SIL Electronic Book Reviews 2003-021
URL: http://www.sil.org/silebr/silebr2003-021
Negation and polarity: Syntactic and semantic perspectives
Oxford and New York: Oxford University Press, 2000. Pp. 288. cloth $90.00, paper $47.00. ISBN 0198238746 (hardback), 0198238738 (paper).
Reviewed by Steve Nicolle
BTL (East Africa) and SIL International
There is a danger with collections such as this that the contributions will be either too diverse or too uniform. The result in both cases is an absence of internal dialogue. With one exception, this volume succeeds in generating a lively internal debate. Most of the papers assume some version of Chomskian syntax (since this is where much of the research into negation and polarity has been taking place); this shared perspective generates dialogue, for example, between Principles-and-Parameters and Minimalism, and between the relative importance of syntactic and semantic factors. There is also an emphasis on data throughout, so even when the dialogue is at an abstract, theoretical level, it remains rooted in linguistic facts, albeit almost exclusively from European languages and Japanese. In short, this is recommended reading. A summary of the contributions follows.
Liliane Haegeman (“Negative preposing, negative inversion and the split CP” pp. 21–61) seeks to explain the contrast between (1a) and (1b), and similar contrasts in West Flemish and Italian (p. 21):
| (1a) | With no job would Mary be happy. |
| (1b) | With no job, Mary would be happy. |
She proposes that in (1a), which can be paraphrased as ‘Mary wouldn’t be happy with any job,’ the preposed PP with no job is a focus operator and that in (1b), equivalent to ‘Mary would be happy without a job,’ it is a topic. Negation in (1a) has scope over the whole sentence, as (2a) demonstrates (it licenses the polarity item ever), and is subject to the NEG-criterion (Haegeman 1995) at the sentence level which triggers inversion. In (1b) the NEG-criterion is not relevant, as the scope negation is confined to the preposed PP and does not license a polarity item in the associated clause (2b):
| (2a) | With no job would she ever be happy. |
| (2b) | *With no job, she would ever be happy. |
In Japanese, both (1a) and (1b) are invariably expressed by sentential forms, and Japanese also exhibits a different distribution of negative polarity items (NPIs) from English (for example, NPIs are licensed by conditional clauses in English but not in Japanese). Yasuhiko Kato (“Interpretive asymmetries of negation” pp. 62–87) proposes a unified analysis of these two sets of asymmetries between English and Japanese within the Minimalist framework. He argues, against Haegeman’s position, that the NEG-criterion follows from his proposed analysis, rather than being basic (as in Principles and Parameters).
The volume continues with two more chapters which apply different approaches to the analysis of the same problem—in this case the role of syntactic conditions, notably c-command, in the licensing of NPIs. Ljiljana Progovac (“Coordination, c-comand and ‘logophoric’ n-words,” pp. 88–114) notes that ‘n-words’ (that is, negative expressions like nobody) in negative-concord languages such as Serbo-Croatian can remain unlicensed in conjuncts and adjuncts, just as some reflexive expressions may. She proposes that in negative-concord languages, n-words which participate in negative marking of the predicate require a c-commanding negative, just as reflexives that participate in reflexive-marking require a c-commanding antecedent; n-words that do not mark a predicate as negative do not require such a negative licenser (pp. 108-9).
In contrast, Jack Hoeksema (“Negative polarity items: triggering, scope and c-command,” pp. 115–146) argues, from “problems in areas such as conjunction, topicalization, subject-object asymmetries, VP-internal positions, NP-specifiers and adjuncts, and predicational structures” (p. 143), that the scope of negation is semantically derived, rather than being determined by c-command or any other s-structure constraint. Hoeksema’s chapter is possibly the most accessible contribution to the volume, and provides a clear introduction to NPIs and their study. It could usefully be read in conjunction with the introduction by anyone not familiar with the background to research into NPIs.
Laurence R. Horn (“Pick a theory (not just any theory): indiscriminatives and the free-choice indefinite,” pp. 147–192) contrasts not only (presuppositional and optionally scalar) and not just (non-presuppositional and obligatorily scalar). He uses the scalar nature of not just as the basis of an analysis of any, arguing that the negative polarity reading of any in I didn’t do anything and the free-choice any in Anything can happen are both non-quantificational, indefinite, end-of-scale determiners. The difference in interpretation between the NPI and free-choice uses of any parallels that between generic and ordinary uses of indefinite descriptions (p. 181).
Semantic and pragmatic explanations are also proposed by Paul Portner and Raffaella Zanuttini (“The force of negation in wh exclamatives and interrogative,” pp. 193–231) in their analysis of expletive negation in Paduan (an Italian dialect spoken in the city of Padua), in which the negative expression seems to lose its semantic force (p. 193):
| (3) |
|
I will pass over the final two contributions. William A. Ladusaw’s “Thetic and categorial, stage and individual, weak and strong” (pp. 232–242) originally appeared in SALT 4 (Papers from the Fourth Annual Conference on Semantics and Linguistic Theory, 1994); Masa-aki Yamanashi’s “Negative inference, space construal, and grammaticalization” (pp. 243–254) departs from the synchronic and generative tone of the rest of the volume and discusses the cognitive origin of Japanese negative markers in spatial terms. The book concludes with a comprehensive “further reading” list.
References
Haegeman, Liliane. 1995. The syntax of negation. Cambridge: Cambridge University Press.