| Previous studies (e.g., Kweon, 2001 a; Park & Goldner, 2005) testing L2 learners' knowledge about wanna contraction reported that the learners incorrectly accepted subject extraction questions with contraction (e.g., *Who do you wanna advise Mary?). Such studies relied on grammaticality judgment and production data to assess the learners' grammatical knowledge. Recent L2 research, however, employs psycholinguistic measures to supplement judgment and production data in order to better understand the nature of L2 acquisition.; This dissertation aims to explore L2 learners' knowledge about the wanna contraction constraint from a processing perspective. Using an on-line self-paced reading task coupled with an off-line grammaticality judgment task, data were collected from 103 Japanese-speaking learners of English who were placed into three different proficiency groups and from 41 native speakers of English as a control group. The experimental stimuli consisted of six conditions: Object extraction, subject extraction, and no extraction, all of which were presented with and without contraction. Subject extraction with contraction (e.g., * Who do you wanna advise Mary?) was the only ungrammatical condition.; The results indicated a developmental pattern of the learners' knowledge about the contraction constraint. Furthermore, the on-line processing data showed a parallel result to the judgment data in that those who distinguished subject extraction with and without contraction in grammaticality judgments also distinguished the two conditions in the self-paced reading task. Similarly, those who were likely to overgeneralize contraction to subject extraction, that is, accept subject extraction with contraction, were found to show no difference between the two conditions in the on-line reading time data. Thus, the presence and the absence of the constraint were reflected in the native speakers' as well as learners' on-line reading. A detailed analysis of the reading data also suggests different reading profiles between the two language groups. The findings are discussed with reference to psycholinguistic proposals made about L1 and L2 sentence processing, and the concept of no contraction constraint will be reexamined. |
