You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster D67, Thursday, November 9, 6:15 – 7:30 pm, Harborview and Loch Raven Ballrooms

Accented speech attenuates code-switching costs in bilingual listeners: An auditory electrophysiological study

Carla Fernandez1, Janet van Hell1;1Pennsylvania State University

Code-switching, defined as the interchangeable use of two languages within a sentence, is considered one of the hallmarks of bilingualism. Recently there has been growing interest in the neurocognitive mechanisms that drive the comprehension of code-switched sentences. Studies that used the Event-Related Potentials (ERP) technique have typically found that comprehending code-switched sentences entails a cost in terms of lexical-semantic-integration (as represented by the N400 effect) or sentence level restructuring and reanalysis (as represented by the LPC). These previous neurocognitive studies on intra-sentential code-switching presented sentences visual, even though code-switching occurs more frequently in spoken discourse than in writing. Moreover, in natural code-switched discourse, bilingual interlocutors often have an accent in one of their languages, which may affect the comprehension of code-switched speech. Taking this into account, in an ERP study, we analyzed code-switching in the auditory modality and examined the effects of accented speech on listeners' comprehension of code-switched sentences. In this study, 31 highly-proficient Spanish-English bilinguals who identified as Hispanic listened to code-switched sentences (that switched from English to Spanish, or vice-versa) or non-switched sentences produced by a Spanish-accented speaker of English or a non-accented speaker in both Spanish and English (10 Spanish-English bilinguals had rated, and agreed on, the speakers' accentedness in the two languages). We hypothesized that Spanish-accented English would alleviate processing costs in the comprehension of sentences that switch from English to Spanish, as the Spanish-accented English in the first part of the sentence would facilitate the comprehension of the upcoming switch into Spanish. On the other hand, if the same sentence begins with non-accented Spanish and then switches to Spanish-accented English, the non-accented speech at the beginning of the sentence would not (strongly) co-activate two languages, and a subsequent switch to Spanish would induce switch-related processing processing cost. This pattern was indeed observed: in sentences that switched from English to Spanish, the switch-related N400 effect was smaller when the first part of the sentence was produced in Spanish-accented English relative to non-accented English. Furthermore, for sentences that switched from Spanish to English, no difference in switching costs were expected for the Spanish-accented and non-accented speaker, as both sentences start in Spanish. Indeed, a switch-related LPC effect was observed in both speaker conditions, and the magnitude of the LPC did not differ in the two speaker conditions. These results demonstrate that accented speech can attenuate processing costs associated with the comprehension of code-switched sentences. More specifically, accented speech can facilitate the comprehension of upcoming code-switches into the language compatible with the accent, and serve as a cue to modulate code-switching costs when produced by speakers that have a similar language background as the listeners's background.

Topic Area: Perception: Auditory

Back to Poster Schedule