The role of phonological overlap and cognates in dual logographic bilinguals’ phonological processing

• Published in: Lingua Vol. 311; p. 103831
• Main Authors: Liu, Xiaoping, Zhang, Yan, Zeng, Bingjie, Liu, Ye, Wang, Xiaolu
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01-11-2024
• Subjects: Cognate facilitation effect; Interactive; Logographic; Modular; Phonological; Logographic; Modular; Cognate facilitation effect; Interactive; Phonological
• ISSN: 0024-3841
• DOI: 10.1016/j.lingua.2024.103831

•A phoneme monitoring task was used to test Chinese-Japanese bilingual lexical access for the first time.•Phonological overlapping effects are modulated by L2 proficiency.•For cognates, semantic overlap marginally interacts with L2 proficiency in phoneme recognition.•Indiscriminate application of L1 pronunciation in L2 learning should be overcome.•Greater importance should be attached to holistic perception of word sound rather than to a word’s component.
Most previous studies on Chinese-Japanese bilingual lexical access have shown that cognates can facilitate L2 Japanese word recognition, and phonological overlap cannot. However, the findings from this research challenge those findings. In this study, both high- and low- proficient JFL (Japanese as a Foreign Language) learners performed phoneme monitoring tasks in their second language. Orthographically similar words with high or low phonological and semantic similarity were fully crossed. Our results suggest that phonological overlap across Chinese and Japanese interferes with phoneme processing speed and accuracy, especially those of the low-proficient JFL learners. Yet as their L2 proficiency improves, this inhibition effect can become smaller. In addition, cognate semantic overlap marginally interacts with L2 proficiency to affect phoneme recognition, though cognates do not directly play a role in phoneme processing. The theoretical implications of our research suggest that phonology functions as a non-selective element in the phonological access phase for bilinguals literate in two logographic languages. Moreover, cross-logographic processing does not conform to the cascading model; instead, it can be more accurately described by the modular or interactive models of bilingual processing.