Unity of science against the rational construction of knowledge?

Most scientists are confronted daily with the heterogeneity of scientific approaches and with the need to solve increasing difficulties to combine knowledge coming from various scientific areas. But these difficulties are ignored when philosophy of science postulates the possibility of establishing a priori rational principles of a unified science. In that case, it is logically considered that the problems met to integrate scientific knowledge originating from different sources result from institutional gaps or organisational failures, or from the lack of skill of the researchers. Subsequently, the analyses of these problems are handed over to the sociology of science. Therefore the debate on the unity of science may screen the heuristic value of approaches based on an epistemological regionalism to support the rational construction of knowledge. Actually, to study the on going building of scientific knowledge, the question of the ultimate state of the science does not matter. What is important is to acknowledge that here and now1) to deal with a similar phenomenon, theories based on distinct conceptual architectures -and sometimes contradictory hypotheses- coexist (for example Cartwright 1999);2) social sciences, nature sciences and technical disciplines may produce knowledge with distinct properties, generating different kinds of interactions with its environment (Hacking 2001, Hottois 1996) ;3) each research programme has its own area of demonstration and builds knowledge through specific paths (Bachelard 1949).If these statements are to be considered seriously –and so are they by most scientists- it is no more obvious to put together different scientific areas: the integration of scientific knowledge is an epistemic situation that needs clarification. The presentation, based on specific examples in ecology and economics, will show the perspectives open by such a view point to analyse the revision of research programmes and to assess the results of research projects. By doing so, this presentation would like to contribute to stress the need for further development of internal analyses of on going science that would associate scientists and philosophers. As a matter of facts, a vision of a unified science that would ignore the difficulties resulting from the current conceptual heterogeneity of scientific approaches would partly deprive the building of rational knowledge from the support of the internalist analyses of philosophy.