Abstract

The present study replicated Banse & Scherer(1996) to obtain acoustic profiles and estimate recognition accuracy in vocal emotion expression for Korean people. In experiment 1 professional actors' portrayals of 14 emotions(joy, elation, sadness, grief, anxiety, fear, irritation, anger, disgust, contempt, boredom, guilt, pride, interest) were subjected to digital acoustic analysis to obtain profiles of vocal parameters for different emotions. The Korean profiles in general replicated those obtained in Banse & Scherer with the exception of fear. In experiment 2 vocal emotional portrayals were presented to judges to investigate the accuracy of recognition. The overall recognition rate was around 50%, close to the one in Banse & Scherer, but correspondence in recognition rates for the individual emotions was not high between the two studies. Vocal expressions of grief, anger, boredom were easily recognized while those of joy, elation, anxiety, disgust were not. Emotions with the same valence tended to be confused with each other. Female vocal expressions and vocal expressions of the same sex were more easily recognized than male ones and those of the different sex respectively. However differences in recognition accuracy between male and female judges were not found.