Second Language Acquisition of Variable Use of the Nominative and Accusative Case Morphemes in Korean: A Corpus Study

Abstract

The goal of this dissertation is to examinethe the second language (L2) acquisition of variable use of nominative/accusative case morphemes by English- or Japanese-speaking adult learners of Korean.

In Korean, case marking of subjects/objects is optional. Previous research proposes that the naturalness of case (un-)marking is constrained by semantic features for prototypical subject/object-hood (i.e., markedness features) such as animacy and definiteness (Lee, 2006b) and by discourse features such as newness (Sohn, 1999) and various focus subtypes (Lee & Choi, 2010). Regarding the L2 acquisition of these conditioning features, the Interface Hypothesis (henceforth IH, Sorace, 2011) posits persistent non-nativelikeness at the syntax-discourse interface, whereas the Feature Reassembly Hypothesis (henceforth FRH, Lardiere, 2009) predicts that L2 features realized in dissimilar configurations from those of the L1 pose substantial difficulty but are acquirable in principle given evident morphological contrasts.

Using a spoken corpus of media discourse, this study investigates whether learners’ L2 morpheme use displays nativelike sensitivity to these markedness and discourse features. The data were collected from eight advanced adult learners of Korean whose L1 is English or Japanese and four native controls. No effect of the markedness factor was observed among either the L1 controls’ or the L2 speakers’ results. The discourse factor, on the other hand, was found to have significant effects on all three groups’ overall performance, in that overt case marking is more prevalent among discourse-new or contrastive focus DPs, suggesting that these learners have acquired the interaction between discourse context and case marking. However, further analysis revealed that while the native speakers tend to integrate an exclusivity sub-distinction for contrastive focus in discriminating the two morphemes, neither English nor Japanese speakers recognized this sub-distinction, except for one English speaker. Still, the Japanese speakers, whose L1 has a similar case system to the L2, performed better than the English speakers in overall frequency of case (un)marking and on other sub-patterns.

These findings indicate that adult L2 knowledge of at least some conditioning factors for variability phenomena at the syntax-discourse interface is attainable; the degree of L1-L2 dissimilarities plays a significant role in L2 development, providing additional support for the FRH.