Implicit L1 Negative Transfer in Unmonitored Contexts: A Cross-L1 (Mandarin vs. Sinhala) Study of Common Additive Negation Mistakes and Immediate Self-Correction Among Highly Proficient L2 English Speakers

Existing research in second language acquisition (SLA) largely contends that L1 negative transfer diminishes with increasing of L2 proficiency and has predominantly focused on formal contexts of learners with intermediate to lower proficiency. Insufficient attention has been paid to the implicit transfer and immediate self-correction among highly proficient L2 English users in unmonitored settings, particularly regarding common features across non-cognate L1 backgrounds. This study examines highly proficient L2 English users with Mandarin and Sinhala L1s via six-month participatory observation and recording, collecting thirteen typical samples of their informal conversations. Focusing on the recurrent mistake patterns in expressing "additive negation" (i.e., *"I also don’t..." and "*I also can’t...") and rapid self-correction, and drawing on six key SLA theories, this paper finds that such implicit transfer stems from the shared L1 word order features in additive negation and the disconnection between explicit grammatical knowledge and implicit L1 thinking habits. This study contributes to filling the gap in understanding shared mistakes among highly proficient L2 English speakers from distinct L1 backgrounds in unmonitored production, offering empirical insights into the development of implicit competence in L2 English acquisition.