On Recent Debates on the Tense Alternation Generalization: A Reply to Fujii et al. (2023)

Abstract

A recent article by Fujii et al. (2023) criticizes Akuzawa and Kubota’s (A&K; Akuzawa and Kubota 2020, 2021; Kubota and Akuzawa 2020) reassessment of a syntactic generalization about morphologically finite complement clauses known as the Tense Alternation Generalization. In this paper, we offer a response to Fujii et al. (2023) by critically examining their arguments. Our response consists of three components. First, we review A&K’s semantic analysis of finite control that dispenses with TAG, in order to provide a background for the discussion. Then, we elaborate on three empirical issues with TAG identified by A&K that remain unaddressed in Fujii et al. (2023). Finally, we show that both of the key claims of Fujii et al. (2023) fail to achieve the goal of defending TAG as a viable syntactic generalization. This leads us to the conclusion that A&K’s argument that TAG should be abandoned remains fully valid. A larger issue that emerges from this discussion pertains to the division of labor between syntax and semantics in analyzing (finite) control phenomena. The semantic proposal by A&K can be thought of as an attempt to reinterpret the core insights of the syntax-dominant approach represented by Fujii (2006) as a reflection of independently motivated underlying semantic properties.