Speaker: Daniel Greco (Yale University)
Date and Time: Thursday, March 21st 2024, 16:30-18:00
Venue: Online only
Title: Idealization in Epistemology
Abstract: I’ll present some material from my recently published book, Idealization in Epistemology: A Modest Modeling Approach. After explaining what I mean by “modest” modeling, and why I take it to provide an attractive framework for thinking about epistemology, I’ll apply that framework to two specific debates. First, I’ll consider the objections to Bayesian models of learning that they go wrong in representing the inputs to learning as certain (strict conditionalization), or, even when not certain, as immune to undermining defeat (Jeffrey conditionalization). I’ll argue that these objections should trouble us much less once we’re modest modelers. Second, I’ll consider the argument that it’s computationally infeasible for limited agents like us to make extensive use of probabilities in thought. I’ll argue that this argument targets an implausibly immodest vision of the cognitive role of probabilistic thinking. When aimed at an appropriately modest conception of the role of probabilities in both descriptive and normative decision theory, the argument fails.