The Plasticity of the Native Language in Adult Second Language Learning

Open Access
Bice, Kinsey L
Graduate Program:
Master of Science
Document Type:
Master Thesis
Date of Defense:
October 13, 2014
Committee Members:
  • Judith Fran Kroll, Thesis Advisor
  • second language learning
  • native language change
  • ERPs
  • comprehension
  • production
  • lexicon
  • N400
Learning a second language (L2) in adulthood is a difficult task that many attempt with varying levels of success. Past research has focused on hard maturational constraints as the constraining factor in adult L2 learning, but recent studies challenge those accounts by demonstrating that some adult L2 learners have achieved native-like processing in the L2. The present study shifts the focus in adult L2 learning from the L2 to the native language (L1), by testing a new hypothesis that the L1 must change in the process of L2 learning, and that the learners who are better able to tolerate L1 changes achieve higher L2 proficiency. Tolerance to L1 change incorporates two types of changes, L1 costs and L2 sensitivity. L1 costs are thought to be modulated by inhibitory control, whereby L2 learners who are better able to inhibit the L1 would have lower L1 performance compared to monolinguals, but overall higher L2 proficiency. The other prediction was the L2 learners who reveal the influence of the L2 during L1 processing, or L2 sensitivity, would have higher L2 proficiency. To test this hypothesis, L2 learners of various proficiency levels were tested in the L1 and the L2 on tasks of comprehension (lexical decision task) and production (semantic fluency task) and compared to monolinguals in the L1 tasks. ERP measurements were taken during the comprehension task to investigate neural sensitivity to the L2 that might not otherwise be present in behavioral measures. Statistical analyses revealed that L2 learners and monolinguals benefited from the presence of cognates in the L1, but the ERP waveforms and scalp distributions revealed that only the L2 learners demonstrated neural activity consistent with cognate facilitation, in the form of a reduced N400 for cognates compared to noncognates. Interestingly, the L2 results from the comprehension task showed that the beginning L2 learners did not produce an N400 for cognates compared to noncognates, but the advanced L2 learners had a clear N400 effect. Results are discussed in terms of the hypothesis and past research finding L2 sensitivity during L1 processing.