Dominic Cummings, a senior adviser to the UK Secretary of State for Education, recently provoked a flurry of complaints by allegedly claiming that "a child's performance has more to do with genetic makeup than the standard of his or her education." In response, he insisted that he had "warned of the dangers of public debates being confused by misunderstanding of such technical terms." Whatever you may think of that defense, it's worth looking a little more closely, because Cummings' technocratic, effectively eugenic, definitely gene-focused approach is dangerously close to affecting public policy.
Cummings has been called "arguably the most brilliant" special adviser in the UK government, but he also seems to be viewed as something of a loose cannon who has been blamed for leaks and criticized for the use of "colourful" language. He was awarded a First in History at Oxford, and has a background in activism against the UK adopting the Euro, and then controversial stints in the Conservative Party apparatus.
As an education adviser, he wrote what is called, oddly, a "private thesis" for the UK's Secretary of Education. What Cummings himself calls an "essay" runs 237 pages and is modestly titled "Some thoughts on education and political priorities" (10.3 MB pdf). It is a ... remarkable document. From the third paragraph:
Less than one percent [of English school-leavers] are well educated in the basics of how the 'unreasonable effectiveness of mathematics' provides the language of nature and a foundation for our scientific civilisation. Only a small subset of that less than 1% then study trans-disciplinary issues concerning complex systems. This number has approximately zero overlap with powerful decision-makers.
That's the flavor of the thing; basically, baffle them with bullshit. It's full of name-dropping: in the first four pages, we find mentions of Murray Gell-Mann, Thucydides, Rudyard Kipling, Dean Acheson, Carl Sagan, E.O. Wilson, Mark Zuckerberg, Linus Torvald and many more. Later on, Malcolm Gladwell is solemnly criticized because he "downplayed the importance of genes" as demonstrated in a New York Times review by Stephen Pinker. (This amuses me since I suspect that the three of them thoroughly deserve each other's company.)
What provoked the fuss about this, beyond the frisson that comes from the document having been leaked by the Guardian, was Cummings' association of genes with IQ. He does indeed qualify his discussion of heritability, noting on page 196 that heritability "is a population statistic -- it does not mean that for every individual x% of one's IQ score is accounted for by genes."
But that's not really the problem with his approach. Let's start with Cumming's full-throated endorsement of the BGI project to identify the genes for intelligence. He approvingly quotes the physicist Steve Hsu (linking to a Google Tech Talk):
Hsu says: 'I'm doing my best to increase the number of future humans who will be "fully awake". My current estimate is that one or two hundred common mutations (affecting only small subset of the thousands of loci that influence intelligence) are what separate an ordinary person from a vN [von Neumann].'
Most of Cummings' genetics comes, as he admits -- nay, boasts -- from Hsu and Robert Plomin (who is "much more cautious" than Hsu about "engineering higher intelligence"). This is summarized in a nine-page Endnote on "intelligence, IQ, genetics, and extreme abilities." The concluding sub-section on "Genetics, school achievement, 'added value' measures" includes the hackneyed trope, beloved of Victorian schoolmasters, that:
Instead of thinking about education as instruo (build in) we should think of it as educatio (draw out).
How true that is. Especially for ... oh, you guessed it. From page 76:
We know (thanks to studies such as SMPY and Terman) that although cognitive tests are noisy, we can quite accurately identify the top ~2% in IQ on whom we rely for scientific breakthroughs and even crude tests at 13 predict differences within the top 1% [italics in original] in future scientific achievements ... We should give this ~2% a specialist education as per Eton or Kolmogorov [see here and here], including deep problem-solving skills in maths and physics.
And there you have the blatantly eugenic association of genes with intelligence, intelligence with worth, and worth with the right to rule.
Steve Jones, Emeritus Professor of Genetics, critiqued Cummings based, it seems, on the first newspaper reports. What he wrote was fine, but Cummings did include enough caveats in his full document to wriggle out of the criticism and find points of agreement with Jones. Steven Rose, Emeritus Professor of Biology, offered a more detailed analysis in New Scientist, concluding:
Whatever intelligence is, these failures show that to hunt for it in the genes is an endeavour driven more by ideological commitment than either biological or social scientific judgement. To suggest that identifying such genes will enable schools to develop personalised educational programmes to match them, as Cummings does, is sheer fantasy, perhaps masking a desire to return to the old days of the 11 plus. Heritability neither defines nor limits educability.
At root, Cummings' simplistic genetics is not the issue. His technocratic eugenics emphatically is. And that deserves to be brought into the open and shredded.