tokyo

Nanjing University in southern China is still the country's most productive source of scientific papers and has increased its lead over China's two top universities in Beijing, according to Science Citation Index figures.

The SCI's statistics, from the US-based Institute of Scientific Information, show that Nanjing, which has long emphasized basic research, published 570 SCI papers in 1996, 118 more than in the previous year.

For several years, Nanjing University has outperformed Peking and Tsinghua universities, in the northern capital of Beijing, in output of SCI papers (see Nature 378, 543; 1995). And this gap has widened: Peking and Tsinghua, which rank second and third respectively, each produced just under 300 papers listed in the citation index, and increased their output only by a few tens of papers during 1995 (see table)

Table 1 Papers published by Chinese universities in international journals in 1996*
Full size table

The figures on the numbers of SCI-listed papers produced in 1996 by universities and research institutes in China were released last month by the Chinese State Science and Technology Commission and the Chinese Institute of Scientific Information in an annual exercise that looks also at the number of engineering and domestic (Chinese-language) research papers.

Nanjing's comparatively high output — which remains far behind the thousands of SCI papers typically produced by leading universities in the West — results from a policy of encouraging basic research. In the 1980s, the university's president Qinyue Qu, an astronomer, introduced bonuses for researchers who published SCI papers funded by local industry.

Officials at Peking and Tsinghua universities are sceptical about Nanjing's performance. They argue that it is merely the anticipated outcome of a highly targeted approach aimed at producing more research papers in international journals, and that numbers of SCI papers do not necessarily indicate the general quality or standing of a university's research output. Nevertheless, they admit that their output of SCI papers is low and needs to increase.