Select Committee on Science and Technology Written Evidence


Memorandum submitted by Professor Charles M Goldie, University of Sussex

  My contention is that the Royal Society, which has long had few Fellows from the mathematical sciences, has in recent years drifted into almost total neglect of the area within it that I know and belong to, statistics and probability. That is, the self-replicating nature of the Fellowship has been allowed to drift into a state where the natural tendency of Fellows to favour their own fields has led to a serious distortion and imbalance.

  This may be an instability that has led to other fields being similarly neglected or, conversely, over-represented; I wouldn't know. But I hope that lack of similar complaints from other branches of mathematics, or science generally, will not cause you to discount this letter. You are hardly likely to hear from over-represented areas. From within any similarly under-represented scientific fields there are very great obstacles against the making of such a complaint. Those who have the standing to do so are likely either to have been elected Fellows themselves, and thereby naturally tend to feel that all must be well, or to be in contention for Fellowship, when their views would tend to be discounted by their being considered interested parties, and in any case they would certainly not express them for fear of damaging their chances. My own position is that I am not a contender for Fellowship, but I have researched and taught in my own field since graduating from Imperial College in 1964, and occupy a central and senior enough position to be able to speak with confidence. Furthermore, towards the end of this letter I shall list evidence that supports my claim.

  Insofar as the Royal Society is now partial and discriminatory in its coverage of science it is clearly not fit to be a conduit for public funds of the sort it has become. However, I want to add straight away that I fully accept that only the Royal Society can address and act on any problems concerning its own membership. No other body can or should tell it what to do. My hope is that your enquiry will lead to it recognising that it has a problem concerning the subject balance of its membership, and that the Fellows will be moved to correct it.

  I write to you as Chair of HoDoMS (Heads of Departments of Mathematical Sciences in the UK) because the dearth of Royal Society Fellows has significant adverse consequences for my member university departments. Those departments cover pure and applied mathematics, statistics and operational research, and we extend throughout the UK including Scotland, Wales and Northern Ireland. Although there has not been time for me to consult my members I am confident they will back me up in my contention of a general lack of Royal Society Fellows in our field, and of specific neglect of my own area once they see the evidence. The main adverse consequence is a major contribution to the general invisibility and lack of "clout" that afflicts mathematical departments for a number of reasons, and a resulting failure to win resources. When Vice-Chancellors have major decisions to make concerning science they look for outside advice, and whom do they seek it of? FRSs, naturally, as those are the individuals who have the badge of authority that they can rely on. There are currently numbers of smaller mathematics/statistics departments in danger from the disaster that affected AS Level Maths last summer—see Lucy Hodges's article "What does a decline in maths really add up to?" in the Independent of 7 March—and Vice-Chancellors feel free to lay about them against a field that to them seems to have few world-class figures to raise opposition.

  The evidence that I promised consists of a complete list of the FRSs in my own area, statistics and probability, considered in relation to the size and strength of that area. Here is the list, of those still living, by date of election. It involves my own judgement of whom to include, but I had no difficulties of demarcation and am confident that any senior colleague would back up my choices of inclusion and exclusion in virtually every case.

D J Finney
1964D G Kendall
1966Sir Richard Doll
1967C. R Rao*
1971Sr John Kingman
1973Sir David Cox
1974Sir Walter Bodmer
1976J M Hammersley
1978P Whittle
1981J A Nelder
1984D Williams
1985G E P Box**
1989F P Kelly, Sir Richard Peto
1994D J Aldous**, P. McCullagh**
1997B W Silverman
1998S R S Varadhan*
2000W Ewens*, P G Hall*
2001A F M Smith

  * Commonwealth citizen, employed abroad

  ** British citizen, settled in USA

  You will see from the table that there have been only two individuals in this country elected in the 12 years 1990-01 inclusive. What is the size of the field in this country? Those towards the statistical end are likely to belong to the Royal Statistical Society while those more in probability (there is a continuum of interests in between) will usually join the London Mathematical Society (a national society despite its name), and many, like me, are in both. The RSS has 7,200 members, perhaps two-thirds in this country, and virtually all of those are in the subject area that I have used to construct the above table. The LMS has 1,500 members in the UK, of whom only a hundred or two work in probability/statistics and are not also in the RSS. At a very rough guess one arrives at about 5,000 individuals in my field in the UK. The table shows that this population gives rise to 14 FRSs in this country. That is surely a low proportion, but the really telling figures come from extrapolating the recent rate of election from the field. Assuming continuance of the two-every-twelve-years rate of election, and assuming, generously, that newly elected Fellows live for 36 years after election, one arrives at a steady-state total of six FRSs in this country from probability and statistics, a field of perhaps 5,000 scientists. This is catastrophically few. The Institute of Physics has 30,000 members, perhaps 20,000 in this country. It is inconceivable that physics could have only 24 FRSs in the UK. The Royal Society, which has 1,216 Fellows, would be an order of magnitude smaller.

  All those listed in the above table are male, by the way. The problem of its Fellowship being overwhelmingly male is one that the Royal Society recognises, but the field of statistics and probability shows up as a signal instance of the imbalance: to my knowledge it has not ever had a female FRS.

  Could the Royal Society's neglect of my field be explained by the field being moribund, no longer producing good new science? No: there is simply no way that any such claim could be maintained. In last year's Research Assessment Exercise the panel for Unit of Assessment 24 (Statistics and Operational Research), under which most researchers would fall, awarded the highest average grade of the three mathematical panels. A small proportion of the relevant individuals will have been assessed under Pure Mathematics (Unit of Assessment 22), but the average grade under that panel was again high. All panels used overseas experts. Overall improvement in research in the mathematical sciences has led the relevant research council, the EPSRC, radically to expand the funding allocated to its Mathematics Programme, within which my area wins its proportionate share in the general competition.

  Another indication of health and vigour is that both learned societies, the Royal Statistical Society and the London Mathematical Society, have recently expanded into new, impressive, central premises, and the energy, confidence and breadth of their activities is plain for all to see.

  I have made my case but I ought to reconcile what might seem my contradictory statements that certain maths/statistics departments are under threat, and yet research in the area is extremely healthy. There is no contradiction, because it is the smallest and most marginal departments that tend to face difficulties, and they do not have a substantial research presence. Concentration may be in progress, but that is a whole other topic. Concerning it, let me just remark that when Stephen Timms, Minister of State in the DfES, came at my invitation to address the HoDoMS Annual Conference last week, he told us that the country needs 40 per cent of maths/statistics graduating cohorts for the next few years to enter school-teaching. This ambitious target cannot be delivered by the grander research-intensive departments alone. It needs universities and HE Colleges with local catchment areas to continue to contribute, up and down the country. Concentration will render the delivery of an adequate supply of trained manpower impossible.

April 2002

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2002
Prepared 6 August 2002