Andrew Sullivan, super-blogger and race science advocate, once wrote a column in the Sunday Times in which he anticipated that neuroscience and genetics would help ‘find out for a fact whether there are any measurable genetic differences in IQ between different ethnic groups’.1 None have emerged but biologists on the make continue with the quest for the holy grail of genetic determinism: a racially distinct intelligence gene.
The first big stab came from Bruce Lahn, a Chicago-based Chinese biologist, who made the bold claim that he had indeed found the golden ticket. Lahn, a professor of human genetics at the University of Chicago, left China for America in the late 1980s, after taking part in pro-democracy protests. He sparked a firestorm in 2005, when he claimed in two papers in the journal Science that there had been recent genetic changes, linked to brain size and intelligence, among non-Africans.2 A Science report on his research had the headline: ‘Evolution: are human brains still evolving? Brain genes show signs of selection.’ It was illustrated with a picture of a muscular white man carrying an impressively large shaved head, echoing the pose of Rodin’s The Thinker. The caption read: ‘Big thinker? Certain forms of two brain genes may confer a selective advantage.’3 The journal honoured Lahn’s research as ‘Breakthrough of the Year’ and ran a glowing profile.
This delighted the far right, which viewed it as proof that non-black people were more intelligent than black people. The white supremacist publication American Renaissance featured a story by the overtly racist British psychologist Christopher Brand, in which he wrote that Lahn had ‘found that sub-Saharan blacks were the most distinct of the racial groups they studied, in that they had a markedly lower frequency of both variants. This is consistent with the distinct black African profile of smaller brains and lower IQ.’4
When I reviewed the evidence a few years ago, a quick Google search found websites with headlines such as: ‘DR BRUCE LAHN, U of Chicago Genetics, proves NIGGERS not fully HUMAN!’5 Sections of the mainstream right also seemed overjoyed. John Derbyshire wrote in The National Review that Lahn’s ‘bombshell papers’ showed that geneticists ‘have been lying through their teeth about the supposed genetic similarity of all races’. And he added: ‘[I]f different human groups, of different common ancestry, have different frequencies of genes influencing things like, for goodness’ sake, brain development, then our cherished national dream of a well-mixed and harmonious meritocracy with all groups equally represented in all niches, at all levels, may be unattainable.’6
Lahn was obsessed with genetic links between race and character. He once proposed an article on ‘why Chinese are boring’ and wondered whether there was ‘some selection’ against rebellious individuals in imperial China.7 He insisted evolution of the human brain was continuing, which meant some populations might have higher intelligence than others,8 and thought he might have turned some of these opinions into fact. He researched two genes (ASPM and microcephalin) that are associated with microcephaly, a condition in which there is a reduction in brain growth. He claimed different variants of these genes would be associated with positive brain growth, which would correlate with intelligence: ‘We’re seeing two examples of such a spread in progress. In each case, it’s a spread of a new genetic variant in a gene that controls brain size. This variant is clearly favoured by natural selection.’9
Lahn’s team examined the DNA of 1,184 people from fifty-nine populations and found that the new mutations spread more quickly beyond sub-Saharan Africa, particularly among European and Middle Eastern groups and that they were ‘probably’ associated with ‘higher IQ’. Lahn acknowledged the evidence was ‘iffy’ but none the less, speculated that a mutation of microcephalin emerged 37,000 years ago and spread to 70 per cent of humanity, coinciding with ‘the introduction of anatomically modern humans into Europe, as well as the dramatic shift towards modern human behaviour such as art and the use of symbolism’.10 Lahn guessed that interbreeding with Neanderthals might have prompted this genetic change and that this exchange made the Eurasian portion of the human population smarter. One reason he offered for his Neanderthal theory was that the new gene variant was more common in Eurasia than sub-Saharan Africa, ‘and we know that Neanderthals evolved outside Africa’, although he admitted that ‘[b]y no means do these findings constitute definitive proof’.11
The modern version of the other gene (ASPN) emerged 5,800 years ago, Lahn claimed, and spread to 30 per cent of humanity, again mainly in Eurasia. This, he said, coincided with the ‘development of cities and written language’, suggesting a causal link.12 Lahn never explained precisely why evolution for higher intelligence would emerge from very limited interbreeding with Neanderthals who, after all, most experts believe were probably not quite the equal of, let alone superior to, humans when it came to creative cognitive capacity. It is also worth remembering that the Neanderthals died out a couple of thousand years before Lahn’s start date for microcephalin and the interbreeding with Neanderthals that left a genetic trace came at least 12,000 years earlier.
Another problem was that the achievements Lahn picked seemed all too convenient. We could just as easily point to the cultural leaps of symbolic art and self-adornment made by the sub-Saharan Africans 77,000 years ago, the construction of Göbekli Tepe in Anatolia 13,000 years ago or the first urban centres 7,000 years ago. But this would not fit with his schema. Pointing to several historical errors, the Harvard historian of science Sarah Richardson noted that Lahn’s account was poorly grounded in history, geography and demography, relying instead on loose and sweeping generalisations about human cultural history.13
The most damning attacks came from other scientists, who savaged Lahn’s premises and conclusions. Geneticists at the Broad Institute in Cambridge, Massachusetts reanalysed his data and disputed his view that natural selection acted to change the ASPN gene. Geneticists from the University of California, Los Angeles tested these gene variants in 120 people and found they had no impact on brain size; a conclusion shared by the results of five other studies, including one from Lahn’s own laboratory. Other scientists found these gene variants were probably far older than he suggested. Sarah Richardson concluded there was ‘no evidence of an association between the alleles and either IQ or brain size’ and that the idea that larger brains led to higher intelligence was ‘also not grounded in empirical evidence’.14
The political right continued to regard Lahn as a hero for defying ‘political correctness’ but the response from the scientific world was very different. Following the demolition of his research by fellow geneticists, the two co-authors of his original papers in Science distanced themselves from the papers’ claims,15 while his university withdrew a patent application to use his work to develop a DNA-based intelligence test. Lahn defended himself, saying his work was being held to a higher burden of proof than others’ but as his claims unravelled he retreated from the debate, saying it was ‘too controversial’.
In the wake of Lahn’s research, several public figures had their DNA tested to find out which versions of the two genes they had, including the BBC’s notably cerebral political commentator Andrew Marr, who was amused to discover that he had the sub-Saharan variant. Even Lahn admitted he once tested himself to find whether he had the Eurasian or the sub-Saharan variant and although the results were indecisive, ‘it wasn’t looking good’.16
Other claims seemed to be drawn from Lahn’s faulty research. Steven Pinker wrote in 2009 that he had had his IQ tested and that the results were ‘above average’. He then had his genome sequenced, which prompted him to ask: ‘who wouldn’t be flattered to learn that he has two genes associated with higher IQ?’17 At that stage and subsequently, the only claims that had been made about single or double intelligence genes were Lahn’s, so one can only presume that Pinker had failed to read the many academic critiques of Lahn’s work that had been published by 2009.
Lahn’s failed bid to find racially linked intelligence genes was hardly the final word. There is prestige and money in genetics, which can attach itself to the scientists making the boldest claims. Lahn retreated with his tail between his legs but his case illustrated the need to be cautious about claims relating to the behaviour of any genes or combinations of genes, because the headlines prompted by publicity-hungry researchers do not always match the peer-reviewed reality.
Before delving deeper into the relationship between genes and environment for intelligence, we must understand the implications of decoding the human genome. In previous chapters I’ve stressed how full genome analysis has opened the way to discovering a great deal we never knew about ourselves but it is time to explain what genetics cannot do.
The first decoding of the full human genome in 2001 prompted a spate of wide-eyed reports based on the perception that the hardwired approach to understanding human behaviour had prevailed. Anyone whose knowledge of genetics was restricted to the news media might have decided that nature had trumped environment and culture. There were gushing tales about the discovery of genes ‘for’ traits ranging from television-watching to alcoholism. In every case, the predictions were later scotched. Claims of single genes ‘for’ a behavioural trait showed nothing more than the ignorance of the authors. An isolated gene can be no more than one element in the biomechanical pathway implicated in a trait. The environment, and the way it is experienced, has huge implications for the way those genes express themselves. Yet this kind of nuance was trampled over in the early excitement, even by those who should have known better. Daniel Kosman, editor of Science, was so thrilled that he claimed the nature-nurture debate was over, and that nature had won.18
Those at the core of the human genome project issued warnings against claims of finding genes for any behaviour, stressing that anyone hoping to find single genes for crime, sporting success or intelligence would be disappointed. Craig Venter, the scientist who led the private sector effort to decode the human genome, was particularly forthright: ‘In everyday language the talk is about a gene for this and a gene for that. We are now finding that that is rarely so. The number of genes that work in that way can almost be counted on your fingers, because we are just not hardwired in that way.’19
He stressed one point, that the gene-centric approach to human behaviour was simply wrong:
There are two fallacies to be avoided: determinism, the idea that all characteristics of a person are ‘hardwired’ by the genome; and reductionism, that now that the human sequence is completely known, it is just a matter of time before our understanding of gene functions and interactions will provide a complete causal description of human variability.20
The project produced some surprises along these lines. It emerged that the human genome contained around nineteen to twenty thousand protein-coding genes, one-fifth the anticipated number. In fact, not many more than a fruit fly and fewer than a mouse. The Stanford University biologist Paul Ehrlich noted that the human genome did not have anywhere near enough genes to programme the connections in our brain that control behaviour:
Our ‘gene shortage’ is one reason human infants and young children are so helpless. Their helplessness allows the physical and cultural environments to do the brain programming that our hereditary endowment couldn’t manage. It’s that environmental input that gives us the adaptability that is the hallmark of humanity. We could never have evolved as genetically controlled robots.21
Like Venter, Ehrlich expressed despair at the trend towards genetic determinism, particularly from evolutionary psychologists:
There is an unhappy predilection, especially in the United States, not only to overrate the effect of genetic evolution but also to underrate the effect of cultural evolution. Uniquely in our species, changes in culture have been fully as important in producing our natures as have changes in the hereditary information passed on by our ancestors.22
The decoding of the human genome coincided with fresh evidence, from the field of epigenetics, that genes weren’t the sum total of destiny. To paraphrase the Bible, epigenetics is all about the sins of the father or mother being visited on their children and their grandchildren. It refers to anything that can alter the impact of a gene but have no impact on the DNA sequence itself. Put differently, it is the study of heritable changes to the phenotype23 of an organism that don’t involve genetic changes. It therefore refers not to a change in genes but to whether a given gene is ‘switched on’ by environmental factors.
Epigenetics has prompted a huge volume of scientific research over the past two decades, and also produced its fair share of quackery and pseudoscience. Since the early 2000s, research has shown that certain traits can be passed on to the fourth generation. A study of male rats exposed to cancer-prompting insecticide found that while it produced no genetic changes in their offspring, the risks were nevertheless carried through four generations.24 Another showed that cocaine-sniffing male mice passed poor memories on to their offspring with no evidence of DNA damage, and a third showed pregnant rats exposed to nicotine passed on asthma to their offspring and to their grandchildren.25
How much or little we eat, what we eat, the alcohol we consume, even the levels of stress or isolation we face; all can affect the expression of our genes and those of our children and their children. This area of study has been dubbed ‘The New Lamarckism’ because it rehabilitated the discredited idea that environmental characteristics acquired during a lifetime could be passed on.26 Several studies have shown that fear and anxiety can be epigenetically transmitted to offspring; experiments on mice showed they could be trained to associate particular smells with fearful memories and that these could be passed on to later generations.27 Professor Kerry Ressler, a neurobiologist and psychiatrist at Emory School of Medicine in Atlanta, Georgia, who led a research project on the epigenetic transmission of fear in mice, said similar transmission could take place in people, with the fearful experience of a parent being transmitted to subsequent generations, perhaps through chemical changes in the sperm and the eggs acting as a ‘mechanism of conserving as much information as possible from a previous generation’.28
Several recent studies have indeed shown a transgenerational epigenetic inheritance of fear and anxiety among people, and also of addictive behaviour and a range of physical outcomes.29 Swedish researchers found that paternal-line grandsons of men who had been exposed to famine during adolescence were less likely to die of heart disease, while those whose paternal-line grandfathers had plenty of food were more likely to die of diabetes, and paternal granddaughters of women who had been exposed to famine while in the womb died at a younger age.30
Epigenetics also covers the more immediate impact of the environment on gene expression. A study of children conceived during the 1944–45 famine in the Netherlands found that ‘early-life environmental conditions can cause epigenetic changes in humans that persist throughout life’31; sixty years on, these people had an increased risk of diabetes, heart disease and other conditions because of the epigenetic impact of these diseases.
A particularly intriguing study involved 153 healthy people in their fifties and sixties. University of Chicago researchers identified the most socially secure, and the loneliest, and examined their DNA. They found that 1 per cent of the participants’ DNA (209 genes) responded differently in the two groups, particularly the inflammatory immune response. This study has since been replicated, confirming that social isolation was even more critical than stress as a disease risk factor.32 It has been further backed up by studies finding that the insecurity and isolation that often emerge from poverty seemed to exacerbate the genetic expression of those experiencing it, weakening their immune systems and making them more susceptible to several diseases. Steve Cole, the UCLA scientist who co-led the Chicago study, explained: ‘You can’t change your genes but … you can change the way your genes behave,’ or as he put it in a lecture: ‘Your experiences today will influence the molecular composition of your body for the next two to three months or perhaps, for the rest of your life. Plan your day accordingly.’33
Epigenetics is not about evolution. Epigenetic traits always fade. Rather, it presents an alternative to the binary question of whether a particular trait is genetically innate or environmental. Until recently, it was assumed that some health problems – depression, obesity, heart disease, cancer, schizophrenia – can only be explained by a combination of social conditions and genetic inheritance. Now there’s a third option: the environment can leave a biological imprint that may linger. Our experiences, and those of our parents and grandparents, might not only change our perceptions of the world; they might also change the response of our genes to this world. Performance in IQ tests is likely to be affected by the kinds of poverty-related gene expression – anything from susceptibility to asthma to anxiety levels to depression – that these studies have highlighted.
There have been several recent attempts to discover which combinations of genes and parts of the brain are implicated in intelligence, some more successful than Bruce Lahn’s. In 2010, a team of British psychologists and scientists researching the academic performance of four thousand British schoolchildren said they thought that perhaps two hundred genes could be directly implicated. However, their general point was that this form of intelligence was governed by a network involving thousands of genes, each making a tiny contribution, rather than by a handful of powerful genes, as once thought. ‘Of the gene variants we looked at, a couple of hundred are emerging which seem to have a small but significant relationship with ability in maths and English,’ said Robert Plomin, the King’s College London psychologist who led the research. Plomin, a hereditarian who signed a contentious race-based public statement on IQ,34 acknowledged that this said nothing about how the genes themselves might work, adding: ‘It seems that no single gene has a really big effect … [C]ognitive powers depend on lots of little effects from lots of genes.’35
Elsewhere, Plomin and the geneticist Oliver Davis cast doubt on other genetic presumptions, noting that ‘for most complex traits and common disorders the genetic effects are much smaller than previously considered: the largest effects account for only 1 per cent of the variance of quantitative traits.’36This caution was backed in 2009 by Ian Deary, head of the University of Edinburgh’s Centre for Cognitive Ageing and Cognitive Epidemiology. He acknowledged that one of the toughest jobs in genetics was trying to find genes linked to intelligence. A multiplicity of genes might affect intelligence indirectly, he speculated, by changing the way the brain grows. But, he added, ‘It is difficult to name even one gene that is reliably associated with normal intelligence in young, healthy adults.’37
If the idea of finding a single significant intelligence gene has faded, several research projects claim to have found combinations of hundreds of genes implicated in cognition. A research team from Harvard University and the Universities of Edinburgh and Southampton, including Ian Deary, used data from two genetic studies to compare the variation in DNA in more than 248,000 people, to isolate genes associated with intelligence, which they conflated with IQ. The researchers located 187 regions in the human genome, in the brain and pituitary gland, that they said were linked to intelligence (and also to longevity) and isolated 538 genes that played some role. They applied this finding to a smaller group, after which they claimed to be able to predict, based on individual DNA, nearly 7 per cent of IQ differences.38 The first author, David Hill, said they were also able to identify ‘some of the biological processes that genetic variation appears to influence to produce such differences in intelligence’,39 while Deary added that the study suggested that ‘health and intelligence are related in part because some of the same genes influence them.’40
The Dutch statistical geneticist Danielle Posthuma has been involved with studies of the relationship between genes and intelligence for several years. The largest and most significant was an international meta-analytical study of almost 270,000 people of European ancestry. This study, published in 2018, claimed to have discovered 205 DNA regions and 1,016 brain genes linked to intelligence, mostly located in the brain’s basal ganglia, where learning and emotion are processed. The subjects’ DNA analysis was married to their IQ scores to find the links. The researchers also found that higher scores correlated with lower odds of suffering from schizophrenia, attention deficit hyperactivity disorder and Alzheimer’s disease but higher odds of autism. An interesting aspect of the study was that all but seventy-seven of the intelligence-related genes had not previously been identified as such.41
Both studies found hundreds of genes that correlated positively with aspects of cognition; they also seemed to affect aspects of health and longevity. But neither study found – or looked for – variations in the presence of these genes in different population groups. The closest we seem to have come to a single brain gene that varies from population to population is the KL-VS variant of the klotho gene, which modulates ageing. However, this has not been proved to relate to IQ and the research has been conducted on older people, so it is not yet clear how this allele would affect younger subjects. It therefore doesn’t cut it as the ‘Holy Grail’ of genes sought by Bruce Lahn and his ilk.
Research conducted at the University of California, San Francisco, published in 2014, involved an analysis of KL-VS, which, the researchers said, as well as being associated with longevity, might also improve the brain’s ability to perform certain everyday mental tasks, particularly relating to memory, by increasing the strength of the connections between nerve cells in the brain. ‘Our findings suggest that the KL-VS variant promotes cognition by increasing levels of secreted klotho,’42 the paper’s authors wrote, in reference to a protein produced by KL-VS. One of the co-authors, Lennart Mucke, added that the allele, which was present in about 20 per cent of the population they researched, may ‘increase the brain’s capacity to perform everyday intellectual tasks’.43 They have since synthesised the protein and given it to mice; they conclude that it ‘also enhanced cognition’ and that this synthetic version might be useful in countering dementia. The team, which was studying age-related cognitive decline, found that those with the KL-VS variant tended to live longer and have a lower risk of stroke and age-related heart disease. After analysing data on three groups of ageing white Americans, they also found that the one fifth of volunteers who had this variant tended to perform better in cognitive tests, leading to speculation that it could prompt IQ differences of ‘up to’ 6 per cent.
There are several reasons why these tentative conclusions should prompt more caution than that shown by journalists’ talk of an ‘IQ-boosting gene’.44 First, while the team tested memory, attention, visuo-spatial awareness and language, they did not test IQ. The suggestion that this allele could be implicated in higher IQ must therefore be seen as speculative, and some of the media claims even more so. The Economist wrote that the KL gene ‘could account for as much as 3 per cent of the variation of IQ in the general population …’.45 Phrases such as ‘up to’ and ‘as much as’ should cause further caution, because when they are attached to statistical data they divest the numbers of real meaning: ‘up to 6 per cent’ could mean 6 per cent or it could mean 0.00001 per cent.
Bold claims resulting from a single paper should be handled gingerly, particularly when the data are drawn from a narrow population range. Almost all those tested were white Americans and the authors acknowledged the ‘potential limitations in extrapolating our data worldwide’, adding ‘[I]t is possible that more diverse genetic or environmental influences could alter or mask the effect in other populations.’46 But the sampling problems go beyond its narrow ethnic base. All the volunteers were aged between fifty-two and eighty-five. The authors suggest cognitive impact is not age-related, but we really don’t know. If they had tested volunteers aged between twelve and eighty-five, would the results have been the same? Or would a ‘longevity gene’ have more impact on the cognition of ageing people? A 2018 paper dealing with this gene variant noted that it was associated with the right dorsolateral prefrontal cortex, a part of the brain involved with planning and decision making, ‘which is especially susceptible to shrinkage with age’.47 This might suggest its impact is indeed age-related.
Are there potential implications for race and intelligence? Let’s begin at the extremes; assume that, on average, people with this allele have higher IQs and that the estimate of ‘up to 6 per cent’ is 5 per cent, rather than 0.00005 per cent. Let’s also assume that this allele is unevenly distributed among world populations. If 25 per cent of population A carries this allele and 5 per cent of population B has it, all other things being equal, the average IQ of population A should be 1 per cent higher than that of population B. This, however, would be impossible to assess. As we shall see in future chapters, comparing IQ averages of different populations is fruitless because of different cultural, educational and other environmental influences. And there have also not been large cross-population studies of the presence of KL-VS.
The most we can say is that, given the limited ethnic data on KL-VS, it does not correlate with the dubious ethnic data on average IQ scores. A study of 107 Iranians found that none carried KL-VS, leading researchers to conclude that this allele ‘seems to be [scarcely found] in the Iranian population’.48 Does the nation that produced the wonders of ancient Persia have a reputation for low IQ? I searched for reports of Iranian IQ studies and found several studies showing that Iranians were average; even the much-derided ‘IQ of nations’ chart produced by Richard Lynn put Iranians in the middle.49 Studies of 723 Caucasian and 242 African-American elderly heart patients in Baltimore showed that almost double the proportion of black patients, compared to white patients, carried the KL-VS allele so it would seem that the relationship between KL-VS and ethnicity is not what knee-jerk racists would anticipate or hope for.50 Even if subsequent studies confirm that people with this gene variant have, on average, fractionally higher IQs, this is a long way from saying it evolved for intelligence.
There are three broad perspectives on the contribution of population genetics to intelligence. What is sometimes self-described as the ‘human biodiversity’ view or ‘racial realism’ stresses that intelligence-related genes differ substantially among populations and races, as demonstrated by the disparity in IQ scores. The currently most popular example is the higher than average IQ scores of Ashkenazi Jews.
Other than the contrarian 91-year-old James Watson, no other prominent living geneticist has lent their name to this view. The podcaster Sam Harris claimed, in a debate with Slate’s editor-at-large, Ezra Klein, that several had ‘privately’ expressed their backing for this idea to him but were afraid to come out because of a climate of political correctness. However, at the loud and proud level, its most prominent backers are vloggers and YouTube stars such as Harris, Jordan Peterson and Stefan Molyneux, journalists such as Andrew Sullivan and Nicholas Wade, evolutionary psychologists such as Steven Pinker, Richard Lynn and Satoshi Kanazawa, and disparate academics, such as the political scientist Charles Murray, the anthropologist Henry Harpending and the educational psychologist Linda Gottredson.
A very different view is prominently associated with the Harvard geneticist David Reich, who is a passionate opponent of racist science and has sunk his teeth into Watson, Wade and Harpending, accusing them of conflating knowledge about average genetic differences between populations with guesswork that ‘has no merit’ and no base in serious scholarship.51 He says their claims are racist, have no scientific authority and their speculations ‘correspond to long-standing popular stereotypes – a conviction that is essentially guaranteed to be wrong’.52 Reich’s perspective is that even though the differences within human populations are significantly greater than those among populations, all genetic traits, including those that affect cognition, might differ between populations. However, he adds, we have ‘no idea right now what the nature or direction of genetically encoded differences among populations will be’.53 He disputes the idea of a qualitative distinction between the evolution of physical traits such as skin colour, which involve a handful of genes, and those relating to the brain, which has more than ten thousand genes. Reich points out that some physical traits that have evolved differently between populations also involve a multiplicity of mutations; height is a prime example. He writes: ‘[I]f natural selection has exerted different pressures on two populations since they separated, traits influenced by many mutations are just as capable of achieving large average difference across populations as traits influenced by a few mutations’.54
However, Reich emphasises that the differences that emerge will not follow the boundaries of the traditional social categories of race. We can be sure of this because recent full genome studies of ancient DNA show that the current populations of the world are ‘mixtures of highly divergent populations that no longer exist in unmixed form’ and are therefore ‘not exclusive descendants of populations that lived in the same locations ten thousand years ago’. This knowledge, he says, should warn those who think that the ‘true nature of population differences will correspond to racial stereotypes’55 He adds that the offence of racism is to judge individuals by a stereotype of their group, which is sure to be misleading. ‘Statements such as “You are black, you must be musical” or “You are Jewish, you must be smart” are unquestionably very harmful.’56 The example he offers to show where surprises might be found relates to the greater genetic diversity among sub-Saharan Africans. He refers to the 33 per cent higher genetic diversity in West Africans, compared to Europeans, and proposes this might be a reason why they produce faster sprinters. He generalises from this point, suggesting it might also translate into a higher proportion of sub-Saharan Africans with other extreme genetic abilities ‘including cognitive ones’.57 In other words, it is possible that the range of cognitive potential in Africa, at both ends of the spectrum, is greater than elsewhere.
A third view – held by most of those working in this area – is that significant genetically based intelligence differences between populations are unlikely. Among this view’s most prominent advocates are the evolutionary biologists and geneticists Richard Lewontin, Steven Rose, Steve Jones, Paul Ehrlich, Kevin Mitchell and the late Stephen Jay Gould, IQ theorists such as Richard Nisbett, Eric Turkheimer and the late Leon Kamin and the paleoanthropologists Ian Tattersall and Agustin Fuentis. The strand of their case relating to population genetics has recently been argued by Mitchell, a geneticist from Trinity College, Dublin. Like Reich, he uses height as an analogy, noting that some gene variants make people a bit shorter and others a bit taller, on average. However, the balance has been maintained by natural selection, because growing ever-taller has no benefit. On the other hand, evolutionary forces ‘drove intelligence in one direction only in our ancient ancestors’, as it was our ‘defining characteristic and our only real advantage over other animals’ and was amplified through culture and language. The selective advantage of ever-greater intelligence prompted a ‘snowball effect’ that was probably only stopped by limits imposed by the size of the birth canal and the ‘metabolic demands of a large brain’.58 This is the evolutionary reason for the huge cognitive potential of our complex brains. But while random mutations will affect all genetic programmes, with intelligence they are likely to affect it negatively, because of its genetic complexity. ‘[O]nce that complex system was in place,’ Mitchell said, ‘the main variation would be in the load of mutations that impair it, which will likely have effects on many traits and impair fitness generally.’59 But in any population general fitness will always be selected, ‘meaning that intelligence will get a free ride – it will be subject to stabilising selection, whether or not it is the thing being selected for’.60
Mitchell draws another analogy, this time with a Formula 1 racing car: intelligence, like the car’s performance, is ‘an emergent property of the whole system’. Thousands of brain genes affect intelligence; there is no dedicated module on which natural selection can act without affecting other traits.61 And just as random tinkering is unlikely to improve the car’s performance, so intelligence won’t be affected by a balance of IQ-boosting mutations and IQ-harming mutations. Instead, genetic differences in intelligence may ‘largely reflect the burden of mutations that drag it down’, which is why evolution will tend to select against them.62 Mitchell dismisses The Bell Curve’s idea that genes have at least something to do with racial differences in intelligence, saying that not only is there no evidence for this but that it is ‘inherently unlikely’,63 because the genetic forces acting against population-level variation in intelligence are far stronger than those supporting it.
Because so many genes are implicated in the brain’s development, there will inevitably be some variation in the mutations affecting cognition among populations, clans, families and individual people but this ‘constant churn of genetic variation works against any long-term rise or fall in intelligence’.64 At the population level, natural selection will remove mutations with large effects; those with minor effects might linger. But significant intelligence differences are highly unlikely because this would require ‘enormous’ and persistent differences in selective forces among groups, which would have to act across huge areas, in wildly different environments, and persist over tens of thousands of years of cultural change, which ‘is inherently and deeply implausible’. Instead, what might once have appeared as distinct races, in fact consist of numerous interbreeding population groups with arbitrary divisions ‘and the larger and more ancient, the greater diversity there will be within that group’.65 Mitchell concludes that while genetic variation helps explain why one person is more intelligent than another, there are unlikely to be ‘stable and systematic’ genetic differences that make one population smarter than another.66
Both Mitchell and Reich recognise that most cognition-related genetic differences are between individual people rather than populations. But unlike Reich, Mitchell says that population-level differences are unlikely, and if they do exist, they will be minor. He stresses evolutionary factors acting against significant population differences in intelligence. Reich emphasises the possibility of significant differences in cognition genes among populations but says they won’t follow the lines of traditional race groups. The form they will take is unknown, but he offers the example that greater genetic variation in sub-Saharan Africa may lead to more variation in cognition genes.
The view of Jim Flynn, widely recognised as the world’s leading IQ theorist, falls between these two stools. He applies this to possible differences between male and female IQ scores as well as to the white-black gap in the USA. In the twentieth century, on average, adult men marginally outperformed adult women in IQ tests, but in recent decades women’s average IQ scores have overtaken those of men. The reasons relate to changes in the workplace and education. ‘This improvement is more marked for women than for men because they have been more disadvantaged in the past,’ Flynn says. ‘The full effect of modernity on women is only just emerging.’67 He made a similar point on racial IQ differences in 2018: ‘I think it is more probable than not that the IQ difference between black and white Americans is environmental. As a social scientist, I cannot be sure if they [African Americans] have a genetic advantage or disadvantage.’68 He noted it was possible that the two groups’ ten-point difference in IQ reflected a twelve-point environmental difference, which would mean that black Americans had a two-point innate edge.
Claims relating to brain genes do not only concern intelligence. They can also relate to behaviour. Some of these claims have tipped over into the realm of racism. In July 2016, when the Breitbart website boss Steve Bannon was about to become the chief of Donald Trump’s presidential campaign, he wrote a piece on the shooting of black men by white policemen entitled ‘Black Lives Matter is a left-wing conspiracy’. His money quote: ‘Here’s a thought: What if the people getting shot by the cops did things to deserve it? There are, after all, in this world, some people who are naturally aggressive and violent.’ Bannon implied that the disproportionate number of black men killed by the police was a result of the ‘naturally’ disproportionate aggression and violence of the black victims; probably a reference to an alt-right trope that circulates on websites such as Breitbart and other race-obsessed sites of ‘the psycho gene’ (also called ‘the extreme warrior gene’) that predisposes its black bearers towards violence. Richard Lynn cited this gene as part of his claim that black people have evolved to have ‘greater psychopathic tendencies’69. Research on this allele was devoured with particular relish by Nicholas Wade, whose favourite example of genetic differences among populations involves the MAOA-2R allele, which is more common among African Americans; one in 21 carry it, compared to one in 200 Caucasian-American men and one in 150,000 Asian-American men.
The MAOA case indeed illustrates that genes relating to mental well-being can vary among populations. Other examples I touched on in Chapter 6, in relation to genetic drift, include schizophrenia and multiple sclerosis but any of the ethnically varying diseases, including hypertension, prostate cancer, Gaucher’s, cystic fibrosis and stroke can have an impact on mental well-being. However, when we take a closer look at MAOA it emerges that the claims of the alt-right are not borne out. It does not make a significant contribution to increasing rates of violence among any population, and contrary to some reports does not appear to have been selected for this reason, or indeed selected at all. A study that dealt with MAOA frequency among South Africa’s white Afrikaners noted the role played by genetic drift.70
The place to start is with an earlier recipient of the ‘warrior gene’ label, the far more common MOAO-3R variant, which causes lower levels of production of a protein needed to break down old serotonin in the brain. It has therefore been associated with risk-taking, depression, aggression, violence and anti-social behaviour. The variant is found in 56 per cent of Maori men and 48 per cent of African-American men, compared to 34 per cent of European men. It was used to explain higher rates of Maori criminality, but further studies produced a surprise: its prevalence among Chinese men was the same as among Maoris (56%) but the highest prevalence was in Taiwanese men (61%). Neither Chinese nor Taiwanese men have high rates of criminality, violence or anti-social behaviour.
Research on the ‘extreme warrior’ 2R variant suggests its association with criminal violence has mainly been confined to men who were abused as children. Poverty, maltreatment during childhood, interrupted education, low IQ, being beaten as a child, as well as high testosterone levels, have been cited as contributory factors.71 Other influences include the in utero environment of the foetus, and the amount of attention given to children. It is also likely that epigenetics plays a role in the expression of this allele.72 One study linked its epigenetic expression to women’s alcohol and nicotine dependence during pregnancy. In other words, if you carry 2R, and your mother smoked and drank during pregnancy, you’re more likely to be aggressive.73 Dire social conditions are therefore most likely to significantly raise the odds of being agitated, aggressive, impulsive or depressed, at least among African-American men. Caucasian men carrying this gene have higher rates of anti-social behaviour even if they have relatively normal upbringing.
Different environments can make profound differences to how genes are expressed. In one study, scientists took one thousand baby bees from their hives; five hundred aggressive African ‘killer bees’ and five hundred easy-going European bees. They swapped them, putting the African bees in the European hive and vice versa. The impact on gene expression was dramatic. It was not only that the bees behaved differently if placed in a foster hive; their genes behaved differently, even though the genome itself was unaltered. The way the genome behaved was far more socially fluid than previously realised.74
Some human genes (‘plasticity genes’) are particularly susceptible to environmental influence; MAOA is one. A study on the impact of physical discipline conducted by a team led by the University of Pittsburgh developmental psychologist Daniel Choe found that being hit as a young child was causally associated with anti-social behaviour in adolescents and young men with the 2R and 3R variants.75 Another study showed that ‘delinquency risk for the 2R allele is buffered for males close to their biological or social father.’76 The low-stress uterine environment of a mother who didn’t smoke or drink during pregnancy, supportive, non-smacking parents and, particularly, attentive fathers, seems to mitigate against 2R and 3R prompting anti-social behaviour, at least among African-American men.
More than 95 per cent of black American men do not carry MAOA-2R. Kevin Beaver, a Florida State University biosocial criminologist who has studied this allele, is sceptical of claims that it is a candidate for explaining violence and criminality in black men, because it’s simply ‘not common enough in African Americans’.77 Speculating loosely, let’s say one in twenty members of population X carry 2R, and that 20 per cent of that group suffered the kind of abusive or chaotic upbringing that would make them prone to significantly higher rates of criminal violence, and among those abused carriers, the rate of criminal violence is double that of non-carriers. This would mean that the additional contribution to population X’s proportion of criminal violence made by those with 2R would be 1 per cent.
In reality, it’s more complicated. Different populations appear to respond differently to what is known as ‘low-yield MOAO’ (particularly 2R). One study of Caucasian and African Americans found that when all those who’d experienced childhood abuse were excluded, the presence of low-yield MOAO did not significantly raise the odds of anti-social behaviour among African Americans but did raise it by 41 per cent among Caucasian Americans, compared with those with high-yield MOAO. In other words, among African Americans the key factor was not the 2R gene variant alone but rather the combination of 2R and the abuse suffered in childhood, while among Caucasians 2R seemed to raise the risk of anti-social behaviour regardless of childhood abuse.78 Put differently, there might be proportionately fewer Caucasians with 2R but its negative impact is more severe for them.
The researchers noted there may be both environmental and genetic reasons for this. The genetic reasons may relate to the fact that aggression and violence is influenced by many genes, some of which have yet to be identified, and some of which may be more prevalent among Caucasians. This is a point acknowledged by Wade, who uses the example of the presence in Finns of the HTR2B-Q20 allele, which, as he puts it, ‘predisposes the carrier to impulsive and violent crimes when under the influence of alcohol’.79 This allele is carried by more than 100,000 Finns (2.2 per cent of the population) and although it is mainly associated with drunken violence in men, because of a ‘tendency to lose behavioural control while under the influence of alcohol’,80 the results of a recent study show that men with this mutation are also more impulsive ‘even when sober and they are more likely to struggle with self-control or mood disorders’,81 said its first author, Roope Tikkanen. Finally, a study on the impact of low-yield MAOA on women found it had the opposite impact to that on men, relating to ‘greater happiness’.82 This contradicts earlier studies that found it was associated with depression in both men and women,83 illustrating the point that when it comes to research on the impact of genes, it is wise to treat them with caution.
Reich notes that the history of science has revealed over and over again ‘the danger of trusting one’s instincts or being led astray by one’s biases – of being too convinced that one knows the truth’. This, he warns, can lead to racist stereotypes, of the kind peddled by the likes of Nicholas Wade and James Watson, that are ‘essentially guaranteed to be wrong’. He adds: ‘We truly have no idea right now what the nature or direction of genetically encoded differences among population will be.’84 This is a debate that will run and run. Claims will be made about ‘IQ genes’ and about how their distribution differs among populations. These claims will inevitably be over-cooked or misinterpreted by people who want to stress race-based difference, until the logic and the research is scrutinised by peer review, as happened for Bruce Lahn. There’s a long way to go on this one.