earticle

영어

In this paper, I discuss syntactic licensing condition of bakk-e-anh-da (hereafter, bakk-e) ‘only’ in Korean and shika-nai (hereafter, shika) ‘only’ in Japanese. Arguments of this paper are as followed:(1) Unlike previous studies, syntactic licensing condition of bakk-e and shika is not the same. (2) shika is directly licensed by Neg. (3) bakk-e is a kind of multifunctional expression unlike shika. In other words, there exist two different usages in bakk-e. Usage of bakk-e1 corresponds to one of shika while usage of bakk-e2 does to one of igai-nai or hoka-nai. Moreover, syntactic licensing condition of bakk-e1 and bakk-e2 is different. Namely, bakk-e1 is directly licensed by Neg, unlike bakk-e2. For bakk-e2, it is not bakk-e2 that is licensed by Neg but covert Indeterminate-do such as amu-do ‘anyone’, amugeos-do ‘anything’, amude-do ‘anywhere’. In previous studies, there are two different approaches to syntactic licensing condition of bakk-e and shika. First approach suggests that bakk-e and shika are directly licensed by Neg. Second approach proposes that it is not bakk-e and shika that are licensed by Neg but covert Indeterminate-do for bakk-e and covert Indeterminate-mo such as dare-mo ‘anyone’, nani-mo ‘anything’, dokoni-mo ‘anywhere’ for shika. However, there are some problems for each approach. First of all, first approach cannot explain why bakk-e is able to appear in adversative predicate, whereas shika is not able to. Second of all, second approach cannot explain when bakk-e or shika is in the non-argument positions such as adjunct positions. In other words, the approach can explain only when bakk-e or shika is in the argument positions. Therefore, this paper suggests that the arguments of previous studies should be modified.