음운탈락과 관련된 몇 문제 = Some issues on deletion in phonology

다국어 초록 (Multilingual Abstract)

This paper aims to study on some issues on deletion in Korean Phonology, considering haplology, deletion constraint, and the underlying forms of suffix, -ini, -imy??n. Firstly, The deleted vowel in haplology is not the one in suffix but in stem. The tone in dialect can play an important role to explain this issue. Comparing an(H)＋kanda(LL)→an-kanda(LHL, don`t go) with an(H)＋katt`a(LL)→an-katt`a(HLL, didn`t go), I discover that the second syllables have different tones made by different ways; whereas the high tone of an-kanda(LHL) is come from the stem vowel, the low tone of an-katt`s(HLL) is from the suffix vowel. Secondly, deletion rule cannot be applied more than once in series of phonological process. Because a well-formed hallin(from the underlying form, halt(to suck)+nin) cannot be derived through /r/-deletion followed by cluster simplification, there must be the general constraint on deletion rule. On the ground of this deletion constraint, I suggest that the suffixes, -ini, -imy??n have two types of underlying forms: X and iX.