Bonnafus Temmar Discourse Analysis Human and Social Sciences

&
HUMAN
DISCOURSE
What is the relationship between discourse analysis and its more recent
AND SOCIAL
S. BONNAFOUS & M. TEMMAR (eds) DISCOURSE ANALYSIS AND HUMAN AND SOCIAL SCIENCES
companion disciplines such as sociology, political science and information
SCIENCES
and communication sciences, at their point of convergence between the
symbolic and the social? How are relationships evolving between dis-
course analysis and disciplines like the literary studies, psychoanalysis
and philosophy, which have been the constant companions of linguistics
ANALYSIS
as these emerged and developed? What is the place and role of discourse
analysis in Europe? These are some of the themes dealt with in this book.
A team effort on the part of Centre d’Etude des Discours, Images, Texte,
Ecrits, Communication (Céditec EA 3119), it aims not to present another
view of the history and concepts of discourse analysis, but to encourage
thinking and debate on interdisciplinary practices.
Simone Bonnafous is Professor of Media and Communications at the

University of Paris-Est Créteil. In 1999, she created the research cluster
Céditec, which she directed until 2008. After her doctoral dissertation on
the 1979 convention of the Socialist Party in Metz, she worked extensively
on the discourse about immigration and immigrants. Located at the cross-
roads of linguistics and political science, her best known book is Immigra-
tion surrounded by words (1991). More recent publications have focused
on gender studies (such as the International Women’s Day). Since 2006,
she has held leading positions in higher education administration in France. SIMONE BONNAFOUS &
MALIKA TEMMAR (eds)
Malika Temmar is lecturer of linguistics and discourse analysis at the Uni-
versity of Picardie Jules Verne, Amiens. She has been member of Ceditec
since its foundation in 1999. She works on philosophical discourses (such
DISCOURSE
as Descartes, Condillac, Merleau-Ponty) and media discourses (such as
blogs of literary critics). Her research focuses on concepts and methods ANALYSIS &
of discourse analysis. She has recently co-edited a book on Discourses of
the Economy, to appear in 2013. She has co-edited a special issue with Le HUMAN
AND SOCIAL
discours et la langue on “les sujets de l’énonciation”. Her work Le recours
à la fiction dans le discours philosophique will come out in April 2013.
ISBN 978-3-0343-1241-7
SCIENCES
www.peterlang.com
&
HUMAN
DISCOURSE
What is the relationship between discourse analysis and its more recent
AND SOCIAL
S. BONNAFOUS & M. TEMMAR (eds) DISCOURSE ANALYSIS AND HUMAN AND SOCIAL SCIENCES
companion disciplines such as sociology, political science and information
SCIENCES
and communication sciences, at their point of convergence between the
symbolic and the social? How are relationships evolving between dis-
course analysis and disciplines like the literary studies, psychoanalysis
and philosophy, which have been the constant companions of linguistics
ANALYSIS
as these emerged and developed? What is the place and role of discourse
analysis in Europe? These are some of the themes dealt with in this book.
A team effort on the part of Centre d’Etude des Discours, Images, Texte,
Ecrits, Communication (Céditec EA 3119), it aims not to present another
view of the history and concepts of discourse analysis, but to encourage
thinking and debate on interdisciplinary practices.
Simone Bonnafous is Professor of Media and Communications at the

University of Paris-Est Créteil. In 1999, she created the research cluster
Céditec, which she directed until 2008. After her doctoral dissertation on
the 1979 convention of the Socialist Party in Metz, she worked extensively
on the discourse about immigration and immigrants. Located at the cross-
roads of linguistics and political science, her best known book is Immigra-
tion surrounded by words (1991). More recent publications have focused
on gender studies (such as the International Women’s Day). Since 2006,
she has held leading positions in higher education administration in France. SIMONE BONNAFOUS &
MALIKA TEMMAR (eds)
Malika Temmar is lecturer of linguistics and discourse analysis at the Uni-
versity of Picardie Jules Verne, Amiens. She has been member of Ceditec
since its foundation in 1999. She works on philosophical discourses (such
DISCOURSE
as Descartes, Condillac, Merleau-Ponty) and media discourses (such as
blogs of literary critics). Her research focuses on concepts and methods ANALYSIS &
of discourse analysis. She has recently co-edited a book on Discourses of
the Economy, to appear in 2013. She has co-edited a special issue with Le HUMAN
AND SOCIAL
discours et la langue on “les sujets de l’énonciation”. Her work Le recours
à la fiction dans le discours philosophique will come out in April 2013.
SCIENCES
DISCOURSE ANALYSIS AND HUMAN AND SOCIAL SCIENCES
SIMONE BONNAFOUS &
MALIKA TEMMAR (eds)
DISCOURSE
ANALYSIS &
HUMAN
AND SOCIAL
SCIENCES
PETER LANG
Bern • Berlin • Bruxelles • Frankfurt am Main • New York • Oxford • Wien
Bibliographic information published by die Deutsche Nationalbibliothek
Die Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data is available on the Internet at ‹http://dnb.d-nb.de›.
British Library Cataloguing-in-Publication Data: A catalogue record for this book is

available from The British Library, Great Britain
Library of Congress Cataloging-in-Publication Data
Discourse Analysis & Human and Social Sciences / SIMONE BONNAFOUS & MALIKA
TEMMAR (eds).
pages cm.
ISBN 978-3-0343-1241-7
1. Discourse analysis–Social aspects. 2. Discourse analysis–Psychological aspects.
3. Interdisciplinary approach to knowledge. I. Bonnafous, Simone, editor of compilation.
II. Temmar, Malika, editor of compilation. III. Title: Discourse Analysis and Human and
Social Sciences.
P305.18.I57D57 2013
401‘.41–dc23
2012047670
Céditec et Institut Universitaire de France
Original title : Bonnafous Simone et Temmar Malika (eds), Analyse du discours et

sciences humaines et sociales, Editions Ophrys, Paris, collection “les chemins du
discours”, 2007.
ISBN 978-3-0343-1241-7 pb. ISBN 978-3-0352-0165-9 eBook
© Peter Lang AG, International Academic Publishers, Bern 2013

Hochfeldstrasse 32, CH-3012 Bern, Switzerland
info@peterlang.com, www.peterlang.com
All rights reserved.

All parts of this publication are protected by copyright.
Any utilisation outside the strict limits of the copyright law, without the permission
of the publisher, is forbidden and liable to prosecution.
This applies in particular to reproductions, translations, microfilming, and storage
and processing in electronic retrieval systems.
Printed in Switzerland
Contents
Introduction
Simone BONNAFOUS and Malika TEMMAR . . . . . . . . . . . . . . . . 1
I. Discourse analysis in Europe

Johannes ANGERMÜLLER . . . . . . . . . . . . . . . . . . . . . . . . . 9
II. Discourse analysis and information and

communication sciences: beyond corpora and methods
Claire OGER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
III. Discourse analysis and comprehensive sociology.

A critical look back at an interdisciplinary
research practice
Claire OGER and Caroline OLLIVIER-YANIV . . . . . . . . . . 43
IV. Multidisciplinary work on discourse.

Examples of a “way of doing” discourse analysis
Alice KRIEG-P LANQUE . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
V. Political discourse analysis: content analysis,

lexical statistics, a semantic-enunciative approach
Pierre FIALA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
VI. Analysing a debate. The contributions

of argumentation study to political science
Juliette RENNES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
VII. Discourse analysis and the study of literature

Dominique MAINGUENEAU . . . . . . . . . . . . . . . . . . . . . . . 113
VIII. Discourse analysis and history. Meeting and forgetting

Marie-Anne PAVEAU . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
VI Contents
IX. The role of psychoanalysis

Dominique DUCARD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
X. Discourse analysis and philosophy:

intersecting perspectives
Malika TEMMAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
The authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173

Introduction1
Simone BONNAFOUS and Malika TEMMAR
Discourse analysis became current in France in the late 1960s, thanks

to researchers working in a variety of disciplines such as linguistics,
sociology, history and philosophy. They all shared an interest in lin-
guistic phenomena considered in their social and historical contexts.
Although the concepts and methods of the language sciences have
long been a feature of this field of research, relationships with the other
human and social sciences have always played a part too. Exchanges
between these disciplines, far from signalling a fragmentation of dis-
course analysis, have gone hand in hand with its development as it
came to be recognised and established through research centres, con-
ferences, journals and, more recently, textbooks and dictionaries2. The
1 Thanks to Alice Krieg-Planque and Claire Oger for reading this text and giving
their advice. We would also like to thank Dominique Ducard for his support
during the later stages of publication.
2 Without claiming to be exhaustive, a list might include: Dominique Maingueneau,
L’Analyse du discours. Introduction aux lectures de l’archive, Paris, Hachette
Supérieur, collection Linguistique, 1991. Georges-Elia Sarfati, Eléments d’analyse
du discours, Paris, Nathan Université, collection 128, 2001. Francine Mazière,
L’Analyse du discours. Histoire et pratiques, Paris, Presses Universitaires de
France, Collection Que sais-je?, 2005. Patrick Charaudeau and Dominique Main-
gueneau editors, Dictionnaire d’analyse du discours, Paris, Seuil, 2002. Catherine
Détrie, Paul Siblot and Bertrand Verine editors, Termes et concepts pour l’analyse
du discours. Une approche praxématique, Paris, Honoré Champion, 2001. “Les
analyses du discours en France” (coordinated by Dominique Maingueneau), Paris,
Larousse, Langages, no 117, 1995. “Le discours: enjeux et perspectives” (coor-
dinated by Sophie Moirand), Paris, Hachette/Edicef, Le Français dans le monde,
special edition, 1996. “Analyse du discours. Etat de l’art et perspectives” (coor-
dinated by Dominique Maingueneau), Marges linguistiques. Langages. Représen-
tations. Communication, half-yearly electronic journal of language sciences,
publisher M.L.M.S., <http://www.marges-linguistiques.com>, no 9, 2005.
2 Simone Bonnafous and Malika Temmar
increasing number of works on discourse analysis testifies to the vi-

tality of this domain which has been able to interact with the growth
of new disciplines such as information and communication sciences.
After the “golden age” of the pioneers who featured in the great intel-
lectual debates of the time (on the notions of ideology and subject for
example), since the mid-1980s there has been a certain specialisation,
in terms of topics of study and sectors, such as media, political and
institutional discourse.
The Céditec (Centre d’Etude des Discours, Images, Textes, Ecrits
et Communications, EA 31 19) was created in 1999 by a small group
of teachers in the language sciences, information and communica-
tion sciences and sociology. Since its inception it has always placed
“discourse” – which has its focus beyond diversity of meaning and
conceptual frameworks – at the centre of its questioning, and been
particularly interested in the relationship of discourse to the political
arena and to knowledge.
This interest is the legacy of the Ecole Normale Supérieure at
St Cloud, with its laboratory of lexicometry and political texts. It was
founded in the 1970s by Maurice Tournier, who was a major figure in
lexicology and political lexicometry in France. From these origins and
from the institutional and epistemological links maintained by several
of its members with the centres responsible for founding discourse
analysis in France, Céditec has remained deeply grounded in the lan-
guage sciences and in the political sphere. These centres were the
departments of language sciences at the University of Nanterre and the
groups of philosophers, linguists and psychologists gathered around
Michel Pêcheux. Because of these influences, Céditec has maintained
a keen interest in computer processing of texts, which is one of its
major methodological contributions. It has also continued to invest
substantially in the journal Mots, les langages du politique3.
3 An interdisciplinary journal which brings together researchers in language sci-

ences with those in information and communication sciences and political sci-
ence around a common objective.
Introduction 3
Céditec has in addition benefited from a wide variety of differ-

ent inputs. Among these is the work carried out since the 1970s by
Dominique Maingueneau and Frédéric Cossutta on “constituent” dis-
courses (literary and philosophical discourses) as well as the research
by enunciation linguists like Dominique Ducard, and more recently
the contributions of young researchers who are interested in discourse
from political science, sociology, information and communication
sciences and so on.
Céditec is therefore both the fruit of this particular academic col-
laboration and also the place where these epistemological and scien-
tific developments are to be seen. Even though a good number of its
members do not teach at the Université Paris-Est Créteil, Céditec
was effectively founded within a multidisciplinary university, where
language sciences are not a separate department, but where questions
of enunciation, argument, enunciative polyphony, language acts, etc.
are dealt with as much in arts subjects and humanities as by commu-
nication and social sciences. In other words, “discourse”, as a so-
cially embedded language activity, is the raw material for much of
our teaching; and this has enabled an interdisciplinary team to be
founded at the Université Paris-Est Créteil – a team which is both
close to and different from other discourse analysis teams.
This diversity of disciplinary origins, and also of their scope (from
literature to politics via media, philosophy and institutions), finds its
rationale in a certain definition of “discourse” and the relationship
between the social and linguistic. The development of a declared
epistemological guideline, which has been driven, among others, by
Dominique Maingueneau, has meant that reflection on the history
and evolution of discourse analysis, including its encounters with
other disciplines, is one of our shared preoccupations.
It was in the context of this thinking that Malika Temmar, a sen-
ior lecturer in language science and author of a thesis on fiction in
philosophical discourse, agreed to organise a Céditec seminar in March
2005. This would enable several members of the team to take stock
of the place of French discourse analysis in Europe and to examine
our interdisciplinary practices of discourse analysis, each from its
specific boundaries. It would also permit an overview of the way in

which the relationship between discourse analysis and its founding
disciplines has evolved over time.
This book provides an account of this process and of the shared
reflections resulting from it. It is not a new introduction to the history
and concepts of discourse analysis; this can be found in the books
and journals already mentioned. It is more of a team effort whose
purpose is to recreate what goes to make up the intellectual life of
Céditec and to give an account of what discourse analysis is today in
its cross-boundary practices.
We have therefore followed the order of our seminar by asking
the sociologist Johannes Angermüller, who is also the author of a
Franco-German thesis on the state of the social sciences and the in-
tellectual landscape in France between 1960 and 1980, to start off
this volume with a chapter on discourse analysis in Europe.
We then gathered together five contributions providing a notion
of how discourse analysis fits together with its newfound compan-
ions, information and communication sciences and political science,
at the meeting point between the symbolic and the social. Three texts
have been written by colleagues (Claire Oger, Caroline Ollivier-Yaniv
and Alice Krieg-Planque) working and teaching in political, media
and institutional communication, where discourse analysis now oc-
cupies an entirely new place in training. Their formation and their
various disciplinary roots – from linguistics to sociology – have of-
ten led them to work together on research and writing, as will be seen
in these papers which can be read as casting three different lights on
very closely-related questions. In 2005 Juliette Rennes completed a
political science thesis on the polemics surrounding the rise of women
in public careers under the Third Republic. Her contribution to this
book completes the previous approaches by showing what discourse
analysis contributes to the understanding of political debates and strat-
egies, with the notion of “controversy” being a good example of how
several disciplines can be linked. Finally, Pierre Fiala examines the
way in which corpus linguistics, assisted by the development of au-
tomatic language processing tools and the extension of text databases,
Introduction 5
have helped re-direct language sciences back towards empirical data

and experimental methods, by reformulating the question of mean-
ing and verbal interaction on the basis of data analysis.
The four contributions which follow have an objective that is
both historical and epistemological, since they are concerned with
examining the relationships between discourse analysis and the dis-
ciplines which have continued to be partners of the language sci-
ences as they emerged and evolved. Dominique Maingueneau, for
example, analyses the relationship between discourse analysis and
literary studies, often seen as rather contentious even though the former
derived partly from the latter through the practice of textual commen-
tary. Marie-Anne Paveau, Dominique Ducard and Malika Temmar,
on the other hand, return to the subjects of history, psychoanalysis
and philosophy which, during the 1960s and 1970s, provided the
spaces in which discourse analysis was able to grow. It is evident
from this that “meeting” is sometimes followed by “forgetting”, to
quote Marie-Anne Paveau’s title, and that, on the institutional and
scientific front, discourse analysis is less at home today with some of
the disciplines that are its historical companions than it is with more
recent acquaintances within the social sciences (mainly sociology,
politics and information and communication sciences).
This book was conceived and written by teachers/researchers –
colleagues who do not divorce their research activity from their
teaching. Indeed, each of these feeds into the other. The texts gath-
ered together here were first presented as lectures to an audience of
researchers and young doctoral students, but our aim in publishing
them is to make them accessible and useful to a wider public, not
least those students engaged in research for Master’s degrees and
doctorates.
Without claiming to put ourselves in the place of readers – for
the way in which they understand these texts will depend first and
foremost on the questions they themselves are asking – and without
wishing to reduce the texts to a single dimension, we have neverthe-
less tried to organise them according to three priorities:
– A component which explores avenues that have been relatively

neglected until now. This, for example, allows the history of Euro-
pean traditions of discourse analysis to be examined, and ena-
bles us to see with Johannes Angermüller how its major trends
(“French formalism”, “German hermeneutics”, “Anglo-Saxon
pragmatism”) interconnect today in the networks which gather
together researchers from different countries. The long multilin-
gual bibliography at the end of this book also has an informative
purpose. In the same way, Malika Temmar’s contribution is not
so much a reflection on the relationships between philosophy
and discourse analysis as a demonstration of the way in which
philosophical discourse, as a scientific discourse and like any
scientific discourse, may be an object of discourse analysis. Then,
in the article by Dominique Ducard, we have a timely reminder
of the links between psychoanalysis and discourse analysis: by
means of various examples, he shows how certain enunciative
processes reflect the order of Law in discourse and not merely
subjectivity or inter-subjectivity. Listing references to Freud and
to the linguist Culioli together with the work of Pierre Legendre
and Vincent Descombes, this contribution is a plea for the an-
thropological significance of psychoanalysis.
– A historical and epistemological component. This naturally plays
a part in all the articles, but has a dominant role in those contri-
butions which retrace the complex history of the relationships
between discourse analysis and information and communication
sciences (Claire Oger); discourse analysis and the study of litera-
ture (Dominique Maingueneau); and discourse analysis and his-
tory (Marie-Anne Paveau). All three contributors enable us to
grasp how these have evolved and to understand how these evo-
lutions are indissociable from debates, within each discipline
and between disciplines, about what constitutes a “corpus” or an
“archive”, a “memory” or a “history” (Marie-Anne Paveau); “lite-
rature” and “literary discourse” (Maingueneau); or “subjects”,
“actors” and “public problems” (Claire Oger).
Introduction 7
– A methodological contribution. Although present in all the arti-

cles, this is nevertheless more explicit in four of them. The first
takes the form of a critical look back at an interdisciplinary re-
search practice between discourse analysis and comprehensive
sociology (Claire Oger and Caroline Ollivier-Yaniv). The sec-
ond is an illustration of a “way of doing” discourse analysis (Alice
Krieg-Planque), while the third reflects on discourse analysis and
its close relationship to lexicometric quantification and enuncia-
tive approaches (Pierre Fiala). The fourth presents a combina-
tion of an argumentative approach and a political science ap-
proach (Juliette Rennes). These contributions thus show that the
meeting of various disciplines may be experienced differently
depending on whether one is starting from a “linguistically de-
scribable” question whose social and political implications are
understood (Alice Krieg-Planque), or from an ideological ques-
tion (Juliette Rennes), or from a collaboration between two peo-
ple in which a sociologist and a discourse analyst are led to fre-
quent the same institutional corpora (Claire Oger and Caroline
Ollivier-Yaniv).
However our readers choose to use these texts, we hope that this
present volume will help show that, although clear theoretical posi-
tions are necessary as far as epistemology is concerned, it is the meet-
ings and exchanges between disciplines which are productive in mak-
ing research contribute to our knowledge of society. This, at any rate,
is the path that Céditec wishes to follow, as a team which is open not
only to the whole academic world, but also to the interested public.
Chapter I
Discourse analysis in Europe
Johannes ANGERMÜLLER
Introduction
Since the late 1960s, a variety of trends in discourse analysis have

developed in Europe. Given this diversity, it is not easy to gain an
overview, especially since discourse analysis does not have its own
disciplinary locus [Ehlich, 1994]1. Roughly speaking, the evolution
of this heterogeneous and relatively vaguely-defined field has been
characterised by an intersecting of the different strands, each of which
has its own national significance. Until the 1970s, relatively homo-
geneous fields were developing in coexistence in certain countries,
but by the 1980s these “schools” were beginning to produce more
and more branches outside their countries of origin.
In order to clarify the rather ambiguous situation of discourse
analysis in Europe, I propose to distinguish three major trends –
French, Anglo-Saxon and German – which relate to ancient tradi-
tions of thought, and of which followers can now be found all over
Europe. Although these trends are no longer attached to any particu-
lar country, they serve as a theoretical basis for the various “clusters”
of researchers who have dominated the discourses on discourse in
Europe since the 1970s: “the French school”, poststructuralist dis-
course theory, critical discourse analysis and interpretive discourse
analysis.
1 I thank Jacques Guilhaumou, Reiner Keller and Dominique Maingueneau, as

well as the two editors of this work, for their valuable comments.
10 Johannes Angermüller
Three major theoretical trends
The “French” trend
The “French” trend was inspired by the 1960s controversy over struc-
turalism. It combines the Saussurian viewpoint (1962) with the psycho-
analytical criticism of the “speaking subject” [Lacan, 1978] and a
Marxist analysis of “ideology” [Althusser, 1965]. Discourse analysis
“à la française” is distinguished by its vision of a rigorous and exhaus-
tive description of the life of signs within a society. In view of the
transpersonal organisation of language activity, this trend gives prior-
ity to the written text and to large groups of texts; their underlying rules
of construction need then to be discovered and analytical models ob-
tained. Towards the end of the 1970s, there was a change of direction
in France with the decline of structuralism and a turning towards prag-
matism. From then on, thinking no longer focused upon the Saussurian
duo of langue and parole (language and word), but on the problem of
enunciation, that is to say the rules which cause language acts to be-
come facts of discourse. Although linguistics has a considerable hold
over this disparate field, it also includes many specialists in informa-
tion and communication as well as sociologists and historians, who are
distinctive for the epistemological break they have made with their
object of study and the fact that they have accorded priority to the
materiality of discourse.
The “Anglo-Saxon” trend
The “Anglo-Saxon” notion of discourse draws its inspiration from

American pragmatism and English analytical philosophy, particularly
in the theory of language acts [Austin, 1962]. Unlike the French
enunciative approaches, discourse refers to the level of the language
act in a given communication situation [Levinson, 1983]. Thus Ameri-
can discourse analysis examines the rules which organise interac-
Discourse analysis in Europe 11
tions and conversations between actors [Brown and Yule, 1998]. While
conversation analysts describe the way in which partners end up with
a shared framework /understanding of the situation by each taking
their turn in discursive interaction, ethno-methodologists reveal the
implicit knowledge governing daily interactions [Cicourel, 1973]. In
view of the insistence on deictic and polyphonic organisation of dis-
course, American conversation analysis and ethnomethodology have
made a crucial contribution to the evolution of linguistic pragmatics.
The objects of investigation may be encounters between actors in a
situation where a conflict has to be negotiated. In the United King-
dom, on the other hand, it was M. A. K. Halliday’s functionalist lin-
guistics which enabled a great number of linguists to analyse the uses
of text within society. As the “Anglo-Saxon” trend is characterised
by a clear grounding in empirical material [cf. Grounded Theory,
Glaser and Strauss, 1967], it has given rise to numerous pieces of
research applied to communication problems arising in different in-
stitutional contexts (professional spheres, hospitals or prisons, for
example).
The “German” trend
In Germany, emphasis has for a long time been put on a theory rather
than a method of discourse. Thus Jürgen Habermas’s theory of the
“communication act”, influenced by pragmatic Anglo-Saxon trends,
aims at a model of the conditions for a critique of authority and in-
equality. According to Habermas and his followers, when we com-
municate, we cannot fail to recognise certain rules of discourse, such
as the discursive partner’s equality and the “criticability” of each ar-
gument. These rules are based on a consensus between partners which
serves as a common measure for criticising arguments put forward in
the discourse [Habermas, 1981]. There have been many attempts to
implement Habermas’s discursive ethics in empirical social research.
In social sciences, for example, political discourses have been ana-
lysed in the light of the democratic claims of modern societies. How-
ever, although in Germany the Habermas paradigm has been in de-

cline since the 1990s, discourse has become the object of “compre-
hensive sociology” and more particularly of the “sociology of knowl-
edge” [Knoblauch, 2005] in the phenomenological and interpretive
tradition of Berger and Luckmann [1966] and Max Weber [1921].
Extending the hermeneutic praxis which consists in reconstructing
meaning shared by a community, the representatives of this trend
understand “discourse” to mean sense (Sinn), and knowledge or inter-
pretation (Deutungen) which uphold the unity of a social and cultural
order. Thus discourse refers to the implicit cultural knowledge stored
in a society’s documents and texts. The comprehensive approach is
directed more towards a description of the empirical world and shares
Habermas’s hypothesis that it is discourse which allows discursive
partners to find a common cultural and interpretive meeting ground –
that of the “life world” with a “shared social meaning”. However, it
should be said that discourse analysis in Germany – because of the
role played there by philosophy and sociology – has had more diffi-
culty setting itself up as a field of investigation than in some other
countries, notably France [Angermüller, 2005a].
These three trends, to which we should perhaps add Slavo-Russian

semiotic trends [Mikhail Bakhtine, Youri Lotman], provide a strong
testimony to the diversity of different, even opposing, ideas of dis-
course which currently exist in Europe. They give an initial overview
which will enable us to look in more detail at the people and projects
involved in European discourse analysis.
The major “clusters” in Europe
While these three or four major trends are inherited from long tradi-
tions of thought, the situation of discourse analysis in Europe since
the 1970s has often been sustained by a few “clusters”, bringing re-
searchers together around a particular theoretical position. Today, these

clusters are nodal points where many traditions come together to set
the problem of language activity back in its context.
The French school of discourse analysis
Discourse analysis was established in France in the late 1960s against

the background of the expansion of the social sciences and of their
“pilot science”, Ferdinand de Saussure’s formal linguistics. Linguis-
tics became acquainted with discourse through the work of Zellig
Harris [1963], but it was a few theoreticians in the wake of structur-
alism who introduced it to the intellectual public. Thus, the starting
point for the followers of Michel Pêcheux was the problem of State
ideological apparatus. At the same time, they extended the questions
posed by Lacanian psychoanalysis to a discursive theory of subjectiv-
ity. Michel Foucault, on the other hand – a rather unique figure – was
interested in the problem of “discursive formations” [1969], that is to
say the rules organising the production of utterances. These trends are
sometimes classified under the label of “French school of discourse
analysis” – a term which is certainly incorrect, if account is taken of
the many different positions which exist in France [Maingueneau,
1996]. Among the research groups which have developed since the
1960s in the Paris region, there is the Paris X group (Nanterre) gath-
ered around Jean Dubois; the centre for research in political lexicology
at the Ecole Normale Supérieure in St Cloud; and the Maison des
Sciences de l’Homme, which acknowledges theoreticians like, for
example, Roland Barthes, Pierre Bourdieu and Michel de Certeau.
Since its beginnings in the 1960s, discourse analysis in France has
become distinctive for its tradition of automatic or lexicometric analy-
sis. Software like Lexico 3 (André Salem) or Alceste (Max Reinert)
bears witness to the technological achievements of lexicometry and
corpus linguistics. Although lexicometric analysis aims to identify
regularities, specificities and correlations of graphic forms in a cor-
pus, it is not devoid of intellectual and methodological ambitions.
Researchers are often closely involved with political or trade union

militancy and are interested in analysing political discourse; they also
place an emphasis on methodological and theoretical thinking.
Following the “algorithmic” view of discourse which predomi-
nated at the outset [see Pêcheux, 1969], the 1970s and 1980s opened
up new areas of work such as pragmatico-enunciative linguistics, his-
torical studies [Robin, 1973] and, more recently, the new media, yet
without renouncing the priority accorded to graphic forms of dis-
course and criticism of the “ideology” of the speaking subject. While
theoretical and political divisions have decreased, the directions in
which theory has developed have increased. Thus, through interac-
tion with other European trends, a pragmatic tendency (for example
Dominique Maingueneau) and an interpretive tendency (for example
Jacques Guilhaumou) have erupted onto a scene in which inter-
disciplinarity is a prominent feature. The research centres now active
in the area of discourse analysis have gathered together numerous
researchers from the language sciences and social sciences, as well
as from history and from information and communication. Among
these are the CEDISCOR, Paris III, for example, directed by Sophie
Moirand; the CAD, directed by Patrick Charaudeau; the GTAD at
EHESS, founded by Pierre Achard; and Céditec, Paris-Est Créteil,
directed by Simone Bonnafous. All these centres are evidence of a
highly productive field [cf. Mazière, 2005; Sarfati, 1997].
Poststructuralist discourse theory
Another cluster based on the philosophy of Althusser was formed

from the 1980s onwards around the Master’s programme Ideology
and Discourse Analysis at the University of Essex. Under the direc-
tion of an Argentinian émigré, Ernesto Laclau, and Chantal Mouffe
[Laclau and Mouffe, 1985], this programme became the main Anglo-
Saxon hub open to discourse theories and, more generally, to a whole
array of French political theories and philosophies. Critical of a de-
terminist Marxist view, the Essex theoreticians were concerned to
describe the way in which, within an antagonistic space, hegemonies

are produced through contingent acts of discourse [Laclau, 1990].
With British Cultural Studies, conceived and promoted by Stuart Hall
and his colleagues in Birmingham [Hall, Hobson, Lowe and Willis,
1980] together with representatives of the journal Screen [Coward,
1977, for example], these representatives of what is traditionally called
“poststructuralism” examine the “representation” régimes of postmod-
ern societies based on the theoretical heritage of Saussure, Marx and
Lacan [Lee, 1992; Mills, 1997]. In the tradition of Althusser, the
English poststructuralists understand “discourse” to mean ideology
and culture, whose heterogeneous and plural nature they stress [Saw-
yer, 2002]. Although they are less interested in linguistic tools and
analyses than in epistemological thinking (hence their propensity for
discourse theories), they were part of the poststructuralist movement
emerging from American social science. In this, textual deconstruction
(the Yale School, following the example of Paul de Man) and the
problems of identity movements – such as the critique of Orientalism
undertaken by Edward Said [1978] – have a tendency to dominate
discursive approaches. Nowadays, the poststructuralist trend has a
strong presence in many European countries, particularly in the areas
of cultural and feminist studies and social sciences [cf. Angermüller,
2007; Bublitz, 2003; Stäheli, 2000; Torfing, 1999].
Critical Discourse Analysis, CDA
A third group has formed with the aim of criticising power relations
in existing societies and making a “positive” social change. Practised
by linguists with social and political ambitions, this approach is dis-
tinguishable from the other two – which are also grounded in political
trends – by its objective, which aims to measure and unmask the
“ideology” of political discourse [Wodak and Meyer, 2004]. The ob-
jective is thus to identify the “ideological” content of a discourse
according to an explicit normative standard. As with the other two
trends, references to the work of Foucault are present – especially with
Siegfried Jäger [1993] and Norman Fairclough [1992] – but the meth-
ods differ from the linguistic approaches predominating in France
insofar as they aim to reveal the “ideological” aspect of a text. While
Teun Van Disk practises text linguistics [1988], the others are often
close to interpretive sociology (known as “qualitative” in England and
Germany, for example Wodak [1996]). Jürgen Link, a pioneer of dis-
course analysis in Germany, is interested in the collective symbols of
society which lead to its “normalisation” [1997]. Among the tradi-
tional subjects belonging to this trend, it is possible to cite analyses of
anti-Semitic, racist or sexist texts which are produced by government
authorities or feed into media discourse.
Interpretive discourse analysis

(“wissenssoziologische Diskursanalyse”)
A fourth group revolves around the German sociologist Reiner Keller

and his colleagues at the University of Augsburg, who combine
Foucault’s approach with the sociology of knowledge [2005]. The
sociology of knowledge forms part of the German phenomenological
and hermeneutic traditions, and so the relationship between the sym-
bolic and the social is not a new problem. Keller adopts the interpretive
approach of “qualitative” sociology in order to recreate the collective
meaning of a text. While “qualitative” tools like “grounded theory”
[Glaser and Strauss, 1967] ensure flexibility and openness of the ana-
lytical procedure, Foucault’s work provides a benchmark to stress the
interpersonal organisation of discourse. While Keller et al. give prior-
ity to the shared bases of understanding in discursive exchange, they
mean by “discourse” the store of collective knowledge which the
actors mobilise in order to build social relations. Inscribed into texts
by the members of a “life world” (Lebenswelt), sense and knowledge
(Sinn) can be reconstructed by the analyst in an act of understanding.
Unlike more linguistic approaches, the graphic form disappears when
confronted by the sense that it carries for the social actors. The inter-
pretive approach lies this side of the debate over the speaking subject.
This list is far from exhaustive. One would probably need to add to it
some trends in linguistics [for example the functionalism of Konrad
Ehlich and the historical semantics of Dittrich Busse, 1987], in psy-
chology [Jonathan Potter’s cognitivism, Potter and Wetherell, 1987],
in history [Reinhart Koselleck’s history of concepts, 1979], the Cam-
bridge School rhetoric (Quentin Skinner) and various approaches to
frame analysis [for example Konerding, 1993]. In addition to these,
one should also include the non-European groups who lay claim to
socio-criticism (Marc Angenot at McGill University in Montreal) and
American pragmatism. However, what appears to distinguish the four
trends mentioned above from these latter ones is that the former are
perhaps less rooted in their national and disciplinary fields and more
organised around a set of problems with a common ground. So, de-
spite the considerable methodological differences which exist between
them, these trends have consolidated around a few lines of thought
and the various references to Michel Foucault are an indicator of this.
It becomes evident that discourse analysis in Europe is not a field
organised along the lines of traditional disciplines. Because exchanges
between the different groups are not always easy, textbooks and dic-
tionaries play a crucial role in its development. More than mono-
graphs, it is a few compendiums and dictionaries which reveal the
shape of a field made up of diverse approaches and trends – works
like the four-volume Handbook of Discourse Analysis by Teun van
Dijk [1985], the two Handbücher sozialwissenschaftliche Diskurs-
analyse by Reiner Keller et al. [2001; 2003], the Dictionnaire d’ana-
lyse du discours by Dominique Maingueneau and Patrick Charaudeau
[2002] as well as journals like Mots, founded by Louis Bodin and
Maurice Tournier, Langage et société, founded by Pierre Achard,
Discourse & Society, Discourse Studies, edited by Teun van Dijk,
KultuRRevolution, edited by Jürgen Link and Discourse Analysis
Online, edited by Jonathan Potter et al.
Conclusion
If in European discourse analysis at least three major historical trends

can be identified – “French formalism”, “German hermeneutics” and
“Anglo-Saxon pragmatism” – most of these trends intersect in one
way or another by means of the clusters which bring researchers to-
gether across the whole of Europe. Despite having been open for the
last 25 years to the challenge of pragmatism, the “French school”
remains perhaps the most homogeneous as far as its grounding in
national tradition is concerned [Angermüller, 2005b]. In most of the
other clusters, existing national trends blend with the French trend,
where the name of Foucault often serves as a unifying element. Thus,
criticism of the speaking subject and of essentialism in poststructuralist
trends is strongly permeated by French discourse theories. Interpre-
tive and critical trends, on the other hand, are informed more by a
“qualitative” approach. Finally, the case of “critical discourse analy-
sis” reminds us of the limits of national labels in describing the field
of discourse analysis. Or perhaps this approach could be considered
as the most “Anglo-Saxon” insofar as it does not accord the same
priority to conceptual and epistemological work as the others? The
major divide is doubtless the one which exists between French
pragmatico-structuralist formalism on the one hand and the her-
meneutic and interpretive approaches on the other, these latter often
becoming consolidated in so-called “qualitative” research. Thus, if
the latter place greater importance on the identity of the actor as well
as on the unity of the symbolic ground, those following the French
tradition have a tendency to consider the subject as “divided” or re-
lating to “ideology” while at the same time stressing the constituent
rifts and antagonisms across the symbolic order. It is in that sense
that it is possible to mark out the inner boundary of the discourse
analysis field in Europe: at the point where some see a swarm of
signifiers (the “French school” and poststructuralism), others seek to
discover the signified (critical and interpretive analyses).
Bibliographical references
ALTHUSSER L., E. BALIBAR, R. ESTABLET, P. MACHEREY

and J. RANCIÈRE
1965: Lire le Capital, Paris, Quadrige/PUF.
ANGERMÜLLER J.
2007: “‘Qu’est-ce que le ‘post-structuralisme’ français?’ A propos de
la réception des tendances françaises de l’analyse du discours en Alle-
magne”, Langage et société no. 120: 17–34.
2005b: “‘Qualitative’ Research in France: Reconstructing the Actor,
Deconstructing the Subject.” Forum Qualitative Research, Electronic
journal, issue no. 6(3). <http://www.qualitative-research.net/fqs-texte/
3-05/05-3-19-e.htm>.
2005a: “Sozialwissenschaftliche Diskursanalyse in Deutschland: zwi-
schen Rekonstruktion und Dekonstruktion”, R. Keller, A. Hirseland,
W. Schneider and W. Viehöver (ed.), Die discursive Konstruktion von
Wirklichkeit, Konstanz, UVK, 23–48.
AUSTIN J. L.
1962: How to Do Things with Words. The William James Lectures
delivered at Harvard University in 1955, Oxford, New York, Oxford
University Press.
BERGER P. L. and T. LUCKMANN

1966: The Social Construction of Reality: A Treatise in the Sociology
of Knowledge, Garden City, New York, Anchor Books.
BROWN G. and G. YULE

1998: Discourse analysis, Cambridge, Cambridge University Press.
BUBLITZ H.
2003: Diskurs, Bielefeld, Transcript.
BUSSE D.
1987: Historische Semantik. Analyse eines Programms, Stuttgart,
Klett-Cotta.
CICOUREL A.V.
1973: Cognitive Sociology: Language and Meaning in Social Inter-
action, Harmondsworth, Penguin.
COWARD R.
1977: “Class, ‘Culture’ and the Social Formation”, Screen no. 18, 75–105.
DIJK T. A.
1985: Handbook of Discourse Analysis. 4 vols. I. Disciplines of
discourse. II. Dimensions of discourse. III. Discourse and dialogue.
IV. Discourse analysis in society, London, Academic Press.
1988: News as Discourse, New York, Lawrence Erlbaum.
EHLICH K.
1994: Diskursanalyse in Europa, Frankfurt am Main et al., Peter Lang.
FAIRCLOUGH N.
1992: Discourse and Social Change, Cambridge, Oxford, Polity Press.
FOUCAULT M.
1969: L’Archéologie du savoir, Paris, Gallimard.
GLASER B. G. and A. L. STRAUSS
1967: The Discovery of Grounded Theory. Strategies for Qualitative
Research, Chicago, Aldine.
HABERMAS J.
1981: Theorie kommunikativen Handelns, Frankfurt am Main, Suhrkamp.
HALL S., D. HOBSON, A. LOWE and P. WILLIS (ed.)

1980: Culture, Media, Language, London, Hutchinson.
HARRIS Z. S.
1963: Discourse Analysis Reprints, The Hague, Mouton & Co.
JÄGER S.
1993: Text- und Diskursanalyse: Eine Anleitung zur Analyse poli-
tischer Texte, Duisburg, DISS.
KELLER R.
2005: Wissenssoziologische Diskursanalyse. Grundlegung eines
Forschungsprogramms, Wiesbaden, 2005.
KELLER R., A. HIRSELAND, W. SCHNEIDER and W. VIEHÖVER (ed.)

2003: Handbuch sozialwissenschaftliche Diskursanalyse. Forschungs-
praxis, Opladen, Leske & Budrich.
2001: Handbuch sozialwissenschaftliche Diskursanalyse. Theorien
und Methoden, Opladen, Leske & Budrich.
KNOBLAUCH H.
2005: Wissenssoziologie, Konstanz, UTB.
KONERDING K.-P.
1993: Frames und lexikalisches Bedeutungswissen: Untersuchungen
zur linguistischen Grundlegung einer Frametheorie und zu ihrer An-
wendung in der Lexikographie, Tübingen, Niemeyer.
KOSELLECK R.
1979: “Begriffsgeschichte und Sozialgeschichte”, R. Koselleck (ed.),
Historische Semantik und Begriffsgeschichte, Stuttgart, Klett, 19–36.
LACAN J.
1978: Le moi dans la théorie de Freud et dans la technique de la
psychanalyse. Le séminaire, Livre II, Paris, Le Seuil.
LACLAU E.
1990: New Reflections on the Revolution of Our Time, London, New
York, Verso.
LACLAU E. and C. MOUFFE

1985: Hegemony and Socialist Strategy. Towards a Radical Demo-
cratic Politics, London, New York, Verso.
LEE D.
1992: Competing Discourses: Perspective and Ideology in Language,
London, New Left.
LEVINSON S. C.
1983: Pragmatics, New York, Cambridge University Press.
LINK J.
1997: Versuch über den Normalismus. Wie Normalität produziert wird,
Opladen, Westdeutscher Verlag.
MAINGUENEAU D.
1996: “L’analyse du discours en France aujourd’hui”, S. Moirand (ed.),
Le Discours: Enjeux et perspectives (special issue Le Français dans
le monde, July 1996), Paris, Hachette, 8–15.
MAINGUENEAU D. and P. CHARAUDEAU
2002: Dictionnaire d’analyse du discours, Paris, Le Seuil.
MAZIÈRE F.
2005: L’analyse du discours, Paris, Presses Universitaires de France.
MILLS S.
1997: Discourse, New York, Routledge.
PÊCHEUX M.
1969: Analyse automatique du discours, Paris, Dunod.
POTTER J. and M. WETHERELL
1987: Discourse and Social Psychology, London, Sage.
ROBIN R.
1973: Histoire et linguistique, Paris, Colin.
SAID E.
1978: Orientalism, London, Penguin.
SARFATI G. E.
1997: Eléments d’analyse du discours, Paris, Nathan.
SAUSSURE F.
1962: Cours de linguistique générale, Paris, Payot.
SAWYER K.
2002: “A Discourse on Discourse: An Archeological History of an
Intellectual Concept”, Cultural Studies no. 16, 433– 456.
STÄHELI U.
2000: Sinnzusammenbrüche. Eine dekonstruktive Lektüre von Niklas
Luhmanns Systemtheorie, Weilerswist, Velbrück.
TORFING J.
ï iz
1999: New theories of discourse: Laclau, Mouffe and Z ïek, Oxford,
Blackwell.
WEBER M.
1921: Wirtschaft und Gesellschaft. Grundriß der verstehenden Sozio-
logie, Tübingen, Mohr.
WODAK R.
1996: Disorders of Discourse, London, Longman.
WODAK R. and M. MEYER
2004: Methods of Critical Discourse Analysis, London, Sage.
Chapter II
Discourse analysis and information

and communication sciences: beyond corpora
and methods1
Claire OGER
The meeting between discourse analysis (DA) and information and

communication sciences (ICS) has long been delayed by the respec-
tive directions in which each discipline has developed. However, it
now appears to be a fait accompli, as testified by the growth of re-
search in media discourse analysis, the existence of “mixed” research
teams, and the interest shown in both these research fields by jour-
nals such as Etudes de communication (Université de Lille III), Com-
munication (Université Laval, Quebec), Questions de communica-
tion (Université de Metz) or Mots. Les langages du politique (ENS
Editions). Nevertheless, working regularly with teachers and re-
searchers of ICS sometimes gives discourse analysts the impression
that this interdisciplinary contact is viewed on both sides in a
reductionist way, and even that it is based on misunderstandings.
My intention here is to briefly recall the conditions under which
the two disciplines have come together and refer to the viewpoint
which appears to me to be reductionist, after which I will propose
two examples which indicate more profitable ways of working to-
gether.
1 This chapter is based on a paper given at the conference “Sciences du Langage et

Sciences de l’Homme” organised by the Association des Sciences du Langage,
10 December 2005, École Normale Supérieure, Paris. It has been further de-
veloped thanks to exchanges with Alice Krieg-Planque and the other members of
Céditec.
26 Claire Oger
An overdue meeting between DA and ICS
An abundant bibliography has been devoted to the conditions which

have delayed the coming together of DA and ICS2. I refer the reader
to this and will therefore limit myself here to a very rapid summary
recalling that both have evolved as mutually exclusive “disciplines”3,
from three main points of view:
The theoretical groundings
It has often been stressed [see in particular Bonnafous, 2006: 5] that

their divergent theoretical positions have been responsible for the
delay in bringing the two disciplines together: in the 1970s, all that
many linguists knew about communicational approaches was Jakob-
son’s over-celebrated schema which rightly appeared to them to be
inadequate for dealing with linguistic phenomena. And the critical
position of discourse analysts, based on the work of Foucault and
Althusser, made them highly mistrustful of American mass commu-
nication research, which was largely financed by the American govern-
ment, as well as by the press and advertising.
2 Cf. Bonnafous (1992 and 2006), Bonnafous and Jost (2000), Bonnafous and
Charaudeau (1996), Meyriat and Miège (2002), Miège (2000), Tétu (2002).
3 This is not the place to question the appropriateness of this definition for DA
and/or for ICS nor the possible alternatives (research field, interdiscipline, etc.)
nor the asymmetry which has led ICS researchers to demand the creation of an
independent section of the CNU (Conseil National des Universités) while dis-
course analysts are forced to move from one department to the other, including
from the 7th to the 71st. Let us simply point out that thinking about institutional
designations and labels is directly connected to the thinking we are summaris-
ing here.
Discourse analysis and information and communication sciences 27
“Supporting disciplines”
In addition to becoming part of the language sciences, discourse analy-

sis has been built up in close partnership with what D. Maingueneau
[2005] calls the “supporting disciplines”: history, psychoanalysis and
philosophy. For ICS, on the other hand, in their search for institu-
tional recognition, the other disciplines were more of a breeding
ground for recruitment than research partners; and the teachers, then
the researchers, who arrived to swell the ranks of ICS were often
literary academics or sociologists [Miège, 2000; Tétu 2002].
The preeminence of teaching or research
A final point is that while the discourse analysis project was first
formulated in terms of theoretical research, ICS originated from a set
of problems to do with teaching, particularly vocational teaching
[Miège, 2000; Meyriat and Miège, 2002]. This involved the setting
up of courses in expression and communication on the one hand, and
the creation of University Institutes of Technology in documentation
and journalism on the other.
Thus, the belated rapprochement between DA and ICS was only
possible thanks to simultaneous developments in both disciplines:
whilst critical research was gradually evolving on the ICS side, DA
was turning its attention to pragmatic problems and problems of enun-
ciation. In this movement, the concept of the subject shifted, distanc-
ing itself from the conceptions of structuralism to give the locutor a
margin of strategy and control of meaning that the founders of DA
refused to give. This shift allowed discourse analysts to introduce a
“new interdisciplinarity”, this time at the crossroads of the language
sciences, sociology, political science… and the ICS [Bonnafous, 2006].
However, at the very time when DA and ICS were coming to-
gether through the development of research objects or the institu-
tional trajectories of researchers, with the formation of the teams and
journals referred to above, everyday relationships between teachers
28 Claire Oger
and researchers in the two disciplines were often reduced to disap-

pointing squabbles.
The “corpus versus method” dialogue
Too many linguists continue to think that discourse analysts who have
moved in the direction of ICS have simply followed a more general
trend, distancing DA from political and trade-union discourse4, which
has long been its priority, and causing it to turn towards media dis-
course corpora. Sociological considerations about the importance of
communication phenomena can here stand alongside thinly veiled
accusations of opportunism; in doing this, such discourse analysts
are said to be surrendering only to “the mood of the moment” or to
material career necessities.
As for teachers and researchers in ICS, they sometimes suspect
discourse analysts of indulging in the obscure jargon that they readily
impute to linguists. Or, very often, more benevolently, they tend to see
them as service providers: since the mass of available texts never
ceases to grow and the development of communication technologies
makes these an ever more cumbersome object of research, discourse
analysts arouse expectations in terms of “analysis charts” or “methods”
that are likely to extract meaning from these masses of discourse.
What is more, the confusion between discourse analysis and con-
tent analysis remains surprisingly intense. Since it is impossible to list
here all the clarifications which have been attempted, I will refer only
to the underlying thinking of R. Robin who, as early as 1973 pointed
out that studies carried out in content analysis, “beyond their rigorous-
ness and their immense merits, are nevertheless based on the assump-
tion that the sense is immediate and is unequivocal” [Robin, 1973: 62].
4 If this is understood in its strictest sense as the discourse of politicians or political

organisations.
At the other end of a long bibliography which is perfectly con-

sistent on this point, the recent reflections of S. Bonnafous [2006]
again echo the need to break free from content analysis as well as
from the principles of thematic pre-categorisation and synonymic
functioning which govern it, and which remain unacceptable to dis-
course analysts.
More generally, Alice Krieg-Planque [2000 and 20065] has em-
phasised the fact that discourse analysis, in the area of media studies,
can in no way be reduced to a set of techniques for accessing “repre-
sentations” or “images”. And D. Maingueneau [2005], as well as J.-
M. Utard [2001 and 2004], have recently reminded us that discourse
analysis cannot be reduced to a “method” and that its roots and fun-
damental theoretical choices should be stressed in order to explain its
specificity.
Therefore the representation of relations between DA and ICS as
a kind of exchange between “corpus and method” should immedi-
ately be condemned for epistemological reasons. One could add other
less weighty arguments. The collecting of corpora in no way entails
the need for an alliance, and all those discourse analysts who have
gathered together political, trade-union and institutional texts have
certainly not used them as a reason for actually collaborating with
the organisations or institutions studied, nor with the disciplines in-
volved in them: when discourse analysts apply themselves to the study
of legal or medical texts, they are perfectly well able to gather informa-
tion about the context of their enunciation without any need for co-
operation with legal experts or doctors. And many of them examine
media corpora while keeping their distance from research carried out
in ICS.
If the meeting between DA and ICS is neither inevitable nor an
intrinsic requirement of the objects of study, it nevertheless seems
possible to envisage a more fruitful way of proceeding by going
beyond the terms of a reductionist exchange of “corpus for method”.
5 See also her contribution to this present volume.

30 Claire Oger
I would here like to illustrate two directions which would seem to

indicate that familiarity with the problems of ICS has the potential to
develop the thinking of discourse analysts and enrich it with regard
to a few points which lie at the heart of their approach – like the
constitution of corpora, the conception of the subject of enunciation,
or the circulation and institutionalisation of discourses.
Communicational approaches
The first direction was suggested by the research I have been con-
ducting for several years in collaboration with C. Ollivier-Yaniv. Based
on questioning about communications in the military establishment
and relationships between the armed forces and journalists6, this has
led us to propose that interpretive sociology and discourse analysis
should be brought closer together. The main stages of our coopera-
tion to date have involved comparison of our respective results, con-
ceptual dialogue and methodological hybridisation. We will not go
here into the detail of this development, which is described in an-
other chapter of this book7.
There is just one further point to mention here: it does not seem
at all fortuitous that this shared thinking has been done in the context
of ICS. As members of one of these “mixed” teams referred to, both
recruited from 71st department, our situation reflects the special na-
ture of ICS. This has been pointed out by E. Neveu and L. Quéré
6 It should be stressed that an interest in communication phenomena does not com-

pel the DA researcher to turn to media discourse (and still less towards journalis-
tic discourse). The definition of ICS, as a result of their institutionalisation in
France, enables the researcher to approach a very wide spectrum of subjects
(communication of institutions and organisations, discourses produced on websites
by communities of all kinds, museums and so on).
7 See also Oger and Ollivier-Yaniv (2003 and 2006), and Oger (2000 and 2003).
[1996: 16] and B. Miège [2000: 557–559], who indicate that the rap-
prochement between sociology and discourse analysis is particularly
relevant for the study of communication phenomena.
Indeed, according to B. Miège, the study of communication
phenomena calls for the implementation of “inter-scientific research
methodologies” of which “the most exemplary case and also the most
frequent is the interconnecting of social discourse analysis with analy-
sis of the strategies and practices of social actors”. If B. Miège’s re-
marks serve to underline how “the connection between social sci-
ence problems […] and the problems of language sciences and
discourse […] turns out to be a productive approach” in the context
of ICS, for our part we would like to stress the contribution it can
make to discourse analysis.
If our work provides an illustration of this, the possibilities for
interdisciplinary cooperation opened up by the study of communica-
tion phenomena would appear to be inexhaustible.
Diane Vincent, querying the instance of “trash radio” discourse,
makes a plea for a decompartmentalisation between disciplines and
specialities, showing “how linguists, discourse analysts, conversa-
tionalists, sociologists, etc. (combined in a single person or convened
as a team) can contribute their knowledge to (de)construct social dis-
courses, according to need” [Vincent, 2005: 173]. Isabelle Laborde-
Milaa [2002], in a study of media genres (such as portraits or literary
criticism) based on discourse analysis and text linguistics, sets her
interest in these within a framework which envisages genres as so-
cially situated norms and practices. In such cases, the breaking down
of internal barriers to the “discourse disciplines” [Maingueneau, 2005]
is a response to the rapprochement with sociology – the discourse
disciplines being discourse analysis, conversation analysis, rhetoric,
text linguistics etc.
In other cases, the study of discourses produced on websites might
suggest that discourse analysis should be brought together with inter-
faces, as might be envisaged by the sociology of uses. A recent exam-
ple of pedagogical cooperation provides an illustration of this. As
part of professional Master’s degree teaching, devised with two col-
32 Claire Oger
leagues from the communication sciences research and teaching unit

at the University of Paris 13 (Emmanuel Paris and Geneviève Vidal),
we studied websites devoted to the memory of Auschwitz. This led
us to compare analysis of the interfaces (architecture as a whole, place
and forms of interaction and interlocution) and analysis of the dis-
courses offered on the sites (ethos, situation in relation to history and
memory, forms of discursive authority, institutional inscriptions) with
analyses of the interviews carried out with site designers (justifica-
tion of the project, declared objectives, meta-discursive comments
and so on).
This type of approach would seem to contribute to discourse
analysis thinking in at least two respects:
As regards corpora
I have mentioned elsewhere the way in which the “socio-anthropo-

logical” approach to discourses led to the construction of heteroge-
neous corpora in which discourses produced according to different
systems of constraint may be, not aligned or assimilated, but com-
pared [Oger and Ollivier-Yaniv, 2003; Oger, 2005].
In some cases the heterogeneity of the corpus remains locked
within the limits of the pre-existing institutional discourse: for exam-
ple, while working for my thesis [Oger, 2002] I was able to compare
the discourse produced in the reports of exam juries with an analysis
of secondary corpora: examination subjects, distance-learning guid-
ance notes, best answers and the like.
In other cases, the hybridisation of methods leads to analysis of
the discourses produced during interviews initiated by the researcher,
as was suggested by me with regard to military communication [Oger,
2003]. In the same way, in her analysis of a contemporary reading
model, Julia Bonaccorsi musters a very heterogeneous corpus which
extends across inscribed discourses (scientific research on reading,
reports and circulars, records of conference proceedings) as well as
non-inscribed ones (mediation practices) produced over a wide field –
scientific, public, legislative, political, within associations and so on

[Bonaccorsi, 2004]. Christine Barats, who is interested in the incor-
poration of ICT/ICTE8 into higher education, adopts the same stand-
point, conducting her research on the basis of a heterogeneous cor-
pus: ministerial flow charts, reports and texts emanating from
government or the Ministry of Education, interviews with actors, site
analysis, contracts and internal documents of institutions as well as
monographic studies of institutions [Barats, 2005].
Consequently, the question of how corpora are constituted (defi-
nition, typology, limits) arises once again, while the various aspects
of the language-word-discourse triangle demand renewed attention.
For example, the question of knowing under what conditions the words
of social actors (recorded during interviews, or noted in discussion
forums and blogs) should be considered as word or as discourse does
not have a simple answer, unless we use categories of analysis that
have become outdated, or very confused, by recent developments in
communication.
On the notion of subject
This type of work can also alter and refine questioning on the subject
of enunciation: sociologists, particularly when interested in the mean-
ing that actors give to their activity, tend to marginalise the hypoth-
esis of a subject to whom his or her own discourse may remain opaque.
Although, as has already been pointed out, DA has caused the notion
of subject to evolve by moving away from structuralist conceptions,
a much more circumspect approach to the notion of subject is still a
strong feature of it.
This question has been very deeply analysed from a theoretical
and epistemological viewpoint, both in discourse analysis and in so-
ciology [Authier, 1995: 67–85 and Lahire, 1998: 23, 227–231]. But
it very often remains unresolved in individual corpus analyses. Ex-
8 Information and Communication Technologies for Education.

34 Claire Oger
actly what measure of autonomy and control of meaning is attributed

to the subject? How far are constraints appropriated and what mean-
ing is given to this “appropriation”? Are constraints internalised or
are they margins for manoeuvre? The uncertainties on these points
are rarely raised explicitly and boundaries are rarely marked out as
closely as possible to the utterances themselves.
A closer attention to sociological concepts, problems and methods,
far from being the discourse analyst’s downfall, might well give rise
to fresh questioning about the practices of discourse analysis.
Construction of public problems
Along the same lines of thinking, the interest shown in media dis-
course in no way binds discourse analysts within the limits of media
discourse corpora in the strict sense, nor does it constitute a shift of
focus, which would have distanced them from institutional and poli-
tical discourses. On the contrary, I would like to suggest here a way
in which the approach to the circulation of discourses and their poli-
tical implications might be enlightening for discourse analysts who
set their research and/or teaching in the context of ICS.
It is well know that the “construction of an event” has been the
object of much attention in media discourse, as has thinking about
notions such as “discourse event” in the public arena9. However,
E. Neveu and L. Quéré [1996] suggest not limiting this approach to
the actual construction of media discourse (information by selection,
grading of information, etc.), but, in addition, working at a prior stage
to media discourse, to take in a more all-embracing set of problems.
9 I will not develop this point here but refer the reader instead to the first two
chapters of Mouillaud and Tétu (1991) and to Krieg-Planque (2003). Also (cf.
infra) to numbers 75 and 76 of the journal Réseaux (1996) devoted to the “time
of the event”.
Alice Krieg-Planque’s [2006] suggestion to capture discourses

through the “discursive places” which they reveal and circulate is a
way of applying this approach: circulars, reports, underlying texts,
phrases, formulae [Krieg-Planque, 2003] are so many types of “dis-
cursive place” and studying them (for appearances, reformulations,
reprises, shifts, producing counter-discourses, etc.) enables us to un-
derstand certain of the political and moral positions which shape public
debate.
Similarly, D. Céfai [1996] is interested in the “construction of
public problems”: by this we understand that the way in which “pub-
lic problems” are constructed into discourses embeds them within
“frameworks of relevance” before they even reach the media. This
influences the way in which they are interpreted, as well as suggest-
ing designations and categories of analysis and imposing discursive
formats and evaluations. In short, “the material armature of the pub-
lic arena” creates a scene which strongly conditions the construction
of the event in the strict sense.
It is not possible to repeat here the detail of D. Céfai’s argument,
so I would prefer to give an example of the way in which the final
pages of his article inspired me with an idea for a discourse analysis
project10.
I used this framework of analysis to direct my attention to the
discourses produced in the context of the CNHI project (Cité
nationale d’histoire de l’immigration, a museum of immigration his-
tory which opened in Paris in 2007, in the Palais de la Porte Dorée).
In fact, this example shows the limits of research centred exclusively
on a press corpus (indeed, journalists have not so far proved very
forthcoming on the subject); or even on official institutional dis-
course, as illustrated and in large part summarised by the Rapport
sur la mission de préfiguration du Centre de resources et de mémoire
10 This project was carried out in the context of teaching a Master’s 2 professional
degree (Exhibition Design and Management at the Université Paris 13) and is
based on a close link between teaching and research. It has not yet been pub-
lished.
36 Claire Oger
de l’immigration11. The theoretical framework provided by D. Céfai

enables us to resituate the “polished” discourse of institutions [Oger
and Ollivier-Yaniv, 2006] in a broader environment. The project of
CNHI is a very good illustration of the phases described by D. Céfai
in one of his examples:
– A phase which we will call “discursive effervescence”, during
which individuals or groups have carried on a discourse on the
perception of discrimination, as well as on the memory of colo-
nisation and/or immigration, without forgetting slavery. We are
dealing here with a very wide spectrum of discursive manifesta-
tions – as much in terms of themes as of enunciators, ideologies,
etc. – which extends from the taking up of positions by histori-
ans like G. Noiriel and P. Weil, via the Call by the Natives of the
Republic to polemics on the sketches of Dieudonné.
– A formative phase in which the group of actors carrying on these
discourses comes together. These range from associations like
Génériques or Devoirs de mémoire to the more recent CAPDIV
(Cercle d’action pour la promotion de la diversité en France
[Action Circle for the Promotion of Diversity in France]) or
CRAN (Conseil representative des associations noires de France
[Representative Council of Black Associations]).
– A third phase – when public authorities intervene – can be illus-
trated in particular by the CNHI project. The report, like the four
interviews carried out with officials of the future museum12, pro-
vides striking evidence of the search to produce a consensus (and
even a “discourse of truth”) on this subject13. The work of insti-
tutionalisation being carried out in this phase can be seen and
11 This report, submitted to the Prime Minister by Jacques Toubon, was published
in 2003 in the “Rapports officiels” collection of Documentation française.
12 These four interviews, lasting about two hours, were carried out between Sep-
tember and December 2005 in collaboration with E. Paris (U. Paris 13).
13 Note that in the “Toubon Report” a performance element is linked to this consen-
sus, since it is incumbent upon the discourse produced for the exhibition trail and
the functioning of the Museum to recreate a “national cohesion” that is lost or
threatened.
evaluated in light of the comparison between the body of the

“Toubon Report” itself and its Annex II, which gives an account
of the debates and questions which have arisen in the different
assemblies and meetings during the project’s development phase
(Scientific Committee, but also Associations Forum, 2003 con-
ference, etc.).
– Finally, in this Annex II, a number of associations repeat the de-
mand to be co-workers in the project and on several occasions
express their fear of becoming no more than marginalised part-
ners. This is not the only indication of the opening of a fourth
phase. Indeed, without waiting for the Museum’s inauguration (in
late 2007), and without waiting for the public’s initial reactions to
the proposed exhibition trail and the choices made, an article was
published showing evidence of the way in which a consensus can
break down, revealing tensions and prominent fault lines. Pub-
lished under the title “Immigration et Histoire” on the 29th of
November 2005, and signed by the historians of the CNHI His-
tory Committee (including G. Noiriel, P. Weil and M.-C. Blanc-
Chaléard), this very revealing article denounced the remarks about
immigration made a short time before by ministers in post. In a
rather inflammatory parallel, it compared them unhesitatingly to
the extreme right-wing ideology of the 1930s.
It is thus possible to see how partnerships involving historians or
associations in an institutional project need to be linked to group
enunciation and circulation of discourses, which cannot be envis-
aged within the narrow framework of a homogeneous corpus. The
comparison of the different levels of institutional discourse (body of
the “Toubon Report”, Annex II of the same Report, interviews with
officials of the CNHI) with articles published in the national press
(articles written by journalists on the one hand and by intellectuals
and experts on the other) and discourses produced on websites (of
associations and institutions) reveal several factors. Among these is
the centrality of questions about the role of associations in the project,
the importance of conflicts surrounding its definition and their links
with the changing designations put forward (from “Resource and
38 Claire Oger
Memory Centre for Immigration” to “National Museum of Immigra-

tion History” via “Museum of Immigration History and Cultures”).
As we have seen, this subject echoes several others in this book
and I will finish by referring to the chapter by Juliette Rennes, who
shows what discourse analysis can contribute to a sociological and
communicational approach to controversies. I have attempted to stress
here the advantages that lie in encouraging the development of dis-
course analysis in the line of communicational problems.
AUTHIER-REVUZ J.
1995: Ces mots qui ne vont pas de soi. Boucles reflexives et non-
coïncidences du dire, Paris, Larousse, Volume I.
BARATS C.
2005: “Les dispositifs des TIC dans le supérieur: discours institution-
nels et monographie – Promesse, menace et visibilité”, paper given
at the SIF (séminaire industrialisation de la formation) Conference
“Les institutions éducatives face au numérique”, Maison des Sciences
de l’Homme Paris-Nord, 12–13 December 2005.
BONACCORSI J.
2004: Le devoir-lire. Métamorphoses du discours culturel sur la lec-
ture: le cas de la lecture oralisée, doctoral thesis in Information and
Communication Sciences defended 13 February 2004 at the Univer-
sity of Avignon.
BONNAFOUS S.
1992: “Linguistique et communication, une rencontre obligée”,
Cinémaction no. 63, March, 47–51.
BONNAFOUS S. and CHARAUDEAU P.

1996: “Les discours des médias entre sciences du langage et sciences
de la communication”, Le Français dans le monde, July, 39–45.
BONNAFOUS S. and JOST F.

2000: “Analyse de discours, sémiologie et tournant communication-
nel”, Réseaux no. 100, 525–545.
BONNAFOUS S.
2006: “L’analyse du discours”, in OLIVESI S. (ed.), Sciences de l’In-
formation et de la communication. Objets, savoirs, disciplines, Presses
universitaires de Grenoble, 213–228.
CÉFAI D.
1996: “La construction des problèmes publics. Définitions de situa-
tions dans les arènes publiques”, Réseaux no. 75, 43–66.
KRIEG-P LANQUE A.
2000: “Analyser le discours de presse. Mise au point sur le “discours
de presse” comme objet de recherché”, Communication, Laval Uni-
versity, Quebec, 75–97.
2003: “Purification ethnique”, une formule et son histoire, Paris,
CNRS Editions.
2006: “‘Formules’ et ‘lieux discursifs’: propositions pour l’analyse du
discours politique”. Interview with Philippe Schepens, Semen no. 21,
19–48.
LABORDE-MILAA I.
2002: “L’angle journalistique: aide ou obstacle au positionnement
des étudiants scripteurs?”, Pratiques no. 113–114, 95–113.
LAHIRE B.
1998: L’Homme pluriel. Les ressorts de l’action, Paris, Nathan.
40 Claire Oger
MAINGUENEAU D.
2005: “L’analyse de discours et ses frontières”, Marges linguistiques
no. 9, electronic journal, May 2005 issue, 64–75, <http://www.marges-
linguistiques.com>.
MEYRIAT J. and MIÈGE B.
2002: “Le projet des SIC: de l’émergent à l’irréversible (fin des années
1960 – milieu des années1980)”, in Boure Robert (ed.), Les origines
des sciences de l’information et de la communication. Regards croisés,
Lille, Presses Universitaires du Septentrion, 45–70.
MIÈGE B.
2000: “Les apports à la recherche des sciences de l’information et de
la communication”, Réseaux no. 100, 549–567.
MOUILLAUD M. and TÉTU J.-F.

1989: Le Journal quotidien, Presses universitaires de Lyon.
NEVEU E. and QUÉRÉ L.

1996: “Presentation” of the dossier “Le temps de l’événement”, Ré-
seaux no. 75, 5–21.
OGER C.
2002: Model candidates, cultures and methods. The general culture
test in three selection examinations for training civil service élites
(Ecole de Guerre/Cours Supérieur d’Etat-Major, Ecole Nationale
d’Administration, Ecole Nationale de la Magistrature). Discourse
analysis of jury reports, doctoral thesis in Language Sciences de-
fended 22 November 2002 at Paris XII.
2003: “Communication et contrôle de la parole: de la clôture à la
mise en scène de l’institution militaire”, Quaderni no. 52, 77–92.
2005: “L’analyse du discours institutionnel entre formations discur-
sives et problématiques socio-anthropologiques”, Langage et société
no. 114, 113–128.
OGER C. and OLLIVIER-YANIV C.

2003: “Conjuguer analyse du discours institutionnel et sociologie
compréhensive: vers une anthropologie des discours institutionnels”,
Mots, les langages du politique no. 71, 125–145.
2006: “Conjurer le désordre discursif. Les procédés de ‘lissage’ dans
la fabrication du discours institutionnel”, Mots, les langages du
politique no. 81, 63–77.
ROBIN R.
1973: Histoire et linguistique, Paris, A. Colin.
TÉTU J.-F.
2002: “Sur les origines littéraires des sciences de l’information et de
la communication”, in Boure Robert (ed.), Les origines des sciences
de l’information et de la communication. Regards croisés, Lille,
Presses Universitaires du Septentrion, 71–93.
UTARD J.-M.
2001: “Du contenu aux interactions discursives. Les enjeux de l’ana-
lyse du discours des médias”, in Georgakakis D. and Utard J.-M.
(ed.), Science des médias; jalons pour une histoire politique, Paris,
L’Harmattan, 159–178.
2004: “L’analyse de discours entre méthode et discipline”, in Rin-
goot R. and Robert-Denontrond P., L’analyse de discours, Apogée,
23–52.
VINCENT D.
2005: “Analyse conversationnelle, analyse du discours et interpréta-
tion des discours sociaux: le cas de la trash radio”, Marges linguis-
tiques no. 9, electronic journal, May 2005 issue, 165–175, <http://
www.marges-linguistiques.com>.
Chapter III
Discourse analysis and interpretive sociology.

A critical look back at an interdisciplinary
research practice
Claire OGER and Caroline OLLIVIER-YANIV
The purpose of this contribution is first of all to present the shared

approach we have developed by combining interpretive sociology
and discourse analysis. This rapprochement was initially suggested
by problems of research emanating on the one hand from the sociol-
ogy of political and public institutions and according an important
place to actors’ discourse, and on the other hand from institutional
discourse analysis. The implementation of an approach which goes
beyond the simple comparison of results then led us to place the for-
mation of corpora – or collections of discursive materials – at the
heart of questions of method.
Secondly, we propose to reconsider the various projects on insti-
tutional discourse which we have worked on together, in order to
identify their difficulties and their prospects. We will in particular
insist on the need to clarify the theoretical and epistemological posi-
tions which underlie this approach and determine its coherence.
The construction of a shared approach
Looking back at a number of shared works, we will provide a quick

summary of their main stages, from comparison of results to concep-
tual dialogue, and then to the hybridisation of methods.
44 Claire Oger and Caroline Ollivier-Yaniv
Discourse as a point of convergence between DA

and interpretive sociology
The basis for a rapprochement between the two disciplines is the

importance of the discourse in the two theoretical traditions in which
they are embedded [Oger and Ollivier-Yaniv, 2003a].
Interpretive sociology cannot figure out the representations given
by social actors of their activities, nor their internalised predisposi-
tions, without considering very carefully the discourse produced by
the subjects themselves. For about thirty years now, there have been
a large number of sociological studies putting forward general evi-
dence for a reinstatement of the subject and the individual, following
the structuralist model of the actor. In France, interactionism [Becker,
1985], ethno-methodology [Garfinkel, 1967; de Fornel, Ogien and
Quéré, 2000] and social phenomenology [Berger and Luckmann,
1966] provided the founding stages in this process, in a possibly radical
way since they are sometimes based on an interpretive consideration
of subjectivities and depend on collecting the words of individual
people. They can be related to more complex theoretical approaches,
some of them re-questioning the dialectic relationship between “in-
dividual and society” [Elias, 1970] or between “system and actor”
[Bertaux, 1999]. Other sociologists, such as B. Lahire, in the wake of
Bourdieu and a fundamentally structuralist legacy [Lahire, 2005: 136],
are rethinking the concepts of subject and individual based on a clari-
fication of “the heterogeneity of schemes of action incorporated by
the actor” [Lahire, 1998: 48]; at the same time, Lahire is concerned
to separate himself from the “subjectivist trends in social sciences”
[Lahire, 2005: 160]. Considering the respective positions of these
authors (and many others not referred to in this article), it is possible
to speak of an “‘existential’ turning point” [Touraine and Khosrokha-
var, 2005: 20] in sociology.
Discourse analysis, on the other hand, has undergone developments
which have brought it close to ethno-sociological types of problems.
Without going into the detail of these changes of direction, it is useful
to recall that the “ethno-linguistic turn taken by discourse analysis”
Discourse analysis and comprehensive sociology 45
has led a certain number of researchers to focus their interest on re-

stricted communities, language rituals and more generally the way in
which “saying, ways of saying, ways of doing and being become
entwined to ‘embody’ discourse and lead to ‘incorporation’ by its
users” [Maingueneau, 1992: 123].
This all represents a move away from the conceptions which,
in the early stages of DA, gave pride of place to the “unthought of
the Sense” and the way in which ideology, like the unconscious, con-
cealed its presence behind a “a fabric of subjective evidences”
[Bonnafous and Jost, 2000: 528]: the choice of problems, corpora
and lines of analysis has been profoundly reconsidered because of
this, and also as the various theoretical positions evolved [Maingue-
neau, 1992]. At the heart of this evolution was the conception of the
subject:
According a pre-eminent role to the problematics of enunciation […] leads us to

reflect on the connection between the subject of the discourse and the subject as
understood by sociologists and psychologists [Maingueneau, 1996: 11].
Developments in the two disciplines thus appeared to make a joint

approach possible, indeed desirable. We have chosen to focus on a
subject common to both of us: namely, the military discourse about the
institutional communication (working relationships with journalists,
communication campaigns), considered as much through corpora of
internal documents as from interviews held with their superior officers
[Ollivier-Yaniv, 2000; Oger, 2000 and 2003].
The variability of orders of constraint in discourse
In establishing an approach which can be recognised by both dis-

course analysts and sociologists, and in which real exchanges can
take place, it appeared important to stress the various conditions of
enunciation. More particularly, we were concerned to emphasise the
quite different constraints borne by official and prescriptive institu-
tional enunciation (which we call “framing discourse”) and the dis-
course produced by actors, in particular during the interaction with

the researcher at the time of the interview. By bringing together se-
ries of discourses produced under different constraints, the researcher
is led to deal with different and complementary materials:
– different because certain discursive features stand out to a greater
or lesser degree. For example, the particular context of enuncia-
tion in institutional texts reveals a predominance of prescriptive
discourse whereas interview situations bring to light narrative
sequences. In the same way, interaction with the researcher leads
institutional actors to favour argumentative discourse, while in-
ternal documents are usually dominated by explanatory discourse
[Oger and Ollivier-Yaniv, 2003b].
– complementary because the analysis of interviews allows us to
understand the construction of the “framing discourse.” Thus,
the corpus analysis of institutional texts can direct the interview
towards elucidation or interpretation by the actors of a certain
number of recorded features [Oger, 2003; Ollivier-Yaniv, 2003].
More generally, the interview enables us to study the actors’ dis-
course about the construction of institutional discourse [Oger and
Ollivier-Yaniv, 2005].
Nevertheless, these different orders of constraints cannot be reduced
to different levels of authenticity, following the idea that the inter-
view enables us to gain access to practices which are more real, or
to “go behind” a framing discourse, simply considered as a window.
Whether the approach is made in the field or through a corpus of texts, the sub-
ject described is always a construction. And however much we question those
who produce texts, what they say about their activity and the reasons guiding it
will be neither truer nor falser than what the clues spotted in the texts will reveal
about their underlying logic [Utard, 2004: 45].
In sociology, Bernard Lahire represents a similar position when he

pleads for a “varying scale of context in social sciences” which would
allow “the effects of knowledge appropriate to each mode of con-
struction” to be displayed [Lahire, 1998: 243].
Our approach is therefore based on a refusal of the divide be-

tween the order of doing, the order of saying and the order of saying
over doing; it is based, rather, on an analysis of the way they are put
into words.
Examination of early research and perspectives
The main methodological and theoretical questions that we have been

led to consider and attempt to clarify are the following:
– how to set limits to the investigation?
– how to unravel the difficulties linked to the framework and the
tools of this process?
From corpus to investigation: what are the limits?
Even if we only consider DA, strictly speaking, as soon as research-

ers approach a socio-anthropological problem, they need to reformu-
late certain familiar concepts of their discipline in order to study in-
stitutional discourse. Essential reading of anthropologists, sociologists
and historians would suggest so, just as much as the effort to recon-
stitute a large part of interdiscourse. Without this, in many institu-
tional texts the political implications and interpretation conflicts fade
away to be replaced by the misleading “platitudes” of conventional
discourse [Bonnafous and Tournier, 95: 71]. We have thus tried to
show how reference to notions like discursive formation or commu-
nity needs to be questioned or how the distinction between closed
corpus and open corpus [Oger, 2005] stood out in this particular case.
The complexity is increased when dealing with an interdiscipli-
nary approach such as the one being presented here. In the work we
have undertaken jointly, we have chosen to focus the analysis more
particularly on documents framing communication or official pre-

scriptive texts, clarifying their sense through the practice of inter-
views. But, as pointed out earlier, we have attempted to show that the
“circles” of institutional discourse (signified by the different scales
of constraint weighing upon its enunciation) are manifold and call
for differentiated analyses. Where communication by the military
establishment is concerned, it is therefore necessary to study several
types of discourse, without reducing their specific nature to a would-
be homogeneous discourse. Over and above the discourse produced
in interviews or the official framing texts already referred to, we might
add the articles and dossiers published in the internal press, recruit-
ment campaigns, information provided in ministerial press briefings,
interviews with ministers or communicators and so on.
The questions linked to the forming of such heterogeneous cor-
pora have already been flagged up by other researchers, notably those
working at the crossroads between DA and sociolinguistics. Regard-
ing our object of study, in appearance very different from those which
interest the researchers of the Langage et Travail network, we can
echo the thinking of Josiane Boutet, Bernard Gardin and Michèle
Lacoste [Boutet, Gardin and Lacoste, 1995: 13–14], for whom
in the realm of work, where discourse, texts and other materials of analysis are
largely to be produced by the research process itself […] the question of corpus
is re-situated in a more general problem of investigation.
In the same line of thinking and in relation to a field of enquiry closer

to our own, Virginie André underlines the methodological and re-
flexive implications of her work. This deals with work meetings in a
public company, and particularly concerns the
methods of data collection (observation, recordings, note-taking, informal dis-

cussions with employees to elicit clarification, particularly for anything relating
to what is implicit, to interdiscourse or discourse previously voiced, etc.) [André,
2005: 195].
Consequently, the investigation appears to have no limits, and it be-

comes a great temptation to practise a distribution of work which
unites the divisions and disciplinary specialisations, with the result

that the process of method hybridisation would cancel itself out.
To ensure that our project becomes realistic and realisable, we
propose to consider these heterogeneous corpora as collections of
discursive materials which enable us to account for the different “in-
stitutional circles” [Oger and Ollivier-Yaniv, 2003a]. This will allow
us to demonstrate more effectively the continuum existing between
discourses which are certainly distinct, but too often disassociated:
an article by the head of the French army’s central information and
PR service (SIRPA) in the military press, an interview with this same
officer in the professional press specialising in communication and
an interview he granted the researcher after leaving his post.
This redefining also shows the specific nature of our own activ-
ity, which is, namely, an anthropology of institutional and political
discourse, and not an institutional and political anthropology. The
limits of the investigation we are proposing are above all laid down
by the plan to work on the institution by taking into consideration a
plurality of discursive data, which give it both body and authority,
rather than observational or quantitative data.
The appropriation of frameworks and analytical tools
It thus becomes evident that the approach we have adopted attempts

to go beyond a simple methodological using of one discipline by
another. Indeed, it seems to be the best way to escape from the implicit
hierarchical system among academic disciplines.This hierarchy would
lead to “maintaining mutually exclusive relationships between a socio-
logy of context (which would assume the status of causing utterances)
and a text linguistics (which would appear as a reflection of the con-
text)” [Utard, 2004: 37]. Certain theoretical difficulties identified here
commit us to pursuing the dialogue on several fronts.
First of all, the practice of the interview needs to reflect the inter-
action between the researcher and the person being questioned [Oger
and Ollivier-Yaniv, 2003a]; such reflection also needs to be directed
at the concept of the actor, or even of the subject or individual, who

underlies this choice of method.
In this respect, the still strong imprint of the theoretical models
of early discourse analysis formed an obstacle which later develop-
ments in DA, mentioned above, have helped to reduce. Indeed, it is
difficult to practise interviews if there is a tendency to consider dis-
course solely “not as the expression of the speaking subject and of
his intention, but as the expression of an ideological and political
complex which transcends the subject” [Bonnafous, 2006]. On the
other hand, support may be found in enunciative theories, which lead
to an “‘interest’ in the presence of the subject in his utterance and
manifestations […], in his argumentative strategies, in the evident
forms of his capacity for self-reflection, etc.”; the subject being con-
ceived of as a “being who negotiates margins of action for himself
within the structures and institutions in which he acts” [Bonnafous,
2006].
For the sociologist, mastering the linguistic categories used by
DA is often difficult, as are the theoretical choices underlying the
discipline, which prohibit it being imported as a “chart” or as a key
giving access to “contents”.
We will not recall here all the considerations, widely referred to
in works linked to DA, which clearly distinguish it from content analy-
sis [Robin, 1973]. More recently, J.-M. Utard [Utard, 2004] and D.
Maingueneau [Maingueneau, 2005] have underlined the epistemo-
logical positions and theoretical choices which forbid the assimila-
tion of DA to a method.
However, it seems that a certain number of sociologists still do not
have a clear perception of these two epistemological and methodo-
logical parameters: content analysis, discourse analysis and thematic
analysis are often understood as equivalents, even though the analysis
of data gathered in the context of an interview (or analysis of open
questions in a questionnaire) is largely established in sociology.
Alongside this naïve attitude, different types of sociological ap-
proach can be observed, which include the need to develop work on
collected discourses. For example, N. Heinich, explaining the details
of the method used for her research on literary prize winners, demon-
strates the “counterproofs to the illusion of word transparency” at the
same time as she implements a “thematic analysis” and takes into
consideration “the context of enunciation, in other words, the inter-
view situation” [Heinrich, 1999: 34–36]. To show the plurality of
meaning in an actor’s discourse, she draws in particular on the work
of Didier Demazière and Claude Dubar [Demazière and Dubar, 1997]
who criticise any position aiming simply to reproduce interviews.
Other works inspired by interpretive sociology are based on the
use of software for the automatic processing of data and textual sta-
tistics. These include Alceste and Prospéro [Chateauraynaud, 2003],
but the sociologist’s objective is not then to acquire knowledge of
how discourse functions. More generally, the whole sociology of
controversies and judgment thereby develops a reflexive approach to
the forms of categorisation of language and discourse, though in the
eyes of discourse analysts, both these terms are considered to be used
in a fanciful way [Boltanski and Thévenot, 1999].
Yet even the sociologist who is familiar with a certain number of
theoretical presuppositions belonging to DA may still encounter real
difficulties with the linguistic categories and structuring arguments
involved in the work of discourse analysis. A few pointers on this
subject might be developed at a later stage in our shared research.
And vice versa, the discourse analyst wanting to build up a cor-
pus through the practice of interviews, might usefully be reminded of
the analysis criteria – other than the linguistic and discursive ones –
which are traditionally taken into consideration by the sociologist. In
the first place, putting together the trajectory or route (professional
and/or personal) taken by the author of the action and discourse can
present an initial means of elucidating the discourse produced during
interaction with the researcher. Examples of cross-cutting or inter-
secting of the different interviews can also contribute to this, by re-
vealing incoherences, contradictions and what remains unsaid.
Finally, the conducting of the interview and the types of ques-
tions or follow-ups are a third determining element from our point of
view: “To succeed in getting the subject to move from a normative
enunciation to a descriptive enunciation”, writes Nathalie Heinich

[Heinich, 1999: 33]. B. Lahire emphasises once again the “frequent
distortion between what the actors do and know, and what they say
they do and know” [Lahire, 2005: 142], and encourages researchers,
by preparing an interview schedule, to “help practitioners state what
their practices are as precisely as possible” so that they do not slip
from the practical logic in which they are bound up towards discur-
sive arguments that they have learnt to marshal for situations where
there is a more “‘formal’, ‘public’, or ‘official’ presentation” [Lahire,
2005: 157–159].
From discourse to practice: a problematical inference
Finally and most importantly, it is really the question of the relation-

ship between discourse and practice which is at the heart of this theore-
tical thinking.
Generally, discourse analysts never fail to insist on their attach-
ment to “discursive materialities” and their own metaphors tend to
revolve around the dual theme of density and opacity of discursive
phenomena. Other disciplines, on the other hand, seem often to pierce
through discourse as if it were a thin casing for content, or a trans-
parent surface beyond which one is at liberty to contemplate an extra-
discursive reality.
The frequent references by discourse analysts to L’Archéologie
du savoir is one of the most obvious indications of this position:
What we are concerned with here is not to neutralise discourse, to make it the
sign of something else, and to pierce through its density in order to reach what
remains silently anterior to it, but on the contrary, to maintain it in its consist-
ency, to make it emerge in its own complexity.
So, conversely, rather than piercing through discourse in this way,

discourse analysts aim to explore its density and its complexity.
Seen from this viewpoint, discourse analysts are reluctant to con-

sider that it is possible to arrive at practices through discourse, pre-
ferring to consider discourses assembled in a corpus as observable
practices. As emphasised earlier, interviews give neither more nor
less access to institutional practices than official texts, insofar as they
must be viewed by the discourse analyst as discursive practices.
It will already have been understood, from reading what precedes,
that DA in no way has a monopoly over this theoretical scruple. From
the point of view of our questioning, it has thus appeared possible to
identify, in sociology, a number of pieces of work whose position with
regard to discourse was similar to our own and should therefore be
repositioned within this disciplinary field.
Indeed, if recourse to the interview has become self-evident for a
large number of sociologists, the conception of subject and discourse
(or words) collected in this way is nonetheless variable. If one were to
venture to draw up an inventory of the responses to this issue, two
opposite poles could be distinguished. On the one hand, ethno-
methodologists, describing themselves as radical or alternative socio-
logists, propose a research activity which is fundamentally inductive
and analytical, aiming to paint a picture of society, “while accepting
the idea that the events are recognisable to those participating in them”
[Watson in de Fornel, Ogien and Quéré, 2000: 27]. The words gath-
ered (notably by interview) are both material and a result of the work
of research; in fact any distinction between these two notions becomes
inappropriate [Garfinkel in de Fornel, Ogien and Quéré, 2000: 34].
The approach that we are proposing will find support at the other
theoretical pole, of sociology shaped by an interpretive approach,
which accords importance to the discourse of actors in the action, the
very action which breaks sharply with any “subjectivism” [Le Bohec,
2001: 225]. The position of B. Lahire and the explanation that this
author himself gives of it by questioning, for example, the relation-
ships between “doing” and “saying about doing”, or by distinguishing
“meta-discursive logic” from “practical logic”, succeed in giving a
precise shape to interview work and its results. “The version of reality
obtained will depend on the way in which the interview is conducted
and there is therefore no natural continuity between practice – or

knowledge in action – and their being declared” [Lahire, 2005: 144].
Sociologists will thus be surprised at the “inordinate place accorded
[by other sociologists] to this presupposed and invisible reality as such
[the “thought” reality], in relation to the place ascribed to objective
and “objectivable” realities, which can be analysed from all imagina-
ble scientific points of view, etc.; that is to say, linguistic realities
(verbal, para-verbal, written, gestural, iconic and so on) reduced to the
subsidiary status of a simple “clue” or “sign” of mental activity which
is judged to be fundamental” [Lahire, 1998: 231]. Acknowledging the
density of discursive materialities, as against a sociology of “mentali-
ties”, claiming to pierce through them to reach an invisible and
unknowable “thought”, Lahire therefore joins discourse analysts in
their mistrust of approaches which practise discourse inference to a
supposed “belief” on the part of the actors.
Thus it would be useless to claim to establish, through inter-
views, how actors represent their activity to themselves; at most we
can determine how they represent it, voluntarily or not, in the context
of interaction with the researcher. It would be limiting, therefore, to
want to gain access to the social dimension by only analysing the
corpora of institutional discourse. The maximum we can hope to
achieve is to formulate hypotheses on the subject. It is for both these
reasons that the comparison between different orders of discourse,
which reflect different levels of constraint, appears to us to be par-
ticularly fertile.
ANDRÉ V.
2005: “Oui non: une pratique discursive sous influence”, Marges lin-
guistiques, no. 9, 195–213 [On line] <http://www.revue-texto.net/
Parutions/Marges/00_ml092005.pdf>.
BECKER H.
1985: Outsiders (1963), Paris, Métaillé.
BERGER P.L. and LUCKMANN T.

1987: La construction sociale de la réalité (1976), Paris, Méridiens-
Klincksieck.
BERTAUX D.
1999: Les récits de vie, Paris, Nathan Université.
BONNAFOUS S.
2006: “L’analyse de discours”, in OLIVESI S. (ed.), Sciences de l’in-
formation et de la communication: une introduction, Grenoble, Presses
universitaires de Grenoble, 213–228.
BONNAFOUS S. and JOST F.

2000: “Analyse de discours, sémiologie et tournant communication-
nel”, Réseaux no. 100, Paris, 524–545.
BONNAFOUS S. and TOURNIER M.

1995: “Analyse du discours, lexicométrie, communication et poli-
tique”, Langages no. 117, Paris, 67–81.
BOUTET J. and GARDIN B. and LACOSTE M.

1995: “Discours en situation de travail”, Langages no. 117, Paris,
12–31.
BRES J. and NOWAKOWSKA A.

2005: “Dis-moi avec qui tu ‘dialogues’, je te dirai qui tu es… De la
pertinence de la notion de dialogisme pour l’analyse du discours”,
Marges linguistiques no. 9, Paris, 137–153 [On line], <http://www.
marges-linguistiques.com>.
CHARAUDEAU P. and MAINGUENEAU D.
2002: Dictionnaire d’analyse du discours, Paris, Editions du Seuil.
CHATEAURAYNAUD F.
2003: Prospéro – Une technologie littéraire pour les sciences hu-
maines, Paris, CNRS Editions.
CORCUFF P.
1996: Les nouvelles sociologies, Paris, Nathan Université.
DE FORNEL M., OGIEN A., QUERE L.

2000: L’ethnométhodologie. Une sociologie radicale, Paris, La Décou-
verte.
DEMAZIÈRE D. and DUBARD C.

1997: Analyser les entretiens biographiques. L’exemple de récits
d’insertion, Paris, Nathan.
ELIAS N.
1996: Qu’est-ce que la sociologie? (1970), Paris, Pocket Agora.
FRANÇOIS B. and NEVEU E.
1999: Espaces publics mosaïques. Acteurs, arènes et rhétoriques, des
débats publics contemporains, Rennes, Presses universitaires de Rennes.
GARFINKEL H.
1967: Studies in ethnomethodology, Prentice Hall, Englewood Cliffs.
HEINICH N.
1999: L’épreuve de la grandeur. Prix littéraires et reconnaissance,
Paris, La Découverte.
LAHIRE B.
1998: L’homme pluriel. Les ressorts de l’action, Paris, Nathan 2005:
L’esprit sociologique, Paris, La Découverte.
LE BOHEC J.
2001: “La question des rapports intersectoriels. Les obstacles à l’ana-
lyse des interactions entre journalistes et politiciens”, Emergences et
continuité dans les recherches en information et communication, Pro-

ceedings of the 12th national Conference of Information and Commu-
nication Sciences, Paris, SFSIC & Authors, 2001, 223–231.
MAINGUENEAU D.
1992: “Le tour ethnolinguistique de l’analyse du discours”, Langages
no. 105, Paris, 114–125.
1996: “L’analyse du discours en France aujourd’hui”, Le Français
dans le monde special edition, Paris, 8–15.
2005: “L’analyse de discours et ses frontières”, Marges linguistiques
no. 9, Paris, 64–75 [On line], <http://www.marges-linguistiques.com>.
OGER C.
2000: “De l’esprit de corps au corps du texte: cohésion militaire et
dissolution journalistique”, Langage et société no. 94, Paris, 9–43.
2003: “Communication et contrôle de la parole: de la clôture à la
mise en scène de l’institution militaire”, Quaderni no. 52, Paris, 77–
92.
2005: “L’analyse du discours institutionnel entre formations discur-
sives et problématiques socio-anthropologiques”, Langage et société
no. 114, Paris, 113–128.
OGER C. and OLLIVIER-YANIV C.
2003a: “Conjuguer analyse du discours institutionnel et sociologie
compréhensive: vers une anthropologie des discours institutionnels”,
Mots, les langages du politique no. 71, Paris, 125–145.
2003b: “Du discours de l’institution aux discours institutionnels: vers
la constitution de corpus hétérogènes”, Proceedings of the first inter-
national Francophone Conference on Information and Communica-
tion Sciences (CIFSIC), Bucharest [On line], <http://archivesic.ccsd.
cnrs.sic_00000717.html>.
2006: “Conjurer le désordre discursif. Les procédés de ‘lissage’ dans
la fabrication du discours institutionnel”, Mots, les langages du poli-
tique, no. 81, Paris, 63–77.
OLLIVIER-YANIV C.
2000: “Quels ‘professionnels’ pour la communication et les relations
avec les médias à la Défense? Carrière militaire et communication”,
Langage et Société no. 94, Paris, 75–96.
2003: “La fabrique du discours politique: les ‘écrivants’ des prises de
parole publiques ministérielles”, in Bonnafous S., Chiron P., Ducard D.
and Levy C., Argumentation et discours politique (Cerisy conference),
Rennes, Presses universitaires de Rennes, 89–98.
PHARO P.
2004: Morale et sociologie, Paris, Gallimard.
TOURAINE A. and KHOSROKHAVAR F.
2005: La recherché de soi. Dialogue sur le sujet (2000), Paris, Le
Livre de Poche.
UTARD J.-M.
2004: “L’analyse de discours entre méthode et discipline”, L’analyse
de discours, Paris, Apogée, 23–52.
WEBER M.
1992: “Essai sur quelques catégories de la sociologie compréhensive”
(1913), Essais sur la théorie de la science, Paris, Agora, 301–364.
59
Chapter IV
Multidisciplinary work on discourse.

Examples of a “way of doing” discourse analysis
Alice KRIEG-PLANQUE
There are different ways of embarking on discourse analysis. In these

pages, I would like to present one of the multiple “ways of doing” it,
among those which place the emphasis on multidisciplinarity while
maintaining discourse as the object of research. Work I have already
published, based on very concrete studies of corpus, will serve to
illustrate this strategy. Indeed, a feature of this work is the diversity
of methods and types of knowledge that it implies: formulated in a
schematic fashion, it involves describing corpora using categories
stemming from linguistics and discourse analysis, and interpreting
them in the light of various other disciplines, such as political sci-
ence, history, sociology, anthropology and political philosophy. The
research objectives justify such an approach, since they are concerned
first and foremost with analysing discourse as a product of the real
social and political world.
Specific nature of the approach:

to break down barriers and make connections
between points of interdiscourse
The reader will find here both a summary and a proposal – and they
both exemplify an approach to discourse analysis that can quickly be
placed in context in relation to others. In an article devoted to an
60 Alice Krieg-Planque
epistemological panorama of discourse analysis, Dominique Maingue-

neau [2005] places my approach among those which:
– give priority to “non-topical units”, constructed by the researcher
independently of pre-established boundaries and grouping to-
gether utterances which are strongly embedded in history1 (un-
like “topical units”, which allow a departure from the discourse
as it is presented, and aim more to contain and delimit it2);
– study these non-topical units in relation to the “route” or course
that they may take (unlike a study of these units according to
“discursive formations”, which is more attentive to the systems
which determine and produce discourse).
In fact, our approach consists generally in going beyond the bounda-
ries set by discourses and those who produce them, in order to high-
light instances of reprise, reformulation, regularity, circulation, dis-
persion and echo. Or again, as pointed out by Dominique Maingueneau
[2005: 73], in my own case it concerns “destructuring the framed
units by describing unexpected routes: interpretation is thus based
on revealing unsuspected relationships within interdiscourse”. This
describes more clearly studies 2 and 4 presented below, but the two
others are also related.
In the paragraphs which follow I shall set out examples of the
“way of doing” discourse that I have used. Although it may appear
odd, it has the merit of differentiating between “ways of doing” dis-
course analysis and “ways of doing” used elsewhere (in content analy-
sis particularly, but also in different uses of the “discursive” in hu-
man and social sciences).
1 “Non-topical units” are given priority if, for example, one attempts to assemble
the properties of “colonial discourse” of a certain period, or locate a network of
interdiscourse around the question of the social and environmental responsibility
of companies.
2 “Topical units” are given priority if, for example, one is interested in genres or
types of discourse such as school textbooks or departmental hospital meetings,
or if attention is focused on different registers of utterances adopted according to
linguistic, functional or communicational criteria.
Multidisciplinary work on discourse 61
The denominal adjective “ethnique” (study 1)
I became interested in the adjective “ethnique” on account of its am-

biguous and ill-defined nature (for more detailed reading of this study
see [Krieg-Planque, 2002]). The choice of “ethnique” rather than any
other denominal adjective (“huileux”, “sportif ”, “joyeux”, “matinal”)
is of course linked to the interest it presents for contemporary politi-
cal discourse. Nevertheless, I first of all made a general analysis of
denominal adjectives as a morpho-syntactic category, showing how
this category is particularly well-suited to supporting ambiguity and
underdetermination.
To begin with, I showed that all denominal adjectives present
inherent relational ambiguity, meaning that the adjective does not
unequivocally reveal the relationship established between the noun
from which it derives and the noun governing it. For example, “sauce
huileuse” can equally mean “containing oil”, “having the translucent
look of oil” or “having the consistency of oil”. Second, I showed how
certain denominal adjectives display an enunciative underdetermination
(meaning that they leave the enunciative responsibility for their under-
lying term in a state lacking in definition). For example, in “discrimi-
nation raciale”, the presupposition of the existence of “race” as a
relevant category for describing the world can be attributed to the
locutor just as much as it can reflect an exterior enunciative authority.
These two characteristics appear to be valuable resources that
language offers to its users in work contexts, where the category that the
noun designates is perceived as problematic in its very existence, in a
given public arena. “Ethnie” belongs precisely to one of those catego-
ries perceived as problematic in the contemporary French public arena
(in the same way as “race”, for example), particularly since “ethnie”
tends to function as a negative hetero-descriptor (on this particular
point see [Krieg-Planque, 2005]). In addition, we can formulate the
hypothesis that the enunciative underdetermination permitted by
“ethnique” explains the frequent use of this adjective (possibly with the
prefix “multi-”, “pluri-”, “poly-”, “inter-” and so on) compared to the
more sparing use of the noun “ethnie”.
Overall, this study involves starting out from a category which

at first sight is of more interest to linguistics (the “denominal adjec-
tive” category), but my aim is to show how language events are of
direct interest to discourse analysis. In the study just referred to,
I showed that there can be immediate implications, for a discourse
and for those holding it, in the act of using a denominal adjective.
This study also provides an opportunity to recall what appears to me
to be an essential consideration in discourse analysis: the things that
are indistinct, vague, not-one, ambiguous, ill-defined, polysemous,
heterogeneous, etc., should not be envisaged as “problems to be
solved” but rather as resources which the researcher must describe
and interpret (see in particular [Krieg, 2002: 106–108]).
The lexico-syntactic association

“ne pas pouvoir dire que ne pas savoir” (study 2)
I became interested in the lexico-syntactic association of the verbs

“pouvoir” and “savoir” and the negation contained in the way this
association works in the pattern of the discursive locus “ne pas pouvoir
dire que ne pas savoir”. This is used in established expressions such
as “Nous ne pourrons pas dire que nous ne savions pas”, “On ne
pourra pas dire qu’on ne savait pas” [“We won’t be able to say we
didn’t know”], “Qui pourra dire: je ne savais pas?” [“Who will be
able to say: I didn’t know?”], “On ne peut pourtant pas dire, comme
il y a cinquante ans: ‘On ne savait pas’”, [“Anyway, we can’t say,
as we did fifty years ago: ‘We didn’t know’”] “Cette fois, nous ne
pourrons pas tricher avec l’Histoire. On savait.” [“This time, we
won’t be able to cheat History. We knew.”] and so on. I studied this
series in the context of discourse analysis of the 1991–95 Yugoslav
wars [Krieg-Planque, 2003b].
Several features of these expressions seem to be worth reflecting
upon. I will indicate a few of them here. What is the language act
being achieved through them? Is it a slogan, a moral injunction, an
alarm, a warning, a confession of guilt? What are the different modes
of dialogue by which these expressions refer (more or less implicitly

depending on the case) to the original utterance, which would be the
justification given to people’s passivity in the face of Nazi crimes
during the postwar period after 1945? The commonplace about Na-
zism being the pinnacle of horror in human history, and the equally
well-known commonplace that the inability to prevent this horror
represents the ultimate in guilt and shame – both are at work here.
What categories of actors are alluded to by the pronouns “nous” and
“on”, with their fluctuating and unrestricted values? It may be a “nous”
with an inclusive but vague value (for example “nous les démocrates
européens”) which constructs partitions between human beings (“nous”/
“eux”). The indistinctness of the reference, here as in many other
cases, is remarkable. This set of expressions is also interesting be-
cause of the absence of a grammatical object. The verb “savoir” de-
mands an object which is absent here: “savoir quelque chose” [“to
know something”]. When all is said and done, what do we know
from these utterances? The context of the words does not always al-
low us to identify the object of the predication. This aspect of the
series of expressions relates on the one hand to the question of quali-
fying facts (legal qualification, in particular); on the other hand, it
contributes to making things appear as unrepresentable, because un-
named. This echoes the well-known debates on the unspeakable char-
acter of certain events in human history.
All these points are developed in my thesis [Krieg, 2000: 206–
214] and book [Krieg-Planque, 2003b: 142–151], to which readers
may refer. It will be evident that what I am trying to grasp in this type
of analysis relates to political and moral positions. It will also be
understood that I am attempting to comprehend them by linking them
to categories that are physically observable on the chain (lexico-syn-
tactic association in the present case). Linguistic considerations are
therefore necessary for scrutinising the process but it is also essen-
tial, on occasion, to go beyond them. However, there is no question
of discarding them. This makes contact with other disciplinary ap-
proaches not only possible but expected. Here, for example, research
into the way the discursive locus forged on “ne pas pouvoir dire que
ne pas savoir” is put into discourse, finds a happy convergence with

the “topic of denunciation” as analysed by Luc Boltanski [1993].
The sign “(sic)” and its uses (study 3)
I became interested in the sign “(sic)” and its uses in the contempo-
rary extreme right-wing press through Minute, National Hebdo,
Rivarol and Présent (for the way this research was developed see
[Krieg, 1999]). I arrived at the conclusion that the use of the sign by
these newspapers is part of a more general way in which extreme
right-wing discourse functions. Features of this are the importance
placed on what is implied; weak arguments in the face of strong evi-
dence; denunciation of the lies of which different actors in the public
arena are said to be guilty; and denunciation of the generalised perse-
cution and victimisation from which the extreme right is alleged to
suffer. Thus it is when Minute accompanies a quotation from the Eu-
ropean draft budget with the sign “[sic]” – a draft budget which talks
about “action en faveur des réfugiés qui ont provisoirement [sic] trouvé
asile dans l’Union européene” [“action in favour of refugees who
have temporarily [sic] found asylum in the European Union”]. The
“(sic)” appears as a marker, opening a space ready to receive an idea,
but it leaves this space vacant: the missing idea is implicit and it is up
to the reader to piece it together.
These conclusions about the specific importance of “(sic)” in
extreme right-wing discourse had two implications. On the one hand
a corpus was formed around this lexical sign (a corpus of utterances,
therefore, and not a corpus of texts); on the other, an analysis was
made of the way this term works in the French language. It can be
minimally described as a metadiscursive term generally placed after
a phrase/statement, which depicts the sequence it is commenting on
as both strange (or unexpected, surprising) and as belonging to a dis-
cursive exterior.
The corpus I set up for this work was a corpus of utterances based
on a clearly identified point of entry (mainly “(sic)” but also “[sic]”,
“– sic –”, “resic”, “re-sic”, “re-resic”). This enabled me not only to

understand a whole set of uses in the extreme right-wing newspapers
chosen, but also to make comparisons. Samples were also taken from
other newspaper titles (Le Canard enchaîné, Charlie Hebdo, Le Figaro,
Le Parisien, Le Monde, Libération, L’Humanité, Route, Lutte Ouvrière),
which enabled me to conclude that the extreme right-wing press made
greater use of this sign than others.
The need to grasp the general function of “(sic)” in the French
language arises from the view that words have a value in the lan-
guage system before they become words functioning in a particular
discourse; and also from the idea that only a description of the sign in
the language (helped, certainly, by observing its behaviour in dis-
course) then allows access to its implementation in particular utter-
ances. Furthermore, the mark “(sic)” does not, in itself, characterise
a political or ideological position. So it is not a direct and transparent
mode of access to the arguments defended by its users. It is, on the
other hand, characteristic of a mode of enunciation. And it was this
mode of enunciation that needed to be described first and foremost,
before understanding how it was embedded in a particular discourse,
namely that of the extreme right-wing press.
From the proto phrase “ethniquement pur”

to the expression “purification ethnique” (study 4)
I became interested in the different variants of the phrase “purifica-

tion ethnique” observed in the French press between 1980 and 1994
(see [Krieg, 2000] and [Krieg-Planque, 2003 b]). Among the many
observations to be made on this subject, I noticed that there was a very
particular period, which might be called a “pre-formation” period,
which saw the “genesis” of the expression, from 1980 to the spring of
1992. During this period, nearly all the occurrences encountered were
in the form of the adjectival syntagm “ethniquement pur(e)(s)”, whereas
after that, the phrase was mainly expressed through the variants “puri-
fication ethnique”, “nettoyage ethnique” and “épuration ethnique”
(to which may be added more than 130 other variants which are much
less frequent: “purificateurs ethniques”, “ont épuré ethniquement”
and so on).
During the first stage, when observing forms, I contented myself
with noting that a certain type of transformation was occurring here:
an adjectival syntagm was changed into a nominal syntagm consist-
ing of an action nominalisation and a denominal adjective. This fact,
on its own, was remarkable: the adjectival syntagm, by its incom-
pleteness, made it necessary to mention – however briefly – an ob-
ject, actors and the spatio-temporal conditions of the action. As for
nominalisation, ample research in linguistics and discourse analysis
shows that it is distinctive for its substantial resources with regard to
preconstruction and its powerful driving forces in terms of
underdetermination and ambiguity. To these nominalisation features
may be added those applying to denominal adjectives, which were
pointed out earlier (study 1). Thus it was that, from the period of the
1980s until 1992–95, discourses relating to Yugoslavia were able to
move – and, in fact, did move – from being precise accounts situated
in the time and space of the action, and being very grounded in terms
of enunciation, to the sometimes imprecise invocation of a concept
whose conditions of actantial and actorial realisation could be ab-
sent. Here again, the realities observable in discourse are made possi-
ble by what the language authorises by way of a system of constraints.
Concerning this last example, I have just stressed the importance
of morpho-syntactic categories in political and media discourse. But
other criteria entered into the study when I proceeded to analyse the
precise moment when the adjectival syntagm changed to being a nomi-
nal syntagm (which occurred between May and August 1992), and
the way in which this nominal syntagm found its way into newspaper
columns. This was by means of names belonging to signatories of
articles, elements of paratext (such as titling, provenance shown for
the report or correspondence), methods of reported discourse, type of
article and so on. These categories are dissimilar, and they relate both
to different levels in the discourse and to a variety of approaches
(from morpho-syntactic analysis of the constituent parts to text lin-
guistics via stylistics). But the concern is always to analyse the dis-
course. If other disciplines intervene, they provide, one might say,
“another viewpoint” (which certainly does not mean “an additional
one”). It is in this sense that I have talked about a “stratified multi-
disciplinarity” [Krieg, 2003 a], meaning that cooperation between
disciplines as envisaged in this way would be a relationship in which
the researcher would ask someone with a different approach to give
her view on the results previously obtained by another discipline. In
this sense, there is a stratification of roles. Such an arrangement is
not the expression of a value judgment, but the result of relevance to
a situation, in the same way as on a building site there is an order in
the way the different trade associations participate, without implying
that the carpenter is superior to the roofer or the painter more impor-
tant than the monumental mason.
The question of an “approach” with other disciplines might also
be asked – still in the study under discussion – via the notion of what
is “observable in discourse”. Underlying this is the argument (which
merits debate) according to which discourses can be a place for ob-
serving practices that have contributed to producing them [Krieg-
Planque, 2007]. For example, the monotony of certain sources of
information quoted in articles in the French press referring to Kosovo
in the 1980s reveal the working constraints and routines of foreign
journalists in the Federal Republic of Yugoslavia at the time.
Observing from discourse, interpreting in multidisciplinarity
In all of the work just referred to, the place of entry into discourse often
has a fairly prominent lexical or formal dimension. But on the one
hand, I see this entry more from the angle of discursive materialities
(transformations, derivations, syntactic operations, lexico-syntactic
associations, reprises, marks and indicators of dialogism) than of lexi-
cal forms strictly speaking. On the other hand, even when the point of
entry is strongly lexical, the analysis is based on considerations which
go beyond it. For example, in studying the sign “(sic)” in the extreme
right-wing press, I noticed that Lutte Ouvrière made very little use of
such a mark. However, a reading of this weekly newspaper enabled me
to observe that Lutte Ouvrière had recourse to another method – this
time explicit – of denouncing the adversary’s misleading discourse.
An example of this was the “euphemisation judgment”3.
Moreover, the interest for discursive materialities extends to what
I call “discursive places” [Krieg-Planque, 2006a]: it can lead towards
non-lexical text objects, whose circulation and distribution of places
in interdiscourse can be analysed. In this perspective, it is permissible
to carry out studies of directives, circulars, letters, short sentences,
official reports, confidential notes, memorandums, etc. I proposed
a few ways of applying this for my thesis research [Krieg-Planque,
2003b, in particular 105–133] and have used it more covertly in a text
dealing with the representation of intentionality of mass crime in media
discourse [Krieg-Planque, 2006b]. The conception of discourses likely
to be chosen as objects of research such as I have envisaged is therefore
both wide in the spectrum (from the “word” to the “text”, if one wished
to adopt empirical categories) and narrow in its conception, that is
very engaged with forms. It always calls for enrichment by other
disciplines and, all in all, in accordance with the “way of doing” re-
ported here, it enables the researcher to describe on the basis of dis-
course and interpret in the light of multiple disciplines.
Such a conception of discourses can easily be broadened so as to
account for various social practices. For example, just to cite one of
Céditec’s chosen areas, it is possible by using this method to contrib-
ute to the study of professional communication practices. We can
define communication as a set of elements of professional expertise
which involves anticipating the reprise, transformation and reformu-
lation of utterances and their contents: developing “elements of lan-
guage”; forcing oneself to include terms which function as markers
of discursive formation when speaking, starting to speak at times and
according to formats which can be taken up by the media and/or inter-
3 A process on which I have subsequently had occasion to work, and on other

corpora. See Alice Krieg-Planque (2004).
locutors; creating and integrating fragments of utterances capable of

being easily detached in the form of verbatims or short sentences;
and incorporating elements expected of the institution into one’s own
discourse or when completing forms (indicators, record cards, dum-
mies, calls for tender, contracts and so on). If communication is com-
prehended in this way, it is clear that understanding the work of com-
municators (or of those who, without being officially entrusted with
such a mission, need to include communication in their activity) nec-
essarily involves discursive analysis, at least in part.
A “way of doing” available for other research
Many studies can be envisaged from the viewpoint described here.

Here are three, by way of example:
Study 5. A study of debates on the governmental initiative to revise

legislation on working hours, begun in 2005, might choose as its
entry the lexico-syntactic association “travailler plus pour gagner plus”
[“working more to earn more”]: a study could be made of what we
might call the “supporting reformulations” (different reformulations
which contribute their support) and the “discrediting refutations” (dif-
ferent forms of refutation which undertake to undo the value of the
proposal). Supporting reformulations are found in various proposals
by Jean-Pierre Raffarin’s government to reform the 35-hour laws.
These proposals were submitted to the National Assembly in Janu-
ary/February 2005, before “travailler plus pour gagner plus” became
one of Nicolas Sarkozy’s slogans in the context of the 2007 presiden-
tial elections. Discrediting refutations, too, can be found in various
forms over the same period; forms such as complex dialogic nega-
tion (for example when the First Secretary of the French Socialist
Party, François Hollande, stated that employees “ne vont pas forcément
travailler plus, mais risquent de gagner moins” [“are not necessar-
ily going to work more, but risk earning less”]) or opacifying and
non-opacifying metadiscursive operations (for example in a CGT tract
according to which “c’est au nom d’une prétendue liberté de gagner

davantage qu’il [Le Medef] entend remiser au rayon des accessoires
la loi sur les 35 heures” [“it’s in the name of a so-called freedom to
earn more that it [Medef, the employers’ union] intends to relegate
the 35-hour law to the bottom shelf”]).
Study 6. Discourse analyses referring to the phrase “développement

durable” [sustainable development] can be organised around conces-
sion. Indeed, it is evident that the contradictions revealed – and just
as contested, rhetorically – by discourses referring to “développement
durable” were notably based upon concessives, particularly conces-
sives reduced to the gerund (or to tout + gerund) or to sans + infini-
tive or NG [nominal group] (or to sans pour autant + infinitive or
NG). Such constructions are found in definitions of sustainable de-
velopment presented as such (starting with the definition in the
1987 Brundtland report: “Répondre aux besoins présents sans com-
promettre la capacité des générations futures à satisfaire les leurs.”
[“To answer present needs without compromising the ability of future
generations to satisfy theirs.”]), but also in political or institutional
discourse which refers more implicitly to the notion of sustainable
development (for example in a document by Autoroutes du Sud de la
France: “Notre groupe a pour ambition d’offrir à tous un accès à des
modes de transports rapides, sûrs, économiques et fiables sans
sacrifier les richesses humaines et écologiques d’aujourd’hui et de
demain.” [“Our group’s ambition is to give everyone access to speedy,
safe, economical and reliable means of transport without sacrificing
human and ecological wealth now and in the future.”] )
Study 7. In the domain of research into journalism and the media, the
analysis of discursive realisations of expressions like “faire son tra-
vail” [to do one’s work] and “faire son métier” [to do one’s job] col-
lected in a corpus of utterances by media professionals (“bien pouvoir
faire son travail/son métier…”, “ne pas pouvoir faire son travail/son
métier…”, “ne plus pouvoir faire…”, “vouloir faire…”, “faire hon-
nêtement…”, “faire correctement…”, etc.) enables certain dimensions
of spontaneous representations by the profession to be understood in

a new light (values, norms, ideals and so on). It also illuminates cer-
tain aspects of the practical rules followed by journalists, such as
those revealed, for example, in the manner of Cyril Lemieux [2000].
Taking discourse as the object of study
I would like now to summarise this approach. What primarily attract

attention are forms, marks, signs, processes, operations – in short,
what is linguistically describable. It is nevertheless clear that this at-
tention is pre-determined through familiarity with corpora and through
experiencing the world beyond what is linguistically describable. Once
attention has been attracted and retained, I then try to understand what
linguists have to say about it. Next (and/or at the same time), I try to
reveal what these instances of discourse – which are necessarily based
on instances of language – can tell us about discourses as systems for
explaining the political and social worlds.
This type of approach and the results it produces may also be of
interest to sociologists, political scientists, economists, philosophers,
historians and so on. Yet the focus of study does not become that of
the sociologist, the political scientist, the economist, the philosopher
or the historian. It is about taking the discourse itself as object, and
maintaining this position right to the end, as I explain more fully in
an article devoted to epistemological and methodological proposals
for discourse analysis [Krieg-Planque, 2006 a: 32–35]. Questions are
asked of discourse, so it is from discourse that an answer is expected.
This implies that researchers with other aims in view may be disap-
pointed if they were expecting to gain something personally, for their
own purposes (particularly by considering discourse analysis, I think
wrongly, as a sort of “service provider”, a “tool box” or a “method”4).
4 In relation to this, see the very apt expression “corpus versus method” proposed
by Claire Oger in her contribution to this book (Oger, 2007).
The corpus as reading practice and the shifting

of hypothesis as a means of understanding
Such an approach implies giving a proper and serious place to the

corpus. It is obvious that “corpus” refers to something else in other
disciplines, particularly in history, sociology and political science,
where ways of doing things properly and seriously follow quite dif-
ferent paths.
For my own part, I have until now worked almost exclusively on
written corpora. Not only that, but, for various reasons, I have worked
more on press corpora (all types of press and all types of text). How-
ever, I have also studied various political documents, such as pam-
phlets from trade unions and associations, transcribed parliamentary
debates, parliamentary reports, magazine articles containing debates
and ideas, institutional communication media and various brochures,
as well as, very marginally, comic strips, children’s books and song
lyrics. What I am trying to say here is that there is no preconceived
limit to what is worthy of being “corpussable”. What appears to be
important is that on the one hand there are corpora (and not merely
collections of examples) and, on the other, that these corpora should
be constructed, identified, described, and finally read, known and re-
read. This last point suggests that the work described in this article
implies a regular acquaintance with utterances or, in short, that the
practice of reading is all-important. Finally, it needs to be stressed that
constructed corpora are only as they are by virtue of certain working
hypotheses initially put forward and then regularly shifted. The corpus
developed in this way allows these hypotheses to be “verified” or
“refuted”. Put more simply, it offers the possibility of shifting them,
reformulating them and putting a new one forward – if we wish to
consider that the “science” in which we are involved is, more than an
enterprise or checking technique, a process of knowledge and under-
standing.
BOLTANSKI L.
1993: La souffrance à distance. Morale humanitaire, médias et
politique, Paris, Métailié, Collection Leçons de choses.
KRIEG, A.
1999: “Vacance argumentative: l’usage de (sic) dans la presse d’ex-
trême droite contemporaine”, Mots. Les langages du politique, no. 58,
pp. 11–34.
2000: Emergence et emplois de la formule “purification ethnique”
dans la presse française (1980-1994). Une analyse de discours, doc-
toral thesis in Language Sciences defended 9 November 2000 at Uni-
versité de Paris 13 – Paris-Nord, 3 vol.
2002: “L’adjectif ‘ethnique’ entre langue et discours. Ambiguïté rela-
tionnelle et sous-détermination énonciative des adjectifs dénominaux”,
RSP. Revue de Sémantique et Pragmatique, no. 11, pp. 103–121.
2003a: “‘Procédures’, ‘routines’, ‘contraintes’. L’analyse des discours
médiatiques à la lumière de l’ethnosociologie”, in Damien Chaba-
nal et al., Sciences du langage: quels croisements de disciplines?
Montpellier, Publications de Montpellier 3 – Université Paul-Valéry;
pp. 71–86.
KRIEG-PLANQUE, A.
2003b: “Purification ethnique”. Une formule et son histoire, Paris,
CNRS Editions, Collection Communication.
2004: “Souligner l’euphémisme: opération savante ou acte d’engage-
ment? Analyse du ‘jugement d’euphémisation’ dans le discours poli-
tique”, Semen, no. 17, “Argumentation et prise de position: pratiques
discursives” (co-ordinated by Ruth Amossy and Roselyne Koren),
pp. 59–79.
2005: “Le mot ‘ethnie’: nommer autrui. Origine et fonctionnement
du terme ‘ethnie’ dans l’univers discursif français”, Cahiers de lexico-
logie. Revue internationale de lexicologie et lexicographie, Paris,

Editions Honoré Champion, no. 87, pp. 141–161.
2006a: “‘Formules’ et ‘lieux discursifs’: propositions pour l’analyse
du discours politique”, Semen. Revue de sémio-linguistique des textes
et discours, Besançon, Presses Universitaires de Franche-Comté,
no. 21, “Catégories pour l’analyse du discours politique” (co-ordinated
by Philippe Schepens), pp. 19–47.
2006b: “L’intentionnalité de l’action mise en discours. Le caractère
intentionnel des crimes de masse sur la scène médiatique”, in Marc
Le Pape, Johanna Siméant, Claudine Vidal ed., Crises extrêmes. Face
aux massacres, aux guerres civiles et aux génocides, Paris, La Décou-
verte, Collection Recherches, pp. 88–102.
2007: “Etude des pratiques d’écriture journalistique: jusqu’où aller
avec les sciences du langage?”, in Marcel Burger ed., L’analyse lin-
guistique des discours médiatiques, Lausanne.
LEMIEUX C.
2000: Mauvaise presse. Une sociologie compréhensive du travail jour-
nalistique et de ses critiques, Paris, Métailié, Collection Leçons de
choses.
MAINGUENEAU D.
2005: “L’analyse du discours et ses frontières”, Marges linguistiques.
Langages. Représentations. Communication, no. 9, pp. 64–75.
OGER C.
2007: “Analyse de discours et sciences de l’information et de la commu-
nication; au-delà des corpus et des méthodes”, in Simone Bonnafous
and Malika Temmar eds., Analyse du discours et sciences humaines et
sociales, Paris, Ophrys, Collection Les Chemins du Discours.
75
Chapter V
Political discourse analysis:

content analysis, lexical statistics and
a semantico-enunciative approach
Pierre FIALA
Over the course of the twentieth century, the language sciences, and
linguistics in particular, have progressed by seeking a balance (not
always attained) between empiricism and speculative constructions,
description and formalisation, basic research and applications. In re-
cent years, the emergence and independent development of informa-
tion and communication sciences, cognitive sciences and computer-
aided linguistic engineering has been a strong feature of these changes.
In this new order of things, corpus linguistics has been assisted by the
development of automated language processing tools (ALP) and the
extension of textual databases: it has thus contributed to steering lan-
guage sciences back towards empirical data and experimental meth-
ods, by reformulating the question of meaning and verbal interaction
on the basis of data analysis. In this context, those working on text and
words – linguists, but also political scientists, sociologists, mediologists
and of course historians and literary specialists, not to mention psy-
choanalysts – all of these now have available to them more highly-
developed automated discourse processing tools (ADP). Through them,
they are discovering new research methods, based on different per-
ceptions and models, a wider range of data, more precise descriptions
and, above all, measurements which can be verified and combined
[Habert, Nazarenko, Salem, 1997: 240]. Yet empirical and quantita-
tive approaches to language events, partly or largely automated by the
growth in linguistic engineering, cannot take place without reflecting
critically on data, formulating plausible hypotheses and taking into
76 Pierre Fiala
account the theoretical knowledge accumulated from previous work

and related areas. The remarks which follow deal with the field
of sociopolitical discourse analysis1 (SPDA) in its close relation-
ship with lexicometric quantification and enunciative approaches. We
should keep in mind that discourse analysis originated in France from
within the language sciences, based on a reformulation of the object
of lexicology and semantics. This reformulation stemmed from broad-
ening the notion of text and taking into consideration the ideological
and political positions of the enunciators2 [Mazière, 2005: 127]. So-
ciopolitical discourse analysis founded on lexicometry is therefore an
important paradigm, if not the unique source of discourse analysis in
France.
How to define political discourse?
Ever since the Enlightenment and the French Revolution, political

discourse has been a familiar tool of society. Yet, as soon as one at-
tempts to dislodge it from the vicious circle in which it is contained
(“political discourse is the discourse held by politicians, in political
institutions”)3, or to update the definitions of classical rhetoric4, its
definition is far from being clear-cut.
1 The focus of SPDA is not confined to an institutional definition of political dis-

course in the restrictive framework of the spatial structures of the state and Aris-
totelian divisions of discursive genres. Based on a sociology of linguistic diver-
sity at the core of the political sphere, it aims to link up data which are massive
and quantifiable, circulating and repetitive (programmes, propaganda, formulae
and slogans) with less obvious language events that are rare and less measurable
in appearance (morphological intersections, semantic encounters, implicatures,
connotations or enunciative positioning).
2 Francine Mazière has recounted the main stages and events of this in great detail.
3 This “default” position is often not a problem, and appears to be adequate in
political science or history, where it can be perfectly justified.
4 In relation to this, see the stimulating collection: Bonnafous et al., 2003.
Political discourse analysis 77
Jean Dubois [1962: 489], who was the instigator of vocabulary

and sociopolitical discourse analysis in France, abandoned the idea
of a preconceived definition, to give it a dual characterisation, ex-
ternal and internal:
First of all by virtue of the situation: a discourse is political because it is the

object of a political reading. What defines it as political is not a particular lexi-
con, certain types of arguments or themes […], it is the fact that the speaking
subject, the speaker who constitutes that subject, wants the listeners to read it
politically […] or it is the fact that the readers-listeners read a text politically.
[…]. Secondly, through its internal structure: political discourse is a specific
discourse because two basic rhetorical forms have interpenetrated in order to
create it: it is first of all a didactic discourse aiming to persuade, so that the
reader-listener makes the arguments which are presented his/her own, as worthy
assertions of universal truths […] Next, the political discourse borrows its dual
structure from polemical discourse: on the one hand the author refutes and com-
bats his/her opponent’s statements […] and on the other s/he presents assertions
which are contrary to these.5
Maurice Tournier, who was responsible for introducing political lexico-

metry6, sought a more specific basis for political discourse, as well as
greater precision in what could be observed and more verifiable meth-
ods and results. He insisted on the repetitive nature and sloganisation
which are particular features of political discourse. These have a close
resemblance to the messages of advertising and make it particularly
open to statistical calculation and probabilistic methods:
Most of the time a political message fits into a tissue of organised redundancies.
The term “sloganisation” refers to all those moments when discourse turns back
on itself and repeats what has already been said, thus becoming hardened into a
set of verbal hammerings which constitute the primary expression of the mes-
sage being delivered.7
Using a sociological approach to social fields, Christian Le Bart8 has

attempted to describe discourse as a genre: he depicts its structural
5 In the foreword (p. 16) to: Marcellesi, 1971: 16.
6 Tournier, 1992: 133.
7 Tournier, 1992: 133.
8 Le Bart, 1998.
78 Pierre Fiala
laws, rules of confrontation and effects of dominations, linked to a

given social system and producing invariants, stereotypes and myths.
Among these regularities he distinguishes odd variations, individual
enunciation strategies, examples of particular expertise and different
styles. For all this, does he succeed in identifying the genre of political
discourse and the varieties encountered or produce a method other
than textual commentary to establish the typologies he is proposing?
The same question may be asked about Patrick Charaudeau’s
approach [2005: 256]. His approach is in fact not very different from
Le Bart’s since he starts from a sociological definition of political
discourse as a social practice in power building and management. He
defines its internal and external constraints as being of a totally separate
genre. He then attaches importance to several extensive typologies:
the discursive strategies of the political domain, the various types of
ethos of “politicians” and linguistico-rhetorical processes, finishing
by finally noting the present-day decline of the political as myth.
Caught between the sociological and political science approaches
and the semiological and linguistic perspectives, political discourse
appears to be depicted with a set of heterogeneous and extensive fea-
tures which manage to convey an outline, without being able to de-
scribe the content or form. Its protean essence is elusive, but there is
no doubt about its existence in language practices and forms.
Characteristic of present-day political discourse analysis is the
importance given to the linguistic dimension of the political, grounded
in the language sciences and articulated with other disciplinary ap-
proaches. Sociopolitical discourse is a particular combination of eve-
ryday language and various specialised languages. As such it is pro-
foundly composite, being technical, didactic, polemical, empathetic,
ethical and performative all at the same time. Its utterances are self-
evident, yet it simultaneously conceals what it is, through lexical,
grammatical and enunciative combinations: it unites saying, doing,
being, power and duty, knowledge, and the numerous means of ex-
pressing necessity, desirability, possibility and certainty, negation and
the grammatical person. Paradoxically, it is personalised and collec-
tive, intentional and without a stable subject, assertive and injunc-
tive, finalised, strategic and tactical – yet pierced through with lin-
guistic excesses and slips of the tongue. It declines the discursive
forms of ethos, pathos, commitment, promise, the gnomic present
and the predictive future by combining them. Ritualised and rhetori-
cal, it monopolises signifiers (signifiants) by hammering out slogans
and expressions of uncertain meaning and by the rarity of solemnised
or unexpected utterances; it imposes the signified (signifiés) by the
systematic use of presupposition, implication/insinuation and seman-
tic underdetermination, but also by using procedures of reframing,
redefinition and metaphorical shifts which allow it to resolve, or even
the opposite, to cultivate contradiction.9
Content analysis or discourse analysis?

A distinction that is still valid
Political discourse was a focus of empirical work in the Anglo-Saxon

countries, before it was conceptualised in France. Designed as a propa-
ganda and psychological conditioning tool, it was the subject of the
first quantitative content analyses10 from the mid-twentieth century
on, linked to the mobilisation of American institutions against Nazi
propaganda, and then against Soviet discourse during the Cold War.
This process of global semantic analysis was extended beyond the
political domain as part of the General Inquirer program [Stone et al.,
1966], and even in more recent work on general semantic ontologies11.
Semantic analysis charts were developed as part of Anglo-Saxon so-
cial psychology, independently of specific reflection on language and
discourse: they are intended to measure the cognitive, argumentative
9 For a group reflection on the “Catégories pour l’analyse du discourse politique”,

see Semen 21, coordinated by Philippe Schepens, April 2006.
10 Lasswell et al., 1949.
11 <http://wordnet.princeton.edu/>.
80 Pierre Fiala
and ideological content of any text from charts established before-

hand. In France, the adaptations and presentations of problems12 of
content have moved closer to language problems. But the standard-
ised and generally non-contextual, non-situational organisation of se-
mantic categories and the non-linguistic treatment of elements of dis-
course13 demonstrate, if not a weakness, at least a profound contrast
with the tradition of discourse analysis. This, as we have seen, is
itself linked to politics. It owes its origins partly to the crisis in struc-
tural linguistics and formalism, but also to the work of lexicology
and political semantics on the French Revolution, the history of the
labour movement, and the social crises of the 19th and 20th centuries
[Dubois, 1962; Marcellesi, 1972; Tournier, 1972]. Its relationship to
the crisis of ideologies and political systems at the end of the 20th
century played a further role. The critical orientation of discourse
analysis is very different from the benchmarks of content analysis.
Faced with the restrictive, immanent and static Saussurian definition
of language, SPDA has constructed a discourse object based on a
variety of linguistic methodologies, such as distributionalism,
enunciativism, pragmatics and argumentology. However, it also rests
on more fundamental principles like discursive formation, discursive
genre, language functions, discursive ethos theory and places. SPDA
explores particular instances of language such as reported discourse
and autonymy; and it produces and develops original concepts like
collective speaker and ethos. It is therefore not possible to consider,
as L. Bardin [1998] does in his presentation, all processes of dis-
course analysis to be intrinsic to content analysis, without noting their
contrasts, or even their contradictions.
12 See in particular Unrug, 1974.

13 In France there are computerised tools which claim to relate more particularly to
content analysis (Cordial, Tropes, Sphynx, Prospéro). These purport to synthe-
sise the semantic content of data, as well as to summarise and compare them, and
to chart the dominant semantic fields of a text, representing them from a pre-
established semantic ontology. Although these tools have undeniable advantages,
they also produce generalisations and normative judgments that are sometimes
extremely rudimentary and incorrect.
Lexicometric measurement, an empirical

quantitative approach to SPDA
Lexicometry developed from the late 1960s as a specific component

of SPDA, in which could be found the various perspectives of tex-
tual hermeneutics which had emerged from problems raised by con-
tact between the social sciences and linguistics. Rejecting a priori
theorising about discursive processes, lexicometry initially offered14
a contribution which was both positive, in terms of verifiable results,
and critical as regards theoretical systems. Distinct from linguistic
discourse analyses15 and likewise from content analysis and textual
semiotics, it defined a strategy for interpreting political texts: this
consisted in verifying, by means of statistical measurements16, the
vocabulary used for historical, sociological and socio-political hypo-
theses, outside of linguistics. Lexicometry has developed an origi-
nal empirical approach within the language sciences in that it groups
a diversified set of automated documentary practices and statistical
problems around a restricted number of unifying principles; this has
14 At a time when the means of making calculations were still very limited.
15 “The fact that extra-linguistic and linguistic elements may be articulated in the
utterance does not imply that they are homogeneous; and attempts to treat them
using a single methodology may be dubious. Thus “discourse analysis” does not
seem to have resolved the problem of moving from one order to another. It ap-
pears to be a murky area which conceals the shifting of a linguistic problem into
a content problem” (Des tracts en mai 68, 1978).
16 The various documentary modules – simple indexation, triage of various con-
cordances, inventories of repeated segments; measuring lexical distances between
parts of a corpus by AFC or hierarchical classification, specificities, co-occur-
rences, lexicograms – all these can provide explicit standards for the quantifica-
tion of texts which are just as reliable, if not more so, as the norms for measuring
public opinion in surveys.
82 Pierre Fiala
ensured methodological continuity17 as well as extending its reach.

These principles concern the three phases of the activity: data col-
lection, computerised programs and the interpreting mechanism.
Data collection is not unlike the sample-gathering practise used
in socio-linguistic surveys. The data consist of written or transcribed
texts, in substantial numbers, which have been carefully identified.
The collection is based on various criteria – of representation and
volume, but also situational, enunciative and thematic homogeneity.
Automated comparison means calibrating the textual data, standard-
ising the punctuation and written forms, and possibly morpho-syn-
tactic labelling. What is known as lexicometric reduction consists of
the following operations:
a. Reducing texts to a set of words (written forms, labelled forms
and lemmas).
b. Delineating the textual corpus and reducing it to comprehensive
tables of forms matching their frequency.
c. Constructing an index and concordances.
d. Choosing invariants and variables for partitioning the corpus.
e. Partitioning the corpus and making a statistical comparison of
the distribution of forms in each part.
f. Analysing the strong statistical trends identified by probabilistic
methods.
g. Analysing globally, or by grammatical category, lexical tables,
tables containing marks of individuals, spatio-temporal signs,
verbal modalisation, negations, reported discourse, language in-
stances, argumentative markers, appreciatives and axiologies, etc.
17 The quantitative method in lexicology originated in France with the work of

P. Guiraud and Ch. Muller (1977), and has undergone various developments ac-
cording to the objectives pursued. For example, E. Brunet, in his lexical statistics
work involving the literary corpuses on the Frantext base in Nancy, has identified
the general quantitative characteristics of the language used by major authors,
such as Hugo, Zola and Proust. He has analysed the contrasts between works by
the same author, as well as the global characteristics and diachronic variations of
the whole Nancy corpus (1981). He has not, however, provided a systematic
interpretation of the statistical results.
h. Interpreting by means of the semantic reconstruction of frequency

calculations, lists, distributions and probabilistic indicators, in
diachronic, ideological and stylistic frameworks.
The validation of external hypotheses requires homogeneous corpuses
to be formed, as regards their conditions of enunciation and thematic
unity. However, the variations they present may be capable of being
reduced to simple variables, which can be neutralised according to a
choice of research hypotheses: diachronic variations, enunciator vari-
ations, types or genre of discourse, or divisions internal to the texts.
Whatever the size and nature of the contextual material or field
of archives from which the lexicometric corpus is extracted, this cor-
pus must be carefully distinguished and isolated by precise enunciative
hypotheses. The interpretation of results, at least in the first phase,
rests exclusively on the measured data and not on their textual envi-
ronment, nor on general language characteristics. This rule of en-
closing the corpus, known as an endogenous or intrinsic norm, goes
beyond a simple principle of immanence. It strictly limits the range
of interpretive inferences: statistical comparisons serve only to es-
tablish frequency contrasts between parts of a corpus which is lim-
ited, dated and linked to a theme, with enunciative characteristics.
This principle avoids unwarranted generalisations but does not
prevent some general frequency characteristics of language being iden-
tified. For example, the distribution of high frequencies of grammatical
words reveals structural characteristics which can be verified in most
corpora. The notion of differential use18, based on frequencies, ena-
bles diversity to be thought of in a homogeneous way (historical strata,
sociological groups, effect of situation), in a relationship not of ex-
clusion from the language, but as linked to it19. Political vocabulary
18 Milner, 1989: 121.

19 “Usage is opposed to structure (schema) of language because it is the intrusion of
historicity, of verbal activity by people, into the permanence of language. Rather
curiously, it is also related to freedom of innovation and to freezing due to repeti-
tion; it disrupts the regular pattern and innovates at the same time” (Greimas,
Keane, Cahiers de lexicologie, 56, 1991).
84 Pierre Fiala
and discourse may be described as a set of particular concrete uses of

language, linked to historical and social situations in which social
groups and individuals are engaged or involved in power relation-
ships. In another interpretation phase, nothing prevents the lexico-
metric corpus being re-inscribed in a broader archival field where it
can function as a place of experiment and verification or reveal symp-
tomatic functioning20. Lexicometric interpretation thus takes its place
within the semantics of discursive usages, rather than in general se-
mantics. It is particularly well-suited to quantitative processing of
some of the traces of enunciative operations.
Mapping levels of enunciation21
As is well known, Bakhtine was the first to establish enunciation as

the central notion for approaching the semantics of discourse in gen-
eral. However, in France, it was Emile Benveniste who suggested
principles for founding a quantified approach to political discourse,
with the notion of levels of enunciation, outlined22 by him in 1956
and set out in detail in 1959. This notion described23 the two distinct
20 In what J. Guilhaumou calls the “corpus moment”, a few lexicometric results,

compatible with more general interpretive hypotheses, are taken as samples for
analysis and linguistic experimentation (1989).
21 On this subject, see: “Topographier des plans énonciatifs dans des réseaux
épistolaires” in Semen No. 20, “Le rapport de place dans l’épistolaire”.
22 “A completed personal utterance is therefore formed on two levels: it uses the
denominative function of language for the object references which this function
establishes as distinctive lexical signs, and it constructs these object references
by means of self-referential indicators corresponding to each of the formal classes
recognised by the idiom” (E. Benveniste, 1966 [1956], p. 253).
23 “The historical level of enunciation may be recognised from the fact that it im-
poses a particular limit on the two verbal categories of time and person taken
together” (E. Benveniste, 1966 [1959], p. 239).
modes of enunciation represented by historical narrative and discourse,

as well as the intermediate mode of reported discourse, with its tem-
poral disjunction. The author insisted on the fact that the distinction
between these different levels was understandable in statistical terms.
Enunciation level refers to a combination of markers which are
spatio-temporal, aspectual, modal and personal, deictic or contex-
tual, organising and locating designations, and forming contexts for
discursive exchanges in given genres and specific language practices24.
The image of verbal exchange as a combination which is more or less
constrained, as a configuration of different enunciative levels, reveals
discourse as a series of overlapping surfaces (or curves) embedded in
measurable reference points and which can be evaluated in statistical
terms. It also and above all enables us to identify enunciative net-
works within social groups, where markers and levels of enunciation
have momentarily stabilised around factual subjects or argumenta-
tive issues. It is through describing these configurations and networks
that we can represent the growing constraints in the relationship be-
tween word and place, not only from an interpersonal perspective,
but in its social determinations, which a sociology of languages25 can
identify in terms of language practices.
Where, thirty or so years ago, there were only limited schematisa-
tion tools available, today the instruments for automatic processing
of discourse26 are able to provide mapped tables of measurements
and comparisons for corpuses: thus the Correspondence Factor Analy-
sis used by Hyperbase [Brunet], Lexico3 [Salem] and Alceste [Reinert]
24 Benveniste returned to this in his conclusion to “l’Appareil formel de l’énoncia-

tion” (1970): “[written enunciation] moves on two levels: the writer enunciates
herself by writing, and within her writing she causes individuals to enunciate
themselves” (1970: 88).
25 From the perspective of a sociology of languages inaugurated at the time by
Jean-Pierre Faye (1973), distinct from Marcel Cohen’s early essays and, in cer-
tain respects, foreshadowing the ideas of Bourdieu.
26 In the context and following the example of Automated Language Processing
(ALP), ADP has brought together automated tools, notably lexicometric tools,
for processing discursive corpuses from a comparative angle.
86 Pierre Fiala
are bidimensional projections by approximation, of multidimensional

statistical characteristics of corpuses; corpus partitioning cards
(Lexico3) or the recursive lexicograms of WEBLEX [Heiden] offer
representations of lexical co-occurrence networks and lexico-seman-
tic clusters in a corpus. Computerised tools can produce digital, lin-
ear or pluridimensional representations and typologise those levels
which structure exchanges. Applied to a variety of sociopolitical data,
these tools provide global representations of enunciation levels. The
results enable researchers to contrast the language practices involved
and to identify the characteristics of new varieties of sociopolitical
discourse, where places are less determined than in traditional ex-
changes27.
The primary lexicometric task now is, on the one hand, to develop
questions capable of being quantified (not all are) together with appro-
priate hypotheses, and to accumulate data and comparable results. On
the other, we need to compare different tools, various experiments and
partial results. We also need to demonstrate the mapping of successive
measurements, proceeding by progressively moving closer until we
have grasped the most subtle phenomena, in a transition from static
analysis of repeated segments towards dynamic analyses of termino-
logical constructions on the one hand, and of phraseology circulations
on the other. These are the conditions, among others, which will per-
haps enable us to move from measuring words and texts to compre-
hending an object, which is discourse; for discourse, as much as lan-
guage, results from a construction of multiple viewpoints.
27 The neologism ‘topographier’, the French word for ‘mapping’ combines two
distinct values of the noun ‘topographie’ from which it derives. There is the tra-
ditional sense of “description of a place, of its characteristic features” linked to
the rhetorical figure referring to “the art of representing on paper the configuration
of a terrain”; and, in addition, there is the technical meaning that is usual today,
i.e. “a method which involves drawing a small-scale version of the map or plan
of a terrain, supposing the land to be flat”. From this we obtain the verb ‘topo-
graphier’ which concerns our present thinking, and which consists of “represent-
ing in the form of drawings, maps, graphic schemas, metrics of textual, lexical,
morpho-syntactic, enunciative and thematic features, in order to identify and
compare the characteristic configurations”.
AMOSSY R.
2000: L’argumentation dans le discours. Discours politique, littérature
d’idées, fiction, Paris, Nathan.
ANGENOT M.
1989: 1889, un état du discours social, Québec, Le préambule.
BARDIN L.
1977: L’analyse de contenu, Paris, PUF.
BENVENISTE E.
1966, 1974: Problèmes de linguistique générale I et II, Paris, Galli-
mard.
BIBER D., CONRAD S. and REPPEN R.

1998: Corpus Linguistics. Investigating Language. Structure and Use,
Cambridge University Press.
BLOMMAERT J.
2005: Discourse, A Critical Introduction, Cambridge University Press.
BONNAFOUS S.
1991: L’immigration prise aux mots, Paris, Kimé.
BONNAFOUS S., CHIRON P., DUCARD D. and LÉVY C. (ed.)
2003: Argumentation et discours politique. Antiquité grecque et latine,
Révolution française, Monde contemporain, Presses universitaires de
Rennes.
BRANCA-ROSOFF S. and SCHNEIDER N.

1994: L’écriture des citoyens. Une analyse linguistique de l’écriture
des peu-lettrés pendant la Révolution française, Paris, Klincksieck.
88 Pierre Fiala
CHARAUDEAU P.
2005: Le discours politique. Les masques du pouvoir, Paris, Vui-
bert.
CHARAUDEAU P. and MAINGUENEAU D. (ed.)

2002: Dictionnaire d’analyse du discours, Paris, Seuil.
CHATEAURAYNAUD F.
2003: Prospéro. Une technologie littéraire pour les sciences humaines,
Paris, CNRS.
COLLECTIF SAINT-CLOUD
1978: Des tracts en mai 68, Paris, Champ Libre.
1982: La parole syndicale, Paris, PUF.
1998: Le syndicalisme à mots découverts. Dictionnaire des fréquences
(1971-1990), Paris, Syllepse.
DÉTRIE C., SIBLOT P. and VERINE B.
2001: Termes et concepts pour l’analyse du discours. Une approche
praxématique, Paris, Champion.
D’UNRUG M.C.
1974: Analyse de contenu, Paris, Éditions universitaires.
DUBOIS J.
1962: Le vocabulaire politique et social en France de 1869 à 1872,
Paris, Larousse.
BRUGIDOU M., ESCOFFIER C., FOLCH H., LAHLOU S., LE ROUX D.,
MORIN-ANDREANI P. and PIAT G.
2000: “Les facteurs de choix et d’utilisation de logiciels d’Analyse
de Données Textuelles”, JADT 2000: 5 International Days for Statis-
tical Analysis of Textual Data.
BRUGIDOU M. and LABBÉ D.

2000: “Le vocabulaire syndical français à la lumière de l’analyse des
données textuelles et de la statistique lexicale”, JADT 2000.
FABRE C., H ABERT B. and ISSAC F.

1998: De l’écrit au numérique. Constituer, normaliser et exploiter les
corpus électroniques, Paris, Masson.
FAIRCLOUGH N.
2003: Analysing Discourse, Textual analysis for social research, Lon-
don, Routledge.
FAYE J.-P.
1972: Langages totalitaires, Paris, Hermann.
FIALA P. (ed.)
1999: In/égalité/s. Usages lexicaux et variations discursives, Paris,
L’Harmattan.
HABERT B., NAZARENKO A. and SALEM A.

1997: Les linguistiques de corpus, Paris, Armand Colin.
GUILHAUMOU J.
1989: La langue politique et la Révolution française, Paris, Klincksieck.
GUILHAUMOU J., MALDIDIER D. and ROBIN R.

1994: Discours et archive. Expérimentations en analyse du discours,
Liège, Mardaga.
KRIEG-PLANQUE A.
2002: “Purification ethnique”. Une formule et son histoire, Paris,
CNRS Editions.
LASSWELL H. and assoc.

1949: Language of Politics, Studies in Quantitative Semantics, Cam-
bridge, MIT Press.
90 Pierre Fiala
LABBÉ D. and MONIÈRE D.

2003: Le discours gouvernemental. Canada, Québec, France (1945-
2000), Paris, Champion.
LEBART L. and SALEM A.

1994: Statistiques textuelles, Paris, Dunod.
LE BART C.
1998: Le discours politique, Paris, PUF, Que Sais-je?
MARCHAND P.
1998: L’Analyse du discours assistée par ordinateur, Paris, Armand
Colin.
MARCELLESI J.B.
1971: Le congrès de Tours, Décembre 1920. Études sociolinguistiques,
Paris, le Pavillon.
MAYAFFRE DAMON
2000: Le poids des mots, Paris, Champion.
2004: Paroles de président, Paris, Champion.
MAZIÈRE F.
2005: L’analyse du discours, Paris, PUF, Que Sais-je?
MEYER F.C.
2002: English Corpus Linguistics. An introduction, Cambridge Uni-
versity Press.
MULLER CH.
1977: Principes et méthodes de statistique lexicale, Paris, Hachette.
PÊCHEUX M.
1990: L’inquiétude du discours, texts chosen and presented by Denise
Maldidier, Paris, Éditions des cendres.
SCHIFFRIN D., TANNEN D. and HAMILTON H. (ed.)

2001, 2005: Discourse Analysis, Oxford, Blackwell.
TOURNIER M.
1975: Un vocabulaire ouvrier en 1848. Essai de lexicométrie, Saint-
Cloud, an Ecole Normale Supérieure publication in 4 volumes.
1992, 1997, 2003: Des mots en politique. Propos d’étymologie sociale
1, 2, 3, Lyon, ENS-éditions.
TROGNON A. AND LARRUE J. (ed.)
1994: Pragmatique du discours politique, Paris, Colin.
VAN DIJK T.
1997: Discourse Studies: a multidisciplinary introduction, 2 vol. New
Delhi Sage.
WILLIAMS G. (ed.)
2005: La linguistique de corpus, Presses universitaires de Rennes.
WODAK R. (ed.)
1989: Language, power and ideology, studies in political discourse,
London, Benjamins Publishing Company, 1989.
Journals
Corpus, “Les corpus politiques: objet, méthode et contenu”, no. 4,

2005. Annual journal published by UMR 6039, University of Nice,
Edizioni dell’Orso, 5 editions published.
Langages, No. 13, 23, 37, 41, 52, 62, 71, 81, 117, Paris, Larousse.
Langages et société, “Le politique en usages”, No. 113, September
2005, Paris, MSH.
Langue française, No. 9, 28, 103, Paris, Larousse.
92 Pierre Fiala
Le Français dans le monde, 1996, “Le discours: enjeux et perspec-

tives”, special edition, EDICEF.
Lexicometrica, online journal, <http://www.cavi.univ-paris3.fr/lexico
metrica/>.
Marges Linguistiques, “Analyse du discours”, No. 9, May 2005.
Mots, Les langages du politique, 80 editions devoted to lexicometry,
lexicology and political discourse analysis, 1980–2006, Lyon, ENS-
éditions.
Semen, “Catégories pour l’analyse du discours politique”, No. 21,
April 2006, Journal of semiolinguistics of texts and discourses,
Besançon, Presses universitaires de Franche-Comté.
Texto, online journal, <http://www.revue-texto.net>.
93
Chapter VI
Analysing controversy.
The contributions of argumentation study
to political science
Juliette RENNES
What is a political controversy? The answer to this question depends

partly on which discipline is asking it. In political sociology, the defi-
nition will tend to emphasise the process by which a polemic, relat-
ing to a public decision or political action to be undertaken, moves
away from its initial localised “arena” to affect a more general public
[Cobb and Elder, 1972; Favre, 1992; Cefaï, 1996; Barthe, 2006 ]. In
discourse analysis, stress will be laid upon the cluster of recurrent
contradictory arguments which mark the discursive identity of the
controversy across these different arenas of public debate. In this
chapter, I would like to outline a few ways in which the first ap-
proach can benefit from the second. By basing what I have to say
more particularly on an argumentative and rhetorical analysis of what
might be called equality controversies in democracy, I hope to cast a
discursive light on a certain number of questions raised by political
sociology.
The life of a controversy is generally broken down into several
phases, all of which give rise to a whole battery of sociological ques-
tions. These relate to the conditions under which the controversy
emerges, the configurations of actors supporting it, the series of posi-
tions it creates and the partisan groupings established, as well as to
the differences and shared features of the way it is formed into a
militant, institutional and media narrative. Moreover, this sociologi-
cal questioning deals with the processes by which controversy mobi-
lises public opinion, occupies the political and media agenda and
94 Juliette Rennes
becomes dispersed through various sites of public debate. It also

examines the possible effects of controversy in shaping the militant
and political scene, and the factors which explain its possible closure
or, on the other hand, its revival.
Certain “tool boxes” developed in the wake of discourse analy-
sis may contribute to understanding these processes. If, from a cogni-
tive point of view, it is agreed that political controversy can only
emerge and last as long as it expresses and sets up a tension between
two or more potentially contradictory value systems of comparable
public legitimacy, and that the supremacy of one or other of these
value systems is precisely what divides the opposing “camps” – then
the concepts of interdiscursivity and type of argument can be used to
advantage. If we were to add that this opposition between “camps”
changes its aspect according to the arena in which it is expressed,
then work on the discursive institution will also prove useful. These
three analytical tools, which I propose to examine here, do not con-
cern all discursive aspects of a controversy equally. While the no-
tions of interdiscursivity and discursive institution enable us to un-
derstand discourses at all levels of their textuality (word, utterance,
text structure, argument, rhetorical figures), the notion of type of ar-
gument is more suited to explaining certain discursive sequences.
Interdiscursivity
The concept of “interdiscursivity”, introduced by Bakhtine [Todorov,

1981], has been further developed in France by Pêcheux [1975] and
Maingueneau [1983], and is now widely used in discourse analysis1.
It describes the state of “being in multiform relation with other dis-
courses” as the “constituent characteristic” of all discursive produc-
1 In Critical Discourse Analysis, this concept is also widely used even if Norman
Fairclough prefers the concept “order of discourse” to analyse similar discursive
phenomena [Fairclough, 2003].
Analysing controversy 95
tion [Maingueneau, 2002, p. 324]. This characteristic is reflected at

the level of the word as well as at the level of the discursive unit, and
it refers to the varied traces of “other” discourses – produced prior to
or at the same time as the text analysed. In a particular discourse
issuing from a given ideological formation2, it may concern the pres-
ence of a lexical register, metaphors or arguments which run through
the whole of the discursive production of this ideological formation
or of related ideological formations (there is a shared rhetoric for
demands for equality in democracy, whether these demands concern
women, foreigners, homosexuals, and so on). Or it may be an oppos-
ing discourse that is present, summoned up through processes of refu-
tation, concession or citation, or in polemical retranslations of its
rhetoric: pastiche, parody, derision, demystification, etc. Again, it may
concern doxic discourse on a given subject, through that unrecog-
nised repetition of what has already been said which we know as
“commonplaces”, as well as the various forms of presupposition. From
these examples, it is clear that discursive otherness may be explicit –
signalled by the use of quotations, inverted commas or reported dis-
course – as well as implicit: the reprise of commonplaces or ideo-
logical places generally intervenes without mention of a “discursive
heterogeneity” [Authier-Revuz, 1982].
From the perspective of discourse analysis, a controversy is an
interdiscursive subject where opposing arguments are constantly
present at the horizon of the argumentation process. The traces of
this presence are not only polemical figures, but also what might be
called figures of constraint to indicate that the antagonistic discourse
is something to be taken seriously, something to be reckoned with
because of a common base of values and principles shared by the
camps present. We have said that a sociopolitical controversy is the
crystallisation of a conflict of value hierarchies, each of which has
2 The notion of “ideological formation” simply means here “any set of socially
and historically circumscribed utterances which can be connected to an enuncia-
tive identity. Examples would be communist, feminist, employers’ discourse and
so on. See D. Maingueneau’s article ‘formation discursive’” (Charaudeau and
Maingueneau, 2002: 271).
96 Juliette Rennes
comparable legitimacy in the public sphere. From this viewpoint,

any plea for one value to take precedence over another must take into
account the opposing value hierarchy.
Let us take as an example equality controversies, that is, contro-
versies in which conflicting claims for equal rights are confronted
with specific practical and discursive resistance devices in regimes
constrained by egalitarian and liberal standards. The conflict of values
at stake in this type of controversy involves a struggle for precedence
between equality and other values claimed by democracies in the
past two centuries. There are nuances, variations and circumstances
relating to specific situations, but equality controversies tend to set
one camp, which insists that the egalitarian principle must have
priority in settling disputes, against another which considers that, al-
though this statement is acceptable “as a general rule”, in the particu-
lar issue under debate, the application of this egalitarian principle
strikes a blow against another shared value. This was why “nation-
hood” was invoked to counteract egalitarian demands during contro-
versies over the equality of political rights between resident foreigners
and French national citizens; and, in the same way, “sexual differ-
ence” – or heterosexual conjugality as a dominant value in the or-
ganisation of contemporary societies – was invoked to counteract the
equal status of same-sex and mixed-sex unions, introduced during
the 1980s and 1990s in several Western European countries [Borillo,
2001]. A third example is the way in which “the secularist principle”
was brought into play during French debates on equal access to schools
both for those wearing religious symbols and those not wearing them3.
3 It should be noted, however, that during this controversy, the egalitarian prin-
ciple was invoked both by the opponents and those in favour of the law forbid-
ding the wearing of religious symbols in schools. Even though, in the case of the
latter, the secularist argument was paramount, some of these proponents, mili-
tant feminists in particular, put forward the principle of sexual equality in favour
of the law: authorisation girls to wear the veil to attend school was seen as con-
firming the preconceptions of inequality between sexes conveyed by the wearing
of the Islamic headscarf.
During these disputes, those upholding the primacy of the egali-

tarian principle have to take into account the value put forward by
their opponents: for example, by stating that “sexual difference”, “na-
tionhood ” or “the secularist principle”, as understood by the oppos-
ing discourse, would be preserved within a regime of equality. Alter-
natively, a heterodox definition of these “value-words” might be
proposed, compatible with their egalitarian leanings4. On the other
hand, their opponents tend to propound a polemical definition of
the word “equality”, with the assurance that the egalitarian principle
is well and truly safeguarded in a regime of differential rights, for
example in the form of an “equality within differences”5.
This framework of constraint exercised by the opposing axiologi-
cal system is shaped by tropes and identifiable forms of reasoning.
Latching on to an opponent’s word in order to give it a polemical
definition is a typical device used in legal oratory, known as antana-
clasis. It recurs in equality controversies dealing with those value-
words “equality”, “liberty”, “progress”, “democracy” and in France
“the Republic”. These latter are also the subjects of semantic battles
between the opposing camps, which all wish to claim them. An ex-
ample of this is the way in which French mayors, in a petition launched
in April 1998 opposing the Pacte Civil de Solidarité (PACS), invoked
“the defence of Republican marriage” to counter legal homosexual
unions. Their opponents, on the other hand, used the Republican prin-
ciple of “equality before the law” to argue against any discrimination
based on sexual orientation.
4 During the headscarf debate, the definition of secularism as an attribute of the

educational institution (and not of its pupils) meant it was possible to state that
the expression of pupils’ religious beliefs was compatible with the upholding of
secularism.
5 This was an argument put forward, in particular, by those opposed to feminist
demands for political, civil and professional equality of rights before 1945. It is
to be found again countering homosexual claims for access to marriage and parent-
hood and, in another politico-discursive context, it was used to justify racial segre-
gation in the United States (Fassin and Fabre, 2004).
98 Juliette Rennes
What we might call a designational battle involves a slightly

different process: it consists in trying to impose one’s own designa-
tions over those of an opponent in order to describe the same reality,
so that one’s own predefinition of the question being debated and its
solutions may be shown to good advantage. In controversies over
political, civil and professional equality between the sexes in France
from 1880 to 1945, it was a feminist ploy to attack the designation of
an opponent who talked about the “splendour” of domestic tasks as
the exclusive vocation of the middle-class woman. Renaming a house-
wife’s activities as part of a register of slavery, physical effort and the
repetitive nature of humdrum tasks aimed to reveal the opponent’s
lyrical paean to the “domestic goddess” as a discourse of oppression.
In this way, feminists were concerned to show that domestic activity
did not constitute a vocation or a destiny. In a quite different, more
contemporary context, renaming a “clandestin” [illegal immigrant]
as a “sans papier” [without ID] as has been done in French, helps to
tip the narrative of the guilty party, one who breaks laws, towards
one in which the person is a victim of iniquity, someone who has
been deprived of their right to have an ID but has a legitimate claim
to their rights. Renaming what those in favour of a law against wear-
ing religious symbols in schools call the “voile islamique” [Islamic
veil], as a “fichu sur la tête” [headscarf], as in the case of several
opponents of this law6, is to reduce the opposing arguments to the
simple rationalisation of “collective hysteria” [Terray in Nordmann,
2004] in which bits of cloth are seen as fetishes.
Forestalling the opponent’s objections “by anticipating them and
demolishing them in advance” [Dupriez, 1984, p. 362] is also typical
of arguments involving controversy, which are constantly shaped by
the presence of a counter-discourse to be refuted. This rhetorical de-
vice is known as a prolepsis, and it can lead to concessions which
consist in acknowledging the validity of the opposing arguments while
at the same time preserving one’s own conclusions. For example, one
6 See in particular Bourdieu, “Un problème peut en cacher un autre”, Delphy, “Une
Affaire française” and Terray “L’hystérie politique” in Nordmann (ed.), (2004).
can grant to those opposing homosexual parenthood that the “couple’s

stability” is necessary for “child development”: one can then work to
produce statistical surveys which show that same-sex couples are as
stable (or unstable, as the case may be) as heterosexual couples are.
One can acknowledge, along with the opponent, that secularism in the
public arena of the Republic must be safeguarded, but also aim to show
that the exclusion of young women wearing the veil risks damaging
that secularism, by reinforcing pockets of religious radicalism.
As in this second example, concession can easily tip over into
retaliation: the validity of the opponent’s objection is conceded, but
then it is turned against the opponent and transformed into an argument
that counters his or her viewpoint. In the first half of the 20th century,
controversies over gender equality before the law, for example, saw
feminists turning the argument of “female difference” against their
opponents – an argument which, until then, had always been appealed
to in order to exclude universal rights being applied to women. Invok-
ing difference then became a utilitarian argument about the contribu-
tion of the “feminine” element to a world of power and knowledge
which had until then been dominated by “masculine” values [Rennes,
2007]. The retaliation argument can have the effect of recouping and
recovering opposing ideological positions. By confirming the idea of
female difference which is said to reside in “gentleness”, “altruism”,
the “sense of obedience” and “an eye for detail”, some of the partisans
of sexual equality ended up having difficulty in rejecting the socio-
political consequences that their opponents derived from these “femi-
nine qualities”, namely that persons of the female sex were predes-
tined to occupy subordinate positions. Pierre-André Taguieff, Ruth
Wodak and Teun van Dijk have shown the effects of comparable re-
coveries in relation to the antiracist retaliation against the racial argu-
ment of cultural differences between nationals and foreigners during
the 1980s [Taguieff, 1985, 1990, Wodak and van Dijk, 2000].
The identification of figures of constraint in political controver-
sies in general and equality controversies in particular, can be a means
of identifying the power relationships between the different camps in
the debate. Concessions, retaliations and prolepses are particularly
100 Juliette Rennes
present in dissident discourses which have constantly to contend with

the massive presence of the doxa as hegemonic “other” discourse
[Angenot, 1989, Fairclough, 2003]. When the belief in the “feminine
nature” holds sway over the debate, arguing outside of this belief
will certainly not bear fruit [Scott, 1996]. When the designations of
the heterodox camp tend to assert themselves over all the speakers,
then this is the moment which may signal, if not a turning in the
power relationship, at least a process in which the world view repre-
sented by the previously marginalised discourse comes to be seen as
legitimate.
This was the case when “sans papier” [without ID] began to re-
place the terms “clandestin” and “immigré irrégulier” [illegal immi-
grant] in the mainstream French press. Another example is when the
term “altermondialiste”, which appeared in the discourse of French-
speaking militants opposed to the neo-liberal globalisation of the years
2000, – usurped during 2002 the pejorative “antimondialiste”, which
had dominated the media until then.
Discursive institution
The concept of “discursive institution” is a variant of the concept of

“discursive genre”, and has been introduced by Dominique Maingue-
neau. It involves thinking in combination about “the operations
through which discourse develops its contents, and the institutional
mode of organisation that discourse both presupposes and structures”
[Maingueneau, 1995, p. 40]. This concept helps to identify the vari-
ety of arguments used in a controversy according to the institutional
sites in which this takes place. The same argumentative strategy will
not be used both in the mainstream press and in an academic debate,
in Parliament or in a television programme. What varies is not the
nature of the arguments put forward – from this point of view, all
controversies have great argumentative stability across the different
sites where they take place – but the way they are prioritised: a dis-
cursive institution tends to determine argumentative predilections and
a certain hierarchy of arguments.
Let us take as an example the argument from precedent. For those
in favour of equality, one form of this argument involves basing the
current scenario on a previous case of legal equality obtained by one
of the members of the category discriminated against, in order to
claim equality for this whole category of persons. This argument is
particularly used in deliberative proceedings or legal sites producing
public decisions7. In these same proceedings, opponents will in par-
ticular use what is called the “Pandora’s box”, or the “slippery slope”
argument [Walton, 1992]. The issue then is to dissuade the decision-
makers from adopting the egalitarian measure in the name of its harm-
ful consequences: giving authorisation to a category of people who
until now have been excluded from exercising such rights is in the
long run tantamount to according them all other rights, or creating a
precedent to which all excluded categories may lay claim. If those
opposing the egalitarian measure cannot find common ground with
their adversaries in relation to the proposed action under examina-
tion, they seek to find it by looking at the effects of these measures,
and by trying to persuade those in favour of equality that they would
not themselves accept the consequences of their egalitarian decisions.
Another example of the link between discursive institution and
argumentative predilection is “the argument of essence” [Plantin, 1991,
p. 311] which involves showing that it is in a person’s nature to act in
a particular way – women/foreigners/homosexuals are, by reason of
an internal determinism, destined to occupy their present position in
the social order. This argument is more forceful in essays and pam-
phlets which are less directly concerned by the decision being taken
and more concerned to perpetuate a certain doxa about the naturalness
of the social order. For reasons of form, when making a spoken con-
tribution to debates taking place in deliberative proceedings, such an
7 It thus constitutes a lead argument in dossiers compiled by associations involved

in “cause lawyering” (Sarat, Scheingold, 2001).
102 Juliette Rennes
argument is often merely presupposed. As for the argument about the

perverse effect, which, in equality controversies, involves proclaim-
ing that the apparent beneficiaries of an egalitarian measure are in fact
its main victims, this can be particularly well represented in discourses
addressed to the interested parties themselves. This was the case with
antifeminist fiction under the French Third Republic, which was es-
sentially aimed at a female readership: based on the principle of identi-
fication, the emancipated woman in this type of novel would be faced
with a gradual accumulation of personal misfortunes which would
lead her in the “happy end” to prefer to return to the safety of her
home8. Similar plots are found to a certain extent in the antifeminist
fiction of the 1980s [Bard, 1999].
Types of argument
The “taxonomic frenzy” of old rhetoric [Barthes, 1994] has handed

down to us a classification of types of argument. In Schopenhauer’s
The Art of Always Being Right [2004] as in the re-worked argumen-
tative typologies of the second half of the 20th century9, the clas-
sification of arguments by their logico-discursive structure, which
corresponds to the ancient “commonplaces” (topoï), sits alongside
typologies which also take into account contents and recurrent ar-
gumentative themes10. Putting these typologies into a socio-political
8 Among the plethora of antifeminist novels, we might cite Colette Yver (1874–
1953) who specialised in fiction condemning women who sought emancipation
by exercising a superior profession.
9 For example: Perelman and Olbrechts-Tyteca (1958), Angenot (1995), Plantin
(2005), Walton (1992) or van Eemeren and Grootendorst (1992, French transla-
tion 1996)
10 What we in modern times call “commonplaces” are often “specific places” from
the viewpoint of ancient rhetoric since they are schemas of reasoning with a
specified content (Aristotle, 1990).
perspective is a field which is still wide open to exploration: from

this point of view, heuristic avenues have been opened up by Jeremy
Bentham’s questioning of the “political sophisms” brought to bear in
the parliamentary assemblies at the beginning of the 19th century
[1962]; by Albert Hirschman’s questions about the link between his-
torical circumstances (the democracies which emerged from the En-
lightenment), ideological objectives (the struggle against egalitarian
offensives) and argumentative predilections (constituting a “reaction-
ary rhetoric”) [1991]; by Pierre-André Taguieff’s formalisations con-
cerning the argumentative and cognitive categories of racism and
antiracism [1985, 1990]; by Ruth Wodak’s analyses of different dis-
crimination discourses (see for example Reisigl and Wodak 2001) or
by Marc Angenot’s work on discursive constructions of the 19th–20th
century “grand narratives” [1993], their refutations [2003] and the
prototypical dimension of forms of reactionary thinking represented
by the “pamphlet” genre [1995].
However, certain argumentative forms are not exclusive to any
particular ideology. “The argument of perversity” and “the argument of
futility” which Albert Hirschman identifies as typical of anti-egalitarian
reasoning, in reality have their due place, with some variations, in
progressive rhetoric, as he partly acknowledges in the sixth chapter of
Rhetoric of Reaction. But in between socio-political objectives and
forms of argumentation, there are incontestably magnetisation phe-
nomena which produce effects of typicality. From this point of view,
the argumentative portrait of a political controversy which opposes
two main “camps” can partly be drawn from pairs of conflicting argu-
ments, revealing structural cognitive and ideological splits.
One of the issues which polarises these cognitive and argumenta-
tive oppositions in equality controversies relates to the similarities and
differences between the social groups demanding access to rights and
those who already hold those rights. The spokespersons in favour of
equality tend to rely on the common identity of these two groups: both
of them are “human beings”, “residents of France”, “citizens”, “work-
ers”, “taxpayers”, etc. In other words, it is the principle of “belonging
to the same essential category” – as Perelman and Olbrechts-Tyteca’s
104 Juliette Rennes
[1958] expression has it – which forms the basis of most of their

arguments. This principle can be found in the reciprocity argument,
“based on the principle of symmetry: ‘a’ is to ‘b’ what ‘b’ is to ‘a’”
[Robrieux, 1993, p. 117]. It can be found again in the categorical
imperative of Kantian ethics: do not do to others what you would not
want them to do to you… and its common variations: do not say of
others what you would not want them to say of you, do not demand of
others that of which you would not be capable, do not refuse others a
right which you would not want to be refused yourself. This argument
implies both the anthropological identity of human beings and the
normative principle that their places and roles can be reversed in the
social order, in accordance with what is promised by democracy. The
principle of shared belonging is also to be found in the inclusion
argument (what is valid for the whole is valid for a part): all members
of category x (minors under 16 years, civil servants, citizens, human
beings…) have such a right; now y is a member of category x, so y must
possess such a right. The “two weights, two measures” argument is the
denunciatory version of this same reasoning.
The opponents of legal equality tend to challenge the idea of
belonging to a “same essential category”. The incommensurability
argument maintains that category x (those demanding the right) and
category y (those holding the right) belong to two heterogeneous spe-
cies: French nationals and foreigners, women and men, same-sex and
opposite-sex partnerships, etc. cannot be contained within the same
category. The incommensurability assertion can lead to the paradoxi-
cal argument of reciprocal non-reversibility of places and roles. It is
the segregationist principle of everyone in their place, whether the
communities formed in this way are so by virtue of race, sex or sexu-
ality. According to the principle of formal reciprocity which disre-
gards the asymmetry of the situations involved, if men do not wish to
be “angels in the house”, then women should not ask to be “engineers”;
if French nationals are not going to emigrate to the southern hemi-
sphere where foreigners come from, then foreigners should not emi-
grate to France; if heterosexual couples do not hanker after the “privi-
leges” in terms of freedom, which homosexual couples are imagined
to possess, then homosexuals should not demand the same rights as

married heterosexuals11, and so on.
The question of justice is intertwined with the question of the rule
and the exception to the rule. If the refusal to grant a right to a given
category is based on the argument of the said category’s “natural dif-
ference”, then the partisans of equality are concerned to show that
particular members of the category do not possess this difference (this
woman is capable of pleading a case or conducting an orchestra, that
homosexual couple have been together for thirty years, this foreigner
is better integrated into society than that Frenchman, etc.): it is the
argument of the invalidating case which stresses that the stigma on
which the exclusion is based is in no way general, and therefore could
not form a basis in law. This argument works together with a demys-
tification of the opposing position as a discourse of interest: invoking
the “difference” of the group being discriminated against serves only
to preserve the privileges of the dominant group. What is more, this
invocation is denounced by the partisans of equality as a circular
fallacy, a reasoning whose premises and conclusions are identical:
those being dominated are forbidden access to such and such a social
position, and then, by their absence, they are deemed to be incapable
of occupying that position; their social inferiority is justified by their
natural inferiority and then their natural inferiority is deduced from the
fact that they are socially inferior. The aim of the opponents of equality
is, on the contrary, to preserve the exceptional status of invalidating
cases: erudite women, well integrated foreigners and stable homo-
sexual couples do exist, but their existence does not invalidate the
general group characteristics which justify its particular legal status.
11 On this basis, the spokespersons for groups being discriminated against are re-
proached with “having their cake and eating it” [“vouloir le beurre et l’argent du
beurre”]: feminists for wanting male privileges while preserving female privi-
leges, homosexual militants with demanding social integration on the hetero-
sexual model while holding on to the privileges of borderline sexuality; while the
spokespersons for resident foreigners are accused of claiming the economic and
social benefits of integration, while refusing to renounce their cultural differ-
ences, and so on.
106 Juliette Rennes
They too reveal shameful motives behind the egalitarian discourse:

the partisans of equality, who deny “differences”, are only rationalis-
ing their envy and resentment of those (men, members of the middle
class, whites, etc.) who have been more successful.
Finally, the question of justice is also linked to the question of use-
fulness. John Stuart Mill said that the partisans of equality must answer
to “those for whom it is not sufficient that inequality has no legitimate
justification but who want to be told what specific advantages would
be obtained by abolishing it” [Mill, 1992, p. 164]. The argument of
pacifying effects or economic gains which can be expected from a
mixed or multicultural society are opposed, for example, to the utili-
tarian conservative arguments already mentioned – those of the “slip-
pery slope” and the “perverse effect” which picture the catastrophes
that would result from “democratic fervour” [Rancière, 2005].
The picture of these clusters of contradictory arguments ex-
changed by the opposing parties in the equality controversy has been
no more than outlined here: it needs further research both to com-
plete it and to show the variations marked out within these topics by
the socio-historic configurations of each equality debate and the het-
erogeneous constellations of the mobilised actors. It would also be
beneficial, from an epistemological point of view, to question the
connections and possible effects of repetition between typologies of
strictly militant arguments and scientific methods of analysing non-
egalitarian discourses. The militant analysis of arguments opposing
equality in terms of defending privileges, unrecognised as such by
the actors themselves, is very close for example to the concepts of
“rationalisation” or “false consciousness”, which in psychoanalysis12
and the Marxist sociology of knowledge13, aim at revealing uncon-
scious interests behind theoretical constructions which appear disin-
terested. As for denouncing the conservative fallacies which consist
in deducing the social inferiority of groups being discriminated against
12 Jones, “Rationalisation in Everyday Life” quoted in Laplanche-Pontalis (1997:

387, 388).
13 For example Guterman and Lefebrvre, 1936.
from their “natural” inferiority and vice versa, this forms the basis of
the sociological position formulated by Durkheim: not “to present
social life as the simple result of the individual natures ” because it is
more the other way round, with “the latter resulting from the former”
[Durkheim, 1960 p. 341].
If we recognise that the cognitive categories used in social sciences
rest partly on normative bases shared by the “ordinary operations” of
social criticism [Boltanski and Thévenot, 1991], the analysis of equal-
ity controversies invites us to think jointly about the effects of social
intelligibility produced by militant discourses and the effects of denuncia-
tion to which scholarly discourses contribute. In this sense, the argu-
mentative approach to controversies does not simply provide the politi-
cal scientist with tools that can help him to carry out the study of conflict
processes and to propose an analysis that articulates the repertoires of
collective action with the repertoires of the arguments brought into
play. It can also contribute to methodological reflection for researchers
who find themselves dealing with politically committed discourses.
ANGENOT, M.
1989: 1889: Un état du discours social, Québec, Éditions du Pré-
ambule.
1995: La parole pamphlétaire, Paris, Payot. 1st edition 1982.
2000: Les grands récits militants des XIX e et XX e siècles. Religions
de l’humanité et sciences de l’histoire, Paris, L’Harmattan.
2003: “Contre le socialisme. Essai d’histoire discursive: 1830–1917”,
Discours social, vol. XVI.
ARISTOTLE
1990: Organon V. Les topiques, Paris, Vrin.
108 Juliette Rennes
AUTHIER-REVUZ, J.
1982: “Hétérogénéité montrée et hétérogénéité constitutive: éléments
pour une approche de l’autre dans le discours”, DRLAV, no. 26.
BARD, C.
1999: Ed. Un siècle d’antiféminisme, Paris, Fayard.
BARTHE Y.
2006: Le pouvoir d’indécision, la mise en politique des déchets
nucléaires, Paris, Economica.
BARTHES, R.
1994: “L’ancienne rhétorique. Aide mémoire” in Recherches rhéto-
riques, Paris, Point Seuil, 1st edition. Communications, 1970, no. 16.
BENTHAM, J.
1962: The Book of Fallacies in The Works of Jeremy Bentham [1816].
Published by J. Bowring, NY, Russell and Russell.
BOLTANSKI, L. and THEVENOT, L.

1991: De la justification. Les économies de la grandeur, Paris, Galli-
mard.
BORILLO, D.
2001: “Pluralisme conjugal ou hiérarchie des sexualités: la reconnais-
sance juridique des couples homosexuels dans l’Union Européenne”,
Revue de droit de McGill, 46.
CÉFAI, D.
1996: “La construction des problèmes publics. Définition de situa-
tion dans les arènes publiques”, Réseaux, no. 75.
CHARAUDEAU, P. and MAINGUENEAU, D., ed.

2002: Dictionnaire d’analyse du discours, Paris, Seuil.
COBB, R. W, ELDER, C. D.
1972: Participation in American Politics: the Dynamics of Agenda
Building, John Hopkins University Press.
DUPRIEZ, B.
1984: Gradus. Les procédés littéraires, Paris, 10/18, 1984.
DURKHEIM, E.
1960 [1893]: De la division du travail social, Paris, PUF.
EEMEREN, F. H., VAN and GROOTENDORST, R.
1992: Argumentation, Communication and Fallacies. A Pragma-Dia-
lectical Perspective, Hillsdale NJ, Lawrence Erlbaum.
FAIRCLOUGH, N.
2003: Analysing Discourse – Textual analysis for social research,
New York: Routledge.
FASSIN, E. and FABRE, C.
2004: Liberté, égalité, sexualités. Actualité politique des questions
sexuelles, Paris, 10/18.
FAVRE, P.
1992: “L’émergence des problèmes dans le champ politique”, in Favre
(ed.), Sida et politique, Paris, L’Harmattan.
GUTERMAN, N. and LEFEBVRE, H.
1936: La conscience mystifiée, Paris, Gallimard.
HIRSCHMAN, A. O.
1991: Deux siècles de rhétorique réactionnaire [The Rhetoric of Re-
action: Perversity, Futility, Jeopardy], Paris, Fayard.
LAPLANCHE, A. and PONTALIS, J.-B.

1997: Vocabulaire de la psychanalyse, Paris, PUF. 1st edition 1967.
110 Juliette Rennes
MAINGUENEAU, D.
1983: Sémantique de la polémique, Lausanne, l’Age d’Homme.
1995: “L’énonciation philosophique comme institution discursive”,
Langages, no. 119.
2002: “Interdiscours”, in P. Charaudeau and D. Maingueneau (ed.).
Dictionnaire d’analyse du discourse, Paris, Seuil.
MILL, J. STUART
1992: De l’assujettissement des femmes [The Subjection of Women,
1867], Paris, Ed. Avatar.
NORDMANN, CH., ed.
2004: Le Foulard Islamique en question, Paris, Éd. Amersdam.
PÊCHEUX, M.
1975: Les vérités de La Palice. Linguistique, sémantique, philosophie,
Paris, Maspero.
PERELMAN, C. and OLBRECHTS -TYTECA, L.
1992: Traité de l’argumentation. La nouvelle rhétorique, Brussels,
Presses de l’Université de Bruxelles, 1st edition 1958.
PLANTIN, C.
1991: Essais sur l’argumentation, Paris, Kimé.
2005: L’argumentation, Paris, PUF, Que Sais-je?
RANCIÈRE, J.
2005: La haine de la démocratie, Paris, La Fabrique.
RENNES, J.
2007: Le mérite et la nature. Une controverse républicaine: l’accès
des femmes aux professions de prestige, Paris, Fayard.
REISIGL, M. and WODAK, R.
2001: Discourse and Discrimination: Rhetorics of Racism and Anti-
semitism, London/New York: Routledge.
ROBRIEUX, J.-J.
1993: Éléments de rhétorique et d’argumentation, Paris, Dunod.
SARAT, A., SCHEINGOLD, S.

2001: Cause Lawyering and the State in a Global Era, NY, Oxford
University Press.
SCHOPENHAUER, A.
2004 [1864]: L’art d’avoir toujours raison, Belval, Circé Poche.
SCOTT, J.
1996: Only Paradoxes to Offer. French feminists and the Rights of
Man. Harvard University Press.
TAGUIEFF, P.-A.
1985: “Le néoracisme différentialiste. Sur l’ambiguïté d’une évidence
commune et ses effets pervers: l’éloge de la différence”, Langage et
société, no. 34.
1990: La force du préjugé. Essai sur le racisme et ses doubles, Paris,
Gallimard, 1st edition 1988.
TODOROV, S.
1981: Mikhaïl Bakhtine. Le principe dialogique.
WALTON, D.
1992: Slippery Slope Argument, Oxford, Clarendon Press.
WODAK, R. and VAN DIJK T. (eds).

2000: Racism at the Top: Parliamentary Discourses on Ethnic Issues
in Six European States, Drava Verlag, Klagenfurt, 2000.
Chapter VII
Discourse analysis and the study of literature1
Dominique MAINGUENEAU
The development since the 1990s of “literary discourse analysis” im-

plies a profound transformation in the very conditions under which
literature can be studied. It would therefore be reductionist to see in
this branch of discourse analysis problems a “means of elucidation”
amongst others; rather, it is about progressively putting in place a
method of understanding the literary fact (and not merely literary
works) which does not allow itself to be enclosed within traditional
divisions and disciplines.
Approaches in terms of literary discourse analysis are based on
the notion of discourse. This notion is very difficult to deal with,
however. On the one hand, it possesses certain linguistic values, but
on the other it is liable to be used without very much control, as a key
word for a certain conception of language. For example, when speak-
ing of literary discourse, a certain number of major ideas are brought
together, which alter our approach to literature. We might indicate a
few of these here: firstly, discourse is a form of action, radically
contextualised, governed by norms and dominated by an inter-
discourse; texts are the traces of language activities and discourse
genres, which are inseparable from non-verbal activities.
The idea that words are an activity might seem banal, but it
modifies the tacit models governing the approach to texts which have
prevailed for centuries: literary discourse is one particular activity,
1 These few pages are extremely allusive. For a more detailed presentation, see
Amossy and Maingueneau (ed.) (2003), Maingueneau (2004), Poétique no. 140
(2005).
114 Dominique Maingueneau
but it is also an activity among others, participating in the world it

is supposed to “reflect”. This activity, which is fundamentally inter-
active, means that the positions of producing and receiving cannot
be divorced from one another; works of literature exist not because
literature might be exterior to any interaction, but because it man-
ages the fundamental interactivity of discourse in a specific way, by
very diverse means: courtly poetry is based on the conversation of
high society, while romantic poetry challenges this. Furthermore, in
common with any language activity, literary enunciation cannot es-
cape the sphere of the Law. Words and the right to speak are inter-
locked: where can words legitimately come from, who do they claim
to address, according to what means, what time, what place?
More broadly speaking, to consider the literary fact as “discourse”
is to contest the central character of this fixed point, the creator, this
source “which has no communication with the exterior”, to quote a
celebrated phrase from Proust’s Contre Sainte-Beuve, describing the
creator of a literary work. So we are a long way from the aesthetic
world opened up by Romanticism where individual creativity was,
directly or indirectly, at the heart of literary studies: directly, when
the life of an author was put under the microscope; and indirectly,
when the “context” of the work was studied or the text was read as an
expression of the author’s “world view”. But for literary enunciation
to take place, it is not sufficient to link a creative soul with a recep-
tive one: the very status of these “souls” varies with the historically
defined word institutions which render them possible. The spectator
watching the performance of a trouvère reciting a chanson de geste is
not the 18th century reader of a pamphlet by Voltaire, or of a novel by
Zola a century later.
When we speak today of “literary discourse”, we are thus loath to
determine a centre or, at least, if there is a centre it is in a very different
sense, since it is the system of communication itself. We seek to restore
works to the spaces which make them possible, where they are pro-
duced, evaluated and managed. This is where the conditions of the dire
pierce through the dit, and the dit reflects its own conditions of enun-
ciation (the writer’s status combined with the way he or she is posi-
Discourse analysis and the study of literature 115
tioned in the literary field, the roles attached to genres, the relationship
with the reader which is built up throughout the work, the medium of
distribution, the way utterances are circulated and so on). This is true
not only for the conditions in which the work is created, but also for
the processes of transformation, of re-evaluation over time. The “same”
text will not be the same if, having been judged to be subversive, it is
published anonymously by an Amsterdam publisher or as a series of
extracts with comments in a “small classic” edition designed for school-
children.
Far from forbidding access to what is essential in the work con-
cerned, taking account of literary communication, the apparatus of
enunciation, appears to be the condition, the driving force and what is
at stake in enunciation. Certainly, part of the claim of literature is to
offer works capable of transcending the context in which they were
produced; but the exterior nature of the context is revealed to be de-
ceptive. A work cannot be conceived of as an arrangement of “con-
tents” which allow ideologies or mentalities to be “expressed” in a
more or less roundabout way. In reality, its conditions of enunciation
are reflected right the way through the “content” of a work of litera-
ture. The context is not exterior to the work, in a series of successive
envelopes, but the text itself governs the context. There is not, on the
one hand a world of mute objects and activities and, on the other,
representations which are detached from it, like a more or less blurred
“image” of it. Work on 17th century “courtly” literature, for example
(see in particular the work of [Viala, 1985, 1999] and [Denis, 1987])
foregrounds the rites of a community of users (producers, readers and
publishers): their texts (such as novels, madrigals or portraits) are both
the condition and the product of these rites, in a space (the salon) which
is both inside and outside society, but at the same time its place of
circulation and its single theme. The gallants name themselves after
the shepherds of pastorals. Similarly, it is impossible to analyse 19th
century works without seeing how they reflect the literary institution
of which they are part: Quasimodo and Esmeralda in Notre-Dame de
Paris are characters in a history and at the same time inscribe the trace
of the writer who bears this very story.
Writers produce works, but writers and works are themselves

produced by a whole institutional complex of practices. One way
of acknowledging these shifts is to reason in terms of discursive in-
stitution, by inextricably combining the institution as the act of pro-
ducing, by means of a legitimate construction process, and the in-
stitution in its usual sense of an organisation of practices and
apparatus. The relationship between “institution” and “discursive”
implies a reciprocal enveloping; discourse occurs only if it is mani-
fested through these speech institutions – discourse genres – which
are conceived of through metaphors of ritual, contract and mise en
scène or staging. For its part, the literary institution itself is con-
tinually being reconfigured by the genres of discourse that it makes
possible. This concept of discursive institution thus articulates a) the
institutions, the frameworks of various orders which give a sense to
the particular enunciation: the structure of the field, the writer’s sta-
tus, the text genres, and so on; and b) the movement by which dis-
course is framed, both by progressively establishing a particular
world in its utterance and by legitimising the scene of the utterance
and its positioning in the field which makes that utterance possible.
If the work, through the world it configures in its text, reflects the
conditions of its own enunciative activity by legitimising them, we
can understand the crucial role played by the “scene of enunciation”
presupposed by all texts. This scene cannot be reduced either to the
text, or to a communication situation which could be described from
the outside, as a historian or sociologist would do. In the traditional
epic, for example, the scene of enunciation is that of a mediating
narrator. As witness of the events he is recounting, he addresses a
community whose values he shares and which belongs to the same
imaginary community as the heroes of the story; he is not recounting
a personal fiction but a story that is already known, part of a shared
repository or thesaurus without an author, which is updated and modi-
fied with each enunciation. This is a long way from the enunciation
of Zola’s Naturalist novels, with their didactic style, in which a man
of science reveals an unknown “milieu” to a faceless reader.
The research on the sociology of literature inspired by Bourdieu2

has the great merit of stressing that the production of a work of litera-
ture must not first and foremost be related to society considered as a
whole, but to a limited sector of that society: in the 19th century, this
took the form of a relatively unified “field” obeying specific rules.
More precisely, it could be said that all works include the three levels
of what might be called literary space.
– This space is a network of apparatus in which individuals can set
themselves up as writer and as readers; within these, the generic
contracts considered as literary, are stabilised and guaranteed,
with mediators (publishers, bookshop owners), legitimate inter-
preters and assessors (critics, teachers) and canons (which might
take the form of textbooks, anthologies, etc.) all playing their
part.
– It is also a field, a place of confrontation between aesthetic posi-
tions which occupy it according to their own modes of genre and
idiom. A discursive field is not a static structure, but an unstable
system. Neither is it homogeneous: there are positions which are
both dominating and dominated, central positions and peripheral
ones. But the idea of a truly independent artistic field is a recent
phenomenon, which is probably losing some of its power thanks
to new technologies; it cannot be extended to the historical and
geographical diversity of “literary” production regimes. For ex-
ample, there are regimes which do not oppose schools of thought
with manifestos and doctrines but through the places where they
are exercised (various princely courts, regions and so on): in this
case, the different competing groups share most of the aesthetic
presuppositions and occupy the same genres, but differ by their
“styles”, themes, etc.
2 The most conclusive explanation for this can be found in Bourdieu (1992), who
uses the examples of 19th century writers. For the 17th century, reference can be
made to Alain Viala’s book (1985). For an approach which is more concerned
with the poetic dimension, see Meizoz (2003, 2004). For a didactic presentation
of the set of problems in which Bourdieu is engaged, see Thumerel (2002).
– Finally, this space is an archive where intertext and legends com-

bine: creative activity is immersed in a memory which, in return,
is itself caught up in the conflicts of the field, and these are con-
stantly reworking it. The notion of archive refers here to litera-
ture’s internal memory, a memory which, beyond intertext in the
strict sense of memory of texts, also includes “legends”, in par-
ticular those attached to the lives of famous writers. Each posi-
tion implies taking an original route through this archive. For
example, the 16th century Pléiade rejected medieval genres and
placed value on certain genres from Greek Antiquity, whereas
Romanticism marginalised neo-Classical poetry and valued popu-
lar and medieval genres.
More generally, a notion like “literary discourse analysis” is heavy
with presuppositions.
– In literary discourse analysis, literature does not have the benefit
of being extra-territorial; in other words, discourse analysis is
not restricted to texts considered as “ordinary”. It is not about
confronting the “profane” social sciences with the “sacredness”
of literature, as is commonly done in literary studies, but explor-
ing the multiple dimensions of discursivity without at the outset
assuming that the multiple manifestations of words are incapa-
ble of comparison. So a different process is followed from the
one which involves bringing into play notions borrowed from
psychoanalysis, sociology, anthropology, etc. to “elucidate” lit-
erary texts: it is no longer about projecting one world (the social
sciences) onto another (literature) which is thought to be alien to
it. It is about exploring the world of discourse in all its diversity.
– The language sciences play a part in two ways, which no longer
accord solely with the traditional ancillary relationship of gram-
mar/linguistics with regard to literature. The joint development
of text linguistics and discourse linguistics, inspired by prag-
matic trends and enunciation theories, has considerably enriched
thinking about literary utterances. The advances made as regards
discourse genres, enunciative polyphony, markers of oral inter-
action, argumentative processes, laws of discourse, tropes, ana-

phora, etc. were quickly incorporated into the study of literary
discourse. Nowadays, having recourse to linguistics no longer
means simply having recourse to elementary tools (as in tradi-
tional stylistics); nor to a few very general principles of organi-
sation (as in structuralism). It is now a proper investigative in-
strument. In situations where people once made do with notions
of common descriptive grammar to validate conclusions which
could have been based on intuition, nowadays they can develop
interpretations which intuition alone is incapable of identifying.
By way of example, there is the work of J.-M. Adam [1990, 2005]
on textual organisation, or Alain Rabatel [1998] on point of view.
There is also the work of C. Kerbrat-Orecchioni on theatre [1984,
1996], that of S. Durrer on dialogue in the 19th century novel
[1994] and numerous works on the demonstrative determinant
in literary texts3. “Traditional” stylisticians would use linguis-
tics as a sort of tool box enabling them to shore up their intuitions
as readers; linguistics as such did not have a truly heuristic role
and did not play a part in interpretation. This was just as true for
“academic” linguistics, with its methods of close reading, as it
was for the much more ambitious stylistics of someone like Léo
Spitzer [1970]. With the development of disciplines taking dis-
course itself as their object of study, what was thought to be a
simple aid plays a part in the actual construction of research
protocols and interpretations.
– It is possible to make previously marginalised or ignored corpuses
(which in fact means most of them) readable, and gain access to
figures of auctoriality which do not belong to the conception of
style that is still dominant: to take France alone, courtly literature,
Petrarchan poetry, vaudeville, the comédie-ballet, even classical
tragedy, can only be read if we break free from presuppositions
inherited from the 19th century. To say that particular works have
no great literary interest because there are “too many clichés”,
3 See in particular number 120 (1998) of the journal Langue française.

because it is “official” poetry or “salon literature”, or because

they “lack originality” shows merely that we do not have a model
which is adequate to understand them. The same obviously holds
for what is known as “para-literature”, the type of production
which is largely ignored by faculties of letters.
– Over and above this, it is the very independence of works of
literature which raises problems. To say that literature is a “dis-
course” is, as we have seen, to put interdiscourse first. This means
that works only exist in association with texts of distinct status
and genre: classical literature is inseparable from colleges, sa-
lons and academies all at the same time; 19th century literature
cannot be dissociated from newspapers with a large circulation,
literary magazines, textbooks of national literature and so on.
Literary creation occurs only within specific configurations of
production, circulation and commentary.
The development of such a space for research is not without its insti-
tutional counterpart. We now see a research force emerging from dif-
ferent disciplines, sharing a common intellectual toolkit that is much
more substantial than the one they shared with colleagues from their
“original” discipline. For example, there are discourse analysis re-
search centres where some members work on literary as well as non-
literary texts, by using the same conceptual and methodological tools,
but each time adapted to the corpora with which they are confronted.
In France, in addition to Céditec, there are the universities of Metz
(Celted), Besançon (centred around the journal Semen), and
Montpellier III (Praxiling). Inevitably, the methods of approach and
thinking of a traditional Mallarmé specialist and a literary discourse
analyst will be further removed from each other than those of this
same discourse analyst and a religious, or even journalistic, discourse
analyst.
We are thus forced to reflect on the status of literary research:
does it really define a discipline? How do we situate discourse analy-
sis in relation to traditional divisions?
In some universities there are departments of theology, as dis-
tinct from the history of religion or religious anthropology. The two
types of approach can be distinguished by a central principle: theol-

ogy is the concern of believers and is aimed at believers, which is not
the case with a science of religion, which comes within the scope of
the human and social sciences. Following the example of what hap-
pens in departments of literature, teaching in theology should not be
reduced to a personal commentary on authorised texts. It claims to be
based on archaeology, codicology, history and so on, but inside a
system which assigns a minor role to these; actual religious truth is
essentially of a different order. In literature too, one is enriched by
knowledge which might be called positive; however, that knowledge
is placed in a line of thought where it serves a higher imperative, that
gives it meaning – that of a living, personal contact with a Source:
“Will we allow the social sciences to reduce literary experience, the
highest experience that humans can attain along with that of love, to
surveys about leisure, when it concerns the meaning of life?”4. The
human and social sciences should “elucidate” and nothing more. Fur-
thermore, in one domain as in the other, there is no question of com-
menting upon just any text, but only upon real oeuvres, which are the
heritage of the literary scholar or the theologian, and for which they
are responsible.
There is nevertheless a difference between departments of theol-
ogy and departments of “literature” in that theology is obliged to
assume its fundamentally hermeneutic status, and to strongly affirm
the supremacy of “the sense of spirituality”. Theologians cannot there-
fore call themselves “researchers”, otherwise they are no longer act-
ing as theologians but as philologists, sociologists, etc. Whereas the
institutional position of literary specialists means that they have to
call themselves “researchers”, and not proclaim too openly how their
approach cannot be reduced to that of the human and social sciences.
This equivocal situation was able to continue for as long as lit-
erature was not understood as discourse. Indeed, as long as the “con-
4 D. Sallenave, Le Don des morts (Paris, Gallimard, 1991). This sentence was
inserted as an epigraph by Bourdieu (1992: 9) as representative of an ideology
whose presuppositions he violently contested.
text” was governed by historical types of approach (“literary history”)

which were intended simply as aids to interpreting texts, it was easy
to refer the human and social sciences to the care of minor authors
and to peripheral aspects of literary activity. But once we depart from
this pattern, to work in discourse analysis, the hierarchy between the
hermeneutic relationship and the highlighting of “context” by sup-
plementary knowledge can no longer be made to work effectively.
“Value” and “beauty”, like admiration, personal commentary and the
institutions that all these suppose are, for discourse analysts, phe-
nomena which have to be studied.
Bourdieu has always denounced the “illusory” and self-seeking
relationship which, according to him, literary specialists maintain with
works of literature5. But by wanting to dispel this hermeneutic rela-
tionship, we challenge the very status of literature which, as a “self-
constituting discourse” [Maingueneau, Cossutta, 1995] maintains an
essential link with the basic values of the community. If we were to
criticise certain traditional literary scholars, it would be for wanting
the privileges of the man of science and the hermeneutist both at the
same time, and yet not for wanting to be part of a hermeneutic frame-
work; and for postulating a “literary experience” as a condition of
their approach to literary texts.
The fundamental question is whether the study of phenomena
which nowadays is commonly grouped under the heading of “litera-
ture” should be exclusively literary, condemning any other approach
as “reductionist”. The split between methods of knowing which are
irremediably “exterior” to the work, and real criticism which would
recapture creative awareness, in fact indirectly supports actual insti-
tutional partitioning: this would separate the “true” literary scholars
who, having access to the creative self, would act in an imperious way,
from the adherents of the human sciences who, by their very approach,
5 See in particular the preface to Règles de l’art, where he denounces the “hack-
neyed themes of the academic cult of the Book or Heidegger and Hölderlin-style
revelations worthy of embellishing an ‘anthology of Bouvard and Pécuchet’”
(1992: 9).
would be dedicated to peripheral tasks. This is a division which liter-

ary discourse analysis can only challenge.
All one can reasonably hope for is that traditional literary scholars
stop appropriating the literary activity, just as theologians have had to
stop appropriating the religious activity. However, this restoration of
balance cannot be taken for granted. Accepting a role for the human
and social sciences which is not an ancillary one means in effect chang-
ing some deeply embedded views, together with the institutional re-
alities that they imply.
ADAM J.-M.
1990: Eléments de linguistique textuelle, Brussels, Liège, Mardaga.
2005: La linguistique textuelle. Introduction à l’analyse textuelle des
discours, Paris, Armand Colin.
AMOSSY R. (ed.)
1999: Images de soi dans le discours. La construction de l’ethos,
Lausanne, Delachaux et Niestlé.
AMOSSY R. and MAINGUENEAU D. (ed.)

2003: L’Analyse du discours dans les études littéraires, Toulouse,
P.U.M.
BOURDIEU P.
1992: Les règles de l’art, Paris, Seuil.
DENIS D.
1987: La Muse galante. Poétique de la conversation dans l’œuvre de
Madeleine de Scudéry, Paris, H. Champion.
DURRER S.
1994: Le Dialogue romanesque, Geneva, Droz.
LANGUE FRANÇAISE
1998: “Les démonstratifs” (Gary-Prieur M.-N. and Léonard A.-M
(ed.), Larousse, no. 120.
KERBRAT-ORECCHIONNI C.
1984: “Pour une approche pragmatique du dialogue de théâtre”,
Pratiques no. 41, 46–62.
1996: “Dialogue théâtral vs conversation ordinaire”, Cahiers de
praxématique, 26, 31–49.
MAINGUENEAU D.
2004: Le discours littéraire. Paratopie et scène d’énonciation, Paris,
A. Colin.
MAINGUENEAU D. and COSSUTTA F.
“L’Analyse des discours constituants”, Langages no. 117, 1995, 112–
125.
POÉTIQUE
2005: “Analyse du discours et sociocritique”, Amossy R. (ed.), La-
rousse, no. 140.
SPITZER L.
1970: Etudes de Style, French translation, Paris, Gallimard.
MEIZOZ J.
2003: Le Gueux philosophe (Jean-Jacques Rousseau), Lausanne, Anti-
podes.
2004: L’œil sociologique et la littérature, Geneva-Paris, Slatkine érudi-
tion.
RABATEL A.
1998: La construction textuelle du point de vue, Paris, Delachaux et
Niestlé.
THUMEREL F.
2002: Le champ littéraire français au XX e siècle. Eléments pour une
sociologie de la littérature, Paris, Armand Colin.
VIALA A.
1985: La naissance de l’écrivain. Sociologie de la littérature à l’âge
classique, Paris, Minuit.
1999: “L’éloquence galante, une problématique?”, in R. Amossy (ed.),
Images de soi dans le discours. La construction de l’ethos, Lausanne,
Delachaux et Niestlé, 172–191.
Chapter VIII
Discourse analysis and history.

Meeting and forgetting
Marie-Anne PAVEAU
Like everyone, I have had less occasion to read books on linguistics,

much less, if the truth were told, than I have detective novels.
P. Vidal-Naquet, Histoire et linguistique, introductory statement
The relationship between discourse analysis and history has been one
in which each has progressively become more distant from the other:
the situation in France between the 1960s and the 1980s, when there
was a close connection between them, subsequently gave way to mu-
tual indifference, with the exception of a research group claiming to
be discourse historians.
This state of affairs can be explained by several factors, and the
examination of these will serve as a framework for what follows.
First of all, there is the question of disciplinary identity, since history
as seen by historians is not necessarily the same as that envisaged by
linguists1. Then there is the status, place and scientific purpose pos-
sessed by each methodology in the associated discipline. Finally, there
is the evolution of the two disciplines themselves, each of them hav-
ing taken routes which distance it from the other.
1 I am using the term linguistic almost synonymously with discourse analysis,

since it is essentially “French” discourse analysis, based on sometimes very for-
mal linguistic methods, which has maintained a relationship with history.
128 Marie-Anne Paveau
History for linguists and history for historians
I would like to question the nature of the history evoked by the two
disciplines, since it is far from certain that the history of linguists
tallies exactly with the history of historians.
What is known as “French” discourse analysis appeared in France
in the late 1960s and was based on the analysis of historical corpora,
that is to say sets of texts in which were concealed matters relating to
French history: Cahiers de doléances (D. Slakta); texts of the Congrès
de Tours (J.-B. Marcellesi); Hébert’s Père Duchesne (J. Guilhaumou);
and the media vocabulary surrounding the Algerian war (D. Maldi-
dier)2. From this point of view, discourse analysis is seen as an inte-
gral part of historical practice, since the linguistic perspective should
allow the opacity of words, the ambiguity of enunciation and the
constructed nature of discourse to be explored, thus enabling histori-
cal archives to be interpreted effectively. Archives is an historian’s
word, whereas the linguist is more involved with constructing cor-
pora. And it is probably over this question of the object of the disci-
pline that the two “histories”, that of the linguist and that of the his-
torian, diverge.
Indeed, linguists are permanently reflecting on their object of
study, which has led them to criticise this early pattern of relations
between discourse analysis and history. Analyses were based on choos-
ing pivotal words (words chosen for studying the contexts in which
they appeared) and identifying discursive groups which corresponded
in an overly mechanical fashion to social class and political move-
ments. Hence, in this pattern, there is a “triple pitfall” for “discourse
analysis as historical object” according to J. Guilhaumou:
In the first place, this pattern introduced a clear split between the chosen corpus,
which was really very limited in terms of the analysis procedure, and the non-
2 The precise references of these works are available in the bibliographies of Robin
(1973) and Charaudeau, Maingueneau (2002).
Discourse analysis and history 129
corpus defined in a referential and ideological way by the notion of conditions of

production. Secondly, the choice of pivotal words was based on a judgment made
in light of the historian’s knowledge, and was itself made in the field of
historiographical debate rather than in terms of the current moment. Finally, the
pattern represented separate discursive entities based on ideology and
historiography – entities such as noble discourse, bourgeois discourse, Jacobin
discourse, sans-culotte discourse and so on [Guilhaumou, 2002: on line].
The examples mentioned above correspond to this early pattern since

the “chosen” corpora are de facto isolated from their original milieu,
in other words from their “conditions of production”; we might also
mention G. Provost’s 1969 study of the words socialisme and socialiste
in Jaurès, which well illustrates the “pitfall” of pivotal words, and
the work of J.-J. Courtine on “communist discourse addressed to Chris-
tians” [1981] which constitutes an example of “separate discursive
entity”.
The “judgment in light of the historian’s knowledge” is in fact
where the problem lies in the relationship between discourse analysis
and history, since the conception of the field of investigation (ar-
chives or corpus) is not the same in both disciplines. Historians as-
semble their documents from actual available archives, researched
thanks to their knowledge of the event, period or cultural/social
phenomenon under consideration; for the linguist, on the other hand,
the construction of the corpus is part of a scientific practice, as
J. Guilhaumou explains:
However, it is no longer a question of constructing a corpus immediately, on the

basis of a judgment made in the light of knowledge, as a preliminary description
of conditions of production. It is advisable to first describe configurations of
archives indicative of a theme, a subject, a concept and, when all’s said and done,
of an event. There is then always time to isolate a “moment of corpus” within
them, that is to say a group of utterances based on lexical, syntactic or enunciative
criteria and thus build up a sub-corpus which is open to a subtle linguistic ap-
proach [Guilhaumou 2002: on line].
The study conducted by J. Guilhaumou and D. Maldidier of the phrase

prise de la Bastille [storming of the Bastille] [1994] allows us to gain
a clear understanding of this process (F. Mazière, who analyses it at
length, considers it to be “emblematic” of this second period of dis-

course analysis (2005: 89)). Here, the set of discourses within which
the syntagm prise de la Bastille was to be formed, through repeti-
tions, reformulations and modifications, constitutes this “configura-
tion of significant archives” in which the linguist builds up a “sub-
corpus”. As P. Achard points out, we can expect to gain nothing from
interdisciplinarity if it is thought of as “falling between two stools, a
mish-mash or a merging” (1983: 28). On the other hand, if “each
discipline, at the very point of constructing its object of study, makes
room for the object of the other(s), and not merely for its discourse”
(ibid.), then history and discourse analysis can develop shared and
effective research strategies. This essentially means that linguistics
would fully embrace the historical dimension (in other words, ac-
cording to the linguistic viewpoint, extra-linguistic data which no
longer relate only to the historian’s expertise), and history in turn
would embrace the linguistic dimension of historical events, which
are then no longer limited to their factual or even cultural aspects.
As a final comment on this subject, we should mention an ideo-
logical element which is important for the relationship between dis-
course analysis and history: the initial joining of the two disciplines,
in the 1960s, took place in the context of historical materialism. In
other words, history for linguists was defined in the context of Marx-
ist philosophy, as a scientific study of the relationships of power and
domination (contradictions) in actual social formations. With the
gradual disappearance of Marxism and Marxists from the French
academic scene, history was probably taken less into account in lin-
guistics3.
3 “[…] to understand the disappearance of certain ways of working in French lin-

guistics, one needs to analyse the political and intellectual conditions of the
demarxisation of research in linguistics and, more generally, in the human sci-
ences.”
History discourse, an epistemological issue
What use is history to linguists and linguistics to historians? A con-

tributory discipline always has a function in the life of the source
discipline, and the longevity, effectiveness and harmony of the asso-
ciation depend partly on the nature of this function.
Discourse as an object of history
The explicit wish of the linguists centred around the historian R. Robin,
who are developing links between history and linguistics, is to open
up a new field of research in the discipline of history:
We are seeking to establish the discursive level as a new object of study in the
field of history; and at the same time, in contrast to a certain positivism in lin-
guistics, to postulate the impasses and deficiencies of an internal analysis from
which the interpretation of discourse is removed, including its function, effec-
tiveness and processes of insertion into social formation [Robin, 1973: 22].
The purpose of the relationship cannot be better expressed: to open

up and revive the field of linguistics, based on the desire for a differ-
ent episteme from the Saussurian doctrine of the primacy of language.
J-J. Courtine speaks of “theoretical urgency, epistemological drama-
tisation” [1991: 157] in relation to this. “Wanting to analyse dis-
course”, he says,
[…] was therefore wanting to do far more than simply the work of a linguist. It
was also, in a way, to think of occupying a heroic position in a theoretical and
political struggle – to reintegrate through a liberating act what had been excluded
by an arbitrary decision [1991: 157].
This means promoting the word (parole) and the conditions of its
production.
Without this being the sole motivation of researchers wishing to
introduce a historical dimension into linguistics, it can be said to be
all-encompassing, serving as an epistemological and theoretical back-

drop to the proposals being made.
Indeed, the work that was carried out in the political lexicometry
laboratory at Saint-Cloud from 1968 under the influence of M. Tournier
was driven by the same preoccupation. Basing his thinking on the premise
that words are charged with the values conferred by usage (words are
“inhabited”, according to M. Bakhtine, whose work was translated
into French from 1977 onwards), M. Tournier progressively developed
a conception of words as instruments of ideology: ‘ideology’ because
“words do not have a history in themselves” but are coded by history
“for the needs of its dominances and in order to preserve them”
[Tournier, 1992: 9]; ‘instruments’ because “social dominances are,
more than any rival power, the owners of the reference code, know-
ledge which is imposed, dictated, and subject to the grammars of
acceptability, dictionaries of definitions and officially recognised values
[…]” [1992: 9]. From this perspective, the past is always falsified by
power and M. Tournier speaks of the “consensus about history”:
The characteristic of officially-diffused language is to be based on “a” past which

is taken for “the” past and set up as a model; its game [this time] is to have its
lexicalised affirmations taken for a common inheritance, which is classless, na-
tional, humanist even: an illusion of diachrony [Tournier, 1992: 11].
The historical (in lexicometric terms, chronological) dimension is cen-

tral to some of the work emerging from this laboratory, such S. Bonnafous’
thesis on the denomination of immigrants in the French press at the
turn of the 1980s. Partly based on the notion of “lexical time” put
forward by A. Salem, the investigation reveals the way designations
have evolved (from the “lexicon of work and social conflict” between
1974–1977 to a lexicon “of the political, the municipal, and integration”
from 1980–81)4. What is stressed here is evolution and lexical differ-
ence, and the expression witnesses of history, coined by M. Tournier
in 1997 to describe the role of words, is representative of the lexicologist-
historian’s position (or that of the historian of the lexicon).
4 Bonnafous, 1991: 226.

The implications of the “linguistic turn”
For historians, the appeal to linguistics takes two forms which corre-
spond to two phases in the history of the discipline.
The first political lexicology conference held at Saint-Cloud in
1968 welcomed historians, like A. Prost [1998], who practised lexi-
cal analysis in order to enrich their activity as historians; also be-
cause, at the time, the linguistic paradigm served as a model for all
the human sciences. However, R. Robin explains that attentiveness
to words is a longstanding tradition in history. She mentions the work
of G. Duby and R. Mandrou (1973: 66–67), but more particularly
A. Dupront’s “passionate quest” to found a new discipline of histori-
cal semantics (1973: 68). She observes, however, that for a number
of historians, from the early 1970s, linguistics was a “recipe” used
more for organising raw data than engaging in a real process of dis-
covery. So discourse analysis was quickly overtaken for historians,
in line with an “anti-positivist”5 development of the discipline, a point
to which we will return.
Another type of appeal to linguistics, identified by the slogan
linguistic turn, emerged in the American field after 1979, the date
when L. Stone’s article, “The Revival of Narrative: Reflections on an
Old new History” was published in the journal Past and Present. As
G. Noiriel explains, historians who adopt the linguistic turn point of
view (Q. Skinner, G. S. Jones and H. White, among others), are op-
posed to history as social science, inspired by the arguments of Barthes
and H. G. Gadamer. The trend is not consistent, but it is possible to
state, along with G. Noiriel, that “as its name would seem to indicate,
the ‘linguistic turn’ encompasses all historical work which accords a
certain importance to the question of language” [1996: 126]. It par-
ticularly includes work which puts narrative (the famous “retour du
récit” analysed in detail in [Noiriel, 2003]) at the heart of historical
5 The term positivism is often used in human sciences to criticise research proce-
dures that are blamed for copying the quantitative, even mathematical methods,
of the so-called “hard” sciences, in order to establish objectively true, or “positive”
knowledge.
practice, from the most subtle positions (narrative as a privileged

genre in the writing of history) to the most radical claims: historical
narrative is in effect considered to be fictional narrative, and this
position finds unexpected but epistemologically interesting reinforce-
ment… from a linguist, defending the idea, within a robust anti-
referentialism, that language cannot speak about the world [Harris,
2004]. In any event, the linguistic-narrative claim serves as a weapon
for an anti-positivism, an anti-quantitativism, in short a contesting of
the “presuppositions of science-history” [Noiriel, 2003: 104].
It was probably the eruption of this debate (violent, with impor-
tant political and generational dimensions) which contributed to his-
tory progressively forgetting the apparatus of early discourse analy-
sis, which doubtless took away with it all linguistic procedures,
considered to be too formal and positive. We have here the paradox
of a linguistic turn which throws out the tools of linguistics… With
the exception of “discourse historians”, who are referred to at the end
of this article, “traditional” historians now attach little importance to
the forms of language contained in their archives.
But we shall see that discourse analysis, for its part, has devel-
oped something that resembles its own forgetting, by bringing about
what J.-J. Courtine calls its “grammaticalisation”.
Two views of scientificity
Discourse analysis: rediscovering history
J.-J. Courtine thought in 1991 that discourse analysis had become

“grammaticalised”, that is to say divorced from history, and weak-
ened theoretically. Indeed, notions like interdiscourse or discursive
formation saw themselves reduced and mechanised, with the first
more or less assimilated to intertext and the second to a homogene-
ous set of discourses emanating from speakers of the same group
(regarding these questions see, for example (Maldidier, 1991), intro-

ductory text). At the same time, linguistic modelling asserted itself,
giving priority to the functioning of language and communication,
no doubt to the detriment of the socio-historic dimension of discourse
(the enunciative model of J. Authier, E. Roulet’s pragmatic model,
J.-P. Bronckart’s interactionist model, and so on).
Corpora have evolved, gradually eclipsing the historical dimen-
sion: media discourse, favoured by the development of forms of com-
munication and information (and, at the same time, of the academic
field of information and communication sciences) entered en masse
into what should now be called, in the plural, discourse analyses.
With the exception of A. Krieg’s work on the phrase purification
ethnique (see her contribution to this book), it is rare to find work on
the press which fully incorporates the historic dimension into lin-
guistic analyses. Under the influence of communication ethnogra-
phy, ethnomethodology and DAI (a term suggested by C. Kerbrat-
Orecchioni to describe discourse analysis in interaction), other corpora
have emerged, which might be called “community” or “everyday re-
lational” corpora (family or business interactions, the discourse of
assemblies or meetings, verbal negotiations of all kinds).
But the historical dimension is perhaps not completely absent
from contemporary discourse analyses. Indeed, it would appear that a
shift has occurred from the notion of history to that of memory (a shift,
moreover, which is a cause of much debate for historians). This sprang
from J.-J. Courtine’s proposal in 1981, in which he “invented” the no-
tion of discursive memory, reformulating it as follows in a later work:
How do societies remember? If we accept the idea […] that language is the fab-
ric of memory, that is to say its essential historical means of existence – who can
fail to see that such a question is directly addressed to the language sciences?
And that it calls for an analysis of the material, linguistic modes of existence of
collective memory in categories of discourse? [Courtine, 1994: 10]
The term and the notion of discursive memory have been developed
from the work of historians (in particular the work of P. Nora on
places of memory), as an enrichment of Foucault’s ideas on discursive
formations. In the discourse analysis inherited from the theorisations

of M. Pêcheux, it means thinking about “language reality” in relation
to “historical reality” and so explaining the “historical existence of
the utterance” [Courtine, 1981: 52]. J.-J. Courtine insists on the non-
psychological dimension of this concept, and its links with the disci-
pline of history. Discursive memory is thus defined in conjunction
with the notion of “domain of memory”, which allows discourses to
be situated within the “long period of discursivities” and, conceptu-
ally, allows the reintroduction of history into discourse analysis.
What we mean by the term “discursive memory” is distinct from
any psychological memorisation of the type for which psycholinguists
like to produce chronometrical measurements […] The notion of dis-
cursive memory concerns the historical existence of the utterance
within discursive practices governed by ideological apparatuses; it
alludes to what Foucault […] notes in relation to religious, legal,
literary and scientific texts,
[…] discourses which lie at the root of a certain number of new acts, of words
which recapture them, transform them, or speak about them – in short, discourses
which, over and again, beyond their formulation, are said, remain said and are
still to be said [Courtine, 1981: 52–53].
This proposal found an immediate echo with discourse analysts of

the 1980s, since P. Achard and M. Pêcheux offer definitions and theo-
retical functions of it in the collection Histoire et linguistique pub-
lished in 1984. “[…] discursive memory”, explains M. Pêcheux,
[…] would be that which, faced with a text appearing suddenly as an event to be
read, comes and reestablishes the “implicits” (more technically speaking,
preconstructions, elements quoted and reported, transversal discourses, etc.) which
are required to read it. It is the condition of the readable in relation to the readable
itself [1984: 263].
But it was only from the late 1990s that the notion was reworked, in
particular by S. Moirand, who was keen to put forward the idea of a
“memory of words”, joining the legacy of French discourse analysis to
the older and more philosophical legacy of M. Bakhtine. The latter
thinks of this memory as a sort of semantic chain endowed with two

essential features: it is infinite (it has no beginning or end); and it is dy-
namic (meanings evolve along the chain). S. Moirand refines the distinc-
tion by also proposing the notion of “memory of statements” (mémoire
des dires), leading her to the concept of interdiscursive memory, which
is particularly at work in media discourses on science. Certainly, memory
is not history, but taking account of “the long period of discursivities”,
tracking discursive journeys (contaminé moving from plasma to
transgenic maize) and analogical reprises (barbares being used to
describe those who tear up GMO crops) is to take account, within
discourse, of a transmissive dimension which is the historical one.
The workplace of memory, the “condition of what is readable”,
but also of what is describable and sayable, is probably one of the
most promising perspectives of present discourse analysis – and who
knows, it may well enable discourse analysis to rediscover history.
The discourse of discourse historians: upholding linguistics
Let us finish by referring to the work of discourse historians or “dis-

course archivists” according to the description recently proposed by
S. Bonnafous [2006], in the tradition opened up by the work of R. Robin
[1973]. For them, it is a question of “upholding linguistics”, for the
reasons set out above in relation to the “linguistic turn”. As pointed out
by D. Mayaffre in a defence accompanied by a rather polemical illus-
tration of lexicometric tools and, more broadly, of scientific rigour in
history, “we are living in a period where subjectivity and historical
impressionism shamelessly hold sway” [2001: 1].
Discourse historians (one thinks of J. Guilhaumou and D. Mayaffre,
but also, to a certain extent, of A. Steuckardt, M. Deleplace, F. Mazière,
S. Branca and A. Collinot) are indeed concerned to rigorously describe
the way in which language shapes history: they work with notions
like “archive”, inherited from Foucault, as well as with “linguistic
event” and “thematic journey”, these last two having been defined by
J. Guilhaumou.
The notion of the archive is designed to show how history is

constructed discursively:
Archive does not simply refer here to all the texts which a society leaves behind
it. It is the raw material explored jointly by traditional historians and discourse
historians, but on the basis of which discourse historians do not give priority to
researching hidden social structures: this means that an archive is mainly a sys-
tem of utterances which are not regulated at the outset and which constitute dis-
tinct figures, objects and concepts. Thus each archival system establishes its own
order (Guilhaumou, 2003: 13).
Of course, the term system echoes the terms used by Foucault in defin-
ing an archive: specific rules, practices and systems of utterances.
In this respect it is remarkable that J. Guilhaumou now pleads for
an “ethno-method” in discourse analysis, explaining that “‘ethno’ refers
here both to culture, the local beliefs of members of society beyond
membership of a particular group, and the knowledge these members
have of that culture, in which they are the actors, protagonists, authors
and spectators all at the same time” [2002: on line]. The underlying
principle of this “ethno-method” is a form of inter-subjectivity, which
refers to the co-construction of a discourse-corpus by the researcher
and the subject being researched. The “new corpora” developed in this
way are based on a conception of discourse which combines an “ethical
approach” with the principle of the subject’s autonomy, taking ac-
count of the other in a way which is, in the end, very phenomenological.
ACHARD P. et al., ed.

1984: Histoire et linguistique, Proceedings of the Round Table
“Langage et société”, April 1983, Paris, Éditions de la MSH.
BONNAFOUS S.
1991: L’immigration prise aux mots, Paris, Kimé.
2006: “Les déclarations de Journée internationale des femmes, entre

récit, occultation et performativité”, Communication, Laval, Quebec,
Éditions Nota bene.
COURTINE J.-J.
1981: “Quelques problèmes théoriques et méthodologiques en ana-
lyse du discours”, Langages 62, Paris, Larousse, 9–128.
1991: “Le discours introuvable: marxisme et linguistique (1965–1985)”,
Histoire épistémologie langage, 13/II, Saint-Denis, PUV, 153–171.
1994: “Le tissu de la mémoire: quelques perspectives de travail
historique dans les sciences du langage”, Langages 114, “Mémoire,
histoire, langage”, Paris, Larousse, 5–12.
GUILHAUMOU J.
2002: “Le corpus en analyse de discours: perspective historique”,
Corpus 1, electronic journal, “Corpus et recherches linguistiques”
[on line] <http://revel.unice.fr/corpus/>
GUILHAUMOU J. and MALDIDIER D.

1994: “La mémoire et l’événement: le 14 juillet 1989”, Langages
114, Paris, Larousse, 109–125.
HARRIS R.
2004: The Linguistics of History, Edinburgh, United Kingdom: Edin-
burgh University Press.
MALDIDIER D., pres.

1990: L’inquiétude du discours. Textes de M. Pêcheux, Paris, Éditions
des Cendres.
MAYAFFRE D.
2001: “History and information technology: the French are way be-
hind”, Lexicometrica 3, electronic journal [on line] <www.cavi.univ-
paris3.fr/lexicometrica/archives.html>
MAZIÈRE F.
2005: L’analyse du discours. Histoire et pratiques, Paris, PUF.
MOIRAND S.
2004: “L’impossible clôture des corpus médiatiques. La mise au jour
des observables entre contextualisation et catégorisation”, Tranel 40,
University of Neuchâtel, 71–92.
NOIRIEL G.
1996: Sur la “crise” de l’histoire, Paris, Belin.
2003: Penser avec, penser contre, Paris, Belin.
PÊCHEUX M.
1984: “Rôle de la mémoire”, in Achard P. et al. (ed.), 261–267.
PROST A.
1988: “Les mots”, in Rémond R. (ed.), Pour une histoire politique,
Paris, Seuil, 255–287.
PROVOST G.
1969: “Approche du discours politique: socialisme et socialiste chez
Jaurès”, Langages 13, Paris, Larousse, 51–68.
RICOEUR P.
1997 [1986]: L’utopie et l’idéologie, Paris, Seuil.
ROBIN R.
1973: Histoire et linguistique, Paris, A. Colin.
TOURNIER M.
1992: Des mots sur la grève. Propos d’étymologie sociale 1, Paris,
Klincksieck.
141
Chapter IX
The role of psychoanalysis
Dominique DUCARD
Preliminaries
The most precise definition of psychoanalysis given by Freud em-

phasises the consistent and fundamental link between rehabilitation
technique, therapeutic method and meta-psychological theory:
Psychoanalysis is the name 1) of a method of investigating psychic processes

which are inaccessible by other means; 2) a method of treating neurotic disor-
ders, based on this investigation; 3) a set of ideas about psychology acquired by
this means and gradually built up into a new scientific discipline.1
In other words, the episteme as theoretical knowledge is inseparable

from techne, implying skill and the art of conducting a cure. Thus,
the question of exporting analytical concepts and applying them to
other domains and practices is posed right away.
However, we should not forget that psychoanalysis has a bearing
on anthropology and that the pathological clarifies what is “normal”:
Freud’s thinking was nourished by the science of his time, as well as
by mythologies and literature, and he sometimes found models here
for comprehending and representing his own discoveries. Greek leg-
end provides an example of this in the story of Oedipus, whose name
was also given to a complex. In this way, Freud invited his detrac-
tors, faced with the scandal of infant sexuality, to read the wonderful
1 Freud, 1923 (1984).

142 Dominique Ducard
accounts which figuratively reveal the drama of human beings grap-

pling with their impulses:
If you have been led to suppose that everything recounted by psychoanalysis

about early infant sexuality originates from the extravagant imaginings of ana-
lysts, you must at least acknowledge that this imagination has produced the same
results as the fantastical activity of primitive humanity, which is crystallised in
its myths and stories.2
So the first point will be to recall this principle: the symbolic inven-
tions which demonstrate what establishes humans in their relation-
ship to the world are the result of an imaginary activity which brings
unconscious elements of the sexualised psyche into play, represent-
ing them by the anamorphosis of fiction. The recounting of dreams is
a prime example of this work of figuration: when communicated
through narrative enunciation in a given situation, they are deformed
and transformed in the process of being verbalised and censured.
The second point, which relates to this last aspect, regarding enun-
ciation, will therefore be to suggest that saying (le dire) is as impor-
tant, and sometimes more important, than the said (le dit): words are
always double, turned towards an interior-exterior (the self and the
world) of which they are the form of expression, and addressed to
another than the self in a relationship of inter-subjectivity of the order
of an asymmetrical reciprocity. For the observer, the first facet corre-
sponds to the question “what is he saying?” or “what does he mean?”,
and the second to the question “what does he want for the other (or
from the other) that he talks to him in this way?”. These questions may
undergo multiple variations according to the situation. The subcon-
scious, by definition inaccessible directly, is only able to seize hold of
itself by dressing up in masks and disguises provided by fiction or
ordinary discourse. The trifling events of daily life do not escape this,
as can be seen from analysing lapses, obstacles, omissions, tics or
ritualised behaviours, habits and obsessions of all kinds. So another
point needs to be emphasised: not only is it possible to decipher fan-
2 Freud, 1926 (1985: 74).

The role of psychoanalysis 143
tastical scenarios in what a person recounts and to perceive a means of

expressing his desire in the way he addresses his interlocutor, present
or imagined, but it is also possible to detect in meaningless words or
phrases, apparent absurdities, defects and slips of speech, some clue
as to this particular subjectivity. Lacan focused the attention of psy-
choanalysts on those tricks played on us by the signifier and the de-
tours we are unknowingly forced to make by a few key signifiers,
which could nonetheless seal the subject’s fate.
To all this may be added the way in which each person goes
about his social occupations (familial, professional, political); the way
in which he adapts, with enthusiasm or displeasure, to the functions
assigned to him; carries out the tasks that devolve on him, whether
imposed or freely chosen; and takes on commitments in more or less
extended circles – in short, what determines the role that person plays
as a woman or man within a community.
Choice of viewpoint
These various points have been briefly recalled in order to locate the
place which psychoanalysis can have in an approach to word and
discourse events. They can be reduced to the three dimensions of the
sign in Peirce’s semiotics, according to the relationship established
between the representative (representamen) and its “object” of repre-
sentation: indexicality, iconicity, symbolicity. We can interpret a par-
ticular signifying form as an index or mark of subjectivity (for exam-
ple, a language tic reduced to an obsessional trait), as a representational
image (for example, a particular metaphorical network reflecting a
paranoiac modality of fear of the other), or as a symbol or sign of
interpersonal gratitude and social pact (for example, the use of a “we”
which unites and identifies). These three functions of the sign are not
of course mutually exclusive; they indicate three ways in which the
process of meaning operates.
My objective will be to show that psychoanalysis can play a part

in discourse analyses. This will not involve a review of the literature
on the subject, nor an account of the various applications of psycho-
analytical theory where it offers textual commentaries. The observa-
tions here will be more closely focused on revealing what seems to
accord with the shared perspective of this present book. I should state
at the outset that it is not about “psychoanalysing” the authors of
discourse by applying a particular set of interpretive rules, be they
Freudian or other, nor by indexing linguistic forms under key words
relating to a list of disorders or personality types. This method, fo-
cusing on the author’s supposed intentions, whether conscious or
unconscious, has been criticised for imposing a reading chart and
carrying out a sort of content analysis akin to amateur psychoanaly-
sis, which is contrary to an analytical practice that respects Freudian
techne. The other tendency would be to place oneself beside the sin-
gle reader who reads himself and discovers himself in the mirror of
the text, that is to say by resolutely making himself the subject and
object of the investigation by self-analysing his transference. My in-
tention here is to occupy the intervening ground in a study which
does not deny the subjectivity of reading and interpreting decisions
but which postulates that all texts, that is to say any meaningful ar-
rangements of linguistic markers, contain traces of the operations
and representations by which they are generated and produced, and
which are reconstructed by the person on the receiving end. Some of
these traces reflect modes of thinking and communicating for which
psychoanalysis can offer schemas to aid comprehension. Let me stress
once again that my aim is not to psychoanalyse an author or text, but
to identify, within discursive forms, the modalities of subjectivity
which fit with what psychoanalysis teaches us about language activ-
ity. In relation to this, it is appropriate to cite Barthes’ question where
he seeks to emerge from a semiology of the quoi [what] to privilege
a semiology of the qui [who]:
The interplay of places between the other and me is an object of psychoanalysis.

But how can this research be given a semiotic interpretation (or aspect)? How may
a discourse-scene be analysed? How can places and their degrees of proximity be
classified, how can the positions of enunciation – of interlocution – be determined?

This is the object of this new linguistics (or semiology) which is in search of itself.3
At the time, Barthes, who was wondering how to understand the in-
ternal movement of a discourse, the changes of intensity and trajec-
tories and the power games, set his hopes partly on pragmatics – he
was then working with F. Flahaut and F. Récanati4 – and partly on
Nietszche’s “active philology”, with its more musical analytical model.
My own methodological benchmark will be the theory of enunciative
operations developed by Antoine Culioli. He defines linguistics as
the study of language through studying languages, texts (oral or writ-
ten sequences) and situations. Following in his footsteps, but with a
different perspective, centred on texts as they are situated and insti-
tuted, I am seeking to understand the signifying activity of language
as an activity of representation and interpretation. In the wake of
Barthe’s questioning, I will therefore place the emphasis on inter-
subject relations in discursive practices. This was the route indicated
by Lacan when he invited disciplines which he labelled “inter-sub-
jectivity sciences” (and also called “conjectural sciences”5) to com-
municate between one another.
3 Barthes (2002: 211).

4 Flahaut (1978) and Récanati (1979).
5 I shall not deal with studies which use computer processing to identify psycho-
social characteristics or personality traits, for the purposes of hypothesis and to
compare them with other data (pieces of evidence, interviews, confessions, biogra-
phical elements, and so on). I would recommend instead the scientific approach
followed by M.-C. Noël-Jorand (a discourse analyst and psychoanalyst) and his
team, who, by using the Alceste software and without hazarding any interpreta-
tion, have discovered certain specific structures, called “language satellites”, in
pathological subjects. These are a sort of second discourse which, in a way, comes
to “parasitise” the main discourse and whose theme is independent. The method
has been tested with other discourses, outside of any nosographic classification,
and shows the permanence of these significant groupings. See Noël-Jorand,
Reinert, Giudicelli, Dassa (1997), and Noël-Jorand, Reinert (2003).
Specular communication, subject and institution
Lacan’s theorising went by the name of mirror stage or, better still,
mirror phase, referring to that genetic and structural moment when the
small child forms a unified image of his own body by capturing a
mirror image of himself in the image of another, by seeing his specular
reflection in a solid mirror or in the image of another – a moment which
marks the emergence of the ego into the dependency of the alter ego.
This moment, which cannot be reduced to a stage of development, is
part of the way the self is formed by the imagination – that fictional
thread, according to Lacan, through which the series of imagos (in
which the subject is alienated but also constructs his identity) makes
its entry into a set of identifications. In order to do this, the strategy
incorporates a third party, whose function is to indicate and name the
subject taking hold of this image, as of an entity which goes to form
his identity. This third party, who in a way mediates the dual relation-
ship, ensures that the imagination is symbolically regulated through
recourse to language and to a legislating discourse: it indicates places
and shows the way. Thus Lacan distinguishes between two modalities
of inter-subjectivity, of an imaginary order and a symbolic order. The
first concerns a self conceived of as “the sum of the subject’s identi-
fications […] like a layer of different coats borrowed from what I shall
call the bric-a-brac of its accessories store.” The second concerns the
“organised system of symbols, claiming to cover the whole of an
experience, and to give it life and meaning”6: this is shared, historical
discourse which has an underlying value (parents, family, community
– whether cultural, social, political or religious – and so on): the
orthodoxa. The scenario of the specular phase, roughly summarised,
must be considered as a theoretical fiction enabling the imaginary
game of communication to be set against a scene where the characters
in the theatre of the self move around, under the guidance of a producer
who is more or less directive, with a possible deus ex machina.
6 Lacan (1954–1955) (1978: 215 and 62).

I have referred to this model in a variety of discourse studies,

and more specifically to Pierre Legendre’s essential thesis7 as part of
his endeavour to found a dogmatic anthropology: “society is a func-
tion for the subject”, a function of figuration and institution of im-
ages for the social subject. This function is possible because society,
with the cultural authorities which mediate it, is a second mirror by
virtue of projection of the first mirror, the mirror of the specular phase
when inter-subjective relations (identity and otherness) were being
shaped. It is through this function that society is apprehended as a
Text, which can be assimilated to all discourses with normative and
prescriptive value, without being limited to the most obvious case
which is that of legal discourse. Before clarifying this point, let us
cite the guiding questions of P. Legendre’s vast and prolific investi-
gation:
In plain language, what relationship is there between word, institution and sub-
ject? And how can this relationship be translated into a link with reality, into a
social link, or a link with Reason? What is the source of the credit accorded to
words, of faith in the images which support them and allegiance to the invisible
architecture which inscribes each individual in the social construction of the Text?8
The attempt to think about the subjectivation of the institution and

the institutionalisation of the subject through recourse to psychoa-
nalysis is based on increasing the specular mechanism on the social
scene where the dogmatic order is formed – the order of norms and
rules legislating the world of belief in Power and legitimising the
regime of submission to the Law. And we should note in passing that
the proofs of this dogmatic order “all relate”, according to P. Legendre,
“to a semiology of Western representations about the institution”.9
The social imagination invests the subject with an edifying character-
istic. So this different organiser, representing the Rule, whatever its
mode of manifestation, the one who prescribes, legitimises, censures,
7 See Legendre (2001).

8 Legendre (1998: 18).
9 Legendre (1974: 59).
authorises or forbids, in the name of what look like ideals, responds

to the other ideal, the one which underpins early identifications and
confirms and directs the subject’s identification-otherness to his mir-
ror image (“you are this”). In this respect, P. Legendre’s work on
Canon Law demonstrates how legislation, handed on by an art of
casuistry and the mise en scène of the question of school of thought
in the proceedings, aims to resolve the fundamental uncertainty: what
must one say? what must one think? by contributing conclusions dic-
tated by the social Text.10
Transposed to other cultural and social locations, this set of prob-
lems should enable us to ask questions about the imaginary places
we think we are obliged to occupy or that we unknowingly occupy,
and about the signs, having the value of badges, to which we give
credit or in which we recognise ourselves. Saussure emphasised the
institutional character of “language”, insisting that it was necessarily
fated to be always in circulation and to make itself an object of trans-
mission subjecting each locutor to its own order. Discourse analysis,
which should enable us to “think about the enunciation strategy which
links a particular textual organisation to a specific social location”11,
is interested in the institutional functioning of discourse but does not
really question what institutes the meaning of these discourses and
what institutionalises the word (parole) within a genre of discourse.
A semiological field is thus opened, composed of various discourses,
icons, emblems, ceremonies and scenes, which act as so many
“specular lassoes”, according to G. Le Gaufey’s expression12, in which,
individually and collectively, subjects and their words are caught.
10 See Legendre (1974).

11 Maingueneau (2002: 43).
12 Le Gaufrey (1997). The author has borrowed the term “lasso” from the logician
David Lewis (Parts of Classes, Cambridge, Mass., Basil Blackwell, 1991). In de-
fining specular, he uses it to name the power of the image to draw people together
and the relationship of resemblance it has to its object, both of which determine
its function and its very nature. “Vaster still than the dominion of the concept, the
specular lasso reigns over images, and even beyond, over all that founds the unity
of our representations in the multiplicity of our perceptions” (p. 261).
Enunciative instances
Enunciation and adopting a position
The transposition of the specular mechanism onto the social scene,

where it then becomes a general model for exchanges in the commu-
nication circuit, with its badges and emblems, has its linguistic coun-
terpart in the theory of enunciation, which offers us the means to
make a linguistic analysis of our interpretive objective.
Let us then recall a few basic elements. The production of utter-
ances corresponds to the construction of more or less complex predi-
cative relationships, which reflect various, mixed representations. In
order to become utterances, these relationships are situated in an
enunciative space formed into a validation domain which includes a
system of reference (space-time) and an inter-subject relationship.
The enunciating subject, who is the subjective origin of the utter-
ance, is accountable for the act of enunciation he is carrying out.
Thus, in an assertion, in the strict sense, the subject commits himself
by taking up a position. The assertive schema is recapitulated by a
sequence of the type “I am keen to say that I know (I believe) p to be
true”, in which are combined commitment, declaration and mode of
validation. A. Culioli has completed this recapitulation by detailing
the components of the schema, and the statement, thus set out and
explained, becomes “I am keen to say that I believe/think/know/
understand p to be the case”.13 The distinction between what A. Culioli
points to as being an “instance of representation”, which is more pre-
cisely an instance of validation of representations, and what might be
called an instance of subjective position might lead us to re-examine
together the linguistic conception and the Freudian (notably Lacanian)
conception of enunciation, when there is a difference or disassocia-
tion between instance of validation and position.
13 See Culioli (1999: 92–99) and (2002).

The term instance, in all its various meanings, here has the great
advantage of being able to refer in linguistics to the referents of pro-
duction of a discourse (situation, enunciator-co-enunciator, locutor-
interlocutor). Moreover, in its legal usage it indicates a series of acts,
periods and formalities with the object of introducing, investigating
and judging a lawsuit, but also the authority, the constituted body
which holds the power of decision. It is also used in psychoanalysis
terminology for the different dynamic components of the psychic
mechanism (ego, id and superego of the second Freudian topics: in-
stance of censure, of the superego), or again for the subconscious
structuring of the signifier for Lacan: “the instance of the letter in the
subconscious”14. The instance is thus a place to occupy for an opera-
tion to be carried out.
Imaginary places and symbolic authorities
Several successive pieces of research were undertaken as part of my

participation in the Céditec team’s “workshops”. Through an exami-
nation of phenomena considered to be particularly significant, these
show how, in my view, the semio-linguistic approach to texts relat-
ing to institutional practices allows markers of operations to be at-
tributed to instances and interpretable positions by reference to im-
aginary places and symbolic authorities.
As part of a joint workshop on “Evaluation discourses and uni-
versity and educational institutions”, research was conducted on re-
ports of viva voce examinations in different disciplines. One element
of this research which is worth mentioning here was the analysis made
of the concession device, which was recurrent and insistent. Conces-
sion (il est vrai, certes,… mais…) [certainly, that is true… but…] is
based on operations of validation (saying what the case is, effective)
and valuation (appreciative judgments of good/bad, useful/useless,
normal/abnormal, happy/unhappy, and so on) and reflects the place
14 Cf. Lacan (1966: 249–289).

of the reporter-judge. Concession also shows conflicts of representa-

tion which are resolved by a sort of transaction, in the legal sense of
the term: any contestation is averted and at the same time some sort
of arrangement or compromise is arrived at. It was important then to
show that the imaginary place of the enunciator, who is both judge
and jury – whether the reporter is the thesis supervisor or a peer
amongst peers – is subject to a third instance, namely the university
institution, which is both speaker and recipient of the discourse be-
ing held. The point of view put forward goes beyond a commentary
in terms of discursive strategies or argumentative methods deployed
in an interaction dependent on a formal “ceremony”.
Let us remind ourselves of the role attributed in traditional rhetoric
to the ethos of the orator, that is to say his character as it is manifested
in discourse, in accordance with the subject. Its aim is to establish a
contract of trust between him and his listeners in order better to per-
suade them. In my view, the notion of imaginary place, which re-
flects the “imaginary formations” put forward by Michel Pêcheux in
his critical re-reading of the communication schema,15 is supported
by what has been said about the specular mechanism and appears as
an interpretation of what emerges from an identification of enuncia-
tion markers, in terms of their position and instance. The expression
symbolic authority seems perhaps to indicate what, in an inter-sub-
ject relationship, relates to a third-party, common, transcendent in-
stance. The word authority, adopted for all the meanings listed by
dictionaries, would be sufficient to justify its use in relation to the
badges of power, by indicating any power of imposition16. The word
15 Pêcheux (1969).
16 We should also mention the different meanings of the word authority given by the
Oxford dictionary: “1. The power or right to control, judge or prohibit the actions
of others. 2. A person or group of people having this power, such as a government,
police force, etc. 3. A position that commands such a power or right (often in the
phrase in authority). 4. Such a power or right delegated, especially from one
person to another. 5. The ability to influence or control others. 6. An expert or an
authoritative written work in a particular field. 7. Confidence resulting from great
expertise.”
badge, then, can indicate the distinctive marks of ideal values el-
evated into guiding principles, norms or models of identification (in
the name of what do we speak?, but also in the name of what do we
act?). Emblems are the figurative attributes of these (by analogy with
what are, for example, the official emblems of the Republic: flag,
national anthem, currency), sometimes gathered into a description
which takes the place of an effigy.
The second example will be taken from a study conducted while
preparing for the Cerisy-la-Salle conference on Argumentation and
Political Discourse17. One of the conference sessions was especially
concerned with the problem of argumentation in its relationship to
what is sometimes called manipulation. We had to try to respond to a
simply worded question: are there linguistic markers of manipula-
tion? In other words, are there forms of enunciation whose use en-
sures that one person has a hold over another, without his or her know-
ledge? Interest was immediately focused on a figure of speech
considered as an obvious form of misleading appearance and duplic-
ity in the exercise of argumentative discourse: the so-called rhetori-
cal or oratorical question, or false question. This study linked up more
generally with figures of meaning which are close to the methods that
old rhetoric classifies as varieties of figurative discourse (Gr. logos
eskhematismenos, Lat. sermo figuratus, oratio) which are often pre-
sented as being motivated by pretence or sham.
Rhetoric describes an oratorical or false question by opposing it
to a real question which ascribes to another person (the present or
imagined interlocutor) the function of raising a doubt or replying to
some information. The description made of this interrogative con-
struction, which is not a real question but a quasi-assertion that does
not appear as such, underlines an inter-subjective relationship which
17 International conference Argumentation and Political Discourse. Ancient Greek

and Roman Civilisation, the French Revolution and the Contemporary World,
organised by Céditec and the philosophy and rhetoric team of the Université
Paris 12-Val de Marne, International Cultural Centre of Cerisy-La-Salle, 3–9
September 2001. See Bonnafous, Chiron, Ducard, Lévy eds. (2003) and my study,
Ducard (2004: 73–88).
involves the subjection of the other. Since my investigation was being

carried out in the field of political public debate and had revealed
intensive use of this device, it seemed appropriate to pose the problem
of the specificity of political words in the discourses and discussions
of Senate members (the corpus being examined on this occasion).
As shown by linguistic analysis of the rhetorical question known
as a biased question (Culioli), underlying this form of enunciation is
an operation validating an assertion which presents itself under the
guise of an interrogative schema but does not reflect the off-centre
position of the enunciator when a request for validation by recourse
to another is made in an inter-subject relationship. The other is then
placed in the position assigned to him by the enunciating subject and
receives from that subject a mirror image of himself. The rhetorical
question is a figure of consent showing proof that the assertion it
contains is true. Of course, this figure does not belong only to public
political debate but it confirms the status of a politician’s words, which
depends on reasons and convictions that institutional conflict situa-
tions tend to radicalise, to the detriment of another rationality, namely
that of discussion.
Three modes of validation
Following references to Pierre Legendre’s dogmatic anthropology

and his studies on the institutionalisation of subjects and signs, I shall
briefly indicate the ideas put forward by Vincent Descombes on what
he calls, in an eponymous work18, the institutions of meaning. The
central thesis of V. Descombes’ book, which he calls “anthropologi-
cal holism”, is that the “objective mind” of institutions is a condition
of possibility of the “subjective mind” of individuals. The essence of
the demonstration, approached by means of analytical philosophy,
aims to confirm the idea that the impersonal use of signs (as laid
down and which makes meaning possible), precedes and, in a way,
18 Descombes (1996).
goes beyond personal usage. This philosopher refers to Charles Tay-

lor’s opposition between “inter-subjective meanings” taken in a rela-
tionship of the self to others (or of the self to the self as another) and
the “shared meanings” determined by institutions and those ways of
thinking and acting laid down in and by society19. The institutions
of meaning, beginning with language, which Saussure considered a
“semiological institution”, impose themselves in communication be-
tween subjects and govern relationships of thoughts, words and ac-
tions. The “subject of institutions” is therefore “the agent whose ac-
tion finds in the institution its model and rule”20. And “the objective
mind” marks “the presence of the social in the mind of each one of
us”; it has this power of authority, social, moral and mental. We might
say, following P. Legendre, that it has a dogmatic value for subjects.
“To understand the authority of the objective mind over subjects”,
says V. Descombes,
[…] it is necessary to think quite differently about the function of (impersonal)

instituted meaning in the formation and communication of thoughts. The priority
of the impersonal over the personal is not the priority of the text over the reader
or transcriber. It is more the priority of a rule over the activity it governs.21
I have adopted this idea of the impersonal to deal with the fact that, in
enunciative activities by subjects, a relationship is established which
is not of the order of an inter-subjective relationship nor one which
makes the latter dependent upon a higher order. Thus it is, for example,
19 V. Descombes gives one of Mauss’s definitions of the word “institution”: “What

in fact is an institution if not a set of completely instituted acts or ideas that
individuals are faced with and which are more or less imposed on them? There is
no reason to reserve this expression for basic social arrangements, as is normally
the case. So we understand this word to mean customs and ways of behaving,
prejudices and superstitions, as much as political constitutions and essential le-
gal organisations; for all these phenomena are of the same nature and only differ
by degree.” (Fauconnet and Mauss, article “Sociologie” in the Grande Encyclo-
pédie, quoted p. 296).
20 Idem, p. 302.
21 Idem, p. 333.
that the rhetorical question depends on a simulated inter-subjective

modality and can be based on a guarantor relating to the impersonal:
the figure of a consent guaranteed by common sense (an “isn’t it?”
being equivalent to “anyone knows that” or “everyone will acknow-
ledge that”). By way of conclusion to an article on justification and
“kettle” reasoning22, I cited the three demands put forward by
Habermas23 for mutual linguistic comprehension: a demand for truth
with regard to the objective world; a demand for precision with re-
gard to the world of social community; and a demand for sincerity
with regard to the personal subjective world. I put them parallel to
the three “claims to validity” of discourse according to the three di-
mensions of the world in K.O. Appel’s semantic-pragmatic model.24
Criticising the idealised version of communication according to these
conceptions, I envisaged examining the question of validation accord-
ing to a triad of three enunciative modes: referential, inter-subjective
and subjective. It would be more appropriate – but that will have to
be the focus of deeper reflection based on the study of specific cases –
to question the meaning to be assigned to utterances according to
whether the instance of representation and validation is subjective,
22 Ducard (2004, pp. 89–108).

“Kettle”-type reasoning was indicated by Freud when analysing the account of a
dream known as “Irma’s injection”. After noting a series of conflicting, even
contradictory explanations, he told the following anecdote: “The whole plea –
for the dream was nothing else – reminded one vividly of the defence put for-
ward by the man who was charged by one of his neighbours with having given
him back a borrowed kettle in a damaged condition. The defendant asserted first
that he had given it back undamaged; secondly, that the kettle had a hole in it
when he borrowed it; and thirdly, that he had never borrowed a kettle from his
neighbour at all. So much the better: if only a single one of these three lines of
defence were to be accepted, the man would have to be acquitted.” (The Inter-
pretation of Dreams, London, George Allen & Unwin Ltd, [translated and edited
by James Strachey] 1961: 106–121).
23 Habermas (1986).
24 The three “claims” according to this author are: the claim to truth with reference
to what is said; the claim to normative precision with reference to the claim to
truth in its performative formulation, regarding partners in communication; and
the claim to sincerity or veracity. See Appel (1994).
inter-subjective or impersonal. For, as I pointed out in passing in this

conclusion, it is not sufficient to share the same objective world (in-
cluding directly choosing the position indicated above) in order to
agree on the reality to which we are supposed to be referring. We
may add to this the ruses of reason and tactics of all kinds, as well as
causes that reason itself is unaware of or disregards.
I have deliberately limited my remarks on the role of psychoa-
nalysis by focusing on enunciative activity in its relationship with
specular imagination and symbolic authority. Aiming to understand
this activity, with its underlying representations and interpretations,
goes beyond the objectives set by studies of discursive practices – or
rather it duplicates them insofar as interpretive semiology as we un-
derstand it requires us to go through formal analysis. This in turn is
subjected to anthropological questioning which takes account of the
knowledge contributed by psychoanalysis to human and social events.
Since words are necessarily both private and shared, to use a co-
ownership metaphor, this questioning inevitably crosses paths with
the problems raised in discourse analysis.
APPEL K.-O.
1994: Le logos propre au langage humain, Paris, éd. de l’Éclat.
BARTHES R.
2002: Comment vivre ensemble. Cours et séminaires au Collège de
France (1976–1977)), Paris, Seuil IMEC.
BONNAFOUS S., CHIRON P., DUCARD D. and LÉVY C. ed.
2003: Argumentation et discours politique. Antiquité grecque et latine,
Révolution française, Monde contemporain, Presses Universitaires
de Rennes.
CULIOLI A.
1999: “Accès et obstacles dans l’adjustement intersubjectif”, Pour
une linguistique de l’énonciation. Domaine notionnel, T. 3, Paris, HDL,
Ophrys, 92–99.
2002: “Je veux! Réflexion sur la force assertive”, Hommage à André
Green, Paris, PUF.
DESCOMBES V.
1996: Les institutions du sens, Paris, éd. de Minuit.
DUCARD D.
2004: “De la subjectivité dans le raisonnement: justification et pseudo-
explication”, Entre grammaire et sens. Études sémiologiques et
linguistiques, Paris HDL, Ophrys, 89–108.
2004: “Une discussion biaisée: la question rhétorique dans le débat
parlementaire”, Entre grammaire et sens. Études sémiologiques et
linguistiques, Paris HDL, Ophrys, 73–88.
FLAHAUT F.
1978: La parole intermédiaire, Paris, Seuil.
FREUD S.
1984: “Psychanalyse et théorie de la libido” (1923), Résultats, idées,
problèmes, Vol. II, Paris, PUF.
1985: La question de l’analyse profane (1926), Folio Essais, Paris,
Gallimard.
HABERMAS J.
1986: Morale et communication. Conscience morale et activité
communicationnelle, Paris, Cerf.
LACAN J.
1966: “L’instance de la lettre dans l’inconscient ou la raison depuis
Freud”, Écrits I, Points Seuil, 249–289.
1978: Le Séminaire Livre II. Le moi dans la théorie de Freud et dans

la technique de la psychanalyse (1954–1955), Paris, Seuil Essais.
LE GAUFEY
1997: Le lasso spéculaire. Une étude traversière de l’unité imaginaire,
Paris, E.P.E.L.
LEGENDRE P.
1974: L’amour du censeur, essai sur l’ordre dogmatique, Paris, Seuil.
1998: Leçon I. La 901e conclusion. Étude sur le théâtre de la raison,
Paris, Fayard.
2001: De la Société comme Texte. Linéaments d’une anthropologie
dogmatique, Paris, Fayard.
MAINGUENEAU D.
2002: “Analyse du discours”, Dictionnaire d’analyse du discours,
Paris, Seuil.
NOËL-JORAND M.C. and REINERT M.

2003: “Comparison of textual analysis to two lectures written three
years apart by an author: the Language Satellites”, Psychological Report
92, 449– 467.
NOËL-JORAND M.C., REINERT M., GIUDICELLI S. and DASSA D.

1997: “A new approach to discourse analysis in psychiatry, applied to
schizophrenic patient speech”, Schizophrenia Research 25, 183–198.
PÊCHEUX M.
1969: Analyse automatique du discours, Dunod, Paris.
RÉCANATI F.
1979: La transparence et l’énonciation, pour introduire à la pragma-
tique, Paris, Seuil.
159
Chapter X
Discourse analysis and philosophy:

intersecting perspectives
Malika TEMMAR
By comparison with other disciplines in the human and social sci-

ences, philosophy has, in a way, long been resistant to discourse analy-
sis. Although contemporary philosophers have expanded ways of un-
derstanding “the philosophical utterance” – hermeneutic approaches
in the case of Ricoeur, deconstructivist for Derrida, analytical or posi-
tivist for Wittgenstein and Carnap – they take little account of philo-
sophical forms of expression themselves. In fact, analyses of these
expressive forms are very few and far between1 and tend increas-
ingly to be developed in the field of research2. However, this is hap-
pening in contexts which do not lay claim to discourse analysis and
do not necessarily seek to connect philosophical forms of expres-
sion with the institutional conditions in which they emerge; nor do
they make use of the modern linguistic tools available in language
sciences. In the early 1990s, as part of a seminar at the Collège Inter-
national de Philosophie which started out from the fundamental re-
sistance of philosophy to setting itself up as an object of study3,
1 Two main references may be noted: J.L. Galay, Philosophie et invention textuelle:
essai sur la poétique d’un texte kantien, Paris, Klincksieck, 1977; and P.A. Cahné,
Un autres Descartes: le philosophe et son langage, Paris, Vrin 1980.
2 The first international conference on “Philosophers’ style” was held in Besançon
(organised by B. Curatalo and J. Poirier in November 2005).
3 (Cossutta, 1995: 34). “Philosophy supposes that one is already situated in its own
particular field in order to be able to stand back from it and take it as an object of
study […] from the philosopher’s point of view, no knowledge can be based on
philosophy which is not philosophical.”
160 Malika Temmar
Frédéric Cossutta developed a project to analyse philosophical dis-

course4. Led by him, the activities of the Groupe de Recherche sur
l’Analyse du Discours Philosophique (GRADPhi5) produced several
publications and conference papers6. This research is all based on the
discursive dimension of philosophy; it therefore works upon catego-
ries of text analysis in a variety of ways, revealing different relation-
ships between discourse analysis and philosophy. Sometimes discourse
analysis is seen primarily as a “tool” through which philosophical
clarification of texts can be achieved, and sometimes priority is given
to understanding philosophical discursivity, with its particular char-
acteristics, yet without neglecting the philosophical aspect. Divisions
are made according to which discipline researchers belong to, which
means that the chosen method of investigation is variable. Whereas
F. Cossutta (1996) and A. Lhomme (1998) see discourse analysis as
having a mainly philosophical objective, D. Maingueneau (1996), on
the other hand, analyses philosophical discourse as a discourse among
others; F. Cicurel (2004), as a courseware designer, is concerned with
the conditions in which the philosophical text is transmitted (prac-
tices and methods); J. F. Bordron (1987), coming from semiotics, gives
priority to reading the philosophical text from the viewpoint of con-
structing its meaning; and J. Angermüller (2004) links the concept of
“field” to the French philosophical scene of the 1960s and 1970s. My
own concern is to attempt, through the study of philosophical dis-
course, to bring out the way in which the choice of certain discursive
forms matches the presuppositions associated with different philo-
sophical conceptions. Thus, for example, discursive methods of
4 Frédéric Cossutta, idem.

5 The Groupe d’Analyse du Discours Philosophique [Group for Analysing Philo-
sophical Discourse] is made up of philosophers, linguists, discourse analysis spe-
cialists, historians of philosophy and sociologists. This group met until 1998 at the
Collège International de Philosophie, and continued to exist thereafter without
institutional support. It is now attached to Céditec at the Université de Paris-Est
Créteil.
6 Cf. Bibliography.
Discourse analysis and philosophy 161
ostension of objects are offered for consideration according to the

way the philosopher sees the world [Temmar, 2003].
What might a philosophical discourse analysis consist of? To
what extent can such a study be distinguished from the approaches
that previously explained philosophical utterances? If the philoso-
pher appears, by definition, to use a discourse regime which has a
claim to truth, how can philosophical discourse analysis account for
the characteristics of philosophical enunciation of truth?
How can the relationships between “discourse analysis” and “phi-
losophy” be formulated? These relationships appear to be all the more
structural since reference to certain 1960s French philosophers is widely
presupposed through the setting up of what is called “the French School
of discourse analysis”, Whether one thinks of Foucault or Althusser,
these philosophers have, in many ways and according to different
methods, if not developed discourse analysis in France, at least made
this type of approach possible. In a very simplified way, it is possible
to identify a few markers which show a kind of line of descent between
philosophers and discursivists. With L’Archéologie du savoir, Foucault
set out a discourse analysis project, as part of an interrogation of the
relationships between discursive practices and social practices. He put
in place a series of concepts like “discursive formation”, “archive” and
“discursive practices” which would be taken up and reworked by
certain linguists, like D. Maingueneau for example. Instead of the
notion of discursive formation, Dominique Maingueneau (1991) in-
troduced the notion of “archive” to bring together utterances relating
to a position. At the same time he stressed (through polysemy of the
etymon of archive, the Greek archeion) that these utterances are in-
separable from a memory and an institution which confers them with
their authority, while being legitimised through them7. The concepts
of “interdiscourse” and “preconstruction” [Pêcheux, 1975] are part of
the Althusserian concern to form discourse analysis into a practice
insofar as it allows ideology to be critically analysed.
7 Maingueneau et al., 2002: 62.

162 Malika Temmar
Beyond their crucial role in what was to become “discourse analy-

sis”, it seemed as if a few philosophers played a decisive part in its
development. Where discourse analysis was concerned, the division
between “philosophy” and “French School of discourse analysis” was
made with a scientific determination to remain resolutely linguistic.
With scientific discourse analysis, philosophy went from being a theo-
retical framework for discourse analysis to being the object of analy-
sis itself.
My objective is mainly to present the philosophical discourse
analysis pioneered by Frédéric Cossutta8, and more particularly to
take from it the discursive model presupposed by this type of analysis.
The examples shown are ones that have emerged from my studies of
philosophical texts. As an active participant in the work of the philo-
sophical discourse analysis group (GRADPhi), I have become very
committed to this field of investigation of discourse analysis [Temmar].
Categories of philosophical discourse analysis
General principles: institution and discursive establishment
For philosophical discourse analysis (in the same way as for other
types of discourse), philosophy is dependent on its modes of expres-
sion: a substantial element in the formation of philosophical mean-
ing is at stake in the discursive process. For the first time, a system of
analysis has been constructed which explains the unwillingness of
philosophy to be comprehended from outside, and the general condi-
tions to be fulfilled by such an approach have been put in place. Philo-
8 This article proposes to offer a summary of philosophical discourse analysis, and

I shall first expand on the categories and problems focused upon by GRADPhi.
For a more detailed study of these categories the reader is invited to refer to
the work of F. Cossutta as well as the work of the group (cf. Bibliographical
references).
sophy does not escape the demands inherent in communication (the

need to negotiate a relationship with one’s public and the social insti-
tutions governing the distribution of words); so, by paying particular
attention to the way in which philosophy resolves the problem of its
expressiveness, this approach connects the conditions under which
philosophical discourse is instituted with the conditions under which
it is established: “Its discursive institution mediates the relationship
between works and context, its discursive establishment mediates the
relationship between expressive forms and speculative schemas”9.
In an article entitled “L’énonciation philosophique comme institu-
tion discursive”, Dominique Maingueneau develops a conception of
philosophical discursivity as “discursive institution”. He shows how
[…] in the process through which philosophical enunciation is instituted one

(cannot) dissociate the operations through which discourse develops its contents
[establishment] and the method of institutional organisation which discourse
both presupposes and structures [institution]. Discourse is at one with the way it
manages its own emergence, and the word event which it institutes; it represents
a world in which its enunciation is a participant.10
The handling by philosophers of various forms of address, which

submits this type of discourse to a regime of “hetero-genericity”, is
certainly not accidental – it is tied up with philosophical doctrine. Out
of multiple genres, the philosopher seeks the one which will ideally
coincide with his doctrine. Descartes, for example, wrote discourses,
principles, treatises, a fable, objections to replies, meditations, medi-
tations re-written in geometrical form, a dialogue, letters etc., all in a
meditative genre (very eloquent in the 17th century) and there is a very
clear convergence between the philosophy expounded and the manner
of its presentation. In Méditations métaphysiques, it was the introspec-
tive approach which allowed access to knowledge. If “nothing can be
known before understanding”11, then meditation, which undoubtedly
facilitates “withdrawal” into the self, will be a favoured tool and an
9 Cossutta, 1995: 19.

10 Maingueneau 1995: 40.
11 Descartes, Règles pour la direction de l’esprit, 11.
164 Malika Temmar
integral part of the philosophy itself. Truth is then to be sought by

simple mental introspection and Descartes made meditation the cor-
nerstone of his philosophy.
It is by taking account of the context in which philosophical work
is set that philosophical discourse can be made to appear in the his-
torical context in which it emerged, together with the persons for
whom it was intended and the social background against which the
philosopher enunciates his philosophy. To this may be added particu-
lar generic choices (meditation, treatise, dialogue, criticism for ex-
ample) depending on the communicational constraints under which
the philosopher produces his discourse.
Based on its conditions of institution, it would be fitting for philo-
sophical discourse to construct a conceptual world which claims its
autonomy and particularity by setting itself in relation to other philo-
sophical doctrines or traditions. In this philosophical work of build-
ing up meaning (discursive establishment), the citational relationship,
to take this discursive method as an example, is never reproduction/
repetition: it is taken up in a process which means that, while posit-
ing the discourse of the other, the philosopher posits himself through
it by means of distant integration12. Philosophical discourse is en-
gaged in this dual activity: it is rich in detachable utterances (differ-
ent sorts of formulae, sentences, maxims) which circulate and are re-
inscribed in various texts. The doctrinal space is not closed, but appears
as a place of “transit” and of re-inscription of utterances or formulae
coming from other philosophies. But it is also through the activity of
citations that autonomy is set up, together with the specificity of a
particular philosophical discourse in relation to another. A study of
the forms of reprise of formulae in Merleau-Ponty (from classical
philosophy as well as from Cartesian philosophy or Husserlian phe-
nomenology) shows that these reprises share a philosophical particu-
larity which invites us to consider that there is a form of community
between philosophers, in the sense that each discovers himself and
12 Malika Temmar, “Les forms de reprises chez Merleau-Ponty”, in Formules, sen-

tences et maxims en philosophie (awaiting publication).
learns to be himself through the work of another, in a thinking rela-

tionship which constitutes a fraternity13. The specificity of philosophi-
cal discourse in relation to other discourses of the human and social
sciences has still to be determined.
Linguistic tools
Philosophical modes of expression are studied by analysing the lin-

guistic and discursive operations which go to make up a philosophical
text. For this reason, the enunciative markers, the generic context
chosen by the philosopher, examples of discursive heterogeneities, or
polyphonies, are all subjected to closer examination. These discursive
phenomena are taken into account insofar as they form an “enunciative
scene” (Maingueneau) through which the philosopher represents his
thought. Since discursive operations are seen in the dynamics of a
philosophical text (and linked to its doctrine), discourse analysis fa-
vours a linguistics of operations (the linguistics of Culioli for Bouacha
and Cossuta) in preference to static linguistics which only takes ac-
count of lexical groups or semantic fields, and is content with drawing
up lists. In fact, philosophical discourse analysis privileges a prag-
matic and semiotic conception of discourse. The methodological frame-
work of which this type of approach is part therefore presents two
particularities: the first is using a linguistic and discursive model which
links linguistic operations creating discourse with the institutional
framework in which discourse is produced; and the second is taking
account of enunciative markers in order to study philosophical dis-
cursivity.
13 Ibid.
166 Malika Temmar
Language operations: a model of philosophical discursivity
Although in some respects philosophical discourse is comparable to

other discourses in the human sciences, one of its essential character-
istics is use of a discourse regime with claims to truth; more than any
other discourse, it also describes itself as having universal features.
This is why specialists in philosophical discursivity place importance
on philosophical enunciative operations uniting to deliver a univer-
sal message, with a claim to being generic [Ali-Bouacha, 1995].
The analysis of philosophical discourse attempts to explain the
coherence of a philosophical text insofar as it is suspended from the
presence of an enunciator who guarantees the enunciation; and through
him there is reference to the biographical and institutional dimensions.
Through a complex work of writing, the philosopher represents the
thought process within the text – the linearity of what Merleau-Ponty,
for example, calls the philosophical “path” of research, the philoso-
pher in the process of developing his philosophical itinerary. In philo-
sophical texts by Merleau-Ponty14, certain passages are ratified in
relation to the enunciating instance, in which case the locutor refers to
himself as “we” (rarely as “I”). The segments in which these persons
appear are interesting because they are very often accompanied by
time indicators like past, present and future, and they comprise tem-
poral organisers which ensure progression by assigning a “programme”
to the discourse. These indicators consist of many elements whose
series of links guarantees the unity of the philosophical discourse.
Like the path, the philosophical journey must lead to a port15, and so
the text appears as a progressive development logically sustained by
remarks and objections. Within the sentences of this sequence of text,
progress is ensured by the demonstrative ethos, the presence of logical
connectors (adverbs, coordinating conjunctions, verbal locutions)
14 More particularly in Le Visible et l’invisible and Phénoménologie de la percep-

tion.
15 Merleau-Ponty, Le Visible et l’invisible, p. 52.
which express and reinforce the structuring of ideas. These elements

determine the vigour of the argument and the continuity of the unify-
ing thread which gradually leads to a conclusion of the reasoning16.
The enunciator also introduces other points of view which may
be those of other enunciators, or even of other schools of thought,
thus creating polyphonic effects, setting up in correlation the figure
of a co-enunciator. The latter is a participant in the development of
the philosophical discourse, through his possible resistance (objec-
tions, prejudices or lack of understanding). Taking account of his
presence leads the enunciator to implement discursive strategies aim-
ing to “transform” the recipient. Thus, the philosophical co-enunciator
plays a role in the text’s construction. In the case of the phenomeno-
logical Merleau-Pontian text, it is through showing examples of ex-
periments which are inter-subjectively controllable that the philoso-
pher invites the reader to re-execute what is said (thought). The choice
of genre (meditation, dialogue, treatise) will more or less reinforce
the way the reader is inscribed in the text; this is most often represented
by frequent recourse to the generic “I” or possible re-enunciation
[Ali-Bouacha, 1995]. In the Méditations Métaphysiques, for exam-
ple, the persons used in the discourse do not offer to take responsibil-
ity for openly directing the reader in one way or another. Descartes
often chooses to use generic “I’s” or “false starts” which constantly
call upon the reader to participate through identification. He prefers
to play upon the formative power of the “I” in order to put his reader
in a position to be able to “carry out” what is said. The “I’s” that he
uses remain to be assumed by any “I” embodied by the reader17. The
reader thus participates in the inner structuring of the text, and this is
discernible in writing which combines didactic, pedagogical and pole-
mical functions.
Philosophy claims to deliver a universal message. But as Frédéric
Cossutta notes, the virtual universality won by a philosophy is al-
ways a partial universality (in fact):
16 Temmar, 2003.
17 Temmar, 2001.
168 Malika Temmar
It cannot make its inscription and historical dependency vanish other than through
the conceptual work by which it attempts to escape this contingency or, failing
that, (at the same time) by limiting it (containing it) – contingence(s) and
conceptualisation to which is linked the development of schools of thought.18
Hence the importance of “discursive strategies” which provide the

co-enunciator with access to the seat of philosophical practice: uni-
versality won by the institution of a doctrine.
Conclusion
The categories of philosophical discourse analysis, although indirectly

linked to epistemological and probably philosophical choices, aspire
to an autonomy which integrates them into the disciplinary field of
discourse analysis. Certainly, in return, these tools (of discourse analy-
sis) are only brought to bear insofar as they take account of the doc-
trines themselves. In fact, Frédéric Cossutta shows very clearly how
an analysis of philosophical writing on its own would be blind: “In the
final analysis, the model has no worth in itself, but in the way it casts
light on the text”. The requirements he gives for thinking about the
conditions that make philosophical discourse analysis possible are
evidence of this:
1 – To avoid analysing discursive operations (reductions) in a unilateral
way. In the case of fiction, for example, one cannot consider philo-
sophical fictions (Descartes’ “evil genius”, the statue of Condillac,
the meditative scene etc.) without taking into account the com-
plexity of their forms, or without thinking of them in relation to
the other fundamental discursive operations which structure philo-
sophical activity (like the initiatory dimension, didacticism, etc.).
18 Cossutta, 1995: 54.

2 – By refusing the “postulate of decontextualisation”, which leads to

reducing all identifications to a simple typology, philosophical
discourse analysis allies itself to a contextual study of observed
structures. It links them as much to “operations developed in
parallel as to those which connect them to a more global context;
from the micro-context, this broadens out towards the macro-
context formed by a short work or a complete work”19. Although
the study of philosophical discursivity is careful to ensure that its
method and the actual object of investigation closely intersect,
this does not mean that it ceases to place importance on discursive
phenomena which are interesting to consider in relation to other
discourses. Philosophical discourse is in many respects compara-
ble to other types of discourse and its analysis may therefore be
a valuable method of casting light on and understanding other
texts from which it is continually borrowing certain categories of
analysis. The study of argumentation, for example, is a basic
component of philosophical texts and may be of interest for other
texts in the human and social sciences. Through the establishment
of doctrines and the implementation of concepts, the philosophi-
cal text offers a prime example for studying how concepts emerge,
together with the way they are taken up and enter into circulation.
ANGERMÜLLER, J.
2004: “Discours et champs intellectuels: l’antagonisme entre ‘huma-
nistes’ et ‘prophètes’ et le discours des sciences humaines dans les
années 1960 et 1970”, in Amossy R. and Maingueneau M. (ed.),
L’analyse du discours dans les études littéraires. Actes du colloque de
Cerisy, Toulouse: Presses Universitaires du Mirail, pp. 83–92.
19 Cossutta, 1995: 34.

170 Malika Temmar
BOUACHA, M. A. and COSSUTTA, F. (ed.)

1998: La polémique en philosophie, Dijon, Presses Universitaires de
Dijon.
BOUACHA, M. A.
“De l’ego à la classe des locuteurs: lecture linguistique des Médita-
tions”, Langages, no. 119, Paris: Larousse, pp. 79–108.
BORDRON , J-F.
1987: Descartes: recherches sur les contraintes sémiotiques de la
pensée discursive de Descartes, Paris, PUF.
CICUREL, F.
1998: “Une mise en scène de la polémique: les dialogues philosophiques
contemporains”, in Magid Ali Bouacha and Frédéric Cossutta (ed.) La
polémique en philosophie, Dijon, Presses Universitaires de Dijon.
2004: “Le Lu et le lire ou l’espace de la lecture”, in J. M. Adam, M. A.
Bouacha, J. M. Grize (ed.), Textes et discours: catégories pour l’ana-
lyse, Dijon, EUD.
COSSUTTA, F.
1995: ed. “L’analyse du discours philosophique”, Langages, no. 119,
Paris, Larousse.
1996: ed. Descartes et l’argumentation philosophique, Paris, PUF.
1989: Éléments pour la lecture des textes philosophiques, Paris: Bordas.
1998: ed. Lire Bergson: “Le possible et le réel”, Paris: PUF.
LHOMME, A.
1998: “Formuler l’informulable: analyse d’un paradoxe pragmatique”,
in Cossutta, F. (ed.), Lire Bergson: “Le possible et le réel”. Paris:
PUF, pp. 101–137.
MAINGUENEAU, D. and PATRICK CH.

2002: Dictionnaire d’analyse du discours, Paris: Le Seuil.
MAINGUENEAU, D.
1995: “L’énonciation philosophique comme institution discursive”,
Langages no. 119, pp. 40–62.
1996: “Ethos et argumentation philosophique: le cas du ‘Discours de
la méthode’”, in Cossutta F. (ed.), Descartes et l’argumentation philo-
sophique, Paris: PUF, pp. 85–110.
PÊCHEUX, M.
1975: Les vérités de La Palice: linguistique sémantique, philosophie,
Paris, Maspero.
PHILIPPE, G.
1995: “Embrayage énonciatif et théorie de la conscience: à propos de
l’Etre et le néant”, Langages no. 119, pp. 95–108.
TEMMAR, M.
2001: “Les stratégies discursives d’ostension du morceau de cire de
Descartes”, in C. Gronemann, C. Maas, S. A. Peters, S. Schrader (ed.),
Körper und Schrift, Bonn: Bonn University Press, pp. 56–67.
2001: “Der Text als ‘Bedienungsanleitung’. Eine philosophische Dis-
kursanalyse der metaphysischen Meditationen von R. Descartes”, in
Diskursanalyse: Theorien, Methoden, Anwendungen, Hamburg: Ar-
gument, L.I.T, pp. 181–207.
2003: “Effets de presence et ostension dans le discours phénoméno-
logique merleau-pontyen: une approche discursive”, Studia Phaeno-
menologica, Romanian Journal for Phenomenology 3/4, Éditions
Humanitas Romanian Society for Phenomenology, Bucharest, pp. 163–
184.
2007: “Le discours philosophique au Carrefour des genres”, Le
Français aujourd’hui, no. 159.
2007: “La fiction dans le texte philosophique”, in B. Curatolo and
J. Poirier (ed.), Le Style des philosophes, Presses universitaires de
Dijon.
The authors
Simone BONNAFOUS, Professor of Information and Communication

Sciences, Université Paris-Est Créteil, foundator of Céditec
Malika TEMMAR, Senior Lecturer in Language Sciences, University

of Amiens, member of Céditec
Johannes ANGERMÜLLER, Professor of Discourse, University of War-

wick, UK, member of Céditec
Claire OGER , Senior Lecturer in Information and Communication

Sciences, Université Paris 13
Caroline OLLIVIER-YANIV, Professor in Information and Communi-

cation Sciences, Université Paris-Est Créteil, member of Céditec
Alice KRIEG-PLANQUE, Senior Lecturer in Information and Com-

munication Sciences, Université Paris-Est Créteil, member of Céditec
Juliette RENNES , Senior Lecturer in Politics and Information and

Communication Sciences, E.H.S.S. Paris
Pierre FIALA, Senior Lecturer in Language Sciences, Université Paris-

Est Créteil, member of Céditec
Dominique MAINGUENEAU, Professor of Language Sciences, Uni-

versité Paris 4, member of Céditec
Marie-Anne PAVEAU, Professor of Language Sciences, Université

Paris 13
Dominique DUCARD, Professor of Language Sciences, Université

Paris-Est Créteil, member of Céditec

Bonnafus Temmar Discourse Analysis Human and Social Sciences

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Bonnafus Temmar Discourse Analysis Human and Social Sciences

Caricato da

Copyright:

Formati disponibili

&

Simone Bonnafous is Professor of Media and Communications at the

Simone Bonnafous is Professor of Media and Communications at the

British Library Cataloguing-in-Publication Data: A catalogue record for this book is

Library of Congress Cataloging-in-Publication Data

Céditec et Institut Universitaire de France

Original title : Bonnafous Simone et Temmar Malika (eds), Analyse du discours et

ISBN 978-3-0343-1241-7 pb. ISBN 978-3-0352-0165-9 eBook

© Peter Lang AG, International Academic Publishers, Bern 2013

All rights reserved.

I. Discourse analysis in Europe

II. Discourse analysis and information and

III. Discourse analysis and comprehensive sociology.

IV. Multidisciplinary work on discourse.

V. Political discourse analysis: content analysis,

VI. Analysing a debate. The contributions

VII. Discourse analysis and the study of literature

VIII. Discourse analysis and history. Meeting and forgetting

IX. The role of psychoanalysis

X. Discourse analysis and philosophy:

The authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173

Simone BONNAFOUS and Malika TEMMAR

Discourse analysis became current in France in the late 1960s, thanks

increasing number of works on discourse analysis testifies to the vi-

3 An interdisciplinary journal which brings together researchers in language sci-

Céditec has in addition benefited from a wide variety of differ-

specific boundaries. It would also permit an overview of the way in

have helped re-direct language sciences back towards empirical data

– A component which explores avenues that have been relatively

– A methodological contribution. Although present in all the arti-

Discourse analysis in Europe

Since the late 1960s, a variety of trends in discourse analysis have

1 I thank Jacques Guilhaumou, Reiner Keller and Dominique Maingueneau, as

Three major theoretical trends

The “French” trend

The “Anglo-Saxon” trend

The “Anglo-Saxon” notion of discourse draws its inspiration from

The “German” trend

ever, although in Germany the Habermas paradigm has been in de-

These three trends, to which we should perhaps add Slavo-Russian

The major “clusters” in Europe

searchers together around a particular theoretical position. Today, these

The French school of discourse analysis

Discourse analysis was established in France in the late 1960s against

Researchers are often closely involved with political or trade union

Poststructuralist discourse theory

Another cluster based on the philosophy of Althusser was formed

describe the way in which, within an antagonistic space, hegemonies

Critical Discourse Analysis, CDA

Interpretive discourse analysis

A fourth group revolves around the German sociologist Reiner Keller

If in European discourse analysis at least three major historical trends

ALTHUSSER L., E. BALIBAR, R. ESTABLET, P. MACHEREY

BERGER P. L. and T. LUCKMANN

BROWN G. and G. YULE

HALL S., D. HOBSON, A. LOWE and P. WILLIS (ed.)

KELLER R., A. HIRSELAND, W. SCHNEIDER and W. VIEHÖVER (ed.)

LACLAU E. and C. MOUFFE

Discourse analysis and information

The meeting between discourse analysis (DA) and information and

1 This chapter is based on a paper given at the conference “Sciences du Langage et

An overdue meeting between DA and ICS

An abundant bibliography has been devoted to the conditions which

The theoretical groundings