Hartry Field

New York University

Epistemology from a “Naturalistic” (but not Reliabilist) Perspective

17 May 2019, 16:00

Faculdade de Letras de Lisboa

Sala Mattos Romão (Departamento de Filosofia)

Abstract: I’ll advocate an obvious-sounding approach to epistemology, that involves developing general models of possible epistemic practices and critically evaluating which of those practices are likely to do best at achieving various truth-oriented goals. Despite its obviousness, there is an apparently serious problem with this idea, a generalization of the one in Lewis’s discussion of immodest inductive methods: each practice seems bound to evaluate itself as best, in which case the “critical evaluation” cuts no ice and one just ends up with whatever practice one starts with. A lot of the paper will be a critique of the line of thought behind the apparent problem, and of a certain picture of “epistemic rules” on which it rests. Once we’ve cleared away the problem, we can see the virtues of the approach, including the fact that it avoids unproductive issues that arise from fetishizing epistemic vocabulary such as knowledge and justification. The critical evaluation in the approach is truth-oriented, but avoids the many problems of reliabilism: both its refusal to recognize any “internalist” considerations and the fact that no notion of reliability seems adequate to encompass all the different factors we want our inductive practices to satisfy. The methodology fits best with a kind of normative anti-realism, about which I hope to say a bit at the end, and which provides another respect in which the approach is “naturalistic”.