A STUDY ON DETERMINERS

Abstract

This is a study on determiners including articles. The theoretical framework for this study is generative-transformational theory of grammar. In this study, the constituent analysis, the feature analysis, the constituent-feature analysis and the transformational analysis on determiners have been examined. The basic assumption of the constituent analysis is that determiners should be represented in terms of the constituent in the deep structure, and the relevant lexical items are inserted by the first lexical look-up. This theory shows a number of shortcomings, especially it goes against generative grammar's foremost assumption of linguistic universality. Presenting counter-arguments against the constituent analysis, Postal suggests that determiners be represented in the deep structure as syntactic features on the head noun and derived into their surface forms through Segmentalization. The Segmentalization rule makes the analysis closer to the linguistic universality. Stockwell, Schachter and Partee claim that determiners should be analyzed in terms of both constituents and features. Their analysis of the determiner in terms of the constituent-feature analysis is not fully justified. The transformational analysis was suggested by Thonpson, and this analysis roughly says that the definite article should be transformationally derived. She claims that the full analysis of the definite article must involve features of the discourses beyond sentences. Karttunen formulates three categories dependent on semantic component in order to interpretate the use of nonlinguistically anaphoric definite articles. In conclusion, it has been found that the transformational analysis incorporating the feature analysis will be the adequate theory of determiners in the future.;本 論文은 變形生成 文法에 立脚한 英語의 限定詞(冠詞包含)에 대한 硏究이다. 本 硏究에서는 the Constituent Analysis, the Feature Analysis, the Constituent-Feature Analysis 및 the Transformational Analysis에 대한 先行 硏究mf 檢討했다. The Constituent Analysis에서는 限定詞가 深層構造(deep structure)에서 constituent로 나타나야 하며 적당한 lexical item이 first lexical look-up에 의하여 insert 된다고 보았다. 이 理論은 특히 많은 단점을 내포하고 있으나 특히 linguistic universality에 어긋 난다는 것을 나타내었다. Postal은 限定詞가 深層構造(deep structure)에서 head noun의 syntactic feature로 나타내고 Segmentalization을 통하여 表層構造(surface structure)로 유도 된다고 보았다. 이 Segmentalization rule은 feature Analysis가 linguistic universality에 한 걸음 앞선 理論이라는 것을 나타내 준다. Stockwell, Schachter와 Partee는 限定詞는 constituent와 feature 양쪽 다에 의하여 분석되어야 한다고 주장하공 있으나 이 constituent-feature analysis도 역시 충분하게 限定詞를 설명 하지 못하고 있다는 것이 드러났다. Thompson이 주장한 transformational analysis에서는 정관사(definite article)는 transformationally derived 되어야 한다고 보고 있다. 그러나 정관사(definite article)의 完壁한 說明을 위해서는 sentence boundary를 넘어서야 된다는 것이 나타났다. 本 論文에서는 the Transformational Analysis에 立脚한 the Feature Analysis가 가장 타당함을 보여준다. 그리고 限定詞에 대한 硏究가 變形文法 一般理論에 깊이 연관됨을 알 수 있다.