On linear segmentation and combinatorics in co-speech gesture: A symmetry-dominance construction in Lao fish trap descriptions

Abstract

Within the unified speech-with-gesture packages that we call utterances (Kendon 1972, 1980), speech and gesture play dierent roles, as deter- mined in part by intrinsically dierent semiotic properties. Co-speech representational gestures typically dier from the accompanying spoken linguistic code in displaying holistic and synthetic structuring of mean- ing, and in lacking hierarchical combinatoric properties or the linear- segmented distribution of information characteristic of spoken language (McNeill 1992: 19-23). The aim of this article is to demonstrate that co- speech gestures nevertheless can and do show combinatoric principles and linear-segmented organization. The claims I want to make are that (a) a given gesture may be structurally related, both in form and function, to a gesture in a neighboring utterance; (b) information may be segmented and supplied linearly with gestures (and not simply as a direct reflection of such linear-segmented structuring in the spoken code); and (c) multiple gestures can occur simultaneously, where the gestures show significantly dierent relationships (both semantic and pragmatic) to the utterance under way. With video-recorded data from speakers of Lao (a South- western Tai language of Laos), I document a two-phase routine that I refer to as the symmetry-dominance construction. Phase 1 is a two-handed symmetrical gesture; in the subsequent phase 2, one hand holds in posi- tion (representing given/topical/backgrounded information from phase 1), while one hand executes a new gesture (representing new/focal/ foregrounded information). This construction shows features of linear segmentation and combinatoric structure that arise from both the aor- dances and the constraints of the manual/visuospatial modality. The symbolic system of spoken language is profoundly linearized, due to severe constraints on the number of dimensions along which speech sounds may simultaneously represent information. (This relates to limi- tations in the capacity of speech sounds to bear iconic and indexical meaning; see below.) Speech phonology facilitates the construction of contrastive lexemes, and patterns of morphosyntax allow linear strings