통사 범주와 문법적 기능: 두자격법을 중심으로

Abstract

This paper presents evidence showing that Korean verbal endings are syntactic units, contrary to the traditional morphological analyses, and argues that they are not syntactic heads, but markers indicating the role of the syntactic construction to which they are attached. One piece of evidence is the asymmetry in deletion between derivational suffixes and the so-called verbal endings. The former are not deleted in any contexts, while the latter might be deleted in such appropriate contexts as coordination constructions The other piece of evidence comes from Korean phonology. Admitted that verbal endings were actually syntactic units, we could answer the question why the phonologically unconditioned tensification happens at verbal endings immediately following what are called stem; Phononlogically unconditioned tensification occurs at syntactic boundaries. The analysis proposed in this paper captures the functional parallelism bewteen nominal and verbal endings and resolves the so-called 'twucakyekpep (dual lexical category)' into syntactic category and its grammatical role. The conclusion drawn in this paper leads to a linguistically significant generalization on the configuration of syntactic construction. That is, every (local) structure is composed of a head phrase (possibly) followed by a (sequence of) marker(s).