Abstract

The traditional account of knowing requires that the object of knowledge is true and is believed justifiably by the knower. Let us say that what a person believes is justified only if it is supported by evidence. In this thesis it is argued that justification conferring evidential relations are the logical relations of deduction and induction. ;The standard objection to this view holds that there are clear cases of justified believing in which the justifying evidence is not related to the object of belief either deductively or inductively, thus we must recognize another sort of evidential relation. The traditional case is that of perceptual beliefs justified on sensory evidence. ;The leading way of meeting this difficulty is by defending a principle asserting that believing itself generates epistemic warrant and, in the right circumstances, justification for what is believed. This strategy, often called "epistemic conservatism" can be found in various guises in the work of such prominent contemporary philosophers as Chisholm, Lehrer, and Davidson, as well as in the work of some naturalistic epistemologists. It is argued that each of these attempts fails. The conclusion is that this strategy fails and that we should look again at deduction and deduction as our evidential relations. ;Here it is argued that a plausible inductive argument can be constructed. Employing a model for inductive arguments formulated by Henry E. Kyburg, Jr. in his work on epistemological probability, an argument is offered and defended that reliable data garnered from our "interior" states can be used in establishing a statistical premise necessary for an inductive justification of perceptual beliefs.