This study investigated L1 (first language) influence on syntactic and semantic reconstructions by English, Spanish and Korean speakers learning L2 (second language) Chinese Applicative Constructions manifested as Double Object Construction (DOC) and Double Unaccusative Construction (DUC). Specifically, I considered the learnability problem, posed by the superset-subset relation between learners' native language and L2 Chinese regarding the two constructions.
A total number of 117 L2 participants learning Chinese and 20 native speakers of Chinese participated an empirical experiment composed of a syntactic judgment task and a semantic interpretation task. It was found that if the target property of the L2 Chinese was in subset relation with the corresponding property in learners' L1, without explicit instructions, L2ers were unable to unlearn the property that existed in L1 but did not exist in L2, supporting Yuan's (2014) Dormant Features Hypothesis. In cases where the target property of the L2 Chinese was in superset relation with the corresponding property of speakers' L1, without explicit instructions, L2ers were able to reconstruct L2 grammars to better accommodate the target property. However, the reconstructed L2 grammars only had face value and learners' syntactic representations remained non-native-like even for learners with advanced L2 Chinese proficiency. This finding lends support to Hawkin's (2000; 2005) Representational Deficit Hypothesis.