Charles Murray, a sociologist by background and a datanaut by inclination, has carved out a prominent place in American intellectual debate by the simple expedient of writing clearly about difficult subjects. He is an Enlightenment Regular Guy, who does not want Americans to lose ground, or be split apart or be cast asunder by imperious elites and their lucrative patterns of frustration. He crunches data, and writes his conclusions in plain text, with helpful explanations about the harder statistical bits. No wonder some people hate him for it.
Having “The Bell Curve” on my university library shelves 26 years ago seemed somewhat daring. I was bewildered by the passions it arose. He had found a dataset and analysed it carefully, using histograms rather than correlation coefficients. I enjoyed the powerful clarity of the findings, and ruefully acknowledged that “bell curve” was a snappier phrase than “standard normal distribution”. I wish I’d had the talent to write it. Perhaps many other academics felt their noses put out of joint by a job well done.
We owe the inspiration for this book to Murray’s wife, who was so outraged by the attack he received at Middlebury College that she urged him to enter the fray on more contentious topics. Cherchez la femme. On the logical premise that “I might as well be hanged for a sheep as a lamb” Murray has obligingly bundled up all the taboo subjects, examined them and explained that they are not so frightening after all. This time he is not crunching new stuff (beyond some interesting investigations of class differences), but mostly explaining what a whole torrent of new research may mean for all of us. In that sense he is following up on his work on Human Excellence, identifying those thoughts and findings which later ages will find of note in ours. These are exciting times, and although we cannot be sure that this is a whole new chapter in our understanding of ourselves, it certainly feels it might be.
Critics will quickly note that Murray’s aim is seditious. He wishes to destroy the proposition that in a properly run society, people of all human groupings will have similar life outcomes. Clearly, they won’t, and the fast flourishing genetic revolution is what provokes Murray to provide a progress report, one he hopes will be out of date shortly. Incidentally, writing a book about the genetics of human behaviour is a selfless act. This book took a long time to write, working through complex new research, but Murray is aware it might have a shelf life of a few months. Given that his explanations of basic issues are helpful, I think it will last far longer.
Let me therefore state explicitly that I reject claims that groups of people, be they sexes or races or classes, can be ranked from superior to inferior.
Murray is very clear about this, though probably wrong. Whether such a ranking is possible is an empirical question, not a moral one. The Scots may be a dour lot, easily distinguishable from a ray of sunshine, mean, resentful, rough and prone to lachrymose sentimentality; or they may be the inventors of the modern world. Opinions differ, but those opinions can be tested by reference to the historical record. I see no reason at all why we should not compare their achievements to those of the Finns, the Bantu, the Tibetans and the Tutsi, and pronounce on their pecking order, and respond to the other evidence-based ranking which might be proposed. Human excellence varies. Books have been written about this. Comparisons are illuminating, and sometimes life-saving. I rate the Allies higher than the Nazis. Whatever the emotions aroused by the provocative language employed by critics, empiricism must prevail.
In reading the book, I was aware that I would want to show that I was above needing to read it. Wrong. Murray handles his scalpel politely but accurately. For example, I had questioned the way Janet Shibley Hyde deals with sex differences (saying she included variables which had never been considered sexually different so as to swamp those that were) but failed to make the point that Murray easily comes up with, which is that to atomise behaviour by individual variable comparisons obscures overall effects, which he calls “profiles”. In a nutshell, there should be poly-variable sex difference risk scores. He discusses when individual differences need to be summed up or considered in isolation.
For example, do you note that these two faces are so similar that individual comparisons of features might mostly show very few significant differences, yet most of us could tell which is man and which is woman?
Murray nails it. Murray also gives you notes on arguments contrary to what he has just concluded in his book, and also rejoinders to those arguments. Proper.
Another advantage of having a good explainer is to be “reminded of/learn for the first time” things which are important but get lost in the detail. Unprompted, I can give an overview of five factor personality theory, and whether collapsing them into a general factor is warranted (probably yes, for high level analyses, I think) but Murray makes the important point that, in terms of variance accounted for, they rank thus: neuroticism (emotional stability), extraversion, openness, agreeableness, and conscientiousness. Stability matters.
Are sex differences already there at birth? Here is Murray’s account of Baron-Cohen’s 2002 paper.
The most dramatic example of a finding from infancy, which led to considerable publicity, was a 2002 study presenting evidence that newborn girls no more than two days old after birth showed stronger interest in a human face while the newborn boys showed stronger interest in a mechanical mobile. It is a single, unreplicated study with a sample of 102, not proof to take to the bank, but its finding was in line with many other studies that have found personality sex differences in infants.
I judge this a balanced account. My own assessment is truculent: on the “last man standing” rule, this result stands until refuted. Baron Cohen told me the reason it hadn’t been replicated is probably that it takes over a year to get new-borns at the right stage of new-born-ness and alert-ness to get this large sample, requiring the researcher to be on call for long hours for 12 months.
Murray deals deftly with the research showing that sex differences increase in societies which remove barriers to women’s advancement:
Both sexes become freer to do what comes naturally.
Sex differences in Maths are very small on average, and get bigger as the tests get hardest. At the very highest level the ratio is 14 to 1. (page 52)
In the table, I counted perfect scores of 150 as being in the 99th percentile. When they are broken out separately, it turns out that from 2009 to 2018, 97 males and 7 females got perfect scores: a ratio of 13.9 to 1.
In general, whatever the level of ability, including those of IQ 186, men prefer to work with “things” women prefer “people”.
At the turn of the twenty- first century, it was known that the incidence of depression was higher among women, that women ruminate more than men, and that there was probably some relationship between those facts. Two decades later, important components of the biological processes of depression are understood and progress continues to be rapid.
Turning to Race, Murray bins it in favour of “ancestral population”. However, he properly continues to use it in his Propositions, so it may be a euphemism too far. He gives a clear, cautious account of race differences, and speculates as what polygenic risk scores may eventually achieve when they are calculated for all ancestral populations.
Murray is perfectly clear that, whatever we call it, there are continental differences, and 7 turns out to be a good number for the races of humanity.
He also shows that if you drill down into the detail you will discern subgroups, but some continental races are more samey than others. For example, look at the vertical lines above, each one representing a single person and thus what proportion of that person is made up of the ancestral population. All the Africans are all African. Europeans are most of them wholly European, as are most of the East Asians, East Asian; most of the Americans, American. As to the Middle East, most of the lines have a mix of Mideastern, European, and Central or South Asian ancestry. The people of the Middle East are an admixture, others far less so. Do that make some groups purer than others? In the statistical sense, yes. They are less stratified. Whether that is good or bad for health and behaviours depends on other tests and other values.
Oddly enough, the main reason why we know that race really exists is that it gets in the way. That is, in searching for links between medical disorders and SNPs, those links which work in one race don’t always work so well in other races. Eventually each continental race will be studied in enough detail to look at the genetic differences which may account for health and behavioural differences.
More and more unique SNPs are being found, and much more are like to be found, and need to be found, before they will all make sense.
However, we can already guess that some significant differences will be found on genetic markers of interest by comparing within continent comparisons (say English with Italians) with between continent comparisons (say Italians with Chinese). When that is done, races show strong communalities within themselves, and differences between races. The scatterplots are shown on page 187 and 188. The overall results are shown on page 190.
Within continents, the allele frequencies are tightly bound (97% of the variance for brain volumes), across continents less so (59% between Africans and Asians, 62% between Europeans and Africans, 74% between Asians and Europeans). So, we have the usual pattern: Asian, then European, then African. The gap between Asians and Africans is the biggest. That general pattern includes all five measures of cognitive performance. These findings lead one to expect that ancestral populations differ in ability and behaviour as a result of genetic differences.
The book is a master-class in explaining, and is far closer to text-book than meta-analysis, though it performs that latter function. Sadly, Murray cannot name his many advisors who looked at drafts of his book and made helpful suggestions. Contemporary academia is poisonous on race, sex and class. Happily, there are many knowledgeable people who were able to help him give an accurate and balanced account, without needing to share in the lime light. Veritas liberabit vos.
Murray is a good top-level guide to genetic discoveries precisely because he is outside the field looking in, with the purpose of being an explainer. Good writers in science quickly make you feel you knew the subject anyway. He is to behavioural science what Feynman’s lecture notes were to physics. Which reminds me of a Feynman quote highly relevant to what Murray is doing in the this book: Reality must take precedence over public relations, for nature cannot be fooled.
Clearly, humans are under selective pressures. We are changing fast, very fast compared to what was assumed before the genetic revolution, and still pretty fast when looking back at our adoption of agriculture. Those ancestors who left Africa also screwed around with Neanderthals and Denisovans, picking up interesting introgressive variants in the process, perhaps assisting adaption to cold climates, and accounting for Tibetan’s ability to live healthily at high altitudes.
As to continental differences in the genome, Murray reports (page 179):
An early analysis of local adaptation using the HapHap database was published in 2006 by a team of geneticists (first author was Benjamin Voight). They examined regions of the genome under selection pressure for three populations: Yoruba (a Nigerian tribe), Europeans from a mix of Northern and Western European countries, and East Asians (a mix of Chinese and Japanese). Of the 579 regions, 76 percent were unique to one of the three populations, 22 percent were shared by two of the three, and only 2 percent were shared by all three populations. In the authors’ judgment, the degree to which selection occurred independently is probably underestimated by these percentages. In any case, these events represent quite recent selection—“average ages of ~6,600 years and ~10,800 years in the non- African and African populations respectively,” in the authors’ judgment.
For Europeans, that is 235 generations ago.
As a consequence of this separation, medical researchers have to restrict their investigations to specific continental groups, or they will fail to find reliable disease variants.
For example, a polygenic score based on a test population of English and Italians usually generalizes accurately for French and Germans, not so accurately for Chinese and Indians, and least accurately for the genetically most distant populations from sub-Saharan Africa.
Part II has described a parallel universe. In the universe inhabited by the elite media and orthodox academia, it has been settled for decades that race is a social construct. In that universe, the lessons taught by Richard Lewontin and Stephen Jay Gould back in the 1970s and early 1980s still apply. In the universe inhabited by geneticists who study human populations, the 1990s saw glimpses of a new perspective, and the new century opened up fascinating stories that had previously been closed.
Murray refers to being intelligent as matter of luck. I would like to nit-pick about that, because that phrase is inimical to one of his key purposes, which is to strengthen respect for the institution of marriage. Intelligence is not a matter of luck. It depends on the sensible choice of partners your parents made. It would only truly be a matter of luck if humans mated randomly, say by gathering in vast crowds every full moon for the express purpose of almost random fornication.
To begin to wind up, in a treasure-trove of a book, I have picked up just a few things that struck my eye.
Taking all jobs together, the predictive validity of IQ scores for overall job performance is about +.50 (it’s higher than that for high- complexity jobs). You can square that figure and point out that IQ explains only 25 percent of the variance in job performance. If you’re an employer, however, and are told that a standard deviation increase in IQ is associated with half a standard deviation increase in overall job performance, a predictive validity of +.50 is a big deal.
I like that one. At another time it is worth discussing whether we should continue to disparage correlation coefficients by always “squaring to minimize”. I think this is a ploy by mystical environmentalists who imagine that by downplaying the variance accounted for by a biological predictor they somehow own the (larger) unexplained variance.
Usually, the class you attain is more determined by your intelligence than the class you were born into. That advantage is not a genetic lottery, that is not unless your parents mated randomly: it is a benefit conferred on children by thoughtful parents. To say “we don’t choose our parents” obscures the important fact that most parents choose each other, and have forgone other options.
Murray persists in arguing thus:
The new form of unfairness is that talent is largely a matter of luck, and the few who are so unusually talented that they rise to the top are the beneficiaries of luck in the genetic lottery.
I think he has this wrong. Certainly, being born bright is not a personal achievement. Choosing a mate wisely is, and should not be under-valued. If the Murrays are doing well, then all power to their mating choices. Although Murray himself did not choose his parents, he was not lucky. He descended from canny people.
Murray treats epigenetics kindly, but politely has his doubts. The evidence so far supports his doubts.
Murray is also kind about 50 years of attempts to boost intelligence and good behaviour by interventions short of full adoption, and concludes that we should not be entirely down-hearted. Designer drugs already work, and will get better. CRISPR will work in the future, and also get better. The old stuff which keeps being tried again and again needs to be laid to rest, apart from those which are simply kind helpfulness, which we should accept anyway.
Consider educational attainment, a rough proxy measure for IQ, as an example. In just the five years from 2014 through 2018, the percentage of the variance that could be explained from genetic material alone went from zero to 15 percent. For some, the appropriate reaction is “Wow!” For others, 15 percent is not much, and the appropriate reaction is “So what?”
I am on the “wow” side. For those who find 15% to be not much, I like pointing out that class of origin probably accounts for only 3%.
Murray deals respectfully with Turkheimer’s view that polygenic risk scores do not elucidate causes.
In my field, applied social science, predictive validity trumps causal pathways. The Turkheimer position about our ignorance of causal pathways is certainly correct now and may be correct for decades to come. But applied social science has never been about causal pathways (until now, it’s never been an option) and perhaps never will be. It’s about explaining enough variance to make useful probabilistic statements.
Too damn right. Astronomy came to greater public attention when Halley predicted the return of the comet which now bears his name, without having a complete understanding of the causes of gravity.
As matters stand, the environment is routinely treated by many social scientists as almost mystically complicated.
Are we interested in g×E interactions for ancestral populations? Every major ancestral population lives in every conceivable kind of environment. They live in countries in different parts of the world. Within most of those countries, they have varying socioeconomic status, varying numbers of generations of acculturation, and, for that matter, varying degrees of admixture with other ancestral populations. They live in countries that they rule and countries in which they are minorities. As minorities, they live in countries where discrimination against their ethnic group is severe and countries where it is negligible. do ethnicity and environment interact in complex ways? The natural variation in the environments where ancestral populations live is so great that the raw material for answering that question is plentiful.
This is a disarmingly simple proposal about how polygenic risk scores could be used to answer questions about the power of the environment. A version of that research can be done with admixture studies, but this should be even more convincing. Murray is also right to chide those who use complexity as a get-out-of-jail card when their chose explanations are found wanting.
So it is with human nature: The important thing is not the heritabilities of specific traits but the way that the heritability of a variety of linked traits forms an interpretable mosaic.
Murray sums up with a series of broad personal reflections. In brief, he discerns a sea change in how we perceive ourselves, moving from blank slate fundamentalism to a realistic acceptance of our biological origins and propensities, not as all-determining, but as guides to reasonable expectations. Human nature becomes an interpretable mosaic.
He is clear about his own values:
For me, what matters most is not material equality, but access to the wellsprings of human flourishing, which in turn requires that society be structured so that people across a wide range of personal qualities and abilities are able to find valued places.
Murray is strongly against the conflation of high intelligence with high human worth. Agreed, but what if brighter people are, on balance, kinder and more thoughtful in their treatment of other people? What if they are less likely to support capital punishment? What if brighter people turn out to be more altruistic? What if surgeons actually contribute more to human happiness than housewives? The terrible empirical “ifs” accumulate. There might be something other than middle class snootiness which accounts for it being better to live surrounded by middle class people.
Murray has few good words for the mass of the American upper class, other than recognising their usual charm at the individual level. Murray judges them injurious, and in contrast stresses the good, competent and likeable nature of the average citizen (with which I mostly agree), recommending the upper classes get out and meet more average citizens. This is his style, and a core value. He has no fear of being a Whig, Hernnstein the Tory, either detecting gradual historical improvements or devoutly wishing them to happen.
In “An Essay of Dramatic Poesy” Dryden was more sanguine: “If by the people you understand the multitude, the hoi polloi, ‘tis no matter what they think; they are sometimes in the right, sometimes in the wrong; their judgment is a mere lottery”.
Harsh, but we need to accept human nature: crowds can get things wrong.
Murry closes thus:
It is time for America’s elites to try living with inequality of talents, understanding that each human being has strengths and weaknesses, qualities we admire and qualities we do not admire, and that our good opinion seldom turns on a person’s talents, but rather on a person’s character. We need a new species of public policy that accepts differences and works with people as they are, not as we want to shape them. I hope this book contributes to that process.
It is very strange that an author who goes to such lengths to be kind, considered and balanced should be excoriated. Stranger still that the attacks should be so rigidly extreme when the text itself is mild, cautious and proudly admiring of the average citizen. Murray is not a polemicist: he just keeps the score, and explains his judgments. He does not eschew the correct nomenclature of digging instruments. I think he makes good calls, and if you want to see the steps in his arguments, he lays them out for you in the appendices.
The purpose and test of this book is whether it will be read. I hope so. The writing invites reading. The tone is balanced, restrained, and friendly to those for whom all this research may be news. When the topics are complicated and technical, anyone can baffle. Being legible is harder. Anyone who wants to know the score on the possible causes of sex, race and class differences will be amply rewarded in understanding if they read this book. It deserves to mark a turning point in public understanding of the biological factors in human behaviour.