The St. Elizabethan World

Abstract

I take the following to be a platitude about knowledge: there would be no problem of knowledge if everything always had been, were, and always would be just as we believe it to be. In one way or another the problem of knowledge is motivated by of our familiarity with the gap between appearance and reality. So it is an intuition that survives our reflections on knowledge that what we might call “an incorrigible world,” any world ω where for all propositions p, if S believes that p at ω, then p at ω, is a world where one knows to be true everything one believes to be true. One might hope that our best account of the nature of knowledge might explain why this platitude is platitudinous. If not, one would at least hope for the preferred account to be compatible with it. Unfortunately, coherence accounts of knowledge do not fulfill this hope, for they imply that there are some incorrigible worlds that are epistemically inaccessible. I will consider the coherence analysis offered in The Structure of Empirical Knowledge, by Laurence Bonjour as a representative case, showing why that analysis is inconsistent with the above platitude, and suggest why this is general problem for coherence theories. I end by suggesting that a solution to the problem is to abandon the presupposition present in many coherence theories that coherence properties are world-invariant.