|Title:||The Morphosyntax and Morphophonology of Japanese Predicates||Add Dissertation|
|Author:||Kunio Nishiyama||Update Dissertation|
|Email:||click here to access email|
|Institution:||Cornell University, Department of Linguistics|
|Linguistic Subfield(s):||Morphology; Phonology; Syntax;|
|Abstract:||This thesis argues for a non-lexical view of the agglutination characteristic of Japanese predicates. Inflectional features are concatenated syntactically or post-syntactically, without any phonological features. Phonological features are supplied at Morphological Structure after syntax, through the mechanism of phonological insertion rules ordered by subset relations (Distributed Morphology, Halle and Marantz 1993).
After reviewing several approaches to morphology in chapter 2, I discuss in chapter 3 two seemingly distinct subclasses of Japanese adjectives. I claim that there is no syntactic (i.e., categorial) difference between them. Rather, the difference is morphological in the sense of Distributed Morphology. Since the difference can be captured only at the morphological level, the analysis provides strong support for the view of morphology as an independent level. The analysis also gives morphological evidence for Bowers' (1993) theory of predication, which claims that whenever there is predication, there is a syntactic projection associated with predication.
Japanese verbs show overt morphological alternations in transitive-intransitive pairs. Chapter 4 claims that such alternating morphemes head a syntactic projection that determines transitivity of the clause, supporting recent views on clause structure where the external argument is introduced by a phrase that selects lexical VP.
Based on the results of chapter 4, I argue in chapter 5 that V-V compounds in Japanese are formed in the syntax. More specifically, V-V compounds are cases of verb serialization of the sort widely known from Kwa languages of West Africa (Collins 1997b). The analysis provides another piece of support for the clause structure argued for in chapter 4. A natural consequence of the (post-)syntactic view of morphology of this thesis is that V-V compounds cannot be formed in the lexicon. I show that previous lexical accounts for these constructions cannot cover the full range of data in a principled way.