The field of Epistemic Discourse Analysis

On 2–3 April 2012, a Symposium on Knowledge and Discourse was held at Pompeu Fabra University, Barcelona. New versions of some of the papers presented at that Symposium are published in this special issue of Discourse Studies under the general title ‘Epistemic discourse analysis’.

Epistemic discourse analysis (EDA) is defined here as the systematic and explicit study of the ways knowledge is interactively ‘managed’ (activated, expressed, presupposed, implied, conveyed, construed, etc.) in the structures and strategies of text and talk.

Since the articles in this issue focus on the epistemics of text and talk in linguistics, discourse analysis, conversation analysis and cognitive psychology, EDA obviously is a multidisciplinary field. Yet the field is even broader if we consider other relevant disciplines of the humanities and social sciences – as well as their limitations.

Epistemology provides insights into the nature of knowledge traditionally defined as ‘justified true belief’, though at present with many variations as to the relative and contextual nature of ascribing truth to beliefs as well as many approaches to criteria or standards of justification and reliability. Especially relevant for the epistemic analysis of text and talk is that one of the three major sources of knowledge is (reliable) discourse, traditionally and strangely called ‘hearsay’ in epistemology. Indeed, most of the knowledge about the world beyond our every experience we learn through text and talk, especially in public communication.

Since the 1970s, cognitive psychology and artificial intelligence (AI) have jointly emphasized that the natural production and comprehension, as well as the computer simulation, of discourse requires vast amounts of generic knowledge, variously described in terms of conceptual networks, prototypes, schemas, scripts or multimodal neural activation in memory and the brain. Local and global discourse coherence is thus strategically established by the derivation of inferences based on such knowledge during the construction of subjective mental models in episodic memory, representing the event or situation the discourse is about, as is shown in this issue in the articles by Sandi Chapman and Raksha Mudar, and by Danielle McNamara. Memory of what we have heard or read is thus not so much a process of recalling contents or structures of text or talk, but reactivating the mental model construed during their knowledge-based processing. Similarly, the influence of the parameters of the communicative situation (such as setting, participant identities, acts, goals and knowledge) on the structures of discourse is mediated by context models of the ways participants subjectively construe the communicative situation.

Where social psychology might be seen as the discipline that should be particularly focusing on the relations between discourse and knowledge, it has generally ignored both (with the exception, in very different ways, of discursive psychology in the UK and social representation theory in France). Yet knowledge presupposed in (public) discourse is socially shared, acquired, changed and normalized by members of epistemic communities in myriad forms of interaction and communication. Knowledge constitutes the very basis of all forms of social cognition, including social attitudes, ideologies, social identities and attribution, among many other phenomena traditionally studied in social psychology.

Similarly, the (macro) sociology of knowledge ignored the fundamental discursive nature of the ways social groups, organizations and institutions (re)produce knowledge, for instance through the mass media, schools, laboratories, universities, academies and other epistemic institutions involved in the construction and reproduction of epistemic communities.

Cognitive anthropology, among other topics – such as cultural models and knowledge schemata – has traditionally focused on the cultural variation of knowledge(s) and knowledge criteria in different cultures. Indeed, culture is often defined in terms of shared knowledge. Yet it usually neglected to relate explicitly such knowledge with a systematic study of the culturally variable ways text and talk are involved in the acquisition, communication and sharing of such knowledge in epistemic communities.

Linguistics, especially in its functional paradigms, has dealt with knowledge first of all in terms of (old/given and new/salient) information and its distribution in sentence structure, as is the case for the study of topic and focus (see the article by Eva Hajičová). Second, as shown in the article by Ilana Mushin, in research on evidentials it has been observed that some (mostly indigenous) languages have special morphological markers that express the (perceptual or communicative) source of the knowledge expressed by the speaker. More generally in all languages, speakers may signal the relation of their talk to reality (see the article by Wallace Chafe), and the sources, and hence the reliability, of their knowledge as distinct from mere belief or opinion as it is expressed or implied in text and talk. The linguistic analysis of modal expressions more directly shows how speakers mark their own statements as more or less true.

While crucial for our understanding of the relations between knowledge and language use, these earlier linguistic studies were often limited to the grammar of word and sentence structures, and ignored the many other ways the structures and strategies of text and talk ‘manage’ knowledge in interaction and communication. The fundamental distinction between old and new information, and specifically the strategic management of common ground, more generally organizes the structure of text and talk as a fundamental means of the acquisition of new episodic or generic knowledge on the basis of socioculturally or contextually given knowledge, as we know from conversation, news reports and textbooks.

Thus, within the broader scope of the study of discourse, we should attend to the role of definitions and the descriptions of situations, events, actions and persons, and the way subjective mental models of such states of affairs are ‘linearized’ in sequences of sentences or turns in conversations, at various levels of generality and specificity, degrees of granularity and in different contexts and genres. Because of the shared nature of sociocultural knowledge of epistemic communities, or on the basis of previous text or talk, discourse is essentially incomplete and implicit. Yet within the constraints of these fundamental pragmatic principles (not to tell recipients what they know already) there is still vast variation of the information we may or should express or leave implicit in discourse, again depending on genre and context. These and many other aspects of the ways information or knowledge are managed in text and talk remain on the agenda of the study of discourse and conversation.

Conversation analysis, as is also obvious in the articles by John Heritage and Lorenza Mondada, has made valuable contributions to this epistemic study of talk, focusing not only on the interactive nature of the way knowledge is conveyed, but also on the normative and moral dimensions of conversational epistemics: Who has access to, and the right, primacy, responsibility or obligation to express, presuppose or suppress knowledge in talk, and in what order or priority?

This is only a brief summary of the multidisciplinary nature of epistemic discourse analysis. Beyond the classical approaches in linguistics, psychology and conversation analysis, the articles in this issue, written by the most eminent scholars in the new field of epistemic discourse studies, show how this new area of the humanities and social sciences can be given shape.

Teun A van DijkEditor, Discourse Studies, Barcelona, May 2013