South Africa has already implemented many of the recommendations made by SciDev.Net's director in a recent editorial (see Can Africa pioneer a new way of doing science?).

Various integrated large-scale projects, put in place by the Department of Industries through the Innovation Fund and the National Research Foundation, have been in operation for close to ten years.

These projects are modelled on the European Framework Programme, although this programme has had little acceptance and marginal success. In South Africa, the introduction of 'mode 2' science — pushing for patents and industrially applicable research — has seen the average age of scientists publishing academic papers rise from the mid-30s to the late 50s, and our share of research published worldwide drop from 0.7 to 0.4 per cent since 1994.

The increasing unattractiveness of a career in science is largely due to two factors: new employment policies and the scarcity of tenured posts.

In the rush to fund socially relevant research, the need for research projects to have talented and dedicated leadership and for the scientists working on them to have a predictable career structure, has been overlooked. Many young scientists returning from a long stint of postdoctoral research abroad find it difficult to move into tenured positions and to set up new research groups.

Individuals who have succeeded in setting up groups in South Africa have shied away from local funding because it is not 'user-friendly' and comes with an inordinate amount of bureaucracy. The well-meant objective of fostering social responsibility in research is therefore difficult to implement and not really suitable for generating a culture of research.

In his editorial, David Dickson makes the point that stimulation of scientific curiosity would remain an important facet of research in Africa. In fact, it should be paramount. Science policies should remain primarily concerned with the discovery and support of scientific talent. Social responsibility should not be the overriding factor because it cannot be measured and is difficult to teach, and it most certainly should not be the basis for academic promotion.

A saint of social responsibility would be rather ineffectual, if not useless, without a good track record in science. Science policymakers must accept that funding research can be risky, and that outstanding merits of social relevance alone cannot guarantee success. It would be dangerous therefore if the importance of social responsibility in science were overemphasised — generating a false feeling of social accomplishment in young scientists — and basic, more 'old-fashioned', sciences such as mathematics, physics and chemistry were ignored.