Second language processing in reading and translation

The purpose of this dissertation is to investigate the processing mechanisms of non-native English speakers at both the sentence level and the morphological level, addressing the issue of whether adult second language (L2) learners qualitatively differ from native speakers in processing linguistic input. Using psycholinguistic on-line techniques (self-paced reading and eye-tracking paradigms) plus an off-line translation task, four experiments explore the use of syntactic and semantic information in L2 comprehension and the sensitivity to morphological violations in L2 processing by Korean late learners of English. A translation paradigm is used to examine how morphosyntactic and semantic knowledge are integrated in the language learner's mind. Task effects are examined by comparing cognitive processes involved in reading for comprehension and reading for translation. The degree to which L2 proficiency modulates processing behaviors is also considered. Expeirments 1 and 2 investigate how Korean learners of English integrate syntactic and semantic knowledge during processing L2 input in a translation task, using active/passive structures in English. The results demonstrate that L2 leanrers use both syntactic and semantic processing routes and that their interpretations are sometimes unfaithful to original context but often contain correct morphosyntax when output from the two routes conflict. Experiment 3 examines L2 learners’ reading behaviors compared to native speakers and the possible influences of reading goals on reading behaviors of L2 readers. The results reveal that reading time patterns of L2 learners are similar to that of native speakers, although L2 learners are influenced by semantic information to greater extent than native speakers. The results also indicate that the use of syntactic information in L2 speakers becomes more automatic as L2 proficiency increases, and that the translation task forces low-proficiency learners to increase attentions to syntactic information during on-line comprehension. Experiment 4 explores in an eye-tracking paradigm whether L2 learners show sensitivity to subject-verb agreement violations similar to native speakers during on-line comprehension, and how task effects modulate the sensitivity. The results show that L2 learners are able to display sensitivity to morphoglical violations during reading, and that the translation task forces L2 speakers to perform deeper processing, resulting in more native-like processing. All together, the four experiments in this disseration make contributions to L2 processing research, mainly revealing (a) L2 learners’ syntactic representations computed during on-line comprehension are not underused (under certain task conditions); (b) L2 learners’ sensitivity to subject-verb agreement violations are not reduced (under certain task conditions); and (c) reading goals and L2 proficiency affect the depth of processing in L2 comprehension. The overall results are discussed in lights of ‘good-enough’ and goal-dependent (Christianson et al., 2001; Ferreira &Patson, 2007; Swets et al., 2008) language processing in L2, pointing to quantitative, rather than qualitative differences between L1 and L2 processing