The practical explication is employed to explain why accidental fulfilment of the conditions for knowledge leads us to withhold the ascription of it, and what is meant by accidental in this context. The inquirer wants her informant to have some detectable property, X, possession of which correlates well with being right about p, and for this correlation to be law-like, and for the continuation of the correlation in any given instance to be non-accidental. At this point, a dilemma arises: either X must entail that S is right about p (too strong), or X must give a high probability of being right as to p (but then it is possible for X to be present but for S to be wrong on p). This property of p, inherent in the explicated concept of knowledge, thus mirrors a feature which, to judge by the discussion of the Gettier problem, the analysed concept has too.
Keywords: accident, defeasibility, explication, Gettier, knowledge, probability