SECOND LANGUAGE PROCESSING OF DERIVATIONAL AND INFLECTIONAL MORPHOLOGY IN ENGLISH

Abstract

This dissertation investigates how later second language (L2) learners process derivational and inflectional morphology to explore whether later L2 learners can develop lexical or syntactic representations qualitatively similar to native speakers’. As the current literature does not provide a conclusive answer to this question, two experiments were conducted to address this gap.

Experiment 1 investigates how later learners process English derived words. Native English speakers and native Korean-speaking advanced learners of English were compared in their performance on an overt-priming lexical decision task. The results show that not only the native speakers but also the L2 learners were able to morphologically decompose derived words in English. Both groups’ data showed a robust morphological priming effect with semantically transparent morphologically related prime-target pairs (e.g., politeness-polite). Furthermore, both groups’ priming effects were similarly affected by major determinants of the morphological decomposability of derived words (whole-word frequency, base frequency, and morphological productivity).

Experiment 2 examines how later learners of English process plural inflection in English. In a self-paced reading task, native English speakers and native Korean-speaking advanced learners of English were tested on their (in)sensitivity to plural errors in two different structures: a simple DP/QP structure (e.g., those long Latin word*(s)) and a partitive structure (many of her book*(s)). The results indicate that despite the striking differences in plural marking between English and Korean, the learners were sensitive to plural errors in both structures, and that their sensitivity to the errors was affected by the structural distance of the feature-checking dependency related to plural inflection. These findings suggest that (1) L2 learners can acquire target-like L2 inflection knowledge, even if such inflection knowledge is not present in their L1, contra the representational deficit account (e.g., Hawkins & Hattori, 2006), and that (2) they compute hierarchically structured representations during real-time language comprehension, contra the shallow structure hypothesis (Clahsen & Felser, 2006).

Taken together, the results of the two experiments suggest that, although there are many differences between L1 and L2 learners’ knowledge of the target language, advanced L2 learners can develop target-like mental representations of certain aspects of lexical and syntactic knowledge.