Abstract
This paper reports the results of a listening test using Japanese word speech samples of various types and degrees of emotional expressions. The speech samples consist of ‘neutral’ speech as well as speech with three types of emotions (‘anger’, ‘joy’, and ‘sadness’) of three degrees (‘light’, ‘medium’, and ‘strong’). Furthermore, we use three words of three accent type groups: flat, mid-high, and head-high. Based on the listening-test results, we calculate the rate of agreement of the listener's receptivity with the speaker's intention in the type and degree of the emotion to the total number of the listeners. We divide agreement rates into three groups depending on levels. We then analyze the relationships between agreement rates and accent types. The results suggest that there is a significant difference between agreement rates and accent types, and that the emotion of ‘strong joy’ can be perceived most accurately when a speaker utters a ‘mid-high’ accent word.
