In lieu of an abstract, here is a brief excerpt of the content:

705 XV The Distorted Treatment of Phrase Theory in Modern Formal Grammar In Chapter XIV we saw how discourse and sentence structure shape phrase structure, rather than phrase structure shaping sentence structure. By contrast, in Chomsky’s minimalist approach, beginning with morphemes , in successive stages, ever larger phrasal groups are formed until sentential “phrases” are reached. In this bottom-up approach the properties of each lexical item as set within the lexicon determine the ways it can be joined (“Merge”) with other items to form the successive complexes and determine how these can themselves be joined. At various stages in this process certain restricted kinds of “Movement” (“transformations”) can occur. Chomsky has provided successively more sophisticated construction manuals for constructing sentences as static entities, but what we need instead is a grammar which models what is spoken as something live, not as something static. We are deceived through always looking at the written word on paper. Yet even in the act of reading and understanding we have something dynamic. The constructional model of grammar has been reinforced in two ways. First, the X-Bar theory of lexical phrases with heads appeared to identify striking parallelisms between noun, adjective, verb, and prepositional phrases. However, as I show in section 1, noun phrases with complements arise in two quite different ways. A first group express relations—for instance, “the father of Isaac” and “a wing of a bird”. The relative terms involved are attributive complements of the copula, as in “Abraham was a father” and “This is a wing”. 706   Rewriting the philosophy of grammar A second group, along with adjective phrases, recapitulate the structures of verb phrases in which they are both semantically rooted. As for prepositions, we have seen how their essentially synlexical character, linking some other element to a noun, adjective, or verb, makes it quite wrong to treat them as heads of lexical phrases parallel to nouns, adjectives, and verbs. (Whatever I say of prepositions in this chapter applies to postpositions, as well.) Semantic syntax is what is fundamental for each language, and morphosyntactic restrictions consequential. Second, it was thought fruitful to extend the application of X-Bar theory to supposed “functional phrases” with “functional heads” alongside lexical phrases, and to link the theory of Movement with this larger framework. In this way the earlier mistakes, made possible by Chomsky’s conception of the autonomy of syntax vis-à-vis semantics, were compounded. In section 3 I show that this notion of functional phrases with functional heads is deceptive, since there is no real analogy between lexical heads and socalled functional heads, and it explains nothing which is not better explained in other ways. I discuss Chomsky’s ideas in their “Principles and Parameters” form, because the same objections apply pari passu to the corresponding ideas presented in minimalist guise for reasons indicated in chapter XIII, section 7 (b).1 Section 1. A Preliminary: Kinds of “of”-Phrases (a) Two sorts of “of”-phrases Complementary phrases can be considered in two different ways: (a) morphosyntactically , whereby the complement concerned (e.g., “tigers”) is the complement of a preposition, and the prepositional phrase as a unity (e.g., “of tigers”) is considered to be the complement of some noun, adjective, or verb; and (b) semantically, whereby essentially the complement concerned (“ti1 . In non-transformational formal grammars such as Lexical Functional Grammar (discussed later) and Generalized Phrase Structure Grammar, X-Bar syntax takes varied forms and applies only to lexical phrases. Chomsky never gives it any precise statement, and in itself it explains nothing, as Geoffrey K. Pullum demonstrates in “Assuming Some Version of X-Bar Theory”, Chicago Linguistic Society 21, no. 1 (1985): 323–53. In functionalist grammars, it is rejected altogether. Dik has an extended discussion of degrees of nominalization (Theory of Functional Grammar: Part 2, 162–68), and regards the verbal “the enemy destroyed the city” as underlying the nominal “the enemy’s destruction of the city”, the relation to be explained in functional and semantic, not morphosyntactic, terms. The same root and branch rejection of X-Bar syntax even in regard to lexical phrases is to be found in Halliday, Introduction to Functional Grammar, 1st ed., 23–29, partly anticipated in Halliday, “Categories of the Theory of Grammar”, Word 17, no. 3 (1961): 241–92, Van Valin and LaPolla, Syntax, 53–67. Van Valin remarks that Role and Reference Grammar “rejects X-bar syntax and assumes instead a multi-projection, semantically based theory of clause structure which...

Share