Gesture and Speech in Spontaneous and Rehearsed Narratives

In the first of two experiments, professional actors were videotaped as they answered questions about their personal experiences, feelings, and beliefs. Their narratives were transcribed, and these scripts were given to actors of the same sex, who were videotaped as they tried to portray the original respondent in a convincing manner. Spontaneous portrayals did not differ from rehearsed versions of the same narratives in the proportion of time the speaker spent gesturing or in their rates of filled and silent pauses. However, spontaneous narratives were accompanied by more lexical movement and contained a greater proportion of intraclausal silent pauses. In a second experiment, naive subjects were shown both spontaneous and rehearsed versions and tried to select the spontaneous one. In different judgment conditions subjects (1) saw the video without sound, (2) only heard the sound track, or (3) both heard the sound and saw the picture. In all three judgment conditions, identification of the spontaneous version was better than chance, but subjects were reliably more accurate from the combined sound and picture than from either the sound or the picture alone.