ABSTRACT

This paper focuses on the syntactic status of three elements which can be shown to have strikingly similar distributions in Chinese, Japanese and Korean, the morphemes de, no and kes, and attempts to see how both the similarities and the differences in the distribution of these elements can be given a unified account. The suggestion is made that the similarities and the differences in the patterns found can be systematically explained if it is assumed that the three elements are all essentially of the same type in origin but synchronically in different stages of grammaticalization and syntactic change. The paper also introduces and argues for a formal movement-based model of grammaticalization and shows how such an approach is able to capture the apparently gradual nature of syntactic change in a possibly more precise way than traditional feature-based cline descriptions. Applied here to the delnolkes paradigm, the approach allows for a new understanding of the development of formal nouns and functional categories in the NP/DP, and also results in an interesting prediction concerning the existence of double nominalizers which appears to be borne out by various cross-linguistic patterns.