RAE scales will never balance

It is right to reward and develop centres of excellence, but how can this really strike the right balance in research funding ("United assault on Clarke's plans", THES , April 11)?

According to the research assessment exercise criteria, the difference between 4, 5 and 5* units is a difference in the proportion of internationally excellent researchers as opposed to nationally excellent researchers. A 5* does not indicate that every member of staff in a 5* unit is internationally excellent but rather that "more than half" are.

In philosophy, only Oxford University will achieve 6* status using the criterion that requires units to have achieved 5* status in two successive assessments. Oxford's 2001 RAE result means that minimally 25 staff of the 49 entered were internationally excellent. The remainder demonstrated national excellence.

The difference between 4 and 5-rated departments is that the former have "some" internationally excellent researchers and the latter have "up to half". Thus, if an average-sized 4-rated philosophy unit entered eight staff, three might be internationally excellent.

It follows that cutting funding from ten 4-rated philosophy units to better fund one 6* unit threatens support for 20 to 30 international researchers.

Some universities are clearly preparing to take a strategic view of research. They will focus funding only on units that have a clear potential to achieve a 5 rating. What will become of the internationally excellent research in the 4-rated units? In philosophy, 73 per cent of units rated 4 in the 1996 RAE were rated 5 in 2001. This suggests that previous funding levels were effective in raising the amount of international research.

It could be argued that funding a few select units at the expense of many others creates "critical mass". But this assumes, falsely, that all research is primarily collaborative or team-based, requiring researchers to be located in a single institution. What does that leave for disciplines such as philosophy in which research is often conducted by individuals working alone?

Daniel Hutto
Head of philosophy
University of Hertfordshire