In several place, most notably Alciphron 7, Berkeley seems to think that the meanings of many, if not all, terms are given by the rules for correctly applying them. He doesn’t seem to mean the conditions under which they are true. Rather, he seems to mean the rules actual speakers apply in deciding to use the word. We’re not talking about mere disquotation; we have to give conditions that speakers can actually use when deciding whether to utter sentences. So, to use one of Berkeley’s favorite examples, the meaning of the symbol ‘i‘ in algebra is given by the formula i^2=-1. Similarly, it seems to me, the meaning of ‘force’ in Newtonian physics is given by the inference rules which allow us to derive theorems about force.
Now consider the predicate ‘true’. Surely the rule for applying this predicate is just the disquotation schema:
‘p‘ is true iff p
(Pretend there are corner quotes around each side of the iff.) The view that the disquotation schema is just all there is to be said about truth is called the deflationary theory of truth, and was endorsed by a number of 20th century thinkers. Would Berkeley endorse that view? I think he’s under at least a little bit of pressure to do so, in virtue of his general views about language, especially since ‘true’ clearly doesn’t stand for an idea. However, here’s a reason for thinking he shouldn’t be a deflationist. In his notebooks, Berkeley flirted with a proto-Humean bundle theory of the self. In later entries he decisively rejects this view, and in the Principles he is at pains to secure the claim that there is, in addition to our perceptions, a thing which perceives. Part of the case he has to make is that the word ‘spirit’ is meaningful despite not standing for any idea, and he does so, I think, by examining the ‘notion’ of perception. This is a notion of a two place relation, so we get the perceiver almost for free: nothing can be a perception unless something perceives it. Now, it’s clear that we can use the word ‘perceive’ (or some other similar sign) in this way if we want, and this will allow us to associate fairly well-defined assertability conditions with sentences containing ‘spirit’ or ‘self’. However, it seems that we could just as easily adopt a different notion of perception which did not require perceptions to have perceivers, and we could have well-defined assertability conditions for this as well. If the deflationary theory of truth is correct, then it is not clear what it could mean to say that one of these notions was the right one and the other was wrong. In other words, this would make Berkeley’s rejection of the bundle theory of the self simply a decision to adopt one ‘language-game’ rather than another. Personally, I don’t think the good Bishop would be at all pleased with this result.
(Cross-posted at blog.kennypearce.net)