[Social sciences in the face of AIDS: between silence and too much talk?]

During the recent years, many behavior surveys have been performed, for instance in San Francisco, Chicago, New York (U.S.A.), Ile-de-France (France), or elsewhere. They have evidenced profound change in sexual behavior among certain high-risk groups. Such a change is probably unprecedented and could contribute efficiently to the struggle against the AIDS epidemic. Nevertheless the problem is far from simple. AIDS infection risk growth is much higher when the number of sexual partners rises to 5 from 1 than to, say, 50 from 5. In the same way infection risk reduction is much lower when coming to "occasional use of condom" from "absence of use" than when passing to "systematic use" from "occasional use". Furthermore, it appears that no direct causal relationship exists between increased information on risks and change in sexual behavior. In fact the probability of an individual integrating the provided information into his/her sexual practice depends to a large extent on the acceptance of the information by his/her group. Norms within and leadership of the group are crucial for accepted preventive practices. Other factors, more personal, play also their role, the question of their impact in the long run is raised. In France, and other countries as well there is a tradition for social sciences to restrict their activities to observation and analysis, mainly in view of theory-building. Certain social scientists are extremely reluctant to commit themselves to work for supporting decision-making. Nevertheless, for the time being, social sciences must be present at the debates on how to manage, and stop, the AIDS epidemic.