Recent discussions of the nature of knowledge have yielded two divergent paths to follow. On the one hand are reliabilists and other externalists, and on the other hand are justificationists and internalists. Sometimes members of the former camp claim to be offering a theory of justification, but for present purposes I am going to treat such claims as idle explanatory wheels of their theoretical machine. If they turn out to be right, it will make their theories no better than they would have been as theories of knowledge, and if they are wrong, their theories of knowledge might still survive the amputation of this aspect.

We shall see in this chapter that ordinary reliabilist theories of knowledge cannot explain the value of knowledge over true belief, but this point is not my ultimate target. My ultimate target is to determine whether a justificationist account of knowledge can do any better. I will argue that it can, but that the picture of justification required for success is quite unusual. The conclusion will be a slightly disturbing account of what a theory of justification must look like if it is to have a chance of accounting for the value of knowledge.

EXTERNALIST ACCOUNTS

To see the difficulty of providing a good answer to the question of the value of knowledge over true belief, consider the following theory of knowledge. On this theory, what closes the gap between true belief and knowledge is objective likelihood of truth for belief.