Exhaustive control as movement: The case of Wolof

Abstract

This paper investigates control constructions in the Niger-Congo language Wolof, which offers several insights into the phenomenon of control. First, I show that one and the same predicate can take infinitival complements of different sizes, giving additional suport to the claims in Wurmbrand (2014c, 2015), Wurmbrand and Lohninger (2023). Next, I present arguments in favor of Grano’s (2012, 2015) claim that Exhaustive Control (EC) and Partial Control (PC) are derived via different strategies, specifically, that EC is the result of movement (Hornstein 1999 et seq.). Control in Wolof is only exhaustive, both with cross-linguistically typical EC predicates and with typical PC predicates, and, notably, all control constructions in Wolof restructure, and all control verbs are monotransitive, properties that usually characterize EC, but not PC predicates. This confirms a correlation between EC, restructuring, and monotransitivity argued for by Cinque (2004, 2006) and Grano (2012, 2015). While Cinque’s and Grano’s approaches treat EC predicates as functional verbs, I argue that this bundle of properties cannot be a simple consequence of monoclausal syntax and propose that movement of the subject from the infinitival into the matrix clause must be available in bi-clausal constructions as well, supporting the view that at least one type of control is derived via movement, and does not involve PRO. An additional argument for this claim comes from ditransitive verbs: I show that Wolof does not have object control, and attribute this property to the larger size of infinitival complements in ditransitive constructions, resulting in the subject movement into the higher clause being impeded.