미얀마인 초급 학습자의 한국어 어두 폐쇄음 산출 양상

Abstract

This study investigated the characteristics of Korean word-initial stops produced by Myanmar beginner learners and analyzed the causes of their pronunciation errors. To this end, perception tests involving native Korean speakers were conducted, and acoustic analyses of the learners' productions were performed, examining the correlation between pronunciation errors and acoustic features. The perception test results showed that 39.6% of the utterances were perceived differently from the target sounds, indicating low production accuracy and a potential for difficulty in communication. Acoustic analysis revealed that aspirated stops produced by Myanmar learners had the longest VOTs, fortis stops were shorter, and lenis stops were the shortest. However, the VOT difference between lenis and fortis was not statistically significant. Furthermore, f0 values did not exhibit clear distinctions among the three phonation types, and their distributions largely overlapped. Native Korean speakers primarily distinguished aspirated and fortis stops based on VOT in the learners' productions. However, lenis stops were not clearly differentiated from other phonation types. This suggests that Myanmar learners do not effectively utilize f0 for phonation type distinction, a crucial cue for native speakers, especially in differentiating lenis stops. Additionally, their lenis VOTs were not realized as sufficiently long as those typically produced by native Korean speakers. These findings clearly demonstrate that Myanmar learners have difficulty accurately producing the subtle acoustic differences in Korean lenis, fortis, and aspirated stops, especially concerning their VOT and f0 realization patterns. The results indicate that Korean pronunciation education for Myanmar learners should prioritize teaching the effective utilization of f0 for phonation type distinction, while also emphasizing accurate target VOTs for each phonation type and requiring intensive practice.