The Role of Linguistic Knowledge in the Encoding of Words and Voices in Memory
Stephen Winters, Karen Lichtman, and Silke Weber
129-138 (complete paper
or proceedings contents
When listening to their first language, listeners store linguistic and voice information together in an integrated fashion in memory. But is this also true when listening to an unknown language, or an L2? In this study, three groups of listeners—native Germans, monolingual English speakers, and English L1/German L2 learners—performed a continuous word recognition memory task. These listeners heard several voices producing a sequence of German words, and were asked to identify whether each word was a new or a repeated item in the sequence. All groups more accurately recognized repeated items when they were presented in the same voice as on the initial presentation, rather than in a different voice. This indicates that listeners store word and voice information together in memory, even when listening to a language they do not know. However, recognition accuracy interacted with the phonetic characteristics of each word: all native English listeners performed better on words containing only phonemes found in English, while German L2 learners outperformed monolinguals on words containing velar fricatives. Native German listeners also appeared to encode words containing German-specific vowels in a more speaker-independent fashion than nonnative listeners. Talker-dependent processing of words thus seems to be a general perceptual mechanism—even when listening to an L2—while the ability to construct talker-independent representations of words containing non-native phonemes may only emerge with significant amounts of L2 experience.
Selected Proceedings of the 2011 Second Language Research Forum: Converging Theory and Practice
edited by Erik Voss, Shu-Ju Diana Tai, and Zhi Li
Table of contents