The Epistemic Virtues of Consistency

The lottery paradox has been discussed widely. The standard solution to the lottery paradox is that a ticket holder is justified in believing each ticket will lose but the ticket holder is also justified in believing not all of the tickets will lose. If the standard solution is true, then we get the paradoxical result that it is possible for a person to have a justified set of beliefs that she knows is inconsistent. In this paper, I argue that the best solution to the paradox is that a ticket holder is not justified in believing any of the tickets are losers. My solution avoids the paradoxical result of the standard solution. The solution I defend has been hastily rejected by other philosophers because it appears to lead to skepticism. I defend my solution from the threat of skepticism and give two arguments in favor of my conclusion that the ticket holder in the original lottery case is not justifed in believing that his ticket will lose. Recently, there has been a lot of commotion about whether or not a person could be epistemically justified in believing a set of statements that she knows is inconsistent.1 Richard Foley, Peter Klein, and Henry Kyburg have each given arguments that seem to show that one can be epistemically justified in believing each member of a set of statements that is known by her to be inconsistent.2 That's an extremely important, yet disturbing conclusion. Many of us think that being consistent is an epistemic virtue. With the exception of Gilbert Harman's negative coherence theory of justification, all coherence theories are committed to the claim that coherence is at least a necessary condition for epistemic justification. Many foundationalists have explicitly made commitments to the assumption that consistency is a necessary con dition for justification.3 It seems that at least part of what makes a belief or set of beliefs justified is that it fits together with one's other justified beliefs. If you know a set of your beliefs is inconsistent, you know you've made a mistake. To continue believing in the face of this known inconsistency seems only to add to your epistemic troubles. If Foley, Klein, and Kyburg are correct, many of us will be forced to give up a strongly held and seemingly quite plausible assumption about the virtues of believing consistently. Their arguments rest on their answers to the Preface Paradox and an epistemic version of the Lottery Paradox. In this paper, I will examine a purely epistemic version of the Lottery Synthese 109: 121-141, 1996. ? 1996 Kluwer Academic Publishers. Printed in the Netherlands. This content downloaded from 204.52.215.70 on Sat, 20 Jul 2013 16:01:26 PM All use subject to JSTOR Terms and Conditions