How do gesture and speech production synchronise?

Abstract

The temporal relationship between the onset of gesture and speech was analysed in two experiments using dual-task paradigms. In single-task conditions, the participants responded to pictures either with predefined referential gestures or with vocal utterances. In dual-task conditions, both responses were required. The experiments manipulated the difficulty of the vocal task during response selection (Experiment 1: set size of two or eight items) or response preparation (Experiment 2: phrases of one or five words). As expected, these manipulations influenced the latencies of the vocal responses. Interference effects (i.e. differences between single- and dual-task conditions) were observed in both response modalities. Gesture initiation preceded speech onset but it was delayed when the difficulty of planning concurrent vocal responses increased. The results are discussed in relation to various models of dual-task performance. They are found to be most compatible with a central executive model, which assumes a strategic response deferment in dual-task conditions.