Phonetics and Phonology Interplay in Loanword Adaptation: English Alveolar Fricative into Korean.

Abstract

The purpose of this investigation is to determine the general principles that govern Korean-speaking listeners’ perceptual strategies regarding the adaptation of English alveolar fricatives into Korean. The English voiceless alveolar fricative is variously borrowed into Korean as /s*/ (as in ‘say’ [s*ei]) or /s/ (as in ‘stay’ [sɨtei]). The systematic pattern for English onset /s/ is that prevocalic /s/E are borrowed as /s*/K, and preconsonantal /sC/E is borrowed as /s/K. In investigating English /s/-initial loanwords in Korean, this dissertation postulated two hypothesis, Hypothesis 1, the consonantal information hypothesis, and Hypothesis 2, the vocalic information hypothesis. Hypothesis 1 predicted that /s/-internal properties (i.e., frication portion itself) play an important role in Korean listeners’ English /s/ perception; Hypothesis 2 states that Korean listeners predominantly rely on /s/- external properties, particularly properties of the following vowel. Hypothesis 2 was postulated to apply to vowels that are acoustically present as well as vowels that are perceptually epenthesized. Four production and three perception experiments were conducted to investigate aspects of these hypotheses. The experiments confirmed that Korean listeners were insensitive to consonantal information in their perception of the English fricative, in spite of temporal and spectral differences between English /s/s in prevocalic and preconsonantal positions. Korean listeners have difficulty discriminating /sC/from/sɨC̯/. This difficulty is interpreted as an indication of a perceptual illusion, which in turn leads to speculate that the illusory vowel influences fricative perception. Phonotactic resolution (i.e., vowel epenthesis) is selectively applied in order for the epenthetic vowel to be perceptually least salient in favor of minimal perceptual consequences of an illusory epenthetic vowel. These experimental results serve as information about listeners’ perceptual modification strategies in facing phonotactically illegal sequences. They also suggest that phonetic and phonological knowledge about Korean interact in loanword adaptation, and that listeners incorporate this knowledge when they perceive non-native speech sounds.