Language and Literature has already published a number of contributions about the pros and cons of critical discourse analysis, doubtless because this is a movement that has been highly influential within stylistics (see Fairclough, 1996;Toolan, 1997;Widdowson, 1995Widdowson, , 1996. The debate seems to be ongoing, with the latest contribution by Henry Widdowson appearing side by side with articles by critical discourse analysts such as Norman Fairclough (2000) and Ruth Wodak (2000), in a prestigious volume entitled Discourse and Social Life edited by Sarangi and Coulthard (2000).Widdowson's 'Critical Practices: On Representation and the Interpretation of Text' ( 2000) is an attack on critical discourse analysis from the perspective of literary criticism, but a literary criticism that seems outdated, untouched by the developments in literary theory of the last few decades. His main point is that critical discourse analysts isolate texts from their contexts, but since this claim is merely based on two short extracts taken from Fairclough (1989) and Lee (1992), it seems a rather weak argument, as Widdowson himself (2000: 168) at least partly acknowledges in a footnote. It is further undermined if we remember how Fairclough's model right from the beginning emphasized the importance of the social conditions of production and interpretation (Fairclough, 1989: Ch. 2) and included an intertextual dimension (Fairclough, 1992: Ch. 4), and if we consider the more recent publications in critical discourse analysis such as Fairclough and Wodak (1997) or Chouliaraki and Fairclough (1999). Thus Fairclough and Wodak insist that 'discourse is historical': Discourse is not produced without context and cannot be understood without taking the context into consideration . . . utterances are only meaningful if we consider their use in a specific situation, if we understand the underlying conventions and rules, if we recognize the embedding in a certain culture and ideology, and most importantly, if we know what the discourse relates to in the past.