Glides as Consonants in Korean

Abstract

This paper is to examine the underlying representation and the structural status of glides in Korean, focusing on the syllable-structure status of on-glides in the onset position. Supporting the two hypotheses : Onset Hypothesis and Coda Hypothesis, in this paper, it has been argued that: (1) Korean glides are orthographically like vowels but phonologically and phonetically like consonants in the phonetic representation; (2) in Korean syllable structure, onset and nucleus form a constituent (X), and a single glide before a vowel is like a consonant in the onset, and on-glides preceded by a consonant are part of the onset based on articulatory and acoustic evidence as well as evidence from glide formation and insertion, language games, phonotactic constraints, partial reduplication, and vowel harmony; (3) a consonant + glide sequence is treated as a consonant cluster, deleted or being simplified in casual speech; (4) a vowel + glide sequence /ɨj/ acts like nucleus plus coda in Korean because consonant clusters are not allowed in the coda on the surface, thus the off-glide cannot be followed by any tautosyllabic consonants in the coda position. Therefore, Korean glides in the underlying representation may be vowels. In the course of derivation, however, the glides are like consonants in the phonetic representation in that glides cannot occur in the nucleus of the syllables, and they do not maintain steady states acoustically. In addition, in the articulation of glides, they are produced with a construction that is greater than the corresponding vowels.