Retour vers liste

Détail de la contribution

Auteur: Katja JASINSKAJA

Co-Auteur(s): Henk ZEEVAT, University of Amsterdam, The Netherlands

Titre:
The (A)symmetry of the additive particle And


Abstract/Résumé: In this talk we will develop the analysis of 'and' as an additive particle akin to 'too' and 'also'. The main idea is that 'and' signals that its conjuncts (a) share a property P, or answer the same topic question ?xP(x); and (b) are distinct instantiations of that property, or distinct answers to that question. We had shown before that this approach can account both for the fact that 'and' normally excludes subordinating discourse relations (elaboration, explanation) between its conjuncts, and for the famous exceptions to this generalisation noted by Horn. Furthermore, this view helps better understand the systematic relationships between additive, adversative ('but'), and correction markers (e.g. Spanish 'sino') and their polysemy patterns across languages, as the similarities and differences between the markers can be expressed in terms of the number and the type of variables (wh-words) in the topic question. However, it is obvious that the additive view of 'and' is essentially symmetric: nothing in the conventional function of 'and' assigns one conjunct a different status than the other. This means that the cases where 'and' shows asymmetric behaviour should be explained with reference to independent pragmatic mechanisms. In this talk we will discuss various external factors that make 'and' asymmetric, concentrating particularly on the cases that have been proffered in recent literature as counterexamples to the additivity-based approach. One of them is Winterstein’s (2010) observation that while in (2) the conjuncts of 'and' must refer to two distinct events (e.g. John had a great time playing football and then watched a movie), in (1) they refer to the same event (John had a great time watching a movie). This is in contrast to the asyndetic versions (3) and (4), where the "same event" interpretation is preferred regardless of the order. (1) John watched a movie and had a great time. (2) John had a great time and watched a movie. (3) John watched a movie. He had a great time. (4) John had a great time. He watched a movie. Winterstein proposes a solution embedded in Merin’s (1999) probabilistic reinterpretation of argumentation theory of Ducrot (1973). We will show how Winterstein’s solution can be integrated in our additivity-based approach. Another set of challenging data has been recently presented by Ariel (2012). We will show that part of the cases discussed by Ariel fall under the same probabilistic argumentative solution. The other part can be explained with reference to temporal iconicity and plain propositional informativity.