Doubling in a staged architecture of realization
In this talk, I look at verb doubling phenomena and how they bear on questions concerning the architecture of post-syntactic realization, particularly the timing of linearization, prosodification, and verb doubling. In addition to widely discussed cases of verb doubling that arise from movement (with more than one copy of a moved verb being pronounced), I consider cases of verb doubling that appear instead to be motivated entirely by the need of an otherwise-unsupported clitic for a host, with examples from Breton (Celtic), Hakka (Sino-Tibetan), Ingush (Nakh-Dagestanian), and Malayalam (Dravidian). Based on the profile of clitic-triggered doubling, I make two related arguments. First, the profile of verb doubling is best accounted for in a model of linearization that involves a constraint-based evaluation of candidates, rather than a deterministic linearization algorithm. Second, linearization proceeds in parallel with the mapping from syntactic hierarchy to prosodic structure, but crucially prior to Vocabulary Insertion, with doubling arising as a trade-off between optimal linearization and prosodic well-formedness. This architecture has potential implications for certain cases of multiple exponence in morphology, for the somewhat different typology of multiple copy realization as a consequence of nominal movement, and for broader questions concerning whether syntactic structure is visible to various proposed post-syntactic operations, including Vocabulary Insertion itself.
The zoom link: https://yale.zoom.us/j/97771780410