Martina Martinović presented her work on control and restructuring in an invited talk at the University of Chicago on November 20th. The title and abstract are below:
Control and restructuring in Wolof
This research explores the phenomenon of control in the Niger-Congo language Wolof, which has the following properties. First, Partial Control is not possible in Wolof; all control predicates exhibit only Exhaustive Control. Second, all control predicates in Wolof exhibit restructuring properties, both those that cross-linguistically generally restructure, and those that have been argued to never restructure. I argue that these properties give support to Grano’s (2012, 2015) claim that there are two strategies for establishing control: one that results in Exhaustive Control (for Grano, following Cinque (2004), this is raising), and another that results in Partial Control (involving a PRO). I argue that Wolof only has the former strategy. While Wolof does not have direct evidence that the strategy resulting in Exhaustive Control involves raising, it does offer some indirect support for this claim. First, I show that all non-finite complements involve a reduced clausal size, and propose that this allows for raising to proceed from all such structures. Second, only predicates that do not take direct objects participate in control: subject control over an object, as well as object control, are not possible. It has long been noted that restructuring/Exhaustive Control verbs are only monotransitive verbs (Kayne 1989, Cinque 2004), which has been taken as an argument for the functional status of such verbs. I argue that lexical verbs must also be allowed to involve raising of embedded subjects. Raising to object, even if it exists (as convincingly argued in Postal 1974), appears to be cross-linguistically rare. If Exhaustive Control is raising, and raising to object is not found in Wolof, this explains the absence of object control. And finally, I show that there is no independent evidence for PRO in Wolof.