Speaker: Prof. Hanti Lin (UC Davis)
Date: September 20th 2017
Venue: Seminar room No.9, 1st floor of Research Building No.2
Title: Hume’s Dilemma and the Normative Turn—Or How It Is Possible to Justify at Least Some Kind of Induction
Is it possible to justify at least some kind of induction? Hume’s dilemma tries to answer in the negative; a simple version goes like this: “To justify an arbitrary kind of induction, the *empirical* thesis that it will (always or often) lead to a true conclusion has to be argued for, either demonstratively or inductively; the demonstrative route is impossible, while the inductive route is circular.” I want to resolve this dilemma by defending a quite general escape route. Here is the idea: (i) to justify induction of a certain kind, we can argue for a non-empirical, *normative* thesis instead, a norm that guides some inductive practices; (ii) unlike empirical theses, normative theses might be justified a priori and demonstratively, without relying on empirical studies or inductive inferences. Call this the normative turn, which has been implemented in various ways by some (formal) epistemologists, such as Bayesians, learning theorists, and Reichenbach (who is probably the earliest pioneer of the normative turn). Unfortunately, those people tend to set aside Hume’s dilemma quickly and rush to develop their own implementations of the normative turn. In their hands, the normative turn is mostly practiced but not really defended. So I want to defend the normative turn—to consider possible ways Hume’s dilemma might be thought to strike back, and to address those worries by reference to the general features of the normative turn, without commitment to any particular implementation.
A note on the mathematical prerequisite: I will keep it to a minimum. You only need to have propositional logic in mind, and I will prepare all the others for you, pictorially.