Distributional Differences Between Non-finite and Finite Complement Clauses: The Case of the English TIME-Subject Construction

Abstract

Abstract In his target article, Günter Rohdenburg aims to refute the general hypothesis that non-finite complements cannot be less restricted in their distribution than finite ones. Among the grammatical contexts that he adduces in pursuit of this aim is the TIME-subject construction, illustrated by sentences like The Kennedy years saw six military coups overthrow popular regimes, which may more canonically be paraphrased as The world saw six military coups overthrow popular regimes during the Kennedy years. This construction allows non-finite clausal complements (small clauses and present or past participle clauses), but not finite ones. In our response paper, we investigate if, and to what extent, this presents a genuine exception to the general hypothesis. We show that there are several TIME-subject constructions forming a family of related, but structurally and functionally different constructions in the sense of construction grammar and argue that the occurrence of non-finite clauses in these constructions is an extension (in the-conceptual, and likely in the historical sense) from nominal and/or prepositional complements. This provides a motivation for their presence in the construction as well as for the absence of finite complement clauses. We also show that there are occasional extensions to finite complement clauses, pointing to the possibility that finite complementation may develop as a regular feature of these constructions in the future, bringing them in line with the general hypothesis.