Abstract:

Hearing someone's voice, we immediately recognize the owner of the voice if we know that person. Even when we do not know him, we make up a certain mental image of the person. How accurate is the image triggered by the voice alone? Is the relationship between voice and the image triggered by the voice language-specific or universal? These are the issues this study addresses. Korean speech samples from 8 males and 8 females were recorded. Two pictures were taken for each speaker: one showing the whole body with the reference background so that physical characteristics may be easily compared, and the other showing only the face close-ups. 151 non-Asian subjects without any knowledge of Korean language were asked to match the voices with the corresponding pictures. The results were compared with the previous results obtained from 361 native speakers of Korean. While Korean subjects showed highly accurate matching patterns, non-Asian subjects showed much lower accuracy. The tendency to match a certain voice with a certain picture, regardless of being correct or not, was very high with Korean subjects, but much lower with non-Asian subjects. These results seem to suggest that the perception of voice is language-, or culture-specific.