Exploring bimodal lexical access in unimodal bilinguals through videos with captions

Abstract Studies on bimodal bilingualism showed that bimodal lexical access is not costly compared to unimodal lexical access, but that it can be even advantageous. We asked whether the same can be observed in unimodal bilinguals as long as some bimodal conditions are provided. We exploited the ecological bimodal setting of subtitled videos and designed four versions of a semantic categorization task, with unimodal (spoken or written language) and bimodal (speech and captions) stimuli. Regardless of the status of the language (L1/L2) and regardless of the bimodal stimuli being also bilingual, answers to bimodal stimuli were systematically faster than speech-only stimuli and slower than written-only stimuli. These results indicate that: i) bimodal stimuli were processed differently from unimodal stimuli, ii) both modalities were taken into account simultaneously, iii) the integration between modalities occurred automatically, even if focusing on one modality only would have been advantageous.