Paralanguage and human-computer interaction. Part 2: comprehension of synthesized vocal segregates

Abstract

Abstract Abstract. The comprehension of synthesized vocal segregates was studied using a paired-associate learning task. Twenty-one naive subjects attempted to learn associations between six vocal segregates and assigned meanings, where the segregates were synthesized using the simple sound facilities common to all PC-compatible machines. In the natural condition sounds were paired with the correct meanings, whereas in the random condition sounds were randomly assigned to incorrect meanings. Subjects in the natural condition showed a significant initial advantage over those in the random condition and significantly better overall performance across learning trials. The results show that uninformed subjects are predisposed to make the correct sound-meaning associations for segregates synthesized with inexpensive hardware, and that associations which are not initially identified are quickly learned. Potential applications are discussed.