Published on:

A crucial new book by a pioneer of behaviour genetics

My Review in The Times of Robert Plomin’s new book:

For a long time there was an uncomfortable paradox in the world of behaviour genetics. The evidence for genes heavily influencing personality, intelligence and almost everything about human behaviour got stronger and stronger as more and more studies of twins and adoption came through. However, the evidence implicating any particular gene in any of these traits stubbornly refused to emerge, and when it did, it failed to replicate.

Ten years ago I recall talking to Robert Plomin about this crisis in the science of which he was and is the doyen. He was as baffled as anybody. The more genes seemed to matter, the more they refused to be identified. Were we missing something about heredity? He came close to giving up research and retiring to a sailing boat.

Fortunately, he did not. With the help of the latest genetic techniques, Plomin has now solved the mystery and this is his book setting out the answer. It is a hugely important book — and the story is very well told. Plomin’s writing combines passion with reason (and passion for reason) so fluently that it is hard to believe this is his first book for popular consumption, after more than 800 scientific publications.

His story is crucial, because in the final chapters he exposes his own genes to readers as a test of the arguments he is making. So we learn that Plomin, a professor of behavioural genetics at King’s College London, “grew up in a one-bedroom flat in inner-city Chicago without books”. Nobody in his family went to university, yet he was an insatiable devourer of books.

An intelligence test identified Plomin’s ability and got him into schools where he could develop his talent. Here lies one of the sources of his passion: he thinks that if children are to be enabled to fulfil their potential, then you cannot believe that they are the product of their upbringing or education. You must understand that they have innate aptitudes that can overcome environmental disadvantages. Nothing, he believes, is bleaker than environmental determinism.

Plomin’s research on twins and adoptees has relentlessly proved the truth of this assertion, so long denied by the dogmatists of the “not in our genes” era. Five key insights emerged, some so counterintuitive as to leave your head spinning.

First, most measures of the “environment” show substantial genetic influence. That is, people adapt their environment better to suit their natures. For example, Plomin discovered that the amount of television adopted children watch correlates twice as well with the amount their biological parents watch rather than with the amount watched by their adoptive parents.

Plomin hesitated before publishing this remarkable finding on the “nature of nurture” in 1989. Knowing what had happened to anybody who discussed genes and behaviour, from EO Wilson to Charles Murray, Plomin realised that telling the world that television watching habits are genetically influenced would be ridiculed by social scientists and the media, however strong his evidence. He feared it would be professional suicide. Yet his insight has since been replicated more than 18 times.

Our personalities are also influenced by the environment, but Plomin’s second key insight is that we are more influenced by accidental events of short duration than by family. Incredibly, children growing up in the same family are no more similar than children growing up in different families, if you correct for their genetic similarities. Parents matter, but they do not make a difference.

Plomin says these chance events can be big and traumatic things such as war or bereavement, but are mostly small but random things, like Charles Darwin being selected for HMS Beagle because Captain Robert Fitzroy believed in “phrenology” and thought he could read Darwin’s character from the shape of his nose. Environmental influences turn out to be “unsystematic, idiosyncratic, serendipitous events without lasting effects”, says Plomin.

Moreover, surprisingly, heritability increases as we get older. The longer we live, the more we come to express our own natures, rather than the influences of others on us. We “grow into our genes”, as Plomin puts it. An obvious example is male-pattern baldness, which shows low heritability at 20 and very high heritability at 60.

Two other findings are that normal and abnormal behaviour are influenced by the same genes, and that genetic effects are general across traits; there are not specific genes for intelligence, schizophrenia or personality — they all share sets of genes.

This last point leads to the breakthrough in identifying which genes make the difference. The first attempt at finding genes that influence behaviour and psychology made use of the “candidate-gene” approach. Find a gene that might be involved and see if it matters. With few exceptions, such as the APOE gene and Alzheimer’s, this approach was a dismal failure. The results were sparse and failed to replicate.

Along came the genome-wide array technique: to search for lots of different mutations at the same time in a large sample of people, hoping to pick up subtler effects. Again, nothing: Plomin’s first try yielded no genes associated with intelligence. Then came the first gene-chips in the early 2000s and he was able to look for 10,000 mutations at the same time. Still nothing. “I was beginning to think my luck had run out — after a decade of work, this was the third false start.”

The problem was that everybody thought they were hunting big game — genes with hefty influence on particular traits. It turns out they should have been looking for much smaller quarry: genes with very slight influence, but many more of them. We now know that you need a sample size of 80,000 people before you can detect the very slight changes that each genetic mutation causes, but when you get to such a scale, you find thousands of relevant genes, each adding only a small percentage to the chance of having a particular trait. It’s gold dust, not nuggets.

However, the effects are additive, and once you have lots of genes, you can start to explain significant portions of the variance among individuals. Plomin illustrates this with height. Being, like me, 6ft 5in tall, he is not surprised to find that his polygenic score, based on thousands of genes, puts him at the 90th percentile for height. The genes in the sample so far only explain 15 per cent of the variance in height of individuals, which may not sound like much, given that height is 80 per cent heritable in western societies. But get this: it’s a better predictor than any other factor — such as the height of the parents, or the height of the person as a child, let alone medical history or socio-economic status — and it works from birth, or even conception.

Plomin is very interested in the possibilities of polygenic scores, which will make it possible partly to predict psychological traits such as depression, schizophrenia and educational achievement. The score is the result of passing a sample of your blood over a silicon chip that tests for thousands of mutations, then adds them together, giving an aggregate score for how many of the thousands of single-letter code changes you have that each very slightly makes you more likely to do well in school, for example.

The predictions such scores give are probabilistic, not certain, but they are improving. Plomin argues that genes can probabilistically predict things about an individual, distinguishing her from her siblings, and can do so from the start of life. School attainment is now better forecast by a polygenic score than any other way of predicting it — it is better than knowing how the parents did at school, better than socio-economic status, better than the type of school (which turns out to have little effect once you control for the fact that selective schools choose innately more talented children).

Plomin thinks parents who give a newborn child such a test and find out that no matter how hard the child is helicopter-parented he is unlikely to be a genius would probably be doing that child a favour. “Parents should relax and enjoy their relationship with their children without feeling a need to mould them,” he argues.

It’s far fairer, Plomin says, to find out what children will be good at and bring that out than to be able to create inequality based on income or opportunity. And, in a point he does not emphasise enough, the fact that intelligence or personality are caused by many thousands of genes, each of minuscule effect, means that it will be impossibly difficult to create a super-intelligent designer baby.

I had only one quibble with this mind-blowing book: the title. I hate the word “blueprint” in association with genetics, for two reasons: first, it is an anachronistic metaphor relating to a technology that nobody uses any more. Second, it gives a misleading impression that each part of an adult is determined by a different set of genes, as is the case with an architectural blueprint, in sharp contrast to the genes of general effect that Plomin is so careful to identify. We are cakes baked to a recipe, not buildings assembled to a blueprint.

By Matt Ridley | Tagged:  rational-optimist  the-times