Mechanistic explanations and animal model simulations in neuroscience

Abstract

The 'new mechanists' assume that discovering neural mechanisms is a major aim of neuroscience where this constitutes a process of explaining. They strive to articulate the norms for good explanations. I argue that: the normative project of the ontic mechanistic explanation is unrealistic; the epistemological version of mechanistic explanation accounts for the practical epistemological constraints of neuroscience but fails normatively; hence a dilemma: either methodologically strong but unrealistic or epistemologically realistic but methodologically weak view of mechanistic explanation. I propose that the solution is in abandoning the idea that the study of mechanisms in neuroscience aims mainly at explaining. Model-building and simulating neural phenomena and mechanisms do not necessarily aim at providing explanations. Additionally, some of the cognitive goals attributed to explanation are actually served by simulation. Thus, the new mechanist approach needs to be revised so as to do better justice to the role of simulation in neuroscience.