At a meeting of the British Mathematical Colloquium Scientific Committee last Wednesday, John Hunton remarked that he was on the EPSRC SAT for mathematics, and offered to lead a discussion of recent EPSRC policy. I found this very informative, and I am grateful to John (who said that what he was reporting from the EPSRC side was not confidential) for doing this, and for agreeing to take a look at what I wrote before I posted it. Of course, he is not in any way responsible for the content, even though it has been greatly improved by his comments. (As an aside, I think that John does overestimate the effect that this post will have. Even in my department, my views are not taken very seriously!) In particular, nothing in this document should be attributed to John except where I have said so.
There is a general feeling that EPSRC has lost the confidence of the mathematics community (and, from what we hear, that of other parts of science as well). This is of serious concern to all of us. John did his best to address this feeling.
This post contains a lot of acronyms and jargon. This is quite usual for funding bodies; I have appended a short glossary at the end. For the rest, I will start with some background on the body which was meeting and the back story of EPSRC, and then discuss what I learned from John. I have mostly not reported the comments of the audience (except mine, at one point).
The BMC Scientific Committee
The British Mathematical Colloquium is the only annual nationwide meeting on research-level pure mathematics in Britain; it has been going since 1949. A detailed account of the history is available here. Each meeting of the Colloquium is an independent event, but it is supported by the London and Edinburgh Mathematical Societies. The Scientific Committee exists to provide some continuity and pass on good practice.
When I was a student, attendance at the BMC was more-or-less obligatory. Indeed, I could have been killed in an accident on the motorway travelling to the 1970 BMC in York, but walked away without a scratch. (I go by train now!) The BMC has been hit by the huge increase in the number of conferences, especially subject-specific conferences, and is trying out different formats to encourage ordinary mathematicians and students to attend more regularly.
The Committee is not a random selection of mathematicians, and has absolutely no mandate to express a view on EPSRC. Moreover, the discussion reported here was not formally part of the BMC committee meeting, although it did involve the same people. It is composed of representatives of the two mathematical societies, present and future Colloquium organisers, representatives elected by the Colloquium business meeting, and by invitation a representative of the Young Researchers in Mathematics, whose meeting this year I reported on here. I was there because the Colloquium will be at Queen Mary in 2014: I intend to post about this sometime soon.
So these were for the most part fairly senior mathematicians, or at least people who are involved in working for the good of the mathematical community.
The current research council was produced by a reorganisation in the early 1990s. The previous research council, SERC, had subject committees; I did a three-year spell on the Mathematics Committee. The committee had two functions: first, to assess applications for research funding in mathematical sciences; second, to take strategic decisions for the benefit of the subject. In the new regime, the applications are rated by ad hoc panels drawn from the community (so that there is only limited continuity), but the referees are chosen by non-mathematicians at EPSRC on the basis of keywords; strategic decisions are all taken by non-mathematicians.
I have a confession to make here. It was during my time on the Mathematics Committee that we divided the subject informally into nine interrelated and overlapping “themes”. This was just for our own internal convenience, to help allocating submissions to committee members, and so that we could monitor applications from different sections of the community, and give encouragement where necessary.
Inevitably these “themes” have been fossilised and govern the functioning of EPSRC mathematics programme to an unreasonable extent. They are treated as atoms which cannot be divided and do not form bonds.
Currently fellowship applications in the mathematical sciences are accepted only in the area of statistics and applied probability. Why? As I have argued here before, a superficial reading of the IRM report suggested that statistics is in need of support. But unfortunately one of our “themes” was statistics and applied probability, so they decided that the recommendation would be met by taking this action. Unfortunately this has a negative effect. There are quite a lot of areas of mathematics, ranging from analysis to mathematical finance, whose practitioners can disguise themselves as applied probabilists. Statisticians don’t have this luxury. The fellowships have not yet been announced; I would not expect to see many statisticians on the list.
EPSRC management are not keen on small grants (even though these are crucial to mathematics) because “it’s like throwing money over a wall”.
John Hunton’s comments
- SAT consists of mathematicians. It is not a decision-making body; it is appointed by EPSRC but all decisions are made by civil servants.
- SAT talks to the mathematics programme at EPSRC, but individual subject programmes are low in the EPSRC hierarchy, so views fed in on major strategic issues don’t necessarily get to the decision making levels even if the subject programme is sympathetic (as it not infrequently is).
- Compared to other research councils, EPSRC gave a big hostage to fortune in the last Comprehensive Spending Review by promising extensive reforms, and are now trapped by their undertakings: hence “Shaping Capacity”.
- The pressure to restrict so drastically the areas eligible for fellowships has come from much higher in the organisation; other subjects are going through the same thing. We have been told that these restrictions will be weakened in the near future.
- There is concern about how “Shaping Capacity” will be implemented in the research grant system. Possibly referees will be asked to evaluate the “importance of the research, in accordance with the strategic plan”. If they cannot or (as I hope) refuse to do so, what will happen? Will civil servants make these decisions?
- Another concern is that one argument to justify restricting the area of fellowships is that it will produce future leaders in the areas we want to strengthen. But mathematicians, especially future leaders, do not stick rigidly to the area of their doctorate; indeed, a postdoc is the best possible period for exploring other areas.
- My comments about the indivisibility of the nine successors to “themes” (I don’t know what they are called now) and the lack of provision for research connecting two themes seem to have been broadly correct, although John feels that some progress has been made on linkages between areas.
What I find a little depressing is the ease with which good practice can be set aside, and the uphill struggle the community faces in trying to get back to these good practices.
- The Engineering and Physical Sciences Research Council, the main source of funding for research and postgraduate research training in British universities in their area, which covers all of “hard” science except for astronomy and particle physics, as well as engineering.
- Strategic Advisory Team. The one referred to here is in mathematical sciences (similar bodies exist in other subjects). It is composed of mathematicians, selected by EPSRC by a complicated procedure which nobody both understood and was prepared to explain. As the name suggests, they can give advice, but have no decision-making function. Part of what they do is to convey the opinions of the community to EPSRC. Individual members also have the job of passing information in the other direction; that is what John did, and it was much appreciated.
- Council of the Mathematical Sciences. A high-level body representing the London and Edinburgh Mathematical Societies, the Institute for Mathematics and its Applications, the Royal Statistical Society, and the Operational Research Society. Part of its job is to engage in political debate and lobbying on matters of common concern. Our experience is that they do a very good job of speaking with a single voice on behalf of the mathematical sciences.
- International Review of Mathematics, a review of UK mathematics by an international panel, which was held last year. The report is here. I have referred to it in a couple of previous posts. The Chief Executive of EPSRC has, I understand, publicly disowned some of the recommendations which he doesn’t like, with the words “We are not doing social engineering.”
- Shaping Capacity
- An EPSRC initiative according to which the bulk of research funding will not be responsive (as now) but will be targeted. The current situation, where Fellowship applications are not accepted in any part of mathematics except Applied Probability and Statistics, is an example of this policy in action.