A. R. Jonckheere

Psychologist/statistician for whom ideas were all

The modern British university is a production line for advanced training and the creation of new technical knowledge. It was not always thus. Fifty years ago in its better forms it could still be a community of scholars. A. R. Jonckheere was a splendid representative of this type of university.

Aimable Robert Jonckheere was born in 1920 near Lille in the north of France to a French father of Belgian origin, a businessman turned astronomer, and an English mother. When he was seven his parents split up and then divorced. His mother returned to England with his two older sisters, but a year or two later returned and brought him over too, to live in Enfield. On leaving Edmonton County School he became apprenticed to a firm of actuaries. He was also a pacifist. So, on the outbreak of the Second World War, he had to leave his actuary position and became a farmworker in Jersey. He was forced to stay there during the war in near-famine conditions, discussing German literature with officers in the German occupation force while at the same time helping to protect an escaped Russian prisoner of war who had swum from Guernsey where occupation was harsher.

Under some special post-war arrangement, on his return to England he was able to enter University College London, where he was to remain all his working life, and indeed was planning to come and teach on the day before he died. He took Psychology but his studies were far more eclectic. Influenced by J.B.S. Haldane, he built on his actuarial training to obtain a deep understanding of statistics. However, possibly the greatest intellectual impact on him was A.J. Ayer's seminars. He absorbed Ayer's critical style perhaps too thoroughly and applied it in all domains of knowledge that he encountered.

In the early 1950s with his statistical background he turned to psychometrics, the study of the measurement of personality and intelligence, and acted as a statistical collaborator for both Hans Eysenck and Cyril Burt. Burt is a discredited figure now, but technically he was highly competent. He was Editor of the British Journal of Statistical Psychology, where he was later to be followed by Jonckheere.

At the time Burt was working on his famous data, or better infamous data, on the intelligence of identical twins reared apart, which for many years provided the cornerstone of the idea that general intelligence, "g", has a very high genetic component. Although Burt did not involve Jonckheere in the project he frequently mentioned that Jonckheere must meet his (Burt's) collaborators, such as "Miss Conway", who have never since been identified. Unsurprisingly, Jonckheere never actually met them and they appear to have been as much a figment of Burt's imagination as the data on which he worked.

This was the time in which Jonckheere married his first wife, Joy. (They were to be divorced in 1968.) Intellectually the period culminated for Jonckheere in his development of a new statistical test for detecting trends in so-called categorical data. The test is now extensively used and indeed is incorporated in the most widely adopted statistical package for the behavioural sciences, namely SPSS. During this period Jonckheere was also working on a doctorate, which, naturally for him, was in a different field, namely animal psychology.

While Jonckheere retained his interests in the technical aspects of psychometrics, as a psychologist he had become much more interested in how cognitive processes actually develop in the child, rather than how they are measured in some fairly crude way. Being bilingual in English and French, he was invited to spend the academic year 1956-57 in Geneva at the International Centre for Genetic Epistemology. There he collaborated with Jean Piaget and Benoit Mandelbrot on how children acquire concepts of probability.

Piaget was so taken by Jonckheere that he refused to allow him to leave until he could find someone equally clever. Jonckheere's response was to find the brilliant South African émigré Seymour Papert, later to prove a theorem that held up the development of neural network modelling for 10 years.

On his return to London Jonckheere developed a style of organisation of his working life that changed relatively little over the years. He had been made a lecturer in 1951 with special responsibility for lecturing on statistics and animal psychology. By the late 1950s he was lecturing on a wide variety of other topics too. His lectures were much appreciated, but these days would be considered dubious in approach. Outside statistics, he often lectured without notes and indeed without much direct preparation. One of his favourite procedures was to take a paper recently produced on the topic of the lecture and held to be influential, to explain it very clearly, paying particular attention to the methods used; he would then demonstrate how in a variety of ways its conclusions were not justified given the results and methods. This he did to even more effect in the seminars he ran weekly for junior staff and postgraduates where a paper would be dismembered in anywhere from two to three hours.

Possibly, though, his greatest influence was as intellectual adviser on issues to do with design and analysis of studies and with their conceptual underpinnings. He was regularly consulted by a range of researchers who worked in fields from education to human reproduction to the visual arts, from beginning postgraduates to figures such as Basil Bernstein and Ernst Gombrich.

Even as a student, one knew that provided you had a problem of sufficient intellectual interest you could go to discuss it with "Jonck" and he would devote hours of his time, unless the pressure of the queue of other people wanting to consult him, who could even include someone like Giacometti, got too great. If you were sensible you limited the issue discussed; otherwise all parts of your grand conception might start to be unravelled.

In many of these discussions he would make a contribution of sufficient importance that it would be normal academic practice for it be acknowledged by his being an author of the eventual paper. This was complete anathema to Jonckheere. Having his name on a paper would imply intellectual responsibility for other aspects of the work over which he had had no control. Indeed he is said to have been very angry when a tough and clever American colleague put his - Jonckheere's - name as first author, as he deserved, very deliberately without sufficient consultation. (Later he became very proud of the paper.) The result of this attitude was that the few papers he published later in his career were highly technical.

A second result was that, when highly appreciative junior colleagues attempted to have him put up for a chair, the idea was quashed on grounds of productivity, even though indirectly his input must have spawned hundreds of papers.

The one person who never seemed too bothered about this was Jonckheere himself. He was a true intellectual, little concerned with personal advancement. What he was interested in was ideas, and how ideas may be adequately assessed. And the ideas which he analysed ranged over science, philosophy, the visual arts, music and literature, in all of which he was deeply absorbed. And, ideally, the analysis was done collectively with others. Work merged seamlessly into pleasure and resulted in innumerable friends.

Prime among them was the philosopher Sophie Botros with whom he lived very happily and whom he later married.

Tim Shallice