Composition and projection of co-speech gestures

Abstract

In this paper I argue against the common assumption in recent literature that content-bearing gestures co-occurring with speech project in a single uniform way determined by their co-speech status. Instead, I propose a composition-driven, modality-neutral approach whereby which projection mechanisms are available or enforced for a given piece of compositionally integrated content, spoken or gestural, is determined by how it composes in the syntax/semantics. I adduce experimental data supporting this view from an acceptability judgement task comparing co-nominal gestures to adnominal adjectives and appositives. More broadly, this paper establishes the need to treat gestures as bona fide linguistic objects at all levels of representation in order to understand how they contribute to the meaning of utterances as well as which aspects of grammar are modality-(in)dependent.