published_or_final_version / English / Master / Master of Arts

Photocopy of typescript. / Digitized by Kansas Correctional Industries

It is a central claim of modern linguistic theory that linguists do not prescribe, but describe language as it is, without pronouncing on correctness or judging one variety better than another. This attempt to exclude evaluation is motivated by a desire to be ' politically correct', which hinders objective analysis of language, and by an ill-advised imitation of the natural sciences, which obstructs the discipline's progress towards becoming a science in its own right. It involves linguists, as users of a valued variety, in self-deception and disingenuousness, distances them from the concerns of the ordinary language user, and betrays a failure to understand the involvement of social values in language, the nature of language itself, and the limits of linguistic science. On a wider scale, linguistics reflects society's devaluing and mechanisation of language. Despite growing concern expressed in the literature, and the incoherence that becomes apparent when linguists attempt to address social problems using a theory that regards language as an autonomous object, newcomers to the discipline continue to be taught that anti-prescriptivism is the natural corollary of a scientific approach to language. This thesis suggests that the way out of these difficulties is to rethink the meaning of ' theory' in linguistics. If we take the reflexivity of language seriously, building on M.A.K. Halliday's notion of 'linguistics as metaphor', we are reminded that a linguistic theory is made of language. Metalanguage must use the experiential and interpersonal meaning-making resources of everyday language. It follows that a linguistic theory cannot escape being evaluative, because evaluation is an inherent part of interpersonal meaning. If we fail to notice our own metalinguistic evaluation, this is because language disguises its evaluative meanings, or perhaps we are just not used to thinking of them as part of the grammar. To achieve clarity about the involvement of value in language, we need to turn our metalanguage back on itself - 'using the grammar to think with about the grammar' . Some ways of doing this are demonstrated here, turning the resources of systemic functional linguistics on linguists' own language. The circularity of this process should be seen not as a drawback but as a salutary reminder that linguistics is an interpretive rather than a discovery process. This knowledge should help us revalue language and make a place for evaluation in linguistic theory, paving the way for a socially responsible and productive linguistics.