Foundationalism and the idea of the empirical.

This thesis is about foundationalism in epistemology. It distinguishes between different forms of foundationalism and defends one particular version of this doctrine. Chapter 1 gives an account of the motivations for foundationalism, including the so-called epistemic regress argument. It criticizes recent accounts of the core doctrines of foundationalism, such as those of Michael Williams and Ernest Sosa, and proposes a different account according to which foundationalism is the view that (a) some of our beliefs must be non-inferentially justified, (b) perception is a source of non-inferential justification, and (c) perception is a basic source of such justification. Chapter 2 gives an account of traditional foundationalism and tries to identify both what is right with it and what is wrong with it. It argues that the basic insight of traditional foundationalism can be detached from some of the other doctrines with which it was associated by the traditional foundationalists. That insight concerns the role of perceptual awareness or acquaintance as a regress-terminating source of epistemic justification. Chapter 3 exploits this idea in defending a more modest form of foundationalism according to which ordinary perceptual beliefs may be foundational. Chapters 3 and 4 focus on two influential arguments against the view that ordinary beliefs about the world around us can be non- inferentially justified by perception. The first argument trades on the alleged fallibility of perceptual justification, the second on its defeasibility. It is shown that neither argument poses a genuine threat to the more modest version of foundationalism that I defend. Chapter 5 compares perception with other sources of non-inferential justification such as memory and testimony. It defends the view that perception is a privileged source of non- inferential justification, even if it isn't the only source of such justification. It also contrasts foundationalism with traditional forms of externalism such as reliabilism and explains why the latter should not be counted as a form of foundationalism