Introduction to Thematic Section on Text Linguistics

In 1992 the Danish government passed a law supporting Danish pure science by DKK 2 bn. At the Aarhus School of Business we saw this as an opportunity for creating a Centre for Text Linguistics and we submitted an application for financial support for a five-year period. The abstract of the project proposal read as follows: “The purpose of the Centre for Text Linguistics is twofold: On the one hand it will promote research in text linguistics in Denmark and on the other it aims at integrating different methods and approaches within the field. The basis of the studies the Centre wants to promote is firmly linguistic, but that does not mean that it will not remain in contact with neighbouring fields such as cognitive and computer sciences. The current scene of text linguistics offers many different approaches, but it is believed that the generally ecumenical spirit of Danish linguistics will offer an ideal context for unifying and testing such approaches. In order to achieve the goals of the Centre, the activities will be both research projects, publishing and diffusion of research results (both newsletters and more in-depth studies), and the organising of international conferences and symposia on text linguistics and related subjects. It is also hoped that the Centre will be able to attract young scholars by offering seminars, working facilities and a generally inspiring environment. The essential feature of the Centre is, however, its international orientation by which it is hoped to create a framework where scholars of different nationalities and theoretical backgrounds can meet both on a formal and an informal basis and exchange views and discuss issues of mutual interest.”1


Introduction to Thematic Section on Text Linguistics
In 1992 the Danish government passed a law supporting Danish pure science by DKK 2 bn. At the Aarhus School of Business we saw this as an opportunity for creating a Centre for Text Linguistics and we submitted an application for financial support for a five-year period. The abstract of the project proposal read as follows: "The purpose of the Centre for Text Linguistics is twofold: On the one hand it will promote research in text linguistics in Denmark and on the other it aims at integrating different methods and approaches within the field. The basis of the studies the Centre wants to promote is firmly linguistic, but that does not mean that it will not remain in contact with neighbouring fields such as cognitive and computer sciences. The current scene of text linguistics offers many different approaches, but it is believed that the generally ecumenical spirit of Danish linguistics will offer an ideal context for unifying and testing such approaches.
In order to achieve the goals of the Centre, the activities will be both research projects, publishing and diffusion of research results (both newsletters and more in-depth studies), and the organising of international conferences and symposia on text linguistics and related subjects. It is also hoped that the Centre will be able to attract young scholars by offering seminars, working facilities and a generally inspiring environment.
The essential feature of the Centre is, however, its international orientation by which it is hoped to create a framework where scholars of different nationalities and theoretical backgrounds can meet both on a formal and an informal basis and exchange views and discuss issues of mutual interest." 1 A specific purpose of the Centre was the integration of methodological elements from the most important of the traditions which have already been established internationally within this new branch of linguistic 9 research. Having its base in Aarhus of Denmark, the Centre should be international and a number of internationally renowned researchers should be affiliated to the Centre. Researchers originating from 12 European countries thus representing a large number of text linguistic "schools" accepted to work with the Centre under specific stipulated conditions.
As the only linguistic project, this was selected for international refereeing and it received very fine evaluations from three international linguists of great reputation (Oswald Ducrot of France, Deirdre Wilson of England and Janos Petöfi of Italy). In spite of the fact that the committee could thus find nothing to detract from the project's scientific soundness and viability, it was not to be found among the 23 projects receiving funding, and we found no other way to raise the sufficient funds. However, we have continued to believe in the idea of organising some kind of international network for promoting text linguistic work.
This thematic volume of Hermes can be seen to illustrate the basic, strategic rationale of the Centre for Text Linguistics. In fact, the contributions we have collected in this volume give a good illustration of the multiplicity of approaches represented in our network. They range from epistemological issues and general theoretical questions to concrete exemplifications of particular approaches.
François Rastier who is Director of Research at the Centre de Linguistique Française (University of Paris-La Sorbonne) and responsible for the research team Sémantique des Textes founded in 1992 presents a new theoretical framework for linguistics redefined as the science of texts. Departing from the view that textual structures are first of all semantic structures (and that a science of texts therefore must assume the form of text semantics), Rastier puts forward a new research perspective which is synchronic (the study of modern French texts), serial (based on large text corpora) and focused on written texts. The general framework of this research perspective consists of four autonomous components interacting during the production and reception of texts (thematics, dialectics, dialogics and tactics), four modes of describing texts (paradigmatic, syntagmatic, hermeneutic and referential), and three degrees of systematicity (the functional system of language, sociolectal norms, idiolectal norms).
Vijay Bhatia from the City University of Hong Kong presents three major orientations to genre theory, e.g. 1) genre as a typified rhetorical action (Bazerman), 2) genre as a staged, goal oriented social process (Halliday & Hasan, Martin, Kress), an orientation based on systemicfunctional linguistics and social semiotics, and finally 3) genre as a conventionalized communicative event, a tradition to which the so called genre analysis, developped by John Swales and Vijay Bhatia himself, belongs. The common ground of these new traditions within the field of genre studies consists of five features: the understanding of genres as embedded in conventionalized communicative settings, as dynamic structures evolving over periods of time, as responding to new sociocognitive needs, as based on the versatile concept of communicative purpose and as based on the genre knowledge of different professional communities.
Anne Reboul and Jacques Moeschler put the provoking question: "Should we go on making discourse analysis?", and their answer is a simple NO! To get thus far they present a thorough epistemological analysis. In principle, there are two ways to do text linguistics. You may either consider texts and discourse as ordered sets of utterances or you may consider discourse as a irreducible unit having a structure independently of the form and content of the utterance that makes it up. The authors label these two approaches analysis of the discourse (l'analyse du discours) and discourse analysis (analyse de discours) and throughout the article they argue against the legitimacy of discourse analysis. According to the authors, a reductionist approach to discourse would be both more scientifically sound and more efficient. Their main points are that the notion of coherence, which, being considered as a constituting property of texts, is at the very heart of discourse analysis, is at best a pre-scientific notion, which has never been either defined or characterised independently of the phenomenon it is supposed the describe or even explain. First, they define the epistemological notion of "scientifically relevant natural categories". Then, taking as their starting point some rather uncontroversial definitions of discourse, utterance and sentence, they argue that discourse is not such a category, because it can be reduced to a set of utterances. In fact, no discourse structure can be shown to be totally independent of the content of its utterances. The interpretation of the discourse parts depends not directly on coherence assumptions but on "local" and "global intentions".
Thus, the basic idea is that discourse interpretation is constructed according to the same principles as utterance interpretation, the only difference being that the number of information to take into account is manyfold bigger so that the interpretation process is correspondingly much more complex. But just as utterance interpretation, discourse interpretation is to be explained in terms of relevance in Sperber and Wilson's sense. At best coherence can then be seen as a principle guiding the construction of the global intention, so not only discourse analysis should be abandoned, so should the notion of coherence as a theoretical or scientific concept.
Co Vet illustrates how a more formal approach to text linguistics can be brought to yield new insight into some classical linguistic problems. He is concerned with the interpretation of temporal and aspectual relations in narrative texts. This is a subject which has been much debated over the last few years. Within the formal paradigm there are two main positions. One group claims that everything is to be explained in terms of anaphoric relations created by the verbal tense system. The other group argues that the recourse to general linguistic, and in particular to encyclopedic knowledge, is quite fundamental. The problem is then, of course, how to present such knowledge in the formal system. Various systems have been proposed, and Vet's point is that most of them pay too much attention to purely pragmatic aspects. After having yielded very detailed analyses of key examples from the relevant linguistic literature, he concludes that linguistic knowledge is not sufficient for predicting the right interpretation of temporal and aspectual relations in text fragments. However, much more information can actually be retrieved from the linguistic material than is normally acknowledged by researchers in the domain. Vet's analysis may thus be seen as a compromise that somehow rehabilitates the linguistic analysis proper.
Michael Herslund's contribution is an exemplary demonstration of how detailed analyses of small linguistic units can yield new insight into basic textual mechanisms. Herslund carries out a very detailed analysis of the French phrase en plein which is shown to have interesting textual functions mostly ignored by the dictionaries. In fact, en plein is normally considered as a kind of geometric expression, which occasionally functions as an intensifier, but it seems in no way to be predestined to have any kind of argumentative function. This is never-theless, what seems to be its basic value. Analysing a huge number of authentic texts, Herslund shows that en plein most often underscores a contrast or a contradiction. Furthermore, by appealing to some generally accepted topos about this contradiction, the phrase authorises a particular conclusion. In this sense, en plein can be seen as an argumentative, cohesive device.