A Corpus-informed Approach to Multimodal Pragmatics: Insights from an Exploration of Requests in B2 Spoken Language on Zoom

Abstract

Embodied actions have a key role to play alongside spoken interaction in pragmatic development when learning a second language. Multimodal corpora have much to offer researchers, language educators and learners within pragmatics. This methodological study outlines a multimodal corpus pragmatic approach to analyse a video-mediated dataset, gathered via the videoconferencing platform Zoom. Two English language learners at B2 level participated in the study. Elicitation tasks were employed through open roleplays centred on request sequences. The data was transcribed and then manually annotated in tiers with xml coding, first for spoken request sequences, then for embodied features relating to those requests (gesture, head movement, gaze patterns and facial expression). The corpus software AntConc and the audiovisual tool ELAN were used for transcription, annotation and recall. A function-to-form corpus pragmatic approach was used to investigate the spoken and embodied resources of the participants’ request sequence. Preliminary findings show that the participants augmented request sequences at key pragmatic junctures during video-mediated interaction with embodied resources. These include the use of paralanguage (e.g., exhaling, pauses, loudness, rapid speech), gesture (literal, deictic, stress and non-literal movements), gaze patterns and gaze aversion connected to politeness, and the use of face and eyes to convey hesitation in response.