국어의 변(邊)자음

Abstract

Traditionally, consonants are classified in terms of place and manner of articulation. In this article, it is claimed that a third parameter called mode of articulation is needed in some languages. In Korean the two modes are the mode of occlusive articulation and the mode of centrifugality. The two modes are in a tangential relationship, the latter mode deriving from the former mode. Thus, as the central region in the oral tract is clogged, articulation is pushed toward the peripheral regions, that is, labial and velar areas. In places of articulation then, Korean consonants are divided into two classes: central and peripheral. In this paper, several phonological phenomena showing the preference of the peripheral articulation are presented. (1) In progressive assimilation, central consonants assimilate to peripheral consonants, but not vice versa (e.g., 준비→줌비, but 삼도→*산도). (2) In umlaut, the intervening consonant must be a peripheral one (e.g., 어미→에미, but 언니→*엔니). (3) epenthetic consonants in hiatus and additive consonants in texting are usually peripheral consonants (e.g., 소아지→송아지, 하네요→하네욤, 아닌데→아닌뎁. (4) In consonant cluster simplification, of two coda consonants, it is usually the peripheral consonant that survives and the central consonant that deletes (e.g., 없다→업다, *엇다, but 젊다→점다, *절다). (5) In Chinese characters whose coda is ㄹ, the following central consonant is tensified, but not if it is a peripheral consonant (e.g., 일도(一度)→[일또] [ilt'o]; 일장(一章) [일짱] [ilʧ‘aŋ], *[ilʤaŋ], but 일과(日課)[ilgwa], *[ilk'wa]; 일분(一分) [ilbun], *[ilp'un]. Why this is the case is not yet known.