KNOWLEDGE, DEFEASIBILITY AND THE GETTIER PROBLEM 1 Knowledge, Defeasibility, and The Gettier Problem 1 Klemens Kappel, Division of Philosophy, University of Copenhagen, Denmark, kappel@hum.ku.dk 1. Introduction Knowledge is defeasible in a way that relates directly to the famous Gettier Problem, or so I argue in this paper. To many readers it might not seem terribly surprising that knowledge is defeasible, but perhaps this is because they have the defeasibility of doxastic justification in mind. Since knowledge requires doxastic justification (whatever the exact nature of this), then if doxastic justification is defeasible, as many would assume, then so is knowledge. What I suggest in this paper, however, is that knowledge is defeasible in a way that does not depend on the purported defeasibility of doxastic justification. Even if doxastic justification were not defeasible, then knowledge would nonetheless be defeasible for the reasons that I will lay out below. And perhaps more surprising, but less important in the present context, even if doxastic justification is defeasible, this does strictly speaking not make knowledge defeasible, as I shall explain below. I hope that all of this is at least mildly surprising. Moreover, I will argue that there is an intimate connection between the sort of defeasibility that I have in mind and the Gettier Problem that has troubled modern epistemology since its inception by Edmund Gettier in 1963. 2 The basic idea I want to propose and defend is quite simple: some event that would constitute a Getter- style obstacle to a true and justified belief being an instance of knowledge is a defeater to that item of knowledge. And since such events are always possible, knowledge is defeasible. In this paper, I develop this view in some detail. 3 2. Defeater-structure Consider first what it is for something to be a defeater of something else. Here and in the ensuing discussion I will assume the following: Defeater-structure. Some object O and properties P and Q have a defeater-structure just in case (i) O has property P in virtue of having property Q, (ii) unless there is a defeater. A defeater is 1 A precursor of this paper was presented at the conference Defeasibility in Epistemology, Ethics, Law and Logic, Goethe- Universität, Frankfurt am Main, September 24-26, 2010. I would like to thank participants at that event for comments and inspiring discussion, especially (in alphabetical order): Claudia Blöser, Michael Janvid, Hannes Ole Matthiessen, Hans Rott, Jay Wallace, Michael Williams, Marcus Willaschek. 2 In his (Gettier 1963) 3 The main discussion of the Gettier Problem draws some material from my longer and more detailed discussion ’Knowledge, Truth by Luck, and What Solves the Gettier Problem’ (in preparation).