1. The idea that Wittgenstein's later philosophy provides a refutation of scepticism about other minds is prima facie at odds with the emphasis that his remarks give to the uncertainty of our ordinary psychological language-game. The thought that the uncertainty of our language-game is ‘constitutional’ (RPPI §141; RPPII §657) may sound more like an endorsement of scepticism about other minds than a refutation of it. Similarly, the suggestion that ‘the following […] is true: I can't give criteria which put the presence of the sensation beyond doubt; that is to say: there are no such criteria’ (RPPI §137) seems to go against the idea that Wittgensteinian criteria are intended to allow us to establish the mental states of others ‘with certainty’.

Stanley Cavell was the first to draw attention to the prevalence of the theme of uncertainty in Wittgenstein's remarks and to question interpretations, such as Norman Malcolm’s, which found in them an argument for the necessity of criteria for mental states, which ‘establish beyond question’ the existence of the mental states of others. Cavell's reading of Wittgenstein sets out, on the one hand, to diagnose the way in which the sceptic about other minds misrepresents the obstacle to our achieving certainty and, on the other, to articulate the true nature of our predicament vis-à-vis others. The result is an interpretation which rejects the thought that Wittgenstein is committed to some form of common-sense rejection of scepticism about other minds, and replaces it with the claim that Wittgenstein accepts something akin to sceptical doubt already haunts our ordinary language-game. It is this claim I want to examine in this chapter.

2. The hope of the Malcolm reading of Wittgensteinian criteria is that the criteria on the basis of which we apply mental concepts to others are such that it does not make sense to suppose that criteria are satisfied and the person is not in the relevant mental state. Cavell points out that the claim that this concept of criteria provides a refutation of scepticism is made problematic by Malcolm's need to acknowledge that criteria cannot be expressed in the form of necessary and sufficient conditions, for it is always possible, at least in principle, to doubt whether criteria are satisfied.