Picking up the two volumes by Teun van Dijk (Discourse and Context, 2008 and the book under review) for the first time, I had the following sentence by Asher in my mind: "Context is one of those linguistic terms which is constantly used in all kinds of contexts but never explained" (Asher 1994:731). Van Dijk tries to lead readers out of this confusion. He outlines how social contexts affect text and talk by presenting a new multidisciplinary theory of context. The book contains a preface and six chapters, including an introduction and a conclusion, as well as an extensive bibliography, author index, and subject index.
Van Dijk argues in detail that "there is no direct link between situational or social structures and discourse structures" (p. vii). In other words, context defined as the relevant properties of social situation (p. 4) has no effect on discourse. No direct relationship can be found between aspects of the social situation and discourse. He argues that if there were such a direct relationship between social situations and discourse, all people would speak in the same way (p. 4). If so, how do we relate social situations and discourse, and account for the uniqueness of each discourse? Van Dijk wishes to develop a theory of context along these lines, a theory which closely examines the way participants understand social situations as an important point for the understanding of context. To van Dijk, context is "what is defined to be relevant in the social situation by the participants themselves" (p. 5). What influences the cognitive processes of discourse production and comprehension is nothing but this subjective representation.
The notion of mental models has been borrowed from cognitive psychology in the theory proposed by the author. A model is "a subjective representation of an episode" (p. 6). Contexts are mental models and represent personal experiences, namely the experience of the current communicative episode. The author calls mental models of communicative episodes context models, or simply contexts. It is said that "contexts are not some (part of a) social situation but a subjective mental model of such a situation" (p. 7). Contexts, as elaborated by van Dijk, are dynamic constructs. These constructs are mental models and not visible. They become visible only indirectly when these constructs control interaction and discourse.
In Chapter 2, the author also adds a social dimension to his theory of context. This means that his theory of context includes not only individual intentions, goals, knowledge, and experiences, but also categories and beliefs. He scrutinizes some studies in social psychology to define social situations. The category of Setting is examined in detail, including the relevant information about Place/Space and Time in the framework of ecological psychology. Socially shared beliefs are defined in terms of mental representations. Social representations such as beliefs, ideologies, and knowledge can be learned, shared, changed, and applied, and this is only possible when group members represent them mentally.
A theory of discourse, according to van Dijk, should spell out how participants show their shared knowledge in talk, both interactionally and cognitively. Chapter 3 focuses on the sociological dimensions of the theory of context. It is this dimension of the theory which articulates how discourse participants can form and apply definitions of communicative episodes in their ongoing text and talk. I do agree with the author that in addition to sociological aspects of the theory of context, sociocognitive aspects also play an important role. It means that if participants want to construct and apply a mental model of the communicative situation, they need not only interactional strategies of focusing on relevant social categories or dimensions of participants, but also powerful cognitive ones.
Van Dijk begins the fourth chapter, "Context and culture", with a thought-provoking quotation from Hanks: "What is context? Everything and nothing. Like a shadow, it flees from those who try to flee from it, evading the levels and categories of theory, and pursues those who try to flee from it" (Hanks 1995:140). Although the notion of context is not the same in all languages and cultures, in the author's analysis the fundamental structures of categories of context models are not that different. He examines how...