Journal of Philosophical Research (2006) 31: 143-152 Why Certainty is not a Mansion Elly Vintiadis Abstract In this paper I address Peter Klein‟s criticism of Wittgenstein in Certainty: A Refutation of Scepticism. Klein claims that, according to Wittgenstein, we attribute knowledge of a proposition p to a person only if that person is not certain of p. I argue that a careful reading of Wittgenstein‟s On Certainty reveals that there are two kinds of objective certainty that Wittgenstein had in mind; propositional objective certainty and normative objective certainty. Klein fails to distinguish between the two and uses what I call propositional objective certainty to make his point against Wittgenstein. I claim that when Wittgenstein said that knowledge and certainty belong to different categories he was talking of normative objective certainty and, therefore, that Klein‟s criticism is misplaced and attributes to Wittgenstein a position that is not his. In Certainty: A Refutation of Skepticism Peter Klein tries to refute skepticism by arguing that we can have absolute certainty based on non-deductive evidence and that the mistake from which the skeptical problem arises is the claim that in order for evidence to confirm a contingent empirical proposition the evidence must entail that proposition.(Klein 1981, xiii, 116) In the process of defending his position, Klein argues against some alternative approaches to certainty, one of which he attributes to Wittgenstein in On Certainty. The position he attributes to Wittgenstein is the view that we attribute knowledge of a proposition p to a person only if that person is not certain of p.(Klein 1981, 117-121) Klein claims that Wittgenstein‟s view is mistaken because, even though when we attribute knowledge to someone we usually do not also attribute certainty to him, it would not be false (though it might be odd and misleading) to say that someone knows and he is certain. To illustrate this Klein draws the parallel between this and saying “This is not a house, it is a mansion” and says that though it may be misleading to call a mansion a house, it is nonetheless not false. So Klein argues from analogy between: 1. X knows that p only if X is not certain of p. and, 2. X is a mansion only if X is not a house. He claims that, just as in the case of (2), though it would be strange, because it is not customary, to say of some structure that it is both a mansion and a house, so too “it may be misleading to call a certain proposition (i.e., a proposition which is certain) a known one, but it is nevertheless on that account not false that certain propositions are known propositions.”(Klein 1981, 120) If I understand him correctly, Klein‟s point is that though ordinarily uncertainty may be conversationally implied by knowledge claims it is not logically implied by them and that therefore Wittgenstein in On Certainty, as far as this is concerned, is wrong. In what follows I do not intend to argue against Klein‟s position vis-à-vis skepticism nor do I intend to touch upon larger issues that are of epistemic interest