Where do most spontaneous representational gestures actually occur with respect to speech?

Abstract

Introduction : "Speech-accompanying spontaneous gestures have attracted the attention of psycholinguists and cognitive psychologists in recent years. Theorists assert that spontaneous gestures, along with the speech they accompany, are a window into the nature of speakers' mental representations and processes (Goldin-Meadow et al. 1993; Kendon 1986; McNeill 1985, 1987, 1992). Researchers have shown that speech-accompanying gestures are inextricably intertwined temporally and semantically with speech (Kendon 1980; McNeill 1992). Various models of production of speech and representational gestures (i.e., iconics, metaphorics and abstract deictics) have been presented (Beattie & Aboudan 1994, Butterworth & Hadar 1989; Kita 1993, Krauss et al. 1996; McNeill 1989, 1992; Nobe 1993, 1996). There have recently been debates on the credibility of each model (see Hadar & Butterworth 1997; Hadar & Yadlin-Gedassy 1994; Hadar et al. 1998; and other chapters in this book). Unfortunately, the current chapter does not have any space (and thus it is not the writer's intention) to evaluate all models mentioned above. It is, however, a fact that many agree that the temporal relationship of representational gestures and speech is very significant in constructing gesture- and speech-production models. Moreover, the 'tendency' of gestures to occur with respect to their accompanied speech does matter. This chapter reports on how the timing between gestures and speech is measured, and how this gesture tendency is calculated and presented as evidence for the construction of speech and gesture models. Comprehensive models of speech and gesture should account for as many gestures as possible. If the number of gestures which are accounted for by a model is rather small, we should not regard it as a general theory (Nobe 1996). In particular, this chapter reports that the timing and the tendency to produce gestures may not be well represented in some contributions to the literature, owing to their adopting a measurement involving 'ratios'. I do not intend to claim that this unit is useless; I agree that it is useful in some cases. However, the ratio measure may sometimes ignore a substantial number of gestures. Ratios were used to support an inference by Beattie & Aboudan (1994) that the potentially different social contexts of experiments might cause us to observe different databases. Although differing social contexts may influence the production of gestures, using this type of measurement alone may have the effect of not explaining many observed gestures. I would like to point out that if researchers try to construct comprehensive theories which explain as many gestures as possible, they should be aware of the fact that such measurements and presentations involving ratios may fail to achieve this important goal. The following sections will discuss these issues in details."