Effects of Web-based Auditory Training on the Perception of Korean Sounds by Mandarin Learners of Korean

Abstract

This dissertation is concerned with the perceptual acquisition of Korean sounds by adult second language (L2) learners and suggests web-based auditory training as an effective learning tool to enhance L2 learners’ perceptual accuracy of Korean sounds. Experiment 1 explores the role of experience in the perception of Korean three-way laryngeal contrasts of stops and affricates by 44 native Mandarin listeners differing in Korean language experience. The results revealed that most Mandarin listeners did not reach the same level of perceptual accuracy as native Korean listeners; however, advanced learners perceived contrasts more accurately than beginner and intermediate learners. This indicates that the greater the adult learners’ experience in the L2, the better their perceptual accuracy of L2 contrasts. The goal of Experiment 2 is to examine whether L1 inventory size affects L2 perception by comparing 28 English and 28 Mandarin listeners’ identification accuracy for Korean vowels and codas. The results showed that English listeners, whose L1 has a rich vowel and coda inventory, outperformed Mandarin listeners, who have a small L1 vowel and coda inventory, suggesting that L1 inventory size has a significant effect on L2 perception. Experiment 3 investigates the effects of web-based auditory training and compares the efficacy of attention on the perception of Korean vowels and codas. For this study, 45 Mandarin listeners were randomly assigned to one of two training groups to receive either vowel- or coda-training, both exposed to identical stimuli, or to a control group to receive no training. Results demonstrated that both training groups outperformed the control group for the identification of target sounds, suggesting a beneficial effect of web-based training of L2 sounds. In addition, both the vowel- and coda-trained group yielded significant improvement in identifying Korean vowels; however, only the coda-trained group achieved a significant increase in coda identification accuracy. These findings indicate that attention type differentially affects perception depending on the target structure and highlights the benefits of explicit over implicit training for at least some L2 sound structures. Furthermore, learners were able to generalize knowledge to a novel phonetic context, providing evidence that robust learning has occurred during training.