Stray Erasure and the Coda Condition in Korean
Abstract
Stray Erasure and the Coda Condition in Korean* KIM Sun-hee The Korean consonant system consists of fifteen obstruents and four sonorants. Among the consonants, only seven, namely p, t, k, m, n, g and 1 appear at the end of a syllable, whereas all consonants except g can appear at the beginning of the syllable. But these seven consonants are not the only ones that appear at the final position in the underlying representation of a morpheme, for example, /pakV 'outside' or /os/ 'clothes'. Furthermore, morpheme-final consonant clusters also occur in the underlying representations, for example /yatalp/ 'eight'. I show in this paper that these two phenomena, traditionally accounted for independently by separate rules, are both the result of the same phonological process: Stray Erasure, a universal convention, in accordance with a language-specific Coda Condition within the framework of Nonlinear Phonology. Key words: nonlinear phonology; Korean; stray erasure; coda. Parmi toutes les consonnes du coréen, seules, p, t, k, m, n, g, et 1 peuvent apparaître en fin de syllabe. Or ces sept consonnes ne sont pas les seules qui puissent apparaître en position finale dans la représentation sous-jacente d'un morphème, comme /os/ 'vêtement'. Il existe, d'autre part,