this paper argues that the Korean /n/-insertion before glide [j] is a consequence of feature [nasal] insertion to avoid a heterornorphemic consonant and glide [j] sequence in syllavle initial position. The /n/-insertion is not to remedy a hetrosyllbic...
this paper argues that the Korean /n/-insertion before glide [j] is a consequence of feature [nasal] insertion to avoid a heterornorphemic consonant and glide [j] sequence in syllavle initial position. The /n/-insertion is not to remedy a hetrosyllbic C.Y sequence, as asserted in previous approaches (Kim-Renaud 1974, Ahn 1985, Park 2005, Lee 2004, Oh 2006, Lee and Lee 2007, etc.). I argue that the /n/-insertion is irrelevant to the Syllavle Contact Law, by which two heterosyllabic vonsonants should go from high to low in sonority. I rather claim that the Korean /n/-insertion occurs to avoid tautosyllabic and heteromorphemic CY sequences, which are not permitted in Korean. Also, I claim that the /n/-insertion is not a segment insertion but a feature insertion. The insertion of feature [nasal]is the most faithful way to fix the marked heteromorphemic CY structure: a feature insertion is more faithful than any segment insertion; feature-filling is more faithful than feature-changing. Also, the output form of the feature [nasal] insertion performs double functions; it makes a morpheme boundary coincide with a syllable boundary and breaks the illegitimate sequence. Thus, the [nasal] insertion is the most optimal way to prevent the illegitimate sequence heteromorphemic CY in Korean.