Listener adaptation to lexical stress misplacement

Abstract

Speech including unfamiliar accents can result in decreased processing efficiency. However, listeners are able to overcome the difficulty of processing accented speech with adequate exposure, through the process of perceptual adaptation (Norris, McQueen & Cutler 2003; Bradlow & Bent 2008, among others). The current study addresses the role of word-level prosodic information in listener adaptation to accented speech. Specifically, it investigates adaptation to lexical stress misplacement in English, and examines how it compares with adaptation to segmental mismatches and accentedness at the whole utterance level. 91 native speakers of English were exposed to English words in isolation with canonical and non-canonical stress position, while performing a speeded cross-modal matching task. Results suggest that adaptation to lexical stress misplacement is largely comparable to adaptation to non-canonical productions at the segmental or utterance level in terms of speed and generalizability across lexical items. Results also indicate that adaptation to lexical stress misplacement is generalizable across talkers to some extent.