Domain extension: a study of restructuring predicates in Japanese and the role of head movement

要旨

This thesis offers an account of how an opaque domain becomes transparent by making use of head movement. The study covers three types of Japanese predicates that generally show clause-union effects called 'restructuring' where the apparent clause boundary disappears and more than one clause acts as a unit. I argue that whether restructuring is obtained or not depends on whether a verb head of the embedded clause moves to the matrix clause to extend the domain, building on Baker's (1988) and den Dikken's (2007) work.The three predicates examined are (i) the motion verbs ik 'go' and ku 'come' in V-ni-ik/ku 'go/come to do V' forms, (ii) the causative morpheme (s)ase, and (iii) the predicate mi 'try' in V-te-mi 'try Ving' forms. I first show that these predicates exhibit restructuring properties and behave similarly on the surface. I then show that the motion verbs behave differently depending on where they license the object. Focusing on long-distance nominative Case assignment of the object, a restructuring property, I propose that two structures are available for the nominative object under sase and mi, whereas only one structure is available for the object under the motion verbs. This structural difference is based on scope facts of the nominative object under each predicate. I argue that structural optionality arises when verbal head movement occurs, and that the intervening morpheme ni under the motion verbs blocks such head movement, whereas te under mi and the null morpheme under sase do not. I thus show head movement to be crucial in calculating the domain for Case assignment, and as such, to be crucial in determining the domain for scope of the nominative object; a correlation that has not been claimed in the literature. The thesis consequently supports the idea that head movement is a syntactic operation and not a PF operation by showing that head movement determines a domain that must be calculated for the purpose of syntactic relations. It also reveals whether Japanese employs head movement depends on a type of predicate, more specifically, a type of morpheme that the predicate requires. The study provides an insight into how morphology and syntax interact, and in turn how this affects domain transparency.

主題