Geometry Bulletin Board

Draft findings of the International Review of (UK) Mathematics 2010


In December 2010 a panel of 16 (non-UK) experts assessed mathematical research in the UK by international standards. This International Review of Mathematical Sciences, called the IRM, was commissioned by the EPSRC, the government funding body for physical sciences and mathematics. It can be regarded as partly an independent review of EPSRC’s priorities and performance, and partly a review of the performance of the universities in developing their departments. Other players in the mathematical landscape, the societies and research institutes, are also considered. And finally, I assume at the explicit request of the EPSRC, the panel considered interactions with industry and other “users” of mathematical sciences.

The draft report of the panel was presented by Margaret Wright, the panel chair, at a town meeting in late January.

UPDATE: I didn’t attend the town meeting; however, Peter Cameron was there, and his account is: here. And Paul Glendinning’s is: here.

More new resources:

The report is a long, careful document, and covers a lot of ground. Even so, one notices that the panel’s remit runs up against some artificial boundaries. It looks like computer science as a whole was outside the panel’s purview, as was engineering; so combinatorics is divided from algorithms, statistics is divided from machine learning, and probability and OR are separated from control theory. Of course, there have to be some boundaries, but these particular divisions feel like unfair handicaps given the government’s current emphasis on demonstrating societal “impact”.

The panel distilled their work into seven findings, as well as 12 broad recommendations (for different combinations of funders, universities, societies, institutes, the mathematical community, and industry) and 15 field-specific recommendations. Because the report is still in draft, I feel cautious about publicizing the recommendations; but I should say that they aren’t anodyne. The panel is clear-sighted about how current practices and plans may affect UK mathematical research for good or bad. The findings are:

  • Overall, mathematical sciences research in the UK is excellent on an international scale, with world-leading researchers in every subfield and closely connected application area considered by the panel.
  • The high quality of UK mathematical sciences research depends critically on the diverse and distributed research community, where ‘diverse’ includes research area, group size and institution size,and ‘distributed’ refers to geographical location.
  • Actions taken by EPSRC since the 2004 International Review of Mathematics and the 2004 Review of Operational Research have greatly contributed to invigoration of the mathematical sciences, including improved structures for PhD education.
  • Newly established institutes and centres dedicated to furthering research in specific topics, interdisciplinary research and connections with industry have improved the UK’s international visibility and standing in the associated areas.
  • The institutes make a significant contribution to the visibility and quality of UK mathematical sciences research, as do activities of the learned societies.
  • Despite improvements, most UK-educated PhDs in the mathematical sciences are not adequately trained to be competitive on the international academic job market; hence a large proportion of postdocs and junior faculty consists of researchers trained outside the UK.
  • Action about gender diversity is not a sufficiently high priority for the UK mathematical sciences research community.

The first five findings are positive, so there’s not much to say; basically I agree with them.

Finding six, about UK PhDs, doesn’t hold as strongly for Cambridge and similar universities as it may “on average” across the UK. Nevertheless, everyone will benefit as we move towards more taught graduate-level courses and longer PhDs. Even now, it’s not uncommon for Cambridge students to take four years for their PhDs, and I certainly like my students, however advanced, to attend suitable lectures given for the Part III course.

As for finding seven, I have no doubt that Cambridge illustrates it perfectly. Some constructive pressure in the right direction could help us.