초록 열기/닫기 버튼
Many Studies have been done to explain the Case alternation in the Korean long-form negation construction (hereafter, LNC). In Yang (1995, 1996) and K. Lee (2002), 'anh-' is reanalyzed into 'ani ha-' and 'ha-' as a head is supposed to have property to license an accusative case to '-ci' in the LNC optionally. However, some problems occur with their analyses, due to the optional property of 'ha-'. In order to overcome the problem, Choi (2006) suggests base-generated X0-adjunction structures for the LNC in which 'ha-' is treated not as a head but a part of 'anh-'. Under the structures and Chomsky's (2000, 2001) Agree system, he explains the Case alternation in the LNC, using Case minimality and c-command suggested by Kayne (1994) and postulating 'anh-' as an accusative licensor. However, there seems to be a problem in his explanation of the Case alternation in the LNC through 'minimality'. In this paper, we take Choi's structures for the LNC, his postulation that 'anh-' can be a case licensor and the concept of the phase suggested by Chomsky (2005) and show an alternative for the Case alternation in the LNC. The LNC containing a transitive verb or an unergative verb as its predicate has two phases, namely, v*P and CP while the LNC containing an adjective or an ergative verb as its predicate has only one phase, namely, CP. In the former case, only accusative Case can be licensed within the v*P phase because the phase is the domain of accusative Case and 'anh-' is an accusative Case licensor. On the other hand, in the latter case, both nominative and accusative Cases can be licensed within the CP phase because the phase is the domain of nominative Case and 'anh-' is the accusative Case licensor.
