Primary and secondary pragmatic functions of pointing gestures

This article presents a study of a set of pointing gestures produced together with speech in a corpus of video-recorded locality description interviews in rural Laos. In a restricted set of the observed gestures (we did not consider gestures with special hand shapes, gestures with arc/tracing motion, or gestures directed at referents within physical reach), two basic formal types of pointing gesture are observed: B-points (large movement, full arm, eye gaze often aligned) and S-points (small movement, hand only, casual articulation). Taking the approach that speech and gesture are structurally integrated in composite utterances, we observe that these types of pointing gesture have distinct pragmatic functions at the utterance level. One type of gesture (usually big in form) carries primary, informationally foregrounded information (for saying where or which one). Infants perform this type of gesture long before they can talk. The second type of gesture (usually small in form) carries secondary, informationally backgrounded information which responds to a possible but uncertain lack of referential common ground. We propose that the packaging of the extra locational information into a casual gesture is a way of adding extra information to an utterance without it being on-record that the added information was necessary. This is motivated by the conflict between two general imperatives of communication in social interaction: a social-affiliational imperative not to provide more information than necessary (Don't over-tell), and an informational imperative not to provide less information than necessary (Don't under-tell). (C) 2007 Elsevier B.V. All rights reserved.