Listener flexibility to lexical alterations for foreign- and native-accented speech

Abstract

Listeners regularly encounter speech produced by speakers born and raised in linguistic communities different from their own. Given that individuals from different linguistic communities often speak in a distinct fashion, it is not uncommon for listeners to be exposed to accent-induced variability. This can lead to discrepancies in the realization of words, rendering such speakers difficult to understand. As a result, listeners tend to increase their flexibility in their interpretation of variations in the phonological form for known words. To examine whether or not a foreign accent also leads to an increased flexibility in label-object mapping, we tested participants in a preferential looking paradigm. Using a between-subject design, participants either listened to a foreign- or to a native-accented speaker. In some trials, the target word labeled an object that was depicted on the screen (match trials), while in other trials the target word referred to an object that was closely related (but not an exact match) to one of the objects depicted on the screen (mismatch trials). Listeners’ looking behavior after target word onset was examined across accents and trial types. The results of this study show that lexical alterations affect word recognition across foreign- and native-accented speech.Listeners regularly encounter speech produced by speakers born and raised in linguistic communities different from their own. Given that individuals from different linguistic communities often speak in a distinct fashion, it is not uncommon for listeners to be exposed to accent-induced variability. This can lead to discrepancies in the realization of words, rendering such speakers difficult to understand. As a result, listeners tend to increase their flexibility in their interpretation of variations in the phonological form for known words. To examine whether or not a foreign accent also leads to an increased flexibility in label-object mapping, we tested participants in a preferential looking paradigm. Using a between-subject design, participants either listened to a foreign- or to a native-accented speaker. In some trials, the target word labeled an object that was depicted on the screen (match trials), while in other trials the target word referred to an object that was closely related (but not an exac...