Gestural viewpoint signals referent accessibility AbstractThe tracking of entities in discourse is known to be a bimodal phenomenon. Speakers achieve cohesion in speech by alternating between full lexical forms, pronouns, zero anaphora as they track referents. They also track referents in co--speech gestures. In this study, we explored how viewpoint is deployed in reference tracking focusing on representations of animate entities in German narrative discourse. We found that gestural viewpoint systematically varies depending on discourse context. Speakers predominantly use character viewpoint in maintained contexts and observer viewpoint in re--introduced contexts. Thus, gestural viewpoint seems to function as a cohesive device in narrative discourse. The findings expand on and provide further evidence for the co--ordination between speech and gesture on the discourse level which is crucial to understanding the tight link between the two modalities.

No abstract

The literature on bimodal discourse reference has shown that gestures are sensitive to referents' information status in discourse. Gestures occur more often with new referents/ first mentions than with given referents/subsequent mentions. However, because not all new entities at first mention occur with gestures, the current study examines whether gestures are sensitive to a difference in information status between brand-new and inferable entities and variation in nominal definiteness. Unexpectedly, the results show that gestures are more frequent with inferable referents (hearer new but discourse old) than with brand-new referents (hearer new and discourse new). The findings reveal new aspects of the relationship between gestures and speech in discourse, specifically suggesting a complementary (disambiguating) function for gestures in the context of first mentioned discourse entities. The results thus highlight the multi-functionality of gestures in relation to speech.

No abstract

No abstract