Anyone who asserts that educational attainment is in large part inherited needs a lesson in modern genetics, says a professor of biology
Is intelligence genetically determined? You might think so, if you saw the headline-grabbing report from UK education advisor Dominic Cummings. The 240-page essay was a parting gift to his boss, education minister Michael Gove. In claiming that educational ability is largely inherited, he reignited an old controversy that many thought had been put to rest.
The heritability claim depends on two assumptions: that we can define and measure intelligence; and that we can unpick the contributions of genes and environment.
Attempts to measure intelligence stretch back to early last century, when the French psychologist Alfred Binet devised a series of tests for schoolchildren of different ages, to help teachers identify those who could benefit from extra help. Defining the average score for each age as 100, he then compared the performance of each child with the average for their age group to calculate the child's Intelligence Quotient, or IQ.
By the 1920s, however, tests had been developed for adults and their purpose had changed. They were now believed to provide a fixed measure by which a person's capacity could be rated against others – no longer a way of helping, but of defining.
IQ had become a surrogate measure, a linear scale along which everyone could be grouped as if by height. IQ theorists insisted that it captured a genuine and fixed unitary property, located somehow in the brain.
Elastic IQ
However, unlike height, which can be measured absolutely with a ruler, the IQ scale is more like a piece of elastic, to be stretched to fit social expectations and adjusted accordingly. Despite attempts to develop tests that apply across all cultures, they inevitably reflect assumptions about what constitutes "correct" answers and about how people should perform.
Thus test questions are devised to ensure that on average boys and girls score similarly, but when in the same tests middle class children scored better than working class, and in the US white students outperformed African-American, this was assumed to reflect real differences. It didn't.
Unsurprisingly, IQ test scores correlate with school performance (and with parental socio-economic status), because this is what they are designed to do, but they only weakly predict a person's future career success. This, as critics were quick to point out, is because IQ doesn't capture such elusive features as practical intelligence, creativity, emotional intelligence, social intelligence, musical or artistic talent… the list goes on.
Less rigid psychometricians began to speak of multiple intelligences. Furthermore, neuroscientists, well aware that the processes involved in intelligent behaviour must include perception, attention, memory and reaction speed, have always been sceptical that it can be reduced to a single measure or brain process.
Does intelligence, however we define it, depend on genes? The answer is trivially obvious: yes, absolutely – as does everything else in a living organism. But, as with everything else, it also depends on the environment within which the developing child grows. The question which has obsessed genetic determinists is whether it is possible to partition out the effects of genes and environment.
Heritability equation
Before modern genetics, researchers developed an equation, dubbed heritability, which attempted to do just this. As with IQ, heritability isn't an absolute measure. Instead, it describes the contribution that genes and environment make to the variance of some trait – for instance IQ – around the mean for the population. To put it in the form of an equation, if V is the variance, G the genetic contribution and E the environmental one, then V = G + E + (GXE). Put simply, genes and environment are supposed to work additively with a small component (GXE) for their interaction.
It is from this formula, and comparisons of IQ scores between identical and non-identical twins, that psychometricians have by and large settled on a figure of 50 per cent for heritability. Robert Plomin, Gove's behavioural genetics advisor and a prominent spokesman for this technique, puts it higher, at around 70 per cent. This is the figure cited by Cummings.
However, the calculation is almost meaningless. It depends on there being a uniform environment – fine if you are studying crop or milk yields, where you can control the environment and for which the measure was originally derived, but pretty useless when human environments vary so much.
Thus some studies give a heritability estimate of 70 per cent for children in middle class families, but less than 10 per cent for those from poor families, where the environment is presumably less stable. And it is a changing environment, rather than changing genes, which must account for the fact that the average IQ scores across the developed world have increased by some 15 points over the past century, to the puzzlement of the determinists.
Untenable assumptions
But even more importantly, the calculation only works if the interaction between genes and environment, GXE, is small. Forty years ago this might have been a reasonable assumption. But despite the efforts of Plomin and his colleagues, it has become untenable in the light of the new genetics that arose from the sequencing of the human genome.
If you would like to reuse any content from New Scientist, either in print or online, please contact the syndication department first for permission. New Scientist does not own rights to photos, but there are a variety of licensing options available for use of articles and graphics we own the copyright to.
Have your say
Only subscribers may leave comments on this article. Please log in.
Only personal subscribers may leave comments on this article
All comments should respect the New Scientist House Rules. If you think a particular comment breaks these rules then please use the "Report" link in that comment to report it to us.
If you are having a technical problem posting a comment, please contact technical support.
