CAREER: Integration of Linguistic Knowledge and Language Processing

University Of Delaware, Newark DE

Investigators

Abstract

Work in linguistic theory typically characterizes linguistic knowledge as a system of static mental representations. The main goal of psycholinguists, on the other hand, has been to provide a more or less explicitly procedural characterization of how speakers perform linguistic tasks such as comprehension or production. These different emphases are reflected in the widespread view of linguistic knowledge and language processing as the products of independent systems of the mind-brain: a declarative grammar, and procedural systems dedicated to specialized tasks. This division is fostered by the typically independent training which students receive in linguistics and psycholinguistics. This project investigates the hypothesis that the grammatical component of linguistic knowledge and language processing are products of the same incremental procedural system. If true, this simplifies the architecture of language and closes the gap between linguistics and psycholinguistics. The project also develops a program of training for graduate and undergraduate students which provides students with the skills required for active involvement in an integrated (psycho-)linguistics. The research component of the project has two main foci. In syntactic theory, the project will extend the coverage of a model of incremental left-to-right grammar, in which grammatical structures are assembled in the same order as sentences are comprehended and produced. The focus will be on constituency in head-final languages and on the phonology-syntax interface. The remainder of the research component investigates the extent to which human sentence processing is fully incremental, and the syntactic search mechanisms that make this possible, using mainly reading-time measures. A cross-linguistic study of pronoun interpretation will compare the immediacy with which constraints on anaphora apply in verb-initial and verb-final constructions, using a probe-recognition measure. Finally, a pair of studies using high-density ERP recordings will investigate the extent to which well-known electrophysiological measures of the disruption of syntactic parsing reflect the syntactic search processes underlying normal, successful parsing. Complementing the research program, the project will develop an integrated program of training in theoretical and experimental linguistics, which emphasizes an active learning approach. Graduate and advanced undergraduate students will study experimental linguistics in laboratory-based course sequence, which includes training in experimental design and analysis which focuses on the specific needs of linguists. Beginning undergraduate students in large enrollment introductory courses will use linguistics as a vehicle for developing skills in active research and scientific theory testing, substantially through the use of instructional technology resources.