In the first part of this paper, assuming, with Yoon (2005), that category-neural elements like man (only), i.e. delimiters, are invisible with regard to syntactic structure building, we claim that the nominal affixes except delimiters appearing between a number affix and a Case marker form a syntactic category, which we call postpositional element (PE). The reasons are as follows. First, they scope over coordinated NPs, which suggests that they constitute a syntactic object. Second, they appear in a fixed position among nominal affixes in the order of N-Num-PE-K. Third, they have syntactic and semantic functions such as inherent or semantic case marking and categorial selection. In addition, certain complexes such as an-ey (in) (which is composed of a noun + a postpositional element), wiha-ye (for) (which is composed of a verb + a verbal particle) or kath-i (like) (which is composed of an adjective +an adjectival particle), must not be considered as a postpositional element because the noun, the verb, and the adjective in these complexes, show their nominal, verbal or adjectival properties. Therefore postpositional elements in Korean have to be restricted to the nominal affixes appearing between a number affix and a Case marker. In the second part, we observe the following three characteristics of postpositional elements, thereby showing that they are different from those of prepositions in other languages such as English or French. First, a demonstrative phrase like yeki i (here this) can scope over a postpositional element phrase (PEP). Second, structural Case markers can appear after a postpositional element. Third, syntactic or semantic features of a noun whose projection NP is a complement of PE, percolate through the PEP. Finally, we show that all these characteristics can be accounted for by Hong's (2006) hypothesis that PEP in Korean is a nominal extended projection à la Grimshaw (1991).
(0)
(0)