Movement and feature-checking in Korean : relative clauses, topicalisation and case-marking

• Main Author: Kim, Yong-guk
• Published: University College London (University of London) 1997
• Subjects: 410; Linguistics
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.267001

The purpose of this thesis is to consider the following phenomena in Korean, within the framework of the minimalist program (Chomsky 1995): (i) the movement of non-restrictive adnominal modifiers, (ii) topicalization, and (iii) Double Nominative Constructions (DNCs). First of all, following in essence Cinque (1992), I propose that there is a functional category, Agreement Phrase (AgrP) whose specifier position is occupied by the pre-nominal modifiers. I argue for the existence of non-restrictive adnominal modifiers (Relative Clauses (RCs) and pre-nominal adjectives) which move overtly out of the scope of the Determiner in head-final languages like Korean. I claim that the RC or the attributive adjective is base-generated in [Spec, AgrP] due to agreement features (honorific and plural in the case of Korean and Japanese). A restrictive adnominal modifier remains in [Spec, AgrP] due to a FOCUS feature. A non-restrictive modifier, having a NON-FOCUS feature moves to [Spec, DP] whose head D⁰ has a NON-FOCUS feature, to check its NON-FOCUS feature. Secondly, I attempt to unify two contradictory accounts (non-movement or movement) in topicalization in Korean within the minimalist program (Chomsky 1995). Thirdly, it is my argument that, following much of the literature on this topic, there are three kinds of DNCs in Korean and that the three types of double nominative constructions are derived from a single underlying construction, i. e. the locative construction. The first NP marked Nominative moves to [Spec, AgrsP], to check its Case feature by the corresponding Case feature in the head of Agrs, while the second NP in DNCs, which originates as the object of the verb, remains inside VP and has its inherent case feature checked by the verb without moving. In addition, I show that the derivation in the DNCs is the same as that found in English Genitive, Existential and Locative sentences. In connection with DNCs, I claim that in Double Accusative Constructions (DACs) the first NP and the second NP are base-generated independently in different positions from each other, just like in DNCs, but that the second NP in DACs is structurally case-marked in [Spec, Agr₀P], unlike the second NP in DNCs.