How common ground affects the use of gesture space

Abstract

Recent work has started to ask to what degree common ground, i.e. shared background knowledge, may affect the articulation of co-speech gestures, specifically as regards viewpoint. We extend that work by asking how gesture articulation is affected, and how the spatial location of those gestures changes as a result of common ground. Using naturalistic data, we find that with quotes grounded in past interactions, co-speech gestures are fewer and smaller, whereas with quotes ungrounded in past interactions (UGPI), co-speech gestures are more numerous and larger. Moreover, when speakers repeatedly reference the same entity, the co-speech gestures used become smaller with each repeated reference, thereby suggesting a kind of ‘online’ grounding. We also find that the use of gestural space is more consistent for UGPI utterances. This suggests that common ground affects both the articulation of gestures, and the way that gesture space is used.


Back to Table of Contents