Volume 34, Issue 2 -

It has been claimed (Inoue, Kazuko. 1976. Henkei Bunpō to Nihongo . [Transformation grammar and Japanese]. Tokyo: Taishūkan.; Kuno, Susumu. 1973. The structure of the Japanese language . Cambridge, MA: MIT Press.; Saito, Mamoru. 1985. Some asymmetries in Japanese and their theoretical consequences . Cambridge, MA: MIT dissertation.) that Japanese “gapless” topic constructions and relative clauses are derived by base-generation. Evidence in favor of the base-generation analysis comes from the observation that there does exist any derivational source of the “gapless” construction. Contrary to this widely accepted view, this paper argues that the Japanese “gapless” construction is derived in terms of movement enforced by labeling. Under the proposed analysis, obligatory raising of a phrase to the topic/relative head position, which is required by labeling, accounts for the fact that there does not seem to exist any derivational source of the “gapless” construction at first sight. It is shown that our movement analysis is supported by island, reconstruction, and parasitic gap facts. If our analysis is on the right track, it gives further support for the Free Merge coupled with a labeling algorithm approach (Chomsky, Noam. 2013. Problems of projection. Lingua 130. 33–49. and Chomsky, Noam. 2015. Problems of projection: Extensions. In Elisa Di Domenico, Cornelia Hamann & Simona Matteini (eds.), Structures, strategies, and beyond: Studies in honor of Adriana Belletti , 1–16. Amsterdam: John Benjamins.).

This paper develops a new analysis of the antepenultimate accent principle that determines the default location of the pitch accent in Japanese words (namely, on the syllable containing the antepenultimate mora). The chief innovation is that this analysis also applies to compounds, where it predicts the location of accent in compounds with “short” N2 (one or two moras) – so-called “preaccentuation” at the end of N1, which often does not coincide with the penultimate mora. In addition, the paper sketches an extension of the analysis subsuming the N2-initial accent characteristic of compounds with “long” N2 (three or four moras).

Back in the 1970s, Kazuko Inoue observed that some active sentences in Japanese allow a prepositional subject. Along with impersonal sentences pointed out by S.-Y. Kuroda, such examples suggest that the nominative subject is not an obligatory element in Japanese sentences. While this observation supports the hypothesis that important characteristics of the Japanese language follow from its lack of (forced-)agreement, Japanese potential sentences require the nominative ga on at least one argument. The present article argues that the nominative case particle ga is semantically vacuous even where a ga -marked phrase is indispensable or the ga -marked phrase is construed as exhaustively listing. Stative predicates require a ga -marked phrase because they can ascribe a property to an argument only by function application. The exhaustive listing reading arises by conversational implicature when the presence of a ga -marked phrase signals that a topic phrase is being avoided. The discussion leads to a semantic account of subject honorification whereby the honorification only concerns the semantic content of the predicate, and does not involve agreement with the subject. It is also shown that sentences with a prepositional subject allow zibun only as a long-distance anaphor, which indicates that they do lack a subject with the nominative Case.

This work attempts to defend the so-called Uniform Hypothesis for passives in Japanese by re-examining the empirical phenomena involving quantifier float (Q-float). We will attempt to reinforce the following main tenets of the Uniform Hypothesis — (i) that all ni -passive sentences in Japanese involve complementation and the theta-marking of their subjects, and (ii) that they crucially lack NP movement. We will also closely examine the syntactic, semantic, pragmatic and prosodic properties of Q-float in general.

In this article, which is taken almost verbatim from parts of Agreement Beyond Phi (Miyagawa, Shigeru. 2017. Agreement beyond phi . Linguistic Inquiry Monograph 75. Cambridge, MA: MIT Press.), I focus on a puzzling observation about subject pro across languages: in languages such as Japanese and those of Romance, the subject pro behaves exactly like a pronoun in being able to freely refer to entities in the discourse with reasonable context, and also to refer sentence internally to a subject, an object, or other phrases. However, in Chinese, the subject pro is extremely limited in its reference potential: it is able to refer to a discourse entity in very narrow contexts, and sentence internally, its antecedent is limited to the subject. I show that the Chinese subject pro demonstrates the principles of Strong Uniformity, by depending on ϕ-feature agreement for sentence-internal reference, and when that option isn’t taken, switches to the Topic feature to refer to a discourse entity.

This paper takes up some constraints on right dislocation in Japanese (JRD) and shows that some of them are sensitive to the presence of an overt correlate for the right-dislocated element while others are not. Adopting a commonly proposed biclausal analysis of JRD, we propose that there are two ways to derive JRD patterns. One involves movement of the right-dislocated element and the other involves no such movement. We show how these two derivations account for island sensitivity and the interpretation of nominal modifiers and adverbs in JRD. The main function of overt correlates in all these cases is to signal where dislocated elements should be interpreted in various syntactic structures. For constraints that are not affected by the presence of overt correlates, we focus on the ones related to phonological phrasing and domains to which multiple RD applies. We show that a binarity constraint can be observed, where it is required for right-dislocated elements to be grouped into two phonological phrases. We then discuss two additional constraints on multiple RD: one on the composition of elements in a phonological phrase, and the other on the domain where multiple RD takes place.