Abstract The L1-transfer pattern may be different between bilinguals and monolinguals as the former has multiple L1 candidates to transfer. This study compared how Mandarin monolingual learners (MDN), Shanghainese-Mandarin bilingual learners (SHM), and Japanese natives produce Japanese stops in word-reading and paragraph-reading tasks. The L2 Japanese learners varied in the years of learning (1–3 years). Shanghainese differs from Mandarin in that the word-medial voiced stops are prevoiced, which may allow facilitative transfer to Japanese voiced stops. As a result, SHM in general showed more target-like pronunciation of voiced stops than MDN. Regarding the L2 experience, third-year SHM produced more target-like word-medial voiced stops, whereas first-year SHM produced less target-like word-initial voiceless and word-medial voiced stops. These results suggest that the overlap between the target L2 and one of the learners’ L1s may lead to finer phonetic realization, but the facilitative transfer is subject to bilingual learners’ L2 experience.
Read full abstract