Dr. Ksenia Ershova of MIT will be giving an in-person Speaker Series talk on Thursday, October 20th, 2022. This will be at 3:30pm at ANNS 111, and will also be accessible through Zoom.

This event will be followed by a reception from 5:00-6:00pm at the Linguistics Department Library.

Below is the title and description of the talk.


Unexpected consequences of polysynthesis: deficient probes, dynamic phases and the role of C

Polysynthetic languages are notorious for morphologically complex predicates which encode agreement with several arguments at the same time. This talk explores two disparate, but unexpectedly linked phenomena in West Circassian, a polysynthetic language: (i) the behavior of deficient phi-probes in nominalizations and (ii) variable islandhood of nominal arguments.

In nominalizations, all the verbal phi-probes are structurally present, but they can only agree with phi-deficient anaphors and not fully specified nominals. Seemingly unrelated, DP islands fail to show islandhood effects in cases of long-distance wh-movement from embedded CPs. I argue that both of these puzzles can be explained by the interaction of the heads in the verbal extended projection with the clause-typing head C. In a polysynthetic language with multiple phi-probes, the probes must be licensed by C. In the absence of C, the probes are phi-deficient, as in nominalizations. A consequence of this licensing – which is established by agreement – is that it can, per Richards’ (1998) Principle of Minimal Compliance, unlock lower phases for wh-movement. The result is that nominals which are by default islands are made transparent for subextraction in contexts where licensing by C is done prior to A’-extraction – in West Circassian, this manifests itself in configurations where movement is triggered by a successive-cyclic edge feature on C, rather than a contentful wh-feature.

The proposal has consequences for our understanding of deficient probes, licensing, and locality constraints on syntactic operations. By demonstrating how two individually complex and seemingly unrelated phenomena may be accounted for with the same syntactic parameter, this talk emphasizes the importance of research programs with a long-term focus on a single language.