Cross-Linguistic Influence on L2 before and after extreme reduction in input: the case of Japanese returnee children


Thematic Section: Modulators of cross-language influences in learning and processing

cross-language influence, transfer, immersion, morpho-syntax, lexicon

Maki Kubota, Department of Language and Culture, UiT The Arctic University of Norway
Caroline Heycock, School of Philosophy, Psychology, and Language Sciences, University of Edinburgh
Antonella Sorace, School of Philosophy, Psychology, and Language Sciences, University of Edinburgh
Jason Rothman, Department of Language and Culture, UiT The Arctic University of Norway and University of Nebrija

This study investigates the choice of genitive forms in English monolingual children and Japanese-English bilingual returnees; namely, children who returned from L2 (English) dominant environment to their L1 (Japanese) environment. There were two objectives in our study. First, we compared relative preferences for genitive forms (s-genitive: the table’s leg vs. of-genitive: the leg of the table) and verb/argument word orders in English between Japanese-English bilinguals and English monolinguals. The results showed that bilinguals differed from monolinguals only in the genitive conditions, specifically in those that required processing of semantic factors that are in conflict. These findings suggest that general processing difficulties in resolving such conflicts provide a better explanation for the observed behaviour than does CLI from L1 per se to L2. The second objective was to investigate how severe change in language input over time from the point of re-immersion in the L1 community affects returnee bilinguals’ L2 grammars over their first year after return to Japan. We examined if there were any changes in the evaluation of genitive forms and of the verb/argument orders in both their English and Japanese, and if so, whether change(s) could be explained by increased CLI effects. Results showed that there was no change in the preference for verb/argument orders; there was a change in the preference for genitive forms over time, but it was restricted to a single condition—namely, a condition that shares similar linear structure to the Japanese no-genitive and involves conflicting semantic cues. Combining the results from across the two studies we show that the dual effect of processing complexity and influence from dominant to non-dominant language work in tandem to explain monolingual to bilingual differences as well as longitudinal changes within bilinguals over time.