Cross-modal interactions in language production: evidence from word learning

Svetlana Pinet,Clara D Martin

DOI: https://doi.org/10.3758/s13423-024-02552-x

2024-08-09

Abstract:Literate adults are able to produce the same word in different language modalities-for instance, through speaking and writing. Yet how speaking and writing interact is not well understood. The present study takes a new perspective on the question of the co-activation of phonological and orthographic representations in speaking and writing by examining the acquisition of novel words. We tested how novel words get integrated into modality-specific lexicons by biasing novel word acquisition toward speaking or writing and assessing cross-modal transfer at the first stages of learning. Participants learned novel words paired with pictures of novel objects and practiced them overtly through speaking or typing. At test, typed training led to higher recall accuracy than spoken training whether words were recalled through typing or speaking. Performance in typing (RT and durations) benefited more from typed than spoken training. Crucially, performance in speaking did not benefit specifically from spoken training and was similar after spoken or typed training. Results are compatible with an asymmetric integration in the phonological and orthographic lexicons according to the modality of training, with representations created in the orthographic lexicon directly transferring to the phonological lexicon, while the opposite doesn't seem to occur. Cross-modal transfer dynamics are discussed according to the level of lexical activation.