The Autonomy of Experimental Traditions as an Epistemological Issue

Abstract

Until recently, a common underlying assumption in philosophy of science was that experimental science, as well as other non-theoretical traditions in science (e.g. natural history), in order to be understood as part of science, have to be subordinated to theoretical aims. In the last twenty years, this assumption has been challenged from different perspectives. However, an important philosophical question remains. Roughly the question is: on what epistemological grounds can experimental traditions in science (and other non-theoretical traditions) be considered autonomous?

In this paper, I address this question by identifying and rejecting two central assumptions of traditional philosophy of science, the Newtonian and the Laplacian presuppositions. The Newtonian presupposition assumes that one can distinguish between contingent and law-like aspects of scientific explanations, in such a way that scientific explanations can be grounded, at least in principle, on laws with universal application. The other assumption, the Laplacian presupposition, consists in the belief that reason is disembodied or, at least, that the embodiment of reason has no major epistemological significance. This presupposition in particular supports the assumption that science is constructed by agents that have no epistemologically significant limitations in their computational and memory capabilities. I claim that the same evidence pointing to the need of abandoning these presuppositions suggests a characterization of the nature of the autonomy that is characteristic of experimental traditions. I argue that the sort of scientific reasoning that is constitutive of experimental traditions is predominantly (and irreducibly) based on heuristics, that inferences are context-dependent, and that ontological and epistemological issues are closely knitted in historically rooted aims and methods.