A to Y of South African science

Funds for science research hinge on a complex grading system that recognises achievement and prejudice under apartheid

The University of Cape Town, South Africa's acknowledged top research institution, dominates the most recent list of A-rated scientists - but it should watch out for the University of Natal, which has the largest number of the country's most brilliant young scientists.

The system of evaluating and rating scientists in South Africa was started in 1984 with the aim of recognising academic achievement and helping top scientists attract research funds. There are 1,007 rated researchers in six categories.

At the top are "A-rated" scientists who, according to the National Research Foundation, are "without doubt" accepted by their peers as being world leaders in their field, in terms of the quality and impact of their recent work.

There are 46 such South African stars, but the number tends to fluctuate annually between 38 and 48.

Not far short of half of them - 20 - are from the University of Cape Town, which is particularly strong in the fields of mathematics and applied mathematics (four top scientists) and geological sciences (three).

The University of the Wi****ersrand comes next with nine A-rated scientists, followed by Stellenbosch (five), Pretoria (four), Natal and Rand Afrikaans University (three each), and the universities of North West and South Africa with one each.

In terms of areas, mathematics and applied mathematics are strongest (nine top scientists), then engineering fields (eight), and then geology or geological sciences (six).

Virtually all the A-rated scientists are white men. Not quite so for South Africa's 17 exceptional young "P-rated" scientists: three are black and two are women. They are researchers, usually under 35, who have a PhD and are "recognised internationally as having the potential to become future world leaders in their fields".

Among these brilliant young brains are five researchers from the University of Natal, three each from Wits and Cape Town, two each from Pretoria and Stellenbosch, and one each from Rhodes and the Western Cape. This time engineering fields dominate.

There are four other categories of scientists. "B-rated" scientists - of whom there are currently 234 - enjoy "considerable international recognition" as researchers of quality and impact. There are also 538 "C-rated" scientists, defined as established researchers who, individually or in a group, produce research of international standard.

In addition, there are 123 Y-rated scientists, who are young researchers with doctorates who show the promise of establishing themselves as researchers within five years.

In recognition of South Africa's apartheid past, there is also a category of L-rated scientists, of whom there are 48. They are people who have shown potential in the past and "who can demonstrate that they were prevented from realising their potential, but who now show promise of establishing themselves as researchers" within five years.

The evaluation and rating process is very intense and stringent, says Gudrun Schirge, manager of the NRF's evaluation centre. Most universities know the system very well by now and tend to filter out applications before they are sent to the NRF. Still, generally about a quarter of applicants in any given year are not successful.

"However, once researchers have made it into the evaluation and rating system, they usually maintain or improve their ratings," Ms Schirge says. Very few drop out of the system, though a number leave of their own free will if, for instance, they move out of higher education or go abroad.

The most important criterion in rating scientists is the quality of their research outputs over the past five years, she adds. They must be employed full-time by a higher education institution.

The ratings are done by local and international peers. Usually about six peers - 20 per cent local and 80 per cent international - are asked to comment on the quality of the research of each applying scientist. A committee of experts then assesses their comments, and consensus is reached on the rating.