Professor George Barnard, emeritus professor of mathematics at the University of Essex and Patron of Humanists UK, died in July 2002 aged 86.
Barnard was born and went to school in Walthamstow, east London. His father was a cabinet maker and his mother had been in service. He won a scholarship to St John’s College, Cambridge, where he read mathematics, and then from 1937 to 1939 was at Princeton University studying mathematical logic. At the end of the war, he went to Imperial College, London, where he became professor of mathematics in 1954, leaving in 1966 for the newly created University of Essex, from which he retired in 1975. He then spent much of each year, until 1981, at the University of Waterloo, Canada.
His work was marked by its range, from the theoretical to the practical: one paper might be on a statistical study of the strength of condoms, with the next on some arcane point of inference. He often returned to the mathematical logic of his youth while applying ideas to quality control, for which he was awarded the Deming prize for the American Society for Quality Control in 1991. He was the first to appreciate what has come to be known as the likelihood principle, which says roughly that, in assessing some data, one should pay attention to the possible explanations for the data, rather than, as many statisticians still advocate, seeing how the data stands in relation to other data that might have occurred but did not. This important concept was to have an enormous influence on statistical thinking and practice.
He was always anti-establishment and critical of religions. In argument, he was less interested in defending his position than in arriving at the truth: at a dinner of the Royal Statistical Society in the 1970s when a young member of the audience interrupted him: “That was not what you said 10 years ago,” Barnard thought for a while and then responded, “Good, that shows that I have learned something in 10 years.”
Based on his obituary in The Guardian, 9/8/02.