Criteria of adequacy in functional linguistics

In this article, I shall put forward a set of criteria of adequacy (based on, but going considerably beyond, those proposed by Dik) which, in my view, a functional account of language should make serious attempts to fulfil. I shall argue that the overarching aim of functional linguistics should be a comprehensive theory of language, and that there are dangers in building models of grammar as ends in themselves, if these are not accompanied by equally well-developed models of the relationships between the grammar and other components of the more comprehensive model. I then briefly assess a selection of five functionalist approaches in relation to the proposed criteria of adequacy. The approaches, chosen to cover as wide a range as possible, are Functional Discourse Grammar, Role and Reference Grammar, Systemic Functional Linguistics, and finally two usage-based cognitive-functional approaches, Emergent Grammar/the interactional study of grammar at work, and Word Grammar. In conclusion, I propose that continued work towards a higher overall level of adequacy in functional linguistics would be greatly facilitated by (i) more widespread use of corpus-based data analysis, (ii) increased attention to the interfaces between different linguistic approaches, and (iii) familiarity with the work of other groups of scholars with an interest in language, in particular psycholinguists, psychologists, sociolinguists, sociologists, neurolinguists and neurologists.