Date of Award

1-1-2012

Document Type

Campus Access Dissertation

Department

Linguistics

First Advisor

Anne Bezuidenhout

Second Advisor

Robin Morris

Abstract

The general goal of this dissertation is to explore two decades-long second language (L2) acquisition issues: whether and to what extent the L2 learner's first language (L1) and Universal Grammar (UG) respectively play a role in L2 development. Targeting English verb phrase ellipsis (VPE) and wh-extraction out of VPE, four acceptability judgment tasks (AJT) are administered among four groups: three groups of Spanish speakers and a control group of English native speakers. The AJTs were run to investigate research questions related to the L2 acquisition issues: (1) Do L1=Spanish L2=English adult learners at an initial stage of learning transfer their L1 = Spanish grammar to their representation of L2=English? (2) Can L1=Spanish adult learners at a higher proficiency level reset the VPE parameter to the L2=English values? (3) Can L1=Spanish adult learners at a higher proficiency level acquire the constraint prohibiting wh-extraction from VPE-internal islands without positive evidence (which is identified as the Poverty of Stimulus problem)? The results of the first two experiments showed that Spanish-speakers transferred three of the four L1=Spanish VPE descriptive properties (related to functional categories) to the L2=English VPE at an earlier stage of L2 acquisition, but at a later stage successfully reset the VPE parameter to English values. The other two experiments found that L1=Spanish adult learners successfully acquired the abstract constraint regulating VPE-extracting of wh-phrases. These results from the four experiments suggest that L1=Spanish adult learners transfer their L1 grammar at an initial stage and at a later stage succeed in acquiring the target L2 grammar through a successful parameter resetting and overcoming the Poverty of Stimulus problem. This leads to the conclusion that both L1 and UG play a role in L2 development without any limit placed on both. This conclusion is followed by the evaluation of four hypotheses regarding the two L2 acquisition issues above. The only hypothesis that survives the testing against the three findings of this study is the Full Transfer Full Access view presented by Schwartz and Sprouse (1994, 1996). The transfer of L1 properties related to functional categories is not compatible with the Partial Transfer Full Access hypothesis by Vainikka and Young-Scholten's (1996). On the other hand, both the successful parameter resetting and the Poverty of Stimulus phenomenon contrasts with the predictions from the Failed Functional Features Hypothesis by Hawkins and Chan (1997) and Bley-Vroman's (1988, 1989, 1991, 2009) Fundamental Difference Hypothesis (FDH); however, there is a single exception that the success in parameter resetting might be entertained by the FDH in another way, that is, through the use of general cognitive abilities such as problem-solving based on positive evidence.

Rights

© 2012, Eun Hee Lee

Download

Share

COinS