Back to list

Detail of contribution

Auteur: Kathryn DAVIDSON

Aristotle isn’t only in Arizona: Coordination in American Sign Language

Abstract/Résumé: In "Aristotle Goes to Arizona and finds a language without 'and'", Gil (1991) discusses coordinators in Maricopa, which can be interpreted either as logical conjunction ('and') or logical disjunction ('or') depending on the context. In this paper I first present a similar phenomenon in American Sign Language (ASL), which has two strategies for coordination that can be interpreted either as disjunction or as conjunction depending on the discourse context, prosodic marking, or other cues like disambiguating lexical material. I first show that syntactically, each general use coordinator can be used to coordinate phrases at multiple levels (clauses, predicates, noun phrases), and in various clause types (alternative interrogatives, declarative clauses). I then argue against a semantic ambiguity approach to account for the conjunctive and disjunctive readings; instead, a Hamblin-style alternative semantics is proposed, where the disjunctive and conjunctive force comes from external quantification over a set of alternatives. Finally, because of its syntactic and semantic properties, general use coordination poses an interesting pragmatic question for the analysis of scalar implicatures, the inference most people make that if 'Mary had coffee or tea' then 'Mary didn't have coffee and tea.' Since Grice (1979) and Horn (1989), many analyses of such implicatures have involved the hypothesis that the lexical items form a scale where use of the weaker term or triggers the implicature that the stronger term and would be false. General use coordinators have separate semantics for conjunction and disjunction, but not separate phonological forms, so it's possible to test the role of the contrasting form of scalemates in generating scalar implicatures. To test for scalar implicatures generated by general use coordination, behavioral data was collected based on felicity judgements of video descriptions by a native signer (ASL) or speaker (English) describing a pictured scene. Data from 10 native signers of ASL indicate significantly decreased calculation of scalar implicatures based on the general use coordinator in ASL compared to implicatures based on scales with contrasting forms in ASL () and compared to the contrastive scale in English for 12 native speakers using the same task design. These results support the view that the instantiation of scale members with contrastive lexical forms is crucial for triggering this type of semantic/pragmatic inference. In particular, contrasting underlying meanings and even prosodic disambiguation (e.g. via eyebrow movement) is not enough to create a scale for the purpose of reliably generating scalar implicatures.