If we wish to establish a sustainable research agenda within mental health, is it wise to ask a group of eminent researchers how we should go about it? It will not be a great surprise if the answer that such a group will come up with is that we should have more research. It appears to be part of the grand British academic life that, from time to time, we have the great and the good coming together to make pronouncements on the state of play, and how the future should be planned on the basis of what has gone before. Usually, the answer is more of the same but with extra helpings all around. The paper by Thornicroft et al (Reference Thornicroft, Bindman and Goldberg2002, this issue) is in that grand tradition, with some of the UK's pre-eminent researchers in the field of social psychiatry coming together to identify a potential research agenda for mental health. Predictably, they conclude that we need more research, much the same way as previous attempts to take stock of research within the sector have concluded.

The purpose of this paper is not entirely clear. Although the authors set out by saying that they want to identify ‘the important gaps in research coverage, particularly in areas key to the National Service Framework for Mental Health and the NHS Plan’, what they appear to end up highlighting are the problems with research infrastructure in mental health in general, rather than anything specifically linked to the national policy initiatives. Also, it is difficult to establish how the authors actually arrived at their conclusions because there is only a passing mention of their methodology, that is conducting a series of expert assessments. We are not privileged to know who these experts were, how they were chosen, what kind of expertise in mental health they had and how they carried out their assessments. In any other research, say, for example, if we were interested in establishing patient preferences of the direction of mental health research, these most rudimentary methodological details would have been made available before submitting the findings for publication. Given that at least one of the source materials used in the assessments was prepared by some of the authors of this paper, there are also bound to be questions about the independence of the findings reported here.

Notwithstanding these criticisms, most people would find the 11 recommendations advocated by the authors to be sensible, but hardly new. The continuing attachment to positivistic and empirical approaches to mental health research within which the randomised controlled trial rules the roost might be comforting within the current research culture, but this hardly breaks new ground in advancing the case for patient-centred or meaningful outcome research, a major weakness within social and community psychiatry in Britain at present. In particular, there is little mention of the need for patient involvement in setting the research agenda, in addition to patient participation, perhaps indicative of the experts' view about who has the expertise in mental health.

Finally, exercises like this would be far more rewarding if academics and researchers were prepared to adopt a more realistic view of our relevance to the way mental health services are developing and the impact our academic work has on patients' lives. One of the most sobering thoughts for academics must be that, by and far, the fundamental changes that are currently sweeping across mental health services in this country, as a result of government policies such as the National Service Framework and the NHS Plan, are not the crowning achievements of research or other academic initiatives. These changes, of a magnitude and scope that is unprecedented within British psychiatry, are the result of patient and community aspirations, given articulation through a political process, within which equal weight has been given to values as well as evidence, the latter not necessarily a product of empirical research. Surely, there is a lesson here for all of us, more important than the recommendations arising out of any academic exercise.