Predicaments of Criticism

Abstract

In this article, some of the pitfalls of critical analyses in socially relevant forms of linguistic research or discourse analysis are spelled out. Though the point of reference is Norman Fairclough’s version of critical discourse analysis, the observations can easily be extended to various related types of work. The argument is made that, in order for critical studies of discourse to be convincing – and hence to have any impact at all – the methodological link between empirical data and conclusions should be kept clear. Some basic guidelines, based on a general theory of linguistic pragmatics, are hinted at in the course of the critical reflections: e.g. whatever can be observed on the basis of formal analysis should never be ignored, and form-function relationships should never be regarded as stable.