Speaker: Giovanni Roversi (MIT)
Title: Bárány & Sheehan (2021): Challenges for dependent case
Time: Tuesday, May 4th, 1pm – 2pm
Abstract: In this paper, Bárány and Sheehan take issue with the idea that Dependent Case can be the only manner of assigning case we need to assume to exist in grammar. To do so, they present two empirical challenges: i) case alternations in clitics in Romance causatives, and ii) global case splits. For the first one, an old insight by Kayne (1975) was that the causee clitic was ACC if the embedded predicate was intransitive, and DAT if it was transitive; this is readily explainable in a dependent case world. However, Bárány and Sheehan show how not only DPs, but also CPs and PPs can work as case competitors, triggering DAT on the causee clitic, which is problematic for a theory that holds that only nominals should be case competitors for other nominals. Then they discuss global case splits, that is, phenomena where the case of an argument is determined not only by properties of that argument, but rather by properties of more than one argument, e.g. both the subject and the object; the empirical basis comes from Kashmiri, Sahaptin and Wampis. They go through two possible Dependent Case approaches, one based on movement and one based on a fine-grained DP-periphery, and argue that both either can’t derive the data or can but at only at a great theoretical/complexity-wise cost. Instead, they propose a Case-by-Cyclic-Agree account for these data, that they claim works better than the Dependent Case ones.