Constrained scrambling in CCG: A case study in Japanese

[winterkoninkje]

I just posted the slides for my NASSLLI talk on my website. As the abstract says there:

Combinatory Categorial Grammar has a number of desirable properties such as giving a natural account of ellipsis, supporting algorithms for on-line and partial parsing, and deriving semantics in synchrony with syntax. It provides these benefits in large part by abandoning traditional notions of constituency. Despite this, providing a clean account of scrambling in non-configurational languages is not much simpler than in traditional formalisms. We present a new family of λ-calculi for the semantic side of CCG. These calculi cleanly capture a number of cross-linguistic patterns, including modelling the existence of constraints on permissible phrase orderings. Specific applications of the new calculi are given for nominative, accusitive, and dative case marking in Japanese.

It was only a 30min talk, so there aren't a lot of details about the formalism itself. Hopefully those will be published later this year. Overall NASSLLI was excellent this year. I'll give a redux in a day or two, once I've had time to relax/recover.