국어의 보충어와 부가어 판별 기준

Abstract

This paper aims at distinguishing complements and adjuncts in Korean. A complement is an element which completes the meaning of its head, while an adjunct restricts the meaning of its head as a simple modifier. In syntactic perspectives, complements are generally considered to co-occur obligatorily with their heads. The question is, however, that it is not easy to judge if an element in an actual sentence is a complement or not. So firstly, I concentrate on showing that referentiality can be a useful criterion for distinguishing complements and adjuncts. Secondly, I elucidate complements of noun. We can divide Korean nouns into several subcategories according to the meaning selection of each. Among them, a class which I call ‘individual-level predicational nouns’ can be regarded to include nouns which depend their existence (referentiality) on property nouns which again depend their existence on entities. Thirdly, I attempt to shed light on complicated aspects of verbal complements. I separate sentential-level predicates from lexical-level ones in Korean sentences, and in any case, I regard any predicate in a sentence as belonging to X0-level, since the level change does not guarantee direct projection from its lexical head. On this configurational basis, I conclude that a complement of a verb can be identified by ‘Relativization Test’. In the last part of this paper, I try to link grammatical cases in Korean to their thematic roles. The most tricky point in doing that is to impose a grammatical case to each complement of a sentential-level predicate. For this, I propose a possibility of explanation in modeled prototypical types of argument patterning in Korean.