What Else Justification Could Be1

According to a captivating picture, epistemic justification is essentially a matter of epistemic or evidential likelihood. While certain problems for this view are well known, it is motivated by a very natural thought – if justification can fall short of epistemic certainty, then what else could it possibly be? In this paper I shall develop an alternative way of thinking about epistemic justification. On this conception, the difference between justification and likelihood turns out to be akin to the more widely recognised difference between ceteris paribus laws and brute statistical generalisations. I go on to discuss, in light of this suggestion, issues such as classical and lotterydriven scepticism as well as the lottery and preface paradoxes. I. RISK MINIMISATION Some philosophers have claimed that, alongside standard Gettier cases, lottery cases provide further, vivid counterexamples to the traditional analysis of knowledge as justified, true belief (see Hawthorne, 2003, pp9, Pritchard, 2007, pp4). They reason along the following lines: Suppose that I hold a single ticket in a fair lottery of one million tickets. Suppose that I am convinced, purely on the basis of the odds involved, that my ticket won’t win. Do I know that my ticket won’t win? Intuitively, I don’t know any such thing, even if it happens to be true. Presumably, though, I have plenty of justification for believing that my ticket won’t win – after all, given my 1 I have benefited from discussions of this material with a number of people. Particular thanks to Philip Ebert, Daniele Sgaravatti, Stewart Shapiro, Ernest Sosa and Crispin Wright. Thanks also to an