Perception of the Prosodic Formative of Multimodal Affective States

<p>The features of the expression and perception of vocal expressions as one of the forming multimodal affective states of a person are studied. The experiment, designed on the basis of the Russian-language version of the Geneva Emotion Recognition Test (GERT), involved two groups of women aged 18-45, 72 women each. One group was shown audio-video clips of 14 affective states, lasting 3-5 seconds, played by 10 professional actors according to the Stanislavsky system. The other group was presented with audio clips extracted from the same clips through headphones. It was demanded to recognize the state of the actors using an interactive image of the Geneva Emotion Wheel. It is shown that vocal expressions differ from the corresponding multimodal states by a low average accuracy of expression and identification, a different configuration and a high spread of estimates, a closer dependence on the degree of arousal, individual ways of expression and a fuzzy extended structure of categorical fields. The range of possibilities that vocal expressions have, ready to be included in cross-modal integrations, is indicated.</p>