Processing benefits during code-switched reading in Cantonese-English bilinguals: Evidence from eye movements

Code-switching is the practice of mixing languages within or across sentences. Cued language switching tasks have typically reported that there are both production and processing costs associated with code-switching. Recent work, however, suggests that such costs are absent in words that are commonly switched and among frequent code-switchers. Most of the code-switching literature has focused on switches between alphabetic writing systems, meaning that we know very little about code-switching between alphabetic and logographic systems. This pre-registered experiment investigated switch costs in 48 Cantonese-English bilingual adults whose eye movements were recorded as they read 80 sentences. Using a 2 × 2 within-participants design, both the context language (Cantonese vs English) and the target word language (Cantonese vs English) were varied between languages for each sentence, yielding both code-switched and unilingual conditions. Surprisingly, we observed a switch-benefit on target word processing. That is, gaze durations, go-past times, and total reading times were shorter on target words (both Cantonese and English) in code-switched sentences compared to unilingual sentences. We additionally observed longer reading times on target words when the context language was Cantonese. The language of the target word itself did not impact reading times on these words. This experiment enhances our understanding of code-switching by providing novel evidence for its benefits during sentence processing.