ÁñÁ«ÊÓƵ

It was a relief to read Andrew Oswald's article on the research excellence framework ("Data be damned: REF's blueprint for systemic intellectual corruption", 22 September).

I am among a small group of academics who believe that the research assessment exercise and its predecessors were wonderful innovations because they cut across established prejudices in the academy: the sciences could no longer be automatically superior in every university, nor all departments in Oxbridge automatically better than all departments elsewhere. But the class system is insidious, and now small groups of academics are insisting on their special sensibility when it comes to judging quality, scorning the plebs who use instruments to measure.

These measures are not perfect, but neither are the sensibilities of an aristocracy. We need both metrics and peer review, and we need the REF panels to be forced to debate in those cases where there is disagreement between the approaches. It is so easy.

But the way arguments against multiple indicators always set the failures of the plebs against the baseline of unquestioned aristocratic judgement is to be reminded, once again, of country houses and sherry in the library.

Harry Collins, Distinguished research professor, Cardiff University