Professor of Writing and Linguistics
Georgia Southern University, United States of America
A grammar that is not intimately connected to communication is highly artificial and abstracted from anything that could describe or produce human language behavior (Pennington, 2002). Yet in both (i) the traditional models of language that inform the grammars which most language teachers and students are familiar with and (ii) the formal models that inform the grammars of mainstream linguistics, the facts of performance and interaction are commonly mentioned only in passing and at worst relegated to a realm which is entirely outside that of grammar or language itself. Since the reason for language to exist is communication with other human beings, a grammar which does not start from the social features of communication, on the face of it has low validity as a model of language. Formal syntax, in being referenced to abstract, unverifiable cognitive structure (e.g. Chomsky, 1995), is far removed from performed, communicative language and so cannot in any obvious way serve as a model of it. If one argues, as formal linguists do, that the reality of language is its mental existence, e.g. in cognitive categories and abilities to make grammaticality judgments, one must nevertheless ultimately show how this internal language relates to external, performed language. In spite of its postulation of interfaces with surface forms, formal linguistics, after more than a half century of development, is still very far from making this connection, which requires linkages of speaker, hearer, and context.