Slide Slam L1
The influence of language similarity on multilinguals’ inhibitory control skills
Sarah von Grebmer zu Wolfsthurn1,2, Leticia Pablos1,2, Anna Ivanova Gupta3, Niels O. Schiller1,2; 1Leiden University, 2Leiden Institute for Brain and Cognition, 3University of Konstanz
Managing and communicating in different languages is a routine aspect of successful communication for multilinguals. A particular notable aspect of multilingual language processing is the co-activation and competition of multiple lexical candidates from different languages within a multilingual system. Here, language similarity was reported as a modulating factor, whereby increased co-activation was linked to linguistically similar languages. Simultaneously, increased language similarity was associated with greater difficulties while mitigating co-activation effects. In order to mitigate language co-activation, multilinguals employ an inhibitory control mechanism. Yet, few studies have directly investigated the direct effect of language similarity on inhibitory control skills. Here, we focused on the question whether and how language similarity affects inhibitory skills in the context of a Stroop paradigm. We tested two groups of late language learners of Spanish with different language combinations with varying degrees of language similarity: our first group were 32 Italian native speakers, and our second group were 25 Dutch native speakers with a B1/B2 level of Spanish. The Italian-Spanish group represented the linguistically “similar” group, whereas the Dutch-Spanish speakers represented the linguistically “dissimilar” group. We employed a Stroop paradigm where we exploited the conflict between the semantics of a Spanish target word (e.g., [izquerda] “left”, [derecha] “right”) and the location of the target word of the screen (left vs. right). We first predicted the classical Stroop effect for both groups, with faster response times for congruent trials (i.e., target word and location matched) compared to incongruent trials (i.e., target word and location did not match). Next, we predicted a main effect of language similarity on response times: we expected Italian-Spanish speakers to be overall faster compared to the Dutch-Spanish speakers. Finally, we predicted an interaction effect between language similarity and condition in the form of a larger Stroop effect for Italian-Spanish speakers compared to Dutch-Spanish speakers. This would yield increased parallel activation and mitigation difficulties for linguistically similar languages such as Italian and Spanish. Our results suggested the following: first, we found a classical Stroop effect, i.e. both groups were significantly faster for congruent than for incongruent trials. In contrast to our hypothesis, Dutch-Spanish speakers were overall faster compared to Italian-Spanish speakers. Further, we did not find evidence for an interaction effect of language similarity and condition: the Stroop effect was comparable across groups. The results therefore suggest a processing advantage for the linguistically dissimilar compared to the linguistically similar language pair in terms of successfully resolving the inherent stimulus conflict of the Stroop task. However, language similarity did not affect Stroop effect size, suggesting a similar degree of inhibitory skills across both groups and therefore a limited role of language similarity in modulating inhibitory control. Our study adds novel evidence to the role of language similarity in a critical component of successful multilingual language processing. In turn, this has important implications for the theoretical understanding of inhibitory control in multilinguals.