Toward a Psycholinguistically Plausible Parser

Child language researchers understand that a successful theory of language acquisition must be firmly rooted in linguistic theory. However, it is not clear that current competence-based grammars can be used to provide accurate characterizations of the learning process. These formalist approaches to language structure have the advantage of being carefully stated and fully generative. Their disadvantage is that they tend to force the language acquisition researcher into making a series of nativist assumptions that are seldom supported by the empirical data. The major alternative to formalist theory is functionalist theory (Dik, 1978; Foley and Van Valin, 1984; Givon,1979). A strong point of functionalism is its emphasis upon the predictability and reasonableness of grammatical markings. However, a major weakness of current functionalist theories is the fact that they are not stated in a way that leads to concrete predictions about language processing. The goal of the present work is the formulation of a process-oriented functionally-based system that can serve as the basis for models of language comprehension, production, and acquisition. Stated more succintly, the goal of this research is the formulation of a functionalist process model. Until such a model is fully formulated, the current controversies between nativism and empiricism and between formalism and functionalism seem largely premature. After a full process model has been fully formulated, it may well turn out that some of the issues currently being debated will be supplanted by newer more detailed issues.