Typicality ratings of male and female voices

Abstract

Researchers have suggested that human voices are represented in memory in terms of prototypes [e.g., Kreiman and Papcun (1991); Papcun et al. (1989)]. Others have suggested that speech utterances are stored in memory via detailed exemplar-based representations [e.g., Lachs et al. (2000)]. The goal of the present study was to provide the first step toward assessing the viability of a prototype view of voice. Ten hVd utterances were recorded from each of 20 male and 20 female speakers. The utterances were blocked by speaker gender and presented to male and female listeners who rated each stimulus on a 1–7 typicality scale from ‘‘least typical voice’’ to ‘‘most typical voice.’’ There were significant effects of the type of vowel and speaker voice on the ratings, as well as interactions of vowel type with gender of subject and speaker voice. The results are discussed in terms of the strength of evidence for a graded category structure of voice categories that would be consistent with a prototype perspective of long-term memory representations of voice.