Abstract

Epistemologists assume that the primary task of epistemology is to explicate justification, and that there is a univocal, unified, pre-theoretical notion of epistemic justification that all are trying to account for. I argue that epistemologists are mistaken on both counts. The notion of epistemic justification is tied fundamentally to the what I will call an "epistemic accident." Consider someone who holds a true belief, p. There are two ways that the holding of this belief can be accidental. Given that she believes it, it could be accidentally true, or given that it is true, it could be an accident that she believes it. Either of these kinds of accident undermines epistemic justification. However, the two kinds of accident are distinct, and accounting for them gives rise to two quite different conceptions of justification. ;Furthermore, since each conception captures one kind of epistemic accident, they each have equal claim to being the "correct" conception of justification. I argue that, unfortunately for theorists of justification, it is unlikely that a combined theory can account adequately for both kinds of accidents. Thus, I conclude that what epistemologists call "justification" is really a composite of several distinct and theoretically irreconcilable evaluations. ;I recommend that we abandon the quixotic task of explicating "the" notion of justification and instead focus on more fundamental epistemic evaluations whose objects and other parameters are clearly defined. To identify such evaluations, I provide a taxonomy of the evaluations proper to epistemology. By appeal to such a taxonomy, one can identify the objects one is attempting to evaluate and the perspective from which one is doing the evaluation, thus avoiding the kind of conceptual confusion present in the debate over justification. ;This raises the issue of epistemic normativity, because many of the evaluations distinguished in my taxonomy do not obviously bear any appropriate relation to truth. Though a much-neglected issue in epistemology, a common assumption among epistemologists seems to be that epistemic normativity derives from the intrinsic goodness of true beliefs. I deny this assumption and argue that truth alone cannot explain the rich normativity of epistemic evaluations