ABSTRACT
I respond to Joachim Horvath’s paper ‘Gettier’s Thought Experiments’, which offers a ‘suppositional’ reconstruction of Edmund Gettier’s arguments against the justified true belief analysis of knowledge. I make three main objections to Horvath’s reconstruction. First, it exacerbates the problem of deviant realizations of Gettier’s cases, thereby unnecessarily invalidating Gettier’s arguments and making them hard to repair. Second, its treatment of Gettier’s use of fictional names is highly artificial. Third, it analyzes Gettier’s arguments in terms of a notion of truth relative to a supposition subject to truth-value gaps and gluts, while providing no metalogical framework for handling them. These features undermine Horvath’s claim that his reconstruction is simpler and more faithful to Gettier’s text than its rivals. I also explain how my new account of the semantics and pragmatics of counterfactuals in Suppose and Tell provides simpler and more elegant treatments of deviant realizations and anaphora between antecedent and consequent than in The Philosophy of Philosophy, thereby streamlining my account of thought experiments in terms of counterfactual conditionals.
