Fairclough approaches the study of discourse primarily from a social perspective. Fairclough deals with the ways language constitutes social relations and social reality, and describes how the “social reproduction of realisations of domination” (Faiclough 1995a, p. 24) are effected through discursive means. Fairclough claims that “naturalised implicit propositions of an ideological character are pervasive in discourse, contributing to the positioning of people as social subjects” (Faiclough 1995a, p. 21), which suggests that through discourse the social standing and the power relations of the communication partners is unwittingly accepted. Fairclough (1995) claims that the reproduction of power in discourse happens unnoticed and that all communication partners accept power relations exhibited in a text without questioning the validity of such power relations.

Fairclough (1992) also stresses the educational implications of CDA, advocating Critical Language Awareness (CLA). CLA is based on Critical Language Study (CLS), a framework for the critical analysis of discourse. The CLS framework was developed in the footsteps of critical linguists Fowler, Hodge and Kress. Fairclough’s (1992) framework of CLS accommodates the following five theoretical supports:

(1) discourse shapes and is shaped by society – i.e. “the use of language is socially determined and […] language varies according to the social situation it is used in” (Fairclough, 1992, p. 8);

(2) discourse constitutes knowledge, social relations and social identity; in other words, discourse has effects on the society: language represents our world and experiences as well as contributes to the formation of social relations and constitutes social identities;

(3) discourse is shaped by relations of power, and is infiltrated with ideologies, which assumes that society has effects on discourse: language use and language varieties are valuated according to the power their users possess in society;

(4) the shaping of discourse is central in power struggles – discourse conventions have social effects, thus controlling discourse means possessing dominance in the society and

(5) CLS shows how society and discourse shape each other, that is, CLS describes the complex interrelations holding between discourse and society. Along these theoretical notions, Fairclough (1992) provides a practical application of this framework in the field of language education. He terms this framework Critical Language Awareness, which is a vital component of language education. As described above, Fairclough’s (1992) framework primarily focuses on social relations of power and ideology and lays the groundwork for future CDA approaches through the five theoretical CLS supports. Fairclough (1992), nonetheless, ignores some other noteworthy aspects of discourse such as historical and cultural context.

Another school, the Vienna Group, including Wodak, Lutz and Matouschek, primarily researches implicit prejudiced utterances, i.e. discourse that pretends to be non-prejudiced. The Vienna Group wishes to provide the analysts of such discourse with analytical tools “to enable the analysis of implicit prejudiced utterances as well as to identify and expose the codes and allusions contained in prejudiced discourse” (Fairclough and Wodak 1997, p. 266). This approach is suitable for revealing that text producers use allusions that their readers can understand only if they know the objects of communication or the cultural and social background which is referred to. This strategy enables writers/speakers “to back away from responsibility easily as they have not made their statements explicit” (Fairclough and Wodak 1997, p. 266). Given this, the most important feature of this approach is to try to “integrate systematically all available background information in the analysis and interpretation of the many layers of a written or spoken text” (Fairclough and Wodak 1997, p. 266). The drawback of this approach, however, for the current research purposes, is that this analytical tool has been specifically designed for prejudiced discourse, not argumentative newspaper discourse.

Another intertextual and hermeneutics-based approach is that of Maas. His premise is that “[t]ext analysis becomes discourse analysis, whereby discourse correlates to a historically formed social practice” (Maas qtd. in Fairclough and Wodak 1997, p. 267). In Maas’ framework, discourse analysis investigates the rules that “constitute a specific discourse, which thus make a certain text [e.g.] a fascist text. Each text relates to other texts, synchronically and diachronically [...] [here] a non-context-oriented analysis is doomed to fail” (Fairclough and Wodak 1997, p. 267). This contextual and intertextual approach is suitable for certain genres and for certain topics only, where intertextuality plays a vital role. In the current research, intertextuality has a considerably less marked role as far as the interpretation of discourse is concerned, given that Hungarian political issues are not widely publicised in English language media.

The last approach discussed here is by van Dijk: his CDA theory supposes that discourse is a social strategy, and van Dijk offers a perspective for the analysis of power and ideology that is wider – through the inclusion of social, cultural, historical as well as interpretative and explanatory dimensions – than that of Fairclough’s framework. Van Dijk (1993) views discourse in its social, cultural and historical context in dynamic terms supposing that each contextual feature affects not only all other features but also the text produced under these circumstances and claims that CDA should interpret and explain textual phenomena in the light of the actual social, cultural and historical context.

Van Dijk criticises many CDA traditions for failing “to show how societal structures influence discourse structures and precisely how societal structures are […] enacted, instituted, legitimated, confirmed or challenged by text and talk” (Fairclough and Wodak 1997, p. 265). Van Dijk, as part of his sociocognitive approach to discourse, which “assumes that different types of schemata [i.e. mental representations of one’s knowledge] […] are important for text production and text comprehension” (Fairclough and Wodak 1997, p. 266), also claims that cognition, an intermediary between language and society and culture, should be taken into consideration. In his view, “no direct relation can or should be constructed between discourse structures and social structures” (Fairclough and Wodak 1997, p. 265, emphasis added by the author). Van Dijk combines quantitative and qualitative analytical tools and takes account of a wider range of contextual features of discourse than any of his predecessors, thereby creating the most thorough and well-structured CDA framework. This framework, on the one hand, embraces all contextual features identified by Translation Studies research with reference to the translation of political texts (cf. Section 4.3) and, on the other hand, allows for the analysis of the interrelation and interplay between these contextual features surfacing on a textual level.

In the next section, van Dijk’s (1993, 1997, 2001, 2003) CDA will be introduced with special emphasis on one part of the framework, the Discourse–Society Interface, which functions as an analytical tool in the van Dijkian CDA. The description of van Dijk’s CDA is justified by the fact that the components of the Interface can be adapted to the principles and purposes of the current research (cf. Section 4.5.1).