Hegemony, Cognition, and Critical Discourse Analysis

Critical discourse analysts root out the ideological and hegemonic underpinnings of common sense.  Hegemony, as a concept consists of a social and a cognitive dimension; as van Dijk observes that the minds of the dominated can be influenced in such a way that they accept dominance, and act in the interest of the powerful out of their own free will.  There are elements of social cognition involved that allow the same values and beliefs to be present in individual minds as schemas of long-term memory (Hougarrd and Oakley 14).  For the majority of critical discourse analysts, the evidence of hegemony as manifested discourse is the site of analysis; the proof of hegemony is found in text and speech, and how individual minds perceive, accept, and construct hegemony is unanalyzable or inconsequential.  There are a few scholars in the discipline, however, who are beginning to investigate the cognitive aspects of ideology and hegemony.  Tuen van Dijk investigated cognitive elements in discourse analysis as early as 1984.  In Prejudice and Discourse, van Dijk analyzed the formation of stereotypes and prejudices from a cognitive perspective and determined production and reproduction to be the result of the storage of individual experiences as event models in episodic and long-term memories (Wodak 184).  Van Dijk’s work today focuses on context, which he believes is cognitively constructed as he states:

If contexts control discourse at all, this is only possible when we conceive of them as cognitive structures of some kind. And only in this way are we able to define the crucial criterion of relevance, that is, in terms of a selective focus on, and subjective interpretation of some social constraint as defined by the participants. This also explains why discourse may be influenced by alternative, fictitious or misguided definitions of the social situation, as long as the speaker or writer sees it that way. Thus, it is not objective gender, class, ethnicity or power that control the production or comprehension of text and talk, but whether and how participants interpret, represent and make use of such external constraints, and especially how they do so in situated interaction (163)

Van Dijk believes the theory of mental models introduced by Laird-Johnson (Mental Models 1983) is the best method to abstract how participants “interpret and represent” context, he states in Discourse, Context, Cognition

…that mental models are not only relevant for discourse production and understanding, but also for any other kind of meaningful interaction and understanding. Hence, they cannot be reduced to text or talk. In discourse processing they explain understanding, as well as a host of other properties of discourse, such as anaphora, local and global coherence, topics, presuppositions, and so on. They also explain how we may falsely recall information from discourse that was never explicitly mentioned in such discourse at all, or how we may remember some event but do not remember whether we have read about, heard about it or seen it as images on television (170).

Van Dijk believes context is the mental model of a discourse participant’s experience of a communicative act (170).  He has begun work on a context model “that is the basis of our pragmatic understanding of discourse” (170).  Context models are integral in the production and reception of discourse.  Van Dijk’s context model is in the early stages of development and does not as of yet have a final structure, but the potential of his work for the future analysis of context is exciting. 

George Lakoff, in his book Don’t Think of an Elephant, develops an analysis of the establishment of mental frames in the 2004 U.S. presidential election.  Lakoff found that George Bush was successful due to his team’s establishment of a Republican friendly frameset in the initial debates and that John Kerry was unsuccessful due to his inability to shift the frame of discussion back to his own agenda (Wodak 185).  As Wodak points out, it is unlikely that such a frame shift actually changed viewer’s ideological beliefs, but the work is valuable in demonstrating the cognitive effect of rhetorical strategies.  In 1990, Eve Sweetser introduced a model of domain mapping based on Laird-Johnson’s mental model theory to account for the complex pragmatic accounts of modal verbs in the analysis of discourse.  Sweetser’s work was some of the first to introduce cognitive models into the study of pragmatics (1990).   A number of conversation analysts (a branch of discourse analysis that focuses on grammatical, structural, and pragmatics in speech) have adopted Fauconnier and Turner’s theory of conceptual blending (also referred to as conceptual integration by conversation analysts) as a model for the study of the pragmatic implications of speech.  Todd Oakley and Seana Coulson have apply conceptual blending to the pragmatic study of metaphor (2008), Barbara Dancygier studies the levels of blending in fictional narratives (2008), Robert Williams analyzes blending in instructional discourse (2008), and Anders Hougaard studies blending and interactional sequences of conversation (2005).  Conversation analysts and linguists specializing in pragmatics are leading the way in applying cognitive models such as conceptual blending to the analysis of discourse.  Critical discourse analysis can also benefit from the application of conceptual blending to not only enhance analysis of textual elements themselves, but also as a model for the cognitive space of ideology and hegemony.