Interaction between talker's voice, linguistic information, and attention during speech perception

Natalya Kaganovich, Purdue University

Abstract

This dissertation consists of two studies. The first study combined behavioral and electrophysiological measurements to investigate interactions during speech perception between native phonemes and talker's voice. In a Garner selective attention task, participants either classified each sound as one of two native vowels ([ϵ] and [æ]), ignoring the talker, or as one of two male talkers, ignoring the vowel. The dimension to be ignored was held constant in baseline tasks and changed randomly across trials in filtering tasks. Irrelevant variation in talker produced as much filtering interference (i.e., poorer performance in filtering relative to baseline) in classifying vowels as vice versa, suggesting that the two dimensions strongly interact. Event-related potentials (ERP) were recorded to identify the processing origin of the interference: an early disruption in extracting dimension-specific information or a later disruption in selecting appropriate responses. Processing in the filtering task was characterized by a sustained negativity starting 100 ms after stimulus onset and peaking 200 ms later. The early onset of this negativity suggests that interference originates in the cognitive effort required by listeners to extract dimension-specific information, a process that precedes response selection. In agreement with these findings, our results revealed numerous dimension-specific effects, most prominently in the filtering tasks. The second study used a dual-channel paradigm to test whether the asymmetrical dependency between talker and linguistic dimensions previously reported in literature is due in part to the power of talkers to capture attention. Participants were asked to classify speech sounds (two English vowels produced by two male talkers) in a target channel (one ear) and to ignore speech sounds in a distractor channel (opposite ear), either according to which talker spoke, ignoring the vowels spoken, or according to which vowel was spoken, ignoring the talkers. Distractor sounds changed along the dimension that listeners had to ignore in the target channel. We collected behavioral data from the target sounds and event-related potentials in response to distractor sounds. We found that irrelevant talker change produced relatively greater sustained negativity over the P1 and N1 components than irrelevant vowel change, an indication that greater attention was allocated to the distractor channel when it contained a talker change. This negativity persisted over the P2-P3 components in the filtering task over frontal, central, and parietal sites. These findings support the hypothesis that a change in talker captures attention even when it is irrelevant for the task at hand. We proposed that such attentional capture is due either to a high social significance of the human voice or to its ability to uniquely define an auditory source.

Degree

Ph.D.

Advisors

Francis, Purdue University.

Subject Area

Speech therapy|Cognitive therapy

COinS