Javier Rodríguez

Regulatory science, which generates knowledge relevant for regulatory decision?making, is different from standard academic science in that it is oriented mainly towards the attainment of non?epistemic (practical) aims. The role of uncertainty and the limits to the relevance of academic science are being recognized more and more explicitly in regulatory decision?making. This has led to the introduction of regulation?specific scientific methodologies in order to generate decision?relevant data. However, recent practical experience with such non?standard methodologies indicates that they, too, may be subject to important limitations. We argue that the attainment of non?epistemic values and aims (like the protection of human health and the environment) requires not only control of the quality of the data and the methodologies, but also the selection of the level of regulation deemed adequate in each specific case (including a decision about which of the two, under?regulation or over?regulation, would be more acceptable).

In this paper I distinguish two ways of using the expression ‘epistemological naturalism’. In one sense, naturalism amounts to a denial that epistemology should be understood as a kind of first philosophy providing the foundations for science from outside. In a second sense, naturalism holds that human knowledge is a natural phenomenon and that epistemology should be seen as a chapter of natural science. Moreover, naturalism in this second sense usually incorporates some additional specifications that build up a very restrictive concept of science. Two different projects for naturalizing epistemology lie behind these two meanings of ‘naturalism’, and the two projects, although sometimes seen as complementary or even equivalent, are not necessarily so. After drawing this distinction, I set out the difficulties faced by those naturalistic stances that do not discriminate sharply enough or do not establish the appropriate hierarchical relationship between both projects. Finally, I argue for a conception of epistemology which is best described as a radical antilfoundntionnlism. This conception sticks to the requirements that follow from naturalism when understood in the first sense above, and takes the cluster of theses that give birth to the second as subsidiary and revisable.