| Description |
One of the biggest challenges facing adult language learners is acquiring the sound system of the second language. While it has been shown that adults can acquire novel phones in their second language, the relevant features of the input available for use remain to be determined. Two proposed features of the input are investigated in this dissertation. The first study investigates the role of phonological distributional information in the acquisition of allophones. Native English speakers were exposed to the Spanish [b]- [β] alternation in either overlapping or complementary distribution. It was expected that if participants inferred the phonological relationship from the phonological distribution, participants exposed to the pair in overlapping distribution would outperform participants exposed to the pair in complementary distribution on an ABX task. The two groups perform the same on the discrimination task, suggesting that phonological distributional information alone may be insufficient for adult allophone acquisition. The second study, investigating whether perceptual gains from high variability phonetic training (HVPT) can generalize to higher level tasks, employs a freely available online implementation of HVPT (English Accent Coach [EAC]) with actual second language English (ESL) learners in two scenarios: part of ESL class sessions or at the times and locations of the learners' choosing (simulating extracurricular practice). In each setting, ESL students trained using EAC and a control group of ESL students complete a pretest and a posttest, each containing a discrimination task and an artificial-lexicon learning task. In the classroom setting, students from both groups demonstrated no improvement on either task from pretest to posttest. In the extracurricular setting, the HVPT group outperformed the control group on the discrimination posttest, but there was no difference for either group on the lexical posttest. Together, these results suggest that (1) efficacy of HVPT for perception improvement relies on adherence to some protocols and (2) the ability to use perception knowledge for a lexical task may not develop in parallel with perception. The results are discussed with regard to their implications for language teaching and their contributions to our understanding of what features of the input adults are able to take advantage of for L2 phonological development. |
