Verb-echo answers in Japanese do not call for syntactic head movement: Arguments for a pragmatic account*

Abstract

This paper tackles the issue of whether syntactic head movement exists in Japanese. Sato & Hayashi (2018) and Sato & Maeda (2021) propose that Verb-Echo Answers (VEAs), an instance of fragment answers, in Japanese are derived via the so-called Verb-stranding TP-Ellipsis (VTPE; i.e., TP-ellipsis accompanied by verb-raising to C), thereby claiming that head movement exists in Japanese as a syntactic operation. In response, this paper argues that pro and Argument Ellipsis (AE) in Japanese sufficiently account for the key observations presented in their works. Specifically, a careful examination of the discourse in each question-answer pair reveals that the seemingly problematic scope patterns in VEAs do not call for the VTPE analysis. We also show that the unacceptability of voice mismatches in VEAs can be explained by a discourse-based analysis within the Question Under Discussion framework. Further, we provide an extensive discussion on the alleged evidence against the pro/AE analysis concerning adjunct-inclusive readings. We show that negative scope reversal effects, which Sato & Maeda (2021) argue occur in VTPE, do not occur between adjuncts and negation in the novel data. Given this, we discuss possible ways to account for the availability of adjunct-inclusive readings in VEAs with no recourse to VTPE, and suggest avenues for future research. The proposed analyses of VEAs shed new light on intriguing aspects of ellipsis phenomena, which involve complex interactions between syntax, semantics, and pragmatics.