A recent study delves into the intricate development of second language (L2) lexical semantics, specifically focusing on advanced Chinese learners of Japanese and the significant influence of their length of residence (LOR) in Japan. The research highlights persistent challenges in acquiring accurate word meanings, particularly when confronted with cross-linguistic differences in verb-noun collocations, even among highly proficient learners. The findings suggest that while extended immersion aids in integrating L2 semantic knowledge, deeply ingrained L1 representations can continue to impede ultimate attainment. The L1-L2 Semantic Divide: A Persistent Hurdle Differences in how languages categorize meanings between a learner’s first language (L1) and their second language (L2) can create significant obstacles to acquiring native-like word meanings. This phenomenon is particularly evident in verb-noun collocations, which are combinations of words that frequently occur together and often carry idiomatic or context-specific meanings. Even advanced learners, those who have spent considerable time in a target language environment, often struggle with these nuances. The study, conducted by researchers at Waseda University and Case Western Reserve University, involved 66 native Chinese speakers learning Japanese and 32 native Japanese speakers. Participants were tasked with an acceptability judgment task, evaluating the naturalness of Japanese collocations. These collocations were categorized into three types: L1-only collocations: Valid in Chinese but not in Japanese (e.g., eakon-o akeru, literally "open the air conditioning," which is acceptable in Chinese but unnatural in Japanese, where eakon-o tsukeru "turn on the air conditioning" is used). L2-only collocations: Valid only in Japanese and not in Chinese (e.g., eakon-o tsukeru "turn on the air conditioning"). L1-L2 collocations: Valid in both Chinese and Japanese (e.g., doa-o akeru "open the door"). Findings: Length of Residence Matters, But Not Universally The results revealed a clear impact of LOR on learners’ judgments. Native Japanese speakers (NSs) consistently rated L1-only collocations as unacceptable and L2-only collocations as acceptable, reflecting their native intuition. However, the learners’ performance varied significantly based on their time spent in Japan. Medium-LOR learners (3-7 years of residence) exhibited non-target-like patterns across all three collocation types. This suggests that even after several years of immersion, their L2 semantic understanding was still heavily influenced by their L1. They tended to over-accept collocations that were natural in Chinese but incorrect in Japanese, indicating a persistent reliance on L1-based semantic mappings. Long-LOR learners (over 10 years of residence) demonstrated more native-like judgments for L2-only collocations, aligning with the expectations of acquiring language-specific patterns through repeated exposure. However, they continued to differ from native speakers in their acceptance of L1-only collocations. This indicates that while extended residence facilitates the internalization of L2-specific usages (positive evidence from input), the influence of L1 can remain a persistent barrier, especially when there is a lack of explicit corrective feedback (negative evidence) against L1-based patterns. The Role of Input and Evidence The study’s findings underscore the differential roles of positive and negative evidence in L2 acquisition. Positive evidence, derived from encountering correct language use in the input, is crucial for acquiring L2-only collocations. The high frequency of these collocations in Japanese input likely contributed to the improved performance of long-LOR learners. This aligns with usage-based models of language acquisition, which emphasize the impact of input frequency on learning. Conversely, the persistent difficulty with L1-only collocations suggests a lack of sufficient negative evidence. When a collocation is acceptable in the learner’s L1 but incorrect in the L2, learners may not readily recognize its unacceptability without explicit correction. The absence of this negative feedback in naturalistic input environments allows L1-based semantic representations to remain entrenched, hindering the complete restructuring of L2 lexical knowledge. Developmental Trajectories and Ultimate Attainment The research provides insights into the developmental stages of L2 lexical semantics, building upon models like Jiang’s (2000) three-stage framework: formal, lemma mediation, and integration. Medium-LOR learners appear to be largely in the lemma mediation stage, where L2 word use is influenced by L1 translations. Long-LOR learners show progress towards the integration stage, particularly with L2-only collocations, but their continued acceptance of L1-only collocations suggests that this integration remains incomplete. This study contributes to the ongoing debate about ultimate attainment in L2 acquisition. While previous research has shown that long-term residence can lead to near-native proficiency in syntactic domains, this study suggests that semantic knowledge, particularly at the lexical level, may be more resistant to full native-like acquisition. The findings imply that achieving complete native-like semantic competence, especially in areas where L1 and L2 diverge significantly, may be exceptionally challenging for adult learners, even with extensive exposure. Broader Implications and Future Directions The study’s results have significant implications for language pedagogy and our understanding of L2 learning. It highlights the importance of not only providing ample exposure to the target language (positive evidence) but also incorporating explicit instruction on linguistic mismatches and providing corrective feedback (negative evidence) to address L1 transfer. Future research could explore the efficacy of different types of feedback and instructional methods in helping learners overcome these semantic challenges. Longitudinal studies tracking learners over extended periods, combined with neurolinguistic approaches like event-related potentials (ERPs), could further elucidate the dynamic processes of semantic restructuring and the potential for achieving native-like proficiency in L2 semantics. Understanding these nuances is critical for developing more effective strategies to support advanced L2 learners in achieving their language learning goals. Post navigation Exploration of the Viewing Intention Mechanism of Chinese Traditional Culture Animated Films Based on SOR and TPB
