언어사용역에 따른 고모음 무성음화의 변이

Abstract

This study analyzes whether Korean mothers tend to apply a less high vowel devoicing when speaking to children than when speaking to adults. In particular, we test whether such a trend is the effect of registers or mediated by factors such as the vowel duration. The results of this study contribute to revealing the phonetic features and nature of Child directed speech and to understanding Korean high vowel devoicing. The data used in the study used speech data in which 10 mothers explained their vocabulary to children and adults, respectively, through reading books and spontaneous speech. As a result of mediated regression analysis, the vowel duration was applied as a mediator between the realization of devoicing sounds according to the register, and it was found that the register was a variable affecting high vowel devoicing. Additionally, as a result of analyzing whether there are other variables that can affect the realization rate of high vowel devoicing, it is found that high vowel devoicing is realized more frequently in [ɨ], and when vowels are located at the end of speech. In the case of preceding consonants, it occurs most often in fricative, and it can be seen that high vowel devoicing occurs more often when the preceding consonants are lenis. This study shows that Korean child directed speech can facilitate children's acquisition of high vowels by applying high vowel devoicing. However, it was revealed that this was a mediating effect due to the slow speaking rate rather than the independent effect of the usage area. In addition, this study showed that the non-categorical phonetic characteristics of the independent variable, the length of the vowel, affected the high-pitched voiceless sound in Korean.