In this section I seek to wrap things up by drawing together some thoughts about the differences between a theory of language and a theory of meaning, strictly understood, and also by making note of the important role that a grammar of the verb "means' and its cognates has to play in the latter endeavor. At this point, it might be felt that the restriction of the true—false poles to clausal negation still allows for such stalwarts as no-frills subject—predicate sentences to be construed as unproblematically truth-conditional — a class that on any view constitutes an enormous chunk of language. But even this is not invariably so. As Dummett observes, negation in ordinary English is standardly effected by negating the predicate. He concedes that the logical sentence-operator is not meant to conform precisely to the structure of English sentences in which "not" occurs (cf. 1978: 27-28, and 1981a: 421, 424). In natural discourse truth and falsity are thus tied directly to the predicate — that is, to the main verb and its complement — and not to a sentence or truth-evaluable proposition as a whole. Likewise, as the scope of assertion is determined by the application of truth and falsity, here again we naturally fasten upon the predicate.