Douglas A. Vakoch
Dept. of Psychol., State Univ. of New York at Stony Brook, Stony Brook, NY 11794-2500
The present study extends traditional approaches to determining the discriminating features of vocal expressions of emotions by visually representing some of the auditory characteristics of speech. This allows the use of a more extensive array of vocal stimuli that if the stimuli were rated in listening tasks. Such an approach is particularly appropriate because a number of the important determinants of different emotions are conveyed in the fundamental frequency and its patterning. Thus an examination of the intonation contours provides substantial information to differentiate emotions. Subjects sorted visual representations of the pitch contours of a range of emotions into groups that looked similar. The dissimilarity of any two representations was determined by counting how often subjects placed the representations in the same group. These measures of dissimilarity were used as input for multivariate analyses. Different emotions were located along continua using multidimensional scaling, and cluster analyses identified discrete groupings of emotions.