The drawbacks of a peer-review system (THES, Leader, January 9) whereby one's rivals can pass judgement on the fate of one's work are obvious. Similarly, few would be surprised at the tendency of committees to reach consensus on the least controversial of projects. "Safe science" is applauded as a sound investment because the results are so predictable that any element of risk is negligible.

An operational inadequacy of the system is that one is asked to devote time to reviewing the work of one's peers for no gain. Though up to six referees can be canvassed, I know of instances where only two replied. How can one application attracting six reports be assessed next to one covered by only two?

Another bureaucratic injustice is that panel selection is far from democratic. Certain scientists with no obviously special qualifications, and in accordance with no obvious guidelines, have greatness thrust upon them, whereas others, even long-term recipients of major grants, are never invited to serve. Such haphazard selection is ripe for old-boy exploitation and ensures panels will not be representative of the scientific community.

This issue of those serving on the inner-sanctum panels is a critical one, since it is here that the final decisions are made. Yet these discussions are never fully disclosed. Thus seemingly strong referees' reports can be trumped by comments not held to account and delivered anonymously in camera.

ÖйúAƬ

Modest alternatives to the current system could include circulating the minutes of in-camera meetings of the parts of the discussion relevant to the applicant; a jury service-like rota for serving on panels for which all tenured academics would be eligible after a certain time; and having all applications based on the same number of reports by restricting all judgements to two only. It might even be worth while motivating referees by paying them.

There are those with still more radical suggestions. David Horrobin outlined some 18 months ago in The Lancet a scheme whereby all available money could be shared equally among an agreed constituency of scientists. Another possibility would be that grey-beards with a lifetime's experience of science and now on the cusp of retirement should serve on the panels: after all, they would have more time to do a proper job and, more importantly, would not be competing for funds themselves. Or perhaps one could even open the limited funds to a lottery system for all applications judged worthy in an initial trawl, irrespective of a low or high priority that could never be proved.

ÖйúAƬ

Unless something is done quickly, the brightest and best will become demoralised and disillusioned. Public-sector research will spiral into mediocrity. The kind of derring-do and innovative thinking that gave this country a reputation for truly astonishing science will be channelled into the open arms of a private sector that cannot afford not to have vision. But the bulk of basic research offering, to use Francis Bacon's terminology, lucifera (light shedding) rather than fructifera (fruit bearing) gains will be increasingly pushed to the wall.

Susan Greenfield, Professor of pharmacology, University of Oxford