In 1998, Nature published a total of 84 papers from Japan — the highest number ever. That record is likely to be exceeded this year. This continually upward trend would seem to indicate that Japanese researchers are taking a more active role in the international science arena. However, a report released last week by the Science and Technology Agency states otherwise (see page 314).

The report, based on a survey across a wide range of research institutions in Japan, reveals that a significant proportion of Japanese researchers do not publish internationally. According to the report, 40 per cent of researchers submit papers only to Japanese journals. The reason behind this, say researchers who responded to the survey, is the fact that many institutions still place a strong emphasis on the numerical output of research papers as a criterion for evaluating research performance. Such a system forces many Japanese researchers, including highly talented ones, to shy away from international journals with high impact factors and to turn instead to Japanese publications, which provide a better opportunity for their papers to get published.

But the system alone cannot be blamed for Japanese scientists' lack of involvement in international science. They themselves need to make an effort to enhance their communication with researchers in other countries. The survey reveals that only 20 per cent of researchers communicate regularly with overseas colleagues. A large percentage of them do not communicate at all, and less than half use e-mail and faxes for their communication.

In the age of the Internet, there is every reason to encourage more proactivity on the part of both Japanese and Western scientists in establishing contacts, despite the obstacles posed by language. And for Japanese scientists to rise more frequently to the challenges of international scientific communication, their institutions need to focus more on quality rather than quantity in their assessment of scientific output.