Critical Discourse Analysis

The umbrella term ‘Critical Discourse Analysis’ shelters a broad family of analysts, but all have this in common: they approach texts from a certain prior point of departure,  often an avowedly political one. That is the ‘critical’ in the term. “The way we approach these questions”, says van Dijk, one of the doyens of Critical Discourse Analysis, “is by focussing on the role of discourse in the (re)production and challenge of dominance. Dominance is defined here as the exercise of power by elites, institutions or groups, that results in social inequality, including political, class, ethnic, racial and gender inequality.”” (van Dijk, 1993, p 249; emphasis in the original). To be aware of the exercise of power, and its resulting social inequality, requires a political theory about social life; and to have such a theory is vital. Otherwise, the CDA argument runs, one risks analysing non-problems or trivialities, or telling only part of the story, and missing its political significance. In the worst case, one’s merely technical analysis, by refusing to recognise political forces at work in the data,  may implicitly condone or perpetuate them. 

Within this broad family of analysts there are those who come from a post-structuralist background, to some degree independent of the linguistics traditions which inform a good deal of critical discourse work. In the post-structuralist tradition much use is made of Michel Foucault’s insights into the operation of power in discourses, and, increasingly, psychoanalytical concepts from the school of Jacques Lacan. An example of this sort of critical discourse analysis can be found in the work of Ian Parker (see, for example, his programmatic statement, Parker, 2003), and in the narrative analysis of Wendy Hollway (see, for example, Hollway and Jefferson, 2000), among many others. Other critical discourse analysts come from linguistics background, and bring with them an array of linguistic tools with which to unfold their data. 

For an illustration of the more linguistically-oriented kind of critical discourse analysis, consider this exemplary analysis, taken from a joint account of CDA by two of its best known (but of course not uniquely representative) proponents and theorists, Norman Fairclough and Ruth Wodak (1997). They give a 125-line long extract from a questionand answer radio interview with Margaret Thatcher during her time as Britain’s Prime Minister. It is not an event-led news interview; she is being asked generally, if I can offer a rough gloss, about her political beliefs and aspirations. Fairclough and Wodak present their analysis in eight facets, of which I select the two most emblematic examples. Inevitably this will impoverish what they say, but it will give a flavour of these authors’ CDA style, on two central CDA themes: power and ideology. I will quote part of the transcript to help illustrate their analysis.

Power. Fairclough and Wodak see Thatcher’s display of power in a number of discourse features: her use of longish monologues; her interruption of her interviewer (not illustrated in the extract above); and her use of linguistic devices such as parallel constructions (“it has to be strong to have defence”… “you HAVE got to be strong on law and order”… “you have to be strong to uphold the value of the currency”). Such rhetorical devices, the authors claim, are ‘the prerogative of professional politicians’ (ibid, p 272). CDA’s willingness to use extra-textual claims (in this case, about what generally politicians do) is shared by many, but not all, kinds of discourse analysis. 

Using thir knowledge of the political scene, the authors are able to say that by using such privileged talk, Thatcher not only “circumvents and marginalises [the radio presenter’s] power as interviewer”, but also exercises her power over the radio audience. They go on to observe that “Thatcherism can … be partly seen as an ongoing hegemonic [power] struggle in discourse and over discourse, with a variety of antagonists – ‘wets’ in the Conservative part, the other political parties, the trade unions, and so forth” (p 273). This is a good illustration of how CDA is able to make the kind of generalisation that allows it to link the immediate data back to the analysts’ prior political commitments.

Ideology. The authors note that, in the extract above, Margaret Thatcher formulates a free-market ideology explicitly; but their analysis aims to adds value by showing how she expresses the ideology more subtly. This stretch of her words (and some 20 further lines not shown here), they say,  “is actually” [i.e. not as one might first naively think, without analytic help] “built around a contrast between government and people which we would see as ideological: it covers the fact that ‘people’ who dominate the creation of ‘thriving industries’ and so forth are mainly the transnational corporations, and it can help to legitimize existing relations of economic and political domination” (pp. 265-266). 

Fairclough and Wodak do not specify exactly where in the extract Thatcher’s failure to mention transnational corporations was significant (that it is a “fact” that her words “cover”). This is an important analytic point. Claiming that something is a fact, and that it is significantly absent from a stretch of discourse, is a harder claim to ground than pointing to something that is significantly present (after all, there is an infinity of things that may be facts, and which are absent from any given stretch of talk or text; whereas what is there is at least there). Different discourse analysis traditions solve the problem in different ways. CDA notices absence not by working it out from the logical or pragmatic implications of the utterances around it, or from of the reaction of those who are there to hear it, as other schools of analysis do. It works it out by virtue of prior theorising about the political or social nature of the world to which the utterance refers. In this case, Fairclough and Wodak have a prior theory or account of what is happening in the British economy, what ‘thriving industries’ refer to, that these industries are owned by transnationals, and that this ownership is important in the discussion that Thatcher is currently having with her interviewer. They have a further belief, or expectation: that, if given an opportunity, a speaker ought to express the politically relevant facts of the matter (as the analysts see them, and whether they are logically or pragmatically implied or not, or whether the speaker’s local interlocutors hold them to it or not). Margaret Thatcher was given the opportunity, and did not mention transnationals; therefore, it is analytically safe, as well as useful, to claim that she is masking their role in the economy.

If we translate these snippets of analysis back into the four core features of DA (data found naturally; interpreted in co-text; non-literally understood; actions achieved), we see that CDA will insist on a very wide sense of “co-text” in its interpretation, and on drawing out implications which may not be visible to those who do not share the analyst’s prior political commitments, or hesitate to apply them to the data. Its prime candidate for “social action” is the action, taken to be unequally shared in society, of constituting the social world.  CDA is attractive to scholars who have the view that discourse analysis must ally itself to a social theory, and must be aware of inequalities in society. This is shared, in a more dilute form, in the next influential discourse analysis I shall look at.