Stance-taking in multimodal interaction

One of the key achievements of taking part in an interaction is positioning yourself on some issue, idea or object, relative to other people; to be able to agree or disagree with someone about politics, or art, or something as mundane as food preferences. The exchanging and co-creating of evaluations and attitudes - the practice of stance-taking - is a critical component of social interactions. Stance is a fundamental component of interaction; it is both constitutive of interactions, in that stance-taking forms part of an interaction (and other actions / utterances are dependent upon it), and is co-constructed by participants through ongoing interactional means. This study considers multi-modal interaction and the achievement of stance through non-verbal behaviour and through the combination of non-verbal and verbal behaviour. Data for this study is drawn from the AMI Corpus (corpus.amiproject.org) non-scenario meetings. This paper argues that gesture and words can be seen to be deployed by participants to index social actions, particularly evaluative stance, and that word+gesture combinations (understood as any combination of verbal and non-verbal behaviour) can be used as single indexical expressions by speakers in conversation. This paper also explores implications for developing gestural detection and deployment of evaluative stance in HCI applications, such as in avatar-based immersive virtual environments.

Authors: Guy Edwards, Katharine Parton

Event: SF08: Embodied Interaction in Mobile, Physical and Virtual Environments Workshop

← View all submissions for this event.