ftive component. The negative component is about how epistemology hasbeen done to date. The positive component is about how epistemologyshould instead be done. Take these components in turn.As Quine sees it, traditional epistemology requires that knowledge hadfoundations. These foundations are supposed to be beliefs that are justified but which are not justified by other beliefs. Traditional epistemologyexplores which beliefs are foundational in this sense. Suggestions aboutwhich beliefs count as foundational beliefs include certain perceptualbeliefs and beliefs about one’s current experiences. Traditional epistemology also requires that there are certain epistemic principles that show howjustification can be transmitted from foundational beliefs to other beliefs.Non-foundational beliefs are to be justified by being derived from thefoundational beliefs via these epistemic principles. Traditional epistemology then seeks to show how all and only justified beliefs belong to a structure as so described. These beliefs would either be foundational beliefs, orbeliefs suitably related by the epistemic principles to foundational beliefs.The goals of traditional epistemology are to show what a justified beliefis, which beliefs count as justified because of their place in a foundationalstructure, and how to improve our set of beliefs—which beliefs to retainand which to revise. A particular challenge to the traditional epistemologist’s project is posed by scepticism about justification: the view that few,if any, of our beliefs are justified. Traditional epistemology seeks to rebutthis scepticism. Quine takes the project of traditional epistemology to runfrom Descartes (1641) to at least Carnap (1928). (We should not overlook,though, the major differences between the epistemologies of philosopherssuch as Descartes and Carnap. Carnap took perceptual experiences tobe among the foundations of knowledge. Descartes did not. As notedabove, his foundations consisted only in what he took to be truths that arei