On the left-hand side of Fig. 2.1 are approaches to the study of language which start from a particular theoretical standpoint. Each approach is grounded in a particular view of language and its relationship with society, by which I mean both context (however defined) and participants. These “perspectives” lean towards using methods which support either a more descriptive or a more critical aim.
Critical Discourse Analysis
CDA is an approach to texts which uses diverse analytical tools to uncover the covert enactment of power through language. Analysts in this tradition are explicit about having an agenda (Fairclough, 1989, 1992a; van Dijk, 1985,
2008; Weiss & Wodak, 2003; Wodak, 1996, 2000). They argue, rightly in my view, that any analysis process, even one which purports to be purely descriptive, starts from a point of view, and the analyst of necessity selects an approach and a data set which reflect a particular research intent. Given this, writers in the CDA tradition find it appropriate to make this intent a principle of the approach, as an overt declaration of reflexivity. Critical work in the field of media representation has shown effectively how news stories, far from being an objective representation of “what is out there”, offer a version of the world that is shaped by political, financial, institutional, personal and temporal constraints. However, it has been argued (e.g. Graham, 2005) that in following its agenda for change, much work in CDA is inclined to make its own claims to truth and objectivity. Widdowson (1995), in a frequently cited debate with Fairclough about the theoretical validity of CDA as an analytic approach, criticises Critical Discourse Analysts for bias in seeking out examples of text that support their own political viewpoint.
While CDA is characterised by its attitude to text analysis rather than by a specific method, different CDA approaches (Fairclough, 1989; van Dijk, 2008; Weiss & Wodak, 2003; Wodak & Meyer, 2001) offer fully developed analysis frameworks, which have in common that they place the individual text at the heart of a set of social practices and cultural norms that tend to both reflect and perpetuate the interests of those with power and access to voice. Writers such as Fairclough and Kress use SFG as a tool for the close analysis of written text, and I return to SFG as analysis tool and perspective later in the chapter.
For the BP research, although I started from the principle that media representation does not reflect “the truth”, but is instrumental in constructing one (albeit influential) version of the truth, I did not wish to pursue an agenda for change from the outset, as is the declared aim for CDA. It was important to me that I came into the research process with as few preconceptions as possible. I fully expected to find evidence of power and convention rooted in interest in the texts I studied. However, in developing a research approach that started with the identification of patterns in the texts, however derived, I intended a necessary critical perspective to follow, rather than precede, my analysis. I agree with Coupland and Jaworski’s (2001: 145) observation that “In all but its blandest forms, such as when it remains at the level of language description, discourse analysis adopts a ‘critical’ perspective on language in use”. The research approach I propose in this book offers a critical view of representation, but one that emerges from its descriptive aims, rather than from an emancipatory agenda. In its understanding of text and socio-cultural context, my approach owes more to the work of Barthes than of Fairclough, and I discuss in Chap. 3 a comparison of the Barthesian framework I draw on and the Faircloughian conceptualisation of discourse.