Is code-switching costly? Evidence from Belize and Benin

This OSF page belongs to a poster presented at 1st International Meeting on Multilingualism - The Experimental Study of the Psychosocial Brain: "Is Code-Switching Costly? Evidence from Belize and Benin". Code-switching/code-mixing (CSCM) is assumed to be cognitively costly [1]. However, growing evidence suggests that such costs, when observed, may not arise from CSCM per se, but from ecological factors such as community practices, individual language experience, and the degree to which switching patterns are locally entrenched [e.g., 2]. This project investigates this issue in two multilingual ecologies: Belize and Benin. In both settings, speakers navigate two or more typologically distinct languages. In Belize, CSCM may involve Spanish, English, Kriol, and Mayan languages; in Benin, it may involve French, Gungbe, Fongbe, and other local languages. Importantly, these language combinations include languages with grammatical gender (Spanish, French) and without grammatical gender (English, Gungbe/Fongbe), creating a potential conflict site that we exploited experimentally. Using a mobile eye-tracking lab, we examined processing effort during CSCM through pupillometry. In Belize, participants (n = 61) completed a visual-world task comparing pupil responses to Spanish-English code-switched utterances with unilingual Spanish speech. They were instructed to identify a target object on a screen, prompted by Encuentra elMASC/laFEM [target] (‘Find the [target]’). In Benin, participants (n = 64) heard Gungbe unilingual sentences introducing a referent, followed by a description in Gungbe or French (e.g., É nyí fífánɔ/douxMASC/douceFEM, ‘He/she/it was sweet’). On 30% of the trials, participants subsequently judged whether an AI-generated image matched the description. In both experiments, the critical manipulation targeted grammatical gender assignment strategies previously documented in these communities, varying in distributional frequency. We therefore predicted comparable pupil responses for unilingual and CSCM utterances overall, but greater pupil dilation for less locally entrenched switching patterns. Preliminary analyses of the Belize data reveal no general additional processing cost for CSCM relative to unilingual speech, nor do less frequent switching patterns appear to incur greater cost than more entrenched combinations. These findings support the view that CSCM is not inherently costly. Our presentation will report on a comparative analysis between the Belizean and Beninois data, further shedding light on how ecological factors may shape the cognitive processing of CSCM. [1] van Hell, J. G., Litcofsky, K. A., & Ting, C. Y. (2015). Intra-sentential code-switching: Cognitive and neural approaches. In J. W. Schwieter (Ed.), The Cambridge Handbook of Bilingual Processing (pp. 459–482). Cambridge University Press. doi.org [2] Beatty-Martínez, A. L., & Dussias, P. E. (2017). Bilingual experience shapes language processing: Evidence from codeswitching. Journal of Memory and Language, 95, 173–189. doi.org