In this chapter a theory of non-inferential knowledge will be developed. The restriction to non-inferential knowledge is not essential, and in a subsequent chapter the analysis developed here will be extended to inferential knowledge. But the case of non-inferential knowledge is peculiarly important. Thinking about the threatened infinite regress in the classical analysis of knowledge seems to lead to the conclusion that there must be non-inferential knowledge. Furthermore, we seem forced in the case of this sort of knowledge to look for some non-classical solution to the problem. (The classical or ‘Cartesian’ postulation of self-evident truths can rather easily be shown to be an insufficient account of the basis of all that we think we know.) If we can find a non-classical solution to the problem of non-inferential knowledge, where such a solution is clearly required, we may then try to extend the solution to cover all cases of knowledge.

What are the paradigms of non-inferential knowledge?

It would be extremely convenient if, in the course of discussion of non-inferential knowledge, one could point to uncontroversial examples of such knowledge. The examples could then be used in testing whether particular philosophical accounts of non-inferential knowledge are correct or not.

In the case of knowledge simpliciter, such uncontroversial cases are available. I know that the earth is round. I know that there is a piece of paper in front of me now. If a philosophical theory of the nature of knowledge yields the consequence that I do not know these things, then this is a conclusive (I do not say logically conclusive) reason for rejecting the theory.