ABSTRACT
The views of leading advocates of Lexical-Functional Grammar (henceforth LFG), namely Joan Bresnan and her associates, can perhaps best be characterised as essentially Chomsky an , subject to an expansion of the domain of enquiry and a corresponding modification of the apparatus of description. In the first place, it is argued that the competence hypothesis is desirable in principle, on the grounds that, properly formulated, a linguistic grammar could provide a unified theory of the grammatical
knowledge assumed to underlie verbal behaviour. In other words, the fact that linguistically motivated grammars of the type developed by Chomsky have not been very successfully incorporated in psychologists' models of acquisition, comprehension or production is taken to be an indication not that the competence hypothesis itself is wrong but that Chomsky's notion of the structure of competence is wrong because based on too narrow a domain of evidence. Once again, therefore, the need to develop a linguistically well-motivated theory of grammar whose principles do not conflict with the results of psycholinguistic research is taken to be the major objective for theoretical linguistics. But such a theory, once developed, is assumed to have a clear claim to psychological reality, and this contrasts sharply with the position adopted by advocates of GPSG.
