This week, Paulina Łyskawa (MA 2015, now at the University of Maryland) is giving a virtual talk for York University. Paulina, a syntactician with an interest in heritage languages and endangered languages, recently completed her Ph.D. under the joint supervision of Masha Polinsky and Omer Preminger. Her talk, "When your grammar is not enough: Agreement with coordinate structures", will be taking place on Zoom on Friday, March 26, from 3 to 4 PM. See the email for a link.
Natural languages reflect the grammatical features of [person], [number] and [gender/noun-class] features (henceforth, phi-features); in many languages, the phi-features of a nominal phrase determine the overt morphophonology of the verb or auxiliary, thus controlling agreement morphology:
(1) Phi-feature agreement in Polish
Papugi krzyczał-y.
Parrot(3pl.fem) scream.past-3pl.fem
'The parrots were screaming.'In my work, I investigate the ways natural languages negotiate agreement with coordinate structures, e.g.:
(2) The cat (3sg) and the dog (3sg) are (3pl) barking loudly.
In a typical coordination structure, involving two nominal conjuncts, each conjunct bears its own independent set of phi-features. This constitutes a surplus of information, as far as the needs of the rest of the clause are concerned, because the verb or auxiliary does not have separate agreement slots to reflect the phi-features of each conjunct. This surplus of phi-features in a coordinate structure needs to be somehow reduced. The question I address here is whether this reduction mechanism is part of the grammar, as is usually assumed, or is instead extra-grammatical. I argue that by situating the feature reduction mechanism outside of the grammar, we have a better handle on the empirical state of affairs (in particular, the preponderance of variability when agreement is controlled by coordination), while maintaining a simpler linguistic theory overall.
No comments:
Post a Comment