Dretskean sensitivity and induction
摘要
The sensitivity principle in epistemology has faced numerous, considerable, and relentless challenges since it emerged in Nozick’s Philosophical Explanations (Nozick, Philosophical Explanations 1981). In this paper, I develop a version of sensitivity, based on Dretske’s notion of conclusive reasons (Dretske, Australasian Journal of Philosophy 49, 1971), that responds to the complaint that sensitivity is either incompatible with or makes an unprincipled mess of inductive knowledge. There are three key moves in formulating reasons-based Dretskean sensitivity (RDS). First, sensitivity is conceived in terms of reasons rather than beliefs tracking truth. Second, focus shifts from whether S would have those reasons in the relevant counterfactual worlds to whether those reasons would be the case. Third, closer attention is paid to the structure of reasons, such that S can know that p on the basis of a reason R because R tracks p, can know that R on the basis of experience because experience tracks R, even if experience does not track p. While critics of Nozick are correct to point out that, typically, S would have their actual world observational basis for belief—their experience—if their true induction-based belief were false, this does not preclude inductive knowledge according to RDS.