The clarity and consistency of the Engineering and Physical Sciences Research Council's pronouncements about its controversial shaping capability policy have come under scrutiny after its two top figures appeared before the Science and Technology Committee in the House of Lords.
Under the policy, the EPSRC will decide whether to expand, shrink or maintain the 111 subjects into which it has divided its portfolio on the basis of national importance and existing capacity, as well as existing UK excellence.
The first decisions, announced in July, prompted many vociferous complaints about a perceived lack of consultation. The EPSRC responded in October with a pledge to work with learned societies before making any more decisions.
Just days after that promise, however, the research council waved another red rag in the direction of its bullish critics with its announcement of how shaping capability would influence peer review of grant applications. Its website stated that "national importance" would become a "primary assessment criterion alongside research quality".
But in an evidence session before peers on 29 November, David Delpy, the EPSRC's chief executive, and John Armitt, its chair, repeated the message of their 17 November letter to Nature: that research excellence would remain the "pre-eminent" or "primary" criterion for peer review.
The puzzled chair of the committee, Lord Krebs, principal of Jesus College, Oxford, quoted an EPSRC document that stated: "Research quality and national importance will be primary criteria of equal weighting."
Lord Krebs said: "That does not seem to fit with what you have just said, which is that research excellence is the criterion and, in the margins, you might apply national importance as a tie-breaker."
Professor Delpy responded that Lord Krebs must have downloaded an "old web statement". On being told that it was dated 3 November 2011, he said: "I mean what you have quoted is incorrect, and I am surprised that it is still on the web."
Indeed, the EPSRC website does now make clear that excellence remains the main criterion of its peer review process. Lord Krebs told Times Higher Education that the research council appeared to have quietly shifted its stance since it published the document he had quoted.
"The statement gave me cause for serious concern that scientific excellence would be traded off against national importance," he said. "The chief executive and the chair tacitly acknowledged that the EPSRC has changed its position, and this is to be welcomed."
A simple clarification
However, a spokeswoman for the EPSRC insisted that the revisions made to the website reflected nothing more than the research council's desire to correct "misunderstandings". She added that more changes might be made if a series of regional workshops this week reveal that the "community would benefit from further clarity".
During the committee meeting, Professor Delpy was at pains to play down the significance of the changes. He pointed out that referees and applicants had always been asked to comment on the "importance" of research proposals. He also insisted that the national priorities identified by the EPSRC enjoyed widespread backing, stemming as they did from prominent reports, government policy aspirations and the existence of "certain industries that we believe the UK is internationally leading in and that the research base should be supporting".
Professor Delpy said he did not believe that the introduction of national importance as a criterion would "inhibit creativity", but he promised to act if he got "any indication" that it was doing so.
Mr Armitt told the committee that the EPSRC was talking to peer reviewers about "how best to reflect on" national importance before the criterion is introduced in the spring.
Another new element in EPSRC peer review will be an assessment of how well grant applications fit with the shaping capability agenda. Professor Delpy said the research council would probably make such judgements itself while leaving it to panels to decide whether they wanted to amend their ranking of applications accordingly.
But critics have not been placated. Lord Rees, master of Trinity College, Cambridge and former president of the Royal Society, described as "entirely unrealistic" the idea that reviewers or applicants would be able to fulfil the EPSRC's requirement to identify the national importance of proposed work up to 50 years ahead. "It could be laughed off, but the concern is that young people putting in their first grant application would be perplexed and baffled and rather terrified by what they have to do to meet this requirement."
Mr Armitt said it was up to universities to help young researchers make good applications. He suggested that a mathematician could pass the national importance test by setting out how his research would add to "the stock of knowledge in his discipline". Nevertheless, he admitted, it was "difficult to differentiate the real differences" between national importance and the existing "pathways to impact" criterion, which requires applicants to set out how their research could have benefits beyond the academy. "We are talking about impact in relation to national priorities," he said.
Not reassured
The research council's moves have not convinced Anthony Barrett, Glaxo professor of organic chemistry at Imperial College London and a vocal critic of shaping capability.
Saying that the research community was not reassured by the latest pronouncements, he called for more clarity over the EPSRC's pledge in its latest guidance that no research not deemed "excellent" will be funded, regardless of national importance or strategic fit, while also promising that "outstanding" proposals in any area will always be funded. "The chief executive and staff at the EPSRC have a track record of the misuse of clear words such as 'consult'," he said. "How much better is an outstanding proposal than an excellent one?"
Meanwhile, mathematicians have continued to protest against the confinement of postdoctoral fellowships in maths to applicants working in statistics and applied probability.
Professor Delpy said this area had consistently been identified as weak, but "the consequence of leaving this to the community to decide was that, in the past five years, not one fellowship in mathematics was awarded in the statistics area". He told the committee that the call for the next round of fellowships would include two additional subject areas.
But an internal briefing note produced for the Council for Mathematical Sciences says that postdoctoral fellowships - the subject of the Lords' questioning - would still be confined to statistics. It urges the EPSRC to "urgently reconsider" that decision.
The note also describes as "misleading" Professor Delpy's claim to the committee that the EPSRC was "not a major player" in maths research because it funded just 10 of at least 370 postdoctoral fellowships available in the subject each year. It calls the EPSRC's analysis flawed and says that, in reality, it funded about 40 per cent of awards.