Classical NEG Raising: An Essay on the Syntax of Negation

20. Conclusion

Conclusion We summarize a few of the consequences of general importance that we believe follow from the arguments in the previous nineteen chapters. We need not linger over the fact that we have provided considerable new evidence for the syntactic nature of Classical NR, especially the evidence based on island facts, Horn clauses, and parenthetical clauses. Nor will we recapitulate our demonstration that the composed quantifier argument has no force against our syntactic view of Classical NR and that some of the data from that domain (islands, Horn clauses, and parentheticals) actually strongly support our syntactic view. Instead, we highlight some of the more general theoretical implications of our work. First, following Postal 2005, we have argued that there are two types of NPIs: unary-NEG structures and binary-NEG structures (reversals). These are illustrated in (1): (1) a. I didn’t see any pyramid. (unary-NEG structure) b. Everybody who knows any physics . . . (binary-NEG structure) In (1a), the underlying structure of any pyramid is [[NEG SOME] pyramid]. NEG raises away from any pyramid to a position right-adjacent to the finite Aux. In (1b), the underlying structure of any physics is [[NEG2 [NEG1 SOME]] physics]. The two NEGs are deleted in a NEG deletion chain initiated by every. In chapter 4, we gave various diagnostics supporting this distinction (several of which expand the set found in Postal 2005). In addition, the distinction between unary- and binary-NEG NPIs played a crucial role in analyzing the scope of NPIs (see chapter 9), the distribution of NPIs in islands (see chapter 12), and the interaction of NPIs with the excluded middle property (see chapter 16). As we showed in chapter 14, the distinction between unary- and binaryNEG NPIs receives striking support from the analysis of Horn clauses: specifically , only unary-NEG NPIs trigger Negative Inversion in Horn clauses. 20 218 Chapter 20 Consider this result from the point of view of current theories of NPIs. One common approach analyzes them as existential quantifiers: (2) a. (G. Carlson 1980:800) “Any is an existential quantifier which must appear within the scope of negation.” (“. . . for those instances where not is the triggering element.”) b. (Linebarger 1980:217) “The three arguments presented here in support of the E-theory (in which polarity any is represented as an existential quantifier in the scope of NOT) . . .” A second approach analyzes them as indefinites introducing a variable: (3) (Ladusaw 1996a:211) “Since Klima 1964, these NPIs have been termed as ‘indefinites’, and it is best to interpret this term in the sense of Heim 1982.” Under either approach, the NPI is analyzed as an indefinite, a point highlighted in the following passage: (4) (Kadmon and Landman 1993:357) “The NP with any has the usual semantic properties of an indefinite. We won’t make a choice here concerning the proper way of treating indefinite NPs. If indefinite NPs are best regarded as existential quantifiers, then so is the any NP; if indefinite NPs are best treated as new variables (Heim 1982), then the any NP too is a new variable. From either assumption the existential behavior of PS [polarity sensitive] any follows without problems.” Under either approach, the behavior of Negative Inversion in Horn clauses is a mystery, for the simple reason that indefinites never trigger Negative Inversion (see chapter 14): (5) a. I lent some money to a/some man I know. b. *To a/some man I know did I lend some money. Second, while it is widely assumed that natural language negation is a clausal modifier (negating the proposition denoted by the clause), as in Klima’s (1964) influential work and many others, we have described a broad range of English negation facts with no appeal to clausal negation at all. Under our analysis, the standard cases usually taken to motivate English clausal negation actually represent NEGs raised from verbal/adjectival phrases to a position right-adjacent to Aux. And we have assumed a rich system in which NEGs also originate on other categories, most notably D (SOME). We argued in Conclusion 219 chapter 14 that the correct analysis of Horn clauses in particular is inconsistent with analyzing the NEG right-adjacent to Aux as clausal negation. We take the success of our accounts to support the view that clausal negation —with its semantic correlate, propositional negation—does not have the primary role in natural language negation that is often assumed...