A Study Of Subject/Object Asymmetry In Acquiirng Long-Distance Wh-/Questions

This study examines whether the subject/object asymmetry observed in L1wouldalso occur in the acquisition of English long-distance wh-questions by Chinese learnersof English with different proficiency levels. Two opposite asymmetry patterns werereported in previous related studies as a result of methodological discrepancy and thelimitation of material design. Correspondingly, there are two parsing-based accountsfor each asymmetry pattern. Specifically, they are the Generalized Theta Attachment(GTA) and the Distance-Based Theory (DBT) of linguistic complexity, based on gapreanalysis and dependency distance respectively. This empirical study is an attempt toacquire valid data to explore this possible subject/object asymmetry in a morepanoramic view, thus to make clear which account can provide a better explanation forthe processing of long distance wh-question in a unified way and to identify what is thedeterminant factor of the target structures’ complexity. This thesis addresses thefollowing three questions:1） Does the subject/object asymmetry observed in L1acquisition also occur inthe processing of English long-distance wh-questions by Chinese learners of English?2）Which type of extraction is preferred by Chinese learners, subject or objectextraction? Does the preference pattern vary in the learners with different proficiencylevels?3）Do the data-collection measures affect the results in the preference patterns？Two experiments, a Grammaticality Judgment Test (GJT) and an ElicitedComprehension Test (ECT), were carried out with four groups of Chinese learners withdifferent proficiency levels: the9thgrader; the12thgraders; the3rdyear English-majors and the graduates. There were30students in each group. The results aredemonstrated as follows:1）The current study did find an asymmetry between subject wh-questions andobject counterparts among Chinese learners of English in terms of bothGrammaticality Judgment Test (GJT) and Elicited Comprehension Test (ECT). Thefindings from the current grammaticality judgment data has added new evidence thatthe performance on judging grammatical long-distance wh-questions by Chinese EFLlearners of English with different proficiency levels are influenced by processingeffects because the investigated questions were all grammatical and different from oneanother only in their extraction types or the finiteness of embedded clauses, but the learners treated them differently.2) Chinese learners with different proficiency levels exhibit the same strongpreference towards subject extractions regardless of the finiteness of the embeddedclauses, and they performed better on subject wh-questions, which is contradictory tothe findings of the previous studies with advanced learners in L2literature (Juffs andHarrington,1995; White&Juffs1998; Juffs,2005) mainly by means ofGrammaticality Judgment on one hand. Therefore, this finding argues against thegreater difficulty of subject-extracted wh-questions within the framework of GTA. Onthe other hand, it is in conformity with that of L1study (Wilhelm&Hanna,1992;Philip et al.,2000) in Elicited Comprehension task: Comparable to the L1children, theparticipants at earlier stages were also more likely to misinterpret the objectextractions as the subject counterparts, which confirms the relative ease in processingsubject wh-questions and is entirely consistent with the predictions of processingdifficulty from the framework of DBT. Therefore, it can be concluded that theprocessing difficulty of long-distance wh-questions is mainly determined by thefiller-gap dependency distance.3) The preference pattern can not be affected by the data-collection method itself.The learners in my study showed the same preference for the subject wh-questions inboth tests, with each test representing one data-collection method, which can refute theassumption that the conflicting results stem from the different data-collection methodsadopted in the previous studies.