DescriptionPublished as Coyote Papers: Working Papers in Linguistics from A-Z, Studies on Arabic, Basque, English, Japanese, Navajo and Papago
AbstractJackendoff (1977) makes an assumption that a clause is a projection of the lexical category V, and he analyze the traditional category S as the major phrasal category of V, i.e., V3 . Such an analysis is necessary in order to capture, for example, the parallelism in the grammatical relations between clause and noun phrase. He furthermore proposes that the major phrasal category, V3, dominates sentential adverbs, sentential appositive relatives, parenthetical clauses, subject NPs, and auxiliaries. Syntactically, however, these constituents do not seem homogeneous. Section 1 of this paper shows that sentential adverbs (S-adverbs, henceforth), not being homogeneous, must be divided into two syntactically different groups, and claims, on the basis of this distinction, that it is necessary to add one more layer, V4, to Jackendoff's maximum layer, V32, so as to put one group of S-adverbs, as Jackendoff claims, under V3, and the other group (in addition, some other kinds of constituents) under V4. Section 2 discusses that these two and some other groups of adverbs occur in different environments: some groups of adverbs, but not others, may occur in given types of clauses. The discussion about which group of adverbs occurs in which type of clauses will give a crucial clue for deciding internal structures of each type of clauses. Section 3 extends the V4 system to the analysis of adverbial subordinate clauses, and, in passing, refers to the relation of the V4 system to "a bounding category" in the sense of Chomsky (1979) and the notion of "command." The argument on the bounding category leads us to interesting phenomena concerning WH-Movement, which will be dealt with in Section 4.