem with all of this body of research which purports to demonstrate the influence of parents and family background on psychology and behavioural outcomes for offspring is that they fail to control for the heritability of personality and intelligence, an obvious confounding factor.

The Non-Shared Environment

However, not everything is explained by heredity. As a crude but broadly accurate generalization, only about half the variation for most psychological traits is attributable to genes. This leaves about half of the variation in intelligence, personality and mental illness to be explained environmental factors.

What are these environmental factors if they are not to be sought in the shared family environment?

The obvious answer is, of course, the non-shared family environment – i.e. the ways in which even children brought up in the same family-home nevertheless experience different micro-environments, both within the home and, perhaps more importantly, outside it.

Thus, even the fairest and most even-handed parents inevitably treat their different offspring differently in some ways.

Indeed, among the principal reasons why parents treat their different offspring differently is precisely because the different offspring themselves differ in their own behaviour quite independently of any parental treatment.

This is well illustrated by the question of the relationship between corporal punishment and behaviour in children.

Corporal punishment

Rather than differences in the behaviour of different children resulting from differences in how their parents treat them, it may be that differences in how parents treat their children may reflect responses to differences in the behaviour of the children themselves.

In other words, the psychologists have the direction of causation precisely backwards.

Take, for example, one particularly controversial issue, namely the physical chastisement of children by their parents as a punishment for bad behaviour (e.g. spanking).

Some psychologists have sometimes argued that physical chastisement actually causes misbehaviour.

As evidence, they cite the fact that children who are spanked more often by their parents or caregivers on average actually behave worse than those whose caregivers only rarely or never spank the children entrusted to their care.

This, they claim, is because, in employing spanking as a form of discipline, caregivers are inadvertently imparting the message that violence is a good way of solving your problems.

Actually, however, I suspect children are more than capable of working out for themselves that violence is often an effective means of getting your way, at least if you have superior physical strength to your adversary. Unfortunately, this is something that, unlike reading, arithmetic and long division, does not require explicit instruction by teachers or parents.

Instead, a more obvious explanation for the correlation between spanking and misbehaviour in children is not that spanking causes misbehaviour, but rather that misbehaviour causes spanking.

Indeed, once you think about it, this is in fact rather obvious: If a child never seriously misbehaves, then a parent likely never has any reason to spank that child, even if the parent is, in principle, a strict disciplinarian; whereas, on the other hand, a highly disobedient child is likely to try the patience of even the most patient caregiver, whatever his or her moral opposition to physical chastisement in principle.

In other words, causation runs in exactly the opposite direction to that assumed by the naïve psychologists.^[5]

Another factor may also be at play – namely, offspring biologically inherit from their parents the personality traits that cause both the misbehaviour and the punishment.

In other words, parents with aggressive personalities may be more likely to lose their temper and physically chastise their children, while children who inherit these aggressive personalities are themselves more likely to misbehave, not least by behaving in an aggressive or violent manner.

However, even if parents treat their different offspring differently owing to the different behaviour of the offspring themselves, this is not the sort of environmental factor capable of explaining the residual non-shared environmental effects on offspring outcomes.

After all, this merely begs the question as to what caused these differences in offspring behaviour in the first place?

If the differences in offspring behaviour exist prior to differences in parental responses to this behaviour, then these differences cannot be explained by the differences in parental responses.

Peer Groups

This brings us back to the question of the environmental causes of offspring outcomes – namely, if about half the differences among children’s IQs and personalities are attributable to environmental factors, but these environmental factors are not to be found in the shared family environment (i.e. the environment shared by children raised in the same household), then where are these environmental factors to be sought?

The search for environmental factors affecting personality and intelligence has, thus far, been largely unsuccessful. Indeed, some behavioural geneticists have almost gone as far as conceding scholarly defeat in identifying correlates for the environmental portion of the variance.

Thus, leading contemporary behavioural geneticist Robert Plomin in his recent book, Blueprint: How DNA Makes Us Who We Are, concludes that those environmental factors that affect cognitive ability, personality, and the development of mental illness are, as he puts it, ‘unsystematic’ in nature.

In other words, he seems to be saying that they are mere random noise. This is tantamount to accepting that the null hypothesis is true.

Judith Harris, however, has a quite different take. According to Harris, environmental causes must be sought, not within the family home, but rather outside it – in a person’s interactions with their peer-group and the wider community.^[6]

Environment ≠ Nurture

Thus, Harris argues that the so-called nature-nurture debate is misnamed, since the word ‘nurture’ usually refers to deliberate care and moulding of a child (or of a plant or animal). But many environmental effects are not deliberate.

Thus, Harris repeatedly references behaviourist John B. Watson’s infamous boast:

“Give me a dozen healthy infants, well-formed, and my own specified world to bring them up in and I’ll guarantee to take any one at random and train him to become any type of specialist I might select—doctor, lawyer, artist, merchant-chief and, yes, even beggar-man and thief, regardless of his talents, penchants, tendencies, abilities, vocations, and race of his ancestors.”

Yet what strikes me as particularly preposterous about Watson’s boast is not its radical environmental determinism, nor even its rather convenient unfalsifiability.^[7]

Rather, what most strikes me as most preposterous about Watson’s claim is its frankly breath-taking arrogance.

Thus, Watson not only insisted that it was environment alone that entirely determined adult personality. In this same quotation, he also proclaimed that he already fully understood the nature of these environmental effects to such an extent that, given omnipotent powers to match his evidently already omniscient understanding of human development, he could produce any outcome he wished.

Yet, in reality, environmental effects are anything but clear-cut. Pushing a child in a certain direction, or into a certain career, may sometimes have the desired effect, but other times may seemingly have the exact opposite effect to that desired, provoking the child to rebel against parental dictates.

Thus, even to the extent that environment does determine outcomes, the precise nature of the environmental factors implicated, and their interaction with one another, and with the child’s innate genetic endowment, is surely far more complex than the simple mechanisms proposed by behaviourists like Watson (e.g. reinforcement and punishment).

Language Acquisition

The most persuasive evidence for Harris’s theory of the importance of peer groups comes from an interesting and widely documented peculiarity of language acquisition.

The children of immigrants, whose parents speak a different language inside the family home, and may even themselves be monolingual, nevertheless typically grow up to speak the language of their host culture rather better than they do the language to which they were first exposed in the family home.

Indeed, while their parents may never achieve fluency in the language of their host culture, having missed out on the Chomskian critical period for language acquisition, their children often actually lose the ability to speak their parent’s language, often much to the consternation of parents and grandparents.

Yet, from an sociobiological or evolutionary psychological perspective, such an outcome is obviously adaptive.

After all, if a child is to succeed in wider society, they must master its language, whereas, if their parent’s first language is not spoken anywhere in their host society except in their family, then it is of limited utility, and, once their parents themselves become proficient in the language of the host culture, it becomes entirely redundant.

As sociologist-turned-sociobiologist Pierre van den Berghe observes in his excellent The Ethnic Phenomenon (reviewed here):

“Children quickly discover that their home language is a restricted medium that not useable in most situations outside the family home. When they discover that their parents are bilingual they conclude – rightly for their purposes – that the home language is entirely redundant… Mastery of the new language entails success at school, at work and in ‘the world’… [against which] the smiling approval of a grandmother is but slender counterweight” (The Ethnic Phenomenon: p258).

Code-Switching

Harris suggests that the same applies to personality. Just as the child of immigrants switches between one language and another at home and school, so they also adopt different personalities.

Thus, many parents are surprised to be told by their children’s teachers at parents’ evenings that their offspring is quiet and well-behaved at school, since, they report, he or she isn’t at all like that at home.

Yet, at home, a child has only, at most, a sibling or two with whom to compete for his parents’ attention. In contrast, at school, he or she has a whole class with whom to compete for their teacher’s attention.

It is therefore unsurprising that most children are less outgoing at school than they are at home with their parents.

For example, an older sibling might be able push his little brother around at home. But, if he is small for his age, he is unlikely to be able to get away with the same behaviour among his peers at school.

Children therefore adopt two quite different personalities – one for interactions with family and siblings, and another for among their peers.

This then, for Harris, explains why, perhaps surprisingly, birth-order has generally been found to have little if any effect on personality, at least as personality manifests itself outside the family home.

An Evolutionary Theory of Socialization?

Interestingly, even evolutionary psychologists have not been immune from the delusion of parental influence. Thus, in one influential paper, anthropologists Patricia Draper and Henry Harpending argued that offspring calibrate their reproductive strategy by reference to the presence or absence of a father in their household (Draper & Harpending 1982).

On this view, being raised in a father-absent household is indicative of a social environment where low male parental investment is the norm, and hence offspring adjust their own reproductive strategy accordingly, adopting a promiscuous, low-investment mating strategy characterized by precocious sexual development and an inability to maintain lasting long-term relationships (Draper & Harpending 1982; Belsky et al 1991).

There is indeed, as these authors amply demonstrate, a consistent correlation between father-absence during development and both earlier sexual development and more frequent partner-switching in later life.

Yet there is also another, arguably more obvious, explanation readily at hand to explain this association. Perhaps offspring simply inherit biologically the personality traits, including sociosexual orientation, of their parents.

On this view, offspring raised in single-parent households are more likely to adopt a promiscuous, low-investment mating strategy simply because they biologically inherit the promiscuous sociosexual orientation of their parents, the very promiscuous sociosexual orientation that caused the latter to have children out-of-wedlock or from relationships that were destined to break down and hence caused the father-absent childhood of their offspring.

Moreover, even on purely a priori theoretical grounds, Draper, Harpending and Belsky’s reasoning is dubious.

After all, whether you personally were raised in a one- or two-parent family is obviously a very unreliable indicator of the sorts of relationships prevalent in the wider community into which you are born, since it represents a sample size of just one.

Instead, therefore, it would be far more reliable to calibrate your reproductive strategy in response to the prevalence of one-parent households in the wider community at large, rather than the particular household type into which you happen to have been born.

This, of course, directly supports Harris’s own theory of ‘peer group socialization’.

In short, to the extent that children do adapt to the environment and circumstances of their upbringing (and they surely do), they must integrate into, adopt the norms of, and a reproductive strategy to maximize their fitness within, the wider community into which they are born, rather than the possibly quite idiosyncratic circumstances and attitudes of their own family.

Absent Fathers, from Upper-Class to Under-Class

Besides language-acquisition among the children of immigrants, another example cited by Harris in support of her theory of ‘peer group socialization’ is the culture, behaviours and upbringing of British upper-class males.

Here, she reports, boys were, and, to some extent, still are, reared primarily, not by their parents, but rather by nannies, governoresses and, more recently, in exclusive fee-paying all-male boarding schools.

Yet, despite having next to no contact with their fathers throughout most of their childhood, these boys nevertheless managed somehow to acquire manners, attitudes and accents similar, if not identical, to those of their upper-class fathers, and not at all those of the middle-class nannies, governoresses and masters with whom they spent most of their childhood being raised.

Yet this phenomenon is by no means restricted to the British upper-classes.

On the contrary, rather than citing the example of the British upper-classes in centuries gone by, Harris might just as well have cited that of contemporary underclass in Britain and America, since what was once true of the British upper-classes, is now equally true of the underclass.

Just as the British upper-classes were once raised by governoresses, nannies and in private schools with next to no contact with their fathers, so contemporary underclass males are similarly raised in single-parent households, often to unwed mothers, and typically have little if any contact with their biological fathers.

Here, as Warren Farrell observes in his seminal The Myth of Male Power (which I have reviewed here, here and here), there is a now a “a new nuclear family: woman, government and child”, what Farrell terms “Government as a Substitute Husband”.

Yet, once again, these underclass males, raised by single parents with the financial assistance of the taxpayer, typically turn out much like their absent fathers with whom they have had little if any contact, often going on to promiscuously father a succession of offspring themselves, with whom they likewise have next to no contact.

Abuse

But what of actual abuse? Surely this has a long-term devastating psychological impact on children. This, at any rate, is the conventional wisdom, and questioning this wisdom, at least with respect to sexual abuse, is tantamount to contemporary heresy, with attendant persecution.

Thus, for example, it is claimed that criminals who are abusive towards their children were themselves almost invariably abused, mistreated or neglected as children, which is what has led to their own abusive, behaviour.

A particularly eloquent expression of this theory is found in the novel Clockers, by Richard Price, where one of the lead characters, a police officer, explains how, during his first few years on the job, a senior colleague had restrained him from attacking an abusive mother who had left her infant son handcuffed to a radiator, telling him:

“Rocco, that lady you were gonna brain? Twenty years ago when she was a little girl. I arrested her father for beating her baby brother to death. The father was a piece of shit. Now that she’s all grown up? She’s a real piece of shit. That kid you saved today. If he lives that long, if he grows up? He’s gonna be a real piece of shit. It’s the cycle of shit and you can’t do nothing about it” (Clockers: p96).

Take, for example, what is perhaps the form of child abuse that provokes the most outrage and disgust – namely, sexual abuse. Here, it is frequently asserted that paedophiles were almost invariably themselves abused as children, which creates a so-called ‘cycle of abuse’.

However, there are at least three problems with this claim.

First, it cannot explain how the first person in this cycle came to be abusive.

Second, we might doubt whether it is really true that paedophiles are disproportionately likely to have themselves been abused as children. After all, abuse is something that almost invariably happens surreptitiously ‘behind closed doors’ and is therefore difficult to verify or disprove.

Therefore, even if most paedophiles claim to have been victims of abuse, it is possible that they are simply lying in order to elicit sympathy or excuse or shift culpability for their own offending.

Finally, and most importantly for present purposes, even if paedophiles can be shown to be disproportionately likely to have themselves been victimized as children, this by no means proves that their past victimization caused their current sexual orientation.

Rather, since most abuse is perpetrated by parents or other close family members, an alternative possibility is that victims simply biologically inherit the sexual orientation of their abuser.

After all, if homosexuality is partially heritable, as is now widely accepted, then why not paedophilia as well?

In short, the ‘cycle of shit’ referred to by Price’s fictional police officer may well be real, but mediated by genetics rather than childhood experience.

However, this conclusion is not entirely clear. On the contrary, Harris is at pains to emphasize that the finding that the shared family environment accounts for hardly any of the variance in outcomes among adults does not preclude the possibility that severe abuse may indeed have an adverse effect on adult outcomes.

After all, adoption studies can only tell us what percent of the variance is caused by heredity or by shared or unshared environments within a specific population as a whole.

Perhaps the shared family environment accounts for so little of the variance precisely because the sort of severe abuse that does indeed have a devastating long-term effect on personality and mental health is, thankfully, so very rare in modern societies.

Indeed, it may be especially rare within the families sampled in adoption studies precisely because adoptive families are carefully screened for suitability before being allowed to adopt.

Moreover, Harris emphasizes an important caveat: Even if abuse does not have long-term adverse psychological effects, this does not mean that abuse causes no harm, and nor does it in any way excuse such abuse.

On the contrary, the primary reason we shouldn’t mistreat children (and should severely punish those who do) is not on account of some putative long-term psychological effect on the adults whom the children subsequently become, but rather because of the very real pain and suffering inflicted on a child at the time the abuse takes place.

Race Differences in IQ

Finally, Harris even touches upon that most vexed area of the (so-called) nature-nurture debate – race differences in intelligence.

Here, the politically-correct claim that differences in intelligence between racial groups, as recorded in IQ tests, are of purely environmental origin runs into a problem, since the sorts of environmental effects that are usually posited by environmental determinists as accounting for the black-white test score gap in America (e.g. differences in rates of poverty and socioeconomic status) have been shown to be inadequate because, even after controlling for these factors, there remains a still unaccounted for gap in test-scores.^[8]

Thus, as Arthur R. Jensen laments:

“This gives rise to the hypothesizing of still other, more subtle environmental factors that either have not been or cannot be measured—a history of slavery, social oppression, and racial discrimination, white racism, the ‘black experience,’ and minority status consciousness [etc]” (Straight Talk About Mental Tests: p223).

The problem with these explanations, however, is that none of these factors has yet been demonstrated to have any effect on IQ scores.

Moreover, some of the factors proposed as explanations are formulated in such a vague form (e.g. “white racism, the ‘black experience’”) that it is difficult to conceive of how they could ever be subjected to controlled testing in the first place.^[9]

Jensen has termed this mysterious factor the ‘X-factor’.

In coining this term, Jensen was emphasizing its vague, mysterious and unfalsifiable nature. Jensen did not actually believe that this posited ‘X-factor’, whatever it was, really did account for the test-score gap. Rather, he thought heredity explained most, if not all, of the remaining unexplained test-score gap.

However, Harris takes Jensen at his word and takes the search for the ‘X-factor’ very seriously. Indeed, she apparently believes she has discovered and identified it. Thus, she announces:

“I believe I know what this X factor is… I can describe it quite clearly. Black kids and white kids identify with different groups that have different norms. The differences are exaggerated by group contrast effects and have consequences that compound themselves over the years. That’s the X factor” (p248-9).

Unfortunately, Harris does not really develop this fascinating claim. Indeed, she cites no direct evidence in support of this claim, and evidently seems to regard the alternative possibility – namely, that race differences in intelligence are at least partly genetic in origin – as so unpalatable that it can safely ruled out a priori.

In fact, however, although not discussed by Harris, there is at least some evidence in support of her theory. Indeed, her theory potentially reconciles the apparently conflicting findings of two of the most widely-cited studies in this vexed area of research and debate.

First, in the more recent of these two studies, Minnesota Transracial Adoption Study, the same race differences in IQ were observed among black, white and mixed-race children adopted into upper-middle class white families as are found among black, white and mixed-race populations in the community at large (Scarr & Weinberg 1976).

Moreover, although, when tested during childhood, the children’s adoptive households did seem to have had a positive effect on their IQ scores, in a follow-up study it was found that by the time they reached the cusp of adulthood, the black teenagers who had been adopted into upper-middle-class white homes actually scored no higher in IQ than did blacks in the wider population not raised in upper-middle class white families (Weinberg, Scarr & Waldman 1992).

Although Scarr, Weinberg and Waldman took pains to present their findings as compatible with a purely environmentalist theory of race differences, this study has, not unreasonably, been widely cited by hereditarians as evidence for the existence of innate racial differences in intelligence (e.g. Levin 1994; Lynn 1994; Whitney 1996).

However, in the light of the findings of the behavioural genetics studies discussed by Harris in ‘The Nurture Assumption’, the fact that white upper-middle-class adoptive homes had no effect on the adult IQs of the black children adopted into them is, in fact, hardly surprising.

After all, as we have seen, the shared family environment generally has no effect on IQ, at least by the time the person being tested has reached adulthood.^[10]

One would therefore not expect adoptive homes, howsoever white and upper-middle-class, to have any effect on adult IQs of the black children adopted into them, or indeed of the white or mixed-race children adopted into them.

In short, adoptive homes have no effect on adult IQ, whether or not the adoptees, or adoptive families, are black, white, brown, yellow, green or purple!

But, if race differences in intelligence are indeed entirely environmental in origin, then where are these environmental causes to be found, if not in the family environment?

Harris has an answer – black culture.

According to her, the black adoptees, although raised in white adoptive families, nevertheless still come to identify as ‘black’, and to identify with the wider black culture and social norms. In addition, they may, on account of their racial identification, come to socialize with other blacks in school and elsewhere.

As a result of this acculturation to African-American norms and culture, they therefore, according to Harris, come to score lower in IQ than their white peers and adoptive siblings.

But how can we ever test this theory? Is it not untestable, and is this not precisely the problem identified by Jensen with previous posited ‘X-factors’.

Actually, however, although not discussed by Harris, there is a way of testing this theory – namely, looking at the IQs of black children raised in white families where there is no wider black culture with which to identify, and few if any black peers with whom to socialize?

This, then, brings us to the second of the two studies which Harris’s theory potentially reconciles, namely the famous Eyferth study.

Here, it was found that the mixed-race children fathered by black American servicemen who had had sexual relationships with German women during the Allied occupation of Germany after World War Two had almost exactly the same average IQ scores as a control group of offspring fathered by white US servicemen during the same time period (Eyferth 1959).

The crucial difference from the Minnesota study may be that these children, raised in an almost entirely monoracial, white Germany in the mid-twentieth century, had no wider African-American culture with which to identify or whose norms to adopt, and few if any black or mixed-race peers in their vicinity with whom to socialize.

This, then, is perhaps the last lifeline for a purely environmentalist theory of race differences in intelligence – namely the theory that African-American culture depresses intelligence.

Unfortunately, however, this proposition – namely, that African-American culture depresses your IQ – is almost as politically unpalatable and politically-incorrect as is the notion that race differences in intelligence reflect innate genetic differences.^[11]

Endnotes

[1] Thus, this ancient wisdom is reflected, for example, in many folk sayings, such as ‘the apple does not fall far from the tree’, ‘a chip off the old block’ and ‘like father, like son’, many of which long predate either Darwin’s theory of evolution, and Mendel’s work on heredity, let alone the modern work of behavioural geneticists.

[2] It is important to emphasize here that this applies only to psychological outcomes, and not, for example, economic outcomes. For example, a child raised by wealthy parents is indeed likely to be wealthier than one raised in poverty, if only because s/he is likely to inherit (some of) the wealth of his parents. It is also possible that s/he may, on average, obtain a better job as a consequence of the opportunities opened by his privileged upbringing. However, his IQ will be no higher than had s/he been raised in relative poverty, and neither will s/he be any more or less likely to suffer from a mental illness.

[3] Similarly, it is often claimed that children raised in care homes, or in foster care, tend to have negative life-outcomes. However, again, this by no means proves that it is care homes or foster care that causes these negative life-outcomes. On the contrary, since children who end up in foster care are typically either abandoned by their biological parents, or forcibly taken from their parents by social services on account of the inadequate care provided by the latter, or sometimes outright abuse, it is obvious that their parents represent an unrepresentative sample of society as a whole. An obvious alternative explanation, then, is that the children in question simply inherit the dysfunctional personality attributes of their biological parents, namely the very dysfunctional personality attributes that caused the latter to either abandon their children or have them removed by the social services. (In other cases, such children may have been orphaned. However, this is less common today. At any rate, parents who die before their offspring reach maturity are surely also unrepresentative of parents in general. For example, many may live high-risk lifestyles that contribute to their early deaths.)

[4] Likewise, the heritability of such personality traits as conscientiousness and self-discipline, in addition to intelligence, likely also partly account for the association between parental income and academic attainment among their offspring, since both academic attainment, and occupational success, require the self-discipline to work hard to achieve success. These factors, again in addition to intelligence, likely also contribute to the association between parental income and the income and socioeconomic status ultimately attained by their offspring.

[5] This possibility could, at least in theory, be ruled out by longitudinal studies, which could investigate whether the spanking preceded the misbehaviour, or vice versa. However, this is easier said than done, since, unless relying on the reports by caregivers or children themselves, which depends on both the memory and honesty of the caregivers and children themselves, it would have to involve intensive, long-term, and continued observation in order to establish which came first, namely the pattern of misbehaviour, or the adoption of physical chastisement as a method of discipline. This would, presumably, require continuous observation from birth onwards, so as to ensure that the very first instance of spanking or excessive misbehaviour were recorded. Such a study would seem all but impossible and certainly, to my knowledge, has yet to be conducted.

[6] The fact that the relevant environmental variables must be sought outside the family home is one reason why the terms ‘between-family environment’ and ‘within-family environment’, sometimes used as synonyms or alternatives for ‘shared’ and ‘non-shared family environment’ respectively, are potentially misleading. Thus, the ‘within-family environment’ refers to those aspects of the environment that differ for different siblings even within a single family. However, these factors may differ within a single family precisely because they occur outside, not within, the family itself. The terms ‘shared’ and ‘non-shared family environment’ are therefore to be preferred, so as to avoid any potential confusion these alternative terms could cause.

[7] Both practical and ethical considerations, of course, prevent Watson from actually creating his “own specified world” in which to bring up his “dozen healthy infants”. Therefore, no one is able to put his claim to the test. It is therefore unfalsifiable and Watson is therefore free to make such boasts, safe in the knowledge that there is no danger of his actually being called to make good on his claims and thereby proven wrong.

[8] Actually, even if race differences in IQ are found to disappear after controlling for socioeconomic status, it would be a fallacy to conclude that this means that the differences in IQ are entirely a result of differences in social class and that there is no innate difference in intelligence between the races. After all, differences in socioeconomic status are in large part a consequence of differences in cognitive ability, as more intelligent people perform better at school, and at work, and hence rise in socioeconomic status. Therefore, in controlling for socioeconomic status, one is, in effect, also controlling for differences in intelligence, since the two are so strongly correlated. The contrary assumption has been termed by Jensen ‘the sociologist’s fallacy’.
This fallacy involves the assumption that it is differences in socioeconomic status that cause differences in IQ, rather than differences in intelligence that cause differences in socioeconomic status. As Arthur Jensen explains it:

“If SES [i.e. socioeconomic status] were the cause of IQ, the correlation between adults’ IQ and their attained SES would not be markedly higher than the correlation between children’s IQ and their parents’ SES. Further, the IQs of adolescents adopted in infancy are not correlated with the SES of their adoptive parents. Adults’ attained SES (and hence their SES as parents) itself has a large genetic component, so there is a genetic correlation between SES and IQ, and this is so within both the white and the black populations. Consequently, if black and white groups are specially selected so as to be matched or statistically equated on SES, they are thereby also equated to some degree on the genetic component of IQ” (The g Factor: p491).

[9] Actually, at least some of these theories are indeed testable and potentially falsifiable. With regard to the factors quoted by Jensen (namely, “a history of slavery, social oppression, and racial discrimination, white racism… and minority status consciousness”), one way of testing these theories is to look at test scores in those countries where there is no such history. For example, in sub-Saharan Africa, as well as in Haiti and Jamaica, blacks are in the majority, and are moreover in control of the government. Yet the IQ scores of the indigenous populations of sub-Saharan Africa are actually even lower than among blacks in the USA (see Richard Lynn’s Race Differences in Intelligence: reviewed here). True, most such countries still have a history of racial oppression and discrimination, albeit in the form of European colonialism rather than racial slavery or segregation in the American sense. However, in those few sub-Saharan African countries that were not colonized by western powers, or only briefly colonized (e.g. Ethiopia, Liberia), scores are not any higher. Also, other minority groups ostensibly or historically subject to racial oppression and discrimination (e.g. Ashkenazi Jews, Overseas Chinese) actually score higher in IQ than the host populations that ostensibly oppress them. As for “the ‘black experience’”, this meanly begs the question as to why the ‘black experience’ has been so similar, and resulted in the same low IQs in so many different parts of the world, something implausible unless unless the ‘black experience’ itself reflects innate aspects of black African psychology.

[10] The fact that the heritability of intelligence is higher in adulthood than during childhood, and the influence of the shared family environment correspondingly decreases, has been interpreted as reflecting the fact that, during childhood, our environments are shaped, to a considerable extent, by our parents. For example, some parents may encourage activities that may conceivably enhance intelligence, such as reading books and visiting museums. In contrast, as we enter adulthood, we begin to have freedom to choose and shape our own environments, in accordance with our interests, which may be partly a reflection of our heredity.
Interestingly, this theory suggests that what is biologically inherited is not necessarily intelligence itself, but rather a tendency to seek out intelligence-enhancing environments, i.e. intellectual curiosity rather than intelligence as such. In fact, it is probably a mixture of both factors. Moreover, intellectual curiosity is surely strongly correlated with intelligence, if only because it requires a certain level of intelligence to appreciate intellectual pursuits, since, if one lacks the ability to learn or understand complex concepts, then intellectual pursuits are necessarily unrewarding.

[11] Thus, ironically, the recently deceased James Flynn, though always careful, throughout his career, to remain on the politically-correct radical environmentalist side of the debate with regard to the causes of race differences in intelligence, nevertheless recently found himself taken to task by the leftist, politically-correct British Guardian newspaper for a sentence in his recent book, Does Your Family Make You Smarter, where he described American blacks as coming from a “from a cognitively restricted subculture” (Wilby 2016). Thus, whether one attributes lower black IQs to biology or to culture, either answer is certain offend leftists, and the power of political correctness can, it seems, never be appeased.

References

Belsky, Steinberg & Draper (1991) Childhood Experience, Interpersonal Development, and Reproductive Strategy: An Evolutionary Theory of Socialization Child Development 62(4): 647-670

Draper & Harpending (1982) Father Absence and Reproductive Strategy: An Evolutionary Perspective Journal of Anthropological Research 38:3: 255-273

Eyferth (1959) Eine Untersuchung der Neger-Mischlingskinder in Westdeutschland. Vita Humana, 2, 102–114

Levin (1994) Comment on Minnesota Transracial Adoption Study. Intelligence. 19: 13–20

Lynn, R (1994) Some reinterpretations of the Minnesota Transracial Adoption Study. Intelligence. 19: 21–27

Scarr & Weinberg (1976) IQ test performance of black children adopted by White families. American Psychologist 31(10):726–739

Weinberg, Scarr & Waldman, (1992) The Minnesota Transracial Adoption Study: A follow-up of IQ test performance at adolescence Intelligence 16:117–135

Whitney (1996) Shockley’s experiment. Mankind Quarterly 37(1): 41-60

Wilby (2006) Beyond the Flynn effect: New myths about race, family and IQ? Guardian, September 27.

The Contemporary Heretic

Category: Behavioural Genetics

Judith Harris’s ‘The Nurture Assumption’: By Parent or Peers

Judith Harris, The Nurture Assumption: Why Children Turn Out the Way They Do. Free Press, 1998.

The Myth of Parental Influence

The Non-Shared Environment

Corporal punishment

Peer Groups

Environment ≠ Nurture

Language Acquisition

Code-Switching

An Evolutionary Theory of Socialization?

Absent Fathers, from Upper-Class to Under-Class

Abuse

Race Differences in IQ

Endnotes

References