Professor Bernard Silverman
Laureator: Professor Rosemary Bailey
Friday 27 June 201
Chancellor, it is my privilege to present Professor Bernard Silverman for the degree of Doctor of Science, honoris causa.
Bernard showed his mathematical colours early by winning a gold medal at the 1970 International Mathematical Olympiad, the only person in that year from this side of the Iron Curtain. He then studied Mathematics as an undergraduate and Statistics as a postgraduate at Cambridge. After a short spell in industry, where he designed the first ever programmable pocket calculator, the Sinclair Cambridge, he began his academic career at the University of Oxford. He soon moved to the University of Bath, where in due course he was given the first ever personal professorship in the history of that university, at the age of 32, and became Head of the School of Mathematical Sciences, helping to build Bath into a university with an international reputation in mathematics. This was followed by ten years in senior positions at the University of Bristol. In 2003 he returned to Oxford as Master of St Peter’s College.
He has been active in many professional societies. The Royal Statistical Society awarded him the Guy Medal in Bronze in 1984, and the Guy Medal in Silver in 1995. He was elected Fellow of the Royal Society in 1997. He served as president of the US-based Institute of Mathematical Statistics from 2000 to 2001.
Bernard’s research in statistics has been wide-ranging, both methodological and applied, with many collaborators. At the core has been the development of new methods of analysing data that would have been unthinkable before the rise of computer power during our lifetime. In density estimation and in non-parametric regression the aim is to estimate a whole curve, rather than just a few parameters. When we move on to functional data analysis, the data themselves are curves, not just a list of numbers. This research requires a deep understanding of the underlying mathematics; a thorough engagement with the statistical ideas of obtaining good but incomplete data and using them to make good estimates of underlying mechanisms; technical knowledge of the computational aspects so that the methods can be implemented and used by other people; and an awareness of other fields of enquiry where the methods may prove useful.
As well as developing difficult theory, he has led the way in showing the applications of his methods. His published papers include work on cot death, on materials science, on positron emission tomography, on how ants build their nests, on the pathology of mediaeval bones, and on human genetics. When I started collaborating with ecologists on biodiversity, one of the first references that they showed me had him as a co-author. He has carried out a wide range of consultancy work for industry and for government. He took part in an enquiry into the 2001 foot-and-mouth epidemic and in a report into the science of genetically modified crops.
In January 2010 Bernard became President of the Royal Statistical Society. He stepped down the following month, on the announcement of his appointment as Chief Scientific Adviser to the Home Office. This was to avoid a perceived conflict of interest, as the Royal Statistical Society issues expert reports from time to time on statistical aspects of public life, such as benefit fraud or clinical trials. Indeed, in 1990 Bernard had himself been a co-author of the Royal Statistical Society’s report on the UK’s official statistics.
Of course, Bernard took up his Home Office post precisely because he wanted to put his mathematical and statistical expertise to good effect in a range of areas that affect us all. He is meticulous about collecting data and using it as the basis of evidence: for example, how do you sample cross-border passengers in an unbiased way? Once you measure queues accurately, the next step is to use those measurements as the basis for action, and his work on the application of queueing theory and algorithms has made a big improvement in the time it generally takes us to cross the border and in the way that resources at the border are managed. It was instrumental in the smooth running of the 2012 Olympics.
Bernard’s responsibilities range from social research, statistics and economics to questions in biological and physical science relevant to fighting crime and preserving national security; they also include the scientific aspects of drugs, and the regulation of animal experiments. On a completely different issue within his very wide brief, the Protection of Freedoms Act 2012 brings the rest of the UK into line with Scotland in that DNA profiles on those arrested but not convicted cannot be retained indefinitely. This is called the ‘Scottish model’ and the English have had to adopt it. Bernard was personally involved in Home Office research on the subsequent behaviour of arrestees which led to the conclusion that it was fair to keep their DNA profiles for three years, but not longer, so that is now in the Act. The team which manages the National DNA database reports to him, so he has had to make sure that this policy is implemented, which has meant the removal of over a million records from the database.
Bernard’s most recent project is in connection with the new legislation on Modern Slavery. The hope is to use the disparate range of data – for example from the police, from social services, from voluntary organisations – to build up a full picture of the scale and nature of the problem, including the ‘dark figure’ of cases that never come to our attention. This is a crucial aspect of tackling this issue in the longer term.
I cannot over-emphasise how important it is to have someone with his level of statistical expertise and general scientific awareness working for the Home Office.
Chancellor, in recognition of his major contribution to statistics, I invite you to confer on Professor Bernard Silverman the degree of Doctor of Science, honoris causa.Awards