In 5 experiments, male and female undergraduates viewed gestures and tried to select the words that originally accompanied them; read interpretations of gestures* meanings and tried to select the words that originally had accompanied them; tried to recognize gestures they previously had seen, presented either with or without the accompanying speech; and assigned gestures and the accompanying speech to semantic categories. On all 4 tasks, performance was better than chance but markedly inferior to performance when words were used as stimuli. Judgments of a gesture's semantic category were determined principally by the accompanying speech rather than gestural form. It is concluded that although gestures can convey some information, they are not richly informative, and the information they convey is largely redundant with speech

No abstract