Communicating common ground: How mutually shared knowledge influences speech and gesture in a narrative task

Much research has been carried out into the effects of mutually shared knowledge (or common ground) on verbal language use. This present study investigates how common ground affects human communication when regarding language as consisting of both speech and gesture. A semantic feature approach was used to capture the range of information represented in speech and gesture. Overall, utterances were found to contain less semantic information when interlocutors had mutually shared knowledge, even when the information represented in both modalities, speech and gesture, was considered. However, when considering the gestures on their own, it was found that they represented only marginally less information. The findings also show that speakers gesture at a higher rate when common ground exists. It appears therefore that gestures play an important communicational function, even when speakers convey information which is already known to their addressee.