중국인 고급 학습자의 한국어 후설모음 발음에 대한 음향학적·지각적 분석

Abstract

This study examined the acoustic properties of the korean back vowels /ɯ/, /ʌ/, /u/, and /o/ as produced by advanced Chinese learners of Korean, and examined their acoustic differences compared to native Korean speakers. In addition to conventional static formant analysis, the study incorporated an analysis of formant trajectories over time. Furthermore, a perception experiment with native Korean listeners was conducted to investigate how these acoustic differences affect perception. According to the results of the conventional static formant analysis, advanced Chinese learners produced back vowels within a narrower acoustic range compared to native Korean speakers, while showing a wider distribution in terms of tongue height. Additionally, in the acoustic space, each vowel only partially overlapped with the distribution of native Korean speakers, indicating that there are differences in how the target vowels are produced. As observed in previous studies, native Korean speakers tended to raise /o/, resulting in acoustic merging with /u/. Similarly, the learners also exhibited acoustic merging between /o/ and /u/; however, unlike the native speakers, their productions showed little distinction in tongue advancement (F2). The analysis of formant trajectories also revealed that Chinese learners exhibited significant variation in both F1 and F2 trajectories during the production of each vowel. This indicates that the articulation of back vowels by advanced Chinese learners involves active movement of the articulatory organs. Such variations in formant trajectories suggest that learners attempt to approximate the target articulations of native Korean speakers through active articulatory adjustments. However, despite these efforts, the learners ultimately failed to fully reach the target phonemic range. Through formant analysis, acoustic differences were identified between advanced Chinese learners and native Korean speakers in the production of back vowels. Based on this, a perception experiment conducted with native Korean listeners showed that the vowel /u/, which exhibited the greatest difference from native speakers, had the lowest perception accuracy. Additionally, patterns of vowel perception confusion were also observed. An analysis of the interaction between acoustic differences and perceptual errors revealed that advanced Chinese learners produced back vowels within a narrower acoustic space, resulting in shorter acoustic distances between vowels compared to native Korean speakers. As a result, the increased proximity among vowels led to greater acoustic similarity, which contributed to perceptual confusion among native Korean listeners. Furthermore, the articulatory movements involved in the production of back vowels by the learners caused variations in formants, which may have blurred acoustic boundaries and further contributed to perceptual confusion. Although continuous acoustic changes were observed during the production of back vowels by advanced Chinese learners, statistical analysis showed that only the vowel endpoint had a significant effect on perception. In the perception of back vowels, the formant values at the endpoint served as critical acoustic cues, and the factors influencing perception varied depending on the type of misperception. When vowels were misperceived as unrounded, the F1 value—reflecting tongue height—was the primary cue influencing perception, whereas when they were misperceived as rounded vowels, the F2 value—which reflects the frontness or backness of the tongue—also had a significant influence. These findings suggest that the perception of back vowels is shaped by the interaction of multiple complex acoustic features.