Code-blending with depicting signs

Abstract

Abstract Bimodal bilinguals sometimes use code-blending, simultaneous production of (parts of) an utterance in both speech and sign. We ask what spoken language material is blended with entity and handling depicting signs (DS), representations of action that combine discrete components with iconic depictions of aspects of a referenced event in a gradient, analog manner. We test a semantic approach that DS may involve a demonstration, involving a predicate which obligatorily includes a modificational demonstrational component, and adopt a syntactic analysis which crucially distinguishes between entity and handling DS. Given the model of bilingualism we use, we expect both DS can be produced with speech that occurs in the verbal structure, along with vocal gestures, but speech that includes a subject is only expected to be blended with handling DS, not entity. The data we report from three Codas, native bimodal bilinguals, from the United States and one from Brazil conform with this prediction.