Discourse in Late Modernity: Rethinking Critical Discourse Analysis.

Lilie Chouliaraki and Norman Fairclough. Edinburgh: Edinburgh University Press

Paperback, 1999 ISB 0 7846 1082 0; £15.95

reviewed by Michael Toolan; to appear in Journal of Sociolinguistics (2001)

From the humming keyboard of Norman Fairclough (he has more recently published an analysis of the discourse of Tony Blair’s ‘New Labour’ party) comes this carefully considered statement of the function and place of Critical Discourse Analysis (henceforth CDA) within the fields of socially-minded analyses ranging from philosophy of science to sociology. Previous CDA studies have implicitly linked it to stylistics, critical linguistics, and media analysis, and CDA has accordingly been adopted and applied by various broadly Humanities-based research communities. Here the connections being asserted are ones with social theory and social science; the broad message is that CDA is relevant to and a potential resource for social scientists. As the title’s reference to late modernity indicates, the kind of social science that CDA makes common cause with is the theorised and revisionist social science articulated by Bourdieu, Habermas, Giddens, and similar. Thus following Bourdieu they argue for a transcending or blending of both interpretivist and structuralist social science, in ‘constructivist structuralism’: “a way of seeing and researching social life as both constrained by social structures, and an active process of production which transforms social structures” (1).

This orientation is put to work in a reanalysis of various texts previously analysed by others, beginning with some that had been explored by Dorothy Smith from her feminist sociological point of view. The implicit message is that Chouliaraki and Fairclough’s CDA can say more about, or explain more fully, the discourses that others have previously grappled with. This ‘trumping’ rhetorical progression recurs in the book: the following chapter, for example, turns the CDA lens on Giddens’ narrative of late modernity as the global restructuring of capital. These early chapters are a difficult read for those not versed in the detail of the arguments of Giddens, Habermas, and others, but the authors offer many challenging and sometimes profound observations about, not so much the ‘linguistic turn’ of culture and theory, but the commodificatory turn that languages themselves have undergone. Part of this issues in the ‘marketisation’ of language and ‘mediatisation’ of politics (spectacularly ugly terms), discoursal hybridity, the fracturing of identity as social life becomes detached from tradition and the consequential ‘struggle to find a voice in discourse’, and globalisation of discursive practices. And the account of these given here is, I think, more fluent and compellingly integrated than in some of Fairclough’s earlier books: it is summative and authoritative, with conclusions that are both insightful and provocative:

Late modernity is characterised by an enhanced reflexivity (for example, in the construction of identities) which is in part linguistic reflexivity – awareness about language which is self-consciously applied in interventions to change social life (including one’s own identity). As commodities become increasingly cultural in nature they correspondingly become increasingly semiotic and linguistic, and language becomes commodified, subject to economically motivated processes of intervention and design (which entail linguistic reflexivity). (p.83)

On every page there is evidence that the authors have thought long and productively about CDA’s connections with and differences from social theories such as Habermas’ account of the rationalisation of the life-world and the uncoupling of systems from that life-world, Giddens’s account of the ‘detraditionalisation’ of society subjected to globalizing forces, and similar analyses. Perhaps inevitably there are some longeurs in the more rarified or abstract reaches of the discussion, with a concomitant stylistic falling away (in one long paragraph on p.104, evaluating Bourdieu, I found that old favourite struggle(s) used no fewer than twelve times). But overwhelmingly the survey and critique is fluent and constructive, regularly linked to specific situated examples–as in the discussion of contemporary de-socialized supermarket (and internet) shopping, or of the language of a works meeting in a textile manufacturing company, or of the rhetorical strategies and alignments in an exchange of public letters between a professor and the mayor of a Californian city over the contentious arrest of a young man during a demonstration. These exemplificatory analyses are begun with an illuminating discussion of a print advert for The Big Issue, a UK magazine set up to campaign on behalf of homeless and unemployed people: the generalizing commodification entailed in advertising sets up dissonances with the magazine’s intention that the homeless and the unemployed be seen anew as agents with dignity and an identity, and CDA helps to describe and explain how those tensions are discoursally negotiated.

Readers of this journal will perhaps find most interesting those sections of the book where the authors turn to linguistics or social linguistics more directly. This begins in chapter 6’s comparing and contrasting of Bourdieu and Bernstein, and the claim, in the course of assessing Bernstein’s work, that CDA has a crucial mediating role between systemic-functional linguistics and sociology (a role which, pace the systemic linguist Hasan, these authors do not feel is adequately articulated by extant systemic-functionalist theory). A sense of the relevance of CDA to both Bernsteinian pedagogical theory and to systemicist linguistic theory is particularly clear in such remarks as the following:

CDA of a communicative interaction sets out to show that the semiotic and linguistic features of the interaction are systematically connected with what is going on socially, and what is going on socially is indeed going on partly or wholly semiotically or linguistically. Put differently, CDA systematically charts relations of transformation between the symbolic and the non-symbolic, between discourse and the non-discursive. (p.113)

By way of drawing on the theoretical strengths of both Bourdieu and Bernstein, the authors propose that CDA takes the risk of incorporating the former’s notion of ‘field’ (related to the developed CDA concept of ‘order of discourse’) and the latter’s construct of ‘voice’; the chapter concludes with the CDA theoretical framework thus enriched, but a putting to the test of that framework must be sought elsewhere.

While chapter 7 explores versions of the openness and contingency, or anti-systemic-ness, of late modern life (something of a review-article built around Laclau and Mouffe, 1985), the final chapter 8 resumes the conversation with Hallidayan or systemic-functional linguistics (henceforth SFL). This is appropriate not merely since in earlier books (e.g., Language and Power, 1989) Fairclough has taken his linguistic description largely from systemic-functionalism, understandably seeing it as one of the most applicable descriptive linguistic resources for analyses of genre, power, and identity in discourse, but also because like CDA practicioners, some systemicists are interested in charting the relationship of semiotic to social change, and in understanding “the generative power of the semiotic” (139).

The authors’ brief resume of key theoretical terms and assumptions of SFL is very fair: thus they report that SFL sees language as a stratal semiotic system, with meanings interfacing with social life, expressions with bodily mechanisms, and the lexicogrammar as an intermediate stratum between expressions and meanings. The theory’s three macrofunctions are introduced, together with the grammar’s three major systems, and the important distinction between realisation and instantiation is sketched in. Further to this, the place of field, mode, and tenor, and of dialect, register and genre, are all acknowledged, together with mention of semantic differences within a social practice characterized in terms of the Bernstein-derived notion of contrasting ‘coding orientations’.

So far so good. But still CDA has a complaint: that there is a ‘consistent gap’ between the theoretical account of the dialectics of the semiotic and the way this is actually operationalised in analyses, and here (using again that ‘trumping’ manoeuvre mentioned earlier) the gap and lack is demonstrated via re-analysis of the discourse of a doctoral thesis defence, previously analysed by Halliday (1994) and Hasan (1994). For Chouliaraki and Fairclough, SFL privileges the semiotic over the social, and the system over the text:

The apparatus of SFL also pushes the analyst to the side of system, for despite claims that every text (instance) ‘perturbs the system’ (Halliday, 1992), the analysis of texts is overwhelmingly an account of what choices the text makes from the potential of the system, of the text as an instantiation of the system. All texts are seen as making particular choices from the meaning and wording potentials, being ‘in’ a particular register (Halliday 1992) and ‘members’ of a particular genre (Hasan 1994). This leads to difficulties with hybrid texts–texts which mix discourses, genres or registers. (p.143)

The immediate cause of disagreement might seem rather slight: whether a confrontation about sexist bias in academic life which comes at the end of the thesis-defence should be construed as part of the defence text and genre, or as a separate text, a ‘post-mortem’ text, or similar. Is the confrontation subtextual, para-text, or separate text? Unquestionably, it is an instance of hybridity in the larger thesis-defence interaction; but while CDA would foreground that genre- and type-deconstructing hybridity, Halliday and Hasan seem to prefer to treat it as a separate text, a detachable or segregatable supplement, and even Hasan’s interesting treatment of this in terms of ‘permeability’ is criticized as a selective and marginal accommodation of hybridity rather than a full addressing of it. Here Chouliaraki and Fairclough make the following sharp point:

The difficulty in dealing more fully and dynamically with hybridity lies in the way the concept of insantiation is operationalised–the claim that a text is an instance of a contextually correlated selection from the potential specified in the system, that a text therefore is an instance of a register (is ‘in’ a particular register) or a ‘member’ of a genre. Although there is indeed a need to show how the instance is anchored in the system, this is making the instance fit the system. In the difficult business of grasping the dialectic of structure and event, we believe it is necessary to be as fully as possible open to the specificity of events, at the same time as reiterating how they are constrained by and reproductive as well as productive of structures. (144)

With these suggestions the authors come strikingly close to the radical perspective propounded in integrational linguistics (in, e.g., Harris, 1996 [esp. chapter 3, where Halliday is discussed], Harris, 1999, and Toolan, unpub. Ms, which discusses possible systemic and integrational linguistic convergences).

More broadly, Chouliaraki and Fairclough suggest that the SFL treatment of hybridity does not recognize that the social structuring is itself semiotic. Thus a joking phase in the thesis-defence is explained interactionally by Hasan as a strategy for easing tension, but the CDA stance wants to add that the articulation between the formal and serious and informal and non-serious is an ‘absolutely characteristic’ negotiation of the public-private dialectic, and how these are combined or articulated is ‘part of the ordering and regulative constraints of genre’ (145). In short, CDA wants to claim that it is (even) more interdiscursivity-minded than SFL, and that it provides a more theoretically rich account of ‘the co-genetic logic of the semiotic and the social’ (149). Several somewhat similar criticisms, arguably from within the same broad theoretical stance as that of those criticised, structure the final pages of the book, to the point that one wonders whether they might have been better suited to an exchange of views in a journal or at a conference, where Halliday, Hasan, Martin and others might respond. And behind all this re-positioning relative to fuller accounts of the co-genetic logic of the semiotic and the social will loom each particular reader’s questions about whether the project of CDA truly speaks to that reader’s interests and aptitudes: for the sociolinguist, at what point might it be felt that the CDA focus on social critique and theory is drawing attention and energy away from the detailed analysis of linguistic phenomenna which is their primary goal and proficiency?

There is some acknowledgement of this at the close, when the authors reiterate that ‘linguists have to be convinced that the social concerns of CDA do not deflect from detailed and careful linguistic… analysis of texts’ (152). In sum, this is a deeply-considered contribution, of particular interest to systemic linguists, with whom among the linguistic academic community it particularly seeks to develop a dialogue.


  • Halliday, M. 1992. How do you mean? In M. Davies and L. Ravelli, eds., Advances in Systemic Linguistics. London. Pinter.
  • Halliday, M. 1994. So you say ‘pass’… thank you three muchly. In What’s going on here? Complementary Studies of Professional Talk, vol.2, in Advances in discourse Processes, vol. 43, Norwood: Ablex Publishing Corp.
  • Harris, Roy. 1996 Signs, Language & Communication. London. Routledge.
  • Harris, Roy. 1999. An Introduction to Integrational Linguistics. Oxford. Pergamon.
  • Hasan, R. 1994. Situation and the definition of genres. In What’s going on here? Complementary Studies of Professional Talk, vol.2, in Advances in discourse Processes, vol. 43, Norwood: Ablex Publishing Corp.
  • Toolan, Michael. 2000. ms. Systemic-functional linguistics and integrational linguistics: any common ground? (Unpublished ms.)