Linguistics professor Jeffrey Health will give a talk titled "Which linguistic theory for CogSci?"

Please visit the seminar website for Zoom link and details.

ABSTRACT
The linguistic theory most familiar to cognitive scientists, the syntax-centric generativist model, has deep-seated problems for which no patches are available. Syntax-centrism alienates the model from processing and especially from speaking, and therefore from any natural cognitive processes. Its emphasis on economy now seems quaint in the context of the brain’s massive storage and computational power. The evo-devo theory that is joined to the theory’s hip makes no sense biologically. Under the microscope, current minimalism consists largely of ad hoc devices to account for language-specific linear ordering: functional projections some of which are meaningless, unnecessary specifier positions for these projections, phonological deletion due to unmotivated “computational efficiency,” and ad hoc processes like “remnant movement” when all else fails. Even with this proliferation of makeshift entities and processes, the model cannot account for basic morphosyntactic phenomena in many nonwestern languages. In this talk I discuss how a directional, speaker-centric model that stands up to crosslinguistic findings might be cobbled together from “cognitive linguistics” on the semantic end and morphophonological (including prosodic) theories on the output end. I describe some linguistic issues that are, and some that are not, amenable to experimental study and to computational modeling.