Functionalism and the Metalanguage - Theory Confusion

It is often said that many functional linguists lack a ‘theory’. Within formal linguistics, it is common to talk of different ‘theories’, such as Government-Binding Theory or Head-Driven Phrase Structure Grammar. When formal linguists look at the work of functional linguists like Givón (1984, 1990, 1995), they do not see a ‘theory’ in the sense of formal theories like Government-Binding Theory, and they conclude that such work lacks a ‘theory’ and that it is therefore ‘atheoretical’. What they find lacking in much functional work is a proposed metalanguage in which languages are analysed or described, a metalanguage in which representations of structure and rules are stated. In this paper, I argue that this view of much functionalist work as ‘atheoretical’ represents a confusion surrounding the notion of ‘theory’, that the term ‘theory’ is being used here in a sense that is in many respects at odds with the way the term is used in other sciences, that this notion of ‘theory’ depends itself on certain theoretical assumptions about language that derive historically from views of Chomsky that he has maintained throughout his writings from 1957 on but which are widely rejected by functional linguists, and that under theoretical views shared by many functional linguists, theories of the sort assumed by many formal linguists are neither necessary nor sufficient for constructing a linguistic theory.