Auditory stream discontinuities interfere with speech processing efficiency

Abstract

Listeners process speech from mixed talkers less efficiently than from one continuous talker. An auditory streaming framework may help explain this effect: talker discontinuity may disrupt listeners' attentional focus, decreasing speech processing efficiency. In two experiments, we examined how auditory interruption of a speech stream interfered with speech processing. In experiment 1, we sporadically presented non-speech distractors (e.g., text message alerts) while listeners identified words spoken by either one consistent talker or multiple talkers. Word identification was slower immediately following non-speech distractors, and this disruption was greater when listening to speech from a single talker than mixed talkers. Experiment 2 tested the impact of interruptions caused by non-speech distractors or sudden switches to a different talker while listeners identified words spoken by a single, consistent talker. Listeners were slower at responding to non-speech distractors than a new talker's speech, but across interruption types, the magnitude of recovery relative to the magnitude of interruption was similar. These results suggest that both speech and non-speech discontinuities interfere with speech processing by disrupting listeners' attentional focus toa talker. Moreover, interruptions appear less disruptive when listeners pre-deployed cognitive resources to accommodate expected discontinuities, such as when listening to mixed-talker speech.