Abstract

Analytic epistemologists, over the last two decades, have been troubled by two vexing and seemingly unrelated problems. First they have been trying to describe the "structure" of justification of our knowledge claims; this enterprise has been called the "foundationalism debate." Secondly though, they have also been trying to formulate precisely the necessary and sufficient conditions for "knowing;" this difficulty has come to be known as the "Gettier Problem." The first project deals with epistemic justification and the second, with the analysis of knowledge. The connections, if there are any, between these two research programs remain most obscure at present. And the silence with which most philosophers greet this question leads one to suppose that perhaps there are no interesting connections after all. I argue, however, that there are strong links between the two projects and that together they point us toward a workable "theory of justification." The foundationalism debate teaches us what makes a belief justified. And the Gettier-related literature is best construed, I contend, as a somewhat confused attempt to show us what kinds of justified beliefs are also likely to be true. We could say then that the foundationalism dispute discloses what makes our beliefs "justified," while the Gettier problem takes our theory of justification somewhat further. Recognizing that a belief can be justified and yet be false, the Gettier literature teaches us how to identify beliefs that are more than merely justified--beliefs that are, shall we say, "completely justified" and therefore especially likely to be true. Consequently, our theory must make a distinction between "justified" and "completely justified" belief and provide us with a model of each. The threading insight that binds the two models and that lies at the heart of the essay is the idea that justification should be understood as an essentially dialectical affair in which one meets the challenges of a critic