Hebrew-L2 speakers process auditory templatic words through their L1 processing mechanism with awareness of L2

Bilingualism involves cross-linguistic influence (CLI) prompted by communicative function, which impacts the activation of the bilingual's L1/L2 language processing mechanisms. The current study examines the extent of CLI when semantic information is reduced. Semitic languages are known for their templatic words composed of intertwined sub-lexical root and template morphemes, entailing non-linear morphological processing. As the roots constitute the semantic core, comprehension was found to impact morphological processing among Hebrew-L2 readers. Herein, we assessed the processing mechanism activated among adult Hebrew-L2 bilinguals in an auditory rhyme judgment task that requires linear processing. The task was provided with Hebrew templatic word pairs comprising accentuated (meta)linguistic information irrespective of semantics: phonological co-occurrence restrictions (root), grammatical information of vocalic melodies (template), and contrastive stress. We hypothesized that CLI in Hebrew-L2 speakers would be reflected in low accuracy rates in rhyming pairs when linguistic information is accentuated, indicating distraction from the linear processing due to activation levels of the L2 processing mechanism caused by competing linguistic cues drawn on transferred linguistic information. We compared the performance of 58 adult Hebrew native speakers with 54 Hebrew-L2 speakers with Semitic and non-Semitic-L1. The findings demonstrate that Hebrew-L2 speakers performed the task using their L1 processing mechanism with varying activation levels of L2, showing awareness of the morphological processing due to the vocalic melody for non-Semitic-L1 and awareness of contrastive stress for Semitic-L1. The results confirm CLI also when semantics is reduced, elucidating how much CLI modulates the bilingual's language processing mechanism.