한국어화자의 영어 어말 유성저해음 실현형 분석

An Analysis of the Realizations of Word-Final Voiced Obstruents in English by Korean Speakers

It is generally known that voiced sounds are more marked than voiceless ones and that word-final position is more marked than word-initial position. Thus, it is not easy for Korean speakers, whose native language has no voiced obstruents in word-final position, to pronounce English word-final voiced obstruents. Using the phonetic transcriptions presented in Speech Accent Archive, this study analyzes the realizations of word-final voiced obstruents in English by Korean speakers. In the analysis the realizations are classified into four types: target voicing, devoicing, deletion, and vowel insertion. The results show that Korean speakers realized English word-final voiced obstruents as voiced by 27.13%, as devoiced by 52.70%, as deleted by 14.61%, and as with vowels inserted by 6.78%. Compared to the production by native speakers of English (Oh 2021), target voicing is less maintained in Korean speakers while more consonant devoicing and more vowel insertion occur in Korean speakers. Moreover, two factors of word type and morphological structure affect target voicing. Target voicing is more maintained in content words than in function words, and less maintained in the plural ending -s at the end of polymorphemic words than in the voiced obstruents at the end of monomorphemic words. Along with the results, some suggestions are presented on English pronunciation education.