A speech about … me

The following speech was spoken by Aad van der Vaart (boss, colleague, successor, and most importantly, friend) as first official reply to my farewell  lecture.

Mr. Rector, Ladies and Gentlemen, Dear Richard,

It is about 35 years ago that we first met, when I was a Leiden student, or maybe when I applied for a PhD position at CWI, where you were the chief statistician, the head of the Department of Mathematical Statistics, in the middle of the 1980s. I ended up doing my PhD at our beloved Leiden University, but I still had the benefit of your presence, both because CWI at that time was the national meeting place for statistics and because you were a one-day-a-week professor at Leiden. I vividly remember the course you gave, centred on Hadamard differentiability, and the enthusiasm and inspiration that came with it. It gave a window on scientific facts known and on things to be discovered. You have been an inspiration ever since.

It is my pleasure to speak here today both as a younger colleague, and in the more formal role of scientific director of the MI, as your last boss so to say. Let me slip into the second role for a moment and list some basic facts of your career. Your training as a mathematician and in statistics was at Cambridge University, but you wrote your PhD thesis in Amsterdam, having followed the love of your life to the Netherlands. Kobus Oosterhoff, my later boss at the Vrije Universiteit, acted as your advisor, and more than once he told me that he had taught you English, while you had taught him mathematics. After completing your PhD, you remained at CWI, but accepted a one-day-a-week professorship in Leiden, where you gave an inaugural lecture under the title “Missed Chances” in 1987. Then in 1988 you were appointed professor of Mathematical Stochastics at Utrecht University. It was only in 2006 that you returned to Leiden University as professor of Mathematical Statistics, thus taking the opportunity to give your valedictory address today in exactly the same room as your inaugural lecture. A lot happened in-between. A bit of trivia is that in 2006 you were the first successor of Sara van de Geer, who had left for ETH, while Sara was your first PhD student, graduating in Leiden. During all those years you spent sabbaticals in Denmark (twice; to learn about the real world, as we heard today from NK, a type of world not available locally), Germany, Australia, and Cambridge, and at the National Institute of Advanced Studies for Social Science, in Wassenaar as a Distinguished Lorentz Fellow. 16 students wrote a PhD thesis under your guidance, and there I count only the ones for whom you were the true advisor, not the number of signatures you put.

Your scientific work is extraordinary broad. In the workshop we ran during the past week we had a taste of the very diverse subjects, represented by scientists from equally diverse communities. It was tough at times, and you were probably the only one who could really appreciate all the talks. There is a rough time order in your interests, and we could divide up your career in the young Richard, the middle Richard and the late Richard. As I realise “the late Richard” does not sound so good in English, let me say Richard I, Richard II and Richard III. During the conference week we have seen photographs from these stages, even one photograph showing you doing intense mathematical research while showing off a naked torso, and, alas time passes, and maybe young and old are also adjectives that could have come to mind.

As Richard I you were concerned with survival analysis and causal inference. It includes your thesis work on the application of stochastic calculus in survival models, which became a well cited book, and it includes what you yourself have described as the “weight-lifters guide to survival analysis”, a beautiful account of the models and methods used to analyse stochastic processes of events in time, widely used to study the effectiveness of medical treatments, and 25 years after its publication still an important reference. A Dutch colleague once described your work to me by confiding “Richard Gill, every time there is no method available to answer a question, then not only does he invent one, but he also builds a completely new mathematical theory for the situation”. Well I suspected that he had no clue about statistics, but the admiration expressed in this description I found hugely impressive. And it is very justified.

As mathematicians we are usually not so very impressed by numbers, I mean counts, of papers, citations, etc., but rather value quality and depth. There is an ample amount of these in your work, but in my role as administrator let me mention a bit of trivia again. In a recent report where the University of Leiden investigates the factors that determine the ranking of the university, an important issue these days, I saw Richard’s name come up on the fairly short list of researchers who make a difference, being an ISI highly cited author. A little footnote was added to his name, though, as in the particular ranking this counted towards the fame of Utrecht University, not Leiden University. Shame on Richard for being a Leiden professor before and after, but not during.

The interests of Richard II came about, I think, as a conscious choice to do something totally different, and of course through genuine interest in the wonders of quantum probability and statistics. Richard was quite ahead of his time. This is something I, and many of the colleagues, have come to understand too late, even sitting through many introductions to Bell inequalities and other strange facts, which often seemed far removed from ordinary statistics. It is not easy to change a successful career midway, but you managed. One achievement that materialised recently was the development of the statistical methods used to analyse the entanglement experiment in Delft. Quantum information is everywhere now, and I wished I had paid more, or better at least, attention.

The interests of Richard III arrived certainly not as a choice, but quickly became a moral obligation. It started with the case of the nurse, who was not a killing nurse, of course, but a nurse in need of help. It was intense and a time full of emotion. Getting involved was a duty attached to being a scientist, a principle that you did not apply only to yourself, but also pointed out to your colleagues, often. The involvement in this particular case developed in interest in statistics and the law in general, and scientific integrity, with concrete cases as deaths in a clinical trial on probiotics to DNA evidence in murder cases, scientific fraud or questionable research practices. One particularly big case was connected to the UN tribunal for Libanon, which was all work done in secrecy, but I do remember the time that an impressive London barrister made a powerful speech in our Snellius building pointing out the importance of the newly developed statistical techniques for international law. Now the big crooks would be caught. Richard himself came under attack too. He may be the only mathematician in Leiden to have been advised by university lawyers how to avoid huge claims, or to have appeared before the integrity committee to fend off accusations of having damaged the integrity of a very well connected colleague. It is admirable how you coped under all the pressure, with difficulties, but still.

We do not know Richard as being extremely well organised. In presentations enthusiasm and inspiration may win from finishing in time. Inspiration, brilliance, creativity are qualities that come sooner to mind than being totally prepared. Nevertheless, a look at Richard’s cv reveals an impressive number of organisational achievements. Numerous program committee memberships and chairmanships, of really big conferences, president of the Netherlands Statistical Society, scientific secretary of the Bernoulli Society, numerous editorships, council member of the IMS, twice, liaison officer, chair of the examination committee, and many, many more. It is really a very long list. I would also mention that the Statistical Science master, which we run since some six years, was conceived from his office, albeit in a joint effort.

And not to forget, while at some time Richard was the only statistician in the MI, look at the big group that he is leaving behind.

Will you leave it behind? I hope not. The renaming of “the late Richard” to Richard III leaves many possibilities for the future. Some of the speakers already expressed it this week: Richard, we continue to count on your presence and your brilliant mind.