Korean EFL Learners Interlanguage Null Objects: A Syntactic-discourse Exploration of Unlearning Patterns

Abstract

It has been suggested that null objects in Korean are licensed by the topic-prominent feature of the language: namely, topic-chaining and the topic NP deletion rule (Huang, 1984). However, null objects do not appear in English because neither topic chains nor topic-drop is allowed. Such syntactic and pragmatic differences between Korean and English often cause Korean EFL learners to produce null objects in their L2. The present study investigated (i) how frequently Korean EFL learners produce null objects, (ii) what causes such a null object phenomenon, and (iii) whether it can be unlearned. The thesis consists of a pilot study and a main study. In the pilot study, 36 Korean high school students and 9 native English speakers participated in a story-retelling task. The results show that null objects were produced more frequently than null subjects and that more proficient learners tended to use null objects less frequently. In addition, null objects were produced more frequently in recoverable contexts than in non-recoverable contexts. The main study investigates on a larger scale detailed properties of null object production by Korean EFL learners. It examines whether the learners production of a null object is related to various factors, such as English proficiency, the recoverability of its reference in the discourse, and the complexity of the argument structure of its verb. A total of 167 Korean college students and 9 native English speakers participated in the main study, and performed two types of writing tasks (i.e., discourse-based and sentence-based tasks). The results show that the Korean EFL learners produced null objects more frequently than null subjects in both the discourse-based and the sentence-based writing tasks. This asymmetry was also observed in the pilot study and is consistent with previous findings (Yuan, 1997

Park, 2004

Hwang, 2005). The thesis argues that unlearning null objects is more difficult and takes more time than unlearning null subjects because it involves acquiring the correct argument structures of each verb. Second, as the learners English proficiency improved, their production of null objects decreased. This result is not consistent with Yuan (1997), but supports Hwangs (2005) research. This suggests that the learners were able to unlearn topic-prominent properties responsible for object-drop (i.e., topic-chains and the topic-NP deletion rule) and acquire the obligatory nature of overt objects in English. Third, null objects were produced more frequently in the discourse-based task than in the sentence-based task. This indicates that object-drop in L2 English is closely related to the recoverability of reference in the discourse context, as it is in L1 Korean. It suggests that null objects in the learner language are licensed by topic chains and the topic NP deletion rule, transferred from L1. Finally, the learners production of null objects was affected by the verbs linguistic characteristics. As for the complexity of complement structures, the more complex the argument structure of a verb, the more frequently null objects were produced. In addition, the learners tended to drop the objects of familiar verbs less frequently than those of less familiar verbs. Finally, the learners produced null objects very frequently when a verb in English allows its object to be optional. The findings of this study carry pedagogical implications concerning how to help Korean EFL learners unlearn null objects. Instruction on English verb complementation, particularly construction grammar-based instruction, will lead Korean EFL learners not to produce null objects. In addition, discourse-based grammar teaching will be valuable and effective since object-drops are closely related to the referential patterns of NPs in connected discourse.