Onnis, L.
Nanyang Technological University
All languages exhibit statistical structure, but in language-specific ways, and thus bilinguals must learn and joggle different sets of statistical regularities all the time. We asked whether bilingual experience predicted higher scores in a dual artificial grammar learning (AGL) task. Bilingual adults (n=95) were assigned to either a Visual or an Auditory condition, in which the two artificial grammars were instantiated with either pseudo-shapes or pseudo-words. In addition, because the protracted experience of learning and using two languages in life may confer higher executive control, we also measured inhibitory abilities as a covariate using a Flanker task. Degree of bilingualism was assessed via a validated Bilingual Language Profile (BLP) and measured as a continuous predictor variable. We used mixed-effects models to predict AGL scores, with Subject, Test Item, and Grammar as random variables, and Modality, Bilingualism, and Inhibitory Control as fixed variables. We found that bilingualism interacted with inhibitory control to predict statistical learning scores (standardized Beta= -.152**). In particular, individuals with weaker inhibitory control benefited from being more balanced bilinguals. The results held across visual and auditory modalities, suggesting that the effect is domain-general. Our findings contribute to better characterising how experiential factors are related to core cognitive abilities.