The Effect of Common Ground on How Speakers Use Gesture and Speech to Represent Size Information

Abstract

Past research has investigated the impact of mutual knowledge on communication by focusing mainly on verbal communication. This study uses a wider focus, which includes speech and gesture. Speakers completed a referential communication task with recipients who did or did not share with them knowledge about the size of certain entities. The results showed that when such common ground exists between interlocutors, speakers' use of gesture and speech is affected. The main finding was that when speakers talked to recipients for whom the size information was new information, they represented this information predominantly in gesture only or in gesture and speech. However, when speakers talked to recipients with whom they shared knowledge about the entities' size, speakers encoded this information mainly verbally but not gesturally. The results are interpreted with respect to past research into common ground and language use, the pragmatics of gesture, and theories of gesture production.

Reduction in gesture during the production of repeated references

“…A qualitative analysis of a number of gestures confirmed that gesture parts depicting new information were larger and more precise (Gerwing & Bavelas, 2004, p. 182). Holler and Stevens (2007) obtained similar results in a referential communication task. They asked participants to locate targets in Where's Wally?…”

“…In addition, we know from various studies that repeated references can be reduced acoustically as well, in such a way that, for example, the second realisation of the word ''building'' in our example may be less intelligible (when heard in isolation) than the initial one (e.g., Aylett & Turk, 2004;Bard et al, 2000;Fowler, 1988). Finally, and most importantly for the current study, a number of studies have shown that repeated references are also accompanied by fewer gestures (e.g., de Ruiter, Bangerter, & Dings, 2012;Holler & Stevens, 2007;Holler, Tutton, & Wilkin, 2011;Levy & McNeill, 1992).…”

“…Alternatively, it might be that repeated gestures are similar to initial ones in general form, but differ in more gradient ways, much like repeated articulations of the same word (''building'') tend to be less clearly articulated. This question has received little attention in the literature, although some studies have looked at some qualitative aspects and generally find evidence for reduction in form (e.g., Galati & Brennan, 2014;Gerwing & Bavelas, 2004;Holler & Stevens, 2007). However, these studies tend to vary with respect to the measures that are used, resulting in an incomplete understanding of how repetition influences how speakers realize their gestures qualitatively.…”

Recipient design in tacit communication

“…This is one instance of audience design (Clark & Carlson, 1982), in which utterances are constructed according to what others in the communicative exchange know and believe (Clark, 1996). Audience design has been reported in speech and actions in both naturalistic and experimental settings (Holler & Stevens, 2007;Jacobs & Garnham, 2007;O'Neill, Bard, Linnell, & Fluck, 2005;Ozyurek, 2002), but there is still considerable debate over when, how, and why it is used. An outstanding issue is whether audience design is structurally linked to initial attempts to communicate, or whether it is used parsimoniously to repair erroneous attempts.…”

On the generation of shared symbols

“…It has been repeatedly shown that our communicative actions are selected according to what we believe addressees of a communicative interaction know and believe [audience design, (Clark, 1996;Holler & Stevens, 2007;Jacobs & Garnham, 2007;O'Neill, Bard, Linnell, & Fluck, 2005)]. For instance, adults tend to modify their speech, gestures, and accompanying body motions when addressing a child (Brand, Baldwin, & Ashburn, 2002;Brodsky, Waterfall, & Edelman, 2007;Grieser & Kuhl, 1988;O'Neill, et al, 2005;Warrenleubecker & Bohannon, 1984).…”