In a standard definition of ‘knowledge’, knowledge requires justified belief. I take this to mean that, in order for me to know that p, my belief that p must be justified, that is; I must be justified in holding it or, to cast the point in terms of reasons, I must have reasons to hold this belief. This is true irrespective of whether the belief is empirical, moral, or mathematical, or whatever, in its content. I speak here only of justification as a necessary condition of knowledge, in order to avoid going into any problems to do with the Gettier paradox. These problems are of no relevance to my argument in this chapter. However, what does it mean for a belief to be ‘justified’?