Philosophical Investigations, reading it together. Also, as an aside, I've been reading Sara Ellenbogen's Wittgenstein's Account of Truth, and there's a passage in there I really like, and though it relates more to the sections we already finished covering (§88-§133, on 'theories' and philosophy), we're close enough to thsoe sections that I feel comfortable posting this. It's a passage on why we cannot construct systematic theories of meaning:
"Wittgenstein ... would have rejected the thought that language should be capable of systematization. Hence, he would have denied that we can give a uniform account of what it means to treat a sentence as assertible which can be applied across the board, in a uniform way, in all the contexts in which we use “is true.” As he argues, the notion that meaning can be explained without reference to anything other than use arises from the fact that “in our discussions, [we] constantly compare language with a calculus proceeding according to exact rules” (B.B., p. 25). And we are inclined to say that the meaning of a word must be fixed and precise in order for it to be intelligible (cf. P.I. #79).
But we should remember that “in general, we don’t use language according to strict rules—it hasn’t been taught us that way either” (B.B. p. 25). In practice, we do not always use names with a fixed meaning—we may use the name “Moses” without a predetermined sense of which descriptions we are willing to substitute for it. And this does not detract from the usefulness of the name in our language (P.I. #79). Similarly, we may say to someone “Stand roughly there,” and the inexactness of the expression does not make it unusable (P.I. #88). The person we are addressing will know what we mean and what he must do to satisfy the request. Examples such as these should suggest to us that what determinate meaning requires is not conformity to a universal standard or model. Rather, it requires an understanding of what is needed by those concerned in a given context.
....When we try to construct a systematic theory of meaning, it appears as though there is something outside of language by reference to which we can explain what meaning is. That is, it appears as though there is something independent of our actual use of words in language which bestows meaning on them—and which does so in a uniform way."