Grayling's question

July 31, 2007
Is solipsism inevitably self-defeating?

Solipsism is sometimes defined as the claim that one's mind is the only thing that exists, and sometimes that it is the only mind that exists. There is a world of difference between the two, but arguments rebutting the latter rebut the former also, though further steps are required to show that the other minds thus shown to exist are not the only things that exist.

Treated as a version of other-minds scepticism, solipsism is vulnerable to a beautiful line of argument proposed by PF Strawson, as follows. To doubt the existence of other minds, a sceptic must employ the concept of other minds, but he can do this only if he can distinguish between his own and other minds so that he can make intelligible the claim that only the former exist. But this in turn can only be done if there is a way of individuating minds, which means having criteria for distinguishing one from another. Whatever the criteria are (and Strawson's own version of the argument turns on the thesis that particulars of a special kind, namely persons, are required to supply them, "persons" being the congeries of psychological and physical attributes which are such that the presence of psychological states are identifiable by their physical correlates), they have to be available so that the sceptic can talk about what he calls "my conscious states." Since he is able to talk significantly about his own such states, there must be such criteria. Then if there are such criteria, the sceptic's doubts about the existence of other minds are empty, because even to articulate doubt about their existence he has to employ the discourse whose very conditions of employment legitimise what he wishes to call into question.

A similar argument can be adapted from Wittgenstein's "private language" and rule-following considerations. This is that because language is a rule-governed activity, and because rules can only be followed on condition that there are checks available for doing so correctly, and because a solitary person cannot provide such checks himself (at best he can only seem to himself to be doing so), it follows that language must be essentially—logically—public, that is, belong to a community of users. And since there is language (required for defining solipsism, for example), it cannot be the possession of a single user (a solitary mind).

What both strategies have in common is the recognition that the Cartesian procedure of inferring the existence of an external world and other minds from the private inner data of one's own consciousness is ultimately incoherent. And this is just another way of saying that solipsism under either of the definitions mentioned above is self-defeating.

Sent in by Hugh Witt, Cyprus.

Send your philosophical queries and dilemmas to AC Grayling atquestion@prospect-magazine.co.uk