The Role of Perceptual and Acoustic Similarity in Learners’ Perception of L2 Vowels

초록

This study investigates the role of L1-Korean-L2-English vowel similarity and acoustic distance in L1-Korean speakers’ perception of L2-English vowels, given that few studies have examined the joint effects on learners’ L2 vowel perception of L1-L2 sound similarity/dissimilarity and the acoustic distance of L2 vowel production from native speaker reference. Thirty-three Korean learners of English produced and identified 12 English vowels (e.g., bet, but, boat) and mapped the same English vowels onto close Korean vowels. Acoustic distances of the learners’ produced vowels from native English speakers’ norms were calculated as Euclidean distance in the F1/F2 space. The learners’ vowel identification accuracy was assessed and L1-L2 perceptual vowel similarity/dissimilarity was calculated based on the learners’ L2-to-L1 mapping patterns. Results of a linear mixed-effects regression model showed a significant main effect of vowel type (i.e., similar, somewhat similar, dissimilar), indicating that the perception accuracy was significantly higher for L1-similar L2 vowels compared to L1-somewhat similar and L1-dissimilar L2 vowels. The main effect of acoustic distance was also significant, with greater distance associated with lower perception accuracy for L1-similar L2 vowels. However, accuracy rates were varied within each vowel type as the identification accuracies of some L1-somewhat similar or L1-dissimilar L2 vowels tended to be high when they bear salient perceptual cues like vowel duration or r-coloring. These results are only partly in line with the propositions of the Speech Learning Model (SLM/SLM-revised, Flege 1995, Flege and Bohn 2021) concerning the relative difficulty and easiness of the acquisition of L1-similar L2 vowels versus L1-dissimilar L2 vowels, indicating that different L2 vowels develop at different rates. © 2025 Elsevier B.V., All rights reserved.

키워드