Japanese downstep revisited

Abstract

This paper presents the results of a production experiment in which downstep in Tokyo Japanese was re-examined. There are three major assumptions that have been widely adopted in the literature: (1) the Major Phrase (MaP) is the domain of downstep; (2) a syntactic boundary blocks downstep, as a result of the insertion of a MaP boundary; and (3) focus blocks downstep, as a result of the insertion of a MaP (left) boundary. The results of the experiment raise questions about these basic assumptions, and call for new theoretical explanations of the data. There are two major findings in the results: (i) no complete register resetting by focus of a syntactic boundary, and (ii) phonetic differences between the effect of focus and that of syntactic boundary. The first finding raises questions as to whether Assumptions 1 and 2 should be maintained, and if so, how they should be modified to capture the results. Recursive prosodic phrasing along the lines of Ito and Mester (2007, 2012, 2013) is adopted to account for the incomplete resetting. The second finding particularly casts doubt on Assumption 3, because the focus effect lacks some of the properties of the boundary effect. The difference between focus and boundary needs to be explained by assuming that the focus effect is independent of MaP-phrasing, as proposed in Ishihara (2011b).