Interfaces and Features in Second Language Acquisition: A Study on the Acquisition of Chinese Negation by English Speakers and Korean Speakers


Student thesis: Doctoral Thesis

View graph of relations



Awarding Institution
Award date10 Sep 2018


The present study investigates the acquisition of Chinese negation by English speakers and Korean speakers with both grammaticality judgment data and learner corpus data. Within the theoretical framework of the Interface Hypothesis (Sorace & Filiaci, 2006) and the Feature Reassembly Hypothesis (Lardiere, 2008), we examine whether L2 learners could achieve nativelike representation at the multiple interfaces involved with Chinese negative structures and whether they could reassemble the mood features, [±realis], encoded with Chinese negation markers (bu/mei), from their L1 configurations. The variables underlying L2 acquisition at interfaces and the process of feature reassembly are also explored. This research also probes into the representation of explicit and implicit knowledge about Chinese negation in L2 grammar.

A written grammaticality judgment test and an aural grammaticality judgment test were administered to L2 learners and native controls. L2 learners included a group of English speakers (N=90) and a group of Korean speakers (N=92). Each L2 group was divided into five subgroups, from elementary to advanced groups, based on learners’ scores in a Chinese proficiency test. Native controls included 15 Chinese native speakers. Production data on the use of Chinese negation markers by L1-English and L1-Korean learners were also extracted for analysis from four existing Chinese learner corpora (overall scale: 15.19 million characters), including the HSK Dynamic Composition Corpus, the TOCFL Composition Corpus, the Error-tagged Chinese Interlanguage Corpus, and the Jinan Chinese Learner Corpus. 

The results of this study provide evidence for examining the L2 theories in question. First, the results from the judgment data and the learner corpus data do not completely support the assumptions of the Interface Hypothesis. While it is found that L2 learners from both L1 backgrounds achieve success at the narrow syntax, there is much variability in their acquisition at the internal interfaces involved with Chinese negation, including the syntax-semantics interface, the syntax-morphology interface, and the syntax-semantics-morphology interface. The Interface Hypothesis thus needs to be refined by considering the variables that modulate the acquisition of interface structures (B. Yuan, 2010).Second, the results of our study in general are consistent with the proposals of the Feature Reassembly Hypothesis. It is shown that L2 learners with lower proficiency tend to map their L1 features encoded with negation onto corresponding L2 items, resulting in their target-deviant performances at the early stages of their learning. With increased Chinese proficiency, both L1-English and L1-Korean learners could assemble the [+realis] feature with mei and the [‒realis] feature with bu although they have great difficulty in detecting the mood features in certain licensing contexts, including past habitual activities, the durative aspect, and hypothetical conditionals.Third, the results of the present study reveal multiple factors contributing to the variability in L2 acquisition at the interfaces involved with Chinese negative structures, including L1 influence, the quantity (input frequency) and the quality of the target input (input consistency and regularity), as well as L2 proficiency. These factors also underlie the detectability and reassembly of the [±realis] features encoded with bu/mei in different licensing contexts.

Fourth, task modality (written vs. aural) seems to play a role in L2 learners’ access to explicit and implicit knowledge about Chinese negation, but the effect of task modality is constrained by other factors such as structural/feature complexity, L2 proficiency, and L1-L2 similarity. 

On the whole, the approach of employing both elicited experimental data and authentic learner corpus data furnishes comprehensive evidence for the acquisition Chinese negation by L2 learners. The findings of our study are of significance to the examination of the Interface Hypothesis and the Feature Reassembly Hypothesis in accounting for L2 acquisition and also to the instruction of Chinese negation in teaching Chinese as a foreign language.

    Research areas

  • Interfaces, Features, Second language acquisition, Chinese negation, English speakers, Korean speakers