The History of Discourse Studies

[Teun A. van Dijk]

Although delimitation debates and histories are not exactly the kind of issues I
find very important, I have followed with some interest the last contributions
about (what Levinson 1983 defined as) discourse analysis vs. pragmatics. Just a
few comments that are intended to show that the history of the study of
discourse is not *that* clear cut, and certainly not reducible to (say)
rule-based text grammar vs. interaction strategy based pragmatics, or other
simple binary divisions.

Thus, if I may briefly recall the scene of the final 1970s, that is, the kind of
discourse studies that could be known in the early 1980s, we already have nearly
the full gamut of what is available today: ethnography of speaking,
interactional sociolinguistics, conversation analysis, various types of
pragmatics (as formal speech act theory or as more general interaction theory),
various text and discourse grammars, the psychology of text processing, and the
beginnings of critical linguistics, to name only the most obvious ones (and
ignoring stylistics, the new rhetoric, argumentation and narration studies,
etc.).

Some of this work is more formal, normative, abstract and rule-oriented, other
work is more "empirical", "natural" (real discourse examples) and strategy
oriented in its descriptions, with many combinations of formal rules and
"natural" strategies: Thus, ethnography of communication from the start was
interested in both (communicative) rules, as well as in the description of
natural communicative events as data. CA used 'real' talk (well, transcripts),
but its interest was rather formal, namely in the (implicit) interaction rules
of 'members'. Psychology of text processing was initially influenced by (some
tenets in) text grammar, but soon developed a theory of natural, flexible,
strategic comprehension. Text grammar very soon (well, at least in my 1977 Text
and Context book) was defined as merely one --more formal-- direction of a
broader multidisciplinary theory of discourse, and already featured a
macro-semantic, pragmatic and a cognitive component with properties very far
removed from those of the rules of generative grammars. Similar attempts at
integration of various directions of research into a more general study of
discourse can be found in other studies.

Despite the obvious differences between the various directions of research,
there was also unity: common to virtually all these approaches was the
fundamental idea that sentences, propositions, interpretations, speech acts,
turns, moves, etc. come in *sequences*, and that the forms, meanings,
interpretations or social interaction functions of these 'units' are mutually
dependent on each other.

And whether using real or (soon quite rarely) invented discourse examples,
virtually all types of discourse study (including pragmatics) made use of
various forms of abstraction or reduction: Most forms of "social" discourse
analysis, despite its claim to study "real" discourse, simply ignored (and
ignore until today) the cognitive "reality" of talk and text, with some
exceptions in cognitive sociology and anthropology. The huge field of the
psychology of text processing (and its vast applications in education, media,
etc.) largely ignored, or only paid lip service, to the interactional or social
dimensions of cognition, communication and interaction. CA for a long time only
paid attention to some formal aspects of sequentiality in talk, largely ignoring
many others (such as speech acts, meaning coherence, cognitive aspects of
knowledge and presuppositions, as well as global and instititional social
structures). Text grammar ignored spoken interaction and social context. Most
scholars who studied talk, ignored text, and vice versa. And so on. It is only
in the 1980s and 1990s that systematic attempts at integration and more
cross-fertilization takes place after the initial monomanic isolation of the
various directions of discourse studies that emerged around 1970.

In sum: (a) The picture of discourse studies around 1980 is rather complex, and
cannot simply be divided into two or three main, polarized approaches, methods
or philosophies; (b) most directions were at least initially monomanically
"doing their own thing", ignoring other approaches, dimensions and in their own
way engaging in various forms of abstraction and reduction; but (c) towards
1980s there are already various attempts at integrating the many approaches into
a general, multidisciplinary "study of discourse".

In other words, the more detailed history of discourse study is still to be
written.


Teun A. van Dijk
University of Amsterdam
Program of Discourse Studies

1999-2000:

Universitat Pompeu Fabra
Institut Universitari de Lingüística Aplicada (IULA)
La Rambla 32
08002 Barcelona, Spain

Phone: +34-93-272.1200 (home)
FAX:   +34-93-272.0106 (home)

E-mail: teun at hum.uva.nl
Home-page: http://www.hum.uva.nl/teun