I became aware of D. R. Cox during my undergraduate studies at Oxford, when I bought his book with Walter Smith on queues, an option that I found particularly opaque. Having previously been fed a strict diet of definitions, lemmas, theorems and proofs, I found the style unfamiliar, not then appreciating it as a typical masterpiece of Coxian synthesis. Its preface states that “as far as possible we have adopted an applied mathematician’s approach … the applied worker is usually quite happy to rely on his intuition here, and it rarely misleads him”—quintessential David. He was described by one of my college tutors as “that chap Cox who writes all those books.”
I met the ‘chap’ himself when I arrived at Imperial College for an MSc in Statistics in October 1980. British universities then had a system of personal tutors for graduate students, whose role was both pastoral and academic, and I had the privilege of a weekly half-hour during which David dissipated any confusion arising from lectures or problem sheets. Despite the flow of visitors to his office from all points of the globe, he rarely appeared pressed for time, though he sometimes asked me to appear as if for a meeting to hint to someone that time was up. During that year I began to fully appreciate the clarity of his writing when I read his book Analysis of Binary Data. Parts of it are so pungently written that I could still recall phrases when asked almost 40 years later to referee a paper that seemed to contradict his earlier views. For his part, David seemed to have read everything and to remember it all, so the quickest way to find something out was to ask him rather than spend hours in the library. His opinions were strong but mildly expressed, at least in public—indeed, after hearing his remarks on some novel development, it could take a while to realize that he saw it as totally misguided.
Many seminar attendees witnessed David snoozing, often in the front row, but afterwards quietly asking a question that cut to the heart of the talk. Fewer saw him fall asleep when he himself was the speaker. After flying overnight across the Atlantic he came straight from the airport to the Huxley Building, just in time to give the statistics PhD students a polished exposition of counting processes and proportional hazards. At the end he put down his chalk, and, while we were taking in the contents of the blackboard, he perched on the very edge of a table, shut his eyes, and nodded off. We gazed in consternation: should we wake him, or should we quietly withdraw, leaving him in a position from which he might fall and hurt himself? In the end someone dared to nudge him awake, he answered a few questions, and left for his office, presumably for some less perilous shut-eye.
Biometrika was edited in the interstices between other activities: commuting on the Tube provided daily slots for reading submissions (and everything else), and sandwich lunch in the tiny Biometrika office doubled as time to dictate editorial letters. David said that acting as editor was a great stimulus for research, but the volume of work became a burden for lesser mortals such as me, when many years later I took over and tried to emulate his quick turnaround of submissions and constructive suggestions to authors. His advice on tricky editorial issues was always spot-on; he saw instantly how an awkward situation could be resolved appropriately and with the least friction. When I handled his submissions after becoming editor myself, I found that he was generally unwilling to change what he had written, so his revisions barely changed—not that they needed to. He wrote so precisely and yet so tersely that checking his argument for (rare) typos could expand it by a factor of two or three.
David set the tone of the statistical community from the 1960s onward. Perhaps because he had witnessed the residual acrimony after Fisher’s titanic battles and the heated exchanges in the 1950s on foundational matters, perhaps because of his unwillingness to impose his views on others, or perhaps simply because he felt no need to do so, he was unfailingly generous, courteous, temperate, and kind, and his example led those in his orbit to behave similarly. He was endlessly supportive and encouraging at all stages of my career, always happy to chat on the phone or receive a visit. For me, and I believe for many others, he set a high-water mark, not only of academic excellence, but also of how a scientist should behave.
A. C. Davison has no financial or non-financial disclosures to share for this article.
©2023 A. C. Davison. This article is licensed under a Creative Commons Attribution (CC BY 4.0) International license, except where otherwise indicated with respect to particular material included in the article.