Retour vers liste

Détail de la contribution

Auteur: Jakob M. STEIXNER

NC in Bavarian: Syntac and Information Structure

Abstract/Résumé: Empirical basis We re-examine Negative Concord in Bavarian and restrictions on its occurrence. Contrary to earlier analyses (Bayer, 1990; Weiß, 1999; Brugger and Poletto, 1993, i.a.), we claim that the emergence of obligatory DN-readings in certain constructions does not directly follow from the licensing mechanism of NC and structural considerations alone. In a sentence like "dass koa Linguist ned oft raalfoad" (lit. `that no linguist not often bicycle-rides'), the adverb ‘often’ in the scope of negation blocks NC whether preceeding or following ‘not’, argued to be symptomatic of a breakdown of local licensing of NC in a specifier-head configuration in Weiß (1999). This analysis, however, incorrectly predicts that "OFT/ foad \KOA Linguist ned raal" (`often rides no linguist not bicycle'), with the same scope and truth conditions, should be ungrammatical too: If the base position of ‘often’ is above NegP, and the topicalised adverb originates from a position below the negative DP, the latter cannot be parsed as in Spec.NegP irrespective of surface order. (Parallel effects appear with other interveners.) To accommodate such data, we propose that the negative interpretation is contributed by a covert negative operator as proposed for strict NC languages in Zeijlstra (2004); Penka (2011). N-words and the negative particle carry a syntactic [+NEG]-feature and require to be licensed by OP¬ under Agree, but do not themselves carry semantic negation. Such an interpretation comes at a cost: The ungrammaticality of Weiss' examples no longer naturally follows from the mechanism that is responsible for NC in the first place. We thus have to seek for independent explanations for their ungrammaticality. We argue that the sentences’ information structural properties provide such an explanation. A reinterpretation of the data in Weiß (1999) along these lines allows to capture Bavarian NC with unified cross-linguistically oriented theories of NC such as the Agreement approach pioneered by Zeijlstra (2004). A pragmatic account of restrictions on NC and the emergence of obligatory DN-readings opens the path for seeking parallels with DN-readings in other NC languages (i.a. Puskás, 2012; Vallduví, 1990; Espinal and Prieto, 2011, on Hungarian and Catalan), which have been described as driven by information structure in the literature. This approach thus offers a window for empirical study of grammar–pragmatics interactions in a wide variety of languages.