This article originally appeared in The Skeptic, Volume 7, Issue 2, from 1993.
When Sir Cyril Burt died in 1971, aged eighty-eight, he was Britain’s most distinguished psychologist, loaded with honours including the first knighthood conferred on a psychologist. Five years later he was accused of gross scientific fraud: inventing figures, research, and even collaborators to support his ideological views.
From 1913 Burt was psychologist to the London County Council, and collected a vast amount of data on schoolchildren. In 1932 he became Professor of Psychology at University College London, and achieved great renown and many distinctions. He retired in 1950 but remained busy, writing, reviewing, examining, lecturing and editing the British Journal of Statistical Research.
Burt was best known for his work on intelligence, which he believed to be innate and largely inherited. His main evidence came from separated identical twins. Since they are genetically identical, differences could only be environmental, and the finding that separated pairs are almost as alike as those brought up together was a strong indicator that heredity played by far the greater part in intelligence determination. This, of course, has enormous political and social implications and towards the end of his life Burt became a controversial figure as it became fashionable to maintain that children were of equal natural endowment and their huge observed differences were mainly due to social inequalities rather than inherent differences. Burt responded with typical energy and combativeness, including contributing to the ‘black papers‘, which criticised modern educational methods and called for a return to more traditional ones.
Some doubts were expressed about his work soon after his death by two very different American psychologists: Arthur Jensen, who was much vilified for saying that programmes intended to boost black and working-class children had produced no significant improvements and were a waste of time and money, and Leon Kamin, whose book The Science and Politics of IQ (Penguin, 1977) is a bitter attack on intelligence testing from a left-wing viewpoint. The event which really opened the floodgates, though, was the publication on the front page of the Sunday Times of 24 October 1976 of an article by its medical correspondent, Oliver Gillie, beginning:
The most sensational charge of scientific fraud this century is being levelled against the late Sir Cyril Burt, father of British educational psychology.
The suggestion arousing the most interest was that the names Howard and Conway appearing on items in his journal were Burt inventions. Gillie reported his fruitless search for the ‘missing ladies’, and later said:
Fudged statistics might make a newspaper story, but non-existent people, invented in order to perpetrate a fraud, really catch the popular imagination … this began to look like a big story.
Professor Leslie Hearnshaw of the University of Liverpool had meanwhile been invited by Burt’s sister to write his biography. “At this time” he said, “my assessment of Burt and his work was almost wholly favourable”. When the accusations burst forth the nature of his work changed dramatically. “Gradually, as evidence accumulated… I became convinced that the charges… were, in their essentials, valid…”, says Hearnshaw in Cyril Burt, Psychologist (Hodder, 1979), and it did indeed seem to establish this beyond all reasonable doubt.
The British Psychological Society held a symposium at which it was treated as proven fact: it was entitled ‘The Burt Scandal’ , and the report of the council meeting which decided to hold it mentions Burt’s ‘fraud’, ‘falsifications’ , ‘deceptions’ and his long-time colleague Charlotte Banks was discouraged from attending by the President himself, apparently for fear of offending Gillie, who was invited to speak despite not being a psychologist or a BPS member.
Burt’s reputation declined dramatically, and some said he had always been a villain and all his work must be discarded as worthless. Hearnshaw, however, had not gone anything like so far, saying that only the later work of a long and productive life was suspect. Burt’s guilt became generally accepted and has often been asserted as established fact. The first substantial suggestions that Hearnshaw’s conclusions might after all be questionable came from 1987 onwards in articles by Ronald Fletcher, Emeritus Professor of Sociology at the University of Reading. He wrote a book, which a publisher accepted but ‘sat on’ for fear of litigation.
Then a British psychologist, Robert B. Johnson, produced The Burt Affair (Routledge, 1989), demonstrating that Hearnshaw’s work is deeply flawed and cannot be relied on. He checked Hearnshaw’s statements about Burt’s allegedly false accounts of his early works, said to be intended to enhance his own role and diminish others’, and found the allegations ill-founded. He then examined the whole case, and summarises his conclusions thus:
…the evidence which has so far been presented is insufficient to support the accusations which have been made… the gross misbehaviour of which [Burt] has been so widely accused has not, in my opinion, been established. A grave injustice has been done.
Hearnshaw’s case certainly seems overwhelming: as one reviewer, Maryson Tysoe, in New Statesman and Society, said: “Every time Johnson recounts a particularly damaging charge, the reader wonders, how the hell is he going to talk Burt out of this one?” But his explanations fit the facts and demonstrate that the case cannot be regarded as proved.
There now seems no doubt that the ‘missing ladies’ existed. They and others probably collected data for Burt at the LCC and University College. But during the wartime evacuation much was misplaced, and when Burt tried to work the data into publishable papers he found a lot was missing. However, useful material was sometimes rediscovered, and Burt, not wishing to ask his secretary to type a paper implying she had mislaid it for years, used it without mentioning its loss and recovery and, having promised to credit those who collected it, added their names. It is common for academic papers to bear the names of people who did no actual writing; indeed, Hans Eysenck wrote up the data of a research assistant who was dead, perfectly properly putting her name to it with his own.
Johnson demonstrates that parts of Hearnshaw’s book are an appalling mess of distortions, wrong and misunderstood references, and total misunderstanding of what Burt was saying. The prize exhibit is a passage presented as a direct quotation from a newspaper. No such item can be found, and when Hearnshaw was asked he replied that it came from a secondary source, he could not remember what it was, and could find nothing relevant in his files. As Johnson says, he had not bothered to check:
It was not merely second-hand evidence… It was third-hand evidence…Yet he had quoted it as if he had seen it himself. If Burt had done this … what would the critics have made of it?
Fletcher’s book was eventually published as Science, Ideology and the Media: the Cyril Burt Scandal (Transaction, 1991). As the title suggests, Fletcher is mainly concerned with ideological and media influences on scientific controversy. He was spurred into action by the Horizon programme, The Intelligence Man (1984) which took Burt’s guilt for granted and included dramatisations depicting him as devious, dishonest, overbearing, dogmatic and egocentric, totally at variance with the testimony of people who knew him well. The book is in legalistic form: putting the case, asking questions of Burt’s critics (Hearnshaw, alas, cannot reply; he died in 1991), and inviting one to return a not guilty verdict.
One of the first matters queried in Burt’s work was ‘invariant correlations’ – where Burt reported identical or almost identical agreement between results despite sample size increases, which is considered virtually impossible. The implication is that Burt believed this would give his conclusions a spurious stability and consistency. But if Burt had so believed he might have been expected to draw attention to the invariances; he didn’t, and it seems that no-one noticed them until after his death. In any case, he would have been incapable of any such belief: he was extremely well-informed on statistics and would have known that, far from strengthening his case, they would have looked very suspicious. Also, they predominate in physical measurements – height, weight, etc. – not those of intelligence.
Having decided Burt was guilty, Hearnshaw suggested that his actions resulted from mental illness, and lists a number of ‘blows’ which, he believed, tipped him over the edge: the loss of research material to enemy bombs; a failed marriage; a dispute with his old department; fierce attacks on his views; ill heath; and the loss of his editorship of the British Journal of Statistical Research. But the loss of material probably wasn’t as serious as suggested; he lost the editorship when he was 80 and, if Hearnshaw was right, already far gone in deceit; although no doubt foolish to try to interfere in his old department after retirement, he had good reason to feel aggrieved; the old often have poor health; and they frequently find their views have become unfashionable.
As for the parting with his wife, Johnson says that for all we know Burt may well have been glad to see the back of her – causing one reviewer, Tom Wilkie in The Independent, to say in shocked tones that his “extraordinary tirade on matrimony left me open-mouthed”. He was missing the point. It’s possible that the marriage failure was a bitter blow, as Hearnshaw maintains; Johnson’s point is that we simply don’t know. As it’s just as possible that he was indeed relieved when it ended. Here as in many other places Hearnshaw simply chooses the interpretation which suits his case and ignores other possibilities.
Also, it’s just as easy to compile a list of successes: the knighthood and many other honours; popularity with students and colleagues who on retirement gave him the biggest party the college had ever catered for; huge audiences at his lectures; important broadcasts; and general fame and renown. Hearnshaw also suggests that at a rough London school Burt developed a defensive ‘gamin’ temperament which reappeared much later. Behind the facade of the polite conventional academic was a naughty schoolboy delighting in defeating his enemies. This is pure speculation: we know very little of his early childhood, not even which school he attended.
The Burt affair is significant for the study of paranormal beliefs for two reasons. It is an excellent example of how thorough research can totally demolish conclusions which appear to be established beyond all reasonable doubt; and, much more important, it is a classic case of the way in which people who form a conviction have no problem in finding evidence to support it and ignoring all contrary evidence. Hearnshaw decided Burt was guilty, suggested mental illness as an explanation, and found causes for it. But his list of ‘blows’ was compiled solely for that purpose. If no-one had ever questioned Burt’s probity, would anyone then have solemnly listed them and asked how he stayed sane?
Johnson’s and Fletcher’s books total over 700 pages, and overlap little. I think any fair-minded person who reads them can only conclude that the smashing of Burt’s reputation was a bitterly tragic injustice.
