Clause-Chaining, Switch-Reference and Action/Event Continuity in Japanese Discourse

This paper investigates the clause-chaining mechanism exhibited in Japanese. Clause-chaining is characterized by a string of clauses with non-finite (or partly non-finite) verb forms followed by a clause with a finite verb. The basic property of clause-chaining devices is to conjoin two or more clauses in a (more or less) chronological order. In this paper, the discourse function of the clause-chaining devices, te, to and zero-conjunction, is interpreted in terms of the notion of action/event continuity. First, general properties of the clause-chaining devices, including various degrees of clause integration, are discussed. Then, the notion of action/event continuity in discourse is introduced, and a hypothesis that the clause-chaining devices in Japanese are coding action/event continuity or discontinuity is advanced. The hypothesis will be tested by examining continuity of referent, aspect, time, and location. Firstly, this paper studies switch-reference properties of the clause-chaining devices along with their correlations with subject marking devices (wa, ga and zero anaphora). As a corollary of the hypothesis, I offer a functional explanation of a well-known phenomenon: that generally only the subject marking device ga appears in a subordinate clause. Later, the paper will investigate the significance of aspectual coding in clause-chaining. Finally, it will investigate how referential, aspectual, temporal, and locational continuity correlate with the overall discourse coherence of action/event continuity.