Do bilinguals avoid ambiguity or minimise processing effort? Insights from an online eye-tracking study in spoken Mandarin

Previous research shows that bilinguals tend to choose more explicit referential forms (e.g., overt pronouns over null pronouns) compared to monolingual speakers, but the mechanisms driving this tendency remain debated. By conducting two experiments examining lexical ambiguity in spoken Mandarin, we tested two hypotheses: Hypothesis 1: bilinguals would rather be redundant than ambiguous in general; Hypothesis 2: bilinguals avoid ambiguity only when doing so helps reduce cognitive load. In Experiment 1, L1 Mandarin L2 English speakers in the UK and more-monolingual speakers in China completed a picture naming task, where we manipulated whether the context in which a picture is named makes the preferred label ambiguous (e.g., do speakers avoid saying “fen3 si1” when describing a picture of glass noodles when it appears alongside a picture of fans which shares the same label?). Experiment 2 extended this task by incorporating online eye-tracking using WebGazer. Our results showed that, contrary to Hypothesis 1, bilinguals were more likely than more-monolinguals to use ambiguous expressions. Eye-tracking analysis revealed that bilinguals tended to direct early attention toward image pairs with more accessible labels, indicating a preference for linguistic choices that are cognitively less demanding; by contrast, more-monolinguals showed proactive monitoring of ambiguity depending on their responses. These findings support Hypothesis 2 and shed light on cognitive constraints in bilingual lexical access and early signs of lexical attrition.