Sociophonetic Variations in Korean Constituent Final -ko and -to

Abstract

The purpose of this dissertation is to examine (i) linguistic and extralinguistic factors that influence vowel raising of /o/ in constituent-final –ko and –to in Seoul Korean and (ii) listeners’ perceptions of this vowel raising and social meanings of the raised variant. The analyses are based on production data collected from one-on-one sociolinguistic interviews and an elicitation task, and on perception data from a matchedguise test. The production data demonstrate that some phonetic and prosodic environments significantly affect the vowel’s height and/or frontness. In the AP-medial position, /o/ shows higher F1 values when it is in an NP particle or followed by /a/, and it shows lower F1 values when it is preceded by /t/ or followed by /i/, /u/, or alveolar, palatal, or velar consonants. Moreover, preceding /t/, following /i/, /j/-diphthongs, /e/, and alveolar, palatal, and glottal consonants lead to higher F2 values. In addition, in the prosodic final position, following /i/, /j/-diphthongs, and bilabial, velar, and glottal consonants increase the F1 values, a final Low tone decreases the F1 values, and preceding /t/ increases the F2 values. Most of the effects of the following and preceding segments can be explained as the result of coarticulation of the vowel and adjacent sounds. Extralinguistic factors influence the vowel raising as well. The production data show that older speakers use a more raised variant than middle-aged speakers in the APmedial position, which implies that the vowel raising of /o/ in constituent-final –ko and –to shows age-grading. In addition, stylistic variations related to the formality of the speech setting and solidarity between interactants affect the vowel, leading to vowel raising (i) in casual speech situations and (ii) in interaction with an addressee with whom the speaker is intimate; these trends are especially salient for younger speakers. In listeners’ perceptions of the vowel raising, it had distinct social meanings in female and male voices. Listeners indexed the raised variant in female voices to (i) outgoingness, (ii) lower economic class (for younger voices), and (iii) lower economic class (by male listeners). Listeners indexed the raised variant in male voices to (i) being cute (for younger and middle-aged voices) and (ii) being masculine (for younger voices). These meanings form separate indexical fields of the raised and unraised variants for male and female voices. Furthermore, some of these perceptions are reflections of social and cultural values of Korean society, and they can be correlated with the production of the vowel variants. The present study contributes to the understanding of vowel raising in Seoul Korean by investigating language-internal and language-external constraints on its production and how it is perceived. In addition, by explaining the (socio)linguistic background of this vowel variation, the findings of this dissertation should be useful in further studies in second language acquisition of Korean and language teaching.