Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Bilingual vs. diglossic code-switching in multilingual listeners: MEG evidence from English and Arabic
Poster Session B, Friday, September 12, 4:30 - 6:00 pm, Field House
This poster is part of the Sandbox Series.
Samantha Wray1, Zainab Hermes2, Benjamin Lang3, Hadi Zaatiti4; 1Dartmouth College, 2University of Chicago, 3University of California San Diego, 4New York University Abu Dhabi
Recent work in bilingualism and executive control (Blanco-Elorrieta and Caramazza, 2020) suggests that a single language system is in use for bilinguals, governed by mechanisms such as word frequency and communicative context between languages. Within these mechanisms is an assumption of a shared language space between each of a speaker’s languages, in which all words compete with each other depending on linguistic context and communicative need. In order to evaluate the interaction between languages in a shared space, previous studies have focused on code-switching between languages such as Emirati Arabic and English (Blanco-Elorrieta and Pylkkänen, 2017) or Korean and English (Phillips and Pylkkänen, 2021). However, Arabic exists on a continuum between standard and dialectal varieties leading to the phenomenon known as diglossia (Ferguson, 1951), in which two registers of the same language exist side by side, each associated with specific functions, and exhibiting substantial structural and lexical differences. Arabic maintains a highly codified formal register Modern Standard Arabic (MSA) and a more familiar spoken register Colloquial Arabic (CA). Speakers of Arabic are exposed to both varieties daily in different contexts and to achieve different functions. This provides an opportunity to further characterize the mechanism of bilingual language control as bilingual speakers of Arabic and English not only need to manage switching between languages (bilingual code-switching), but also between varieties of one language (diglossic code-switching) which share significant representational space. Using magnetoencephalography (MEG), we examine the neural time-course of diglossic switching by comparing sentence-internal switches between MSA, Egyptian Arabic–a variety of CA, and American English. This pilot study describes work in progress (n=4 at time of submission). Participants listened to 164 pairs of sentences which contained a single-word code-switch and 164 sentences with no code-switch. Switch materials included utterances in which the matrix language was either MSA or Egyptian Arabic, while frequency-matched code switched words were either MSA, Egyptian Arabic, or English. Data were analyzed using a temporal response function (TRF) approach (Brodbeck et al., 2023) that set as impulse predictors (a) the presence of a switch of any kind (diglossic or bilingual) and (b) that distinguished between a diglossic and a bilingual switch. A time window of up to one second after the switch was analyzed for its relationship to the impulse predictors, and predictive power of each model was assessed. Preliminary results show that model (b), which modeled diglossic code-switching, did not outperform model (a), which modeled all types of code-switching, bilingual and diglossic. This is consistent with contemporary models of bilingualism that posit a shared language space with activity at the switch modulated by mechanisms such as frequency and communicative context. Further analyses will explicitly examine the contribution of these mechanisms at the moment of switching. Our results have important implications for second language pedagogy, suggesting the current practice of integrating MSA and CA in the same classroom is comparable cognitively to integrating two distinct languages (Arabic and English, for example).
Topic Areas: Multilingualism, Speech Perception