A Constraint-Based Analysis of Association with Focus in Japanese

In this paper, I present an explicit analysis of association with focus in Japanese. The proposed formal analysis, basically couched in Rooth's (1985) alternative semantics supported by the syntactic devices of HPSG, captures the behaviors of focus particles in Japanese in a fairly straightforward manner. Section 1 is devoted to a survey of data and definitions of key terms and concepts in this paper. In section 2, I present my analysis and explain in detail how it accounts for the observed phenomena. 1 Focus and focus particles in Japanese The term "focus", as it is used in the context of modern generative linguistics, typically refers to some prosodic prominence which functions as indicating certain semantic and pragmatic emphasis (Rooth, 1996, 271). Since Rooth (1985), much attention has been paid to a class of function words called focus particles (FPs), representatives of which in English include only, even and also, that interact in a certain way with the phenomenon of focus (ICOnig, 1991). Typically, an FP appears with a focused phrase which is marked by a readily detectable pitch accent. The significance of the interaction between FPs and focused phrases is that it affects the truth-conditional interpretation of the sentences. Thus, the following sentences from Rooth (1985, 2) have different truth conditions although they are composed of exactly the same string of words:1 (1) a. I only introduced BillF to Sue. b. I only introduced Bill to SueF. (la) is true when I introduced Bill and no one else to Sue. ( lb), on the other hand, is true when I introduced Bill to Sue and to no one else. Under the situation where I introduced Bill and Mary to Sue and no one else to anyone else, ( la) is false while (lb) is true. It is clear that the interpretation of the sentence is crucially dependent on what is the focused phrase. That is, the focused phrase is what the FP only is interpreted with respect to. This phenomenon is termed as "association with focus" (Jackendoff, 1972; Rooth, 1985). In (la), for example, we say that the FP only associates with the focused phrase Bill. In English, association with focus is more or less a clear-cut phenomenon. What is focused in a given sentence can be largely determined by phonologically observable evidence. In Japanese, on the other hand, the situation is somewhat obscured. Phonological stress does play a role in indicating a certain kind of emphasis in a sentence. But it is far less clear, as compared to English, whether what we call "focus particles" in Japanese (perhaps so named because they most closely resemble English ones in terms of their function and meaning) interact with something that can be independently defined by 1 In this paper, I mark the focused phrase with the subscript F. The definition of the term "focus" as used in this paper is given below.