Distinct language control mechanisms in speech production and comprehension: evidence from N-2 repetition, switching, and mixing costs

The debate remains unresolved regarding whether identical language control mechanisms underlie production and comprehension in individuals who use two or more languages. Most previous research focused solely on switching costs aggregated at the group level, potentially overlooking a broader set of indices and individual variabilities. To address this, we integrated the n-2 repetition paradigm, language-switching paradigm, and single-language tasks to examine two modalities in Cantonese–Mandarin–English trilinguals: spoken word production (Experiment 1) and auditory word comprehension (Experiment 2). We evaluated a wider range of language control indices, including n-2 repetition, switching, and mixing costs. Group-level analyses revealed significant effects for all indices in production, whereas only mixing costs reached significance in comprehension. Individual differences analyses showed no significant correlations between production and comprehension across three types of costs, with Bayesian statistics providing supporting evidence for the absence of such relationships in most cases. Radar and permutation analyses further showed that language control in production was quantitatively greater than comprehension, and qualitatively distinct in pattern from comprehension. These findings show that language control in production is not only stronger, but also qualitatively distinct from that in comprehension, underscoring the importance of distinguishing between modalities in models of language control.