Leader: Academy is right to question the use of metrics

There are always more promising proposals for research than can be funded, so some way of deciding on the most deserving is essential. But the sheer variety of scholarly life means that there can never be a single way of defining quality.

The British Academy is right to point out that much academic work cannot be assessed by measuring papers and citations. This approach works well for science, where researchers often write several papers a year. But a brilliant historian might be remembered for publishing two great books over several decades. Even science and technology have varying publication patterns. An average research paper in engineering contains far fewer citations than one in biomedicine. At the same time, some social sciences have a publishing model very similar to those of the physical sciences. Economics produces more papers in journals and fewer books and monographs than other disciplines covered by the British Academy. At the other extreme, there are some distinguished arts academics who produce plays or paintings rather than books or monographs.

The academy's call for better peer review is a welcome one. At the moment, it is regarded as part of the bargain between universities and academics that they work as peer reviewers for little or no reward, whether in cash or prestige. Teaching and research bring in cash to the university and are central to career success, while peer review is one of the activities - editing journals is another - that is taken for granted and allowed too little time in academic schedules. Once these important activities count in promotions and pay rises, they will suddenly gain in popularity and esteem.

Peer review matters increasingly to arts and social science researchers. At one time, many did not need grants. With enough peace from students, and a well-stocked library, they could produce original research for little or no cost. But few modern universities will provide staff with endless free time in the hope that some great work might eventually emerge. The personnel management skills they have imported from the private sector are far too unforgiving. At the same time, research has become more of a collective enterprise. Even historians now work in research groups.

However, the larger scale of research in these fields, which parallels the increase in the ambition of science during the 20th century, means new publishing patterns. Even in areas such as the Classics, researchers are producing more articles in journals than before. So the problem the British Academy has highlighted will become less serious as time passes.

A more fundamental difference between science and other research areas is that science became genuinely global many years ago. It is possible to compare the research impact of a chemist in London directly with a competitor in Korea. Much humanities research, for example on law or politics, is tied to national systems and is reported in national publications. This is unfortunate because the present British Government is keen to back research that it regards as world class.

Indeed, it is inevitable that the next moves on peer review for arts and social sciences research will be international. Five of the 20 subject areas being funded by the new European Research Council are in these areas, and the European Union's Framework research programme is paying more attention to them. In response, the European Science Foundation is drawing up lists of the top humanities journals and working on means of assessing humanities research on a world scale. UK universities have unquestioned strengths in these subjects. Proving their quality would build the case for more funding and make UK researchers more attractive partners for international projects.