#### Contact information
Office: 214 Oxley Hall (the
building on the SE corner of 12th and Neil Avenue)
Office phone \#: 292-2217
Office hours are TuTh 11:30--12:30, but if for some reason you can't
make these hours and need to talk with me, we can arrange another time
- Empirically based investigations of type-logical categorial
grammar (TLCG) using a combined Abstract Categorial Grammar/Lambek
calculus foundation with independent lambda-calculi for both
prosodic and sementic proof terms and an innovative hybrid
implicational logic (Kubota & Levine 2012; 2013a,b; 2015a,b,c). We
present an application of this hybrid logic to yield a maximally
simple treatment of Gapping as like-category coordination (Kubota &
Levine 2015a), as part of a full-scale unitary treatment of all
major 'non-constituent' coordination phenomena (Kubota &
Levine 2015b; other work on the syntax/semantics interface carried out in our
research program appears in (Kubota & Levine 2015c).
- Development of a wide-ranging approach to ellipsis phenomena
employing Hybrid TLCG along the lines of Kubota & Levine (to appear,
*Linguistic Inquiry* ), and its interaction with the syntax and
semantics of extraction.
- Exploration of the role of proof-theoretic semantics, in
particular, Dependent Type Semantics along the lines of Ranta and
Bekki, in providing an empirically adequate formal framework for the
interpretive component of type-logical proof terms; Kubota & Levine
2013b illustrates the application of a particular
dependent type formalism to the notorious 'Geach scope parallelism'
problem posed by Right Node Raising (and other coordination phenomena).
- Syntactic conditions on polarity item distribution;
ramifications of the apparently anomalous entanglement of NPI and
auxiliary properties in the distribution of modal *need*, described
in Levine 2013a.
- Fundamental issues in grammatical theory; formal foundations and
well-foundedness of hybrid type logics incorporating both lambek-
and lambda-implicational connectives and the possible embedding of
such logics in first-order linear logic with only a single mode of linear
implication, by means of a rich phenoterm subtyping, as per ongoing
work by Chris Worth.