Visual contributions to talker attunement

Abstract

Listeners can attune to talker-specific speech patterns, as demonstrated by findings of greater accuracy in spoken-words-in-noise identification for words spoken by familiar voices [e.g., Nygaard and Pisoni, Percept. Psychophys. 60, 355–376 (1998)]. In previous demonstrations of this effect, talker familiarity was produced by auditory speech exposure. The purpose of the present study was to examine the contribution of visible speech information to talker attunement. First, the auditory attunement effect was replicated: Participants were given extensive exposure to auditory words spoken by different talkers, and were subsequently tested on novel words in noise spoken by one of the familiar talkers and by a talker not previously heard. Identification of the words was more accurate for the familiar talker. This procedure was then extended to include video presentations of the talker’s articulating face along with each word during the exposure phase. To control for familiarity with the faces, a separate condition included presentation of static images of the talkers’ faces. Visual articulations were not presented during the identification-in-noise task. Thus, the experiment tests whether visual exposure to talker’s articulatory patterns contributes to talker attunement, and, specifically, whether any such visual contributions to talker attunement generalize to auditory speech perception. [Work supported by NICHD.]