Page i

Civil Society and Political Theory

Page ii
Studies in Contemporary German Social Thought (partial list)
Thomas McCarthy, General Editor
Theodor W. Adorno, Prisms
Seyla Benhabib and Fred Dallmayr, editors, The Communicative Ethics Controversv
Richard J. Bernstein, editor, Habermas and Modernitv
Ernst Bloch, Natural Law and Human Dignitv
Ernst Bloch, The Principle of Hope
Ernst Bloch, The Utopian Function of Art and Literature. Selected Essavs
Hans Blumenberg, The Genesis of the Copernican World
Hans Blumenberg, The Legitimacv of the Modern Age
Hans Blumenberg, Work on Mvth
Susan Buck-Morss, The Dialectics of Seeing. Walter Benfamin and the Arcades Project
Craig Calhoun, editor, Habermas and the Public Sphere
Jean Cohen and Andrew Arato, Civil Societv and Political Theorv
Helmut Dubiel, Theorv and Politics. Studies in the Development of Critical Theorv
John Forester, editor, Critical Theorv and Public Life
David Frisby, Fragments of Modernitv. Theories of Modernitv in the Work of Simmel, Kracauer and Benfamin
Jürgen Habermas, On the Logic of the Social Sciences
Jürgen Habermas, Moral Consciousness and Communicative Action
Jürgen Habermas, The New Conservatism. Cultural Criticism and the Historians´ Debate
Jürgen Habermas, The Philosophical Discourse of Modernitv. Twelve Lectures
Jürgen Habermas, Philosophical-Political Profiles
Jürgen Habermas, Postmetaphvsical Thinking. Philosophical Essavs
Jürgen Habermas, The Structural Transformation of the Public Sphere. An Inquirv into a Categorv of Bourgeois Societv
Axel Honneth, The Critique of Power. Reflective Stages in a Critical Social Theorv
Axel Honneth and Hans Joas, editors, Communicative Action. Essavs on Jùrgen Habermas´s
The Theory oI Communicative Action
Reinhart Koselleck, Critique and Crisis. Enlightenment and the Pathogenesis of Modern Societv
Reinhart Koselleck, Futures Past. On the Semantics of Historical Time
Harry Liebersohn, Fate and Utopia in German Sociologv, 18871923
Guy Oakes, Weber and Rickert. Concept Formation in the Cultural Sciences
Claus OIIe, Contradictions of the Welfare State
Claus OIIe, Disorgani:ed Capitalism. Contemporarv Transformations of Work and Politics
Joachim Ritter, Hegel and the French Revolution. Essavs on the Philosophy oI Right
AlIred Schmidt, Historv and Structure. An Essav on Hegelian-Marxist and Structuralist Theories of Historv
Dennis Schmidt, The Ubiquitv of the Finite. Hegel, Heidegger, and the Entitlements of Philosophv
Carl Schmitt, The Crisis of Parliamentarv Democracv
Carl Schmitt, Political Romanticism
Carl Schmitt, Political Theologv. Four Chapters on the Concept of Sovereigntv
Gary Smith, editor, On Walter Benfamin. Critical Essavs and Recollections
Michael Theunissen, The Other. Studies in the Social Ontologv of Husserl, Heidegger, Sartre, and Buber
Ernst Tugendhat, Self-Consciousness and Self-Determination
Mark Warren, Niet:sche and Political Thought
Albrecht Wellmer, The Persistence of Modernitv. Essavs on Aesthetics, Ethics and Postmodernism
Thomas E. Wren, editor, The Moral Domain. Essavs in the Ongoing Discussion between Philosophv and the Social Sciences
Lambert Zuidervaart, Adorno´s Aesthetic Theorv. The Redemption of Illusion

Page iii
Civil Society and Political Theory
Jean L. Cohen and Andrew Arato

Page iv
To Julian and Rachel
Fourth printing, 1997
First MIT Press paperback edition, 1994
©1992 Massachusetts Institute oI Technology
All rights reserved. No part oI this book may be reproduced in any Iorm or by any electronic or mechanical means (including inIormation storage and retrieval) without
permission in writing Irom the publisher.
This book was set in New Baskerville at MIT Press and was printed and bound in the United States of America.
Library oI Congress Cataloging-in-Publication Data
Cohen, Jean L.
Civil society and political theory / Jean L. Cohen and Andrew Arato
p. cm. (Studies in contemporary German social thought)
Includes bibliographical reIerences and index.
ISBN 0-262-03177-9 (H), 0-262-53121-6 (P)
1. Civil society. 2. Civil societyHistory. I. Arato, Andrew.
II. Title. III. Series.
JC336.C65 1990
306.2dc20 90-46723
CIP

Page v
CONTENTS
Preface vii
Acknowledgments xix
Introduction 1
I
The Discourse of Civil Society

1
The Contemporary Revival oI Civil Society
29
2
Conceptual History and Theoretical Synthesis
83
3
Theoretical Development in the Twentieth Century
117
II
The Discontents of Civil Society

4
The Normative Critique: Hannah Arendt
177
5
The Historicist Critique: Carl Schmitt, Reinhart Koselleck, and Jürgen
Habermas
201
6
The Genealogical Critique: Michel Foucault
255
7
The Systems-Theoretic Critique: Niklas Luhmann
299
III
The Reconstruction of Civil Society

8
Discourse Ethics and Civil Society
345
9
Social Theory and Civil Society
421
10
Social Movements and Civil Society
492
11
Civil Disobedience and Civil Society
564
Notes 605
Index 745

Page vii
PREFACE
This book is meant as a contribution to democratic theory. Unlike other approaches to the topic, however, ours does not Iocus directly on political institutions. Nor is
it restricted to the domain oI norma Iive political philosophy, although both institutions and philosophy have their place in the text. Our goal, rather, is twoIold: to
demonstrate the relevance oI the concept oI civil society to modern political theory and to develop at least the Iramework oI a theory oI civil society adequate to
contemporary conditions. In the process we hope to Iill a rather glaring lacuna in recent work in the Iield oI democratic theory. Every theory oI democracy
presupposes a model oI society, yet no one has addressed the question oI which type oI civil society is most appropriate to a modern democratic polity.
1
To put it
another way, the relation between normative models oI democracy or projects oI democratization and the structure, institutions, and dynamics oI civil society has
remained opaque, in part because there is no suIIiciently complex theory oI civil society available to us today. The task oI this book is to begin the construction oI such
a theory.
The concept oI civil society, in a variety oI uses and deIinitions, has become quite Iashionable today, thanks to struggles against communist and military dictatorships in
many parts oI the world. Yet it has an ambiguous status under liberal democracies. To some, it seems to indicate what the West has already achieved, and thus it is
without any apparent critical potential Ior examining the dysIunctions and injustices oI our type oI society. To others, the concept belongs to early modern Iorms oI
political philosophy that
1

Page viii
have become irrelevant to today's complex societies. It is our thesis, however, that the concept oI civil society indicates a terrain in the West that is endangered by the
logic oI administrative and economic mechanisms but is also the primary locus Ior the potential expansion oI democracy under "really existing" liberal-democratic
regimes. In advancing this thesis, we shall demonstrate the modernity and normative/critical relevance oI the concept oI civil society to all types oI contemporary
societies.
There are good arguments Ior each oI these three positions, and we shall address them in detail. We shall try to show that the Iirst two sets oI arguments draw their
strength Irom inadequate versions oI the concept that have been unreIlectively revived in the discussion so Iar in Latin America, Eastern Europe, and the West. One
common ambiguity concerns the relation between the terms "civil" and "bourgeois" society, a distinction that cannot even be made in German (bùrgerliche
Gesellschaft) or in some East European languages. This is not simply a terminological problem, Ior the program oI "civil society vs. the state," challenging statist
dictatorships that penetrate and control both the economy and the various domains oI independent social liIe, seems to stand Ior the autonomy oI both the civil and the
bourgeois. To be sure, the democratic movements in the East rely on the strengths oI new autonomous Iorms oI discourse, associations, and solidarity, i.e., on the
elements oI civil society. But they have not suIIiciently diIIerentiated between the task oI establishing viable market economies (whatever Iorm oI ownership replaces
state property and control), on the one hand, and the project oI strengthening civil society vis-a-vis the state and the liberated market Iorces, on the other. Yet, as we
know Irom the history oI the West, the spontaneous Iorces oI the capitalist market economy can represent as great a danger to social solidarity, social justice, and
even autonomy as the administrative power oI the modern state. Our point is that only a concept oI civil society that is properly diIIerentiated Irom the economy (and
thereIore Irom ''bourgeois society") could become the center oI a critical political and social theory in societies where the market economy has already developed, or
is in the process oI developing, its own autonomous logic. Otherwise, aIter successIul transitions Irom dictatorship to democracy, the undiIIerentiated
2
3
a civil society diferentiated from the
economy

Page ix
version oI the concept embedded in the slogan "society vs. the state" would lose its critical potential. Thus, only a reconstruction involving a three-part model
distinguishing civil society Irom both state and economy has a chance both to underwrite the dramatic oppositional role oI this concept under authoritarian regimes and
to renew its critical potential under liberal democracies.
Let us start with a working deIinition. We understand "civil society"
2
as a sphere oI social interaction between economy and state, composed above all oI the intimate
sphere (especially the Iamily), the sphere oI associations (especially voluntary associations), social movements, and Iorms oI public communication. Modern civil
society is created through Iorms oI selI-constitution and selI-mobilization. It is institutionalized and generalized through laws, and especially subjective rights, that
stabilize social diIIerentiation. While the selI-creative and institutionalized dimensions
3
can exist separately, in the long term both independent action and
institutionalization are necessary Ior the reproduction oI civil society.
It would be misleading to identiIy civil society with all oI social liIe outside the administrative state and economic processes in the narrow sense. First, it is necessary
and meaningIul to distinguish civil society Irom both a political society oI parties, political organizations, and political publics (in particular, parliaments) and an
economic society composed oI organizations oI production and distribution, usually Iirms, cooperatives, partnerships, and so on. Political and economic society
generally arise Irom civil society, share some oI its Iorms oI organization and communication, and are institutionalized through rights (political rights and property rights
especially) continuous with the Iabric oI rights that secure modern civil society. But the actors oI political and economic society are directly involved with state power
and economic production, which they seek to control and manage. They cannot aIIord to subordinate strategic and instrumental criteria to the patterns oI normative
integration and open-ended communication characteristic oI civil society. Even the public sphere oI political society rooted in parliaments involves important Iormal and
temporal constraints on processes oI communication. The political role oI civil society in turn is not directly related to the control or
a 3-parts model State/economy/civil society

Page x
conquest oI power but to the generation oI inIluence through the liIe oI democratic associations and unconstrained discussion in the cultural public sphere. Such a
political role is inevitably diIIuse and ineIIicient. Thus the mediating role oI political society between civil society and state is indispensable, but so is the rootedness oI
political society in civil society. In principle, similar considerations pertain to the relationship between civil and economic society, even iI historically, under capitalism,
economic society has been more successIully insulated Irom the inIluence oI civil society than political society has been, despite the claims oI elite theories oI
democracy. Nevertheless, the legalization oI trade unions, collective bargaining, codetermination, and so on witness the inIluence oI civil on economic society and
allow the latter to play a mediating role between civil society and the market system.
Second, the diIIerentiation oI civil society Irom both economic and political society seems to suggest that the category should somehow include and reIer to all the
phenomena oI society that are not directly linked to the state and the economy. But this is the case only to the extent that we Iocus on relations oI conscious
association, oI selI-organization and organized communication. Civil society in Iact represents only a dimension oI the sociological world oI norms, roles, practices,
relationships, competencies, and Iorms oI dependence or a particular angle oI looking at this world Irom the point oI view oI conscious association building and
associational liIe. A way to account Ior this limitation in the scope oI the concept is to distinguish it Irom a sociocultural liIeworld, which as the wider category oI "the
social" includes civil society. Accordingly, civil society reIers to the structures oI socialization, association, and organized Iorms oI communication oI the liIeworld to the
extent that these are institutionalized or are in the process oI being institutionalized.
Finally, we want to stress that under liberal democracies, it would be a mistake to see civil society in opposition to the economy and state by deIinition. Our notions oI
economic and political society (which admittedly complicate our three-part model) reIer to mediating spheres through which civil society can gain inIluence over
political-administrative and economic processes. An antagonistic relation oI civil society, or its actors, to the economy or the state
vision libérale totale...
exit les classes sociales.

Page xi
arises only when these mediations Iail or when the institutions oI economic and political society serve to insulate decision making and decision makers Irom the
inIluence oI social organizations, initiatives, and Iorms oI public discussion.
The Structure of This Book
We shall argue that what is at stake in the debates animating political and social theory in both East and West
4
is not simply the deIense oI society against the state or
the economy but which version oI civil society is to prevail. There is, however, another issue underlying these debates. Max Weber's disillusioned insistence that we
moderns are living in an age oI disenchantment appears to be more true now than ever beIore. Secular political utopias seem to have gone the way oI the great
mobilizing religious world views oI the previous age. The demise oI the most important radical-democratic and socialist utopia oI our time, Marxism, has already led
thinkers to proclaim the end oI history and the worldwide triumph oI a rather uninspired version oI liberalism. Now that the revolutionary rhetoric oI communism has at
last (and deservedly) been discredited, the question conIronting political theorists is whether utopian thought and corresponding radical political projects are
conceivable at all. Or are the mobilizing ideals embedded in earlier utopias consigned to the dustbin oI the history oI ideas?
The great ideals generated in the age oI democratic revolutionsliberty, political and social equality, solidarity, and justicewere each embedded in totalistic and
mutually exclusive utopias: anarchism, libertarianism, radical democracy, Marxism. Sober reIlection on the history oI the past century and a halI should dissuade
responsible persons Irom seeking to revive any one oI these utopias in their original Iorm. However, a society without action-orienting norms, a society without political
projects, is equally undesirable, Ior the civil privatism or "realism" that would result would really be just another name Ior egoism, and the corresponding political
culture would lack suIIicient motivation to maintain, much less expand, existing rights, democratic institutions, social solidarity, or justice.

Page xii
It is our thesis that the revival oI the discourse oI civil society provides some hope in this regard. For this discourse reveals that collective actors and sympathetic
theorists are still oriented by the utopian ideals oI modernitythe ideas oI basic rights, liberty, equality, democracy, solidarity, and justiceeven iI the Iundamentalist,
revolutionary rhetoric within which these ideals had once been articulated is on the wane. Indeed, civil society itselI has emerged as a new kind oI utopia, one we call
"selI-limiting," a utopia that includes a range oI complementary Iorms oI democracy and a complex set oI civil, social, and political rights that must be compatible with
the modern diIIerentiation oI society. It is this utopian ideal that plays a Iundamental, iI regulative, role in the construction oI our book as a whole, as well as in its
individual parts.
Parts I and II analyze the major theories and criticisms oI the concept oI civil society that have emerged in the nineteenth and twentieth centuries. In the introduction we
present an overview oI the theoretical importance oI the problem oI civil society by situating it in terms oI three central debates in contemporary political theory:
between elite and participatory democracy, between liberalism and communitarianism, and between critics and deIenders oI the welIare state. For the most part, this
discussion draws on American sources. Our intention here is not to prove that the concept oI civil society can resolve all the relevant debates and antinomies but rather
to show that it opens up new and unexpected possibilities Ior synthesis in each case.
But which concept? Bracketing the working deIinition just provided, chapter 1 introduces the concept oI civil society in a deliberately nonsystematic manner, by
reproducing its heterogeneous current usage by intellectuals in or close to a number oI social and political movements. Since our interest is in politics, we believe we
must Iirst learn Irom contemporary discourses in order to contribute something to them. We start our examination oI the political motivations relevant to our task with a
presentation oI Iour ideal typical discourses: Polish (the democratic opposition), French (the "second leIt"), German (the realist Greens) and Latin American (the new
democratic leIt). In each case, the concept and categories oI civil society have become central to eIIorts to articulate normative projects Ior liberalization and
democratization. We do not assume

Page xiii
that the discourses we reproduce are Iully representative oI what is available, and even less that they in themselves can provide or substitute Ior a political analysis oI
the Iour contexts. Only in the Eastern European case do we return to analyzing, this time on the basis oI a variety oI primary and secondary sources, the Iate oI the
intellectual project in the Iace oI complex constraints. We complete this part oI our analysis by comparing and contrasting the Iour diIIerent discourses oI civil society,
and only then do we raise the question oI whether a uniIied conception oI civil society with a critical thrust can be developed Irom intellectual contexts related to
contemporary Iorms oI action. The chapter shows how these heterogeneous and unsystematic attempts diIIer, what they have in common, and why it makes sense to
link them together.
The revival oI concepts oI civil society notwithstanding, it could be argued that twentieth-century developments render key dimensions oI the concept irrelevant. The
norms oI civil societyindividual rights, privacy, voluntary association, Iormal legality, plurality, publicity, Iree enterprisewere, oI course, institutionalized
heterogeneously and in a contradictory manner in Western societies. The logic oI capitalist private property and the market oIten conIlicts with plurality and Iree
association; that oI bureaucratization, with parliamentary will-Iormation. The principles oI a representative, inclusive, political process oI legislation controlled by
society conIlicts with new Iorms oI exclusion and domination in society, in the economy, and in the state. Moreover, given structural changes over the last century, any
attempt to equate "state" with "the political" or "civil society'' with "the private" seems anachronistic. II this is so, can a category oI early modern political philosophy
have any continuing relevance to the contemporary world?
In chapter 2, we present a short conceptual history oI early modern versions oI civil society and a theoretical analysis oI Hegel's masterIul synthesis. These steps
belong to what we take to be the necessary prolegomenon Ior a theory oI civil society on the level oI the history oI theory. Indeed no one could seriously contest
Hegel's position as the most important nineteenth-century predecessor oI and inspiration to twentieth-century analyses oI civil society. The categorial richness oI the
concept oI civil society can be recovered only through an analysis oI Hegel's Iramework, which gathered

Page xiv
into itselI all available interpretations oI the concept. We cannot, oI course, pretend to examine the evolution oI Hegel's political philosophy, the whole body oI his
relevant work, or even the Iull range oI secondary literature dealing with the most important text Ior us, The Philosophv of Right. Nevertheless, the Hegelian theory is
crucial because it reconstructs civil society in terms oI the three levels oI legality, plurality and association, and publicity and because Hegel sees the link between civil
society and state in terms oI mediation and interpenetration. As chapter 1 shows, no contemporary discourse oI civil society has managed to add even a single
Iundamental category to legalitv, privacv, pluralitv, association, publicitv, and mediation, except that oI social movements, and the most sophisticated
contemporary authorsMichnik, O'Donnell, and Cardoso, Ior examplework with all oI these levels.
Hegel's own ambiguities concerning civil society, and perhaps even his recurring statism in Iace oI the alienation oI the system oI needs, can be traced to his inclusion oI
the economy as one oI the levels oI civil society. The importance oI Gramsci and Parsons Ior our Iramework is their demonstration that the basic Hegelian conception
can be improved by introducing a three-part model diIIerentiating civil society Irom both economy and state. We argue in chapter 3, however, that both Gramsci's and
Parsons's analyses suIIer Irom the Iact that they introduce these three domains in terms oI overly monistic and Iunctionalistic Iorms oI theory. In Gramsci's case, this led
to a deep ambivalence toward modern civil society and its Iuture in a Iree socialist society. In Parsons's case, on the other hand, the Ilat combination oI normative and
Iunctionalist approaches leaves us with an explicitly apologetic theory oI the contemporary American version oI civil society. We wish to make the reader sensitive to
the dangers oI both versions oI Iunctionalism.
Together, the Iirst three chapters show that the concept oI civil society continues to inIorm major paradigms oI contemporary social and political theory. Chapter 3 in
particular shows that the theoretical aims oI Hegel's synthesis are better served iI we abandon his own statist bias and iI we diIIerentiate civil society Irom the system oI
needs more sharply than he did. Gramsci and Parsons thus point beyond economism and statism within the terms oI Hegelian political philosophy.

Page xv
Twentieth-century usages oI the concept oI civil society are not without their critics. Indeed, many have argued that the concept oI civil society is anachronistic,
normatively suspect, or both. Accordingly, in part II, we reproduce and assess Iour Iundamental types oI criticism to which we believe all currently available
conceptions oI civil society are more or less vulnerable. To be sure, there would be other ways oI schematizing critical approaches, and other critics to includeno
analysis can avoid selectivity. We have chosen to divide critical perspectives according to Iour models: the normative (chapter 4), the historicist (chapter 5), the
genealogical (chapter 6), and the systems-theoretic (chapter 7). With the exception oI the historicist model, where we reIer to three authors, each approach is typiIied
by a single theorist. We use this procedure to produce as coherent a case Ior each perspective as possible. For the same reason we bring to bear our own criticisms
immanently in each instance, leaving our own position until later. As we proceed, though, we note that several oI the critics have reconstructed one dimension oI the
classical concept oI civil society as inherited Irom Hegel even as they battled against the conception as a whole. In addition, each critic has helped to weaken the case
oI at least one oI the others. This was the case Ior Arendt's notion oI the public sphere as a genuinely political concept (vs. Schmitt), Ior Habermas's rediscovery oI the
biIurcation oI the public in a model oI mediation (vs. Arendt), Ior Foucault's genealogy oI modern power relations (vs. all Iunctionalist models), and Ior Luhmann's
notion oI diIIerentiation (vs. Schmitt and Habermas).
Part III is more systematic and less expository than the Iirst two parts. Keeping in mind the diIIiculties that have emerged Irom contemporary political discussions and
Irom the Iour types oI criticism oI the concept oI civil society, we have produced Iour theoretical studies. These are meant to respond to the most important objections
leIt over Irom the critical conIrontation oI the critics with each other, to outline a reconstructed theory oI civil society, and to reconnect this theory to politics through
analyses oI social movements and civil disobedience.
Chapter 8 starts to work out the normative Ioundations oI a theory oI civil society, using the discourse ethics developed by Habermas and his colleagues. The
presentation oI the discourse ethics has a double Iunction. First, it responds to the normative and

Page xvi
genealogical critics by showing how a convincing justiIication Ior civil society can be provided today. Second, it shows that the project oI the institutionalization oI
discourses is possible only on the ground oI modern civil society. It is in this context that we hope to give a more comprehensive solution to the antinomy oI
rightsoriented liberalism and communitarianism discussed in the introduction, taking into account the claims oI participatory democratic theory as well. The thesis oI
chapter 8 is that the plausibility oI rights and democracy depends on their conceptual and normative interrelation despite the apparently antithetical character oI the two
theoretical paradigms in which each is articulated and deIended.
Because every normative theory oI democracy, and every liberal theory, implies a model oI society, it is incumbent on political theorists to add the dimension oI social-
structural analysis to norma Iive political philosophy. OI course, those who are convinced oI the universality oI hermeneutic methodology would need do no more to
demonstrate the validity oI contemporary theoretical uses oI the concept oI civil society than reconstruct the contemporary discourses oI civil society in a normatively
coherent theory. On such a view, the Iact that the concept oI civil society inIorms the selI-understanding oI social movements is enough to show that it remains an
adequate basis Ior the symbolic orientation oI collective action. But the "discourse oI civil society," including even the best philosophical reIormulation oI it, could be
merely ideological. Whatever the intentions oI social actors, the Iunctional requirements oI modern economic and political systems may make projects based on the
concept irrelevant, corresponding identities unstable, interpretations one-sided. Given the challenges to the very model oI diIIerentiation that is at the heart oI the
discourse oI civil society, it is essential to provide a systematic reconstruction oI its structural presuppositions. Without a social-scientiIic analysis oI the structure and
dynamics oI modern society, we have no way oI evaluating the generality oI a given identity or the global constraints operating behind the back oI social actors.
In addition, the relation between civil society, the economy, and the state requires elaboration. This is the goal oI chapter 9, which starts by mapping out the three-part
model oI civil society intro-

Page xvii
duced by Gramsci in terms oI the Habermasian distinction between the liIeworld and the economic and political subsystems. We then attempt to demonstrate the
modernity oI this construct. Chapter 9 must be read as a sympathetic revision oI the Habermasian Iramework. Our main addition is to integrate the concept oI civil
society into the overall model, making the necessary adjustments. Convinced that the theory oI communicative action represents the most advanced contours oI critical
social theory today, we try to unIold the implications oI the whole conception on the level oI political theory. Indeed, our reconstruction oI civil society should be seen
also as a political "translation" oI Habermasian critical theory, one that has been guided by the dramatic struggles oI our time under the aegis oI his own values and
ours: Ireedom and solidarity. We argue, against Luhmann, that a model oI diIIerentiation and modernization cannot do without an ultimately cultural substratum, where
the rationalization oI normative action coordination occurs. We also show that our model has the advantage oI being able to accommodate the negative phenomena
associated with modern civil society in the genealogical criticism, and much more. We discuss the contradictory institutionalization oI the norms oI civil society while
insisting both on the utopian implications oI the model and on its alternative Iorms oI development. Chapter 9 concludes by outlining a proposal, based on the three-
part model, Ior the reIlexive continuation oI both the welIare state and the democratic revolution.
The last two chapters Iormulate this politics by reIerring to social movements and to one oI their key Iorms oI contestation: civil disobedience. We do not wish to imply
that the politics oI civil society can only take the Iorm oI social movements. Normal institutional Iorms oI political participationvoting, becoming active in political
parties, Iorming interest or lobby groupsare part oI this politics. But the utopian dimension oI radical politics can be Iound only on the level oI collective action. Thus,
in chapter 10, we address the relation between collective action and civil society Irom a slightly diIIerent point oI view than that oI chapter 1. Instead oI Iocusing on the
discourse oI activists, we take up the major theoretical paradigms that have evolved since the 1960s in order to analyze social movements and show that they each
pre-

Page xviii
suppose (in some cases implicitly, in others explicitly) the concept oI civil society. Moreover, we demonstrate that civil society, beyond all Iunctionalist and pluralist
models, should be seen not only passively, as a network oI institutions, but also actively, as the context and product oI selI-constituting collective actors. We then try
to demonstrate that our tripartite structural model is the best Iramework with which to approach "new" and old Iorms oI collective action.
We conclude by reIlecting on the question oI what is and what should and could be the relationships among societal plurality, individual autonomy, social movements,
and a liberal, democratic political system. Social movements are not always internally democratic, and they oIten engage in action that violates the democratic
procedures or laws generated by a nonetheless legitimate political order. What mode oI political voice, action, and representation is legitimate Ior social actors in both
society and state? What is the proper locus oI political activity and how should boundary lines between public and private be drawn? How can the danger oI
permanent mobilization be avoided? Our discussion oI civil disobedience in chapter 11 responds to these questions. Above all, our argument on civil disobedience
seeks to demonstrate that social movements and citizen initiatives are capable oI inIluencing policy and molding political culture without entry into the Iield oI power
politics and without necessarily endangering liberal or democratic institutions. Thus (implicitly returning to the Iirst debate in our introduction), we provide Ior a
Iramework oI democratization in the contexts oI elite democracies, without Ialling into the traps oI Iundamentalist theories oI participation. We also take up once more
the debate between rights-oriented liberals and participatory democrats, this time Irom the perspective oI the appropriate Iorms oI noninstitutionalized politics oI civil
society. We hope to provide, iI not the solution to the antinomies oI contemporary political and social theory, then at least to a way to begin rethinking them.

Page xix
ACKNOWLEDGMENTS
Each chapter oI this book was extensively discussed by the authors beIore draIts were written. The preIace, introduction, and chapters 6, 8, 10, and 11 are primarily
the work oI Jean L. Cohen; chapters 1, 2, 3, and 7 are primarily the work oI Andrew Arato; chapters 4, 5, and 9 are collaborative eIIorts.
We received support Ior this project, individually and jointly, Irom more people and institutions than we can mention here. We start with our separate
acknowledgments.
Jean L. Cohen would like to thank the Russell Sage Foundation Ior the intellectual and institutional support received there while I was a Post Doctoral Fellow in
Residence in 198687. I would also like to thank the Department oI Political Science at Columbia University Ior a leave oI absence that enabled me to carry on this
research. Special thanks to the Councils Ior Research in the Humanities and Social Sciences oI Columbia University Ior summer Iellowships in 1987 and 1988 that
enabled me to conduct research abroad. The Ecole des hautes etudes en sciences sociales, and in particular Claude LeIort and Pierre Rosanvallon, deserve special
mention Ior allowing me to work as Directeur d'etudes associe in political and social theory in 1989. While in Paris, I presented several lectures based on the book
and received very helpIul criticisms. My deepest thanks to Jürgen Habermas Ior sponsoring a two-month research Iellowship at the Max Planck Institut Iür
SozialwissenschaIten in Starnberg in 1981; there I was able to Iamiliarize myselI with his recent work, which has had the greatest inIluence on my thinking and on this
book. I would like to thank

Page xx
the American Council oI Learned Societies Ior Iunding my travel to Dubrovnik, Yugoslavia, in 1984 and 1985, to present lectures at the Course on Philosophy and
Social Science. There, too, I aired my ideas on social movements, discourse ethics, and civil society and received invaluable Ieedback. Finally, thanks are due to the
Vienna Institut Iür die WissenchaIten vom Menschen, and in particular to KrzysztoI Michalski and Cornelia Klinger, who invited me to lecture at their summer school
program in Cortona, Italy, in 1989 and 1990. There I gave seminars on the theme oI civil society to an interesting group oI graduate students Irom the United States,
Eastern Europe, and the Soviet Union, and I proIited greatly by their responses.
Andrew Arato would like to thank the Alexander von Humboldt StiItung Ior its support in 19801981, and the Max Planck Institut Iür SozialwissenschaIten in
Starnberg Ior providing a work base during this time. I very much appreciate the help I received Irom ProIessor Habermas, then director oI the institute, and his
colleagues in getting to know the Iramework oI the Theory oI Communicative Action, which is oIten utilized in this book. I would Iurther like to thank colleagues at the
Institute oI Sociology oI the Hungarian Academy oI Science, Ior their interest in my work on civil society and Ior the many interesting discussions we have shared. I
owe a great debt to the Graduate Faculty seminars oI the New School Ior Social Research at which I had a chance to discuss topics Irom this bookin particular, the
democracy seminar, the staII seminar in sociology, and the philosophy colloquium. ConIerences at the Cardozo Law School on Hegel and Luhmann have provided
excellent opportunities to reIine my ideas. I wish, Iinally, to thank all my students who have so actively participated in courses related to the problems oI civil society.
Many Iriends and colleagues have given us useIul critiques oI draIts oI the manuscript and interesting suggestions in the course oI conversations. We would like to
mention, in particular, Ken Baynes, Robert Bellah, Seyla Benhabib, György Bence, Laszlo Bruszt, Jose Casanova, Cornelius Castoriadis, Juan Corradi, Drucilla
Cornell, Ferenc Feher, Carlos Foreman, Alessandro Ferrara, JeIIrey GoldIarb, Claus Guenter, Jürgen Habermas, Elemer Hankiss, Agnes Heller, Dick Howard,
George Kateb, Janos Kis, György Markus,

Page xxi
Maria Markus, Alberto Melucci, Sigrid Meuschel, Claus OIIe, Guillermo O'Donnell, Alessandro Pizzorno, Carla Pasquinelli, Ulrich Preuss, Zbigniew Pelczynski,
Pierre Rosenvallon, Bernhardt Schlink, Phillippe Schmitter, AlIred Stepan, Ivan Szelenyi, Mihaly Vajda, JeIIrey Weintraub, and Albrecht Wellmer.
We owe a special word oI thanks to our series editor, Thomas McCarthy, as well as to our editor at MIT Press, Larry Cohen. Without their help this book would
certainly not have been possible.
We dedicate the book to our children, Julian Cohen Arato and Rachel Arato.

Page 1
INTRODUCTION
We are on the threshold oI yet another great transIormation oI the selI-understanding oI modern societies. There have been many attempts Irom various points oI view
to label this process: the ambiguous terms "postindustrial" and "postmodern" society reIlect the vantage points oI economic and cultural concerns. Our interest is in
politics. But Irom this standpoint, the changes occurring in political culture and social conIlicts are poorly characterized by terms whose preIix implies "aIter" or
"beyond." To be sure, Ior a variety oI empirical and theoretical reasons the old hegemonic paradigms have disintegrated, as have the certainties and guarantees that
went with them. Indeed we are in the midst oI a remarkable revival oI political and social thought that has been going on Ior the last two decades.
One response to the collapse oI the two dominant paradigms oI the previous periodpluralism and neo-Marxismhas been the attempt to revive political theory by
"bringing the state back in." While this approach has led to interesting theoretical and empirical analyses, its state-centered perspective has obscured an important
dimension oI what is new in the political debates and in the stakes oI social contestation.
1
The Iocus on the state is a useIul antidote to the reductionist Iunctionalism oI
many neo-Marxian and pluralist paradigms that would make the political system an extension, reIlex, or Iunctional organ oI economic (class) or social (group)
structures oI selectivity and domination. In this respect the theoretical move served the cause oI a more diIIerentiated analysis. But with respect to all that is nonstate,
the new paradigm continues the

Page 2
reductionist tendency oI Marxism and neo-Marxism by identiIying class relations and interests as the key to contemporary Iorms oI collective action. Moreover, the
legal, associational, cultural, and public spheres oI society have no theoretical place in this analysis. It thereby loses sight oI a great deal oI interesting and normatively
instructive Iorms oI social conIlict today.
The current ''discourse oI civil society," on the other hand, Iocuses precisely on new, generally non-class-based Iorms oI collective action oriented and linked to the
legal, associational, and public institutions oI society. These are diIIerentiated not only Irom the state but also Irom the capitalist market economy. Although we cannot
leave the state and the economy out oI consideration iI we are to understand the dramatic changes occurring in Latin America and Eastern Europe in particular, the
concept oI civil society is indispensable iI we are to understand the stakes oI these "transitions to democracy" as well as the selI-understanding oI the relevant actors. It
is also indispensable to any analysis that seeks to grasp the import oI such changes Ior the West, as well as indigenous contemporary Iorms and stakes oI conIlict. In
order to discover, aIter the demise oI Marxism, iI not a common normative project between the "transitions" and radical social initiatives under established liberal
democracies, then at least the conditions oI possibility oI IruitIul dialogue between them, we must inquire into the meaning and possible shapes oI the concept oI civil
society.
Admittedly, our inclination is to posit a common normative project, and in this sense we are post-Marxist. In other words, we locate the pluralist core oI our project
within the universalistic horizon oI critical theory rather than within the relativistic one oI deconstruction. At issue is not only an arbitrary theoretical choice. We are truly
impressed by the importance in East Europe and Latin America, as well as in the advanced capitalist democracies, oI the struggle Ior rights and their expansion, oI the
establishment oI grass roots associations and initiatives and the ever renewed construction oI institutions and Iorums oI critical publics. No interpretation can do these
aspirations justice without recognizing both common orientations that transcend geography and even social-political systems and a common normative Iabric linking
rights, associations, and publics together. We believe that civil society, in

Page 3
Iact the major category oI many oI the relevant actors and their advocates Irom Russia to Chile, and Irom France to Poland, is the best hermeneutic key to these two
complexes oI commonality.
Thus we are convinced that the recent reemergence oI the "discourse oI civil society" is at the heart oI a sea change in contemporary political culture.
2
Despite the
proliIeration oI this "discourse" and oI the concept itselI, however, no one has developed a systematic theorv oI civil society. This book is an eIIort to begin doing just
that. Nevertheless, systematic theory cannot be built directly out oI the selI-understanding oI actors, who may very much need the results oI a more distanced and
critical examination oI the possibilities and constraints oI action. Such theory must be internally related to the development oI relevant theoretical debates. At Iirst sight
the building oI a theory oI civil society seems to be hampered by the Iact that the stakes oI contemporary debates in political theory seem to be located around
diIIerent axes than the nineteenth-century couplet oI society and state. It is our belieI, however, that the problem oI civil society and its democratization is latently
present in these discussions and that it constitutes the theoretical terrain on which their internal antinomies might be resolved.
Three debates oI the last 1520 years seem to tower above all the rest. The Iirst continues an older controversy within the Iield oI democratic theory between
deIenders oI elite vs. participatory models oI democracy.
3
The second, Ior the most part restricted to the Anglo-American world, is between what has come to be
called "rights-oriented liberalism" and "communitarianism." While it covers some oI the same ground as the Iirst controversy, the terms oI the second discussion are
quite distinct Ior, unlike the Iirst, it occurs within the Iield oI normative political philosophy rather than between empiricists and normativists.
4
The third debate, pitting
neoconservative advocates oI the Iree market against deIenders oI the welIare state, has animated discussion on both sides oI the Atlantic.
5
Its context is, oI course,
the notorious crisis oI the welIare state that intruded on political consciousness in the mid-1970s. These debates are interrelated, and, as already indicated, there are
overlaps. Nevertheless, each oI them has culminated in a distinct set oI antinomies leading to a kind oI standoII and increas-

Page 4
ing sterility. What no one seems to have realized, however, is that the relatively unsystematic and heterogeneous discourse oI the revival oI civil society can be brought
to bear on these debates and indeed can provide a way out oI the antinomies that plague them. Accordingly, we shall brieIly summarize these debates in this
introduction and show how our book provides a new paradigm Ior thinking about the issues they raise.
Debates in Contemporary Political Theory
Elite Vs. Participatory Democracy
It would not be an exaggeration to say that the debate between elite and participatory models oI democracy has been going around in circles ever since Schumpeter
threw down the gauntlet to the normativists in 1942.
6
Schumpeter's claim that "the democratic method is that institutional arrangement Ior arriving at political decisions
in which individuals acquire the power to decide via a competitive struggle Ior the people's vote"
7
has Iormed the core oI the elite model oI democracy ever since.
Democracy is deIined not as a kind oI society or as a set oI moral ends or even as a principle oI legitimacy but rather as a method Ior choosing political leaders and
organizing governments. The elite model oI democracy claims to be realistic, descriptive, empirically accurate, and the only model that is appropriate to modern social
conditions.
Far Irom indulging in utopian illusions about the possibility oI either conjuring away the phenomenon oI power or the gap between rulers and ruled, this approach
assumes that no society, and certainly no modern one, could Iunction without both. A "realistic" appraisal oI democratic societies must grant that the motor oI the
political system is power just as the motor oI the economy is proIit. The struggle to acquire and use power is at the heart oI the political. What distinguishes democratic
Irom nondemocratic societies is thus the way in which power is acquired and decisions are arrived at: So long as some core set oI civil rights is respected and regularly
contested elections are held on the basis oI a universal Iranchise, so long as alternation in power is accepted by elites and occurs smoothly without violence or
institutional

Page 5
discontinuity, so long as decision making involves compromises among elites and (passive) acceptance by the population, a polity can be considered democratic. The
main concern here is obviously with the ability oI a government to produce decisions, to have them accepted, and to ensure orderly transitions, i.e., stability.
The elite model oI democracy prides itselI on providing an operationalizable and empirically descriptive account oI the practices oI polities considered to be
democratic. There is no pretense here that voters either set the political agenda or make political decisions; they neither generate issues nor choose policies. Rather,
leaders (political parties) aggregate interests and decide which are to become politically salient.
8
Moreover, they select issues and structure public opinion. The true
Iunction oI the vote is simply to choose among the bids Ior power by political elites and to accept leadership. The voters are consumers, the parties are entrepreneurs
oIIering alternative packages or personnel; it is they who create demand, bowing to consumer sovereignty only with regard to the yes/no decision by the voters about
who among the preselected candidates will be their "representatives" (using the latter term very loosely indeed).
9
In short, the empirical theories oI democracy (elite,
pluralist, corporatist, and rational choice models) tend quite openly to reduce the normative meaning oI the term to a set oI minimums modeled on a conception oI
bargaining, competition, access, and accountability derived more Irom the market than Irom earlier models oI citizenship.
Competitiveness in acquiring political power and in making policy decisions is, oI course, the core oI this model oI democracy. The competitive element is deemed to
be the source oI creativity, productivity, responsibility, and responsiveness. The ultimate sanction oI the vote, together with the necessity on the part oI elites to
compete Ior it, will supposedly keep things Iair, encourage authorities to be responsive to a multiplicity oI demands and accountable to the citizenry, and Ioster their
willingness to compromise with one another. To be sure, this model oI democracy rests on certain preconditions that it supposedly should be able to reproduce: high-
quality leadership with a tolerance Ior diIIerences oI opinion, a restricted range oI political decision,
10
and an elite political culture based on democratic selI-
control.
11
These pre-

Page 6
conditions are predicated in turn on the Iact oI social pluralism or cleavage, which the democratic method institutionalizes into non-violent competition Ior oIIice and
inIluence. A Iinal precondition, deemed indispensable Ior a stable political system to be able to make decisions, is that it must be shielded Irom too much participation
by the population: Citizens must, as it were, accept the division oI labor between themselves and the politicians they elect.
12
Accordingly, this model oI democracy
argues that the secret ballot, civil rights, alternation, regular elections, and party competition are central to every modern conception oI democracy iI democracy is to
have any place at all in complex modern societies.
We Iind this last statement to be quite convincing, so Iar as it goes. But the normativist critique oI the elite model oI democracy is also convincing. It is especially
compelling against the elite model's tendency to extol apathy, civil privatism, and the necessity to shield the political system Irom "excess" demands oI the population as
democratic principles, the meaning oI this excess to be determined by the elites alone.
13
The normativists correctly point out that what makes Ior stability and
continuity in a polity is not identical with what makes it democratic. From the standpoint oI participation theory, the elite model oI democracy is both too broad and
too narrow. To deIine a polity as democratic iI it periodically holds contested elections and guarantees civil rights, regardless oI what sorts oI public institutions or
private arrangements exist, is to extend democratic legitimacy to an enormously wide range oI societies while simultaneously shielding them Irom critical scrutiny.
14
At
the same time the concept oI democracy at play here is too narrow, Ior it is deIined by procedures that have little to do with the procedures and presuppositions oI
Iree agreement and discursive will Iormation.
15
Indeed the participation theorists argue that the "realistic" model has denuded the concept oI democracy oI so many oI
its elements that it has lost any connection with its past meaning.
16
What is leIt iI one drops the ideas oI selI-determination, participation, political equality, discursive
processes oI political will Iormation among peers, and the inIluence oI autonomous public opinion on decision making? In short, the price oI the elite model's realism is
the loss oI what has always been taken to be the core oI the concept oI democracy, namely, the citizenship principle.

Page 7
Moreover, by restricting the concept oI democracy to a method oI leader selection and to procedures regulating the competition and policy making oI elites, this
model sacriIices the very principles oI democratic legitimacy on which it is nevertheless parasitic. It loses all criteria Ior distinguishing between Iormalistic ritual,
systematic distortion, choreographed consent, manipulated public opinion, and the real thing.
17
The participatory model oI democracy maintains that what makes Ior good leaders also makes Ior good citizensactive participation in ruling and being ruled (i.e., in
the exercise oI power) and also in public will and opinion Iormation. Democracy in this sense would allow all citizens, and not only elites, to acquire a democratic
political culture. For it is through political experience that one develops a conception oI civic virtue, learns to tolerate diversity, to temper Iundamentalism and egoism,
and to become able and willing to compromise.
18
Hence the insistence that without public spaces Ior the active participation oI the citizenry in ruling and being ruled,
without a decisive narrowing oI the gap between rulers and ruled, to the point oI its abolition, polities are democratic in name only.
19
For the most part, however, when it comes to conceptualizing alternatives, participation theorists oIIer institutional models that are meant to substitute Ior rather than
complement the allegedly undemocratic (and/or bourgeois) Iorms oI representative government that exist today.
20
Whether the theorist harkens back to an idealized
model oI the Greek polis, to the republican tradition oI the late medieval city-state, or to the new Iorms oI democracy generated within the milieus oI the workers'
movement (council communism, revolutionary syndicalism), in each case the alternative is presented as the single organizational principle Ior society as a whole.
Accordingly, the underlying thrust oI these models is the dediIIerentiation oI society, the state, and the economy. Small wonder that participationists in turn are accused
by their opponents oI utopianism and/or antimodernism.
21
To sum up, this debate leaves us with the Iollowing antinomy: Contemporary democratic theory involves either some rather undemocratic adjustments to the
"exigencies oI complex industrial societies" coupled with an abandonment oI the normative core oI

Page 8
the very concept oI democracy, or it proIIers somewhat hollow normative visions that cannot be reconciled with the institutional requirements oI modern society.
22
Rights-Oriented Liberalism Vs. Communitarianism
The debate between political liberals and communitarians reproduces some oI the arguments described above but on a diIIerent terrain. In one respect, both sides in
this debate challenge the elite/pluralist model oI democracy.
23
Both reject the antinormative, empiricist, utilitarian strain in this model and both seek to develop a
convincing normative theory oI democratic legitimacy or justice. The dispute is over how to Iormulate such a theory. Despite this shiIt in emphasis, however, this
debate also culminates in a set oI antinomic positions Irom which it seems unable to extricate itselI.
At the center oI the controversy are two interrelated issues, one epistemological, the other political. The Iirst revolves around the question oI whether it is possible to
articulate a Iormal, universalistic (deontological) conception oI justice without presupposing a substantive (historically and culturally speciIic) concept oI the
good.
24
The second revolves around the question oI how Ireedom can be realized in the modern world. At issue here is whether the idea oI Ireedom should be
explicated primarily Irom the standpoint oI individual rights or oI the community's shared norms.
25
Each side comes up with a diIIerent, indeed opposed, set oI
responses as to what constitutes the legitimating principles oI a constitutional democracy. In the process, however, the very conception oI liberal democracy
disintegrates into its component parts.
Liberal theorists see the respect Ior individual rights and the principle oI political neutrality as the standard Ior legitimacy in constitutional democracies. The core
premise oI rights-oriented liberalism is that individuals qua individuals have moral rights that serve as constraints on government and on othersconstraints that are
under the control oI the rights holder. They have these rights not on the grounds oI some social convention, aggregate common utility, tradition, or dispensation Irom
God, but by virtue oI their having some "property" (moral autonomy, human dignity) that constitutes them as bearers oI rights.
26
The liberal sees indi-

Page 9
vidual autonomy, moral egalitarianism, and universalism as inherent in the idea oI moral rights.
27
As such, rights constitute the heart oI a conception oI justice that
makes plausible the claim to legitimacy oI any modern polity. Law and political decisions are binding to the degree to which they respect individual rights.
28
The communitarian critique oI the rights thesis Iocuses on its individualist presuppositions and universalist claims. With respect to the Iirst, communitarians argue that
the liberal ideals oI moral autonomy and individual selI-development are based on an atomistic, abstract, and ultimately incoherent concept oI the selI as the subject oI
rights.
29
This allegedly leads to a Iocus on nonpolitical Iorms oI Ireedom (negative liberty) and an impoverished conception oI political identity, agency, and ethical liIe.
Accordingly, the communitarians invoke a set oI empirical and normative arguments against these assumptions. First, they argue that individuals are situated within an
historical and social context; they are socialized into communities through which they derive their individual and collective identity, language, world concepts, moral
categories, etc. Hence the empirical primacy oI the social over the individual is asserted against the alleged priority oI the asocial individual to society. Second, on the
normative level, communitarians charge that liberals Iail to see that communities are independent sources oI value and that there are communal duties and virtues
(loyalty, civic virtue) distinct Irom duties to others qua their abstract humanity. Indeed, duties oI loyalty and membership are and must be primary.
As Iar as universalism goes, communitarians claim that what the liberal sees as universal norms grounded in the universal character oI humanity (dignity or moral
autonomy) are in Iact particular norms embedded in shared understandings oI speciIic communities. The individual cannot have a Iirm basis Ior moral judgment without
getting it Irom a community to which one is committed. The strongest claim is that there are no duties pertaining to abstract man but only to members: The proper basis
oI moral theory is the community and its good, not the individual and her rights. Indeed, individuals have rights to the degree to which these Ilow Irom the common
good. Accordingly, the idea oI moral rights is an empty universalism that mistakenly abstracts Irom the only real basis oI

Page 10
moral claims, the community. Only on the basis oI a shared conception oI the good liIe, only within the Iramework oI a substantive ethical political community (with a
speciIic political culture) can we lead meaningIul moral lives and enjoy true Ireedom.
For those communitarians who see themselves as democrats,
30
the concept oI Ireedom thus has to do not with the idea oI moral rights but with the speciIic way in
which agents come to decide what they want and ought to do. Taken together, the empirical and normative criticisms oI the rights thesis imply that Ireedom must have
its original locus not in the isolated individual but in the society that is the medium oI individuation: in the structures, institutions, practices oI the larger social whole.
Civic virtue rather than negative liberty, the public good as distinct Irom the right, democratic participation unlike individual rights (and the concomitant adversarial
political culture), involve a communal practice oI citizenship that should pervade the institutions oI society on all levels and become habitualized in the character,
customs, moral sentiments oI each citizen. By implication, and on the strongest version oI these claims, a society in which claims oI individual rights proliIerate cannot
be a solidary community but must be alienated, anomic, privatized, competitive and lacking in moral substance.
This debate also leads to an apparently unresolvable antinomy. On the one side, the liberal tradition itselI, with its Iocus on individual rights and its illusions about the
possibility oI political neutrality, appears as the source oI egoistic, disintegrative tendencies in modern society and hence as the main impediment to achieving a
democratic society predicated on civic virtue. The other side counters with the contention that modern societies are precisely not communities integrated around a
single conception oI the good liIe. Modern civil societies are characterized by a plurality oI Iorms oI liIe; they are structurally diIIerentiated and socially heterogeneous.
Thus, to be able to lead a moral liIe, individual autonomy and individual rights must be secured. On this view, it is democracy, with its emphasis on consensus, or at
least on majority rule, that is dangerous to liberty, unless suitably restricted by constitutionally guaranteed basic rights that alone can render them legitimate in the eyes
oI minorities.

Page 11
The Defense of Welfare State Vs. Neoconservative Antistatism
The debate between deIenders oI the welIare state and its neolaissez-Iaire critics has also been going around in circles, albeit Ior a shorter time than the controversy
plaguing democratic theory.
31
Arguments Ior the welIare state have been made on both economic and political grounds.
32
According to Keynesian economic doctrine,
welIare state policies serve to stimulate the Iorces oI economic growth and to prevent deep recessions by encouraging investment and stabilizing demand. Fiscal and
monetary incentives Ior investors coupled with social insurance, transIer payments, and public services Ior workers compensate Ior the dysIunctions, uncertainties, and
risks oI the market mechanism and contribute to overall stability. High growth rates, Iull employment, and low inIlation should be the result oI this policy.
The political aspects oI the welIare state would also increase stability and productivity. On the one side, legal entitlements to state services and transIer payments
simultaneously aid those who Ieel the negative eIIects oI the market system while removing potentially explosive needs or issues Irom the arena oI industrial conIlict. On
the other side, the recognition oI the Iormal role oI labor unions in collective bargaining and in the Iormation oI public policy "balances" the asymmetrical power relation
between labor and capital and mitigates class conIlict.
33
The overall increase in social justice would lead to Iewer strikes, greater productivity, and an overall consensus
oI capital and labor that they have a mutual interest in the success oI the political economic system: Growth and productivity serve everyone. The welIare state would
Iinally deliver on the claim oI liberal capitalist societies to be egalitarian and just, by supporting the worst oII and by creating the preconditions Ior a true equality oI
opportunity, which in the eyes oI deIenders oI the welIare state is the only context in which civil and political rights can Iunction in a universalistic manner. Instead oI
being concerned by the anomalous status oI the so-called social rights, Ior a theorist such as T. H. Marshall these represent the highest and most Iundamental type oI
citizen rights.
34
Certainly the remarkable growth rates, relative stability, and increase in the standard oI living in postwar Western capitalist

Page 12
economies have, until recently, made the arguments Ior state intervention convincing to all but a very Iew. In a new context oI more limited possibilities Ior growth,
neoconservative deIenders oI a return to ''laissez-Iaire" criticize both the economic and political claims oI the welIare state model. UnIortunately Ior the latter, their
arguments also carry weight. Indeed, it was not diIIicult Ior these critics to point to the high rates oI unemployment and inIlation and low growth rates that have plagued
Western capitalist economies since the 1970s as prooI that state-bureaucratic regulation oI the economy is counterproductive. They can also point to successes in
these domains where their own policies have been applied.
On the economic Iront, three claims are made against the policies oI welIare states: that they lead to a disincentive to invest and a disincentive to work, and that they
constitute a serious threat to the viability oI the independent middle class.
35
The burden imposed by the regulatory and Iiscal policies on capital together with the
power oI unions to extract high wages allegedly contribute to declining growth rates and, in a context oI severe competition, lead to the perception that investment in
home markets will be unproIitable.
36
The disincentive to work is attributed to extensive social security and unemployment provisions that allow workers to avoid
undesirable jobs and to escape the normal pressure oI market Iorces. The quantity oI available workers shrinks as whole sectors oI the working class are turned into
welIare state clients, while the work ethic declines as workers become simultaneously more demanding and less willing to spend eIIort on their work. Finally, the
independent middle class Iinds itselI squeezed by high rates oI taxation and inIlation. The emergence oI the "new middle class" oI civil service proIessionals and higher-
level bureaucrats only exacerbates these problems because these strata have an interest in reproducing and expanding the client population on which their jobs
depend. WelIare state economic policies are thus antinomic in more than one respect: Policies meant to stimulate demand undermine investment, policies meant to
provide economic security Ior workers undermine the willingness to work, the policy oI tempering the undesirable side eIIects deriving Irom unregulated market Iorces
creates even greater economic problems in the Iorm oI a vastly expanded, expensive, unproductive state sector.

Page 13
On the political Iront, neoconservatives argue that the very mechanisms introduced by welIare states to resolve conIlicts and create greater equality oI opportunity,
namely legal entitlements and the expanded state sector, have led to new conIlicts and have violated the rights and liberty oI some Ior the sake oI others. By impinging
upon the core right oI liberal market systems, namely, private property, state intervention and regulation undermine both the liberty oI entrepreneurs and the incentive
to achieve on the part oI the working population. Far Irom increasing social justice or equality oI opportunity, welIare undermines the preconditions Ior both oI these.
In short, it rewards Iailure rather than success. In the name oI equality, moreover, state intervention in the everyday lives oI its clients poses a severe threat to liberty,
privacy, and autonomy.
In addition, these mechanisms have allegedly generated a set oI rising expectations and increasing demands that lead to an overall situation oI ungovernability.
37
Indeed, the very institutions oI welIare state mass democracy that promised to channel political conIlict into acceptable and harmless Iorms (the end oI ideology) and
to integrate workers especially into the political and economic system oI late capitalism (deradicalization)i.e., the competitive (catchall) party system based on
universal suIIrage, interest group politics, collective bargaining, and extensive social rightslead to a dangerous overload oI the political system and a crisis oI
authority.
38
In short, the rights explosion that so irritates democratic communitarians is even more alarming to neoconservative critics oI "statism." By placing
obligations upon itselI that it cannot possibly IulIill,
39
the state creates rising yet unsatisIiable expectations, becomes overexpanded and weak at the same time, and
suIIers Irom a dangerous loss oI authority. Indeed, on this view, there is a central political contradiction inherent in the welIare state: In order Ior the perIormance
capacity oI the state to be enhanced vis-a-vis the number oI demands, the very Ireedoms, modes oI participation, and sets oI rights associated with it would have to be
curtailed.
40
The neo-laissez-Iaire economic and political alternatives, however, do not escape the Iate oI becoming merely one oI the untenable sides oI an antinomic structure.
"Supply-side" economists seek to

Page 14
dismantle the welIare state in order to eliminate the "disincentive" to invest, but to do so would be to abolish precisely those "buIIers" that stabilize demand.
41
II the
socioeconomic supports Ior workers and the poor are terminated in the name oI reIurbishing the work ethic, the compulsion oI the market will certainly return, but so
will the gross injustices, dissatisIaction, instability, and class conIrontations that characterized the capitalist economies prior to welIare state policies.
OI course, the attack on the welIare state is predicated on the idea that there is an unlimited growth potential Ior marketable goods and services that would be
unleashed once the state is pushed back into its proper, minimal terrain. Privatization and deregulation would allegedly restore competition and end the inIlation oI
political demands. However, the political presuppositions Ior such a policy conIlict with its goals oI social peace and social justice. Necessarily repressive policies
regarding the right to associate and eIIorts to abolish social rights ranging Irom social security to unemployment compensation, not to mention welIare, are scarcely
conducive to consensus. While the "Ireedom-threatening" dimensions oI state intervention, namely the regulation oI proprietors, the supervision and control oI clients,
and the spiraling cycle oI dependency, would end, so would all oI the gains in social justice, equality, and rights. Moreover, eIIorts to restore state authority by limiting
its scope and by shielding it Irom popular demands would not reduce state activism but would simply shiIt it Irom the political to the administrative terrain. For, iI one
reduces the ability oI democratic institutions such as the party system, elections, and parliaments to provide Ior the articulation oI political conIlict, alternative channels,
such as the neocorporatist arrangements proliIerating in Western Europe, will develop. While these arrangements successIully shield the state Irom excess demands,
they hardly indicate a shiIt Irom state to market regulation. Thus the neo-laissez-Iaire alternative to the "crisis" oI the welIare state is as internally contradictory as the
illness it purports to cure.
We are accordingly leIt with the Iollowing antinomy: Either we choose more social engineering, more paternalism and leveling, in short, more statism, in the name oI
egalitarianism and social rights, or we opt Ior the Iree market and/or the reIurbishing oI authori-

Page 15
tarian social and political Iorms oI organization and relinquish the democratic, egalitarian components oI our political culture in order to block Iurther bureaucratization
oI everyday liIe. It seems that liberal democratic market societies cannot coexist with, nor can they exist without, the welIare state.
Revival of the Concept of Civil Society
The early modern concept oI civil society was revived Iirst and Ioremost in the struggles oI the democratic oppositions in Eastern Europe against authoritarian socialist
party-states. Despite diIIerent economic and geopolitical contexts, it does not seem terribly problematic to apply the concept also to the "transitions Irom authoritarian
rule" in Southern Europe and Latin America, above all because oI the common task shared with the oppositions oI the East to constitute new and stable democracies.
But why should such a concept be particularly relevant to the West? Is not the revival oI the discourse oI civil society in the East and the South simply part oI a project
to attain what the advanced capitalist democracies already have: civil society guaranteed by the rule oI law, civil rights, parliamentary democracy, and a market
economy? Could one not argue that struggles in the name oI creating civil and political society especially in the East are a kind oI repeat oI the great democratic
movements oI the eighteenth and nineteenth centuries that created a type oI duality between state and civil society which remains the basis Ior Western democratic and
liberal institutions? And isn't this an admission that the elite theorists, the neoconservatives, or at best the liberals are right aIter all? Put this way, the revival oI the
discourse oI civil society appears to be just that, a revival, with little political or theoretical import Ior Western liberal democracies. And iI this is so, why would a civil-
society-oriented perspective provide a way out oI the antinomies plaguing Western political and social thought?
Several interrelated issues that have emerged in the current revival go beyond the model oI the historical origins oI civil society in the West and thereIore have
important lessons to oIIer established liberal democracies. These include the conception oI selI-limitation, the idea oI civil society as comprised oI social movements as
well as

Page 16
a set oI institutions, the orientation to civil society as a new terrain oI democratization,
42
the inIluence oI civil on political and economic society, and Iinally an
understanding that the liberation oI civil society is not necessarily identical with the creation oI bourgeois society but rather involves a choice between a plurality oI
types oI civil society. All these notions point beyond a restriction oI the theory oI civil society merely to the constituent phase oI new democracies.
The idea oI selI-limitation, all too oIten conIused with the strategic constraints on emancipatory movements, is actually based on learning in the service oI democratic
principle. The postrevolutionary or selI-limiting "revolutions" oI the East are no longer motivated by Iundamentalist projects oI suppressing bureaucracy, economic
rationality, or social division. Movements rooted in civil society have learned Irom the revolutionary tradition that these Iundamentalist projects lead to the breakdown
oI societal steering and productivity and the suppression oI social plurality, all oI which are then reconstituted by the Iorces oI order only by dramatically authoritarian
means. Such an outcome leads to the collapse oI the Iorms oI selI-organization that in many cases were the major carriers oI the revolutionary process: revolutionary
societies, councils, movements. Paradoxically, the selI-limitation oI just such actors allows the continuation oI their social role and inIluence beyond the constituent and
into the constituted phase.
This continuation oI a role oI civil society beyond the phase oI transition can be coupled with domestication, demobilization, and relative atomization. That would mean
convergence with society as the Western elite pluralists see it. But in the postauthoritarian settings actors who have rejected Iundamentalism and raised civil society to
a normative principle show that we do have a choice. While the total democratization oI state and economy cannot be their goal, civil society itselI, as Tocqueville was
Iirst to realize, is an important terrain oI democratization, oI democratic institution building. And iI East European oppositionals were driven to this alternative at Iirst
only by blockages in the sphere oI state organization, there is certainly a good chance that the idea oI the Iurther democratization oI civil society will gain emphasis in
the Iace oI the inevitable disappointments, visible above all in Hungary, (East)

Page 17
Germany, and Czechoslovakia, with the emergence oI the typical practices oI Western democracies. Thus, the actors oI the new political societies would do well, iI
they value their long-term legitimacy, to promote democratic institution building in civil society, even iI this seems to increase the number oI social demands on them.
The idea oI the democratization oI civil society, unlike that oI its mere revival, is extremely pertinent to existing Western societies. Indeed, the tendency to see
extrainstitutional movements and initiatives in addition to settled institutions as integral parts oI civil society is Iound earlier in Western than in Eastern experience, to
which it is rapidly being extended primarily by new and old movements and initiatives. It is quite possible that some oI the emerging Eastern constitutions will embody
new sensitivity to an active civil society, a sensitivity that should in turn inIluence Western constitutional developments. These potential normative gains will conIirm, in
the East as well as the West, the idea that there can be very diIIerent types oI civil society: more or less institutionalized, more or less democratic, more or less active.
Discussions in the milieu oI Solidarity in Poland raised these choices explicitly as early as 1980, along with the choice oI political vs. antipolitical models oI civil society.
In the current wave oI economic liberalism in Poland, Czechoslovakia, and Hungary, another question inevitably arises concerning the connection between economy
and civil society and the choice between an economic, individualistic society and a civil society based on solidarity, protected not only against the bureaucratic state
but also against the selI-regulating market economy. This debate, too, will be directly relevant in Western contexts, as it already has been in Latin America, and
conversely Western controversies around the welIare state and the "new social movements" should have much intellectual material to oIIer Eastern radical democrats
hoping to protect the resource oI solidarity without paternalism.
The aim oI our book is to Iurther develop and systematically justiIy the idea oI civil society, reconceived in part around a notion oI selI-limiting democratizing
movements seeking to expand and protect spaces Ior both negative liberty and positive Ireedom and to recreate egalitarian Iorms oI solidarity without impairing eco-

Page 18
nomic selI-regulation. BeIore turning to this task, we would like to conclude this introduction by illustrating the important, and perhaps decisive, contribution oI our
theory oI civil society to the three theoretical antinomies mentioned above.
43
Civil Society and Contemporary Political Theory
It might seem that our position is already anticipated by one oI the six theoretical traditions involved in the debates depicted above, namely the pluralist version oI the
elite democratic tradition oI political theory.
44
Indeed, the pluralists' addition to the elite model oI democracy is precisely a conception oI a "third realm" diIIerentiated
Irom the economy and the state (what we call "civil society").
45
On the pluralist analysis, a highly articulated civil society with cross-cutting cleavages, overlapping
memberships oI groups, and social mobility is the presupposition Ior a stable democratic polity, a guarantee against permanent domination by any one group and
against the emergence oI Iundamentalist mass movements and antidemocratic ideologies.
46
Moreover, a civil society so constituted is considered to be capable oI
acquiring inIluence over the political system through the articulation oI interests that are "aggregated" by political parties and legislatures and brought to bear on political
decision making, itselI understood along the lines oI the elite model oI democracy.
Although we use many oI the terms oI this analysis in our work on civil society, our approach diIIers in several key respects Irom that oI the pluralists. First, we do not
accept the view that the "civic culture" most appropriate to a modern civil society is one based on civil privatism and political apathy. As is well known, the pluralists
value involvement in one's Iamily, private clubs, voluntary associations, and the like as activities that deIlect Irom political participation or activism on the part oI
citizens.
47
It is this which allegedly makes Ior a stable democratic polity. Moreover, it makes no diIIerence to this model what the internal structure oI the institutions
and organizations oI civil society is.
48
Indeed, in their haste to replace "utopian (participatory democratic) principles" with realism, the pluralists tend to consider
attempts to apply the egalitarian norms oI civil society to social institutions as naive.
49

Page 19
We do not share this view. Instead, we build upon the thesis oI one oI the most important predecessors oI the pluralist approach, Alexis de Tocqueville, who argued
that without active participation on the part oI citizens in egalitarian institutions and civil associations, as well as in politically relevant organizations, there will be no
way to maintain the democratic character oI the political culture or oI social and political institutions. Precisely because modern civil society is based on egalitarian
principles and universal inclusion, experience in articulating the political will and in collective decision making is crucial to the reproduction oI democracy.
This, oI course, is the point that is always made by participation theorists. Our approach diIIers Irom theirs in arguing Ior more, not less, structural diIIerentiation. We
take seriously the normative principles deIended by radical democrats, but we locate the genesis oI democratic legitimacy and the chances Ior direct participation not
in some idealized, dediIIerentiated polity but within a highly diIIerentiated model oI civil society itselI. This shiIts the core problematic oI democratic theory away Irom
descriptive and/or speculative models to the issue oI the relation and channels oI inIluence between civil and political society and between both and the state, on the
one side, and to the institutional makeup and internal articulation oI civil society itselI, on the other. Moreover, we believe that the democratization oI civil societythe
Iamily, associational liIe, and the public spherenecessarily helps open up the Iramework oI political parties and representative institutions.
50
Indeed, these concerns open the way to a dynamic conception oI civil society, one that avoids the apologetic thrust oI most pluralist analyses. Far Irom viewing social
movements as antithetical to either the democratic political system or to a properly organized social sphere (the pluralists' view), we consider them to be a key Ieature
oI a vital, modern, civil society and an important Iorm oI citizen participation in public liIe. Yet we do not see social movements as preIiguring a Iorm oI citizen
participation that will or even ought to substitute Ior the institutional arrangements oI representative democracy (the radical democratic position). In our view, social
movements Ior the expansion oI rights, Ior the deIense oI the autonomy oI civil society, and Ior its Iurther democratization are

Page 20
what keep a democratic political culture alive. Among other things, movements bring new issues and values into the public sphere and contribute to reproducing the
consensus that the elite/pluralist model oI democracy presupposes but never bothers to account Ior.
51
Movements can and should supplement and should not aim to
replace competitive party systems. Our conception oI civil society thus retains the normative core oI democratic theory while remaining compatible with the structural
presuppositions oI modernity. Finally, while we also diIIerentiate the economy Irom civil society, we diIIer Irom the pluralists in that we do not seal oII the borders
between them on the basis oI an allegedly sacrosanct Ireedom oI contract or property right. Nor do we seek to "reembed" the economy in society. Instead, on our
analysis, the principles oI civil society can be brought to bear on economic institutions within what we call economic society. The question here, as in the case oI the
polity, is what channels and receptors oI inIluence do, can, and ought to exist.
52
Indeed, we are able to pose such questions on the basis oI our model without risking
the charges oI utopianism or antimodernism so Irequently and deservedly leveled against worker-based versions oI radical democracy.
It is also our thesis that the tensions between rights-oriented liberalism and, at least, democratically oriented communitarianism can be considerably diminished iI not
entirely overcome on the basis oI a new theory oI civil society. While the idea oI rights and oI a democratic political community derive Irom distinct traditions in
political philosophy, today they belong to the same political culture. They need not be construed as antithetical, although on an empirical level the rights oI an individual
may conIlict with majority rule and "the public interest," necessitating a balancing between the two sides.
53
Nor is it necessary to view these as based on two conIlicting
sets oI principles or presuppositions, such that one could accommodate the Iirst set only insoIar as it is instrumental to the achievement or preservation oI the other. On
the contrary, we contend that what is best in rights-oriented liberalism and democratically oriented communitarianism constitutes two mutually reinIorcing and partly
overlapping sets oI principles. Two steps are necessary to argue this thesis and to transcend the relevant antinomies. First, one must show that there is a philosophical

Page 21
Iramework that can provide a political ethic able to redeem the normative claims oI both rights-oriented liberalism and radical democracy. Second, one must revise the
conception oI civil society as the private sphere, shared by both theoretical paradigms, in order to grasp the institutional implications oI such an ethic.
We also deIend the principles oI universality and autonomy to which the rights thesis is wed, but we deny that this commits us either to the liberal notion oI neutrality or
to an individualist ontology. The communitarians are right: Much oI liberal theory, especially the contract tradition Irom Hobbes to Rawls, has relied on either one or
both oI these principles.
54
However, the Habermasian theory oI discourse ethics, on which we rely, provides a way to develop conceptions oI universality and
autonomy that are Iree oI such presuppositions. On this theory, universality does not mean neutrality with respect to a plurality oI values or Iorms oI liIe but rather
reIers, in the Iirst instance, to the metanorms oI symmetric reciprocity
55
that are to act as regulative principles guiding discursive processes oI conIlict resolution and, in
the second instance, to those norms or principles to which all those who are potentially aIIected can agree. The procedure oI universalization deIended here involves
an actual rather than a hypothetical dialogue. It does not require that one abstract Irom one's concrete situation, need interpretations, or interests in order to engage in
an unbiased moral testing oI principles. Instead, it requires that these be Ireely articulated. It also requires that all those potentially aIIected by institutionalized norms
(laws or policies) be open to a multiplicity oI perspectives. Accordingly, universality is a regulative principle oI a discursive process in and through which participants
reason together about which values, principles, need-interpretations merit being institutionalized as common norms.
56
Thus, the atomistic disembodied individual
allegedly presupposed by procedural (deontological) ethics is most emphatically not the basis oI this approach. Assuming that individual and collective identities are
acquired through complex processes oI socialization that involve both internalizing social norms or traditions, and developing reIlective and critical capacities vis-a-vis
norms, principles, and traditions, this theory has at its core an intersubjective, interactive conception oI both individuality and autonomy. It is thus able to

Page 22
accommodate the communitarian insights into the social core oI human nature without abandoning the ideas oI either universality or moral rights. Indeed, discourse
ethics provides a philosophical basis Ior democratic legitimacy that presupposes valid rights, even iI not all oI these rights are derivable Irom it.
57
While it is oI course individuals who have rights, the concept oI rights does not have to rest on philosophical or methodological individualism, nor, Ior that matter, on
the idea oI negative liberty alone. Although most liberal and communitarian theorists have assumed that such a conception oI Ireedom and oI individualism is
presupposed by the very concept oI rights, we believe that only some rights involve primarily negative liberty while none requires a philosophically atomistic conception
oI individuality. It is here that a revised conception oI civil society, together with a new theory oI rights, must enter into the analysis. For every theory oI rights, every
theory oI democracy, implies a model oI society. UnIortunately, communitarians and liberals also agree that the societal analogue oI the rights thesis is a civil society
construed as the private sphere, composed oI an agglomeration oI autonomous but egoistic, exclusively selI-regarding, competitive, possessive individuals whose
negative liberty it is the polity's task to protect. It is their assessments and not their analysis oI this Iorm oI society that diverge.
But this is only one possible version oI civil society and certainly not the only one that can be ''derived" Irom the rights thesis. Only iI one construes property to be not
simply a key right but the core oI the conception oI rightsonly, that is, iI one places the philosophy oI possessive individualism at the heart oI one's conception oI civil
society and then reduces civil to bourgeois societydoes the rights thesis come to be deIined in this way.
58
II, however, one develops a more complex model oI civil
society, recognizing that it has public and associational components as well as individual, private ones, and iI, in addition, one sees that the idea oI moral autonomy
does not presuppose possessive individualism,
59
then the rights thesis begins to look a bit diIIerent. In short, rights do not only secure negative liberty, the autonomy oI
private, disconnected individuals. They also secure the autonomous (Ireed Irom state control) communicative interaction oI individuals with one another in the public
and private spheres oI civil society, as well as a new

Page 23
relation oI individuals to the public and the political spheres oI society and state (including, oI course, citizenship rights). Moral rights are thus not by deIinition apolitical
or antipolitical, nor do they constitute an exclusively private domain with respect to which the state must limit itselI. On the contrary, the rights to communication,
assembly, and association, among others, constitute the public and associational spheres oI civil society as spheres oI positive freedom within which agents can
collectively debate issues oI common concern, act in concert, assert new rights, and exercise inIluence on political (and potentially economic) society. Democratic as
well as liberal principles have their locus here. Accordingly, some Iorm oI diIIerentiation oI civil society, the state, and the economy is the basis Ior both modern
democratic and liberal institutions. The latter presuppose neither atomistic nor communal but rather associated selves. Moreover, on this conception the radical
opposition between the philosophical Ioundations and societal presuppositions oI rights-oriented liberalism and democratically oriented communitarianism dissolves.
This conception oI civil society does not, oI course, solve the question oI the relation between negative and positive liberty, but it does place this issue within a
common societal and philosophical terrain. It is on this terrain that we learn how to compromise, take reIlective distance Irom our own perspective so as to entertain
others, learn to value diIIerence, recognize or create anew what we have in common, and come to see which dimensions oI our traditions are worth preserving and
which ought to be abandoned or changed.
This brings us to the heart oI our diIIerences with the neoconservative model oI civil society. The neoconservative slogan, "society against the state," is oIten based on
a model in which civil society is equivalent to market or bourgeois society. Another version oI this approach does, however, recognize the importance oI the cultural
dimension oI civil society. We have serious objections even to this second version, whose strategies Ior unburdening the state are aimed in part at the institutions
involved in the Iormation and transmission oI cultural values (art, religion, science) and in socialization (Iamilies, schools). An important component oI the
neoconservative thesis oI "ungovernability" is the argument that the excessive material demands placed by citizens on the state are

Page 24
due not only to welIare institutions themselves but also to our modernist political, moral, and aesthetic culture. The latter allegedly weakens both traditional values and
the agencies oI social control (such as the Iamily) that tempered hedonism in the past.
60
In this view, we need to resacralize our political culture, revive Ialtering
traditional values such as selI-restraint, discipline, and respect Ior authority and achievement, and shore up "nonpolitical" principles oI order (Iamily, property, religion,
schools), so that a culture oI selI-reliance and selI-restraint replaces the culture oI dependency and critique.
61
The cultural politics oI neoconservatism that accompanies
the policies oI deregulation and privatization are thus based on the deIense or re-creation oI a traditionalist and authoritarian liIeworld.
62
Our conception oI civil society points to a diIIerent assessment. First, we try to show that the resources Ior meaning, authority, and social integration are undermined
not by cultural or political modernity (based on the principles oI critical reIlection, discursive conIlict resolution, equality, autonomy, participation, and justice) but,
rather, by the expansion oI an increasingly illiberal corporate economy as well as by the overextension oI the administrative apparatus oI the interventionist state into
the social realm. The use oI economic and political power to shore up (or worse, re-create) the "traditional" hierarchical, patriarchal, or exclusionary character oI many
oI the institutions oI civil society is, on our view, what Iosters dependency. We agree that certain Ieatures oI the welIare state
63
Iragment collectivities, destroy
horizontal solidarities, isolate and render private individuals dependent on state apparatuses. Unrestrained capitalist expansion, however, has the same destructive
consequences. But appeals to Iamily, tradition, religion, or community could Ioster the destructive Iundamentalism oI Ialse communities so easily manipulated Irom
above, unless the achievements oI liberalism (the principle oI rights), democracy (the principles oI participation and discourse), and justice (a precondition Ior
solidarity) are Iirst deIended and then supplemented with new democratic and egalitarian Iorms oI association within civil society.
Moreover, to opt Ior the preservation oI traditions, iI accompanied by a denial oI the universalist tradition oI cultural and political

Page 25
modernity, implies Iundamentalism. Accordingly, the question that Ilows Irom our model becomes: Which traditions, which Iamily Iorm, which community, which
solidarities are to be deIended against disruptive intervention? Even iI cultural modernity itselI is just one tradition among many, its universal thrust is the reIlexive,
nonauthoritarian relation toward traditionan orientation that can be applied to itselI and that implies autonomy (allegedly cherished by the neoconservative) rather
than heteronomy. Indeed, traditions that have become problematic can be preserved only on the terrain oI cultural modernity, i.e., through arguments that invoke
principles. Such discussion does not mean the abolition oI tradition, solidarity, or meaning; rather, it is the only acceptable procedure Ior adjudicating between
competing traditions, needs, or interests that are in conIlict. Accordingly, our model points toward the further modernization oI the culture and institutions oI civil
society as the only way to arrive at autonomy, selI-reliance, and solidarity among peers allegedly desired by the neoconservative critics oI the welIare state.
64
Our conception oI civil society seeks to demystiIy the other strain within neoconservatism, namely, that the only alternative to the paternalism, social engineering, and
the bureaucratization oI our lives typical oI welIare state systems is to shiIt steering back to the magic oI the marketplace (and oI course to renounce distributive justice
and egalitarianism). This "solution" is not only politically untenable and normatively undesirable; it is also based on the Iallacious assumption that no other options exist.
Our Iramework, however, allows in principle Ior a third approach, one that does not seek to correct the economic or state penetration oI society by shiIting back and
Iorth between these two steering mechanisms. Instead, the task is to guarantee the autonomy oI the modern state and economy while simultaneously protecting civil
society Irom destructive penetration and Iunctionalization by the imperatives oI these two spheres. For now, oI course, we have only some oI the elements oI a theory
that can thematize both the diIIerentiation oI civil society Irom state and economy and its reIlexive inIluence over them through the institutions oI political and economic
society. But we believe that our conception has the best prospects Ior Iuture theoretical progress and Ior integrating the diverse conceptual

Page 26
strategies that are currently available. The project it implies would avoid correcting the results oI state paternalism by another Iorm oI colonization oI society, this time
by an unregulated market economy. It would seek to accomplish the work oI social policy by more decentralized and autonomous civil-society-based programs than
in traditional welIare states and the work oI economic regulation by nonbureaucratic, less intrusive Iorms oI legislation, "reIlexive law," Iocusing on procedures and not
results.
65
In our view this synthetic project should be described not only by Habermas's term, "the reIlexive continuation oI the welIare state," but also by the
complementary idea oI the "reIlexive continuation oI the democratic revolution." The Iormer arises in the context oI Western welIare states, the latter in that oI the
democratization oI authoritarian regimes. The two ideas can and should be combined. Thus Iar, the recent revival and development oI the concept oI civil society has
involved learning Irom the experience oI the ''transitions to democracy." The idea oI the reIlexive continuation oI the welIare state and oI liberal democracy should,
however, open the way Ior enriching the intellectual resources oI democrats in the East by what has been learned in a double critique oI already established welIare
states and oI their neoconservative discontents. A theory oI civil society inIormed by such ideas should also inIorm the projects oI all those in the West who seek the
Iurther democratization oI liberal democracies.

Page 27
I-
THE DISCOURSE OF CIVIL SOCIETY

Page 29
1-
The Contemporary Revival of Civil Society
Phrases involving the resurrection, reemergence, rebirth, reconstruction, or renaissance oI civil society are heard repeatedly today. These terms, indicating the
continuity oI an emerging political paradigm with essential trends oI early modernity, are misleading in one important respect: They reIer not only to something modern
but also to something signiIicantly new. A simple chronology derived in part Irom Karl Polanyi might, in an extremely preliminary way, indicate what is at stake.
According to Polanyi, during most oI the nineteenth century, Iorces representing the capitalist selI-regulating market economy were on the oIIensive, claiming an
identity with the liberal society that was in the process oI emancipating itselI Irom the absolutist and paternalistic state. Polanyi, however, rightly stressed that in the late
nineteenth century and through much oI the twentieth century a reversal had taken place. Now, elites representing the logic and goals oI the modern state were
successIully claiming to express the interests oI a heterogeneous set oI social groups and tendencies resisting and challenging the destructive trends oI capitalist market
society. Not even Polanyi, however, Ioresaw that the statist phase would also have its limits. For a period oI more than a decade and a halI now, citizen initiatives,
associations, and movements have increasingly oriented themselves toward the deIense and expansion oI a variously described societal realm, the Iorms and projects
oI which are clearly distinguished Irom statism.
Two crucial ambiguities remain Irom the orientation "society against the state." First, while increasingly signiIicant groupings oI

Page 30
collective actors reject any representation oI their program in terms oI communitarianism, others continue to deIend an idealized Gemeinschaft or premodern network
oI communities, traditional solidarities, and collectives against modernity itselI. Second, there are various neoconservative, neoliberal, and libertarian initiatives (rarely
movements, but with signiIicant Iorce behind them) that identiIy "society" with market economy. Both oI these trends are regressive versions oI antistatism. The Iirst
wishes to retreat behind the modern state, thus eliminating an essential precondition oI modernity itselI; the second wishes to repeat the already Iailed experiment with
the Iully selI-regulated market economy oI classical capitalism. There is no chance oI the Iirst trend registering even temporary successes, although it will continue to
have a role within most social movements. The second trend, wherever successIul, threatens to transIorm history into oscillation between economic liberalism and
paternalist statism.
We believe there are today important elements oI a third project Ior retrieving the category oI civil societv Irom the tradition oI classical political theory. These involve
attempts to thematize a program that seeks to represent the values and interests oI social autonomy in Iace oI both the modern state and the capitalist economy,
without Ialling into a new traditionalism. Beyond the antinomies oI state and market, public and private, Gesellschaft and Gemeinschaft, and, as we shall show,
reIorm and revolution, the idea oI the deIense and the democratization oI civil society is the best way to characterize the really new, common strand oI contemporary
Iorms oI selI-organization and selI-constitution.
Problems oI selI-reIlection and selI-understanding within the movements and the initiatives themselves sometimes prevent them Irom clearly recognizing their own
diIIerence with communalism or libertarianism. At best the diIIerence represents a stake that must be internally contested. Behind the many ambiguities oI meaning tied
up with the concept oI civil society stand such conIlicts. In company with many participants, our book takes a clear stand in these conIlicts on behalI oI a modern civil
society capable oI preserving its autonomy and Iorms oI solidarity in Iace oI the modern economv as well as the state.
Such a project emerges Irom contexts oI social and political conIlicts themselves. In this chapter we present the idea by exam-

Page 31
ining several discourses that have revived the category oI civil society (albeit in diIIerent versions) in order to critically interpret the political contexts oI East and West,
North and South. Without aiming at a complete presentation oI all related views within each context, we deliberately stress perspectives in each that can be compared
with those in the other contexts. We shall attempt to identiIy the common strands, the alternative models, the signiIicant diIIerences as well as the conceptual unclarities
in these Iorms oI interpretation and selI-interpretation. The rest oI this book will, we hope, contribute to the Iurther development oI the discourse oI civil society and
thereby beneIit the actors and interpreters we present in this chapter.
The Polish Democratic Opposition
The opposition oI civil society and state made its most dramatic return in East Europe, particularly in the ideology oI the Polish opposition Irom 1976 to the advent oI
early Solidarity and beyond. The juxtapositions are well known: society against the state, nation against state, social order against political system, pavs reel
against pavs legal or officiel, public liIe against the state, private liIe against public power, etc. The idea was always the protection and/or selI-organization oI
social liIe in the Iace oI the totalitarian or authoritarian state. Adam Michnik provided the theoretical elaboration oI this conception under the heading oI "new
evolutionism."
1
He also discovered the historical conditions oI its possibility: the Iailure oI a potentially total revolution Irom below (Hungary in 1956), and the demise
oI a process oI reIorm Irom above (Czechoslovakia in 1968).
2
Michnik drew two lessons Irom these deIeats. First, the transIormation oI the Soviet-type system oI
East Central Europe was possible only within limits whose thresholds were the alliance system (threatened in Hungary in 1956) and the conIirmation oI the control oI
state institutions by a Soviet-type Communist party (challenged in diIIerent Iorms both in Hungary and in Czechoslovakia in 1968). Second, neither revolution Irom
below nor reIorm Irom above would work as the strategy Ior achieving what was in Iact possible.
The point oI view oI civil society in this context aims at a twoIold reorientation. First, the juxtaposition oI society against the state

Page 32
indicates not only the battlelines but also a shiIt concerning the target oI democratization, Irom the whole social system to society outside oI state institutions proper.
Thus, while the conception surely implies a pushing back oI the state-administrative Iorms oI penetration Irom various dimensions oI social liIe, it has, nevertheless, the
idea oI selI-limitation built into it Irom the start: The leading role oI the party in the (albeit shrinking) state sphere will not be challenged.
Second, the conception also indicates that the agent or the subject oI the transIormation must be an independent or rather a selI-organizing society aiming not at social
revolution but at structural reIorm achieved as a result oI organized pressure Irom below. These two aspects are brought together by the term "selI-limiting revolution"
coined by Jacek Kuron in the period oI Solidarity. At that time, the new conception truly came into its own, showing its Iormidable powers in promoting the selI-
understanding oI new types oI social actors. Nevertheless, it should be noted that the "new evolutionism" or the ''selI-limiting revolution" represented both a strategic
and a normative break with the revolutionary tradition whose logic was understood to be undemocratic and inconsistent with the selI-organization oI society.
3
All
major revolutions Irom the French to the Russian and the Chinese not only demobilized the social Iorces on which they originally depended but also established
dictatorial conditions that were meant to block the reemergence oI such Iorces at their very root Ior as long as possible. The project oI "selI-limiting revolution" has, oI
course, the opposite goal: the construction Irom below oI a highly articulated, organized, autonomous, and mobilizable civil society.
Leaving aside Ior now the overall theoretical cogency oI the conception, we must note some serious ambiguities in its elaboration in the milieu oI the Polish democratic
opposition.
4
Are the terms "society" and "civil society" the same? AIter all, they both reIer to a plurality oI Iorms oI independent groups (associations, institutions,
collectives, interest representations) as well as Iorms oI independent public opinion and communication. Put another way, how can civil society be both the agent oI
social transIormation and its result? One could, oI course, try to resolve the diIIiculty by distinguishing between society and civil society. The latter would represent a

Page 33
version oI the Iormer, institutionalized by legal mechanisms or rights, as in the Gdansk and subsequent agreements oI August and September 1980.
5
But the ambiguity
would remain, because "rights" in an authoritarian state-socialist setting (lack oI independent courts; lack oI a clear, unambiguous legal code; lack oI an organized legal
proIession) are easily revocable not only in principle but also in a political practice that depends on constant demonstration oI this revocability. Moreover, institutional
continuity can apparently be achieved by public enlightenment and selI-organization even without rights, as witnessed by the durability and growth oI autonomous
Iorms oI culture in the twelve-year period aIter 1976.
6
Another set oI conceptual diIIiculties revolves around the interpretation oI the notion oI society, oI social selI-organization in a supposedly totalitarian setting. Here one
view (Michnik) stressed the obliteration oI all social solidarities and the resulting social atomization, except Ior careIully deIined institutional complexes (the church) or
historical periods (1956, 197071, and aIter 1976). Another position, more consistent with the theory oI the new evolutionism, insisted on the Iailure oI totalitarianism,
whatever its intentions, to truly atomize society, or to completely disorganize Iamilies, Iace-to-Iace groups, and cultural networks.
7
This position, however, would have
required the working out oI a paradigm to replace the totalitarianism thesis as the theoretical Iramework oI the "new evolutionism," something never actually attempted.
More serious in principle is the lack oI clarity regarding the type oI civil society that is to be constructed or reconstructed. The conceptual conIusion derives above all
Irom a common unwillingness to take an openly critical attitude toward the liberal model oI civil society, despite participation in a solidaristic workers' movement that
is, in many respects, incompatible with this model. In the 1980s more and more people (e.g., Krol, Spievak, the editors oI Respublica) came to champion a version oI
the liberal model, based on economic individualism and Ireedoms oI property and enterprise as the central rights. Even within the milieu oI those close to Solidarity in
its Iirst great period (198081), there were disagreements over the various conceptions oI civil society. Cultural models (Wojcicki) were counterposed to political
conceptions (the Committee Ior the DeIense oI Workers, or KOR), on the one

Page 34
hand, while the level oI democracy needed in popular movements and institutions was hotly debated, on the other. Whereas it was generally recognized that the new
civil society was to be pluralistic,
8
the need Ior a single, all-encompassing organization to respond to the interest oI this plurality was temporarily accepted.
9
But once
such an organization emerged and managed to survive in the Iace oI "totalitarian" power, could its unitary and all-encompassing tendency be easily disposed oI?
Formulating a dualistic civil society and state Iramework proved even more diIIicult, especially in practical politics. Was civil society, as represented by Solidarity, to
be entirely apolitical, disinterested in "power," or was it to be expanded as a selI-governing republic making a state in the old sense more or less superIluous?
Sometimes aspects oI each conception are to be Iound even in the same author.
10
Would a selI-coordinating system oI society not negate the idea oI selI-limitation iI
the party-state were leIt only as a representative oI Soviet power, in charge oI military, police, and Ioreign policy and partially converted into an expert bureaucracy?
11
II, on the other hand, the dualistic conception requires institutional mechanisms oI compromise between societal organizations and party-state institutions, does the
idea oI building a hybrid system based on a new type oI society next to an unreIormed party-state make sense? And iI a reIorm oI oIIicial institutions, especially the
party itselI, must be hoped Ior and even promoted, iI party pragmatists could be looked upon as partners even iI not allies, could the much insisted-upon independent
identity oI the social movement be maintained?
12
What would be the point oI this iI on many issues party pragmatists and sectors oI the movement are closer to one
another than potentially diIIerent elements oI the antistate opposition? It is insuIIicient to reply that only an organized society, conscious oI its identity, is capable oI
compromise, Ior just this unity tended to demobilize potential partners in the party. The deep identity problems oI the ruling party could hardly be solved in the Iace oI
an organized society successIully reclaiming all legitimacy. Without a new party identity, party pragmatists lost all Ireedom oI action. And Ior the party leadership,
without legitimacy, the only Ireedom oI action leIt was the exercise oI raw sovereign power.
13
Many oI the diIIiculties touched upon here pointed toward the Iailure to rebuild civil society or at least a stable version oI it. Yet the

Page 35
Iailure itselI produced a new set oI social relations that could again be reinterpreted in terms oI a new model oI opposition between state and society. Thus in the
context oI the Iailure oI "normalization," the original "new evolutionist" conception remained the basic Iorm oI orientation Ior theorist-activists such as Michnik.
Undoubtedly the Iact that it was now the turn oI the martial law state to practice (reluctant) selI-limitation reinvigorated the idea that an independent society could
somehow be deIended. "Independent civil society" was not, according to Michnik, annihilated. "Instead oI resembling a Communist system aIter victorious
paciIication, this situation resembles a democracy aIter a military coup d'etat."
14
Despite the reappearance oI martial metaphors such as "a dramatic wrestling match between the totalitarian power and a society searching Ior a way to attain
autonomy" and "the stationary war between an organized civil society and the power apparatus,"
15
the new situation was nevertheless one that indicated the coming
into its own oI the cultural model oI independent society. The major independent activities were publishing, lecturing, discussing, and teaching. For several years, the
hope seems to have been the building oI the moral bases oI democratic structures and practices, i.e., a democratic political culture. While the army-state seemed
powerless against these trends, it was rather successIul in marginalizing its major political opponent: underground Solidarity. The latter, however, linked to the
mechanisms oI independent culture, continued to survive and play a role.
Nevertheless, in this context, the democratic opposition moving within the paradigm oI civil society had to Iace the question oI how and when the survival and even the
dramatic expansion oI an independent culture, more and more pluralized ideologically, could be a Ioundation Ior the reemergence oI aboveground political
organizations capable oI making eIIective demands. The regime's inability to deal with the same economic crisis that was used in 198081 to help erode the resistance
oI the population provided new opportunities Ior the opposition. The strategy to restore the regime's legitimacy through a relatively Iree reIerendum, and thereby to
recover Ireedom oI action to impose an austerity program, Iailed in 1987 in the Iace oI an only partially organized opposition. In this context and that oI the strike
movements during the spring and summer oI 1988, it became clear that the regime needed

Page 36
partners to be able to initiate signiIicant policy, and that only a reconstituted Solidarity could command suIIiciently wide loyalty to become a credible partner.
From the point oI view oI Solidarity's leadership, given the economic crisis and the prospects oI simultaneously weakening both regime and opposition in a continuing
process oI polarization, it would certainly have been counterproductive not to promote and utilize reIorms Irom above, as long these involved real gains in
institutionalizing a genuine civil society.
16
AIter the negotiated "resolution" oI the second strike wave, the issue seems to have been the Iollowing: Could the regime
yield enough concessions that would be adequate trade-oIIs Ior legitimizing the deep austerity measures required Ior successIul economic reIorm? While such
concessions even minimally had to involve legalizing elements oI civil society, it was not at all clear that a version suIIiciently democratic Ior the population and still
acceptable to elements oI the regime could be Iound. It was not clear, Iurthermore, whether the minimum unity oI a society with diIIerent interests and increasingly
diIIerentiated ideologies could be maintained even in an emergency situation in which there were no longer any alternatives other than radical change or social decay.
But could radical change still be conceptualized within the Iramework oI opposing civil society to the state?
16a
The Ideology of the "Second Left" in France
It is not only under authoritarian regimes that the problem oI democratization gets posed in terms oI the reconstruction oI civil society. The category was revived in
France in the mid-1970s as a prime reIerent Ior democratic projects on the part oI signiIicant groups oI intellectuals and a variety oI collective actors.
17
OI course it
was here that the critique oI totalitarianism and sympathy Ior East European dissidence had the greatest intellectual importance.
18
And here, too, totalitarianism was
deIined as the absorption oI independent social liIe oI "civil society" by the party/state, involving the replacement oI all social ties by statized relations. It seems clear
that the French "discourse" oI civil society derived Irom a sympathetic understanding oI developments in the East. But could a

Page 37
category so derived be made applicable to a Western capitalist society with a multiparty parliamentary state?
In France three arguments have been used to justiIy this theoretical move. First, and most like the East, the political culture oI the French leIt (and not only the
Communist party) is seen as deeply connected to the totalitarian phenomenon, i.e., a statist political culture deriving Irom an idea oI revolution based on the Iantasy oI
a society without division or conIlict.
19
Paradoxically, a leIt that represents in its very existence societal diversity, conIlict, and opposition denies just these
presuppositions while hoping to use the state as the instrument oI progress and as the agent oI the creation oI the good society beyond conIlict.
Second, the actual role oI the centralized, modern state in French political liIe is traditionally greater than in most Western democracies. With a good deal oI
exaggeration, one could speak here oI a "totalitarian" statist tendency suppressing many dimensions oI an independent "civil society."
20
Third and Iinally, recalling the
thesis oI Herbert Marcuse, or its more sophisticated French counterpart in the writings oI Cornelius Castoriadis in the 1950s and early 1960s, one might also claim
again with signiIicant exaggerationthat "capitalism has become more 'totalitarian,' engulIing all spheres oI social activity under the single dimension oI economic
activity."
21
The last two theses concerning the state and capitalism converge in another thesis asserting that all autonomous social solidarity is destroyed under the impact oI the
administrative penetration oI society by the (capitalist) welIare state. OI course, this line oI argument does not theoretically assimilate France to a paradigm derived
Irom the analysis oI the East. P. Rosanvallon and P. Viveret warn us that even the three theses taken together do not add up to a conception oI capitalist democracies
as totalitarian in the sense oI Marcuse. But the limitation turns into an advantage: Whereas in the East, in Iully totalitarian society, no internal opposition is allegedly
possible, the totalitarian trends oI French society can be met head on by countertrends involving the reconstruction oI civil and political society.
It is noteworthy that the French discussion has preserved a three-part Tocquevillian distinction among civil society, political society,

Page 38
and state. Civil society is deIined in terms oI social associations cutting across class relations: neighborhood groups, networks oI mutual aid, locally based structures
providing collective service.
22
More dynamically, civil society is seen as the space oI social experimentation Ior the development oI new Iorms oI liIe, new types oI
solidarity, and social relations oI cooperation and work.
23
Political society, on the other hand, is understood as the space in which the autonomy oI groups and the
articulation oI conIlict among them are deIended and the discussion and debate oI collective choices occur.
24
The concept oI political society thus includes the public
sphere as its major dimension, but, given the stress on conIlict (and negotiation and compromise), it is not entirely reducible to it.
Nor are civil and political society to be reduced to one another. To eliminate political society in the conception, or to treat it as civil, is to juxtapose civil society rigidly
to the state. This alternative is variously (and somewhat conIusingly) described by Viveret and Rosanvallon as a choice among liberalism, apolitical and utopian
anarchism, or corporatism as alternatives to statism.
25
Without political mediations, however, the integrity oI civil society in Iace oI the state cannot be indeIinitely
stabilized; the model preIigures a new statist outcome. However, to deIend and extend onlv political society, to seek to politicize all civil structures themselves, leads
to an overpoliticized democratic or autogestionaire (selI-management) utopianism oI which political anarchism and council Communism have been the representative
historical conceptions. It is, however, doubtIul that the Iorms oI selI-organization oI political society can be maintained without the protection and development oI
independent but apolitical Iorms oI solidarity, interaction, and group liIe.
The rigid conceptual division oI civil and political society is diIIicult to maintain in the speciIic Iorm in which it is used in the French discussion. Solidarity and conIlict, as
well as structures oI public communication, are to be Iound on both sides oI the divide. Politically, however, the distinction makes good sense because it implies a
reorientation oI democratic politics away Irom the state to societv without promoting the overpoliticization oI society. Thus the exact translation oI the revolutionary
tradition into the language oI democratic theory is avoided: Viveret and Rosanvallon attempt

Page 39
to think both democratization and the selI-limitation oI democracy. In other words, core components oI the liberal model oI civil society as the sphere oI private,
voluntary association secured by rights are retained in a model that also includes the "democratic" dimensions oI publicity and political inIluence oI nonproIessional
actors, i.e., citizens.
26
The point, however, is not simply to recommend a move (typical oI social democracy) Irom revolution to democratic reIormism. Both poles oI the old duality,
revolution or reIorm, oriented themselves through a structure oI demands to the state
27
and to a society understood in terms oI a class dichotomy. The reorientation to
civil and political society relocates the locus oI democratization Irom the state to society and understands the latter in terms oI groups, associations, and public spaces
primarily. As Claude LeIort argued, the actors the strategy banks upon are not classes but social movements constituted in civil society.
28
These attain a political status
in the conception oI Viveret and Rosanvallon through the mediations made available in political society: the reconstruction oI political parties (replacing the no longer
ideological catchall party) and the renewal oI public forums oI discussion and debate (ending the hegemony oI the established media and oI political communication
that has been reduced to measuring nonpublic opinion, i.e., polls).
The conception oI Viveret and Rosanvallon was designed to promote the selI-understanding oI one dimension oI the French leIt: the so-called "second leIt"oriented
to the Rocard group oI the 1970s in the Socialist party and to the CFDT labor union. As the original conception was Iurther developed, the reconstruction oI civil
society received an even more central role in terms oI the political history oI the period in which the watershed was the Socialist party's coming to power. Civil
society's integrity had to be preserved now even in the Iace oI a socialist-controlled state and political society. Logically, however, since political society was
understood in terms oI mediation between civil society and state, its reorganization presupposed the rebuilding oI more Iundamental social ties. One strong strand in
the then triumphant French socialism could be easily understood to endanger exactly this level through its connection to a Keynesian Iorm oI statism. As Pierre

Page 40
Rosanvallon has IorceIully argued, the welIare state disorganizes above all social networks, associations, and solidarities, replacing these by state-administrative
relations. Not only has the welIare state in the countries oI its highest development proved to be an increasingly ineIIicient and ineIIective strategy oI societal steering;
more importantly, its earlier success has implied a veritable crisis oI solidarity replacing Iorms oI mutuality, selI-help, and lateral co-operation by systemically organized
Iunctions. Thus, the reiIication oI human relations in the context oI social statism Iully matches the eIIects oI the capitalist market economy; a civil-society-oriented
program must thereIore represent not only a third way between social statism and neoliberalism but a way qualitatively diIIerent Irom the other two, which, despite
their opposition, are seen as resembling each other in their eIIects on solidarity relations.
What is extremely vague in the analysis is the nature oI the civil-society-based alternative, beyond the demand Ior a "thicker civil society" involving the creation oI new
networks, new Iorms oI intermediation and association, as the sources oI local and Iace-to-Iace solidarity. Evidently such a general premise is compatible with very
diIIerent Iorms oI civil society. Rosanvallon notes the Iailure oI the communitarianism oI the 1960s and 1970s and seeks to avoid a corporatist version oI the return to
society.
29
He is, however, skeptical concerning the very possibility oI a theoretical answer to the problem oI reconciling individual autonomy and new spontaneous
Iorms oI solidarity, i.e., concerning a model beyond statism, neoliberalism, corporatism, and communitarianism. In general, he convincingly asserts a complementary
relation between a (nonregressive) reduction oI demands on the welIare state and the building oI new Iorms oI sociability. His list concerning the latter, however, is
limited. He notes the existence and importance oI new Iorms oI privately based collective service and oI underground Iorms oI nonmarket, non-state-oriented
structures oI economic liIe,
30
but he understands these as only the Iirst and most primitive Iorms oI what is required. The need Ior new types oI socially generated legal
structures, neither statist nor individualist, is powerIully asserted, but we Iind out little about the nature oI such law or its relationship to existing private and public law.
The projects oI building new social norms, new cultural identities, and

Page 41
a new public sphere are vaguely postulated, but we do not Iind out much about the relationship oI new social actors (movements) to any oI these. Moreover, there is
some serious ambiguity here about the relation oI solidarity and conIlict in constructing a new Iorm oI sociability.
The analysis is more convincing in its treatment oI the problem oI compromise. Rosanvallon postulates the need Ior compromise (1) with capitalist entrepreneurs
(exchanging rationality and mobility in the use oI capital Ior selI-management and Iree time), (2) with the bureaucratic state (exchanging reduction oI demands Ior the
recognition oI Iorms oI autonomous collective services), and (3) within society itselI, involving the construction oI new, democratic Iorms oI public debate, negotiation,
and interest aggregation. It remains unclear, though, how the two post-welIare-state, post-Keynesian, post-social-democratic projects mentioned, regulation by selI-
management and intrasocial regulation, would have a Iundamental eIIect capable oI generating the Iorce behind those Iorms oI compromise. The relationship oI these
projects, presumably representing political (selI-management) and civil (intrasocial regulation) society, respectively, is highly unclear. Here political society is
introduced not so much as a political rearticulation oI civil society but rather as a competing model altogether. But the notion oI political society oscillating between
public discussion and selI-management shows its problematic nature, since the latter notion threatens to assimilate political society to the world oI work or, at best, to
industrial democracy. Correspondingly, the idea oI intrasocial regulation oscillating between individualistic and solidaristic conceptions oI civil society threatens to
surrender part oI what has been achieved: the critique oI the statist logic oI individualism. While the protection oI individual rights has its legitimate place in the
normative conception oI a modern civil society, just as industrial democracy can be reconceived in a way perhaps analogous to a democratic political society,
31
the
moments that need to be stressed in the context oI the critique oI the staII zing and economization oI society, as Rosanvallon recognizes, are solidarity and publicity.
UnIortunately, it is their all too crucial relation that is leIt underdetermined by the tradition oI French analysis we associate with the term "the second leIt." It may very
well be the case

Page 42
that the eventual emergence oI Iorms oI neoliberalism in this milieu can be traced among other things to the theoretical weakness oI the original conception, i.e., to the
diIIiculty oI Iormulating adequate concepts oI civil and political society and their relationship.
A Theory for the West German Greens
A direct intellectual relationship to ''antitotalitarian" or antiauthoritarian struggles Ior democracy is not entirely indispensable Ior interpreting the politics oI Western
democracies in terms oI the category oI civil society. A good case in point is West Germany, where, unlike France, the Eastern dissidents have had only a slight and
ambiguous impact. There was also no need here to diIIerentiate radical politics Irom that oI an authoritarian mass party in the Leninist mold. To be sure, even in West
Germany one could insist upon some impact oI the thought oI the French "second leIt" (especially through the writings oI Gorz), and one could also stress the statist-
authoritarian and even repressive political culture oI the German Social Democratic Party.
32
Nevertheless, in our judgment two related developments, common to all
the Western democracies including the United States, link the German rediscovery oI civil society to the somewhat earlier one in France: the crisis oI the welIare state,
and the emergence oI a neoconservative critique oI "social statism."
The welIare state has oIten been understood not only as a mechanism oI the repoliticization oI the economy but also as a dissolution oI the sharp boundaries between
state and society. However, the crisis oI the welIare state raises doubts concerning the continued eIIectiveness and legitimacy oI state intervention into the capitalist
economy as well as into the various spheres oI civil society: the Iamily, schools, cultural institutions, etc. As a whole series oI radical leIt writers oI the 1970s indicated,
state intervention in the capitalist economy creates insoluble Iiscal and administrative problems in the long run, while political intervention on behalf of the capitalist
economv (especially in the context oI decreasing eIIectiveness) is not easily legitimated in the context oI democratic norms.
33
These projections turned out to be
devastatingly accurate

Page 43
and were in Iact taken up by conservative opponents oI the welIare state under headings such as decline oI productivity, proIit squeeze, dissolution oI tradition and
authority, and ungovernability.
34
However, the original political alternative proposed by some oI the same radical writers, a democratic statism that would exploit the
repoliticization oI economy and society but break its link with the private accumulation oI capital, was abandoned just around the time when the diagnosis concerning
the end oI the welIare-state-guaranteed processes oI growth was conIirmed. In Germany, at least, the reason Ior this surprising development in the selI-understanding
oI one key writer, Claus OIIe, was the emergence oI two distinct programs oI society against the state: the challenges to the welIare state by neoconservatives and by
the new social movements. What the two trends have in common are many aspects oI an economic analysis oI what went wrong with the welIare state. More
importantly, each challenge was ready to move beyond a critique connected to ineIIiciency and dysIunction to develop a distinct, normatively based critique exploring
the negative consequences oI the welIare state, even at its most successIul.
Leaving the economic analysis to the side,
35
the two programs oI civil society against the state that emerged oIIer sharp contrasts. The neoconservative analysis
stresses the erosion oI authority as a result oI the political manipulation oI the nonpolitical spheres oI society, leading to the introduction oI conIlict and controversy into
the very sources oI legitimacy. Authority can be reIurbished, accordingly, only iI uncontestable economic, moral, and cognitive standards are restored. Civil society is
to be restored in this program, but its restoration is understood not only as a deIense against the state but also, more importantly, against politics. The neoconservatives
thus have in mind a model oI depoliticized civil society.
36
In this interpretation oI neoconservatism, the stress is on their identiIying the Ireedom oI civil society with that
oI the market. What remains outside the market must be reintegrated through a conservative retraditionalized cultural model and liIeworld that itselI will help to
integrate market society. However, it is also evident that their model seeks to strengthen the state, speciIically an authoritarian version oI it.
37
The boundaries oI state
and society are to be redrawn in their model in order to provide Ior a smaller but

Page 44
stronger state streamlined Ior Iewer but Iar more eIIective and authoritarian Iorms oI action. Despite explicitly aiming at such an outcome, the neoconservatives have
managed to channel and Iocus a good deal oI antiauthoritarian political sentiment produced by the various consequences oI the welIare state Ior diIIerent spheres oI
liIe.
An alternative program Ior the restoration oI civil society, according to Claus OIIe, must begin by recognizing that "social statism" or "welIare statism" did indeed have
disastrous consequences Ior whole strata, Ior Iorms oI liIe, Ior Iorms oI participation, solidarity, and autonomy. Here his analysis duplicates those oI French "second
leIt" critics oI statism. The program oI the new social movements Ior the reconstitution oI civil society that OIIe calls one oI nonstatist socialism
38
makes no
concessions to economic privatism or to statist authoritarianism. This program
seeks to politicize the institutions oI civil society in ways that are not constrained by the channels oI representative-bureaucratic political institutions, and thereby reconstitutes a
civil society that is no longer dependent upon ever more regulation, control, and intervention. In order to emancipate itselI Irom the state, civil society itselIits institutions oI
work, production, distribution, Iamily relations, relations with nature, its very standards oI rationality and progressmust be politicized through practices that belong to an
intermediate sphere between "private" pursuits and concerns, on the one side, and institutional, state-sanctioned modes oI politics, on the other.
39
Two not entirely consistent Ieatures oI this conception need to be stressed. Behind it lies a deIense oI modern but postmaterial values inherited Irom the new leIt oI the
1960s that contrast participation, autonomy, and solidarity with consumption, eIIiciency, and growth. Thus the model oI civil society here is that oI a culturally deIined
Iramework oI the social, to be distinguished Irom economic and political models. On the other side, however, is a model oI a civil society inherited Irom the
antiauthoritarian dimension oI the Marxian tradition, involving a democratization mainly oI the world oI work. This model is one that French writers tended to call that
oI political society, and OIIe's deIense, unlike theirs, separates the case Ior political and civil society in terms oI alternative and opposed leIt and neoconservative
scenarios. Civil society in the sense oI

Page 45
Rosanvallon and Viveret is here identiIied with the private, and correlatively anything not leIt to the private is to be politicized. Moreover, the new "political" society is
understood by OIIe to represent a model oI democracy alternative to the institutions oI liberal democracy, even iI it remains unclear whether we are to see the two as
opposed or potentially complementary.
The program Ior the restoration oI civil society that OIIe represents has, to a greater extent than that oI the writers oI the French second leIt, preserved its links to the
classical Marxian conception that places political economy within civil society. The model oI politicized civil society recapitulates Marx's early stress on the
reinterpretation oI political democracy and everyday liIe. Even more important, OIIe operates within the terms oI the Marxian critique oI liberal democracy. In his
conception, liberal democracy represents a mediation between state and civil society that is in our time on the verge oI Iailure. Here civil society, however, means
capitalist bourgeois society, and liberal democracy (a particular version oI "political society") is identiIied also as a mediating principle between two supposed
incompatibles, capitalism and democracy.
40
Following Macpherson, OIIe points to the competitive party system as the speciIic mechanism that accomplishes
mediation between state and civil society, reconciling democracy and capitalism in the process. Along with the crisis oI the welIare state, however, the major
contemporary institution oI the competitive party system, the catchall party, has entered into crisis: it never could (unlike its Iorerunners) generate collective identities,
and in a zero-sum society it can decreasingly satisIy the interests oI its diverse constituency when this happens.
The conIlict between democratic legitimacy and nondemocratic economic order can be resolved in one oI two "extrainstitutional" directions,
41
one (representing
governing elites) antidemocratic; the other (representing ordinary citizens) radical democratic. Neocorporatism represents the Iirst type oI solution Ior the articulation
and resolution oI conIlict outside liberal democratic channels. With private organizations taking on public Iunctions, OIIe depicts neocorporatism as a higher degree oI
Iusion between state and society, public and private, than state interventionism itselI.
42
This idea parallels the view oI Viveret and Rosanvallon, according

Page 46
to whom neocorporatism means the disappearance oI political society as such, i.e., all mediations between civil society and state stabilizing their diIIerentiation.
The radical democratic "extrainstitutional" solution Ior the Iailure oI liberal democracy has the opposite consequence: rediIIerentiation rather than Iusion. The
revitalization oI political society or oI a political version oI civil society in the Iorm oI citizen initiatives and social movements represents a renewed model Ior the
diIIerentiation oI state and society. OIIe variously and somewhat inconsistently depicts this option as a response either to the Iailure oI the party system or to the
success (but exclusionary tendencies) oI neocorporatism. In either case, however, we can speak oI the reconstitution oI civil (or political) society outside an
established institutional Iramework that has threatened the disappearance oI all independent Iorms oI social liIe.
The bases on which (political) civil society is reconstituted, iI a Iusion between the spheres oI state and societv has already occurred, remains unclear in this analysis.
43
Since no revolutionary rupture is being contemplated, one must somehow discover the Ioundations oI the new independent structures in the old society on the level
oI norms and/or nonstatized Iorms oI association.
44
OIIe's model oI the reconstitution oI civil society is more emphatically movement-centered than the other two Iorms
oI analysis we have so Iar depicted. Social movements play a major role in all oI them, but only in OIIe's model is there a shiIt oI emphasis toward movement politics
Irom two directions: nonpolitical associations, institutions, Iorms oI liIe on the one hand, and liberal democratic, parliamentary politics on the other. While the issue may
be one oI stress rather than omission, the relationship oI a political version oI civil society to its nonpolitical associational substratum is hardly explored (though without
this the origin oI movements cannot be thematized), while that oI the two paradigms oI politics is explored only in an inconclusive way.
Along with the realist Iaction oI the Greens, OIIe, oI course, presupposes in practical politics the complementarity oI party and movement Iorms oI organization, oI
parliamentary and grass-roots Iorms oI politics. His earlier critique oI liberal democracy, however, oscillated between a conception that asserted an outright contra-

Page 47
diction between liberalism and democracy and another positing liberal democracy as a deIicient democratic bridge between the will oI citizens and the state. Both
versions still leave the way open to the secret hope oI the classical Marxian theory: a political society embodying all economic and political powers in a single
institutional Iramework.
45
Such a utopia beyond the dualism oI state and civil society needs no bridge between the two poles, least oI all a liberal democratic one.
Under the impact oI the new selI-limitation oI contemporary social movements, which seek to limit but not abolish the existing version oI the modern state, OIIe no
longer seems to hold this particular utopian view. His critique oI majority rule
46
allows him to thematize the relationship between the "extrainstitutional" political impulse
oI the new social movements and the need Ior constitutional change within the structure oI liberal democracy. Since this critique is actually aimed at the centralized
Iorms oI majority rule represented by the liberal democratic nation state, OIIe proposes to supplement majority rule not so much with the classical liberal Iorms oI the
protection oI minorities as with various Iederal, decentralized, quasi-aristocratic (in the sense oI selI-elective bodies oI those most concerned), and elsewhere also
Iunctional representative Iorms. OI course, all oI these supplementary Iorms oI democracy would have to rely on some Iorm oI majority rule. What remains unclear
about the analysis is again the problem oI the relation oI the two political societies, this time the centralized and the supplementary ones, and, in particular, how the
oIIicial, institutional, centralized Iorm is to be transIormed or at least made receptive to and capable oI being influenced by the other Iorms. While the suggestion to
make majority rule reIlexive about its own boundaries through a reinstitutionalization oI the pouvoir constituant is important, this (still vague and possibly impractical)
proposal bypasses the question oI the structure oI parliamentary, party democracy. We are leIt with the impression (also present in some oI the other analyses we
have presented) that while liberal democracy is admittedly dangerous Ior the autonomy oI a political version oI civil society, because oI its depoliticizing tendencies,
civil society cannot in the long run be institutionalized without some oI the structural possibilities that, in the West at least, are carried by liberal democracy.

Page 48
Civil Society in the Transition from Latin American Dictatorships
The concept oI civil society has also emerged under several "bureaucratic-authoritarian" regimes as a key term in the selI-understanding oI democratic actors as well as
an important variable in the analysis oI the transition to democracy.
47
This discussion has been the richest, most open-ended, and most synthetic among the ones so Iar
discussed. We can, oI course, only trace Iorms oI discourse that we believe indicate the beginnings oI a new political culture; it is beyond our competence to integrate
this discourse into the diverse political and social contexts involved. Nevertheless, we are struck by the remarkable unity oI the discussion and by its parallels with
developments elsewhere.
The main concern oI Latin American theorists and their collaborators has been the transition Irom a new type oI military-bureaucratic authoritarian rule: First, involving
a period oI "liberalization" (deIined as the restoration and/or extension oI individual and group rights); and second, a stage oI "democratization" (understood in terms oI
the establishment oI a citizenship principle based on at least a "procedural minimum" oI participation). But these transitions are seen as strongly dependent on the
"resurrection oI civil society."
48
Here, civil society stands Ior a network oI groups and associations between (in some versions, including) Iamilies and Iace-to-Iace
groups on one side and outright state organizations on the other, mediating between individual and state, private and public. DiIIerent Irom clan, clique, cabal, and
clientele, the associations oI civil society have themselves a public, civic quality related both to "a recognized right to exist'' and the ability "to openly deliberate
about . . . common aIIairs and publicly act in deIense oI justiIiable interests."
49
Others signiIicantly add the notion oI selI-expression to that oI the representation oI
interests, and they propose to include movements along with recognized associations in the concept.
50
It is oIten suggested that the "resurrection" oI civil society
culminates in the highly mobilized and concentrated Iorm oI "mass mobilization" and "popular upsurge," in which the various layers and strata oI civil society develop, iI
temporarily, a single collective identity.

Page 49
The category oI mass is misleading here Ior two reasons. First, the analysts tell us that in liberalized authoritarian states, civil society typically comes into motion in
distinct and successive layers: intellectual groups, middle-class organizations, human rights organizations, proIessional associations, movements oI industrial workers,
etc. (not necessarily in this order).
51
Even in contexts oI high mobilization, in the recent transitions to democracy the diIIerent groups, associations, and organizations
do not coalesce into one mass, as was characteristic oI the earlier "populisms" that oIten led to dictatorships. Second, the Iorums oI resurrected civil society are
typically "public" as against "mass," ranging Irom intellectual discussions in universities, bookstores, caIes, etc., to popular Iorms oI association and assembly, which
together represent new contexts in which "the exercise and learning oI citizenship can Ilourish in deliberations about issues oI everyday concern."
52
High levels oI
mobilization against recent dictatorships typically used rather than bypassed these public Iorms. This is understandable, since aIter the authoritarian reduction oI public
discussion to state-controlled, restricted ''codes and terms," the restoration oI this sphere achieved high signiIicance, Ior a while at least making the simpliIications
involved in populist discourse less attractive. All the same, the distinctions between higher and lower levels oI mobilization, as well as between uniIied and more
particularized collective identities in civil society, remain important.
Leaving aside some diIIerences among the relevant authors concerning the very meaning and the relative importance oI the concept oI civil society, some important
puzzles and ambiguities characterize the whole line oI analysis. According to an interpretation characteristic oI the most repressive regimes, such as Argentina,
authoritarian regimes atomize, depoliticize, and privatize society, creating a purely manipulated and controlled public sphere.
53
According to another, in some contexts
at least (such as Brazil), civil society or its residues survive authoritarian rule in Iorms oI interest associations, autonomous agencies, local government, and church
liIe.
54
According to a third line oI interpretation, the "resurrection oI civil society" that pushes the democratization process Iorward is possible in either case, with or
without surviving Iorms oI recognized association, with or without

Page 50
memories oI earlier mass mobilization.
55
As Francisco WeIIort Irom Brazil puts it, "we want a civil society, we need to deIend ourselves Irom the monstrous state in
Iront oI us. This means that iI it does not exist, we need to invent it. II it is small, we need to enlarge it. . . . In a word we want civil society because we want
Ireedom."
56
In this interpretation, which recalls arguments made in Poland, the social Ioundations Ior civil society, starting with Iamily and Iriends and continuing with
the church, never disappeared in any oI the southern dictatorships.
The strategy oI "inventing" and "enlarging" is Iavored by the Iact that bureaucratic-authoritarian regimes never manage to solve their problems oI legitimacy.
57
The
constitution or reconstitution oI elements oI civil society, indirectly promoted by reducing both Iear and the costs oI autonomous activity, becomes a means to address
these Iundamental problems.
58
While this eIIort Irom above is always expected to stay within careIul limits, it cannot amount to a complete Iarce iI the goal oI
legitimacy is to be attained, and the elements oI actual democratization that are established in this way are by deIinition unpredictable and cannot be kept within any
given predeIined limits.
59
It is still unclear, however, what diIIerence the state oI development oI civil society under authoritarian rule makes in terms oI the process oI transition or the stability
and character oI the outcome. It seems likely that the character oI a mobilized civil society itselI is aIIected by the alternative patterns: more homogeneous where no
previous structures existed or were preserved, more pluralistic and structured where civil society did not have to be created aIter a high degree oI atomization. And
this diIIerence has many potential consequences.
It maybe helpIul to distinguish, in relation to transitions, processes oI initiation, consolidation, and completion. The exact role oI civil society in the process oI
initiating the transition remains in some dispute. The dominant thesis stresses, on the basis oI much comparative data, that the beginning is primarily a Iunction oI internal
splits in the authoritarian regime, although all analysts concede that iI such a split leads to an "opening" or to liberalization, the resurrection oI civil society cannot be
easily contained and will play an important role in all succeeding steps.
60
However, some inter-

Page 51
preters seem to argue that where mobilization plays a role in the end oI an authoritarian regime, the whole process oI the "overthrow" or "selI-dissolution" Irom the very
beginning is very much a Iunction oI the regime's relationship to civil society.
61
The notion that the problem oI legitimation is the Achilles heel oI the post-1945
authoritarian regimes
62
seems to imply that the instability oI the regimes and the impetus Ior liberalization should be sought in the relationship oI the rulers to groups and
opinion outside oI them.
The Ieatures oI civil society are as important to potential rollbacks, in particular military coups, as to the process oI initiation and acceleration. While some analysts Iear
overmobilization as a pretext Ior coups and a motivation Ior reuniIication oI the ruling elites, the dominant position seems to stress the costs oI a conIlict with mobilized
civil society as a deterrent to hard-liners that reIormers can use.
63
One might add here that not only the level oI mobilization but that oI structure Iormation is important
because it is easier to suppress a society without deep organizational roots than a highly articulated one, even iI the Iormer is superIicially mobilized.
Equally important is the issue oI whether or not the pressure oI civil society, once mobilized, is capable oI pushing to the end a process oI transition to democratic
politics. It seems obvious that an evolutionary strategy involves important negotiating and bargaining processes with those authoritarian rulers who are able and willing
to moderate their rule, while at a later stage any transition to democracy must involve organization Ior elections. It is not obvious in either oI these contexts, however,
how civic associations, social movements, grass-roots organizations, or even media oI communication can substitute Ior the diIIerentiation oI a political element
capable oI strategic considerations. In Iact, a strategy Irom below on its own has nowhere succeeded.
Aside Irom ideologies oI reIorm Irom above, two Iorms oI discourse are available to participants seeking to understand the place oI political organizations in the
transition Irom authoritarian rule; one is dialectical and the other more analytical. According to the Iormer, since bureaucratic-authoritarian regimes suppress or
seriously deIorm all types oI mediation between the private sphere and the state (including popular organizations as well as institu-

Page 52
tions Ior political citizenship), the task oI democratization is primarily to reconstitute these.
64
Indeed, the dialectical version oI the discourse oI civil society oIten comes
to identiIy democratization with the reconstitution oI these mediations. In this version, the political actors capable oI interposing themselves between society and state
emerge Irom the process oI organizing new social associations and movements as their organic continuation. But in their search Ior legitimacy, the regimes themselves
oIten initiate the process oI reconstituting mediations beyond the semipolitical, state-constituted "bureaucratic rings or clusters" oI "social interests" that have Iailed as
eIIective replacements Ior societal pressure groups.
65
As a result, those in opposition Iind themselves having to choose between "the imbecility" oI reIusing degrees oI
social autonomy simply because they are oIIered or even accepted by governments and "the opportunism" oI accepting limited autonomy too quickly, entering into a
predetermined and coopting game without testing the actual possibilities oI democratization.
66
One option beyond these two seems to be the attempt to organize and
deIend the new sphere oI civil society not as mediation but as an end in itselI, as in itselI political: ''II politics was to have a new meaning, a new sphere oI Ireedom Ior
political action had to be developed. For political Brazil, civil society, previously either ignored or seen as an inert mass, began to signiIy that sphere oI
Ireedom."
67
From this point oI view, it is natural to treat even political parties and associations as undiIIerentiated parts oI the heterogeneous Iield oI selI-
organization.
68
In an extreme anti-political version in Brazil, combining the views oI "lay anarchism and Catholic solidarity thought," parties are to be more Ieared than
trusted because oI their propensity to enter the game oI the state. To the extent that selI-organization had to be complemented by policies and legislative measures,
these were to be achieved by movements oI direct participation organized around single issues oI intense concern to their own constituencies.
69
In the Iace oI intact authoritarian power, however, a high level oI mobilization without mediations, symbolized by the Iigure oI civil society as "the political celebrity oI
the abertura,"
70
could have demobilizing consequences. Unable to go beyond polarization, civil society can deIeat state initiatives without generating a compre-

Page 53
hensive alternative oI its own. As in the cases oI both Brazil and Chile, Iear oI the regime can easily be replaced by society's Iear oI itselI, Iear oI the consequences oI
its own impotent power.
71
Both in theory and in practice, a second strategy comes to stress the need Ior an orientation to political society to complete the transition to
democracy. This strategy is intellectually analytical in that it does not see the institutions oI political societyparties, electoral mechanisms, Iorms oI bargaining, and
legislaturesas either parts or as organic continuations oI the processes oI the selI-organization oI civil society.
72
While it seems misleading to identiIy civil society primarily with liberalization, and political society primarily with democratization, it is certainly right to insist that "Iull
democratic transition must involve political society."
73
Without political society, neither the necessary negotiation Ior transition nor the mechanisms oI societal control oI
postauthoritarian states can be established. This has been shown through analyses oI elections and political parties. In those dictatorships where electoral mechanisms
were maintained, even iI greatly restricted, it has been possible to channel social pressure in the direction oI substantial, iI gradual, political change
("decompression"),
74
even in the context oI an intact authoritarian order that has not been weakened Irom the outside. This was the case in Brazil. Similarly, the
continued, iI restricted, existence oI political parties represented in several countries, Irom Brazil to Uruguay and (most recently) Chile, the natural Iocal point Ior
negotiated transitions.
75
Indeed, parties and elections represented opportunities Ior the remobilization oI civil society in several contexts where phenomena oI
demobilization occurred aIter Iailures oI early challenges against authoritarian rule.
76
Wherever it has been possible, the activation oI political society seems to have
been the key to avoiding polarized, zero-sum, or even negative-sum conIrontations between organized civil societies and authoritarian regimes that have maintained
some continuity with the past.
77
Whatever its necessity, the turn to political societv has potentiallv demobili:ing consequences with respect to civil societv, as many participants and observers
have noted. In this context, Cardoso justly calls attention to the double nature oI political parties: Their mediating role is made possible by, but cannot overcome, the
contradictions

Page 54
within them oI movement and administration, oI participation and elitism, oI democratic norm and strategic calculation.
78
At two points, however, the elitist,
administrative, and strategic side may dominate: pacts and elections. OIten possible and necessary as "undemocratic" halIway stations, pacts are rightly stressed by
many as important means oI avoiding violence and its risks in the transition to democracy.
79
It does not seem completely justiIied, though, to claim that, where they are
possible, pacts between the parties oI the opposition and elements oI the regime are also desirable, especially when it is a little too quickly admitted that they are as a
rule exclusionary, nonpublic, and aimed at drastically curtailing conIlict in the political system. Their violation oI the norms oI democracy
80
can have long-term negative
consequences Ior a political culture. With this said, it should perhaps be added that pacts in which certain interests oI the existing rulers are guaranteed have diIIerent
possible consequences Ior civil society, depending on their timing. Coming early in a process oI transition, pacts can secure elements oI liberalization, making possible
the reconstitution oI civil society. In this case, with the emergence oI new actors and the activation oI public spaces, the chances are good that the initial pact will
eventually be swept aside.
81
II a pact comes late, however, aIter the resurrection and possibly the upsurge oI civil society, and especially iI it guarantees power
positions to all contracting parties, including some oI the opposition, its very aim involves an exclusion and demobilization that may be successIul Ior a long period.
OIten the consequence is a revival oI populism rather than processes oI Iurther democratization.
The only "late" pacts that seem to avoid this trajectory are those in which oppositional groupings ask Ior no concessions Ior themselves but only Ior society as a whole.
Above all, pacts that arrange elections and electoral rules can have this character. But elections, even when they themselves do not incorporate strongly exclusionary
rules, can be ambiguous Irom the point oI view oI mobilizing civil society.
Several analysts ask the partially rhetorical question, Why should ruling elites agree to elections that are likely to abolish their rule? The answer given is that these elites
expect to channel politics "away Irom the ebullience oI civil society" and perhaps even to win

Page 55
elections by dividing the opposition and being rewarded by the electorate.
82
When elections are only gradually decontrolled, as in Brazil, the hope is to slow down
the rate oI change while still achieving procedural legitimacy. The hopes oI victory and legitimacy are generally Irustrated, but not those oI demobilization and, where
pertinent, gradualism.
83
The move to electoral parties with their less intense, more inclusive, more abstract Iorm oI political identiIication and their lower degree oI
direct participation tends to devalue and replace movements and associations with their more particular, but also more intense and participatory, Iorms oI organization.
Although this depends on the speciIic electoral rules enacted, the tendency oI modern elections is to reduce the number oI political parties capable oI eIIectively
participating in elections. In turn, and especially in periods oI uneasy transition, potentially successIul parties will oIten restrain movements oI civil society that might
jeopardize the outcome or even the possibility oI elections.
84
The major parties, moreover, share a common interest in obtaining a larger than representative share oI
votes Ior Iorces close to the authoritarian regime, to avoid an overly great victory Ior the opposition.
85
Thus, it can be said not only oI the processes leading to
unrestricted electoral contests that end dictatorships but also oI the elections themselves that they are implicit negotiations between regimes and oppositional parties
that provide space and time to "redeIine their respective roles."
86
And while the weak legitimacy and the plebiscitary possibilities oI partially restricted elections can
indeed lead to societal mobilization and to learning processes outside the oIIicial Iramework, the liberal democratic legitimacy oI open conIrontation provides much less
oI a chance Ior such an outcome. It is possible that where civil society is underdeveloped and passive, or is in the process oI contraction, elections might draw
otherwise uninvolved strata into organized politics; in the context oI a highly mobilized civil society, the reverse may very well occur, with parties turning out to be "not
only, or not so much, agents oI mobilization as instruments oI social and political control."
88
There is little doubt aIter the experience oI several countries that the highest level oI a mobilized civil society cannot be maintained Ior long.
89
But is civil society
equivalent to such mobilization? Is it

Page 56
not a mark oI its weakness that it can exist in some countries only in this Iorm? There is some serious theoretical uncertainty concerning what comes or can come aIter
demobilization. The question is whether there is anything leIt oI a "resurrected civil society" aIter selective repression, co-optation, manipulation, internal conIlicts,
Iatigue, disillusionment, and the channeling oI opposition into the party and electoral systems take their toll and demobilize "the popular upsurge."
90
Here one
interpretation stresses depoliticization, reprivatization, and the emergence oI political ghettoes, which together will endanger democratic consolidation and weaken the
society's ability to resist renewed authoritarianism. The idea that in some countries, notably Chile and Uruguay,
91
an overdeveloped system oI parties contributes to a
dependent and underdeveloped civil society is more consistent with this line oI argument than is the stress in the case oI other countries on the survival oI civic
associational liIe even under authoritarianism. II one identiIies demobilization with the atomization oI civil society, it is hard to see how one can speak oI a transition to
democracy rather than a return to cycles oI democracy and dictatorship, neither oI which can be stabilized, in part because oI the cycles oI politicization and
depoliticization oI civil society within each Iorm oI rule. The idea oI Iinally leaving the cycle
92
must thereIore point beyond the alternative oI a Iully mobilized and Iully
depoliticized and privatized civil society.
Logically, at least, the demobilization oI a popular upsurge is not necessarily the end oI a politically relevant civil society. Nor is it necessary that everything learned in
previous cycles be Iorgotten. In this context, it is signiIicant that some interpreters see the emergence oI a new Iorm oI diIIerentiation between de Iacto societal
pluralism and democratic pluralism as a change in values, as the transIormation oI the collective identity oI groups and institutions.
93
The Iormer type oI pluralism has
been present in most oI the societies in question, but the latter has been a product only oI the recent struggles against authoritarian regimes that have led to the
replacement oI the imagery oI the via revolucionaria by democratic ideologies.
94
AIter the Iailure oI illusory revolutions and the experience oI dictatorships,
democracy came to be increasingly viewed as an end in itselI rather than a means Ior the

Page 57
realization oI sectoral interests.
95
But Ior it to become an end also Ior nonelite groups, a reorientation to civil society had to and actually did occur. "The discovery oI
the value oI democracy is inseparable, within the opposition, Irom the discovery oI civil society as a political space."
96
The question inevitably arises, What will happen
to the value oI democracy as the space oI civil society shrinks to the beneIit oI political society?
Actually, one should distinguish three possibilities: (1) a civil society that loses its value Ior social actors with the restoration oI democracy, a process in which political
society has come to play the major role; (2) an overpoliticized civil society that implicitly, on behalI oI various oI its sectors, seeks to abolish societal plurality itselI
and/or devalues mediations between itselI and the state; and (3) a civil society that has become reIlexive to itselI through its selI-thematization and selI-normatization,
as well as its selI-limitation vis-a-vis political society.
The selI-reIlexive model oI civil society involves not only the idea oI the selI-limitation oI civil society but also its own strengthening. This has consequences Ior both
civil and political society. The model is incompatible with the liberal-individualistic concept oI civil society that implies both its Iull depoliticization and its dependence
on the Iorces oI the market economy: "the social inequality and the Iragility oI the individual beIore business and the bureaucracy." Cardoso proposes an alternative
combining the radical democratic stress on collective subjectivity and selI-organization (without, however, abandoning individual rights) and a reIorm democratic
acceptance oI the necessity oI the state. This "dualistic" synthesis leads to the start, admittedly needing Iurther development, oI a proposal Ior greater social
responsibility on the part oI the management oI Iirms and the bureaucracy, with increasing public control over their processes. Without this, civil society remains
deIenseless and "private in the strict sense oI the word."
97
This redeIinition oI the relationship oI state and civil society in a democracy yet to be created alters the model oI political society as well, and along with it that oI
political parties. Their task now becomes building "movable bridges on both sides oI the antinomy."
98
The idea is not well enough explained in terms oI the notion oI
"countering the widespread idea that the parties are 'inauthentic'

Page 58
and incapable oI serving as a Iilter Ior the aspirations oI the electorate."
99
What seems to be involved instead is the rejection oI a choice between the elitist and the
radical-democratic, between the strategic and the normative-democratic dimensions oI the ambivalence oI modern parties. Rather, it seems to be this ambivalence
made conscious that could allow both the sensitizing oI civil society to the need Ior strategic considerations and the introduction oI elements oI democratic decision
making into state and Iirm.
100
Sketchy as it may be, Cardoso's outline Ior the development oI democratic theory has several virtues. It is a model oI the goal oI transition that does not lose sight oI
the preconditions oI consolidating democracy and remobilizing in its deIense. It corresponds well to the institutional requirements Ior O'Donnell's notion oI building a
civil-society-based democratic political culture. Finally, the model points beyond restricting democracy to the political sphere (i.e., beyond elite democracy or elite
pluralism) to the possibility oI exiting the historical cycle in away that allows the issue oI "more democracy" to be raised without being a subterIuge Ior a dictatorship oI
the leIt or the pretext oI the dictatorship oI the right.
Revisiting Eastern Europe in the Late 1980s
As indicated above, the rediscovery oI civil society in Poland was the product oI two negative learning experiences: the Iailure oI total, revolutionary change Irom
below (Hungary in 1956) and oI comprehensive reIorm Irom above (Czechoslovakia in 1968). Polish reIormers decided that a radical change oI society was still
possible iI a third route was Iollowed. This would have two components: The agent would be organized society "Irom below," and the target would be civil society
rather than the state, within a program oI selI-limitation. Note that by its own standards the new strategy was itselI open to the test oI new learning experiences. AIter
the repression oI Solidarity in December oI 1981, the question inevitably arose oI whether the third and seemingly last route had also been proved impossible in
Soviet-type societies. (Apparently last on the basis oI a dualistic conception that rigidly juxtaposes state and civil society.)

Page 59
Within Poland the dualistic Iormulation has been subjected to stringent critique by Jadwiga Staniszkis. Here we will outline and expand her general line oI attack:
1. The polarization oI society vs. the state in Poland is connected to a political history in which three Ioreign imperial governments represented the state.
2. Polish culture survived the age oI partitions by preserving its own traditions, mentalities, practices, system oI education, and religion in isolation Irom the state (s).
3. The strategy was, however, always a purely deIensive one and is not suited Ior real social change.
4. The post-totalitarian state is more subtle and penetrating, more invisible and corrupting, than the openly repressive states oI the past. Thus the isolation oI state and
society is in principle not possible.
5. The unity oI society is illusory on the empirical level, and a populist and solidarist uniIormity imposed on society (allegedly the case during the sixteen months oI
Solidarity) is undesirable.
6. The unity oI the party-state is also illusory and, Irom a strategic point oI view, hardly desirable. The notion oI inherent opposition between society and state makes it
impossible to exploit internal cleavages and tensions in state and party. ReIormist attempts Irom above and within the ruling structure must then be taken as a priori
illusory, and compromise can be understood only as strategic, i.e., in principle unstable. Party oppositions are continually driven back into the party.
7. Popular mobilization and conIlict under the aegis oI the dualistic conception can amount only to ritualized Iorms oI channeling opposition; they will not be able to
produce any signiIicant change in the existing system.
101
Staniszkis was wrong about the mobilizing power oI the dichotomous conception oI society against the state. Indeed, the conception was in many respects selI-
realizing: While Solidarity was legal (19801981), Polish society was at least tendentially organized around the Iault lines oI the dichotomy oI civil society and (party)
state, despite conIlicts within each pole oI the duality. In retrospect,

Page 60
however, one implication oI Staniszkis's analysis was IulIilled: The dichotomous conception reinIorced a type oI polarization in which compromise solutions became
impossible, however much desired by the sector oI Solidarity led by Lech Walesa. For compromise one needs partners, presumably reIormists, and also (political)
institutions oI mediation. In a context oI radical polarization, actively sought by sectors oI the regime but Iavored by Solidarity's ideology, neither could emerge. The
normatively and aIIectively successIul dualistic conception oI the original project oI the selI-liberation oI civil society was thus part oI the constellation that led to
strategic Iailure.
In the 1980s this project was, amazingly enough, not only not abandoned but extended to two other countries: Hungary and the Soviet Union. Two reasons, aside
Irom that oI the inherent normative validity oI the basic ideas, were responsible. One was geopolitical: Important shiIts had occurred in the international economic and
political environments in which the project had originally led to stalemate. The other was theoretical, involving an expansion oI the original Iramework by introducing
the category oI political society.
The change in the international environment Iollowed Irom the crisis oI the Soviet model oI economic development both on the periphery and even in the center oI the
imperial system. The Soviet Union had exhausted the possibilities oI extensive development based on continuous expansion oI the resources oI raw materials and labor
and was being decisively challenged by the threat oI unlimited technological-military competition with the United States, a competition the Soviet Union could not win.
102
Aside Irom economics, the new situation was marked by three new processes: the Iailure oI normalization in Poland, the emergence oI reIormism Irom above in
the Soviet Union, and the beginning oI the crisis oI Kadarist consolidation in Hungary.
The reIerence to the Soviet Union already indicates that, given the change oI environment, the strategy oI reIorm Irom below as well as Irom above has made a
comeback, despite the expectations oI Polish oppositionists in the late 1970s under the inIluence oI the Czech experience and the atmosphere oI the Brezhnev era.
Remarkably, the strategy oI reIorm Irom above, initiated by segments oI the ruling party, was now complemented by another one: the

Page 61
reconstruction oI independent civil society. Indeed, it is this complementarity that was oIten seen as the mark oI the diIIerence between radical reIorm and mere
reIorm. According to this line oI thought, attempted changes in the Soviet economy Iailed in the past because (1) they targeted only the economy, (2) they did not go
Iar enough even in relation to the economy, and (3) their only agent was the ruling institution above, excluding all Iorces Irom below.
103
All these points belong
together. Assuming that the goal was Iirst and Ioremost an economic reIorm that went "Iar enough" to work, elite reIormers now argued that this is possible only iI
other areas oI liIe were transIormed and other actors than the party-state participated in the overall project. In eIIect, the claim is that civil society is a part oI the
environment needed Ior a new type oI economic coordination that could not be created without movements Ior, and in, civil society.
The thesis applied not only to the system inherited Irom the conservative Brezhnev era by the Gorbachev team, where even the Iormal abolition oI the command
structure would require the mobilization oI pressure outside the ruling apparatus. It applied as well to the reIormed Kadarist system, whose successes were due more
to partial privatization than to the transIormation oI the command system into one oI inIormal bureaucratic controls.
104
From the writings oI Hungarian economists,
legal scholars, political scientists, and sociologists it becomes clear why civil society was implicated on two levels in what was supposedly required Ior "radical reIorm."
First, we have learned that the introduction oI reIorms exclusively Irom above cannot, because oI conservative-bureaucratic resistance, be Iormulated or implemented
in a suIIiciently consistent manner.
105
Nor is such a process protected against rollbacks initiated by bureaucratic counterattacks in contexts oI even minor leadership
realignments. Thus, independent actors are needed Ior more consistent and determined pursuit oI economic reIorm. However, since social movements are not likely to
be the agents oI economic reIorms (because oI the sacriIices involved), political trade-oIIs Ior movements (unions, Iorms oI industrial democracy, ability to strike) and
the institutionalization oI collective economic actors

Page 62
(legality oI interest representations, new Iorms oI property) are necessary.
106
Second, both the relevant trade-oIIs and the institutionalization oI actors point to laws, rights, and associations oI interest representation. These Ieatures oI civil society
are also needed to counter spontaneous reinvasion or repenetration oI the economy, Ireed Irom the prerogatives oI direct economic command, by inIormal, extralegal
types oI bureaucratic regulation that reinIorce the weaknesses oI the inherited ''economy oI shortage."
107
Laws and rights consistently Iormulated and made entirely
public are needed, along with independent courts and judicial procedures, to provide predictability and regularity Ior economic actors and to protect them against the
discretionary power oI the existing apparatus operating through legal inconsistencies and the gaps and loopholes within the law.
108
But laws and rights alone would be
powerless against administrations whose practice is to bypass all Iormal regulation through their control oI the execution and implementation oI the laws. They must be
backed up by established interest associations and an open public sphere. These are also needed to provide a counterweight against the already established,
monopolistic lobbies (themselves rooted partly in the apparatus and partly in the moderately decentralized structures oI industry) that now control the bargaining
processes involving investment, subsidies, tax exemptions, and even prices and that reinIorce the resource-constrained and built-in wasteIul character oI the economy
oI shortage.
When the reconstitution oI civil society was promoted as a component oI reIorm Irom above, especially in the Soviet Union, it was supposed to stay within careIully
deIined limits. The only institutions oI civil society that were to be reconstituted were those most relevant to economic rationality; the independent actors were to
accomplish only the strictly necessary tasks. But both aims were selI-contradictory. Economic laws and rights become such only in the context oI Rechtsstaatlichkeit
(constitutionalism), with Iar more general implications. Associations genuinely competent to exert open economic pressure are also able and motivated to address
other social and political issues. A public sphere that allow criticism oI economic waste, corruption, and resistance to change cannot

Page 63
easily be prevented Irom taking up other issues. All these departures presuppose the reduction oI Iear in society, and the reduction oI Iear becomes the stimulus Ior
new departures. Finally, movements that can be easily restrained cannot play an important role in overcoming resistance to reIorm, while those that can play such a
role cannot be controlled and are unpredictable. The constant Iluctuation in the Soviet Union between measures that lead Iorward and those that revive past practices,
between democratization and authoritarian centralization, is best explained in these terms. The regime wants radical reIorm, it unleashes and even prods the revival oI
civil society, but it also wants to press its prerogative to determine the limits oI what can and cannot be changed, including the structure and dynamics oI civil society
itselI.
Nevertheless, the process oI social mobilization and the building oI at least some dimensions oI what the actors themselves call civil society continues amidst the
Iluctuation. The level oI societal selI-organization today would have been unthinkable a couple oI years ago. But it is not at all clear that the result will be radical reIorm
rather than hopeless polarization and stalemate.
109
II the pathology oI reIorm Irom above is that it replaces a Iormal command system with one oI inIormal
bureaucratic regulation, the step to civil society supplies only the necessary but not the suIIicient condition oI its cure. As the Poles discovered, even an organized and
mobilized civil society cannot, especially in the context oI selI-limitation, act directly on an unchanged party-state and overcome the resistance oI a political-economic
apparatus whose last major stronghold becomes the unreconstructed bureaucratic economy.
This was the lesson that inspired those who imported the Polish project oI radical reIorm into Hungary, especially aIter martial law. Key elements oI the Hungarian
opposition
110
reIormulated the program in terms oI a radical minimalism that nevertheless implied that changes in society must be complemented by necessary, iI less
radical, change in the party-state sphere. At Iirst, this meant redeIining as rights the elements oI already conceded openness and diIIerentiation in Hungarian society and
redeIining the discretionary state (Massnahmenstaat) as an authoritarian Rechtsstaat that is selI-limiting, at least with respect to the rights it grants. The second
version, developed at the time oI increasing crisis and some

Page 64
success in involving intellectuals in oppositional activity, proposed to independent social Iorces that they demand pluralism in the sphere oI private law (civil society)
and a Iully developed Rechtsstaatlichkeit in the sphere oI public law.
111
Finally, in 1987, at the time when the Ioundations oI the Kadarist system were already
cracking, a detailed model oI radical reIorm was proposed. Appearing under the name Social Contract, this involved the restoration oI civil society in all its
dimensions and a reIorm oI the political system to include elements oI genuine parliamentarism, a responsible government, and a reconstruction oI the place and role oI
the Communist party that would preserve some oI its prerogatives, but only within a Iramework oI constitutional legality. It is the structure, rather than the exact
Iormula, that is important to us, Ior it represented a call Ior discussion, negotiation, and compromise. The partisans oI the Social Contract approach attempted to
reconstruct the dualistic project inherited Irom Poland in terms oI a model linking the radical reconstruction oI civil society with a less radical but nevertheless
principled reIorm oI the political sphere. The idea was not to abandon the goal oI parliamentary democracy but to combine two diIIerent rates oI change, one in civil
society and one in the state sphere, in a mutually reinIorcing way, and to provide at the same time the necessary change oI "environment" Ior institutionalizing a genuine
market economy.
The Social Contract retained an important link to the Polish politics oI the "new evolutionism" by maintaining, against other approaches oI the time that still addressed
the regime or its reIormist elements,
112
that groups, associations, and indeed movements outside the oIIicial institutions would have the primary task oI pushing the
reIorms through. In Hungary, though, the idea was paradoxical, given the absence oI anything resembling the Polish level oI societal selI-organization.
113
Oddly enough, the political results in Hungary turned out to be more radical than in Poland. Indeed, aIter the removal oI Kadar in May 1988, the Hungarian
Communist party made a number oI rapid concessions: a de Iacto open public sphere, a law oI association and right to strike, and a law that allowed the Iormation oI
parties, though not initially as electoral organizations. Moreover, by February 1989 the party conceded the need Ior early competitive

Page 65
and unrestricted elections, and in June 1989 it entered into negotiations concerning electoral rules and procedures with eight or nine protoparty Iormations represented
by "the roundtable oI the opposition."
There are two ways oI reading the logic oI these changes. The Iirst (F. Köszeg) takes the point oI view oI the Iledgling organizations oI independent society and points
to the inner dissolution oI the ruling party (due to economic crisis as well as the destabilizing eIIects oI the Soviet hands-oII policy) that made it too weak to resist even
a relatively small degree oI social pressure. Certainly the thesis seems to be conIirmed by the history oI several key concessions, which began with proposals intending
merely co-optation, continued with intense public criticism, and ended with the regime backing down.
114
But this reading does not leave enough room Ior an
important actor outside the opposition, namely the reIorm groupings within the party, which played an active role in several oI the same concessions.
The second reading (J. Kis) sought to correct this underestimation by stressing the attempt on the part oI the increasingly dominant reIormist Iaction to Iind legitimate,
viable partners in society Ior instituting economic reIorms along with new austerity programs. The search Ior partners might itselI have led only to an attempted co-
optation oI the social Iorces in Iormation, but the necessity oI viable partners, given the decline oI the regime's legitimacy, required genuinely independent entities
operating in an open, competitive political terrain.
115
In this analysis, the search Ior partners led the regime, or its dominant Iaction, to the opening oI the space Ior the
emergence oI political society.
It is instructive to compare this situation to the 19801981 period in Poland. Then it was Solidarity that sought a "historic compromise" with the regime, unsuccessIully,
involving the creation oI institutions oI mediation.
116
Its own polaristic conception, and the regime's belieI in the possibility oI "normalization" and in its powers to enact
economic reIorm, played major roles in the Iailure oI compromise. Perhaps at that time, as opposed to 1988, Solidarity, having behind it all oI society, was so strong
that the regime could not allow it any genuine role in the making oI policy. By 1990 important elements oI the old regimes themselves both in Hungary and Poland had

Page 66
accepted the idea oI Iar-reaching compromise with relatively weaker opponents, and this involved the creation oI institutions oI mediation that required the
participation oI independent actors. For this reason, they turned to the actors oI civil society, actively promoting their transIormation and in the process stimulating the
emergence and consolidation oI political agents (they hoped) without any (or weakened) roots in civil society. To make such a change in the existing pattern oI
oppositional politics worthwhile, competitive political procedures leading to elections were conceded. Given the risks oI elections Ior the survival oI the established
regimes, the elites that reluctantly opted Ior this process sought their own survival by introducing elements oI restriction into the compromise (Poland) or by taking up
roles as members oI the new political society in Iormation (Hungary).
117
Our interest is not in the correctness oI such calculations but in the eIIects on civil society oI the turn to political society. Four ideal types oI signiIicant change operate
in East Europe today: reIorm, radical reIorm Irom below (or the "new evolution"), political transition to a new system, and what has been recently called
"revolution."
118
Each has its actors, its pathologies, and its potential Iorm oI selI-correction. Each takes up a diIIerent dimension oI the problem oI civil society. The
strategy oI reIorm, still dominant in the Soviet Union, has as its agents modernizing state actors. The pathology oI this path is that it replaces Iormal bureaucratic
discretion with its inIormal variants, which do not on the whole improve economic Iunctioning and, as in the current Soviet case, may actually weaken it. The imagined
corrective is the turn to civil society, which would involve in the reIorm process collective actors (groups, associations, movements, and publics) outside the state
sphere. In the Soviet Union, even the turn to the electoral mechanisms typical oI political society bypassed and Ior a while even blocked the emergence oI independent
political actors, though it helped the selI-organization and mobilization oI inIormal actors oI civil society. Thus the elections oI early 1989, and the contradictory and
inconsistent sessions oI the Congress oI Peoples' Deputies,
119
tended to lead not to mediation but to a Iorm oI mobilization that is already polarizing and will turn out
to be more so as the economic reIorm continues to stagnate. In the absence oI both violent

Page 67
repression and parliamentary mediation, the conIlicts will more and more take place in the streets.
Polarization, as we have seen in Poland, is the speciIic pathology oI the turn to civil society and its actors, in spite oI the dramatic consequences oI this turn Ior societal
learning processes and, speciIically, Ior the building oI a democratic political culture. Linked to polarization in Poland has been an overuniIication oI civil society in
which a single movement has been the vehicle Ior heterogeneous and even competing social interests and identities, somewhat blocking (even iI against the intentions oI
the participants) the emergence oI societal and, later, political pluralism. In a nationally divided society such as the Soviet Union, a second Iorm oI polarization
between competing ethnic or national groups, or between democratic and national movementshas been an even more negative consequence oI a civil-society-
oriented strategy.
120
In this context, the emergence oI political groupings capable oI negotiation, compromise, and genuine parliamentarism represents a small hope
Ior mediation, which can work only iI the institutional means are Iound to link them to the deepening lines oI social conIlict involving national, economic, and political
issues. The question is how the increasingly mobilized groups oI civil society will be able to manage their conIlicts with the regime and each other. In this context, there
does not seem to be an alternative to the rule oI law and multiparty parliamentarism other than an increasingly destructive polarization that, in the Russian center oI the
crumbling imperium, could eventually take either the Iorm oI a stalemate between societal Iorces and a state they cannot overthrow or a clash between democratic and
conservative-nationalist movements, or even a combination oI these outcomes.
121
In Poland and Hungary, the supposed corrective Ior polarization has already been promoted in the Iorm oI the turn to political society. This implies that the agents oI
the process oI transition will increasingly be the actors oI political society, at least initially including the reIormists in the Communist party. Does this model have its own
potential pathologies, and iI it does, what are its correctives?
As we have seen in the case oI Latin American transitions, one oI several reasons governmental elites turn to or revive political

Page 68
society is to help demobilize civil society. They do this both to protect themselves and the transition Irom an excess oI economic demands and to exclude Irom the
political process actors and Iorms oI mobilization that could lead to their own exclusion. While the elites oI the old ruling parties, or rather their reIormist parts, do not
have the social support to become actors oI civil society (with the very questionable exception oI trade union bureaucracies), they hope that by selI-conversion into
electoral parties with social democratic ideologies they can become actors in the new political society. Thus, clearly, the turn to political society has as its pathology the
demobilization oI civil society and the Iailure to replace its mobilized Iorms by institutionalized ones. This is a serious matter in Eastern Europe, where atomization and
the disruption oI social ties, solidarities, and associations Iar surpassed anything under even the recent bureaucratic-authoritarian regimes, and where civil society
seems to exist Ior the moment only in a mobilized Iorm whose contribution to the restoration oI social integration has been limited. For this reason a constellation that
bypasses institution building in civil society would be highly unIavorable Ior the development oI a democratic political culture, and conversely, where this type oI culture
continues to develop, it could lead to serious legitimation problems Ior new political elites.
The attempts by the reIormist elements oI the old elites to depoliticize and even Iragment civil society are quite understandable. For them, the issue involves not only
maintaining their Iree hand at making economic policy but also their survival as a political Iorce. The root oI the diIIiculty goes deeper, oI course, and may have to do
with basic tendencies linked to modern political society composed oI parties and parliaments. Arising Irom civil society and preserving some oI the marks oI their
origin, and having resisted the label "party," the new leading parties oI Hungary, Poland, and Czechoslovakia have nevertheless given rise to expectations that they
would be able to resist the "oligarchic" tendencies oI modern political parties.
122
They are nevertheless (or as a result) oIten criticized Ior replacing one elite rule by
another, Ior disregarding civil initiatives and social movements and even intensiIying state controls over local government and the public sphere, and Ior bypassing
social consultation beIore making major economic deci-

Page 69
sions.
123
SigniIicantly, attempts to reIute such charges by reIerence to parliamentary sovereignty have only led to new charges oI parliamentary absolutism and even
the exaggerated accusation oI multiparty dictatorship.
Even iI an elite democracy in which popular participation is restricted to periodic votes is not the ideal oI the major elements oI many oI the parties and groups
involved, the present context in many respects points in this direction. Once again, the requirements oI economic transition, which some rigidiIy in terms oI a
nonsolidaristic, individualist version oI civil (i.e., bourgeois) society, are in part responsible.
124
In Hungary even more than in Poland, such trends are reinIorced by
conceptions oI parliamentary sovereignty based on the so-called Westminster model, which are present in all oI the major parties. But will a population used to social
guarantees easily accept the legitimacy oI decisions involving new austerity merely on the basis oI the arrangements oI elites, irrespective oI their Iormal possession oI
an electoral mandate? There is ample experience Irom the history oI Latin American populisms that it will not, election or no election. There is a danger that populism,
which has strong roots in Eastern Europe, will be the response to elitism on the part oI demobilized or undeveloped, semiatomized, unsolidaristic civil societies.
Some Comparisons and Some Problems
It would be illegitimate to try equating the projects just surveyed. The models oI civil society that have emerged in these diIIering contexts have shown important
variations. Indeed, there are obvious diIIiculties with any single interpretive Iramework that seeks to interrogate the meaning oI, and provide orientations Ior, these
varying constellations oI structure and history. Yet a theoretical Iramework that can anchor what is in the end a common discussion across boundaries is indispensable.
A Ialse uniIication would provide only illusory solutions, and we must thereIore explore the whole range oI discourses available today. BeIore doing so, however, we
should at least justiIy our presentation oI the diIIerent projects Ior reconstructing civil society as a single set, beyond the obvious

Page 70
use oI the same terminology in diIIerent contexts. We shall do this in two steps.
First, we argue Ior a common intellectual background on the level oI the circulation oI Iorms oI discourse. In the milieu oI critical social thought, there is noticeable
today a post-Marxist intellectual turn, producing a discussion oI civil society that is truly international. Second, we present two intellectual positions, related to the crisis
oI Marxism but not reducible to it, that are shared by social actors in the Iour political contexts, as our "case studies" demonstrate. These are (1) critique oI the state
and (2) the desire to go beyond the alternative oI reIorm and revolution, in the classical sense oI these terms.
The crisis oI Marxism is a worldwide phenomenon today, Ior a variety oI local and global reasons. In the advanced capitalist countries, the continuing inability oI
Marxist theory to explain the relative stability and repeated reconstruction oI the existing system is one major reason. Another is the decisive end to the era when it
seemed possible (not to mention desirable) Ior the working classor any other single social stratum or groupingto play the role oI the global subject oI social
change. In Latin America, the decisive Iactor was Marxism's association with a revolutionary road that not only Iailed to produce any kind oI socialist commonwealth
but also directly and in some cases deliberately contributed to the end oI liberal democracy and the rise oI right-wing dictatorships. Where so-called socialist
revolutions succeeded, the results are hardly such as to inspire imitation. The Soviet model in the East, in the hour oI its Iall, is now almost universally recognized as
ineIIicient and dehumanizing. This development, reIlected in the actions and intellectual views oI dissidents, has discredited in advance the goals oI most Western and
Southern Communist or ultraleIt groupings that have inherited the mantle oI Marxism. SigniIicantly, Marxian theories and Iorms oI analyses have repeatedly Iailed in
their attempts to understand the structure oI Soviet-type societies and to outline plausible orientations Ior actors seeking to transIorm them.
125
It has always been possible, oI course, to move Irom Marxism to any position Irom liberalism and neoconservatism to religious Iundamentalism. But iI one desires to
avoid replacing a Marxist

Page 71
dogmatism by an anti-Marxist one, iI one reIuses to exchange apologetics Ior one Iorm oI domination with that Ior another, one has to grant the possibility that Marx
did establish some critical vantage points that cannot be abandoned as long as capitalist society persists. In many cases, this means reinterpreting or reconstructing
some oI his major concepts, leading to theoretical projects going Iar beyond the normative and analytical implications oI any version oI the classical Marxian theory,
including the neo-Marxisms oI Lukacs, Gramsci, and the older FrankIurt school. It is these theoretical projects that we wish to describe under the heading oI post-
Marxism.
126
A common position oI all post-Marxisms, in spite oI diIIerent terminologies, is a revision oI Marx's identiIication oI civil and bourgeois society as well as
his various political projects aiming at the reuniIication oI state and society.
127
Post-Marxists not only register, as did Gramsci,
128
the durability oI civil society under
capitalist democracies and the consequent implausibility oI revolution in the classical Marxian sense, but maintain the normative desirability oI the preservation oI civil
society. Yet post-Marxism can be distinguished Irom all neoliberalisms (which in their own way also identiIy civil and bourgeois society) by their attempts to thematize
the radical democratic or radical pluralist transIormation oI existing versions oI civil society.
We maintain that the concept oI civil society, as our various sources so Iar have used it, belongs to the intellectual world and even political culture oI post-Marxism
(and perhaps oI "post-Gramscianism"). The contemporary discourse oI civil society was internationally disseminated, at least initially, by the circulation oI post-Marxist
ideas. The wide reception oI such a concept Ior the Iirst time in our recent history, allowing Ior a dialogue between social critics East and West, North and South, has
been possible because oI shared problems and projects among those contexts.
Two such problems/projects can be Iound in the sources we have cited already. First and Ioremost, there is the critique oI the state and the search Ior a "poststatist"
politics. The inability oI Soviet-type regimes, Latin American dictatorships, and even welIare states to solve all or some key social problems, and the undesirability oI
the solutions that have emerged, is thematized in all the relevant sources. There was a time when the answer to similar diagnoses was

Page 72
a more rational statea dictatorship oI the proletariat, i.e., oI the leIt rather than the rightor (in the case oI the welIare state) simply more state, ''nationalizing" more
spheres oI liIe. It seems that aIter our recent experiences with dictatorships, nationalizations oI big industries, and the consequences oI the penetration oI social liIe by
central bureaucracies, none oI the old answers can carry their earlier weight. It is increasingly impossible to regard the state as either a passive synthesis oI a plurality
oI social Iorces or a neutral instrument in the hands oI whatever class holds the socially dominant position or manages to have its party elected to governmental power.
"Bringing the state back in" should mean recognizing that the modern state has its own logic and that it constitutes an independent constellation oI interests.
129

Contrary to the spirit oI the great nineteenth-century rebellion against the selI-regulating capitalist market economy, the state cannot be a neutral medium through which
society can act upon itselI in a selI-reIlective Iashion.
130
Second, the alternative oI reIorm or revolution has been discredited because both reIormist and revolutionary parties have had a share in our present crises. All oI our
case studies reveal, explicitly or implicitly, the same renunciation oI the utopia oI revolution, oI the dream oI a single, imposed model oI the good society that breaks
completely with the present, that is beyond conIlict and division. Such a model is not compatible even in principle with any modern notion oI democracy. At the same
time, what the case studies express is more than merely incremental reIorm; at the very least, structural or radical reIormism is implied. Yet even these terms coined by
A. Gorz
131
do not exhaust what is at stake. Revolution and reIorm are both today widely understood in terms oI (and condemned Ior) their statist logic, and the idea
oI somehow combining them, as the term "radical reIormism" still suggests, now becomes unacceptable. The term "new evolutionism" is too vague to serve as a
replacement, but either "selI-limiting revolution" or "selI-limiting radicalism" seems appropriate. The idea here, worked out by analysts as diverse as J. Kuron, A. Gorz,
N. Bobbio, and J. Habermas, is that the object oI radical reconstruction and also its (multiple, nonuniIied) subjects shiIt Irom the state to society. Correspondingly,
with regard to the existing structures

Page 73
oI state (and, in the West, capitalist) economies, a new kind oI selI-limitation would have to be and even ought to be practiced. This idea survives in the two
temporalities oI change reIerring to state and civil society, as proposed by Social Contract, and even in the turn to political society that implies a consciously
nonrevolutionary slowing down oI the rate oI change through negotiations and elections. In a Western version, the same idea is expressed quite well by Rosanvallon's
juxtaposition oI the rebuilding oI civil society with necessary compromises on the structures oI the state and the economy. Civil society can help change those
structures but must not abolish all aspects oI their autonomous operation.
Interestingly enough, it is in the most anti-Marxist oI our three constellations, Eastern Europe, that the term "revolution" is most oIten used to indicate transition Irom
authoritarian rule. It must be said, however, that the sense oI the term diIIers Irom those established by the French and Russian revolutions. The search Ior the perIect
and transparent society associated with these revolutions is explicitly rejected as state-strengthening and even unavoidably terroristic. Some authors redeIine the term
in a more conservative sense, seeking to preserve still existing (or imagined) older political cultures or traditions threatened by Sovietization, or conserving someone
else's tradition (e.g., classical liberalism).
132
Others, building on the single case oI the deIeated Hungarian Revolution oI 1956, seek to understand the transitions in the
making as a pure "political revolution" leading to the establishment oI a new Iorm oI democratic sovereignty, a novus ordo seclorum.
133
The Iirst oI these lines oI
thought, in part returning to the premodern notion oI revolution as an attempt to reestablish a previous state oI aIIairs, tends to miss what is genuinely new in the
present-day projects oI transIormation. It can lend credence to views reIerring to "restoration" or "counterrevolution.'' The second misses their explicitly selI-limiting
and evolutionary character. This has been repeatedly maniIested in the search Ior compromise and transitional solutions and the deliberate acceptance oI the slowing
down oI the rate oI change. Amazingly enough, given the nature oI the previous regimes, their successors seek neither a general personal expropriation oI the members
oI earlier elites nor their total exclusion Irom political or proIessional activity. Indeed, these options are

Page 74
avoided in a reIlective and conscious manner even in the Iace oI repeated eIIorts to convert powers oI the past into those oI the Iuture. The selI-limiting revolution
avoids the total destruction oI its enemy, which would inevitably mean putting itselI into the place oI the sovereign,
134
thereby depriving society oI its selI-organization
and selI-deIense.
The term "selI-limiting revolution" (as well as its partial synonyms, "peaceIul" and "velvet" revolution) avoids the weaknesses oI both the ideas oI ''conservative" and
"popular" revolution. Instead oI retreating behind the modern meaning oI "revolution" or repeating its totalizing thrust, this idea extends the selI-reIlexive and selI-critical
discourse oI modernity to its most important political concept, namely, revolution.
135
We have already noted that the more or less common posture oI antistatist, selI-limiting revolution that we discover in our diverse sources is not expressed in terms oI
a single categorical Iramework or a single model Ior reconstructing civil society. At times we Iind several variants are proposed within a single cultural-political context,
and oI course the projects vary even more signiIicantly across contexts. The common core oI all the interpretations, though, is the concept oI civil society, or rather
some oI the components oI this concept. All agree that civil society represents a sphere other than and even opposed to the state. All include, almost always
unsystematically, some combination oI networks oI legal protection, voluntary associations, and Iorms oI independent public expression. A very Iew conceptions seem
to include Iamilies and inIormal groups. Some include movements and even equate civil society with the presence oI social movements; others (such as that oI the
Polish writer Wojcicki) exclude and even Iear this possibility as a Iorm oI unacceptable politicization. In the texts concerning the Iour political projects, however, we
have Iound no comprehensive treatment oI the relation among the categories oI civil society or, Ior that matter, oI the nexus between civil society as movement and as
institution. But there is no question that the stresses in the various contexts and texts are oIten quite diIIerent, even iI little has been added to (or explicitly subtracted
Irom) the classical list oI laws, associations, and publics.
136
There are two major issues that produce important shiIts in categorial Irameworks. First, should the economy be included or

Page 75
excluded Irom the concept oI civil society (the Hegelian vs. the Gramscian model)? And second, should one seek to diIIerentiate civil and political society (the
Tocquevillian vs. the Hegelian model)? Neoliberals and residually neo-Marxist writers tend to agree on including the economic sphere within civil society, albeit Ior
opposite reasons. The Iormer, whether in the West or now increasingly in the East, reaIIirm the identity oI the civil and the bourgeois, Iear a model oI rights in which
property is not in the primary position, and reject the politicization oI society and the Iormation oI social movements that would demand economic redistribution Irom
the state. While legitimately concerned about the consequences oI the link between populism and statism, this intellectual tendency Iorgets the destructive eIIects oI the
selI-regulating market on the cultural Iabric oI society, described so well by Karl Polanyi. Those in Eastern Europe who Iorget this lesson because oI their hatred oI all
Iorms oI state interventionism seek in eIIect to rejoin Europe not as it is today, Iacing ecological and social problems generated by the capitalist economy, but as it
once was, inviting the repetition oI already known disasters.
The second approach, the residually Marxist one typiIied by Andre Gorz and to an extent even by Claus OIIe, presupposes these destructive eIIects but does not
suIIiciently consider the disastrous results oI eliminating economic rationality in the process oI politicizing production and distribution. While neoliberals reduce civil
society to economic society, neo-Marxists either reduce the Iuture (postcapitalist) economy to political society or propose, in the manner oI utopian socialists, some
kind oI socially reembedded economy. In Gorz's Farewell to the Working Class, these two recipes are combined. In the (to us, preIerable) realist Green Iormula oI
OIIe and his colleagues, an economic sphere based on reciprocity, mutuality, and selI-activity (Eigenarbeit) is combined with a macro-economically steered but
nevertheless genuine market economy. In this Iormula, economic activities in the substantive sense are (at least in part) included in civil society, but economy as a
Iormal process is outside oI it.
137
When civil society in the shape oI a social movement is in the process oI organizing and institutionalizing itselI, however, Iew authors argue Ior its unity or even
continuity with economic

Page 76
society. There is no question oI such reductionism, Ior example, in the writings oI Michnik and Kuron. Instead, they have consistently argued Ior the autonomy oI legal
structures, Iree associations, and genuine public liIe conceived in terms oI the promise oI a solidaristic civil society. Undoubtedly the Iact that a minor chord in their
argument is the liberation oI the economy Irom state controls played a major role here. Beyond the utopia oI the complete democratization oI production that Kuron
still proposed in the mid-1960s, the writers oI the Polish democratic opposition are Iorced to Iace the harsh reality that only the restoration oI the market, beyond any
model oI social reembedding, could master the Polish crisis and produce a viable, modern economy. Even iI industrial democracy plays a role in their proposals, it is
recognized that this must be made compatible with the needs oI expert management operating in an environment that allows rational calculation. Understandably, in the
East European context, the harmIul eIIects oI a Iully autonomous capitalist market economy on social solidarity, denied by neoliberal writers, was not directly
thematized by the main authors oI the democratic opposition. Nevertheless, the Solidarity movement, because oI its social nature as well as its ties to a Catholic
syndicalist tradition, has been to an extent sensitive to just these dangers.
SigniIicantly, the intellectual and political journey made by Latin American writers like O'Donnell and Cardoso is in many respects similar to that oI Kuron and
Michnik. As late as 1978, O'Donnell still used "civil society" in the neo-Marxian sense oI bourgeois society. The mediations he then proposed between civil society
and the state (nation, pueblo, and citizenship) corresponded only to the underdeveloped structure oI societies plagued by cycles oI populist uniIication and
authoritarian atomization. Under the impact oI new Iorms oI selI-organization and struggles Ior democracy in the next decade, O'Donnell and P. Schmitter Iully
changed their terminology and began to use "civil society" to describe a sphere between economy and state, characterized above all by associations and publics. The
Iailure oI populist-authoritarian eIIorts, moreover, led to the rejection oI the reverse subsumption, that oI the economy by social or political institutions. In Cardoso's
subtle analysis, the role oI industrial democracy seems to be to establish vantage points oI social control without impairing economic rationality.

Page 77
On the whole, in neither Latin America nor Eastern Europe has the "interIace" oI civil society and market economy been adequately analyzed.
138
Such an analysis,
however, is a precondition Ior any really serious conceptual alternative to the dangers oI economic liberalism and the Ialse promises oI utopian socialism.
139
Without
such an alternative, one can expect more vacillation between market and state as agents oI liberation and renewed neglect oI the destructive eIIects oI both on social
solidarity and individual autonomy.
Equally important is the division oI opinion on the interIace between civil society and state. The French writers we have described tend to consider civil and political
society as two spheres, the second mediating the relations oI the Iirst with the state. In this conception, both civil and political society must be reconstructed to
preserve and renew the Ioundations oI associational liIe and to be able to make those eIIective vis-a-vis the state. In most oI the East European analyses coming Irom
the democratic opposition, and in at least some Latin American writers (e.g., F. WeIIort), the category oI civil society includes and subsumes the levels oI its political
mediations. Finally, in yet other models, the two categories "civil" and "political" appear more as alternatives oI the type oI civil society that is desirable or possible. In
the writings oI Claus OIIe, Ior example, the choice seems to be between neoconservative (depoliticized) or radical democratic (political) civil society. In the argument
oI O'Donnell and Schmitter, there is a succession oI temporal phases, with depoliticized civil society representing the normal phase that can survive even authoritarian
rule, while political civil society is only the exceptional phase oI mobilization or upsurge. Here the cycle oI types oI civil society represents another version oI the
political cycle oI authoritarian and democratic regimes. The move Irom demobilized to mobilized civil society implies the end oI the authoritarian regime; demobilized
civil society, implies Iirst the stabilization oI democracy and only eventually the possibility oI a return oI dictatorship. Even in some Eastern European analyses, a choice
between unpolitical and political interpretations has been proposed (in Poland, by Catholic intellectuals) to highlight the alternative oI antipolitics in a society deeply
tired oI previous Iorms oI politicization.

Page 78
Assuming Ior the moment that the stark alternative between political and civil society is a Iunction oI either undesirable political polarization, in which the
neoconservatives have had the initiative, or an equally undesirable cycle, we are still leIt with two competing models that express the need to combine prepolitical
levels oI social liIe with political Iorms that can provide Ior public liIe outside the Iramework oI public political authority, i.e., the state. These involve, on the one hand,
a model oI civil society that includes a political public sphere among its categories and, on the other hand, a Iramework within which civil and political society are
clearly diIIerentiated. To some extent, the choice is a question oI inherited intellectual traditions. The German tradition stemming Irom Hegel and Marx represented a
culmination oI the diIIerentiation oI the classical topos oI political or citizen society into state and depoliticized civil society. This tradition has room Ior mediations
between civil society and state within each domain but not Ior an independent domain between them with distinct institutions and dynamics. In contrast, the French
tradition derived Irom Tocqueville never totally dissolved the old category oI political society but instead established it alongside civil society and state. Finally, and
most conIusingly, the Italian tradition going back to Gramsci uses all three terms but tends to identiIy political society with the state, echoing the traditional premodern
usage.
Current political requirements are equally important in the choice between the two types oI categorization. In both Latin America and Eastern Europe, the juxtaposition
oI civil society and state was a conceptually dualistic outcome oI a period oI societal selI-organization that led to polarization between democratic and authoritarian
Iorces. Independent society was strong enough to survive and even to challenge the legitimacy oI the authoritarian state. But it was not strong enough to compel
genuine compromise or to secure a transition beyond authoritarian rule. With the emergence oI real possibilities oI negotiation and compromise, and even agreement,
concerning the dismantling oI authoritarian governments in Iavor oI electoral scenarios, the category oI civil society seemed to many writers (Cardoso, Kis, Stepan) to
be unsuitable to depict the organized social Iorces entering into processes oI political exchange with state actors. This led to the

Page 79
resurrection oI the category oI political society (or its stand-ins) even where the inIluence oI Hegel, Marx, and Gramsci was strong. Some writers oIIer normative
reasons Ior the shiIt, insisting that the turn to political society allows a desirable pluralization oI the opposition, whose location on the level oI civil society is said to
involve monolithic uniIication within the one great movement oI society.
140
Thus, the choice between the two Irameworks cannot rest on intellectual history, current political requirements, or even their combination; it presupposes additional
systematic considerations that we shall outline later in this book. For now, we note only that a choice oI either approach has been insuIIiciently motivated thus Iar. In
particular, the structures and Iorms oI action that would correspond to civil as distinct Irom political society have not been systematically analyzed by those who
presuppose the sharp diIIerentiation oI these two domains. To make their case, deIenders oI diIIerentiation would have to have recourse to something like the old
distinctions oI movements and elites, as well as oI inIluence and power, to Ilesh out the diIIerence between the "civil" and the "political." This they may notwish to do,
however, Ior tacit normative or ideological reasons.
Indeed, the two Irameworks seem to have diIIerent relations to analytical and norma Iive considerations. From an analytical point oI view, the distinction between civil
and political society helps to avoid the sort oI reductionism that assumes that political activities with a strategic dimension are easily generated by societal associations
and movements or are somehow unnecessary. Paradoxically, an undiIIerentiated concept oI civil society gives us a stark choice between the depoliticization oI society
(where the political is assigned to the state) and its overpoliticization (where all dimensions oI civil society are held to be political or are to be politicized). The
distinction between the civil and the political, on the other hand, highlights the Iact that neither oI these domains is automatically reconstituted when the other is. Indeed,
there could even be opposition and conIlict between the requirements oI the two projects.
From a normative point oI view, treating political society as a mediation within a many-leveled civil society has the possible

Page 80
advantage oI establishing the priority oI nonstrategic domains oI solidarity, association, and communication. DiIIerentiating the civil and the political seems to put the
domains on an equal normative Iooting. While this latter approach does not make the reconstitution oI civil society an automatic Iunction oI the existence and activity oI
political organizations, it nevertheless tends to relieve the actors oI political society Irom the normative burden oI having to build or IortiIy civil institutions that may limit
their own Ireedom oI action. This is a serious problem, because although the actors oI civil society seem to learn by their Iailures that they cannot achieve their own
goals without recourse to political society, the reverse is unIortunately not the case, as the history oI elite democracies shows.
141
It is only in the long run that the
viability oI a democratic political society may depend on the depth oI its roots in independent, prepolitical associations and publics.
Given the complementary normative and analytical advantages oI the two conceptions, one treating political society as mediation and the other stressing analytical
diIIerentiation oI the civil and the political, we propose to use both conceptions and at times to combine them. We believe that this is appropriate because our
methodology combines hermeneutic and analytical approaches.
The issue oI the relationship between civil and political society is connected to the question oI the locus oI democratization. All oI our relevant sources view liberal
democracy as a necessary condition Ior bringing the modern state under societal control. They also assume that liberal democracy is incompatible with a democratic
pyramid whose base is direct participation. They have, moreover, broken with the old dream oI abolishing the state. Nevertheless, in the West this new emphasis
tends to be coupled with an old one: awareness oI the elitist character oI contemporary liberal democracies. This set oI positions, together with a certain deemphasis
(though not abandonment) oI the idea oI industrial democracy, has led many authors in the West to shiIt the project oI "democratizing" elite democracy Irom the state
to civil society.
142
In the program oI the Greens, as represented by OIIe, this change has also been articulated on the organizational level, in the attempt to combine
party-based with movement-oriented strategies. In general, those who seek to democratize civil society understand this domain as comprised oI movements as well as
institutions.

Page 81
This has been also true oI Eastern Europe and Latin America, where movements have tended to be Iar more global and comprehensive than in the West. Under
dictatorships, though, there was something constrained and artiIicial in the shiIt oI the project oI democratization to civil society: The sphere oI the state (not to mention
the economy) and oI potential parliamentary mediation was placed oII limits not by normative choice but by strategic necessity. The long-range goal oI parliamentary
democracy was as a rule aIIirmed, with the exception oI those appealing to a diIIerent (deIicient or superior, as the case may be) political culture and tradition. When
the crisis oI the regimes made this a possible short-term goal, Ior many the project oI democratization shiIted to political society. Some authors even tried to juxtapose
"liberalization," oriented to civil society, and "democratization," whose locus was to be primarily political society.
143
In Eastern Europe, the elite theoretical
understanding oI Western European liberal democracy was either Iorgotten or abandoned in Iavor oI a civics textbook version. The revival oI economic liberalism also
increased suspicion oI societal organizations capable oI making demands on new political elites that might translate into unacceptable economic costs. Many who seek
to restrict democratization attack social organizations such as Solidarity Ior being undemocratic. Some hold that societal democratization inhibits the creation oI a truly
modern state capable oI eIIective decision making.
144
There are, oI course, countervailing tendencies rooted in the movement character oI the Polish and also, in part, the Hungarian opposition. There is a tendency to
articulate, more in practice than in theory, a dualistic strategy that sees the diIIerent Iorms oI democracy and democratization in civil and political society as
complementary, each indispensable Ior a project oI "more democracy." Cardoso, in Latin America, has come the closest to articulating such a program explicitly.
Initially, at least, the dualism oI union and party in which the victorious Solidarity movement articulated itselI Iavored a similar Iormulation. Even aIter the split oI this
movement-party, the two new organizations that have emerged, the liberal-democratic ROAD (Civic Movement-Democratic Action) and the right-wing Center
PlatIorm, seem to share this dual heritage, as do all the dynamic new organizations oI

Page 82
Hungary (MDF, SzDSZ, Fidesz) and Czechoslovakia (Civic Forum, Public Against Violence). The organizational models oI these new political "parties," none oI
which is Iormally named as such, have at least initially brought them close to the dualistic model sought, generally unsuccessIully, by some oI the new social movements
oI the West, especially the Greens.
Today's trend nevertheless is to proIessionalize and "partiIy" the new parties. Some still talk, though, oI developing more complex ties to the Iorms oI civil society
within the Iramework oI increasing diIIerentiation Irom them. Such ties would presuppose both a programmatic openness oI the political to the civil and a suIIicient
strengthening oI the latter to allow it to Iunction in institutionalized Iorms. What is needed, in other words, are programs that not only establish an ongoing process oI
political exchange with organizations and initiatives outside the party political sphere but also strengthen civil society with respect to the new economic society in
Iormation.
145
Only such a program could oIIer something genuinely new with respect to present models oI Western politics, thereby transcending the bad choice oI
either economic liberalism and elite democracy or direct democratic Iundamentalism.
But even iI such a new civil-society-oriented strategy whose roots can be discovered in the varieties oI political discourse explored here were to emerge, it is not yet
clear why it should be preIerred to a renewed liberalism (very much on the rise) or a radical egalitarian democracy (at the moment on the decline). And iI it could be
shown to be normatively preIerable to those options, it may well be the case that more complex theoretical considerations would show precisely what is attractive
about the politics oI civil society is incompatible with the development oI modernity. To examine these issues with suIIicient seriousness, we now take our leave oI the
discussions oI contemporary actors and turn to theoretical reconstruction and critique oI the concept oI civil society.

Page 83
2-
Conceptual History and Theoretical Synthesis
A Sketch of Early Modern Conceptual History
Present-day political models that use the concept oI civil society not only contradict one another but are also relatively poor in categories. Furthermore, their links to a
rich tradition oI interpretation are not clear. Since this tradition is not thematized, the diIIerences between the new versions oI the concept and their historical
predecessors are also leIt unexplored. Thus, a theoretical scheme inherited Irom the past (or even several pasts) is simply assumed, but not demonstrated, to be
adequate to modern conditions.
In our view, a conceptual history oI the term "civil society" is an important way to begin to address these tasks. Such a history should, Iirst oI all, deepen and extend
the relevant categorical Irameworks in use today. Second, it should allow us to distinguish premodern and modern layers in the concept, indicating what versions have
become questionable and inadequate today. While conceptual history cannot remove the contradictions among contemporary usages, it can help us see what is at
stake in these contradictions and what options have become, at least historically speaking, implausible. Finally, a conceptual history can help root the usages oI a
concept oI civil society in a political culture whose motivational power has not yet been exhausted: the political culture oI the age oI the democratic revolutions.
Conversely, the revival oI the concept today helps validate this particular political culture.

Page 84
The Iirst version oI the concept oI civil society appears in Aristotle under the heading oI politike koinonia, political society/community. It is this term the Latins
translated as societas civilis. The concept represented the deIinition oI the polis, understood as the telos oI the human being as a political animal, :oon politikon.
Politike koinonia was deIined as a public ethical-political community oI Iree and equal citizens under a legally deIined system oI rule. Law itselI, however, was seen
as the expression oI an ethos, a common set oI norms and values deIining not only political procedures but also a substantive Iorm oI liIe based on a developed
catalogue oI preIerred virtues and Iorms oI interaction.
1
Today we can symbolically represent our distance Irom the Greeks by pointing to the absence oI a series oI
distinctions and oppositions in the concept oI politike koinonia. First oI all, the Aristotelian notion did not allow Ior our distinction between state and society. The
polis-oikos duality may seem to indicate the contrary, but the oikos, household, was understood primarily as a residual category, the natural background oI the polis.
Politike koinonia was logically only one koinonia among many (including perhaps the oikos, but more generally all Iorms oI human association Irom occupational
groupings to groupings oI Iriends, etc.); it was more deeply understood as the all-encompassing social system with nothing except natural relations outside oI it.
2
Thus,
there could be no question oI the polis and the oikos representing two svstems oI (diIIerent) social or political relations. First, the oikos was not a legal entity: It was
regulated not by law but by the despotic rule or domination oI its head. Second, the plurality oI households represented no system: They related to one another (in
theory) only through the polis; indeed, through their heads they were in the polis. Economic relations beyond the household were considered merely supplementary
and, beyond a maximum point, pathological.
3
The resulting concept oI politike koinonia was paradoxical. It indicated one koinonia among many and, at the same time, the whole, a whole with parts outside itselI.
The paradox could be resolved because oI the absence oI a second distinction: that between society and community. Koinonia in general denoted all Iorms oI
association irrespective oI the level oI solidarity, intimacy, or intensity oI interaction. In the case oI politike koinonia, this allowed Ior a con-

Page 85
ception that already presupposed the existence oI a plurality oI Iorms oI interaction, association, and group liIe; hence, something oI our concept oI ''society." Yet
plurality and diIIerentiation were dramatically integrated in a model that presupposed a single, homogeneous, organized solidary body oI citizens capable oI totally
uniIied actioncloser to our notion oI community, a "community oI societies." In theory at least, politike koinonia was a unique collectivity, a uniIied organization with
a single set oI goals that were derivable Irom the common ethos. The participation oI all citizens "in ruling and being ruled" represented a relatively small problem in
theory, given this assumption oI a shared set oI goals based on a single Iorm oI liIe.
4
There is hardly any doubt about the idealized nature oI the Aristotelian conception.
5
But what is important Ior us is that it was this conception that entered into the
tradition oI political philosophy. We leave to the side the Iirst Roman translations oI politike koinonia as societas civilis, because, as Iar as we can tell, here the
concept played only a minor role. More important were the medieval Latin adaptations Iollowing the translations oI Aristotle by William oI Moerbeke and Leonardo
Bruni. While some oI the earlier utilizations by Albertus Magnus and Thomas Aquinas tended to restrict societas civilis to the medieval city-state (as the closest
available equivalent oI the ancient polis),
6
such a prudent use oI the concept could not be maintained Ior long, perhaps because the Greek notion also reIerred to the
overarching level oI sovereignty. Only in Italy, however, did city-states approach the status oI Iull sovereignty, and even here only in Iact and not in law. As a result,
when the Greek conception was more generally utilized, the Ieudal order oI Iragmented sovereign units (patrimonial rulers, corporate bodies, towns, etc.) as well as
medieval kingship and empire, all came to be described in diIIerent sources as societas civilis sive res publica.
7
Unnoticed, this usage introduced a level oI
plurali:ation into the concept that could now hardly be uniIied under the idea oI an organized, collective body, the notion oI respublica Christiana notwithstanding.
A second important shiIt, one oI duali:ation, occurred when the concurrent revival oI monarchical autonomy and public law Iavored the adaptation (however
implausible) oI the ancient idea oI republic

Page 86
(with which societas civilis was identiIied) to the Stàndestaat that balanced the new powers oI the prince with that oI the organized, corporate estates that assembled
all those having power and status in Ieudal society. The dualism here was, however, as Otto Brunner has tirelessly insisted,
8
not between state and society: Civil or
political society was understood as a type oI state dualistically organized with the "prince" on one side and "land" or "people" or "nation" on the other, with the latter
terms denoting the privileged estates. II we accept Marx's 1843 judgment that the old corporate society was immediately political, then the history oI the concept oI
civil society beIore absolutism belongs at least in this sense to the Iundamental pattern established by the Greek prototype oI politike koinonia, despite enormous
diIIerences among the social Iormations in question.
The development toward absolutism represents the watershed between traditional and modern meanings oI "civil society." We see the reasons Ior this in two well-
known and complementary developments. First, the development oI princely authority Irom the primus inter pares oI a plurality oI power holders (classical Ieudalism)
and the senior partner oI a dualistic system oI authority (Stàndestaat) to the monopolistic holder oI the legitimate means oI violence laid the Ioundations oI the modern
state. Second, the depoliticization oI the Iormer power holders, the estates and corporate bodies, did not destroy their organized and corporate status. Instead, it
produced a veritable society oI orders. To be sure, the transition to a duality oI state and nonpolitical society could be and was indeed achieved by other, at times
complementary, routes: the emergence oI autonomous religious bodies tolerated by a more secular state (North America)
9
as well as the rise oI new Iorms oI private
economic activity outside the policies oI the mercantile state (Great Britain). In our opinion, however, the shiIt Irom the corporate entities oI the Stàndestaat to those
oI the depoliticized society oI orders was not only historically prior but was also more important, Ior the European continent at least. BeIore the absolutist state could
disorganize and level its corporate rivals in the name oI the universal status oI the subject oI the state, a countermovement already began to reorganize "society" against
the state through associations and Iorms oI public liIe that may have drawn on the

Page 87
resources oI estate independence, religious dissent, and economic entrepreneurship but that embodied new egalitarian and secular principles oI organization.
10
There
is no doubt, at least as Iar as we are concerned, that the "society" oI the Enlightenment, constituting a new Iorm oI public liIe, was the prototype oI the early modern
concept oI civil society.
OI course, political philosophy that sought to preserve the identiIication oI civil and political society did not immediately register the emergence oI a new Iorm oI
societal public sphere. Three or Iour alternatives were developed. First, one could try to continue, as did Jean Bodin, in spite oI decisive historical changes best
registered by himselI, the stàndestaatliche conception oI res publica sive societas civilis sive societas politicus. Reapplied to the constellation oI absolute
monarchy and society oI orders, this conception IalsiIied the new type oI duality now in Iormation, a duality Bodin otherwise deIended. Nevertheless, the model
persisted into the German eighteenth century.
11
Second, one could identiIy the modern state itselI with the commonwealth or civil/political society. This was the option oI Hobbes, who oI course believed that
sovereign power supplied the only "social" bond oI naturally unsocial yet rational individuals.
12
In Hobbes's theory, the social contract creates a state, not society. The
Iusion oI society is accomplished only by the power oI the state. While Hobbes merely came close to the Greek view that construed the concept oI a political society
as an undivided system oI power, he soon came to realize that the ancient concept relied on a notion oI moralized law rooted in ethos, rather than positive law limited
only to enactment or command. Thus the later construction in the Leviathan more or less leIt out the whole concept oI civil society (i.e., the normative idea oI Iree and
equal citizens comprising the body politic). Nevertheless, the identiIication oI state and civil society is preserved down to our own day in some Anglo-American
literature.
The third option involved a breaking up oI the old Iormula societas civilis sive politicus sive respublica by retaining the identity oI political and civil society but
distinguishing both Irom the state. Locke's speciIication oI the product oI the social contract as "political or civil society"
13
seems to continue on the path oI the

Page 88
early Hobbes, representing no break with the tradition. At Iirst sight, his conception even includes an apparent identiIication oI the body politic with government.
14

Locke, however, does clearly seek to diIIerentiate between "government" and "society." He distinguishes between surrendering power to society and to the
government "whom society hath set up over itselI"
15
and even more emphatically (unlike Hobbes) between the "dissolution oI the society" and "the dissolution oI the
government.''
16
Characteristically, however, in this context Locke stays close to the ancient concept when he speaks oI the "one politic societv" in terms oI "the
agreement to incorporate and act as one body." This ability to become and to act as one body is still assigned to the legislative power oI government. The dissolution
oI the legislative power is proposed as the end oI a society, but Locke inconsistently assigns the possibility oI providing Ior a new legislature to the same society when
the legislature is dissolved, or even when it acts contrary to its trust.
Montesquieu's conception was more historically sensitive. It united the eighteenth-century notion oI two contracts (social and governmental) with the Roman law
distinction oI civil and public law (here "political law").
17
Whereas political law regulates the relationship oI governors and governed, civil law regulates the relations oI
members oI society to one another. Accordingly, Montesquieu, Iollowing the Italian writer Gravina, distinguishes between government (l´etat politique) and society
(l´etat civile).
18
Montesquieu's conception oI society appears under a shiIting terminology. In the context oI monarchical government (which represents the modern
state Ior him!) it meant, alternatively, the "intermediate powers," "the political communities," or "societies or communities" inherited Irom the epoch oI estate dualism.
19
Thus, Montesquieu's antiabsolutist strategy relied more on a society constituted by a hierarchic traditional society, one that he wished to repoliticize, than even Locke's
notion oI political society, which contained at least the notion oI an initial equality oI status. With regard to the Enlightenment conception, Montesquieu anticipated,
however inconsistently, the diIIerentiation, Ior polemical reasons, oI state and society, while Locke redeIined the notion oI society itselI in terms oI the idea oI Iormal
equality derived Irom

Page 89
universal natural law. Despite the ideological Ieatures oI their conceptions (in Montesquieu's case, still expressing the world view oI privileged but depoliticized orders;
in Locke's, that oI a new status order increasingly based on private property), these two philosophers provided important conceptual preparation Ior the modern
redeIinition oI civil society. Their constructions pointed beyond the ideological limits oI the original presentations.
It was Hegel who synthesized much oI late-eighteenth-century thought on the subject, in eIIect weaving together the somewhat divergent strands oI "national"
development. It would, however, be erroneous to credit Hegel alone with redeIining the concept oI civil society.
20
BeIore turning to his synthesis and its Iate, then, we
pause to note some oI these other contributors.
(1) The conception we have reIerred to as the Enlightenment notion oI "society" (as contrasted with the state) rapidly developed beyond its origins in Locke and
Montesquieu. Paradoxically, the new notion oIten coexisted with the more traditional identiIication oI civil and political society with the state, as in the case oI
Rousseau (and then Kant).
21
In France, these two trends both shared in the growing opposition to both societal pluralism, in the sense oI group or collective rights
identiIied with social orders, and monarchical absolutism. Thus, one might say that, as the polemical conception oI "society against the state" was Iashioned in the
salons, coIIeehouses, lodges, and clubs oI the time,
22
both the rhetoric oI antiabsolutism (Montesquieu) and opposition to privilege (Voltaire) were united in a single
conception oI a (civil) society opposed to a state whose components were Iormally equal, autonomous individuals as the sole repositories oI rights. This conception
Iully came into its own in a series oI revolutionary conceptions oI natural law. Thomas Paine's Common Sense, the various American bills oI rights, and the French
Declaration of the Rights of Man and Citi:en clearly juxtapose an individualistic, egalitarian society to government (even a constitutional state!), with the society
becoming the sole source oI legitimate authority.
23
(2) In England aIter the Glorious Revolution, Locke's ambiguous separation oI society Irom government was slowly eroded. What counted as "society" was now
organized as a state that involved a gradual Iusion between parliamentary representation and the

Page 90
executive.
24
The term "society" as distinct Irom "the state" came to be reserved Ior high or polite society, a custodian oI manners and inIluence, but not oI any kind oI
political project. In general, the term ''civil society" preserved its traditional identiIication with political society or the state. A new component was added to this
identiIication by the thinkers oI the Scottish enlightenmentFerguson, Hume, and Smith, among otherswho came to understand the essential Ieature oI civil or
"civilized" society, not in its political organization but in the organization oI material civilization. Here a new identiIication (or reduction) was already being prepared: that
oI civil and economic society, reversing the old Aristotelian exclusion oI the economic Irom politike koinonia.
25
(3) The French and British conceptions had a strong inIluence in Germany, in the works oI Kant, Fichte, and a whole series oI lesser Iigures. A certain intellectual
conservatism, however, in political as well as intellectual history, also played a historically important role in Germany in preparing the way Ior Hegel's theory. We have
in mind the preservation oI the Montesquieuian stress on intermediate bodies or powers in the notion oI a neustàndische Gesellschaft in which Stànde or estates (in
particular, der bùrgerlicher Stand) would be based on occupational mobility and merit, rather than birth and inheritance, as well as a Iorm oI a constitutionalism that
represented a modernization rather than the abolition oI the dualism oI the Stàndestaat.
26
Nevertheless, the attempt to modernize the notion oI estates was
overshadowed by the inIluence oI Kant's redeIinition oI civil society as based on universal human rights beyond all particularistic legal and political orders. In Kant's
philosophy oI history, a universal civil society based on the rule oI law was postulated as the telos oI human development. Kant explicitly rejected (in the spirit oI the
French Revolution) any compromise with the corporate and estate powers oI the absolutist era.
27
Instead oI the old concept, Kant and then Fichte put Iorward the
notion oI a citizen society, staatsbùrgerlicher Gesellschaft, which they interpreted in the spirit oI the French Declaration oI 1789.
28
In Fichte especially, according
to ManIred Riedel, two speciIically modern notions appear Ior the Iirst time: the sharp separation oI state and society, and the understanding oI society itselI in
individualist and universalist terms. In making this shiIt, the young Fichte moved Irom liberalism to radical democracy.

Page 91
The two strands oI the German discussion oI civil societythe universalism oI Kant and Fichte and the pluralism oI more the conservative line oI thoughtcome
together in Hegel. But Hegel also brought other strands into his great synthesis: in particular, the Scottish notion oI civilized or economic society. While Hegel's
conception oI civil society may not be the Iirst modern one, we do believe that his is the Iirst modern theorv oI civil society. Moreover, the theoretical inspiration oI
Hegel's synthesis is in our view not yet exhausted. Despite some views to the contrary (Riedel, Luhmann), we shall argue that several important theoretical traditions
that emerged aIter Hegel, with or without conscious reIerence to him, continued to move within the terms oI analysis that he has brought together. For this reason, we
would like to present Hegel not in the context oI a conceptual history that analyzes the hermeneutic structure oI our concepts but rather as the most important
theoretical Iorerunner oI several later approaches that have preserved their potential to provide more global, intellectual orientation even in our own time.
Hegel's Synthesis
All strands oI the history oI the conception oI civil society so Iar presented meet in Hegel's Rechtsphilosophie. He is the representative theorist oI civil society because
oI the synthetic character oI his work and, even more, because he was both Iirst and most successIul in unIolding the concept as a theory oI a highly diIIerentiated and
complex social order.
It is by now a commonplace that Hegel attempted to unite, in a scheme that was to be both prescriptive and descriptive, a conception oI ancient ethos with one oI the
modern Ireedom oI the individual. But it should also be stressed that in his conception, the modern state did, could, or at least should also reconcile dimensions oI the
ancient, homogeneous, uniIied political society with the late medieval plurality oI autonomous social bodies. The ancient republican dimension in his conception, drawn
Irom Aristotle and other classical thinkers, was to rest on the twin pillars oI ethical liIe (ethos or Sittlichkeit) and public Ireedom. The medieval dimension drawn Irom
Montesquieu and a whole series oI German sources involved a renewed stress on intermediate bodies in the Iace oI the modern

Page 92
state.
29
The speciIically modern component was to rest on three major Ieatures. First, Hegel took over Irom the natural-law tradition and Irom Kant the universalist
deIinition oI the individual as the bearer oI rights and the agent oI moral conscience. Second, he generalized the Enlightenment distinction between state and civil
society in a manner that also involved their interpenetration. Third, he took over Irom Ferguson and the new discipline oI political economy the stress on civil society as
the locus and carrier oI material civilization. Astonishingly, he succeeded in building all these elements into a uniIied Iramework, albeit one that was not Iree oI
antinomies.
One contradiction that permeates Hegel's work is that between systematic philosophy and social theory. This is expressed politically as the antinomy oI statist and
antistatist positions running through both the doctrine oI civil society and that oI the state.
30
Hegel's social theory presents modern society both as a world oI alienation
and as an open-ended search Ior social integration. His philosophical system, conversely, pronounces that this quest has ended in the modern state. It is never entirely
clear, though, whether he means a possible and desirable, or a not yet existent but necessary, or an already existing state. But even in the weakest version oI this
argument, when he identiIies the possible and desirable Iorm oI the state with a modernizing and constitutional version oI a bureaucratic monarchy, the statist
implications oI Hegel's system building become clear. Yet, at the same time, Hegel's recurring arguments against monarchical absolutism and revolutionary
republicanism revive an antistatist stress on intermediary bodies limiting bureaucratic sovereignty and providing a locus Ior public Ireedom. This trend in his thought is
compatible only with the repeated implicit (and nowhere systematized) denial that the search Ior social integration can end in institutions like "our modern states," which
can only provide citizens with "a limited part in the business oI the state."
31
The contradiction runs through Hegel's analysis oI civil society in the Iorm oI two interrelated questions: (1) Is Sittlichkeit or ethical liIe possible only as inherited and
unquestioned ethos to which individual subjects must conIorm in order to be consistent with their very identity, or is it possible to think oI ethical liIe in a truly

Page 93
modern Iorm, permitting and even requiring its own questioning and criticism as well as a plurality oI normatively valued Iorms oI liIe? (2) Is civil society to be
conceived as Sittlichkeit or Antisittlichkeit or as a dynamic combination oI both "moments"?
The two questions are oI course deeply related and may indeed be ultimately the same. To answer them, we must begin with some oI the basic categories oI the
Rechtsphilosophie. Hegel diIIerentiated objective spirit (obfektiver Geist), rationally reconstructed intersubjective structures oI meaning ("spirit") embodied in
institutions ("objective"), in three dimensions: abstract right, morality, and Sittlichkeit (ethical liIe). The diIIerentiation among them is not so much that oI contents
(though these do get progressively richer as we move through the three levels) but among three levels oI moral argumentation. Abstract right represents a Iorm oI
argument on the basis oI dogmatically assumed Iirst principles, as in natural-rights theories. Morality, a level clearly reIerring to Kantian ethics, represents the selI-
reIlection oI the solitary moral subject as the proposed Ioundation Ior a universalist practical argumentation. Finally, Sittlichkeit represents a Iorm oI practical reason
that, through selI-reIlection, is to raise the normative content and logic oI inherited institutions and traditions to a universal level. Only Sittlichkeit allows the exploration
oI normative questions (including "rights" and ''morality") on the level oI concrete, historically emergent institutions and practices that represent, at least in Hegel's view
oI the modern world, the institutionalization or actualization oI Ireedom.
32
Ethical liIe is itselI diIIerentiated in a way (entirely unique to Hegel) that combines the two
dualities oI oikos/ polis and state/society in the three-part Iramework oI Iamily, civil society, and state.
33
Civil society (bùrgerliche Gesellschaft) is deIined variously,
but most revealingly as ethical liIe or substance "in its biIurcation (Ent:weiung) and appearance (Erscheinung)."
34
To understand this deIinition oI civil society, we must examine the notion oI Sittlichkeit more closely. Charles Taylor is surely on solid Ioundations in at least one
dimension oI Hegel's text when he interprets the content oI this notion "as the norms oI a society's public liIe . . . sustained by our action, and yet as already there."
35

According to Taylor, "in Sittlichkeit there is no gap between what ought to be and what is, between Sollen and Sein."
36
Hegel's overall

Page 94
scheme repeatedly stresses the total identity oI the (rational) will oI the subject with laws and institutions,
37
making any clash between particular and universal will,
subject and object, right and duty, impossible or at least irrational.
38
Taylor is on less solid ground when he interprets Moralitàt and Sittlichkeit merely in the Iorm oI opposition. Modern ethical liIe as Hegel unIolds it is distinguished
Irom all ancient ethos because it contains the other two ethical dimensionsrights and universalist moralityon a higher, i.e., an institutionalized, level. Indeed,
according to Hegel, an institutional space is created Ior private morality that should not become "matter Ior positive legislation."
39
On this basis, Hegel could have gone
on to recognize the possibility oI institutionalized conIlict between theory and practice, norms and actuality, as the greatest achievement oI the modern world. That he
did not do so allows Taylor to interpret him primarily as an "ancient," entirely against Hegel's own intentions. OI course, Taylor Iocuses on only the main strand oI
Hegel's conception, not the antinomic whole. Hegel's own deIinition oI Sittlichkeit involves a greater stress on its production and reproduction through selI-conscious
action.
40
Are the bases oI such action to be Iound in Sittlichkeit alone, or in Moralitàt as well, or at least, Ior the modern world, in a Iorm oI ethical liIe that has
incorporated morality, along with the tension between is and ought? When we say that Sittlichkeit, as the norms oI a society's public liIe, is already there, Hegel's
authority takes us only so Iar as to register the institutional existence oI the norms in question, possibly in Iorms oI discourse only, or as legitimations and ideologies.
Their oIten "counterIactual" character is noted by Hegel himselI, Ior example, in the case oI the principles and practice oI positive law. UnIortunately, Hegel did not
discover that modern civil society is characterized by the conIlict not only oI moralities (which he at times seemed to note) but also oI the normative conceptions oI
politics itselI. Thus he did not see that it was possible to establish a new Iorm oI Sittlichkeit containing a plurality oI Iorms oI liIe; this would make consensus possible
only on the level oI procedures, but even such a consensus can lead to some shared substantive premises and even a common identity. He certainly does admit the
possibility oI conIlict between institutionalized norm, the actual basis oI moral opposition, and the practice

Page 95
oI institutions. Primarily Ior this reason, his thought and the social world he describes are open to immanent critique.
Because oI the internal division oI its institutional sphere, civil society is the Iramework par excellence where the tension between is and ought emerges. Our aim is to
show that this division hardly disappears in Hegel's theory even in the state sphere, which is supposed to be the one in which all antinomies are reconciled.
41
Although
Hegel periodically implies that no actually existing state should be considered already rational, he nevertheless holds that ethical (sittliche) substance deIined in terms
oI the identity oI rational selI-reIlection and actualized institutions is the "wirkliche Geist einer Familie und eines Jolks."
42
The absence oI civil society and the
presence oI the Iamily and the state, the latter only as people, are the notable Ieatures oI this deIinition oI Sittlichkeit. Consistently enough, civil society reappears in
the next paragraph only as an "abstract" and "external" version oI Sittlichkeit.
43
The section on the transition between the Iamily and civil society speaks oI ''the
disappearance oI ethical liIe" and its reemergence only as a "world oI ethical appearance."
44
Hegel goes on to speak oI civil society "as a system oI ethical liIe lost in
its extremes."
45
Thus civil society is a level oI Sittlichkeit where the oppositions oI ought/is, subject/object, right/duty, and even rational/actual would all reappear. But it would not be
diIIicult to argue that this level oI Sittlichkeit is its very antithesis, a Gegen- or Antisittlichkeit.
46
Much oI Hegel's discussion oI civil society emphasizes the
disintegration oI the supposedly natural Iorm oI ethical liIe represented by the Iamily in a world oI egotism and alienation. Nevertheless, when he speaks oI the ethical
roots oI the state, he speaks oI the Iamily and the corporation, the latter "planted in civil society."
47
Here is the real sense oI seeing civil society as the "biIurcation oI
ethical liIe," as both Sittlichkeit and Antisittlichkeit, where the unity oI substantial ethical liIe (according to Hegel's Iinal judgment on civil society) is attained only in
appearance.
By Iollowing Hegel's unIolding oI the categories oI civil society Irom the system oI needs and system oI laws to the police (general authority) and corporations, and
even beyond to the estate assembly and public opinion, we gain a depiction oI modern society as a dialectic oI Sittlichkeit and Antisittlichkeit. Only the illusions oI
sys-

Page 96
tem building put an end to this movement in the (highly inconsistent) depiction oI the state as Iully realized but no longer naturally given ethical liIe.
48
We should stop to consider the great importance oI a two-sided understanding oI Hegel's concept oI civil society. II we were to interpret it only as alienation, social
integration would have to be conceived exclusively on the levels oI Iamily and state. In relation to civil society, then, the prescriptive or critical dimensions oI the theory
would come to the Iore, but a transcendent version oI critique
49
would have to take the Iorm oI romantic communalism, with Iace-to-Iace relations as its normative
standard, or oI statism, whose selI-legitimation could take various republican or nationalist Iorms. II civil society were interpreted exclusively in terms oI the Iorms oI
social integration that emerge here, however, the descriptive and tendentially conIormist elements oI the theory would come Iorward, and the negative aspects oI
bourgeois civil society that Hegel was one oI the Iirst to point out in detail would be lost Irom view. The richness and power oI Hegel's social theory lies precisely in his
avoiding both a transcendent critique oI civil society and an apology Ior bourgeois society.
Many interpreters oI Hegel see the integration oI modern society as a series oI mediations between civil society and the state. However, this way oI putting the issue is
already a hostage to the statist dimension in Hegel's thought. II we are not to accept Irom the outset that the only important line oI thought in Hegel assumes the state
(but which element oI the state?) as the highest, most complete and universal level oI social integration, the issue oI mediation should be put diIIerently. On a more
abstract level, it should already be clear that mediation is between Antisittlichkeit and Sittlichkeit. On a more concrete level, however, it is the distance between
private and public that is to be mediated, iI we understand the Iormer as the vanishing point where the social integration oI the Iamily is dissolved beIore the
mediations characteristic oI civil society begin. Thus it is our thesis that the mediation oI Antisittlichkeit and Sittlichkeit culminates in a notion oI public liIe that Hegel
only inconsistently identiIied with state authority.
50
AIter Marx's early critique oI Hegel's philosophy oI the state, little would be leIt oI this identiIication, except Ior the
small detail oI the role oI statism, in

Page 97
the critiques oI the capitalist market economy in the next century and a halI, including those by Marx's own Iollowers.
51
In both Hegel's and Marx's work, however,
the statist trend is in a powerIul tension with antistatist options.
As any reader oI Hobbes knows, the road to statism is prepared by the identiIication oI society outside the state with egotistic competition and conIlict. Such is also
the outcome oI the well-known Marxian identiIication oI civil and bourgeois society.
52
The traditional German translation oI societas civilis as bùrgerliche
Gesellschaft is not the only basis oI this theoretical move. Hegel himselI repeatedly identiIies bùrgerlich as bourgeois,
53
and nowhere does he use the adjectival Iorm
in the classical sense oI Bùrger or citoven. When he states that individuals as Bùrger oI civil society, the "external state,"
54
are private persons,
55
he participates in a
Iundamental shiIt in the concept oI civil society away Irom the original meaning oI citizen society. At the same time, iI the bourgeois were to be understood as homo
oeconomicus, then clearly it would represent only one dimension oI what Hegel deIines as the subject oI civil society, the concrete person.
56
OI course, the latter is
Iirst deIined as "a totality oI needs and a mixture oI natural necessity and arbitrary will (Willkùr)." But this is only Hegel's starting point: The system oI needs is the Iirst
level oI civil society. As the argument proceeds through the next levels"the administration oI law" and "general authority and corporation"we encounter the
concrete person again under new headings: legal person, client oI general authority, and association member.
57
It is only on the level oI the system oI needs, the
description oI which Hegel derives Irom political economy,
58
that a radical depiction oI civil society as Antisittlichkeit is consistently upheld. For example, when
Hegel deIines civil society as a system oI Sittlichkeit "split in its extremes and lost,''
59
he has in mind a condition where egoistic individualismone extremeis
integrated by means oI an abstract generality (universal interdependence)the other extremethat is entirely Ioreign to the will oI individuals. Accordingly, civil
society as "an achievement oI the modern world"
60
involves the creation oI a new type oI market economy that integrates the "arbitrary wills" oI selI-interested
economic subjects by means oI an objective and "external" process that achieves a universal result unintended and unantici-

Page 98
pated by the participants.
61
This objective process can be reconstructed by a science speciIic to the modern world, namely political economy, that Hegel regards as
being entirely parallel to the sciences oI nature.
62
Hegel's model oI integration on the level oI the system oI needs takes oII Irom Adam Smith's description oI the selI-regulating market as an invisible hand linking selI-
interest and public welIare. But his arguments are less economic than sociological, even iI the tremendous process oI economic growth implied by the modern market
economy underlies the whole thesis.
63
He sees three levels oI integration in this context: needs, work, and "estates." Needs in modern society become more and more
abstract in the Iorm oI money, which makes everyone's needs commensurable. It is monetarization that makes the general recognition and satisIaction oI needs
possible. Hegel also sees the underside oI the process: The abstraction oI needs allows Ior their tremendous expansion. And the result oI the limitless expansion oI
needs can only be great luxury and extravagance alongside permanent want, i.e., the inability oI some to satisIy even basic needs.
64
Work in modern society mediates
particularity and universality through the process oI value creation (the particular work oI the individual creating products that are commensurable with the products oI
all others) and the division oI labor, leading to the "dependence oI men on one another and their reciprocal relation."
65
Again Hegel sees the underside oI the process,
this time in ''the dependence and distress oI the class" that is tied to Iorms oI increasingly one-sided and restricted work that "entail inability to Ieel and enjoy the
broader Ireedoms and especially the intellectual (geistigen) beneIits oI civil society."
66
Finally, Hegel has a theory oI stratiIication according to which the diIIerentiated
social strata oI civil society that he still calls Stànde (estates or orders) integrate individuals as members oI "one oI the moments oI civil society" with its own rectitude
and status honor ( Standesehre).
67
Hegel insists that his estates are modern, and that individuals become part oI them Ireely, through their own achievement, rather than ascriptively.
68
Nevertheless, it is
clear that he has only partially discovered the speciIically modern principle oI stratiIication, namely socioeconomic class.
69
The working class, to which (as

Page 99
Avineri showed) he restricts the new term class (Klasse), is not included in his scheme oI agricultural, business, and universal (i.e., bureaucratic) estates.
70
This is a
serious omission, especially because Hegel claims that his estates correspond to economic diIIerentiation. In Iact, however, he did not discover the speciIically modern
Iorm oI stratiIication based on socioeconomic divisions oI interest and lines oI conIlict because he did not adequately distinguish between diIIerentiation and integration.
Thus, his theoretical instruments Iailed him when he conIronted an increasingly diIIerentiated class, the victim oI poverty and the alienation oI labor, that he thereIore (as
it turned out wrongly) considered only as being unable to integrate into, and unable to contribute to the integration oI, civil society.
Strictly speaking, integration through estates does not belong to the level oI the "system oI needs," where integration is the Iunction oI objective, unwilled processes.
This is shown by the Iact that the analysis simply duplicates what Hegel elsewhere assigns to the Iamily (the agricultural class
71
), to the corporation (the business
class
72
), and to the general authority (the class oI civil servants
73
). It is only what Hegel considers the underside oI this process oI the emergence oI new,
nonascriptive status groups that belongs to the socioeconomic level oI his analysis. Accordingly, the working class represents a Iorm oI inequality produced by civil
society
74
in which the absence oI inheritance and otherwise unearned income, as well as a speciIic Iorm oI liIe, makes estate membership inaccessible and exposes
individuals to the hazards oI economic contingencies beyond their control.
75
Taken together, need, labor, and diIIerentiation achieve a level oI universality in civil society only at great social cost. Hegel is acutely conscious oI this even iI he does
not and cannot notice the level oI the corresponding potential oI conIlict. Unlike some political economists he knew (in particular, Ricardo), he did not readily
thematize the problem oI conIlict in relation to the working class,
76
perhaps because oI his belieI that estates (i.e., new types oI status groups) alone constituted the
modern principle oI stratiIication.
77
Nevertheless, he did understand the "system integration" oI civil society to be highly unstable, though he did not pose this issue in
terms oI action- theoretic categories. Even so, more than any

Page 100
political economist, he understood that social integration must occur outside the system oI needs in order Ior the market economy itselI to Iunction. Unlike early
modern political philosophers in the natural-law tradition, however, he does not conIine this level oI integration to the exercise oI sovereign power, to the sphere oI the
state, or to the Iamily, another possible choice. It was in conscious opposition to these theoretical options that he developed a theory oI social integration that
constituted one oI the Iounding acts oI modern sociology, or at least oI the paradigm developed by Durkheim, Parsons, and Habermas, among others.
Hegel's theory oI social integration moves through six steps: legal Iramework (Rechtspflege); general authority (Poli:ei); corporation; the (bureaucratic) executive; the
estate assembly or legislature; and public opinion. While the Iirst three oI these are developed as parts oI the theory oI civil society, and the second three belong to the
theory oI the state, or rather constitutional law, the argument turns out to be essentially continuous.
78
We should perhaps think oI these as two lines oI argument, even
iI Hegel's movement back and Iorth between them is so constructed as to avoid the appearance oI such diIIerentiation. It is this double argument concerning social
integration on which we shall concentrate.
As we have shown, the system oI needs in Hegel's theory is itselI integrated, but in a manner that is "external" (outside oI will and consciousness), incomplete (less than
Iully universalist), and selI-contradictory. Integration beyond the system oI needs operates according to two diIIerent logics: the logic oI state intervention into society,
and that oI the generation oI societal solidarity, collective identity, and public will within civil society itselI. Through most oI the text, the unIolding oI the two logics can
be clearly diIIerentiated: One seriesuniversal estate, general authority, crown, executiveexpresses the line oI state intervention; anotherestates, corporation,
estate assembly, public opinionIollows that oI the autonomous generation oI solidarity and identity.
Only in the "administration oI law" is it diIIicult to separate the two lines oI argument. In Hegel's exposition, this level represents the possibility oI the universally (or at
least generally) valid resolution oI the clash oI particulars in civil society. The overcoming oI Gegensittlichkeit as the division oI particular and universal begins

Page 101
here, but in a Iorm that is capable oI generating only a limited collective identity. The legal person identiIies with the collective only in the Iorm oI abstract obligations.
Hegel not only recognizes the noneconomic presuppositions oI economy in the modern sense, in the law oI property and contract,
79
but he also sees that their
implications go Iar beyond the economy. In particular, the publication oI the legal code and, even more, the publicity oI legal proceedings are changes oI universal
signiIicance and validity that make possible the emergence oI a universalist sense oI justice.
80
This argument becomes Iully intelligible in the context oI Hegel's
understanding oI the concept oI the public (Offentlichkeit) that goes beyond the Roman law dichotomy oI public and private. We shall analyze this concept in detail
below, but here we simply stress that Hegel sees a Iunctional relation between modern law and the system oI needs: Each is necessary Ior the emergence and
reproduction oI the other. He also insists, however, that the institutionalization oI subjective right and objective law protects the Ireedom and dignity oI modern
subjects in a way that private persons rather than isolated individuals brought together in a public process can mutually recognize.
81
To Hegel, the institutionalization oI
right as law requires both state action (he strongly preIers statutory codiIication to precedent-based adjudication
82
) and autonomous cultural processes. He is neither
a legal positivist nor a natural-law theorist nor even a historicist. For Hegel, universal rights have more than just a historically restricted validity even iI they emerge in
cultural development and can be universally recognized only through a process oI education (Bildung) that has become possible in civil society.
83
Universal rights do not, however, attain objective existence without being posited as law (geset:t als Geset:), which involves legislation, codiIication, and
administration by public authority (òffentliche Macht). Without autonomous cultural processes that create them, rights cannot acquire validity or recognition. But
without the various necessary acts oI the state and its organs, neither true deIinition nor a systematic relation to other rights is possible.
84
Only the combination oI the
two yields obligatory Iorce. Hegel wisely recognizes the possible discrepancy oI the two moments,
85
cultural and political, "between the content oI the law and the

Page 102
principle oI rightness."
86
Yet within the analysis oI law, he can oIIer only some Iormal and procedural requirements that legislators and judges should not violate, in
particular the requirement oI publicity and the Iormal generality oI law. Presumably he expects a closer Iit between the principle oI right and positive law regarding
substantive legal rules through the ability oI the other institutional mediations oI his theory to create law.
Integration through the State
Hegel cannot maintain the complementarity between societal and statist strategies oI social integration beyond his analysis oI the administration oI law. From this point
on in the argument,
87
the two types oI strategies become identiIied with diIIerent institutional complexes. The statist trend in Hegel's thought, anticipating Marx and
especially Marxism, is clearly connected to the notion oI civil society as Gegensittlichkeit, rooted in the analysis oI the system oI needs.
88
The pathological
consequences oI the system oI needs, involving extremes oI wealth and poverty, want and luxury, as well as a severe threat to the humanity and very existence oI the
class oI direct labor, call Ior measures that allow Hegel to anticipate Ieatures oI the modern welIare state.
89
In particular, a state bureaucracy (the universal class, the
class oI civil servants) is called upon to deal with the dysIunctional consequences oI the system oI needs, in two Iorms.
(1) The universal estate is called upon as the key mechanism to deal with the antagonism oI estates. Here the analysis suIIers Irom a lack oI reIerence to the class
Hegel knows to be both the product oI the modern economic order and the most endangered by it. Nevertheless, the assumption that estates produce both integration
within strata and antagonism between strata does represent an important opening to a sociology oI conIlict. In this context, Hegel maintains that the status-honor and
economic condition oI the estate oI civil service imply that particular or "private interest Iinds its satisIaction in its work Ior the universal."
90
The salaried condition oI
the oIIicial, the requirement Ior open access to oIIices, and the limits against turning oIIices into private patrimonies all inhibit the Iormation oI the sort oI selI-interested,
closed estate that

Page 103
characterized most traditional bureaucracies. The education oI the public servant makes the idea oI public service conscious and deliberate.
91
Thus, according to
Hegel, the universal estate is in a unique position to resolve the antagonism oI estates.
There is no need to repeat Marx's brilliant 1843 critique oI the pretensions oI Hegel's view oI the universal estate, which pinpointed its particular interests and status
consciousness. Hegel managed to delude himselI on this score partly because oI the statist strain in his thought, and partly because he did not see any reason to
consider the social antagonism implied by the existence oI the "class oI direct labor." Being incapable oI intraclass integration, workers in this view do not seem to be
capable oI interclass conIlict. The dysIunctional consequence oI the plight oI this class is seen in the existence oI an anomic mass, the Pòbel, whose integration requires
measures that aim at individuals (i.e., clients) rather than integrated groups. But with the poorest stratum removed Irom the Iield oI analysis, the idea that the
bureaucracy represents a general interest needs to be reconciled only with the interests oI the landed classes.
92
Hegel's discussion oI civil servants takes place in two sections oI his analysis: those on the system oI needs oI civil society and the executive oI the state. This is
justiIied by the Iact that the bureaucracy is both a social stratum and a state institution.
93
But Hegel's theoretical decision disguises the Iact that this estate diIIers Irom
others in two respects. First, it is constituted by the state and not by the societal division oI labor. Second, in the state the bureaucracy Iinds its institutional place in the
executive rather than in the estate assembly. Thus, Hegel's argument concerning the Iortunate double meaning oI the German term Stànde,
94
reIerring both to social
orders and to a deliberative assembly, does not apply. By calling the bureaucracy a Stand, Hegel misses the opportunity to discover the second, primarily modern,
Iorm oI stratiIication whose constitutive principle is political power. Even more importantly, he disguises the statist principle oI the Iorm oI social integration under
consideration.
The way the bureaucracy is to accomplish the integration oI antagonistic estates reveals at least some oI the consequences. The state executive or political bureaucracy
has the role oI "subsuming

Page 104
the particular under the universal" by applying the laws. Hegel accepts the parliamentary assumption that an estate assembly is capable oI generating a public and
general will. But he believes that in civil society all the particular interests will reappear, and that Ior this reason outside the state sphere proper the bureaucracy must
be the agent oI universality. The Iact that he Ieels compelled to admit that the authority oI local communities (Gemeinden) and corporations is needed as a "barrier
against the intrusion oI subjective caprice into the power entrusted to the civil servant"
95
shows, though, that Hegel is aware that reality can be quite diIIerent Irom his
idealized depiction. Presenting the bureaucracy as an estate oI civil society is thus not only a way oI disguising the actual level oI state intervention he advocates but is
also away oI deIlecting the responsibility Ior dysIunctional or even authoritarian intervention Irom the state to a social group and to the subjective caprice oI its
members.
(2) The model oI integration through state intervention is Iurther developed in the theory oI police or general authority (Poli:ei or allgemeine Macht). UnIortunately,
the modern term "police" does not cover Hegel's meaning here. In accordance with earlier absolutist usage, he means more than the prevention oI crime and tort and
the maintenance oI public order. However, Hegel also uses the term "general authority" in senses not covered by the section on the Poli:ei. Thus, it may be best
simply to list his actual uses oI this concept: surveillance (linked to crime and tort);
96
intervention in the economy in the Iorm oI price controls and regulation oI major
industrial branches;
97
and public welIare in the Iorm oI education, charity,
98
public works,
99
and Iounding oI colonies.
100
The idea behind linking these apparently diverse areas is not quite coherent. The Iunctioning oI the system oI needs is linked in Hegel's conception to two rather
diIIerent Iactors: a centriIugal dysIunctionality based on the subjective caprice and carelessness oI individuals, and systematically induced eIIects largely based on
worldwide competition and the division oI labor. The police represent state penetration into civil society to serve the interests oI justice and order by compensating Ior
both oI these phenomena without eliminating their basic causes, which lie in the dynamism

Page 105
oI the system oI needs. As a result, the centriIugal and anomic consequences oI conIlict are diminished but not entirely done away with. "Crime prevention" and the
punishment oI criminals do not eliminate crime but keep it within tolerable limits. Provisions Ior social welIare and public education do not abolish conIlict and
alienation, but they can prevent the decline oI the class oI labor to the status oI a rabble (Pòbel). In these cases and also in the case oI price and production controls,
the goal Hegel espouses is compensation Ior the dysIunctional side eIIects oI the new type oI market economy, a core dimension oI modern civil society. The details oI
his analysis do not always make clear whether he is deIending precapitalist Iorms oI paternalist intervention or anticipating Ieatures oI a modern welIare state. The
general conception, however, involves reactive compensation Ior the eIIects oI a genuine market system more than proactive, statist substitution Ior market Iunctions.
The statist Ieature oI the doctrine oI the police lies elsewhere. Hegel does not systematically distinguish between state intervention in the Iorm oI economic steering
(e.g., price controls in a system oI market prices) and intervention in noneconomic spheres oI liIe (e.g., surveillance). While Irom the point oI view oI market
dysIunction, each oI these measures represents post Iacto compensation, surveillance and other Iorms oI social control are proactive Irom the point oI view oI
noneconomic Iorms oI liIe, substituting, as Tocqueville noted, statized relations Ior horizontal social ties.
101
A similar proactive character can be noted in the roles oI
general authority dealing with trusteeship and education.
102
The problem, oI course, is not that Hegel hopes to prevent orphans and the children oI the poor Irom
Ialling into poverty, but that he deIines the remedies in terms oI a "right" oI society as a whole rather than the rights oI the individuals, Iamilies, and communities
concerned. Once again Hegel replaces horizontal social interaction and solidarity by vertical ties based on state paternalism. Even iI it were true that civil society
destroys the Iamily ties that protected individuals in premodern society, the idea oI the general authority (the state) "taking over the role oI the Iamily Ior the poor"
103
is
a mystiIication oI measures that do not produce but replace social solidarity.

Page 106
Social Integration through Civil Society
Hegel does not claim that on this level the state produces a thoroughgoing uniIication oI society. Moreover, the kind oI universality it achieves here amounts to a Iorm
oI "external" imposition and control.
104
In civil society we encounter the state only in the Iorm oI externality, and the metaphor oI civil society as "universal Iamily" is
entirely misplaced in the theory oI the police or general authority. This metaphor belongs instead to the second strand oI Hegel's conception oI social integration, the
solidaristic strand that runs Irom the Iamily to the corporation, the estate assembly, and public opinion. But, since Hegel (wrongly) considers the integrating role oI the
Iamily to be negated in civil society,
105
the corporation becomes the starting point oI the selI-integration oI civil society. As in the case oI the police and the estates,
one can legitimately question whether Hegel's theory oI the corporation revives a premodern Iorm oI social liIe or anticipates a postliberal Iorm oI social integration.
We shall return to this question, noting here only that Hegel was both harshly critical oI the revolutionary and liberal attacks on the old corporate entities and in Iavor oI
a Iorm oI corporate organization signiIicantly diIIerent Irom that oI the old regime.
106
Indeed, he proposed and deIended a version oI the corporation that was open to
entry and exit, that was based on no ascriptive or hereditary principle, that was voluntary and not all-inclusive, and that did not imply any suspension oI the individual
rights oI members with respect to the corporate body. Unlike the case oI a modern union, however, both employers and employees would be members oI
corporations in the economic sphere. Moreover, Hegel does not restrict corporate organization to that sphere: Learned bodies, churches, and local councils are also
included in the concept.
107
The primary Iunctions oI the corporation in Hegel's theory are socialization and education. The business association in particular is meant to combine vocational training
with training Ior citizenship. Thus all oI corporate liIe, assuming the already mentioned modernization oI its structure, helps to overcome the gap civil society produces
between bourgeois and citizen by educating individuals to internalize the common good and develop civic

Page 107
virtue. In the process, solidarities are expected to develop that would aIIect the motivational structure oI individuals, substituting collective concerns and identiIications
Ior egoistic ones. In this context, Hegel's problem was the same as Rousseau's, namely, how to move Irom the particular to the general, given modern individuality. But
his answer is signiIicantly diIIerent, because Hegel did not believe that the reality oI the modern large-scale state or oI a modern civil society with a dynamic system oI
needs could or should be imagined away, or that individuals who are entirely egotistical in private liIe can attain the general in the political sphere. In his view, generality
can be attained only through a series oI steps that incorporate something oI the public spirit in what is juridically the private sphere. The corporations that Rousseau, his
natural-law philosophical Iorebears, and his revolutionary republican successors sought to banish Irom social liIe, replace the particularity in Hegel's theory with a
limited Iorm oI generality on a level where resocialization is actually possible.
While the corporation represents a crucial step in the development oI the strand oI Hegel's thought that stresses the selI-integration oI society, the antinomy oI his
political position is nevertheless visible in it. Like Montesquieu beIore him and Tocqueville aIter him, he sought an intermediate level oI power between individual and
state; he Ieared the powerlessness oI atomized subjects and sought to control the potential arbitrariness oI the state bureaucracy.
108
But at the same time, in line with
his doctrine oI the state, he wants to deIend a model oI socialization thatwill make the transition to a state-centered patriotism plausible. In this context, Hegel's aim is
to provide a smooth transition based in everyday liIe Irom the Geist oI the corporation as the schoolhouse oI patriotism to the Geist oI the state where patriotism is to
achieve its Iull "universality."
109
Much depends, oI course, on whether the conception oI the state implied here is based on a public, parliamentary generation oI
identity or a bureaucratic-monarchic imposition oI unity. But since the antinomy is not resolved on the level oI the state, the role oI the corporation in political education
also becomes ambiguous. This, in turn, aIIects the the relation oI the corporation to the general authority; as Heiman shows, Hegel was never able to decide between a
medievalist doctrine involving

Page 108
corporate independence and legal personality and a Roman law conception stressing state control and oversight.
110
Whatever the ambiguities oI Hegel's corporate doctrine, the diIIerent center oI gravity here when compared to the concept oI the police cannot be overlooked. Both
police and corporation are at times identiIied as the individual's second Iamily. They also share some Iunctional assignments, such as education. Furthermore, the
normative justiIications produced Ior each are equally convincing. The corporation is a second Iamily small and determinate enough in its purpose to allow genuine
participation by its members. These members, however, include only a part oI the population; while it appears general with regard to its members, the corporation
inevitably represents a particular interest with respect to other groups and those not ''incorporated." Nevertheless, the corporation is capable oI creating internal
motivations, and it does not depend on external sanctions guaranteeing compliance. On the other hand, the regulation oI the police is universalist and ought not to allow
the Iormation oI particular clusters oI interests. However, the activity oI the police does rely on external sanction, involves no participation oI those concerned, and
does not lead to the Iormation oI autonomous motivation.
As the comparison oI police and corporation shows, statism in Hegel's thought is linked not only to some kind oI political opportunism but also to the idea oI
universality, without which no modern conception oI justice is possible. Hegel has good reasons not to make a deIinitive normative choice between police and
corporation, between abstract universality and substantial particularity. These moments are sundered in civil society, and it is Hegel's thesis that they can be reunited
only in the state. It would be only on this level that the corporation, as the second ethical root oI the state (aIter the Iamily), would achieve its universality.
Our reconstruction oI Hegel challenges interpretations suggesting that the antinomies oI civil society are resolved on the supposedly higher level oI the state. Instead,
we would argue that it is more IruitIul to interpret Hegel's thought as dualistic or antinomic on both levels. What we crudely label as "statist" and "solidaristic" trends in
his thought appear in the analysis oI both civil society and state. Accordingly, the doctrine oI the state itselI can be analyzed in

Page 109
terms oI these two trends. Thus, it would be a mistake to oppose to the statist dimension of Hegel´s thought a quasi-liberal conception according to which
civil societv, as opposed to the state, is the onlv source of genuine norms. Such a view would be all the less deIensible because oI the unavoidable element oI
particularism attached to the important intermediary bodies oI civil society. Thus, the transition to a key norm oI modernityuniversalitycannot occur without some
participation oI state institutions. Even iI we were to note that the protection oI the individual rights oI members can be written into the charters oI modern
corporations, the establishment oI universal rights as positive law presupposes, as we have seen, the activity oI the state. But which dimension oI the state? The
question we must consider is whether, in Hegel's theory, the estate assembly and public opinion or the executive bureaucracy and public administration is the locus
and source oI the highest level oI social integration and will Iormation.
In Hegel's conception, we should recall, the police represent the penetration oI the state into civil society. Analogously, the estates assembly represents a penetration
oI civil society into the state. However, the civil society represented in the state through the estate assembly is already organized; to Hegel the presence oI an atomized
civil society in the state would be most regrettable. According to the Iree but convincing translation oI Knox:
The circles oI association in civil society are already communities. To picture these communities as once more breaking up into a mere conglomeration oI individuals as soon they
enter the Iield oI politics, i.e., the Iield oI the highest concrete universality, is eo ipso to hold civil and political liIe apart Irom one another and as it were to hang the latter in air,
because its basis could then only be the abstract individuality oI caprice and opinion.
111
This conception directly links the estates and corporations oI civil society with the assembly oI estates. While Hegel at Iirst stresses the link oI estates to the legislature,
as indicated by the German term Stànde, the more important theoretical Ioundation oI the assembly is in Iact the corporation, the existence oI which is the only real
evidence provided Ior the claim that organization and community are possible in an otherwise atomized civil society. The deputies oI

Page 110
civil society are "the deputies oI the various corporations."
112
Earlier, this statement is limited and expanded. Atavistically, the agricultural estate (suddenly meaning
only the nobility) is to be directly present, as in the assemblies oI the Stàndestaat. The business estate, on the other hand, is represented by the deputies oI
associations, communities, and corporations (Genossenschaften, Gemeinden, Korporationen), which are all incorporated Iorms oI association. Hegel does not even
Ieel the need to indicate and justiIy his exclusion Irom political liIe oI the one class, direct labor, that is supposedly totally disorganized.
113
More important than the
conIormist and conservative elements in his thought, however, are his reasons Ior recommending his particular version oI representative government. According to
Hegel, when civil society elects its political deputies, it "is not dispersed into atomistic units, collected to perIorm only a single and temporary act, and kept together Ior
a moment and no longer."
114
Rather, in the process oI deliberating and choosing deputies, the associations and assemblies oI social liIe acquire a connection to politics
in the same act that gives politics a Ioundation in organized social liIe. It is precisely at this level, at the point where civil society and the state interpenetrate, that Hegel
rediscovers and integrates, without explicitly saying so, the ancient topos oI political society.
The estate assembly has the role oI completing the job begun by the corporation, but on a societywide level oI generality that he (and especially his English translator)
oIten reIers to as "universality." This job is to bring public aIIairs and, even more, public identity into existence.
115
Again parallel to the doctrine oI the corporation, the
legislature is regarded as a mediating organ, this time between the government (Regierung) and the people, diIIerentiated as individuals and associations.
116
The
Iormer is thus prevented Irom becoming tyrannical and the latter Irom becoming a mere aggregate, a mass with an unorganized and thereIore dangerous opinion. Hegel
oI course stresses the role oI the estate assembly in legislation and even constitution making,
117
but his main interest throughout is in the constitution oI the agent oI
legislation and, even more, its proper medium. The category oI publicity indicates that only the genuine representatives oI the public are legitimately entitled to make the
laws. The laws they enact are to be considered

Page 111
legitimate only iI the procedures oI public deliberation are rigorously Iollowed. Since Hegel insists on genuine and unconstrained discussion and deliberation, he
emphatically rejects the imperative mandate, the principle oI the traditional Stàndestaat. The assembly must be "a living body in which all members deliberate in
common and reciprocally instruct and convince one another."
118
Hegel's vehement insistence on genuine publicity in the legislature (as well as the courts) has other important grounds. He wishes to promote knowledge oI public
business in society and (however inconsistently) to make the estate assembly susceptible to the inIluence oI public opinion. Quite like Tocqueville, Hegel is ambivalent
concerning public opinion. DeIined as "the Iormal, subjective Ireedom oI individuals to express their own judgments, opinions, and recommendations concerning
general aIIairs whenever collectively maniIested,"
119
public opinion is internally contradictory and "deserves as much to be respected as despised (geachtet als
verachtet)."
120
Respect is due because oI a hidden strain oI rationality that is, however, buried and inaccessible to public opinion's opinion about itselI because oI its
concrete, empirical Iorm oI expression. Interpreting public opinion is thus the role oI intellectual and political elites.
121
In order to promote the Iormation oI public
opinion, Hegel supports extensive Ireedom oI public communication (especially speech and press), and he worries only slightly about possible excesses. Indeed, he
believes that the genuine publicity oI legislative debates has a good chance oI transIorming public opinion and eliminating its shallow and arbitrary components,
rendering it harmless in the process.
122
Nevertheless, it is also implied here that the debates oI the assembly can transIorm public opinion precisely to the extent that its
essential content and elements oI rationality are raised to a higher level. In this sense, not only does the political public oI the legislature control public opinion (Hegel's
stress), but a prepolitical public sphere plays an important role in constituting public liIe in the political sense.
The concept oI public opinion developed by Hegel is not Iree oI the antinomies oI his political thought. The statist trend in this context is expressed in the concern to
control and disempower public opinion in order to make it compatible with the management oI the state. The solidaristic trend, on the other hand, involves the

Page 112
raising oI public opinion to a higher level oI rationality in a parliamentary Iramework between state and society, itselI exposed to the controls oI publicity. From the Iirst
point oI view, public opinion is ultimately a threat, and the proper relationship to it on the part oI political (including parliamentary) elites is manipulative. From the
second point oI view, public opinion is the condition oI possibility oI political public liIe, and the proper relationship to it on the part oI elites would have to be one oI
public dialogue in which truth would be an open question to be decided by the more convincing arguments rather than the a priori possession oI one oI the sides. The
public sphere oI the estate assembly plays a role in enlightening and educating public opinion precisely because truth here is not known in advance but rather emerges
during the debate itselI, along with the virtues that can serve as examples to the larger audience.
123
One trend in Hegel's thought implies that in those states where the
liIe oI the legislature is genuinely public the structure oI public opinion will itselI change: "What is now supposed to be valid gains its validity no longer through Iorce,
even less habit and custom, but by insight and argument (Einsicht und Grùnde)."
124
At other times, however, the dialogue model oI rational political deliberation is
restricted to the parliamentary public sphere. In these contexts, the statist trend in Hegel's thought, supported by the Ialse analogy between the search Ior scientiIic
truth and the attainment oI normative truth in politics, stops him Irom extending the model to the public sphere as a whole.
At issue here, as well as in Hegel's political theory as a whole, is the ultimate locus and nature oI public Ireedom. We accept the interpretation according to which
Hegel sought to develop a political doctrine in terms oI a whole series oI mediations that relativize the Roman law distinction between private and public law.
125
But
we accept it with two reservations.
First, we see the mediations as two distinct series: civil servants/ police/executive/crown, and estates/corporation/estate assembly/ public opinion. The two express the
conIlicting trends in Hegel's thought. Indeed, the very manner in which they mediate the spheres regulated by private and public law is signiIicantly diIIerent in each
case. The Iirst series involves public law categories taking on both private and public roles. The second indicates private law

Page 113
entities developing structures oI publicity and taking on public Iunctions rooted in these structures.
126
This second pattern is the same as the model in which
constitutional rights constitute the public law rights oI private subjects.
127
Once these two patterns are separated, however, the meaning oI the public sphere in Hegel
becomes uncertain. Is its primary paradigm that oI public authority or that oI public communication? And iI he maintains both paradigms, what is to be their
relationship?
Second, we do not accept the implicit identiIication oI state and public presupposed by the interpretation, or the idea that each succeeding step in Hegel's exposition
represents (even in terms oI his own argument) an unambiguously higher level oI public liIe than the one beIore. For Hegel, undoubtedly the highest purpose oI public
liIe is to generate a rational universal identity that he equates with the patriotic ethos oI the state. What remains unclear is whether the generation oI this ethos is
assigned to a state sphere dominated by the executive and linked only to the projections oI the state into civil society, or to a sphere dominated by a legislature
drawing on autonomous societal resources such as the corporation and public opinion. The issue cannot be decided iI we stress the problem oI mediating between
private and public realms alonemost categories oI Hegel's theory oI Sittlichkeit, beginning with the system oI needs, provide such mediations. But it can be decided
iI we link the process oI generating a modern, rational collective identity to the concept oI public Ireedom that Hegel repeatedly uses in this context, that is, to a
process that allows the eIIective participation oI individuals in the Iree shaping oI the meaning oI a "we." Obviously, public Ireedom is quite a bit more than the kind oI
Ireedom available to the agents oI the system oI needs, who cannot participate in the Iormation oI any collective identity whatsoever. But Hegel also registers serious
doubts about whether the modern state as such can be the locus oI public Ireedom, doubts that run completely contrary to the statist strain in his thought.
We should note once again that, while Hegel nowhere systematizes a conception oI the public sphere (Offentlichkeit), the categories oI public authority, public
Ireedom, public spirit, public opinion, and publicity play key roles in his work. Let us recall Ilting's thesis that the Philosophv of Right seeks above all to synthesize the
negative

Page 114
Ireedom oI modern liberalism and the positive Ireedom oI ancient republican thought. The categories oI the public sphere represent important ways in which
republicanism could be sustained in Hegel's thought aIter his supposed conservative turn. But even here there is an essential diIIerence with ancient republicanism.
Instead oI restricting the Iormation oI public Ireedom to a single social levelpolitical societyHegel works out a modern republican theory in which a whole series
oI levels have key roles to play, including the public rights oI private persons, the publicity oI legal processes, the public liIe oI the corporation, and the interaction
between public opinion and the public deliberation oI the legislature. Not all oI these processes have a public political purpose. Yet they are the stages oI learning
leading to the Iormation oI public identity. What is common to all oI them is the Iree public participation oI those concerned in the Iormation oI decisions.
128
The
public purpose oI the acts oI the police, at times identiIied as general (allgemeine) and even public (òffentliche) power, is beyond doubt Ior Hegel. The same is true
oI the acts oI the executive and, in a Rechtsstaat, oI the crown as well. Yet in these cases Hegel speaks neither oI the Iormation oI public spirit nor oI the actualization
oI public Ireedom. In Iact, it has been noticed that Hegel's most explicit discussion oI public Ireedom juxtaposes the corporation, belonging to civil society, to the
modern state:
In our modern states (modernen Staaten) citizens have only a restricted part in the general (allgemeinen) business oI the state; yet it is essential to provide menethical
entitieswith activity oI general character over and above their private business. This general activity which the modern state does not always provide is Iound in the
corporation.
129
In this passage Hegel not only registers the tension between the modern state and public liIe but identiIies a diIIerent locus Ior public Ireedom than did classical
antiquity. The corporations are, in his words, "the pillars oI public Ireedom (òffentlichenFreiheit)."
130
Yet Ior Hegel the public Ireedom possible in the corporation,
involving a relatively high level oI participation, cannot be primary in society as a whole. Pelczynski and others are surely right when they argue that Hegel believed that
he had proved that "the |modern| state is the actuality oI concrete Ireedom."
131
This argument is supported, in general, by the greater universality oI the

Page 115
estate assembly, this veritable corporation oI corporations, over the inevitably particularistic societal associations. But it also disguises the reality oI the modern state as
a hierarchy oI oIIices, as the monopolistic possessor oI the means oI violence, and as a compulsory association. By reversing the sociologically obvious hierarchy oI
the modern state, making the legislature primary and the executive secondary, Hegel is constructing a legitimation both in the sense oI counterIactually justiIying a
structure oI authority and in the sense oI establishing a set oI normative claims open to critique. These critical potentials come into view, Ior example, when the
assembly Irom which the normative claims oI state are drawn is depicted as its penetration by civil society.
Hegel, the peerless social theorist oI his time, was clearly aware oI the sociology oI the modern state. We are Iortunate to have at our disposal Ilting's careIul
reconstruction oI Hegel's turn Irom an earlier conception stressing the Ireedom oI the citizen in the state to one stressing the Ireedom oI the state.
132
The shiIt may
well have had independent intellectual motivations, which were then reinIorced by Hegel's reaction to the reactionary Karlsbad decrees. Hegel knew and rejected both
absolutist and revolutionary statism, as so much oI the Rechtsphilosophie demonstrates. Is it too IarIetched to assume that a reactionary turn in Prussian politics made
him realize (as did Tocqueville soon aIter) that Ieatures oI two supposedly aberrant versions oI the modern state belonged to its ideal type instead? II this were so, the
shiIt to institutions oI civil society as the pillars oI public Ireedom would be logical and also indispensable Irom the point oI view oI strengthening this dimension in the
parliamentary institutions oI the state. Thus, Hegel in his mature text not only restricted the possibility oI the citizen's Ireedom in the state but also expanded, in Ilting's
words, the liberties (Freiheitsrechte) oI civil society into rights oI participation (Teilnehmerrechte).
The most obvious objection to our reading oI Hegel would be that he himselI did not admit and, Ior systematic reasons, would have rejected the idea oI two
unreconciled strands in his thought. We are not particularly concerned with this criticism (in any case, it is reIuted by Ilting's reconstruction) or with the systematic aims
oI Hegel's work. We are interested only in rebuilding Hegel's conception around what may well be a subtextual antinomy in his

Page 116
political philosophy so that we can trace a new theory oI civil society back to the institutionally most elaborated conception Irom which we can still learn. Thus, a more
serious objection to our reconstruction would insist, as did the young Marx in 1843, that the dimensions we bring into special relieI represent elements in Hegel's
thought that are not modern, in contrast to the modernity oI his conception oI the system oI needs, on the one side, and the bureaucracy, on the other. In this reading,
Hegel's "corporation" is an attempt to save medieval corporate doctrine; his estate assembly, the institutions oI the Stàndestaat; his notion oI public opinion, the earlv
bourgeois public sphere; and perhaps the very idea oI public Ireedom, the ancient city-states. Accordingly, iI we are to look Ior the modernity oI Hegel's social
theory, we would do better to Iocus on the critical aspects oI his depiction oI the capitalist economy (Lukacs) or his anticipation oI the welIare state (Avineri).
OI course, each interpreter Iavorable to Hegel tries to interpret him through a speciIic conception, and even to enlist his alliance. The theory oI civil society we are
trying to develop is no exception to this rule. Nevertheless, we believe, in the context oI both subsequent social and intellectual history, that the categories we stress
were not mere atavisms in Hegel's time and have become even less so in the postliberal (and now also the poststatist) epoch. In this context, the history oI social
theory oIIers an important, iI hardly conclusive, prooI. While the theory oI the system oI needs was IruitIully developed by the Marxian tradition, and the theory oI
bureaucracy became a cornerstone oI the works oI Weber and his Iollowers, the idea oI civil society as the central terrain oI social integration and public Ireedom was
to become just as IruitIul in a line oI theoretical development that had its beginnings in Tocqueville, its continuation in Durkheim, in English, French, and American
pluralism, and in Gramsci, and its culmination in Parsons and Habermas. In our opinion, this tradition oI interpretation has shown at the very least that the basic
categories oI Hegel's Rechtsphilosophie can be thoroughly translated into modern terms. II we are to believe the testimony oI social actors East and West, North and
South, such reconstructed terms oI analysis have not yet exhausted their critical and constructive potential.

Page 117
3-
Theoretical Development in the Twentieth Century
The untenability oI the Hegelian synthesis and the collapse oI its systematic assumptions do not represent the end oI the theory oI civil society. Subsequent theorists,
however, tended to Iocus only on speciIic dimensions oI the multilayered Hegelian concept, developing these to the exclusion oI all others. Marx stressed the negative
aspects oI civil society, its atomistic and dehumanizing Ieatures; but in so doing, he managed to deepen the analysis oI the economic dimensions oI the svstem of needs
and went Iar beyond Hegel in analyzing the social consequences oI capitalist development.
1
Tocqueville removed the ambiguities Irom the discussion oI publicitv,
discovered in voluntary associations a modern equivalent oI the anachronistic corporation, and demonstrated the compatibility oI civil society and democracy, albeit in
a context (America) that he considered to be an uncharacteristic version oI modern society. Gramsci reversed the reductionist trend oI the Marxian analysis by
concentrating on the dimension oI associations and cultural intermediations and by discovering modern equivalents oI Hegel's corporations and estates. Finally,
Parsons Iocused on the dimension oI social integration in terms oI a whole series oI institutions constitutive oI what he called "societal community." More like Hegel
in his systematic aspirations than any oI the others, Parsons attempted to synthesize the normative claims oI tradition with those oI modernity. His concessions to
ideology, again reminiscent oI Hegel, were the price he paid Ior the Iailed attempt.
In this chapter our primary interest is in two twentieth-century attempts to develop theories oI civil society on the Ioundations

Page 118
provided by Hegel. This seems to us the best strategy to test the viability oI a Iorm oI theorizing originally attached to the problems oI early modern states and
industrial society and based on a mode oI empirical generalization whose plausibility rested on surviving ideologies and institutions Irom premodern constellations such
as city-states, Stàindestaaten, and societies oI orders.
The combination oI Parsons and Gramsci is easily justiIied. Both are inIluenced by Hegel, and yet both correct him by diIIerentiating civil society Irom the economy as
well as the state. The one overcomes liberal, the other Marxian, reductionism. Both are inclined to interpret civil society in Iunctional terms, as the sphere responsible
Ior the social integration oI the whole. At the same time, both are aware, even iI ambiguously, oI the norma Iive achievements oI modern civil society. The crucial
diIIerences between them, linked to their diIIerent theoretical traditions and political assumptions, can be Iound in the way they combine normative and Iunctional
theory. Parsons identiIies the normatively desirable with the actual Iunctioning civil society oI the present, thereby Ialling into an unconvincing apology Ior contemporary
American society. Gramsci, Iocusing on the normative desirability oI a Iuture (socialist) civil society, tends to treat the civil society oI the present only in terms oI its
Iunction Ior a system oI domination he completely rejects. His combination oI an excess oI utopia with an excess oI realism does not allow him to adopt a genuinely
critical attitude to the Soviet Union, the country oI the revolution where not only bourgeois but all civil society was suppressed. In the end, then, neither is suIIiciently
critical oI his own ideological tradition, and as a result, neither is Iully able to thematize the duality oI modern civil societyits liberating promise as well as its links to
heteronomy.
Parsons:
Civil Society between Tradition and Modernity
The classical sociological tradition brought to completion by Talcott Parsons rarely used the concept oI civil society, Ior it was undoubtedly considered a remnant oI
pre-social-scientiIic discourse about human aIIairs. All the more remarkable is the reappearance oI the concept in Parsons's work. To be sure, it appears both in a
new disguise and in the context oI a new model oI diIIerentiation.

Page 119
Parsons's concept oI a societal community that is distinguished Irom the economy, the polity, and the cultural sphere represents a synthesis oI the liberal concept oI
civil society as diIIerentiated Irom the state with the stress on social integration, solidarity, and community that typiIies the sociological tradition initiated by Durkheim
and Tönnies. This synthesis, in which both individuation and integration are central, involves, remarkably enough, a partial and conscious return to the Hegelian theory
oI civil society.
2
While (unlike Hegel but similarly to Gramsci) Parsons diIIerentiates the societal community Irom the economy as well as the state, the continuities
between the two conceptions are more striking than the diIIerences.
For Parsons, as Ior Hegel, modern society is structured by normative Irameworks oI plurality (associations) and legality. Publicity and participation are also present,
but as in Hegel's work, they are deemphasized. Moreover, Parsons, like Hegel, is ready to pronounce a single version oI modern society (in his case, the United
States) as more or less the highest realization oI the potentials oI modernity. ''The completion oI the society . . . called modern" will take place when the integration
problems oI this society or type oI society are resolved. Finally, Parsons is conscious oI the debt that modern society bears to the historical project oI the age oI
democratic revolutions, even iI he considers this project to be Iully accomplished (and hence annulled as a project) by the developed Western societies: "The more
privileged societies oI the late twentieth century have to an impressive degree, which would have been impossible to predict a century ago, successIully institutionalized
the more 'liberal' and 'progressive' values oI that time."
3
As Iar as these societies are concerned, the struggle Ior democratization is, on the whole, relegated to the
nineteenth century.
4
This last thesis concerning the actual accomplishment oI the values oI the age oI revolutions opens Parsons's concept oI modern society to the charge oI "bourgeois
apologetics" leveled at all post-1848 usages oI the "utopia" oI civil society.
5
Parsons, though, is ideological only in the sense that Hegel was, namely, in the extent to
which he mixes normative insight with mystiIications concerning existing institutions. Yet, and again like Hegel, the theory points beyond ideology insoIar as it links
these normative insights

Page 120
to the potentialities oI existing society, even iI Parsons himselI does not recognize that these are actualized only partially and selectively.
Parsons's division oI the social system into Iour Iunctions or subsystems appears distinctly unhistorical next to Hegel's speciIication that it is modern development that
produces the diIIerentiation between state and civil society. But Parsons, too, insists that in earlier societies, undiIIerentiated institutional complexes carried out more
than one, and possibly all, oI the major social Iunctions. For example, in tribal societies, kinship was the key social, cultural, political, and economic institution; the
Ieudal bond in the High Middle Ages organized social, economic, and political relations; and the absolutist-mercantilist state was a political and economic entity. The
development oI modernity is thus conceived as the diIIerentiation oI what had been implicitly there in all societies, in institutions that may have had dimensions linking
them to all Iunctions but whose center oI gravity was tied up with a single Iunction. This teleological interpretation oI history may well involve an impermissible
projection oI modern Western categories to premodern and non-Western societies, so that the universal applicability oI a category such as diIIerentiation is thereIore
open to doubt.
6
The relevance oI this category to modern development itselI is, nevertheless, highly plausible.
7
To Parsons, the societal community is the integrative subsystem oI society: Its Iunction is to integrate a diIIerentiated social system by institutionalizing cultural values as
norms that are socially accepted and applied. The diIIerentiation oI the societal community Irom the cultural, economic, and political subsystems was accomplished,
according to Parsons, by the three modern revolutions: the industrial, the democratic, and the educational. Each oI these is represented as a step in "the societal
community's declaration oI independence" Irom the other subsystems, which, however, also acquire their diIIerentiated institutions in the process.
8
Actually, in
Parsons's analysis, the diIIerentiation oI the societal community was begun in the major English antecedents to the three revolutions: (1) the coming oI religious plurality
and toleration, which diIIerentiated religion and the state Irom one another while to some degree Ireeing the societal community Irom a religious

Page 121
deIinition oI Iull membership; (2) the establishment oI purely economic relations through a market economy Ireed oI social, iI not yet political, restraints; (3) the
development oI an aristocratic Iorm oI representative government that diIIerentiated government and its constituency (primarily the aristocracy and the gentry) and
stabilized their relations through parliamentary representation; and (4) the development oI a Iorm oI law that helped to carve out a societal sphere not open to arbitrary
intervention even by the state itselI. In presenting these antecedents, Parsons simpliIies by linking steps in the diIIerentiation oI each oI the Iour subsystems to a single
process, even iI the step has consequences Ior other subsystems as well. Thus, Ior example, the development oI the rule oI law, which he links to the institutionalization
oI the legal proIession and the stabilization oI a system oI independent courts, is also the most important preparation Ior a diIIerentiated societal community.
SigniIicantly, Parsons considers that the process oI diIIerentiation oI the societal community would have been incomplete without all three revolutions. In one version oI
his argument, these revolutions represent the diIIerentiation oI the integrative sub-system Irom one other subsystem in each case.
9
In another version,
10
Parsons insists
that each revolution actually strengthened the other subsystem: the economic in one case, the bureaucratic-administrative in the other. There is no inconsistency here,
however, because Parsons sees diIIerentiation as a reciprocal and non-zero-sum process that involves institution building in all the relevant spheres. But there is one
major inconsistency in his account: The diIIerentiation oI the societal community Irom the market economy is nowhere provided Ior in the doctrine oI the three
revolutions, in spite oI general claims to the contrary. As a result, the argument must surrender its parallel structure; in particular, the dramatic process in which the
societal community declares independence Irom the state, vividly portrayed by Parsons, does not have a parallel in the relationship oI the societal community to the
new type oI market economy. We might suggest that Parsons here came up against a problem he sought to deemphasize: the problem oI capitalism and a century oI
socialist responses to it, symbolizing, as Karl Polanyi noted, society's selI-deIense against the economy.

Page 122
In Parsons's conception, the democratic revolution, whose center was France, certainly did lead to a tremendous strengthening oI the state power that was Iirst built in
the epoch oI absolutism. Nevertheless, Irom the point oI view oI the societal community, the original contribution oI this revolution was the creation oI a new type oI
solidary, national collectivity whose members have equal claim to political rights in addition to the civil rights already aIIirmed in English development.
11
The
emergence oI this new type oI collectivity involves a reversal oI primacy with respect to the absolutist era: "The societal community was to be diIIerentiated Irom
government as its superior, legitimately entitled to control it."
12
Again, no inconsistency is involved in aIIirming the simultaneous strengthening oI state power and the
development oI a more autonomous society capable oI deIending itselI against this power, because Parsons rightly does not consider power to be a zero-sum game.
13
Obviously, Parsons thought oI the industrial "revolution" as entirely parallel to the democratic one. This is true, however, only iI we take the relationship oI polity-
economy to be the the central axis oI interest. Accordingly, the industrial revolution, whose center was Great Britain, completed the trend oI earlier capitalist
development by enormously extending the division oI social labor (in Durkheim's sense) and by diIIerentiating an economically deIined society Irom the state (in
Polanyi's sense), leading to the complementary growth oI both subsystems (as both Durkheim and Polanyi noted).
So Iar the parallel between the two revolutions works. But iI we choose as our axis economysocietal community, as did Polanyi (on whom Parsons otherwise
greatly relies), the parallelism stops. Instead oI diIIerentiation and complementary expansion, the industrial revolution produced an economic society (the market
economy) that threatened to subsume and reduce autonomous social norms, relationships, and institutions. While one would hardly expect Parsons to be sensitive to
the Marxian discussions oI reiIication and commodiIication, it is indeed surprising that he does not examine Polanyi's thesis that a selI-regulating market produces an
"economization" oI society, against which a program oI the selI-deIense oI society emerged in the nineteenth century.

Page 123
Indeed, this program had many Ieatures parallel to the eighteenth-century liberal conIrontation oI society and state to which Parsons's conception oI the democratic
revolution in part reIers.
Aspects oI Polanyi's stress on society's selI-deIense against the destructive trends oI classical capitalism do, oI course, return in the discussion oI Ieatures oI the
twentieth-century welIare state and unionism.
14
But, characteristically, Parsons considers the issue solved with the development oI the welIare state. Indeed, the latter
seems to "transcend" both capitalism and socialism. Uncharacteristically, however, the problem is not considered in the context oI the thesis oI diIIerentiation. One
suspects that this thesis could not have been applied in a consistent and convincing Iashion to the economysocietal community axis.
The thesis oI the educational revolution addresses the same issue once again, though this time in a rather Iuturistic perspective. Curiously, it is in this context that we
Iind some oI Parsons's most critical remarks with respect to classical capitalist development:
The capitalist alternative emphasized, Iirst, Ireedom Irom the ascriptive past, then protection Irom governmental "interIerence." The socialist alternative proposed the mobilization
oI governmental power to institute Iundamental equality, ignoring almost completely the exigencies oI economic eIIiciency . . . Both Iailed to ground themselves in adequate
conceptions oI the societal community and oI the conditions necessary to maintain its solidarity.
15
The American-centered educational revolution, abstractly located on the axis culturesocietal community, implies, according to Parsons, a more consistent Ireeing oI
the social structure Irom all ascriptive patterns oI stratiIication than could be provided Ior by private property (capitalism) or governmental oIIice (socialism), providing
equality oI opportunity (though not ensuring equality oI results). Even more importantly, he maintains that the central institutional complex oI this revolution, the
university, provides Ior the development oI an associational pattern oI social organization that is to be distinguished Irom and counterposed to the bureaucratic and
individualistic Iorms promoted by the state and the market economy, respectively. Thus, he sees the educational revolution, amazingly enough, as a solidaristic
corrective not only to

Page 124
socialism and capitalism but also to the democratic and industrial revolutions. It promises, in short, a potential completion oI modernity capable oI securing the
autonomy and integration oI the integrative subsystem, the societal community, alias civil society.
Parsons's claim that the modern university provides an alternative model oI organization to both market and bureaucracy would be bizarre iI it were not simply a
special case oI his general argument about the associational character oI contemporary American society. BeIore turning to his ideological mystiIication oI aspects oI
this society, however, we must highlight another deIiciency in his conception.
We have already noted that at least one strand oI Parsons's conception, apparently contradicted by another, makes the emergence oI the modern societal community
a residual result oI the selI-diIIerentiation oI the other subsystems in the three revolutions. Within a purely Iunctionalist scheme, such a depiction leads to no internal
contradictions, but Parsons can continue to operate within such a scheme only to the extent that his evolutionary model denies itselI the possibility oI explaining the
actual mechanisms oI social change involving action and conIlict. He can do this only as a sociologist; as a historian, he repeatedly runs into the problem oI social
movements and conIlicts. But the Iunctionalist sociologist is ready with his answer: The radical democratic movement, the socialists, and the new leIt are depicted as
the Iundamentalist wings oI the three revolutions
16
whose projects apparently involve the sort oI short-circuiting oI processes oI problem solving ascribed to "Value-
oriented movements" by Neil Smelser.
17
Parsons, however, Iorgets Smelser's other type oI movement, the "norm-oriented movement" that is capable oI positively
inIluencing social change. This omission on the level oI theory is all the more odd given that Parsons himselI depicted the civil rights movement in the United States in
terms oI this paradigm.
18
As a result oI the theoretical short circuit in Parsons's approach to social movements, there are two issues that he cannot even raise, much less resolve: the problem oI
the agencies involved in the selI-constitution oI the new type oI societal community he describes, and the problem oI the resistance oI an increasingly modern societal
community to trends threatening its diIIerentiation. We shall address these in turn.

Page 125
With regard to the Iirst issue, in Parsons's analysis, agency can apparently only short-circuit social change that is caused by objective processes. In the case oI the
other subsystems, however, the state-makers and jurists, the entrepreneurs and managers, the educators and Iiduciaries are never described as Iundamentalists oI any
kind. Thus, action in the service oI social change is possible, but only on the part oI elites and Ior subsystems other than the societal community whose diIIerentiation
does in this sense become residual.
With regard to the second issue, with the democratic movement, the working-class movement, and the student movement all described as Iundamentalist, we get the
impression that their Iorms oI action as well as their goals aimed at dediIIerentiation in each case, that is, the absorption oI the modern economy, state, and educational
system into a solidaristic societal community whose own modernity would thereby be placed in doubt. All oI these movements did indeed have some elements and
ideologies that were Iundamentalist in exactly this sense. However, Parsons Iailed to see that other dimensions oI the very same movements struggled precisely Ior
social autonomy and, thereIore, Ior the diIIerentiation oI the societal community, along with its norms and institutions. This is simply the other side oI his Iailure to take
into account the tendencies oI the modern state, the capitalist economy, and even modern science to dediIIerentiation, that is, the absorption and penetration oI the
other social spheres. A theory oI modern society that Ialls to see these trends necessarily turns ideological and apologetic.
19
Parsons's theory oI the societal community is an excellent object oI immanent criticism because he both elaborates the normative achievements oI modernity and
represents these as iI they were already institutionalized. Indeed, he Iacilitates the job oI the critic by pointing to integration problems that implicitly throw much doubt
on the claims oI successIul institutionalization. The conception oI societal community represents yet another answer to Hobbes and Austin, maintaining the existence oI
a normative order without the deus ex machina oI sovereignty.
20
The concept itselI, bringing together Tönnies's well-known pairing Gemeinschaft/Gesellschaft,
consciously aims at the same kind oI synthesis oI ancient and

Page 126
modern categories as did Hegel in his doctrine oI civil society. II anything, Parsons's model seems to put greater emphasis than Hegel on elements that modern and
traditional societies have in common. He deIines the societal community in terms oI the two dimensions oI "normativity" and "collectivity." The Iormer is a system oI
legitimate order produced by the institutionalization oI cultural values; the latter is the aspect oI society as a single, bounded, organized entity. We should note that
Parsons, like Hegel, is ready to see the whole as a "politically organized'' collectivity oI collectivities: "Perhaps the prototype oI an association is the societal community
itselI, considered as a corporate body oI the citizens holding primarily consensual relations to its normative order."
21
But in the case oI a modern society, equal
emphasis is placed on the multiplicity oI oIten conIlicting groups, strata, loyalties, and roles; the modern societal community is at best a "collectivity oI collectivities."
Such an overarching collective solidarity, which is suIIicient to produce a capacity, as well as motivations, Ior eIIective collective action,
22
is possible because oI
loyalty to norms based on consensus. Here, too, Parsons assumes a kind oI uniIication hardly characteristic oI modern societies; his notion that ultimately "Values are
mainly legitimized in religious terms" tends to commit him to the view that a legitimate social order rests on shared substantive values. But, once again, he is ready to try
to modernize the conception by (inconsistently) reIerring to a "relative consensus," one that is only "a matter oI degree,"
23
which, however, could hardly play the role
oI representing the decisive Iorum that resolves the conIlict oI loyalties among individuals, and even within each individual. A mere matter oI degree cannot provide that
"high position in any stable hierarchy oI loyalties" that Parsons seeks to ascribe to loyalty to the societal community itselI.
24
II the overall contours oI the conception are open to the charge oI insuIIiciently representing modern society, in its detail the argument is capable oI dealing with this
objection. Once again there are uncanny similarities to Hegel, this time in terms oI the very architectonic oI the presentation: Modern societal community is understood
above all as a Iramework oI laws and associations. As we have already argued with respect to Hegel, there is one notable absencethe system oI needsand a
notable presencethe citi-

Page 127
zenship complex. The latter, understood in terms oI three categories oI rights, is in Iact an outgrowth oI the system oI laws.
For Parsons, the most important step in the emergence oI a modern legal system is the transition Irom law as an instrument oI state policy to law as the "mediating
interIace" between state and societal community, Iormally constitutive oI their diIIerentiation. Such a legal system puts the state in "the dual position oI deIining and
enIorcing certain legally embodied restrictions on its own powers." This paradox could be sustained on the bases oI judicial independence, the corporate integrity oI
the legal proIession, and especially the openness oI the boundaries oI the legal system "permitting tentative approaches to consensus beIore Iull 'legalization' oI a norm
and its enIorcement'' based on appeals to "collective solidarity, moral standards, and practicality."
25
While Parsons assigns deIinite priority here to the development oI
the common law with respect to continental variants, quite evidently the development oI "constitutionalism," that is, the enIorceability oI constitutions even against state
policy, was everywhere structurally related to the diIIerentiation oI a modern societal community and the state.
26
The citizenship complex, an outgrowth oI constitutionalism and the rule oI law, represents its Iurther development in three areas. (1) Embodying universal norms,
modern rights anchor constitutions in principles higher than the traditions oI particular societies. (2) Representing a move Irom objective law to subjective right, modern
citizenship makes constitutional claims actionable on the part oI individuals and groups. As a result, (3) the citizenship complex not only Iurther diIIerentiates societal
community and state but establishes the priority oI the Iormer over the latter in the sense oI both normative principle and political action.
Parsons's deIinition oI "society" as the social system having the highest level oI selI-suIIiciency is resolutely in terms oI politically delimited territorial units, generally
"nation-states."
27
The normative structures that deIine the identity oI a society thus are never Iree oI a dimension oI particularism, even iI the cultural value orders in
which the legitimacy oI norms is rooted oIten transcend the limits oI any given society.
28
The modern citizenship complex, with its egalitarian tendency to Iree
membership Irom all ascriptive

Page 128
characteristics, is rooted in an important attempt to base the norms oI modern societies in not only transsocietal but actually universal values, oI which the Iirst version
was the doctrine oI natural rights. Constitutional rights thus become normative embodiments oI universal principles that represent limitations on the power oI the state
tied to the interests oI a particular politically organized society in the name oI something higher. The democratic revolution, in Parsons's conception, attempted to turn
such philosophical claims Ior the superiority oI the societal community, "the nation," into actual political primacy. The citizenship complex in this argument consists oI
three sets oI components, civil-political-social, that represent the project oI institutionalization oI such primacy. He considers the "structural outline" oI modern
citizenship "complete, though not yet Iully institutionalized."
29
For Parsons, citizenship in the modern sense signiIies equal conditions oI membership in the societal community rather than in the state.
30
Its civil or legal component
consists oI equal rights guaranteeing autonomous Iorms oI action with respect to the statein other words, "negative liberties." Rights involving property, speech,
religion, association, assembly, and individual security along with substantive and procedural equality beIore the law were Iirst Iormulated in the natural-law tradition
and are enshrined in the French Declaration of the Rights of Man as well as the American Bill oI Rights. In Parsons's presentation, they represent the principle oI
constitutionalism reIormulated as subjective rights oI private persons; as such, their Iunction is to stabilize the diIIerentiation oI societal community and state.
31
Political rights are positive rights oI equal participation rather than "Ireedoms" or "liberties"; they involve both indirect participation in representative government through
the Iranchise and rights to influence policy. It is signiIicant that Parsons, at least in the Iirst statement oI his position, included here again the rights oI Iree speech and
assembly.
32
The overlap means that the rights oI participation, especially when so strongly linked to negative rights, do not mean dediIIerentiation, but rather the
emergence oI new mediating structures
33
that indirectly contribute to diIIerentiation through interpenetration as well as new Iorms oI integration. It is these structures
that are supposed to establish the primacy oI the

Page 129
societal community, going beyond the constitutional state (Rechtsstaat) already established by negative rights.
Finally, the social components oI citizenship, not called "rights" by Parsons, consist oI the "resources and capacities" required Ior implementing rights, Ior "realistic"
rather than merely "Iormal" opportunities Ior their equal utilization. At issue are "adequate minimum standards oI 'living,' health care and education.'' Although Parsons
mentions here some sort oI "equality oI conditions," his real concern is to deIend a genuine, as against an "empty," version oI the "equality oI opportunity." Whether he
does this in a convincing manner is a question we must now ask.
According to Parsons, "In one sense the 'social' component oI citizenship is the most Iundamental oI the three."
34
We are not told in exactly what sense this is true oI
this temporally latest addition to the citizenship complex. In any case, Parsons elsewhere notes a lack oI parallelism between the "citizen" and the welIare "client."
35

The Iact that he does not speak oI social rights, that he does not note an overlap here with other parts oI the citizenship complex as in the case oI political and civil
rights, indicates his awareness oI a Iundamental lack oI symmetry. He does make a good case on behalI oI the need Ior a social component oI citizenship. The
theoretical problem is only that this case does not primarily belong to the problem complex oI the diIIerentiation oI societal community and state and the stabilization oI
this diIIerentiation. While one could argue that the autonomy oI the societal community depends on the resources and capacities oI its members, the threat to these
comes not only Irom the modern state but also Irom the modern capitalist economic order. And while at least in one context Parsons mentions the "social" component
oI citizenship in relation to the diIIerentiation oI economy and societal community,
36
the discussion goes nowhere because Parsons wants to deny the Iunctional
necessity or even plausibility oI both rights and Iorms oI participation with respect to the modern economic order.
37
This unwillingness deIinitively links the "social"
component to the role oI client, one that clearly does not belong to any citizenship complex. Even more, this role actually contradicts the idea oI citizenship, which
cannot be made consistent with any Iorm oI paternalism.
Parsons is, in general terms, very much concerned with the diIIerentiation oI societal community Irom both economy and

Page 130
state, but while he argues Ior a principle oI organization speciIic to the societal community, thus establishing the pattern oI diIIerentiation, the structure oI mediation he
provides stabilizes this diIIerentiation only between the societal community and the polity. We have already noted that Parsons considers the principle oI association to
be the Iorm oI organization oI the societal community, parallel to bureaucracy in the case oI the polity and the market in the case oI the economy. The depth structure
oI associations is linked to the mutual solidarity oI members, and this is what distinguishes the societal community Irom the diIIerent individualistic patterns oI the
market and the bureaucracy. Indeed, together with the third type oI individualistic pattern represented by the citizenship complex, the solidaristic dimension oI the
societal community is the secret oI the various syntheses stressed by Parsons, between modernity and tradition, individualism and collectivism, Gesellschaft and
Gemeinschaft.
In Parsons's conception, an association represents a corporate body whose members are solidary with one another, in the sense oI having a consensual relation to a
common normative structure.
38
Parsons believes that this consensus, generally established by prestige and reputation, is the source oI the "identity" oI the association,
oI its becoming a "We." The associational principle involves not only a solidaristic basis oI identity but also a diIIerent determination oI collective action: Here basic
decisions emerge Irom the organization itselI and are not merely applied by it, as in the case oI the bureaucratic principle. To Parsons, all organized Irameworks have
associational components, but only in cases where this is dominant (unlike the modern Iirm, or authoritarian governments) can we speak oI an association.
39
In his
view, the contemporary trend in organization is toward associations rather than bureaucracies, and he maintains that this trend emanating Irom the societal community
penetrates government and business Iirms as well, though in the latter case (concerning which Parsons is inconsistent) without becoming primary.
The emergence oI consensus through appeals to prestige and reputation, deliberately counterposed to the acceptance oI valid argumentation,
40
points to less than Iully
modern associations. Indeed, in several contexts, such as the role oI the associational

Page 131
principle in voting, Parsons explicitly speaks oI "traditionalism" as against rational action.
41
Nevertheless, Ior the contemporary societal community, he is interested in
working out the speciIically modern type oI association. Even in relation to voting, he maintains that associational mobility and the possibility oI belonging to a
multiplicity oI associations partially counteract the traditionalist implications oI all associations (with the possible exception oI the Iamily).
42
These characteristics are
Iunctions oI the Iirst speciIically modern principle oI associations: voluntariness, allowing relatively easy entry and exit, based in the normative principle oI the Ireedom
oI association. The second such principle is the equalitv oI members, constituting a horizontal as against a hierarchical pattern oI organization. The third is
proceduralism, in the sense both oI providing deIinite, Iormal rules Ior regulating discussion and oI voting. Since the Iramework oI discussion and deliberation is
understood as the locus oI consensus building through persuasion, it is possible to see these three principles as the application oI the great modern triad oI liberty,
equality, and solidarity to the model oI association.
Again, the modernity oI the model depends on the interpretation oI the terms "consensus," "persuasion," ''solidarity," and "inIluence." Parsons, as a student oI
Durkheim, is obviously aware oI the diIIerence between traditional and modern solidarity. Solidarity achieved through consensus is in some contexts identiIied
speciIically with the ideal type oI voluntary association.
43
But Parsons also notes the importance oI another, Gemeinschaft type oI solidarity, "a mutual relation oI
diIIuse solidarity" based on "common belongingness."
44
The two models thus seem to be (1) the achievement oI solidarity through discussion and deliberation among
individuals who Ireely choose to participate in an association, and (2) the generation oI consensus among individuals on the basis oI a preexisting, diIIuse solidarity that
is not open to questioning or thematization. UnIortunately, the key concept oI inIluence tends to submerge the Iirst model in the second, and the two are treated almost
interchangeably as the basis oI having inIluence.
The concept oI inIluence has a major structural role in Parsons's theory oI the diIIerentiation oI the societal community. Along with money, power, and value-
commitments, inIluence is one oI the

Page 132
Iour generalized symbolic media oI interchange that replace relations oI direct negotiation or "barter" in the Iour subsystems, regulating their internal relations as well as
their exchanges with one another.
45
While Parsons is less insistent than Niklas Luhmann on the historical processes oI the evolution oI media-regulated Iorms oI
action, his theory does also imply that the real importance oI the media emerges in the modern, diIIerentiated societies they help to constitute. In relation to the
modernity oI inIluence as a medium, there are, however, three unresolved tendencies in his thought. First, the analogy with money and power, and the idea that
inIluence is Iully exchangeable with these media, points to a modern principle oI integration that reduces communication to the production and reception oI codes and
action to an adaptation to interconnections established "behind the backs oI actors." This conception cannot ground the diIIerence between the organizing principle oI
the societal community and those oI the economy and the polity, and it treats integration through solidarity as a Iorm oI control.
46
Second, the argument that inIluence
''must operate through persuasion . . . in that its object must be convinced that to decide as the inIluencer suggests is to act in the interest oI a collective system with
which both are solidary"
47
points to a model that is speciIically modern, yet signiIicantly diIIerent in principle Irom money and power. The diIIerence is clearly indicated
by the idea that, whereas money and power work through altering the situations oI actors, inIluence (along with value-commitment) works by one person's having an
eIIect on another's intentions.
48
Finally, while Parsons is unable to make up his mind about how inIluence as a "generalized medium oI persuasion" actually works,
49

his stress is clearly on the reputation and prestige oI inIluential individuals and not on the "intrinsic" validity oI their argumentation. Here the model easily slips into one oI
traditional integration oI action unless, more consistently than Parsons, one were to speciIy that the ultimate Ioundations oI an individual's reputation, with respect to the
given issues, must be capable oI deIense as well as challenge in terms oI argumentation. While this idea is present in Parsons,
50
it is incompatible with another, namely,
that the ability oI one person to inIluence another is rooted in a background oI diIIuse, Gemeinschaft-type solidarity.

Page 133
Parsons, oI course, Iully assumes that he does succeed in grounding the diIIerentiation oI the modern societal community Irom the state and the economy in terms oI his
categories oI association and inIluence. He thus Iaces Hegel's problem oI thematizing the relevant mediations. With respect to the societal communitystate axis, these
turn out to be the classical ones the pluralist tradition inherited Irom Hegel and Tocqueville: the public, the lobbies, the political parties, and the legislature, which are
the channels Ior societal inIluence on the administration oI the state.
51
Their eIIective operation, according to Parsons, presupposes the system oI mass
communications that he pronounces the "Iunctional equivalent oI some Ieatures oI Gemeinschaft society," again running into all the ambiguities that characterize his
theory oI inIluence. Throughout this strand oI his argument, he presupposes that social constituencies communicate with instances in the political system in ways that
are entirely undistorted by money and power, and that there is a symmetrical relationship oI exchange between "public support" and "public inIluence."
Since Parsons's exploration oI the problem oI the diIIerentiation oI societal community and economy is unsatisIactory, it is not surprising that he does not realize that his
theory, unlike Hegel's, needs a series oI mediations in this context as well.
52
Such mediations do make a limited appearance in various essays. We learn, Ior example,
that the associational trend also penetrates the economy in the Iorm oI proIessional associations and Iiduciary boards. In the case oI the modern Iirm, however, we
also Iind out that the members oI the association (the stockholders) have a passive role, while the board is assimilated to bureaucratic management.
53
As Iar as
workers are concerned, Parsons rejects any model oI democratic participation in management,
54
and he restricts the role oI unions, in the gap between household and
workplace, to that oI improving the economic position oI the working class.
55
Parsons's discussion oI the relation oI the societal community to the economy raises existing capitalist practice to the level oI norm, or at least oI Iunctional necessity.
His theory oI the societal community as a whole, however, consciously (though unsuccessIully) aims at a model that goes beyond the alternatives that might be
described as capitalist economism and socialist statism. The astonish-

Page 134
ing part oI this theory is the claim that such a postcapitalist, postsocialist model is not only the counterIactual normative construction oI a social-political project but is
already actualized, even iI not yet completely, in contemporary American society. Once again, the rational is the real, the real is the rational:
The United States' new type oI societal community, more than any other single Iactor, justiIies our assigning it the lead in the latest phase oI modernizationu We have suggested
that it synthesizes to a high degree the equality oI opportunity stressed in socialismu It presupposes a market system, a strong legal order relatively independent oI government,
and a "nation state" emancipated Irom speciIic religious and ethnic control. . . . Above all, American society has gone Iurther than any comparable large-scale society in its
dissociation Irom the older ascriptive inequalities and the institutionalization oI a basically egalitarian pattern. . . . American society . . . has institutionalized a Iar broader range oI
Ireedoms than had any previous society.
56
The United States, in Parsons's view, is not only the proper home oI the educational revolution with its emphasis on the "associational pattern," but also the most
successIul synthesis oI the results oI the democratic and industrial revolutions. American models oI representative government and Iederalism yield the highest level oI
diIIerentiation between state and societal community. This is the case because this society is the Ireest Irom ascriptive and political deIinitions oI membership and
(much less plausibly) this political system is least encumbered by social restrictions on participation at any level. Representative government makes all societal members
its proper constituency, but the separation oI powers provides the political system proper with broad Ireedom oI action. The structures oI representation, national and
Iederal, adequately mediate, according to Parsons, between the state and the societal community.
Parsons is less able (but, given the inconsistency oI his normative conception, less compelled) to claim a similar degree oI diIIerentiation between the societal
community and the economy. He does seem to admit that since the "social component oI citizenship" in America lags behind that oI European welIare states,
57
market
economic rationality has a greater power over social liIe. Nevertheless, he maintains that American society is also beyond the obsolete,

Page 135
Iailed alternatives oI capitalism and socialism, which he deIines primarily in terms oI an absence oI governmental controls on the economy versus total governmental
control.
58
To be Iair, Parsons's analysis does contain the suggestive notion that neither capitalism nor socialism is grounded "in adequate conceptions oI the societal
community and oI the conditions necessary to maintain its solidarity." However, his depiction oI America as a postcapitalist, postsocialist society is Iocused primarily
on the emergence oI the mixed economy, and he is apparently unaware oI the possibility that modern interventionist welIare states are also capable oI threatening and
displacing social solidarity. It may be that Parsons assumes, in this context at least, that overcoming the dysIunctional eIIects oI capitalism through state regulation and
redistribution, within the limits oI the market economy, establishes social control over the economy. And perhaps he sees such a control operating through the
secondary mediation oI representative government, which provides a more direct Iorm oI control over the state. However, the asymmetry between the two Iorms oI
supposed control is obvious. Any identiIication oI social control with state regulation implicitly violates Parsons's own stress on the diIIerentiation oI these spheres. And
even the idea that representative government is the medium oI social control would bypass, in an illegitimate way, Parsons's depiction oI the internal diIIerentiation oI
the political system and his stress on elites as providing the actual mechanism oI rule.
To be Iair, Parsons also aIIirms the existence oI structural positions Irom whose point oI view an analogous control over state and economy could be conceived.
American society is understood as the most hospitable terrain possible Ior the principle oI associationism, which Parsons presents as the alternative to both capitalism
and statism, symbolizing respectively a modern economy and a modern state Iree oI any social controls. Continuing the line oI analysis began by Tocqueville, Parsons
roots the importance oI a pluralistic version oI associations deeply in American history. The organization oI American Protestantism has Iavored both pluralism and
associationalism, the latter by the internal structure oI organization oI many oI the churches, the Iormer by the multiplicity oI denominations and the relatively long
history oI toleration.

Page 136
But secular patterns contributed greatly to these trends, in particular an exceptionally long history oI voluntary associations and a later, but even more important,
pattern oI inclusion in American society oI a whole series oI ethnic groups that were nevertheless able to preserve their individual identities. The struggle oI American
blacks Ior civil rights, concerning which Parsons wrote one oI his best essays, represented Ior him a great culmination precisely oI the preexisting normative and
organizational patterns oI American history.
59
In this one context, Parsons could see that movements in contemporary modern societies did not necessarily imply Iundamentalism but could actualize universalist,
normative potentials (here the premises oI the democratic revolution) in a manner capable oI creating and preserving particular identities. UnIortunately, though, he
seemed to expect that associationalism would be generalized not by new movements Iollowing in this pattern but only through the social implications oI the so-called
educational revolution and its supposedly collegial pattern oI organization. Parsons did not, however, explain how the associational Iorms oI the university are to
transIorm bureaucratic structures in the rest oI society or how these Iorms can be protected against penetration by economic wealth and political power. One reason
why this issue does not come up, despite Parsons's obvious Iamiliarity with contemporary universities, is that he identiIies it with the claims oI supposed
Iundamentalism. He insisted, Ior example, on seeing only the Iundamentalist, communitarian side oI the New LeIt and the student movement, and not the side that
demanded university democracy (and associational rights) as well as autonomy and diIIerentiation with respect to economic and political institutions. By dogmatically
rejecting these movements, he closed himselI to an important discourse that is in many respects continuous with his own.
60
This issue is important because Parsons Iully recognizes that "associationism" today cannot be deIended on the nineteenth-century ground oI small-town America,
which even Tocqueville considered to be somewhat atavistic.
61
But his various attempts to provide adequate modern alternatives all Iail because he never takes into
account the negative potentials oI contemporary institu-

Page 137
tions. While he is right to notice, beyond the elite theory oI democracy, the element oI social control inherent in representative institutions, he is wrong to bypass their
oligarchic trends and to stylize existing political elites as the "Iunctional equivalent oI aristocracy" that "democracies urgently need."
62
He is right in insisting on the
important normative implications oI the pluralistic traditions oI American society, but his dismissal oI the speciIic selectivity and asymmetry built into the existing
practice oI pluralism is both unsophisticated and misguided.
63
Finally, he is right in not taking the mass-society thesis too seriously, as well as in insisting on the
continued importance oI "kinship and Iriendship" along with "associational activities and relationships,"
64
but he is wrong to think that this removes the ground Irom
another distinction, that between "public'' and "mass culture." Indeed, his views concerning mass culture and mass media could have been based on Ileshing out the
insight concerning the existence oI two trends identiIiable with this distinction, one toward manipulation and the other toward democratic communication.
65
Instead,
aIter noting the possibilities oI overconcentration, manipulation, decline oI cultural standards, and political apathy as possible consequences oI the modern mass media,
he dismisses, or at least vastly deemphasizes, the relevance oI these trends to American society! And aIter presenting the system oI mass communications as a kind oI
market,
66
he inconsistently declares that this system represents "a Iunctional equivalent oI some Ieatures oI Gemeinschaft society."
67
Given his diIIiculties in basing his theory oI association on speciIically modern trends, it is not surprising that Parsons seeks a Iunctional equivalent oI Gemeinschaft. In
this context, however, his choice oI the mass media can amount only to a tacit admission oI deIeat. In Parsons's theory, this implicit deIeat appears in his thematization
oI integration problems in contemporary American society, the solution oI which would complete modernity itselI. Indeed, we should note that he does not admit that
the diIIerentiation oI the societal community and its associational Iorm oI organization are in any respect incomplete. Nor does he consider the cultural values oI
modern societies to be in any sense deIicient or contradictory. Rather, his thesis is that successIul diIIerentiation and reorganization have produced integration gaps or
lags that

Page 138
have not yet been successIully addressed because norms capable oI generating suIIiciently high levels oI motivation, legitimacy, and solidarity have not been adequately
institutionalized. As a result, the societal community is "the storm center" oI Iuture conIlicts that cannot be dealt with through the control oI money and power. On the
other hand, the demand oI new movements Ior participation and community, considered exclusively in their Iundamentalist versions as signs oI integration strains, can
provide solutions only at the cost oI massive dediIIerentiation and regression. Between these two extremes, the direction in which Parsons himselI would look Ior a
solution remains quite unclear.
Abstractly, his theory commits him to the view that only the generation oI new Iorms oI inIluence could lead to a normative consensus that could provide symbolic
resources capable oI integrating the societal community (solidarity) as well as regulating its interchanges with the state (legitimacy) and the economy (motivation).
UnIortunately, because his theory oI inIluence is indeterminate, it is hard to see possible solutions to problems oI social integration that could derived Irom it. The
assimilation oI inIluence to money and power leads, Ior example, to the technocratic solution oI planning and manipulating its sources and conditions oI application,
presumably through the mass media. Alternatively, an interpretation oI inIluence as rooted in prestige and reputation linked to traditional solidarity leads to a
neoconservative option that would hope to restore an authoritarian and possibly religious Ioundation Ior norms that would be closed oII to questioning and criticism.
Finally, an understanding oI inIluence in terms oI rational argumentation as the "intrinsic means oI persuasion" leads to a democratic alternative that would have little
plausibility unless democratization were continued as an open-ended process carried on, in part, by social movements, a possibility that Parsons explicitly rejects.
Indeed, he seems unaware that all these diIIerent options are compatible with one or another oI the contradictory substantive value-complexes inherited by modern
societies, or that their diIIerent Iorms oI institutionalization would presuppose unavoidable organizational changes. Above all, he does not see that they imply the
projects oI three alternative versions oI the modern societal community or civil society, among which social actors may

Page 139
in Iact choose. One suspects that Parsons never makes up his mind among these alternatives, that he aIIirms all oI them, or rather a combination oI them in which the
respective weights are unclear. Thus, he is open to the objection that the democratic elements in his theory imply only a legitimating window dressing Ior a traditional
model oI civil society that has become impossible, or Ior a technocratic model that is the culmination oI the genealogy oI unIreedom.
Actually, though, the situation may have been the reverse. Perhaps the traditional and apologetic elements in Parsons's thought interIere with his genuine insights into
the critical place oI civil society in modernity. This reading is suggested by his last two published essays.
68
Here Parsons demonstrated that his reconstruction oI the
concept oI civil society did not represent a dead end and was capable oI Iurther development. The context, though, was not system construction butimmanent critique,
namely oI R. M. Unger's important book Law in Modern Societv. Unger oIIers a critique oI both Iormalist, market-oriented and substantialist, state-interventionist
structures oI law Irom the point oI view oI the endangered values oI solidarity and mutual recognition. In the Iace oI older models oI liberal and contemporary welIare-
state capitalism, he seeks to justiIy a third, communitarian Iorm oI organization combining substantive justice with a morality based on Iace-to-Iace relations. Yet Unger
cannot save his model Irom the charge oI primitivism. He concedes that while the welIare state has in a sense returned to earlier, bureaucratic Iorms oI law, his own
alternative also completes a historical cycle by returning to customary law. Calling this movement a spiral rather than a circle does not alleviate the diIIiculty.
Despite his own ambiguities with respect to a traditional organization Ior the societal community, Parsons will have nothing to do with communitarianism, which he
identiIies as the absolutization oI the dimension oI social integration (in a highly misleading way, he speaks oI "the absolutism oI law").
69
But he is willing to take up
Unger's challenge to push the critique oI Iormal law (and thus liberal capitalism) and substantive or purposive law (and thus the welIare state) to the point where the
outlines oI a third option become visible. We should notice, even iI he did not, that the two

Page 140
criticized options are not, as in his earlier work, liberal capitalism and socialism, with the welIare state representing their Iinal synthesis. Without noticing, he took Irom
Unger's critical theory the premise that critique must aim beyond all contemporary Iormations.
70
The crucial point, Irom the point oI view oI his own conception oI
civil society as the societal community resting on norms and associations and counterposed to both the economy and the polity,
71
is that he is here able to Iormulate a
two-sided critique oI market and state in terms that avoid all regression to historically obsolete structures oI law and society.
He Iinds the Archimedean point in Unger himselI, who distinguishes between substantive and procedural patterns oI the deIormalization (rematerialization) oI law.
Substantive law involves interventions whose purpose is to bring about speciIic social results beneIiting speciIic interests; procedural law ("the great intermediate and
mediating category"), however, aims only at the equalization oI partners whose negotiation under careIully determined procedures is to reach agreement concerning
means and ends. Unger's preIerence, like that oI many deIenders oI the welIare state (such as T. H. Marshall), is Ior substantive law; he considers procedural law to
be still within the tradition oI Iormal law because oI its preservation oI the principle oI legal generality on the "meta" level oI procedure. To Parsons, oI course, this
element oI continuity that preserves the status oI law as a limitation rather than an instrument oI sovereign power is attractive: The diIIerentiation oI the societal
community Irom the polity depends on it. Procedural law, moreover, preserves the possibility inherent in contract law, recognized neither by legal positivism nor Ior
that matter by Unger, that law can be created by social entities other than the state.
Equally important, Parsons discovered the link oI procedural law to his own concept oI associationalism, contrasted with both the bureaucracy and the market. He
goes too Iar, though, and identiIies all institutions governed by procedures as the domain oI procedural law, Irom courts and parliaments to elections and voluntary
associations. In this manner, even the very corporatism that appears to Unger as a danger to the public and positive Ieatures oI the law is recast by Parsons as an
instance oI indepen-

Page 141
dent law-making by society. From a supposed indication oI the decomposition oI autonomous law, he thus obtains the prooI oI essential continuity. It is a pity that this
initially promising analysis has such a Ilat outcome.
What goes wrong? First, Parsons vitiates his own important point concerning the link oI procedural law and associations by conIusing procedure and procedural law.
While all manner oI institutions can be regulated by procedures, including undemocratic and hierarchic ones, procedural law in Unger's provocative deIinition is in
eIIect reIlexive and deals with procedures (oI equalization) that have other procedures as their objects. Thus, to give an example accessible to neither Unger nor
Parsons, while the associations participating in corporatist bargaining may and generally are regulated by procedures, procedural law would target these procedures to
produce internal democracy and the protection oI individuals and minorities. Again, while the secret bargains oI a limited number oI associations might be arrived at
under Iixed procedures, procedural law might seek to make this process public and open to other interested parties. Thus, procedural law does not merely reIlect the
existence oI associations, as Parsons suggests, but aims at the democratization oI their internal liIe as well as their interrelations.
There are two reasons Ior Parsons's analytical error. First, he identiIies procedural law as "a cooperative . . . Iramework within which 'parties,' whether they be
individuals or groups, can be 'brought together' to adjust their interests with each other under a normative order."
72
This deIinition captures only halI oI what is meant
by procedural law, because it brings procedures under the rule not oI procedures but oI a higher normative order that is not Iurther deIined. II that order were legal
norms, then the deIinition would beg the question with respect to the type oI law (Iormal, procedural, or substantive) that these entail. But we have good reason to
think that what Parsons has in mind is not law at all, but the higher normative (religious-moral) order oI society. With the meta level thus occupied, Parsons apparently
sees no important reason to distinguish between procedures themselves and the procedures that produce or regulate procedures. In other words, he cannot discover
the meaning oI procedural law as a speciIically

Page 142
modern, reIlexive, and intersubjective regulation oI the production oI norms, because Ior him agreements and perhaps laws can be produced only as the
institutionalization oI what is already there on a higher normative level.
Second, while he does not notice that through an immanent critique oI Unger he has implicitly been brought to a position critical oI all existing societies, he deIinitely
seeks to escape this implication on a more concrete level. As always, he is in a rush to pronounce existing American society to be the resolution oI all antinomies, this
time oI liberal capitalism and the welIare state, at least Irom the legal point oI view. II procedural law is the solution oI the riddle, as he insightIully notes within Unger's
text, then the bulk oI American law must be procedural law. This apologetic claim can be sustained, however, only through the misidentiIication oI procedural law and
procedure. Once again, his discovery oI the potentially critical terrain oI civil society, this time on the level oI legal theory, is vitiated by his apologetic treatment oI
American society as representing some kind oI "end oI history." In this respect, Parsons remained a through-and-through Hegelian to the end oI his liIe.
Gramsci and the Idea of Socialist Civil Society
II Parsons can be said to represent a twentieth-century rehabilitation oI the Hegelian idea oI Sittlichkeit in social-theoretical terms, with inevitably apologetic
consequences Ior contemporary civil societies, Gramsci can be said to reIlect a modern renewal oI the leIt radical critique oI civil society. This characterization should
not be taken to imply, however, that he simply Iollows the classical Marxian analysis and criticism oI civil society. Although a Iollower oI Marx, Gramsci generated his
own conception oI civil society directly Irom Hegel.
73
And unlike Marx, he turned not to the system oI needs but to the doctrine oI corporations Ior his inspiration.
Undoubtedly aware oI the Marxian use oI the term bùrgerliche Gesellschaft, Gramsci's interpretation oI Hegel was thus at the same time an implicit critique oI that oI
Marx and Engels. Although he did not know Marx's text that denounced the concept oI the corporation as so much medievalism, Gramsci was keenly aware oI such
an interpre-

Page 143
tation. Nevertheless, reading Hegel's conception primarily on an abstract analytical level, he was convinced that the contents drawn Irom the world oI the old regime
could, were, and had to be given modern replacements. Accordingly, Gramsci recognized the new Iorms oI plurality and association speciIic to modern civil society in
modern churches, unions, cultural institutions, clubs, neighborhood associations, and especially political parties.
The most decisive departure oI Gramsci Irom both Hegel and Marx is his highly original option Ior a three-part conceptual Iramework. Against Hegel' s version, and
more convincingly, Gramsci located both Iamily and political culture on the level oI civil society. Unlike Hegel and Marx, however, he did not include the capitalist
economy on this level. We can only speculate about the reasons Ior the second oI these moves.
74
Gramsci was essentially a political thinker who was interested in
theory Ior the sake oI political orientation. In this he conIronted two great and, Ior him, decisive problems: the Iailure oI revolution in the West and its (supposed)
success in Russia. In neither context did the economistic reduction oI civil society to the political economy, so prevalent in Marxism, allow the problem oI transition to
a genuinely democratic society to be seriously posed. In the West, the reduction led to the disappearance oI the deIensive "trenches" oI the existing system: Iorms oI
culture and association that protect bourgeois society even when the economy is in crisis and the power oI the state has crumbled.
75
Only the "methodological"
76

diIIerentiation oI civil society Irom both the economy and the state allowed a serious thematization oI the generation oI consent through cultural and social hegemony as
an independent and, at times, decisive variable in the reproduction oI the existing system.
In the Soviet Union, where "the state was everything" and civil society was "primordial" and "gelatinous," the collapse oI the state did make revolution possible. But
given the Iact that that the new revolutionary power constituted itselI in a statist ("statolatry") and even ''Caesarist" or "Bonapartist" and "totalitarian" Iorm, the project
oI creating a Iree society that could absorb state power was put in doubt. The very constellation that made revolution possible was apparently the greatest roadblock
Ior developing a Iree society. Thus, in this context as well, Gramsci came to Iocus on the problem

Page 144
oI civil society as independent oI economic development and state power.
There were, oI course, other reasons Ior Gramsci's stress on civil society. One certainly has to do with the peculiarities oI the Italian situation. An acute analyst oI
Italian history and social structure, Gramsci was aware oI the Iailure oI liberalism to attain "hegemony" aIter the Risorgimento. In this assessment he was directly
inIluenced by the great Italian philosopher and historian, Benedetto Croce. Like Croce, he attributed this Iailure, in part, to the power oI the church in Italian cultural
and social liIe. Although the church no longer had direct political power in the Italian state, its power within civil society remained impressive. Indeed, through its
organization oI everyday social liIe in "civil" institutions such as church Iunctions, education, neighborhood Iestivals, and its own press, the Catholic church was able to
occupy many oI the trenches oI civil society and to constitute a powerIul barrier to the Iormation oI liberal, secular bourgeois, hegemony on this terrain. Accordingly,
Italian civil society was prevented Irom becoming Iully modern. At the same time, like many other intellectuals oI his day, and more speciIically under the inIluence oI
Georges Sorel, Gramsci believed that Italy and the West as a whole suIIered Irom a general crisis oI culture. He related the contemporary ''wave oI materialism" to the
crisis oI authority that resulted Irom the ruling class's inability to organize consensus (hegemony) and the corresponding detachment oI the masses Irom their traditional
ideologies. (The ruling class was thus only dominant, not hegemonic.) "The crisis consists precisely in the Iact that the old is dying and the new cannot be born."
77
In
other words, the moment Ior the triumph oI liberal ideology had been missed, while the old action-orienting world views had become anachronistic and were being
increasingly undermined by social and structural developments. Thus, civil society, and especially its cultural institutions, appeared as the central terrain to be occupied
in the struggle Ior emancipation.
Gramsci's conception is presented in a notoriously conIusing terminology.
78
Civil society is variously deIined as the counterpart oI the state (which is said to be either
identical with political society or its main organizational Iorm), as a part oI the state along with and counterposed to political society, and as identical with the

Page 145
state. The idea that runs through all these attempts at a deIinition is that the reproduction oI the existing system outside the economic "base" occurs through a
combination oI two practiceshegemony and domination, consent and coercionthat in turn operate through two institutional Irameworks: the social and political
associations and cultural institutions oI civil society, and the legal, bureaucratic, police, and military apparatus oI the state or political society (depending on the
terminology).
79
It may be helpIul here to recall Norberto Bobbio's insistence that Gramsci battled two Iorms oI reductionism, one reducing the superstructure to the
base and the other, cultural processes to coercion. Within the Iramework oI classical Marxian historical materialism, Gramsci sought simply to assert the independence
and even primacy oI the superstructure. We would go Iurther than Bobbio, arguing that, against Gramsci's intentions, this shiIt rendered the whole doctrine oI base and
superstructure irrelevant.
80
And yet, this irrelevant dualism, now in the Iorm oI an idealist reversal, could have biased Gramsci at times to treat the two dimensions
within the supposed superstructure, civil society and state, either as somehow one or at least as expressing the same principle and logic. One oI his terminologies, the
one integrating both civil and political society in the state, seems to express this option. Nevertheless, when he was Iorced to conIront the consequences oI reducing
social integration to political coercion, he postulated that the opposition between civil and political society (here meaning the state) was indeed one oI two diIIerent
principles, hegemony and domination.
81
One might say, thereIore, that Gramsci developed his doctrine oI civil society in terms oI two "declarations oI independence,"
one Irom the economy and the other Irom the state, and that the resulting trichotomous conception, however inconsistently, burst the bounds oI historical materialism.
As a theorist, Gramsci surely traveled a road Irom Marx to Hegel, even though his political project remained a Marxist one.
82
OI course, the Hegel oI the Philosophv
of Right also proved inadequate Ior his purposes. Not only did he want to use a trichotomous conception diIIerent Irom Hegel's, one that could not lead back to either
economism or statism,
83
but he considered the corporate doctrine, which he located as the heart oI the Hegelian theory oI

Page 146
civil society, to be hopelessly obsolete in the Iorm in which it was originally developed. Gramsci notes that Hegel's "conception oI association could not help still being
vague and primitive, halIway between the political and the economic; it was in accordance with the historical experience oI the time, which was very limited and
oIIered only one perIected Iorm oI organizationthe 'corporative' (a politics graIted directly on the economy."
84
Thus, like Marx, Gramsci is Iully aware oI the modern state's destruction oI the older Iorms oI corporate liIe that constituted a "dual power" in the late medieval world
(in the Stàndestaat, that is). He is even aware, like Tocqueville, oI the existence oI an intermediary Iormthe absolutist state and the depoliticized society oI
orders
85
Irom which the contents oI Hegel's model are drawn. More importantly, however, unlike Marx or even Tocqueville, Gramsci thoroughly understood that
the eIIorts oI Jacobin and bureaucratic statemakers to the contrary, the older corporate Iorms were capable oI modern replacements. He stresses in particular the rise
oI modern unionism and cultural associations.
86
And while modern churches, surrendering their earlier role in the state, also became institutions oI the new type oI civil
society, modern political parties gradually replaced them as the main organizational Iorm Ior intellectuals.
87
Although he is clearly aware that modern statemakers seek to abolish all intermediary associations, Gramsci does not stress the obvious point that their reappearance
in modern Iorm had to be at least partly the result oI what used to be called the struggle oI society against the state. Instead, he tends to argue, in a more or less
Iunctionalist manner, that the demand oI the state Ior consent, and its tendency to organize and educate such consent, is the major reason Ior the emergence and
stabilization oI new types oI associations.
88
OI course, Gramsci viewed the particular content and Iorm oI civil society as the outcome and object oI a class struggle.
From this point oI view, the outcome depends on which social group has been or is becoming hegemonic. Where the bourgeoisie is hegemonic, civil society is
bourgeois society, and its constitutional guarantees (rights) and political expression (parliamentary representation) are window-dressing Ior bourgeois rule.
It is worth noting that the associational Iorms that replace Hegel's corporations can, Ior Gramsci, turn into key vehicles Ior

Page 147
social movements, even iI Gramsci does not emphasize the state/society opposition in this context. Indeed, he not only discovered the modern replacements Ior the
corporation but also added the dimension oI social movements to the concept oI civil society, giving it a dynamism in addition to and independent oI the system oI
needs. What is given with one hand is, however, taken away with the other, Ior the dynamism oI civil society as the terrain oI social movements lasts only so long as
the working class is in opposition. Once civil society becomes socialist, the raison d'être Ior social movements, that is, Ior class struggle, will have disappeared. As we
shall show, one tendency in his thought, namely the Iunctionalist reduction oI the political culture (representative democracy and rights) and oI associational Iorms oI
modern civil society (clubs, interest groups, bourgeois political parties) to the reproduction oI bourgeois hegemony and/or to the creation oI socialist hegemony
(unions, communist parties), locks Gramsci into an overly schematized conception that is at once too realist and too utopian.
We have already noted Gramsci's conviction that the associations and cultural institutions oI civil society in the developed capitalist countries, as the inner "trenches" oI
the established system, have added immeasurably to the stability oI this Iorm oI domination. At the same time, he notes their abolition under contemporary
dictatorships. It is Ior just this aspect oI their rule that he dubs them "totalitarian."
89
Thus, Gramsci seems to register Iive phases oI relation between the state and civil
society: (1) medieval corporatism and dualism (the Stàndestaat); (2) the absolutist dualism oI state and depoliticized, privileged orders; (3) the early modern
dissolution oI the older corporate Iorms that, strictly speaking, exists only in revolutionary terror; (4) the dualism oI the modern state and new Iorms oI associations; an
Iinally (5) the totalitarian Gleichschaltung oI modern associations and cultural Iorms. What is most signiIicant in this typological reconstruction oI the history oI civil
society is that "totalitarianism," as against earlier statist Iorms, is depicted as the dissolution and atomization oI modern Iorms oI social and cultural integration! But why
and how are eIIective Iorms oI social integration, oI the organization oI consent, dissolved? And iI it is dissolved under totalitarianism, does civil society have a second
chance oI being reconstructed?

Page 148
These questions are diIIicult to resolve because oI three systematic ambiguities or "antinomies" in Gramsci's analysis. The Iirst comes Irom his application oI the term
"totalitarianism" to both "progressive" and "regressive" versions; the second comes Irom his discussion oI the normative status oI civil society, which sometimes implies
the consolidation oI a system oI domination through the organization oI consent and at other time the weakening and even eventual abolition oI domination; and the
third comes Irom his conception oI a Iree society, which alternates between a pluralistic civil society and a uniIied state-society.
90
All three antinomies are linked to
the attempt to work out critical theories oI two very diIIerent societies: Soviet Russia (oI which Gramsci remained supportive) and contemporary capitalist societies
and their totalitarian variant (to which he was unalterably opposed).
Without disguising anything about the Iorm oI social organization and repressive political practices in the Soviet Union, Gramsci nevertheless tries to distinguish
between "regressive" and "progressive" versions oI totalitarianism, both oI which involve abolishing the independence oI the institutions oI civil society.
A totalitarian policy is aimed precisely (1) at ensuring that the members oI a particular party Iind in that party all the satisIactions that they Iormerly Iound in a multiplicity oI
organizations, i.e., at breaking all the threads that bind these members to extraneous cultural organisms; (2) at destroying all other organizations or at incorporating them into a
system oI which the party is the sole regulator. This occurs (1) when the given party is the bearer oI a new culturethen one has a progressive phase; (2) when the given party
wishes to prevent another Iorce, bearer oI a new culture, Irom becoming itselI "totalitarian"then one has an objectively regressive and reactionary phase.
91
The policies oI the two totalitarianisms with regard to civil society are depicted as being exactly the same; both suppress cultural meaning, social solidarity, and Iorms
oI organization outside a uniIied party-state, thus ending social divisions. But their intentions are supposedly completely diIIerent. In this context, the deIense oI the
Soviet Union by an antiIascist must seem bizarre. Leaving aside Gramsci's political commitments, however, the whole argument does in Iact Iollow consistently Irom
his Iunctionalist depiction (still tied to classical Marxism) oI the institutions oI civil

Page 149
society in the advanced capitalist countries as Iorms oI organization oI consent whose role is exclusively the stabilization oI dominationits social integration, as it
were. Given this interpretation, smashing these institutions by subordinating them to a monolithic party-state can be represented as at least part oI the negative work oI
social emancipation. (We shall return to the question oI what was supposed to be the positive part oI this work.) Totalitarianism is regressive or reactionary in this
reading only when its purpose is to block "progressive" totalitarianism, rather than create a new culture, in a context where the inner trenches oI civil society are
suIIiciently weakened to raise the prospects oI their progressively motivated abolition. All in all, Gramsci seems to indicate only three possible political positions: a
conservative deIense oI the existing version oI civil society whose Iunction is the social integration oI capitalist domination; a totalitarian-revolutionary abolition oI this
civil society Ior the sake oI building a new culture; and a totalitarian-revolutionary abolition whose purpose is to conserve the existing structure oI domination.
It is also possible to detect in Gramsci the Ioundations (or at least its traces) Ior yet a diIIerent version oI "progressive" politics, one that is radically reIormist rather
than totalitarian-revolutionary. Bobbio develops such an interpretation on the basis oI Gramsci's stress on building a new cultural hegemony by the socialist party in
civil society.
92
The obvious contrast is between the cultural work oI building a new consensus that would erode the old Iorms oI consent and a program oI
revolutionary overthrow using violent means.
It is diIIicult to pinpoint such a strategy in Gramsci because oI his second "antinomy": a Marxian-Iunctionalist conception oI civil society as the locus Ior producing the
hegemony that will stabilize bourgeois domination, and a conIlicting theoretical conception oI a terrain where two alternative strategies Ior hegemony building contest
one another.
93
In the context oI the Iirst position, a strategy Ior building counterhegemony would simply integrate the working class into the established institutional
network oI civil society, which would have to be totally abolished in order to break with the existing system oI domination. In the context oI the second, however,
which postulates the possibility oI building a cultural

Page 150
hegemony incompatible with the existing system, the institutions oI civil society would themselves have a double structure, linked to both domination and emancipation.
A radical-reIormist strategy would have to build on this dual structure.
In the terms oI the Iunctionalist version oI Gramsci's theory, a strategy oI hegemony building could be, and most oI the time probably was, entirely instrumental, given
the diIIiculties the trenches oI bourgeois civil society place on the route to direct revolutionary transIormation. The aim, in this interpretation, is to erode the existing
Iorms oI social integration, to create alternative associations, and to prepare the subject oI revolutionary politics. Given the negative assessment oI the existing civil
society in this interpretation, however, the associations and Iorms oI a counterhegemony would have to be regarded instrumentally: The independent parties and unions
oI the working class would have the Iunction oI producing dysIunction within the existing Iorm oI social integration, helping to produce a crisis in which the opposing
side would have to rely on domination alone. In this interpretation, thereIore, a revolutionary rupture, with Iorce Iacing Iorce, must complete the internal work oI
transIormation.
94
More important Ior our argument, in this context at least, there would be no reason why the independent organizations involved in building a
counterhegemony should play any role aIter the revolution. Gramsci supports this view, especially when he assigns the task oI constructing a new society and
civilization primarily to the state and when he states that it is essential that the old mechanisms oI producing bourgeois hegemony be eliminated. Within the Iunctionalist
interpretation, this would oI course mean the end oI a pluralistic system oI parties, unions, and churches.
The alternative, conIlict-theoretical view oI hegemony building in civil society implies (even iI Gramsci never explicitly drew such a conclusion) a positive normative
attitude to the existing version oI civil society or, rather, to some oI its institutional dimensions. Clearly, a principled version oI radical reformism could be based on
such an attitude. Gramsci's unwillingness or inability to develop such a conception is apparent in the presence oI a more developed Iunctionalist-revolutionary option in
his thought. Indeed, one might say that the more explicit development oI the radical reIorm-

Page 151
ist option would have presupposed a political choice Gramsci never made: a thoroughgoing critique oI the Soviet Union's version oI totalitarianism. It would not be
possible to choose a strategy oI building new institutions oI associational and cultural liIe as alternative bases oI hegemony in the existing society and also as the core
structures oI a new society, while on the whole accepting the ruthless eradication oI those new institutions along with the old under a revolutionary statism.
To sum up so Iar, while Gramsci avoids economic and political reductionism by diIIerentiating the associational and cultural dimensions oI civil society Irom the
economy and the state, the Iunctionalist trend in his thought combined with his strategic political goals and allegiances lead him to construe the institutions oI civil
society in a one-dimensional way. Although autonomous, the associational Iorms (types oI political parties and unions), cultural institutions, and values oI civil society
are precisely those most adequate to reproducing bourgeois hegemony and manuIacturing consent on the part oI all social strata. They are, in short, not dualistic but
thoroughly bourgeois. This version oI civil society must thereIore be destroyed and replaced by alternative Iorms oI association (workers' clubs, the new proletarian
party Iorm, or the "modern prince"), intellectual and cultural liIe (the idea oI the organic intellectual), and values that would be help create a proletarian
counterhegemony that might eventually replace the existing bourgeois Iorms. Yet even the strategy oI building a counterhegemony is just that, a strategy. Gramsci never
sees the institutions and cultural Iorms oI counterhegemony as ends as well as means, because he is unwilling to concede that, within bourgeois civil society, some
immanent possibilities extend beyond the established Iramework oI domination. Thus, in itselI, the Iocus on cultural means (the organization oI consent) in civil society,
as against the coercive means oI the state, does not imply that a radical reIormist project has replaced the revolutionary one. We are still dealing with a theory that
seeks the total replacement oI one Iorm oI society by another.
95
Furthermore, Gramsci's doctrine oI civil society is never advanced in terms that would imply an uncompromising hostility toward statism. This attitude, too, is
consistent with the Iunctional-

Page 152
ist strain in his thought. While at times he conceives oI hegemony as a product oI civil society, just as coercion is a product oI the state, so in other Iormulations both
hegemony/consent and domination/coercion are Iunctions oI the state, the Iormer pair operating on the terrain oI civil, the latter on that oI political society. It is this
second Iormulation that is consistent with the Iunctionalist reduction oI civil society.
96
According to its logic, one must regard hegemony not as autonomously
produced within civil society but as one oI the Iorms in which state power Iunctions eIIectively. The Iorms oI the establishment oI counterhegemony within the old
society can then be seen primarily as marking the way to a new state power that would have to establish on an entirely new basis the terms oI its own operation,
including a new basis in "civilization" Ior consent. Gramsci's remarks on the "civilizing" mission oI the state support this interpretation.
Gramsci stresses the notion oI the state as civilizing agent in two contexts in particular: the historical Iailure oI Italian uniIication, leading to the Risorgimento oI the
nineteenth century, and the problems oI Soviet development in the twentieth century. For our present purposes, we are interested in his analysis oI the Soviet context,
which he also used Ior comparisons with Iascist Italy. Like other Marxists, Gramsci built on Marx's analysis oI Bonapartism ("Caesarism") to analyze the structural
similarities oI modern dictatorships, all oI which use a more or less autonomous Iorm oI state power to organize an otherwise unstable system oI domination. Unlike
Trotsky, however, Gramsci did not argue Ior a speciIic diIIerence in the case oI the allegedly progressive version oI Bonapartism that would come Irom the working
class, somehow dominant and yet not ruling, on behalI oI which state power would act. Instead, he explicates the diIIerence in terms oI building a new culture or
preserving the old. But what is the meaning oI this new culture? Gramsci oIIers two interpretations, only one oI which is consistent with the thrust oI his own theory.
First, he argues that, Ior a progressive Iorm oI statism, "the point oI reIerence oI the new world in gestation" is "the world oI production; work," i.e., the organization oI
"individual and collective liIe . . . with a view to the maximum yield oI the productive apparatus."
97
This argument, in line with both historical materialist premises and
rather short-

Page 153
sighted apologetics Ior Soviet society, is consistent with the acceptance oI the obliteration oI the existing version oI civil society in the name oI a "progressive" agenda.
Indeed, Gramsci speaks in this context oI the repressive activity oI the state, its rationalization and Taylorization oI society, and its reliance on punitive sanctions.
98

The argument, however, is inconsistent with the antieconomistic turn in Gramsci's social theory: II the base does not determine the superstructure, how can the
character oI a new culture and a new society be determined simply by the transIormation oI the economic structure? And while Gramsci may have believed that in
some contexts the social sphere should be reduced by the actions oI the state to a mere complement oI economic transIormation, it is entirely unclear how this was to
be the source oI a new culture, especially one leading to a Iree society.
This last point becomes especially striking in light oI the second interpretation, which presupposes Gramsci's own original position within Marxism. Here the positive
role oI the state that can justiIy even "statolatry" is said to be "the movement to create a new civilization, a new type oI man and even a new citizen . . . the will to
construct within the husk oI political society a complex and well articulated civil society, in which the individual can govern himselI without his selI-government coming
into conIlict with political societybut rather becoming its normal continuation, its organic complement."
99
This criterion oI what constitutes the progressive version oI
statism is very diIIerent Irom the Iirst, namely, the creation oI a complex, well-articulated civil society capable oI selI-government as the hallmark oI a new culture.
Given the totalitarian obliteration oI civil society, however, the thesis is highly paradoxical. It may be that Gramsci had in mind the historical experience oI many early
modern states that abolished the institutions oI traditional European corporate society only to allow and even promote the emergence oI a modern structure oI civil
society. But the analogy does not quite work. Abolishing the old society oI orders was the joint work oI the state and democratic eIIorts Irom below that also
maintained their distance Irom state power. Thus, it is almost impossible to locate historically in most Western European countries (except perhaps in the Reign oI
Terror) that vanishing moment in which the old associations have disappeared

Page 154
while the new ones have not yet emerged. On the contrary, when totalitarian governments abolished civil society, dissolving already modern rather than traditional
Iorms oI culture and association, they speciIically disallowed the Iormation oI new types oI associations independent Irom themselves, including even, and perhaps
especially, the independent social organizations and movements that had helped overthrow the old regime. How convincing was it, then, to expect that a Iorm oI
statism that was more uncompromisingly hostile to civil liIe than any oI its predecessors would create Irom above a ''complex, and well-articulated civil society" that
would be able to govern itselI more or less independently? And what might be the Iorms oI this new type oI civil society that would be created Irom above, as diIIerent
Irom the modern one as this latter was Irom its traditional predecessor? This second question is important because the analogy Gramsci seeks to construct with past
statisms Iails iI we are to assume merely that a "totalitarianism" that dissolves a model oI civil society is progressive iI it recreates Irom above more or less the same
model, or even one oI its variants.
Gramsci does argue that a statolatry "abandoned to itselI" or "conceived oI as perpetual" must be subjected to criticism. How strong this criticism should be, and what
its political consequences might be, he does not say. And yet one gets the strong impression that he is aware oI what must have been a disturbing implication oI his
own thought, namely, that a leIt totalitarianism would not be normatively diIIerent Irom one oI the right iI it made no contribution to the reconstruction oI civil society.
And oI course only a Iool (oI whom there were many in the 1930s, though Gramsci was not one) could have thought that Stalin's Russia satisIied the normative criteria
here assigned to progressive dictatorships.
In this context, it is possible that Bobbio is right in arguing that Gramsci was at least on the verge oI recognizing that abolishing civil society is not the best way to
reconstruct it, even iI one seeks to create a new type oI civil society. II there actually was a radical reIormist strand in his thought, it would have been based on the
insight that the institutions through which a radical movement can build its hegemony are part and parcel oI any meaningIully conceived modern Iorm oI social selI-
government and, as such, have value in and oI themselves. It would have been based, in other

Page 155
words, on a recognition oI the dualistic character oI at least some oI the core institutions oI modern civil society. In short, Gramsci would have had to acknowledge
that the norms and organizational principles oI modern civil societyIrom the idea oI rights to the principles oI autonomous association and Iree, horizontal
communication (publicity)are not simply bourgeois or Iunctional to the reproduction oI capitalist or any other hegemony. Rather, they constitute the condition that
makes possible the selI-organization, inIluence, and voice oI all groups, including the working class. Accordingly, the task oI radical reIorm would be to expand such
structures in a direction that reduces the chances oI their being Iunctionalized to the purposes oI economic or political power. But such a stance would have led to an
outright rejection oI totalitarian revolution, a step Gramsci, unlike many oI his heirs, did not make.
Aside Irom the undoubtedly decisive political reasons why Gramsci did not make this move, what we have called his third antinomy also stood in the way oI his
reevaluating the problem oI civil society Irom a normative point oI view. This antinomy is between a conception oI a Iree society in terms oI a pluralistic, democratic
civil society and one in terms oI a uniIied state-society. The Iirst oI these models, consistent with the conIlict-theoretical strain in his thinking, and especially with the
conception oI the dual structure oI existing civil society, potentially tempers utopia with an imagery oI partial institutional continuity. Here, utopia is the realization oI
existing but blocked normative possibilities. The second model, consistent with Iunctionalism (the one-dimensional critique oI bourgeois civil society and the call Ior
total revolutionary rupture), suIIers Irom an excess oI utopianism and potential links to authoritarianism. One might say that the strain in Gramsci's thought involving the
relentless "unmasking" oI the role oI the institutions and political culture oI bourgeois civil society in reproducing capitalist relations oI domination helped to prepare the
way to an authoritarian position vis-a-vis civil society in general.
In our view, it is this second strain that is dominant in Gramsci's thought. Here, one cannot blame the timidity oI Gramsci's critique oI the Soviet Union, because, in
spite oI his overall sympathy and reluctance to drive his criticism too Iar, he may have had real doubts whether a genuinely Iree society would be created there.

Page 156
ThereIore, in contrast with the totalitarian project in which civil society is absorbed by the state, Gramsci returns to the Marxian program oI abolishing the state, which
he calls, with some variation on the original Iormula, "the reabsorption oI political society into civil society."
100
Marx, in his most explicit critique oI bùrgerliche
Gesellschaft (in "Zur JudenIrage"), wrote only oI an absorption in "society."
101
The diIIerence appears to be all the more signiIicant because, as Gramsci imagines
''the coercive elements oI the state withering away by degrees," he postulates the corresponding emergence oI"ever-more conspicuous elements oI regulated society
(or ethical state or civil society)."
102
Thus, his identiIication oI the new Iorm oI social organization that he most oIten calls "regulated society" with at least a version oI
civil society is quite deliberate.
Regulated society, a society without a state, seems to be deIined by two premises: (1) a premise oI equality and (2) a premise oI the replacement oI law by morality.
In other words, the new society is to be characterized by a spontaneous acceptance oI law by Iree and equal individuals without any coercion or sanctions
whatsoever. This notion comes perilously close to the selI-deluding Marxian utopia oI a society without institutions.
103
But the transition to regulated society that
Gramsci has in mind seems diIIerent. He reIers to a phase in which the state will indeed be a nightwatchman, in the sense oI saIeguarding the "continually proliIerating
elements oI regulated society" and in the process progressively reducing "its own authoritarian and Iorcible interventions."
104
This process is supposed to be identical
to the construction 'within the husk oI political society" oI a complex, well-articulated, selI-governing civil society. Thus, it is hardly an exaggeration to argue that
Gramsci's reIormulation oI the idea oI the road to socialism consists oI the construction oI a new type oI selI-governing civil society that would gradually take the place
oI all state control over social liIe, leading to a withering away oI the state as well as political society. Nevertheless, and amazingly enough, he does not believe that the
new type oI civil society in Iormation and its Iorms oI selI-government would enter into any conIlict with the state whose powers it is to erode and replace. Instead,
civil society would become the "normal continuation" and "organic complement" oI what he calls "political society," namely, the state.
105


Page 157
There are two images here that do not mix. On the one hand, we have a notion something like the emergence oI dual power: Two Iorms oI social organization exist
side by side; one based on democratic selI-government and social solidarity is to replace another based on administrative sanctions and coercion. On the other hand,
Gramsci leaves us with a notion oI a state power gradually converting its Iorm oI domination into an equally eIIective Iorm oI social control through the institutions oI
civil society. Thus, the antinomy between civil society as a consolidation or normalization oI domination and civil society as a genuinely alternative principle to
domination returns once again. This time, the two notions appear as one because the utopian idea oI the total absorption oI the state by civil society would logically
eliminate the distinction between a state power acting through the institutions oI civil society and a Iorm oI selI-government based on these institutions. Until society
reaches utopia, however, the ambiguity would remain, and the elimination oI conIlict Irom the model certainly seems to imply that Gramsci's supposed transition to a
Iree society is ultimately only a statist authoritarianism with a human Iace.
The utopia oI a (modern) civil society absorbing political society and the state, the supposed telos that would resolve the most important oI Gramsci's antinomies, is
incoherent even on its own. First oI all, it remains unclear which absorbs the other in the relationship between civil and political society. Here, the stress in Gramsci's
sparse descriptions seems to be on "political society" as understood by Tocqueville, Ior example, as political organizations rather than the state (as in Gramsci's use oI
the term). Regulated society is selI-governing, even iI its "laws" are enIorced as internalized moral rules that do not need to appeal to external sanctions. This highly
unrealistic postulate has authoritarian implications, at least in the modern world, that its advocates rarely conIront. Even iI we suppose that a period oI statist transition
has eliminated older Iorms oI heterogeneity and plurality, Gramsci's regulated society would have no social space Ior an opposition consisting oI new minorities and
pluralities that may be willing to obey the laws but cannot identiIy with them and may wish to organize themselves in order to reverse them.
106
With the sphere oI
prepolitical association eliminated or Iused with that oI political association, such an

Page 158
organization could not take place in principle. Indeed, the model oI moral rather than legal enIorcement eliminates the space in which such an opposition could emerge
at all: autonomous conscience, which is always to some extent in conIlict with the laws. The postulate oI a morally based acceptance oI law in itselI tends to
presuppose social homogeneity and to exclude pluralistic organization.
107
By deIinition, "pluralism" means some conIlict over policy and is thereIore incompatible with
internalized acceptance oI the decisions oI majorities. Thus, it is not clear how, and on what normative and empirical bases, individuals and groups could have rights
against Gramsci's monolithic regulated society.
The issue could be addressed Irom the point oI view oI the modernity oI Gramsci's idea oI regulated society. Can a civil society be a modern society iI state power is
abolished or absorbed? Does not the duality oI civil society and state (oI which Gramsci is a major analyst), not to mention the diIIerentiation between civil society and
the economy, constitute the modernity oI both terms? It would seem that abolishing the state, which is impossible in Iact but certainly imaginable, would lead not to an
autonomous, plural, civil society in other ways resembling its modern Iorerunner but to a restoration oI traditional political-civil society without modern administration
but also without a modern structure oI rights and liberties carving out autonomous spaces Irom the world oI politics.
108
Given an already established, sturdy, and complex structure oI civil society, albeit oI the bourgeois model, Gramsci's regulated society can be established only through
a revolutionary totalitarian rupture. Most oI the established institutions, including those oI the working class, would otherwise militate against it: The existing plurality oI
Iorms oI liIe, culture, and association, presupposing social conIlict, needs a structure oI laws and rights linked to sanctions. It also requires the mediating and interest-
aggregating outputs oI a modern state. No radical reIormist strategy would in itselI reduce this complexity, and indeed the organization and mobilization oI new social
actors would add to the heterogeneity oI interests and increase the conIlict potential oI society. UnIortunately Ior Gramsci's thesis, a revolutionary-statist destruction oI
the existing version oI civil society would have even less oI a chance to

Page 159
usher in the regulated society. The choice Gramsci actually Iaced was not between radical reIormism and revolutionary democracy prepared by a totalitarian abolition
oI civil society.
109
Rather, it was between civil society tout court and an authoritarian system that would certainly attempt to perpetuate itselI. Gramsci supplied
important concepts to those who would militantly challenge later versions oI such a system, but this was something he neither intended nor even anticipated. And those
who were to undertake the challenge could postulate the value oI an independent civil society only when they completely divested themselves oI the radical democratic
utopia oI the regulated society whose deepest roots involved, as Marx knew but Gramsci apparently Iorgot, a hatred Ior modern civil society.
Excursus on Gramsci's Successors:
Althusser, Anderson, and Bobbio
Gramsci's antinomic intellectual position allows two distinct and opposed roads Ior continuation. While diIIerent combinations among his alternatives are possible, there
is a more than elective aIIinity among an apologetic attitude toward the Soviet Union, a Iunctionalist reductionism with regard to the existing version oI civil society, and
a utopian project (or a normative countermodel) oI a uniIied state-society. With an emphasis on the Iunctionalist-reductionist component, this combination marks the
path oI Louis Althusser and his Iollowers, who insist on maintaining intact the Marxian project oI revolution. Similarly, the internal relation is equally strong among
critique oI the Soviet Union, a conIlict-theoretic and dualistic conception oI the existing civil society, and a pluralist democratic normative model oI civil society. This
combination is pursued by Norberto Bobbio, who has recently Iocused on civil society as the proper Iramework Ior contemporary radical reIormist projects oI
democratization.
Althusser entirely disregards the version oI Gramsci's theory that involves an opposition between state and civil society and resolutely Iocuses on the secondary version
in which civil and political society, hegemony, and domination are all Iunctional aspects oI the state.
110
Political society here becomes the "repressive state apparatus"

Page 160
deIined in terms oI a supposedly unitary structure oI government, administration, army, police, courts, and prisons. "Civil society" (his quotation marks) in turn
becomes a diIIerentiated Iramework, with the "ideological state apparatuses" consisting oI religious, educational, Iamily, legal, trade-union, communications, and
cultural components. Althusser has notorious diIIiculties in showing that all these domains belong to the state.
111
He dismisses, in part rightly, the objection that their
status is private, as distinct Irom the public "repressive apparatus'' oI the state, as so much bourgeois legalism masking the actual functions oI institutions. But this
strategy only justiIies a diIIerentiation Irom the private, economic sphere, not an inclusion in the structure oI the state. To argue that the ruling class holds state power,
the ideology uniIying the various institutions in question, by which they "massively and predominantly Iunction," is the ideology oI the ruling class, and "ideological
apparatuses" are therefore state institutions is both logically Iallacious and empirically questionable. It is logically Iallacious because, even iI the state were the
instrument oI the ruling class, the two terms would still not be identical, which is what Althusser's "syllogism" presupposes. And it is empirically Iallacious because, as
we know Irom the history oI social democracy Ior example, many nonbourgeois strata and groups can occupy state power in capitalist societies, and because the
institutions to which Althusser reIers are characterized by great ideological diversity, internally (Catholic vs. Protestant churches, Christian vs. syndicalist unions, etc.)
and among one another. Despite these seemingly obvious problems, this argument has had extended inIluence.
More important Ior us is Althusser's own inability to stick to a consistent version oI this Iunctionalist position. He rightly repeats the Gramscian position according to
which no Iorm oI power can be stable Ior long without "hegemony over and in the State Ideological Apparatuses."
112
But this thesis is not interpreted according to his
own version oI Iunctionalism when he argues that the Iunction oI the apparatuses and the ideology they supposedly produce is to reproduce the existing relations oI
production.
113
This latter argument separates civil society Irom the state and Iunctionally links its institutions, along with those oI the state, to the reproduction oI the
capitalist economy. Once Ireed Irom the

Page 161
absurd burden oI having to make civil society a dimension oI the state, Althusser can pronounce the ideological state apparatuses, i.e., the institutions oI civil society,
as "multiple, distinct, 'relatively autonomous' and capable oI providing an objective Iield to contradictions, which express, in Iorms that may be limited or extreme, the
eIIects oI clashes between the capitalist class struggle and the proletarian class struggle."
114
Not only does this argument implicitly shiIt between the two
Iunctionalisms (statist and capitalist) available in Gramsci; it is also on the verge oI rediscovering the other, conIlict-theoretic and pluralist-democratic, position
traceable in Gramsci's work.
115
However, since he is so much more dogmatic and traditional in his revolutionary and state-socialist commitments than Gramsci,
Althusser is even less able to travel this road than his predecessor. Even in a highly modiIied Iorm, the Iunctionalist road chosen by Althusser cannot lead to a genuine
reevaluation oI the normative double nature oI civil society.
The brilliant interpretation oI Gramsci by Perry Anderson, who was once a Iollower oI Althusser, is a case in point?
116
Anderson devastates Althusser's
reconstruction oI Gramsci both textually and politically. Politically, he calls the reconstruction disastrous because it cannot distinguish between Iascist-authoritarian and
liberal democratic versions oI capitalist society: Only the Iormer absorb social institutions oI cultural reproduction within the state.
117
But it is also textually wrong to
the extent that it Iocuses on a secondary conceptual strategy in Gramsci's work, disregarding the primary usage that diIIerentiates state and civil society.
Anderson argues that Gramsci developed this secondary usage, in which civil society is absorbed in the state, because oI diIIiculties with his primary one. It is not civil
society alone that wields cultural legitimacy; the state does as well, in particular through its educational and legal institutions (mentioned by Gramsci) and its
parliamentary structures (omitted by Gramsci but strongly stressed by Anderson). Gramsci's answer was to make coercion and hegemony Iunctions oI both civil
society and the state. The diIIiculties oI this conception, which threatens the deIinition oI the modern state as the monopolist oI legitimate violence, supposedly led
Gramsci to include civil society in the state or even to identiIy the two spheres with one another.
118


Page 162
Anderson's own solution, which in a sense combines those oI Gramsci and Althusser, is to maintain the separation oI civil society and state but to insist that, while the
institutions oI civil society produce only cultural hegemony and consent, the structures oI the statebecause oI the all-important role oI parliamentary institutions
produce consent as well as coercion. This idea assimilates the Althusserian notion oI ideological state apparatuses but maintains the Gramscian stress on the
production oI ideology outside the state as a secondary one. By this conceptual move, Anderson in eIIect overcomes the bad option between an overly schematic
diIIerentiation between state and civil society in the main version oI Gramsci's arguments and the complete absence oI diIIerentiation in the secondary version. In the
process, he inadvertently comes close to the Hegelian notion oI parliament as an institution oI mediation between civil society and state, as the place where
simultaneously civil society penetrates the state and a uniIied political will is Iormed. He comes close to such a view but, as we shall see, not close enough.
The new argument, in Iact, does not overcome the limitations oI Marxian Iunctionalism. Anderson is quite clear: Civil society as we know it belongs only to the
Iunctional reproduction oI capitalist society; "the 'private' institutions oI civil society" have no place in "any social Iormation in which the working class exercises
collective power."
119
With this assumption in mind, he is entirely consistent in Iearing and rejecting the whole Gramscian strategy oI trying to build a counterhegemony
within the existing version oI civil society, certainly more consistent than those who hope to use such a strategy as a road to the revolutionary establishment oI a uniIied
state-society. Anderson shares the latter dream and thereIore rejects a radical reIormist road that implicitly assumes the preservation oI key dimensions oI existing civil
society. Because such a strategy is powerless against the ultimate guarantee oI the existing system, the possession oI the means oI violence and repression, it can serve
only to integrate the working class into the established society.
120
The reIerence to violence and repression already indicates a shiIt to the level oI the "state apparatus." A key reason why the building oI counterhegemony in civil
society must Iail is that the main

Page 163
instance oI the ideological reproduction oI the existing system is exercised by parliament, within the sphere oI the state. This instance, however, is reinIorced by its links
to potential violence and cannot be simply bypassed or displaced by alternative institutions. As long as parliamentary institutions are not overthrown, their primacy in
the production oI consent cannot be successIully contested. Such, according to Anderson, is the real answer to Gramsci's puzzle concerning the stability oI liberal
democracies.
This answer cannot escape the antinomy between Gramsci's two views oI civil society, one monistic-Iunctionalist and the other dualist and conIlict-theoretic. The
problem lies in parliament's peculiarity as a mediating institution in the Hegelian sensein the Iact that it appears as the institution through which the state is
"penetrated" by civil society. Because Anderson does not Iully recognize this, he is ambushed by the consequences oI his own argument. Why, he asks, are
parliaments so successIul in generating consent? Why are they so rarely radically challenged under liberal democracies? To his credit, Anderson is suspicious oI
doctrines oI cultural manipulation, oI the generation oI passivity in the work place, and even oI the ability oI welIare-state beneIits to buy consent.
121
Parliaments do
not rely on consent produced by cultural, social, and economic institutions but generate their own. They do so by presenting individuals who are unequal and unIree in
civil society with an imagery oI equality beIore the state and, by way oI their representatives, active participation in the Iormation oI political will. This imagery in turn
produces the ideological code (equality, Ireedom, etc.) on which the secondary activities oI the generation oI consent all depend.
122
The idea oI parliament as the center oI ideological integration brings Anderson close to the Althusserian doctrine oI "ideological state apparatuses,"
123
which he Iinally
manages to make coherent by pointing to a process actually originating in the state that produces the ideological unity oI the diIIerent "apparatuses."
124
But Anderson
is even less able to remain consistently within the Iunctionalist mode than Althusser. On the one hand, the general ideological code emanating Irom parliament is said
merely to mask prevailing Iorms oI inequality and unIreedom. On the other hand, "The code is all the more powerIul because the juridical rights oI citizenship

Page 164
are not a mere mirage: On the contrary, the civic Ireedoms and suIIrages oI bourgeois democracy are a tangible reality, whose completion was historically in part the
work oI the labor movement itselI, and whose loss would be a momentous deIeat Ior the working class."
125
Anderson goes on to describe the autonomy oI
parliament, which makes all such bodies double, expressing both the Iunctional needs oI the cultural reproduction oI capital and the still potent historical achievements
that express the ideals oI the revolutionary bourgeoisie.
Anderson may admire these ideals, but he implies that he does not share them. He certainly rejects the quasi-Gramscian strategy oI using them and the spaces they
provide Ior building an alternative hegemony. It is unclear, however, what he would put in their place, how he would abolish them without promoting yet another
"momentous deIeat Ior the working class" whose members are admittedly still attached to equality and liberty in the sense oI contemporary parliamentarism. Anderson
proposes that this attachment can be broken only in the postrevolutionary experience oI proletarian democracy, "in parties or councils |sic| "where "the real limits oI
bourgeois democracy" can be learned and historically surpassed.
126
UnIortunately, he tells us little about this alternative democracy; more importantly, his thesis
implies that its principles cannot even be convincingly presented to those who now experience democracy in terms oI the established procedures. The link between the
two democracies would thus have to be in principle an antidemocratic one, a rather strange recommendation to those who presently value the beneIits oI liberal
democracies. One is asked to accept a revolutionary strategy on the basis oI a Iaith that somehow it will lead to a qualitatively diIIerent, yet untried and within the
present society entirely untriable, Iorm oI democracy.
That there is no such alternative Iorm oI democracy is the best-known thesis oI Norberto Bobbio. And yet Bobbio is a leIt socialist theorist oI democratization.
Though he is no mere Iollower oI Gramsci, his justly Iamous interpretation oI the Prison Notebooks is the key to his own distinctive theoretical position on the
question oI democracy. According to Bobbio, Gramsci Iought a two-Iront war against those who sought to assimilate civil society (and the state) to the economy
(economic determinists) and those who

Page 165
sought to subordinate it to the state and the cult oI Iorce. He wanted to transcend not only the conditions oI bourgeois society but also "the Ialse way oI transcending
these conditions."
127
In this way, somewhat ahistorically oI course, Bobbio seeks to distinguish Gramsci Irom both social democratic and Leninist politics. To
Bobbio, as we have already argued, Gramsci was a strategist oI "reIorm," in the strong sense oI wanting to transIorm not only politics and economics but also
"customs and culture." Indeed, the whole stress here is on the building oI an alternative cultural hegemony that must precede the conquest oI power, involving not only
the political party but also, and especially, the activity oI all institutions oI civil society involved in the production and diIIusion oI culture.
128
Thus, the center oI the
radical strategy in this interpretation is entirely relocated Irom the state to civil society, where a protracted "war oI position" Ior the conquest oI cultural hegemony
would be Iought.
In his Iamous article on Gramsci, Bobbio seemed to notice no inconsistency between this radical, civil- society-centered strategy and the goal oI a regulated society in
which civil society absorbs the state,
129
nor even that, on this point, Lenin's views oI the distant Iuture (though obviously not Soviet reality) coincided with the essence
(though not the terminology) oI Gramsci's position. Nevertheless, the Gramscian vision oI a monolithic, regulated society in which civil society would absorb the state is
not to be Iound the theory oI democracy and democratization that Bobbio developed in the 1970s and especially the 1980s.
130
On the contrary, his works oI this
period rejected in the strongest terms the idea oI a monolithic direct democracy. Instead oI the radical substantialist approach, Bobbio insists that the normative
procedural principles oI representative democracy constitute the necessary, though admittedly not the suIIicient, criteria Ior any state to be considered democratic. The
real problem Ior radical democratic reIorm, then, is to identiIy the reasons why liberal democracies have not succeeded in keeping their promises, and to articulate a
program Ior their Iurther democratization.
Accordingly, Bobbio states what he takes to be both a normative and a realistic (Ieasible) deIinition oI democracy. Every democratic government has three basic
prerequisites: participation (or collec-

Page 166
tive and general involvement, even iI a mediated one, in the taking oI all decisions applying to the whole community); control Irom below (on the basis oI the principle
that all power not so controlled tends to be abused); and Ireedom oI dissent.
131
Bobbio is, oI course, under no illusion regarding the realization oI the these principles
in existing liberal democracies. He argues that these promises have not been kept even in states where democratic institutions are the most Iully and Iormally
developed. Here, too, as in every modern society, there are at least Iour paradoxes oI democracy that make it diIIicult to realize its principles adequately: "In a
nutshell, these Iour enemies oI democracywhere I am taking democracy to mean the optimum method Ior making collective decisionsare the large scale oI
modern social liIe, the increasing bureaucratization oI the state apparatus, the growing technicality oI the decisions it is necessary to make, and the trend oI civil society
toward becoming a mass society."
132
In sum, we moderns seem to be demanding more and more democracy under conditions that are increasingly unpropitious. Moreover, these paradoxes seem to be
exacerbated in representative parliamentary systems. The phenomena oI political apathy and participation distorted and manipulated by elites with a monopoly on
ideological power have militated against the promise oI participation. Control Irom below is emptied oI signiIicance as the center oI power shiIts away Irom those
institutions that citizens succeed in controlling: The signiIicant instruments and centers oI real power, such as the army, the bureaucracy, and big business, are not
subject to democratic control. Finally, the right oI dissent is severely restricted in capitalist societies in which the dominant economic system never oIIers the possibility
oI a radical alternative.
What, then, is the sense oI calling contemporary Western societies "democratic"? By identiIying the (minimum) deIining principles oI democracy with the classical
(broken) promises oI democracy, Bobbio's works in the 1970s tended to make this question unanswerable. In the 1980s, he conIronted the issue with a procedural
turn in his thought, diIIerentiating minimum deIinition Irom norma Iive promise. He now deIined democracy in terms oI a procedural minimum that includes (1)
participation oI the largest possible number oI those concerned, (2) majority rule in

Page 167
decision making, (3) the existence oI real alternatives (persons and policies) to choose Irom, and (4) the existence oI guarantees oI Iree choice in the Iorm oI basic
rights oI opinion, expression, speech, assembly, and association.
133
Modern democracy is thus liberal democracy by deIinition, even iI Bobbio believes that there is also a built-in conIlict between democracy and those dimensions oI
economic and political activities calling Ior a strongly limited government.
134
Equally important, modern democracy is also a Iorm oI elite or oligarchic, pluralistic,
particularistic, and deIiciently public mass democracy whose democratic character is limited to the space oI politics alone. These characterizations amount, in Bobbio's
view, to a series oI broken promises with respect to the classical model oI democracy, even in its early modern, liberal restatements, all oI which involved a
relativization oI the distinction oI ruler and ruled along with varying emphases on individualism, universalism, publicity, and an educated citizenry.
135
Despite a
heterogeneous set oI causes diminishing the democratic character oI modern politiesthe survival oI secretive or invisible political practices, the capitalist character oI
modern economies, the elective aIIinity between democracy and bureaucracy, the overload oI demand produced by democratic party politics, and the increasing role
oI technical expertise in modern liIeeven these violations oI the classical promise oI democracy do not eliminate the minimally democratic character oI the existing
liberal democracies, which is procedurally deIined by majority rule, electoral competition, and civil liberties.
136
This point, however, could be reversed: The
procedural minimum apparently cannot diminish the elite, particularistic, nonpublic, and depoliticized Iorm oI democracy in modern societies.
Bobbio is certainly not satisIied with this conclusion. He stresses the socializing aspect oI the procedural minimum oI democracy, which promotes values oI toleration
and nonviolence in conIlict resolution and, less convincingly, those oI solidarity and openness to radical cultural learning experiences.
137
More importantly, he strongly
believes that the Iurther democratization oI existing democracies is possible. This issue is addressed on three levels: the possible place oI direct democracy; the role oI
alternative Iorms oI representation; and the possibility oI expanding the space oI democracy Irom the state to civil society.

Page 168
Already in the 1970s, Bobbio insisted that there is no Iull-Iledged, realizable alternative to representative democracy that would satisIy the classical promise oI
democracy better than the existing model does.
138
In a manner quite reminiscent oI Roberto Michels, Bobbio shows convincingly that neither oI the individual
institutions oI "direct democracy"reIerenda, local committees or assemblies, the binding mandatenor even their combination oIIers a Ieasible replacement Ior the
representative system. ReIerenda are simply inIeasible Ior all the issues that must be debated and resolved collectively in complex modern societies. The problems that
a local committee or assembly is competent to discuss are hardly identical to those conIronting a national polity. Binding mandates already exist where strong party
systems are in eIIect (party discipline being the Iunctional equivalent Ior the mandat imperatif), and where they are not in eIIect, the question remains as to the nature
oI an acceptable authority able to revoke a mandate. Finally, an alternative model oI "socialist" democracy based on the dual strategy oI structural reIorm and the
widening oI participation would run up against two additional diIIiculties. First, a structural reIorm that radically aIIects the economy is hard to imagine without invoking
violent means, which have never led to an increase oI democracy. Second, the widening oI democratic participation in the sphere oI economic power comes up
against what appears to be a permanent Ieature or countertrend common to socialist as well as capitalist states, namely, the removal oI economic power Irom the
province oI democratic control Irom below. While it is debatable whether the conditions Iavoring autocratic power in this sphere are historically determined or
objective, Bobbio maintained (at least in the 1970s) that there are good grounds Ior suspecting that the progressive widening oI the democratic base will eventually run
into an insuperable barrier when it tries to pass the Iactory gates.
139
But should representative and direct democracy be seen as exclusive alternatives? In the 1980s, Bobbio began to see them as potentially complementary. First, there
was a possibility oI mixed or intermediary Iorms such as representation with binding mandates. Second, one could also include direct democratic Iorms such as
reIerenda, recall, and local assemblies into representative demo-

Page 169
cratic constitutions.
140
Bobbio remains skeptical toward the intermediary Iorms he mentions, and he rejects any Iurther extension oI the already abused imperative
mandate. Moreover, he considers the role oI complementary direct democratic devices important but necessarily limited. He sees the reIerendum, Ior example, as
appropriate only when a relatively Iew issues oI uncompromisable principle are at stake. Thus, his own model oI democratization relies primarily on extending new
representative rather than direct Iorms.
Within the sphere oI state institutions, ideas oI Iunctional democracy have oIten been proposed as extending the logic oI democracy to a level that has become more
important in modern society (according to Emile Durkheim, the most Iamous exponent) than the territorial one. The best-known proposal Ior such representation
articulated by the guild socialists and Austro-Marxists, among othersinvolves an additional parliamentary chamber representing proIessional associations outside oI
the party system. To Bobbio, such a scheme implies a poor and even dangerous alternative to territorial representation. Such representation oI interest groups would
simply deliver parliament to lobbying by special interests and to the deals they would make. To the extent this is already a "degenerative" tendency oI the existing Iorms
oI parliamentarism, it should not be made worse by being raised to a principle and an institution. While Bobbio does not believe that anything like a general interest
emerges in contemporary parliaments, he nevertheless claims that the political parties that dominate these structures represent a superior Iorm oI mediation between
the individual and state than do interest groups. To the necessarily rigid pattern oI the representation oI group interests he counterposes general visions available in
political movements leading to the potentially creative and Ilexible handling oI issues. Political parties thus represent the diIIerently interpreted, multiIaceted interests oI
citizens, as against the narrow and inIlexible interests oI group members.
141
To choose a party means to choose a general Iramework oI interpretation based on
political opinions. On the other hand we do not choose our interest group; our relation to it is ordinarily not political but is deIined by shared social and economic
interests.

Page 170
Only in this polemical context do we get such a gloriIied description oI the logic oI representation through the party system. Yet the idealized picture distorts what
could have become a more diIIerentiated analysis and proposal. Bobbio could have stressed the possibility oI a Iorm supplementing rather than replacing
representative democracy as it is now oIIicially practiced. Good arguments and models are available Ior such proposals. II neocorporatist bargaining already
characterizes contemporary political processes, as he repeatedly admits, there may be some signiIicant virtue in bringing such negotiations into the light oI the public
sphere, thereby diminishing their corporatist character, to which Bobbio, in light oI the Iascist experience, is understandably allergic.
142
Moreover, a second
parliamentary chamber could take on a secondary role in relation to the Iirst; it could be overruled on the basis oI a qualiIied majority in the territorial chamber, and its
Iunctions could be limited to certain types oI issues. All oI this is important because, as we shall see, Bobbio's alternative strategy oI democratizing civil society may be
Iutile iI the channels enabling democratic associations, organizations, and movements to inIluence the political system are not increased with respect to the ordinary
practice oI party political elite democracy.
Within a general program oI democratization, Bobbio's emphasis is on the expansion oI Iorms oI representative democracy beyond the sphere oI politics. He hopes in
Iact to redeem two ''promises" that were not inherent in either the classical or the liberal model oI democracy: expanding the space oI democratic decision making and
exploiting the potential oI pluralism. In this context, he mentions a variety oI roles that can be democratized (in particular, Iamilial, occupational, educational, and client
roles) as well two major institutions that are not at present organized democratically: the school and (inconsistently) the work place. His justiIication Ior choosing these
is the one used by Durkheim Ior his theory oI Iunctional representation, namely, that it is here that "most members oI modern society spend the majority oI their
lives."
143
At issue is not the invention or re-creation oI new and direct Iorms oI democracy but the "inIiltration" oI new spaces, the spaces oI civil society, by "quite
traditional Iorms oI democracy, such as representative democracy."

Page 171
Interestingly enough, Bobbio's earlier doubts concerning the democratization oI economic liIe are not dispelled; the prospects here are still pronounced uncertain, as
they are Ior the sphere oI administration. Nevertheless, he insists that while the process oI democratization oI civil society has only just started, considerable progress
has already been made in areas such as schooling, where he stresses the participation oI parents in school councils, apparently a relatively new experience in Italy. On
the basis oI such examples, Bobbio maintains that the new index oI democratization in the Iuture will be provided not "by the number oI people who have the right to
vote, but by the number oI contexts outside politics where the right to vote is exercised."
144
This conclusion seems premature on the basis oI the empirical support Bobbio provides, but he has a more theoretical line oI reasoning to back it up. He argues that
pluralism, although not democratic in origin, provides both a reason and an opportunity Ior democratizing civil society. Bobbio insightIully demonstrates the origins oI
modern pluralism and democracy in two diIIerent polemical situations. Originally opposed not so much to autocracy as to monocratic Iorms oI power, pluralism or
polyarchy is in conIlict with monolithic models oI democracy, whether ancient or modern. In other words, given the dominant models oI democracy in the early
modern period, pluralism was antidemocratic. And yet Bobbio is right: Pluralism, based on the heterogeneity oI conIlicting interest constellations, cannot be eliminated
in complex societies. As Iar as he is concerned, this Iact represents a violation oI the promise oI democracy because nondemocratically organized centers oI power
bring particular interests to bear on processes oI decision making and also remove these important centers Irom democratic controls. Yet antipluralist, individualist
Iorms oI resistance on the part oI democracy would oI course be Iutile under genuinely modern conditions. Democracy can counterattack only by bringing extrastate
and even nonpolitical centers oI power under its own logic. In the context oI pluralistic society, the promise oI democracy can be redeemed only through the extension
oI processes oI democratization through the whole Iabric oI human association. And this requires not a Iundamentalist program oI direct democracy but the
introduction oI representative democracy in the relevant polyarchic centers oI society.

Page 172
So Iar the argument is convincing. But Bobbio also claims that the greater distribution oI power characteristic oI pluralism itselI "opens the door to the democratization
oI civil society."
145
One is hard-pressed to Iind an explanation in his text Ior how a pluralistic organization provides targets Ior democratization and even Iacilitates
such process, although he does at one point reIer to dissent promoted or shielded by pluralistic organizations. The claim is, moreover, implicitly contradicted by the
Iollowing assertion:
The process oI democratization has not even begun to scratch the surIace oI the two great blocks oI descending and hierarchical power in every complex society, big business
and public administration. And as long as these two big blocks hold out against pressures Irom below, the democratic transIormation oI society cannot be said to be complete. We
cannot even say whether this transIormation is possible.
146
It seems, then, that some oI the most important centers oI power greatly resist their own democratization. It is unIortunately the case that iI we measure
democratization by the extent to which a single set oI procedural standards extends into diIIerent spheres oI society, the results will inevitably be mixed, and
nondemocratic spaces or centers oI power are likely to remain "so numerous and so large, and their importance so great"
147
as to place the whole project in
signiIicant doubt.
Without wishing to replace Bobbio's somewhat pessimistic conclusion concerning democratization by a more optimistic scenario, we believe that a Iew critical remarks
may help elucidate the reasons why his own civil-society-centered program has reached an impasse. First, Bobbio does not consistently operate with the Gramscian
notion oI a civil society diIIerentiated Irom the economy.
148
As a result, he cannot clearly distinguish spheres whose internal logic Iacilitates radical democratization
Irom spheres whose reproduction is consistent only with subsidiary Iorms oI democratic participation. His overly procedural deIinition oI democracy does not serve
him well in this context: It makes him demand too little oI elites in some spheres (e.g., political parties) and too much oI elites in other spheres (e.g., capitalist
managements).
Second, Bobbio does not pose the question oI the internal relations oI diIIerent democratized spheres. As a result, his progno-

Page 173
sis, according to which spheres oI society can be democratized in an order that more or less reverses their general social importance, seems to reduce unduly the
stakes oI democratization. What is needed is a demonstration oI how and under what conditions newly democratized spheres can inIluence the less democratic spaces
oI society. In this context, his general pessimism concerning the introduction oI new structures into existing versions oI political democracy does not serve him well.
Finally, Bobbio does not distinguish between pluralism as a context oI institutions that can and should be democratized and the plurality oI collective actors that are to
carry out the work oI democratization. His remarks on social movements and civil disobedience do not indicate much conIidence in "extrainstitutional" actors as agents
oI democratization.
149
We are thereIore leIt with the suspicion that he entrusts such processes to the elites presently ensconced in the relevant pluralistic institutions,
including the parties oI the political system. Such a position would be reason enough Ior pessimism; the work oI democratization cannot ordinarily be entrusted to the
beneIiciaries oI less democratic or even nondemocratic arrangements.
We do not share Perry Anderson's critique oI the leIt socialist appropriation oI Gramsci. In our view, it makes little sense to criticize Bobbio on the ground that his
strategy cannot lead to a radical rupture with the institutions oI parliamentary democracy, since he speciIically and rightly rejects the idea oI rupture. Nor does he make
a transition to socialism the goal with respect to which democratic politics can be reduced to a mere means; in general, it seems that the very meaning oI socialism is
transIormed here into that oI the radicalization oI democracy.
150
With all oI this, we are in agreement.
Our criticism oI Bobbio has to do with the unIinished nature oI his program oI democratization, which in part is linked to the undeveloped and even ambiguous nature
oI his conception oI civil society. But even this criticism should not disguise our Iundamental agreement with two oI the most important Ieatures oI Bobbio's conception:
his displacement oI the terrain oI democratization Irom the state to civil society, and his insistence on a nonIundamentalist program in which Iormal and representative

Page 174
democracy provides the general model that should be Iollowed in the various spheres oI society. These achievements, based on a speciIic interpretation oI Gramsci,
link Bobbio to the most important strategies oI emancipation oI the 1970s and 1980s. And yet it remains doubtIul that his conception could provide such initiatives
with an adequate Iramework oI orientation and selI-understanding. At issue are not only his ambiguities with respect to the concept oI civil society, his somewhat too
generous concessions to the elite theory oI democracy, his one-sided conception oI pluralism, and his deemphasis oI social movements in Iavor oI political parties.
These imperIections could be corrected within the terms oI his theory. On a deeper level (and this is a diIIiculty he shares with the Iorms oI discourse within social
movements), it is not automatically obvious that the concept oI civil society taken over Irom Hegel and other nineteenth-century authors can with only a Iew
corrections sustain a program oI democratization and yet avoid the ideological utilization with which Parsons's theory culminates. Bobbio never considers the
possibility that the whole conceptual strategy may be intimately linked to now obsolete nineteenth-century conditions beIore the "Iusion" oI state and society; that even
in its original utilization it may imply not only antistatism but depoliticization as well; that it might represent only a set oI institutional masks Ior deeper and more reIined
authoritarian strategies; and Iinally, that the model oI social diIIerentiation it presupposes is a Ialse and unsophisticated one that is inadequate to the realities oI complex
societies.
In our view, the kind oI theory Bobbio seeks to develop cannot be constructed until these criticisms are considered in detail. We believe Iurther that the several
paradigms oI the critique oI civil society associated with Carl Schmitt, Hannah Arendt, Reinhart Koselleck, Jürgen Habermas, Michel Foucault, and Niklas Luhmann
will yield important contributions to our attempt at theory construction. It is to these critiques that we now turn.

Page 175
II-
THE DISCONTENTS OF CIVIL SOCIETY

Page 177
4-
The Normative Critique:
Hannah Arendt
One oI the most challenging, and certainly the most passionate, critiques oI modern civil society has been presented by Hannah Arendt in a whole series oI books and
essays.
1
Arendt's main, though barely mentioned, antagonist is Hegel. Her attack is concentrated speciIically on the concept oI "society" as an intermediate realm
between private and public, between Iamily and political liIe. "Society" is a realm oI mediations where private interests, activities, and institutions assume public roles,
while public institutions take on private "housekeeping'' Iunctions. Thus, to Arendt, institutions such as Hegel's corporations and police do not stabilize and regulate the
diIIerentiation oI public and private but rather dissolve the sharp line between them and threaten the integrity and autonomy oI both. Unlike Hegel, Arendt does not
seek a synthesis oI modern society and ancient republicanism. Instead, she resolutely deIends the model oI classical political society, politike koinonia, along with its
sharp separation Irom the oikos or private sphere, against modernity, particularly against the modern state (bureaucracy) and modern (mass) society. Her critique is a
normative one based on what she takes to be the values oI classical public liIe (political equality, public discourse, and honor) and private liIe (uniqueness, diIIerence,
individuality). Unlike that oI the young Marx in 1843, whom she in many respects resembles, Arendt's is not an immanent criticism. The actual political reemergence
and reinstitutionalization oI these values requires an almost total rupture with all existing institutions. A history oI decline Irom the emergence oI "society" to mass
society, seen as more or less

Page 178
inexorable, deprives modernity oI its one admitted achievement: the development and enrichment oI the private sphere as a sphere oI intimacy. Thus, like Walter
Benjamin, Arendt consciously practices a Iorm oI redemptive criticism that, Ior the sake oI a possible Iuture, attempts to save some valued aspects oI the past Irom the
perceived disintegration oI tradition, including the tradition oI early modernity.
2
We examine Arendt's critique in detail Ior several reasons. First, she will help us counterbalance the Parsonian conception by providing rich insights into the dark side
oI the institutionalization oI modern civil society. Second, the internal contradictions oI her analysis will help us show that not even Arendt was able to base a modern
theory oI Ireedom on the abolition oI civil society; she, too, is Iorced to assume, however unwillingly, the necessity oI its preservation. Third, a comparison with the
early work oI Reinhart Koselleck and Jürgen Habermas will allow us to show that in the modern world one can make sense oI Arendt's normatively based project,
which revolves around the concept oI the public sphere, only iI it is relocated around the intermediary sphere oI the social that she sought to banish.
The concept oI the social in Arendt's work corresponds to the Hegelian topos oI bùrgerliche Gesellschaft and is, in Iact, counterposed to both the political society oI
the ancients and the civil society oI the modern liberals. While these two conceptualizations emphasized the public sphere in the case oI the ancients and the private in
the case oI liberalism, "the social realm," a creation oI modernity occluded by these two political philosophies, involves a mixture and interpenetration oI the two realms
and their constitutive principles.
3
To understand the mixture, we must Iirst analyze its components.
Arendt's theory oI the public sphere, although systematized around a theory oI action, is derived Irom her understanding oI the model oI ancient republics. She
conceives oI the polis as "the organization oI the people as it arises out oI speaking and acting together."
4
Action in turn is understood as the selI-disclosure and even
selI-renewal oI the actor through the medium oI speech, possible only in presence oI others who see and hear and hence are capable oI establishing the reality oI
subjective expression.
5
Action

Page 179
is thereIore always interaction that both conIirms the plurality oI unique experience and personality and establishes a common world, "relating and separating" human
actors at the same time. This common world is the public sphere.
One striking diIIiculty oI Arendt's conception is that it describes both an anthropologically constitutive condition oI human liIe and a historically speciIic and unique
constellation: the ancient city republic (and its alleged, but admittedly exceptional, modern revivals). In this she Iollows the prejudices oI the Greeks, and she tries to
escape the resulting diIIiculty through her conception oI power.
Action, or rather interaction, is constitutive oI the public sphere,
6
but it is supposedly only power that can keep it in existence.
7
Power in turn is deIined as acting in
concert, on the bases oI making and keeping promises, mutually binding one another, covenanting.
8
While Arendt's model oI action stresses the striving oI the actor
Ior the Iame and even "immortality" that can be achieved through dramaturgic selI-presentation based on the rhetorical skill "oI Iinding the right words at the right
moment,"
9
her concept oI power points to action oriented to normative principles that derive their Iorce Irom the depth-structure oI a Iorm oI communication based on
mutual recognition and solidarity.
10
Thus, the concept oI action can be understood as a general anthropological constituent oI the ''human condition," but the concept
oI power, and along with it a Iully institutionalized public sphere, seems to require a republican model Ior its Iull actualization. And Arendt does in Iact link power more
closely to political speech than to action in its primordial, "rhetorical" sense.
11
The public sphere in Arendt's view presupposes a plurality oI individuals unequal by nature who are, however, "constructed" as politically equal. According to her, the
meaning oI the polis as isonomia (literally, equality in relation to law) is that oI "no rule," in the sense oI an absence oI diIIerentiation into rulers and ruled within the
citizen body.
12
Thus, the public sphere establishes a model oI interaction characterized by noncoercive discourse among citizens who initially hold and Ireely exchange
a genuine plurality oI opinions.
13
This model turns out to be rather restrictive. Based on her diIIerentiation between action and work, praxis and poiesis,

Page 180
Arendt at times goes along with what she takes to be the Greek exclusion oI legislation, decision by voting, and even the Iounding oI cities Irom the properly public,
political activities.
14
When she made her journey Irom Greece to Rome, however, in On Revolution, she made the act oI Ioundationthe making oI constitutions or
the exercise oI le pouvoir constituantthe public political activity par excellence. Yet she kept an important consistency between the two positions, namely, the view
that public liIe must be seen exclusively as an end in itselI. Thus, genuine republican constitution making in the later view ought to have no other purpose than to
institutionalize the public sphere itselI.
15
Arendt thereIore strongly rejected, as contrary to the very principle oI publicity, the idea that actors bring into their common
deliberations the interests, needs, and concerns oI their private lives and households.
Arendt describes the all-important relationship oI public and private in terms oI diIIerentiation, complementarity, and conIlict. She starts by diIIerentiating principles
described variously in terms oI action vs. labor and work, constructed reality vs. natural reality, uniqueness vs. real diIIerence, Ireedom vs. necessity, no rule vs.
domination, or equality vs. inequality.
16
For Arendt, an actual and thoroughgoing institutional diIIerentiation is required Ior the operation oI the principles oI both
private and public Ior two reasons. First, the complementary role oI the private vis-a-vis the public can be perIormed only in context oI their separation. Second, in
each other's terrain the two principles have a strong tendency to vitiate and even abolish one another.
Abstractly, the Ireedom oI public liIe requires the conquest oI necessity, the task oI the private in its economic capacity, as oikos.
17
Thus, the organization oI the
household was such as to provide its head with suIIicient time Ior the exercise oI public Ireedom. But Arendt's stress is on the conditions required Ior the emergence oI
the citizen as an independent subject, possessing substantial and independent opinions. The institutional Iorm oI the private as property (in contrast to mobile wealth)
guarantees this independence by setting up "external" boundaries among citizens and households; its "interior," by oIIering a hiding place Irom the light oI publicity, is
the precondition Ior nurturing the unique aspects oI personality without which liIe becomes entirely "shallow."
18


Page 181
In spite oI the importance oI a diIIerentiated private realm Ior the public, the latter also involves Iear and suspicion oI the Iormer. This is based on the possible
distraction oI the citizen by a model oI private happiness, but even more on the temptation to impose on the polis the despotic Iorms oI rule, inequality, and
diIIerentiation characteristic oI the oikos.
19
While in this context Arendt speaks oI the "permanent threat" oI the private to the public, elsewhere she maintains that in
the ancient world the greater danger was "the tendency oI public power to expand and to trespass upon private interests." This possibility, "inherent in republican
government," could be checked only by institutionalizing private property and eventually by the modern alternative, born in renewed republican experimentation, oI
Iraming laws that publicly guarantee the "rights" oI privacy, that is, the creation oI constitutional rights.
20
While Arendt always maintains her staunch support Ior such rights, she nevertheless argues that they do not suIIiciently protect the diIIerentiation oI public and private
under modern conditions. In particular, neither the speciIically modern Iorms oI the invasion oI the public by the private nor the resulting assaults on privacy and
intimacy by a new and corrupt Iorm oI "public" liIe can be counteracted by public rights oI private persons. Arendt connects both tendencies to a single phenomenon:
the rise oI the social.
Even iI she thus admits tendencies within the public and the private to invade one another's domains, Arendt consistently claims that the ancient republics managed to
maintain the diIIerentiation that belonged to their own constitutive conditions. The actual interpenetration and even Iusion oI the two is a product oI modernity, oI the
rise oI the social realm that constitutes the target oI Arendt's critique oI civil society. The interpenetration, in line with latent tendencies oI both the public and the
private spheres, goes both ways. The state (i.e., the modern territorial compulsory association) takes over Iunctions oI material reproduction, or "housekeeping," while
collective liIe, in the shape oI the nation, takes on the structure and Iorms oI behavior oI a superhuman Iamily. Arendt's Iormula Ior the political Iorm oI the rise oI the
social, the nation-state, expresses this two-sided interpenetration.
21
The result oI mutual interpenetration oI public and private is the disappearance oI any stable boundaries between "two realms |that|

Page 182
. . . constantly Ilow into each other."
22
In the new topos, however, an entirely novel type oI hybrid structure comes into being that will become the dynamic center oI
a process leading to the eventual disappearance oI both public and private.
The origins oI this social realm are analyzed in quite diIIerent terms in Arendt's various works. At least three points oI origin are distinguishable among these: the early
modern political or national economy; the depoliticized court society and the emergence oI salon society; and the modern democratic revolution. In each case, the role
oI the early modern state, created by absolutism, is central. The Iirst explanation, which comes closest to the Marxian tradition,
23
stresses the selI-organization oI the
absolute monarchy "as a tremendous business concern" that Iailed, according to one version oI the argument, to Iind an adequate class basis.
24
In this version, it was
the state that elevated matters oI mere housekeeping into the public realm, in the process deIorming that realm with concerns that were incompatible with its basic
principles.
25
It should be noted that in this context "the social" becomes synonymous with "political economy." Its supposedly almost unrestrainable expansion is
associated with the modern phenomenon oI unlimited economic growth. Here the step to an economy-centered neo-Marxist argument is a rather small one, and
Arendt actually takes this step when she describes limitless economic growth as the expansion oI the private realm at the cost oI the public.
26
The second train oI argument is, in part, Tocquevillian. The thesis is that absolutism destroyed its own class basis by depoliticizing the Stàndestaat in the Iorm oI a
society oI orders whose model and preeminent institution was courtly society.
27
This argument stresses conIormism, secret manipulation, and intrigue as the results oI
"depoliticization" rather than "economization." The most important consequence was that the French nobility was reduced to insigniIicance. In other words, this
development oI the social occurred at the expense oI political society.
These two arguments may indeed be compatible, but they share a common Ilaw: Both seem to imply that, beIore the process oI absolutist depoliticization and/or
economization, diIIerentiated public and private realms existed, each operating according to its own proper logic. Because she relies on a normative model derived

Page 183
Irom the ancient city republics, however, Arendt explicitly contradicts this implicit claim. Rightly or wrongly, she posits the loss oI the Greek understanding oI politics in
the medieval period and the absence oI a public realm in the secular sphere oI the Ieudal epoch. Since she depicts medieval corporate liIe as having patterned all
human activities on that oI the household, it is hardly Ieasible that she could consider the Stàndestaat based on it as a model oI public liIe, in her sense oI this concept.
28
Arendt's third line oI argumentation, developed in On Revolution, proposes a model that avoids this diIIiculty, but in the process she throws into doubt the historical
relevance oI the other two theses. Here Arendt solves the problem oI what precedes depoliticization by crediting the "republican" moment oI the modern revolutions
with re-creating the classical model oI the public. It then makes sense to argue that it was the Iailure to institutionalize this moment and/or the emergence oI the "social"
question led to the subsequent dediIIerentiation oI public and private and their decline. In the case oI the French Revolution, however, the argument concerning the rise
oI the social is an entirely new one. According to Arendt, the revolution in its radical phase opened the political realm to the poor, to the multitude driven by material
need, in the process making matters public that by their very nature belonged to the private realm oI housekeeping and could be solved not by public-political but only
by administrative means.
29
Thus, once again, despite the republican ethos oI the revolutionaries, government turned into administration. OI course, the turning oI
government into administration was anticipated by the monarchic absolutist Iounders oI the modern state. Recalling her earlier argument as a counterpoint, Arendt now
states that iI in the old regime economic and Iinancial problems could be said to have "intruded" into the public sphere, "the people" violently burst upon it.
30
And iI
"high society'' imposed its mores and moral standards on politics, reducing it to intrigue and perIidy, the society oI the poor, driven also by its earlier exclusion Irom
society, transIormed public liIe into its very negation: brutality and violence.
31
Evidently, then, and somewhat inconsistently, Arendt sees the mercantilist economization oI politics, the absolutist depoliticization

Page 184
oI the aristocracy, and the revolutionary socialization oI public liIe as successive and increasingly destructive Iorms oI the rise oI the social realm, which will be
Iollowed by the successive Iorms oI mass society and totalitarianism, involving the complete eradication oI both public and private. Her analysis oI the American
Revolution, however, indicates that the overall trend implied by the thesis oI the rise oI the social does not require the stages just depicted. American history knows
only Iailed attempts at mercantilist economization and even more so at absolutist depoliticization. In particular, Arendt argues that the social question did not burst upon
the public-political stage in America and that here, unlike all other revolutions, the institutionalization oI a diIIerentiated private sphere protected by constitutional rights
was Iully successIul.
32
And yet exceptionalism in these respects obviously did not prevent the United States Irom developing its own brand oI mass society, indeed
Ior many the paradigmatic model.
Like other analysts, Arendt had diIIiculty perceiving the reality oI the modern state behind the institutions oI American Iederalism and pluralism. Yet this reality does
make an appearance when Arendt analyzes the American Iailure to Iound lasting institutions oI republican Ireedom. The reasons Ior this included a Iailure to
institutionalize small-scale structures oI direct political participation and an increasing identiIication oI Ireedom as well as the aims oI government with the negative
Ireedoms oI private liIe protected by constitutional rights. But these points are not on the same level as the arguments dealing with the rise oI the social sphere; indeed,
they imply only the strengthening oI the private at the expense oI the public.
Nevertheless Arendt maintains that the retreat to the values oI private as against public happiness, and the reduction oI Ireedom to civil liberties alone, along with the
rise oI utilitarian criteria in politics and the domination oI public liIe by a uniIorm, homogeneous public opinion, correspond also in America "with great precision to the
invasion oI the public realm by society."
33
As Ior this invasion, we get only two related reasons, which do not add up to an explanation on the level oI the rest oI
Arendt's thesis. To begin with, she speaks oI "rapid and constant economic growth" equivalent to the "constantly increasing expansion oI the private realm"

Page 185
at the expense oI the public.
34
This is a simpliIied version oI a classical Marxian thesis that does not in itselI explain the emergence oI the new structural topos, the
social realm. For this realm, Arendt, unlike Tocqueville,
35
is able to discover only European origins. According to an entirely unconvincing train in her argument, the
immigrant poor oI Europe, conIronting American riches based on economic expansion and technological innovation, brought over the social question Irom its originally
European home. We are led to believe that it was Ior this reason above all that the American dream oI the "Ioundation oI Ireedom" was converted into that oI the
IulIillment oI all material desires.
36
Thus, immigration in America supposedly played something like the role oI the radical phase oI the French revolution; that is, it
converted inadequately instituted republican structures and practices into the rule oI a public opinion whose ultimate interest was in satisIying needs proper to the
private spherethe needs oI consumption.
Irrespective oI the problem oI origins, Arendt depicts the "hybrid" sphere oI the social as an extremely dynamic one with devastating consequences Ior both public and
private. Even to those who, like ourselves,judge her analysis to be highly one-sided, the depiction yields an impressive analysis oI the underside oI the
institutionalization oI modern civil society matched only by Marx beIore her and Foucault aIter her.
37
The key terms in Arendt's analysis oI the deIormation oI the public realm are bureaucracy, welIare state, public opinion, and political corruption. We note that the Iirst
three correspond with some precision to the categories oI Hegel's analysis oI civil society and state that mediate between private and public: civil service, "police," and
public opinion. The category oI corruption in turn leads to a critique oI interest representation in the party system that is implicitly a modern variant oI the Iourth
Hegelian mediation, the corporation.
According to Arendt, bureaucracy is the "social" Iorm oI government par excellence because the social question which is to say questions oI collective welIare, can
have only administrative solutions.
38
Arendt does not, in Iact, deny the need Ior civil service or administration under modern Iorms oI government. She argues only that
when questions oI welIare become the predominant or

Page 186
even exclusive questions in the liIe oI the state (as in the So:ialstaat or welIare state), the result is bureaucracy, in her terminology the rule of the administration, which
can become the most tyrannical Iorm oI all.
39
Bureaucracy is an especially arbitrary Iorm oI government because it involves rule by decree, with the holders oI
discretionary power becoming anonymous and invisible behind the Iacade oI other, apparently more political Iorms oI deliberation and decision making. II tyranny is
"government that is not held to give account oI itselI," then bureaucracy, as rule "by Nobody," goes so Iar as to hide the agents who might be held accountable.
40

Such, according to Arendt, is the case in modern welIare states, where the idea oI democracy is converted Irom that oI public participation to the achievement,
through the most eIIicient administrative means possible, oI the goals oI public welIare.
41
The procedures oI public participation are not, however, merely deIormed Irom above; they are also hollowed out Irom within. The social Iorm oI politics is the
corruption oI politics: It takes three Iorms linked to status, wealth, and need, respectively. Members oI the depoliticized aristocratic orders oI the old regime continued
to act together in court society to improve their status, but they could not do so in the properly political sense oI relying on open speech. Thus, public deliberation and
persuasion were replaced by the "pull, pressure and the tricks oI cliques," the result being mores and moral standards that open the door to intrigue and perIidy.
42
The
peddling oI inIluence replaced the generation oI power. The same pattern occurred in salon society. Indeed, the eighteenth-century Rousseauian attack on "society,"
reproduced by Arendt, was an attack on the hypocrisy oI the court and its analogues, the aristocratic salons, and the hypocritical, unnatural power oI women.
43
But
Arendt does not restrict the notion oI the corruption oI politics to this obvious example. For her, it is as part oI a genuine public liIe that property owners emerge Irom
a protected private realm to pursue public aIIairs. When, however, property is replaced by "wealth," and the pursuit oI political goals by the deIense and generation oI
ever-expanding wealth, the corrupt Iorms oI acting together generated by aristocratic society become the best means also Ior the "bourgeois" to pursue private goals
that cannot by their very nature be validated publicly. Finally, the popular response to

Page 187
the corruption oI status and wealth, the brutality oI people driven by need, itselI corrupts politics and is corruptible by "politicians." Here, too, the proper medium oI
political conIlict and competition is replaced by a principle wholly at variance with it: instead oI the secret interaction oI cliques and maIias, the violence oI those unable
to use political speech.
44
What ties these examples oI political corruption together, in the context oI the depoliticizing bureaucratic rule by Nobody, is the quasi-political interaction oI people in
their private capacity who lack the institutions oI a public sphere that could establish their capacities as citizens. It is nonetheless part oI Arendt's thesis that the
eighteenth-century revolutions sought to establish precisely such institutions. Their Iailure was not simply a result oI the intervention oI bureaucracy and private wealth
into the public sphere, problems that even the ancients had to Iace, as Arendt well knew. The core oI her thesis about the speciIically modern decline oI republican
politics thereIore depends on the eIIect oI the social on the very structure oI the public: the transIormation oI public spirit into public opinion.
Once again, Arendt assigns a pioneering role to "high society," to the absolutist court and its extension in the aristocratic salon.
45
Indeed, it is this cultural development,
unlike the problems oI bureaucracy and poverty, that is unique to modernity and thus a pivotal point in the analysis. It is here, in a space neither private nor political,
dominated by status consciousness and empty uniIorm conventions, that public liIe Iirst acquired, according to Arendt, the Iorms oI interaction characteristic oI a
uniIied, conIormist, corrupted, collective opinion. All those who sought to enter "high society" or "society" were Iorced to submit to this logic, producing conIormity
and assimilation.
46
Courtly and salon society, characterized by the basest pursuit oI private interests, intrigue, unnatural pretentiousness, concern Ior status and style,
and corruption (in the sense oI utter lack oI concern Ior the res publica) became the model oI behavior that was emulated by the rest oI society.
47
But what is the dynamic oI the dramatic extension oI this logic beyond "society" in the narrow sense, the beginnings oI which can be ascribed to the absolutist
suppression oI politically meaningIul speech and oI the plurality oI political opinion within the aristoc-

Page 188
racy? To Arendt, the revolutionary transIer oI the notion oI sovereignty Irom king to people and the concomitant rise oI the politics oI interest are the best symbolic
representations oI the relevant trends.
48
The "compassionate" response by the radical revolutionaries in France to the multitude driven by need led them to substitute
will Ior consent, unity Ior plurality, and a single opinion Ior the conIlict oI opinions, because any accommodation oI consent, plurality, and conIlict seemed to
compromise the most urgent and desperate measures required to solve the "social question." The mythological sovereignty oI the people, in the sense oI a collective
will whose only object was a uniIied general interest, thus became the Ioundation oI a public opinion that could only be threatened by independent public liIe, including
the new, decentralized, and inevitably plural institutions oI the popular strata themselves.
49
And while the dictatorial embodiment oI this supposedly general will did not
arise Irom an actually uniIied or uniIorm public opinion, it was in a position to create such an opinion.
50
While the Iact oI nationalism allows Arendt to extend her critique oI sovereignty beyond dictatorial-populist regimes, the argument again does not work well enough
Ior the United States, where nineteenth-century critics such as Tocqueville uncovered a public opinion oI unrivaled uniIormity and assimilating power. Arendt does
recall a part oI Tocqueville's argument contrasting democracy and republics. A democratic society involves the kind oI social leveling that could open up the way to a
new kind oI plurality, one oI opinion, only in context oI creating genuine republican institutions based on Iree communication even at the micropolitical level. This eIIort
having in large part Iailed, democracy in America came to reveal some oI the despotic characteristics Ieared by the Iounders, with the public spirit, based on a
multiplicity oI opinions, replaced by a uniIied and homogeneous public opinion. Arendt insists that this trend was checked politically through the survival oI some
republican institutions on the national and state levels. Nevertheless, the rise oI a politics oI interest, common to both Europe and America, tended to complete the
destructive process.
Interest as against (genuine) opinion is politically relevant only when belonging to a group, indeed a large group. The representation oI interest more or less binds the
representatives and interIeres

Page 189
with the genuine exchange and Iormation oI opinion. The modern party system in particular, by Iocusing on interest representation, ends up replacing parliamentary
discussion by the competing collective opinions oI disciplined party blocks. The hierarchic and oligarchic structure oI the party thus becomes the model oI
contemporary politics. The welIare state may be democratic in representing the interests oI the many, but it is oligarchic in the sense oI drastically curtailing
participation on all but the highest levels oI the state.
51
The Hegelian attempt to mediate private and public spheres through intermediary social-political bodies thus
winds up reducing the space Ior public Ireedom within the structure oI the state.
The situation is made all the worse, in Arendt's assessment, because the decline oI the public does not beneIit the private; the social tends to destroy the private sphere
as well. In this context, Arendt distinguishes between private property, which constitutes the outer shell oI the protection oI privacy, and wealth. The latter is a means
oI deIorming the public realm but is incapable oI protecting the private.
52
Because oI its Iluidity and absence oI stable location, wealth is supposedly unable to
guarantee a sphere in which the individual is Iree Irom any external gaze or penetration. More convincingly, Arendt argues that, the object oI wealth being its own
accumulation and consumption, its pursuit commits individuals to uniIorm trends, reIlected by a behavioral science, oI not only economic production and distribution
but daily liIe as well. Not only does laboring activity become mindless and uniIorm, but the liIe oI the home is invaded by a process oI homogenization and
commodiIication that destroys the possibility oI any authentic private liIe. In our unlimited drive to consumption, we Iinally consume the material Iramework oI the
private.
53
Mass society, the society oI job-holders and consumers, presupposes the absorption oI the immense variety oI Iamily liIe into a uniIorm, homogenized social
realm that becomes a Iamily writ large.
54
The private sphere resists this absorption by a speciIically modern creation: intimacy. On the level oI a small circle oI interpersonal relations, intimacy involves a
tremendous deepening oI the private sphere, in the sense oI an intensiIication and enrichment oI "subjective emotions and private Ieelings."
55
This Iorm oI privacy is

Page 190
structured in terms oI an opposition not to the public but to the social. Despite its immense contribution to the culture oI modernity, though, the intimate sphere does
not represent a reliable substitute Ior the protection to privacy oIIered by property. The intimate sphere cannot deIend itselI against the modern pattern oI "unnatural
growth" oI the social, because the intensiIication oI subjectivity cannot yield a stable and intersubjective or institutionalized "world."
56
Without tracing any Iurther the Arendtian thesis concerning the decline oI public and private, leading at least ideal-typically to a Iull-Iledged "mass society" and to
totalitarianism, we should note the structure oI her view oI modern society. In this theory, the complex oI the social, constituted by modern bureaucracy and political
economy, conIronts two realms on which the reproduction oI authentic human liIe depend: the public and the private. These realms do appear in modern society, even
iI in a situation that threatens their very existence. Any reconstruction oI the human condition, then, would obviously depend on a new or renewed institutionalization oI
both public and private. Arendt's theory oI the modern revolution, understood broadly, explores the chances oI such a double reinstitutionalization. In the process, she
not only revives the spirit oI ancient republicanism but is Iorced to do so in ways that require taking yet another look at the modern topos oI a diIIerentiated civil
society.
In another context, Arendt links the idea oI a diIIerentiation oI state and society, already associated with "the rise oI the social," to the rise oI a "modern" Iorm oI
republicanism. She explicitly notes that the early modern (especially Lockean) version oI social contract theory reIers to two contracts and to the origin oI two
diIIerentiated entities: "society" and "legitimate government." We should not be misled, however; Arendt explicitly deIends only the principle oI the Iirst contract, resting
on reciprocity, mutuality, and equality and rooted in promising and making covenants. More-over, she interprets the Iirst contract in terms oI a constitution oI bodies
politic locally, regionally, and, ultimately, on a Iederal level, leading to a multiplication oI power. She does not, in other words, see that the principle oI horizontal
covenanting establishes an intermediate sphere between the strictly private and the political

Page 191
public, whose very principle would be voluntary association.
57
SigniIicantly, she is skeptical regarding the second contract, whose principle she understands as
submission and surrender oI power and the creation oI a relation between ruler and ruled, only supposedly "legitimated" by mere consent. Indeed, unlike the tradition
oI natural rights to which she reIers, she seems to consider the two contracts as mutually exclusive.
58
But does the mere existence oI the second contract really vitiate
the Iirst? And is the Iirst contract suIIicient unto itselI in Iounding a body politic? The model oI diIIerentiation Arendt entertains here is in Iact that between political
societv and state, and iI one takes her analysis as a whole, it is not entirely clear whether diIIerentiation thus understood sustains or undermines a modern political
society. The dilemmas oI modern republicanism that she is Iorced to note do indeed lead her back to a model oI diIIerentiation, rather than away Irom it.
Arendt's revival oI ancient republicanism, oI the ideal oI civil society as politike koinonia, in the contexts oI modern revolutions Irom Paris to Budapest, is deservedly
well known. Juxtaposing direct participation to representation, and Iederalism to uniIied sovereignty, she presents us with a model oI pyramidally organized "small
republics," "councils," or "wards" capable oI institutionalizing a Iramework oI public Ireedom and establishing a Iorm oI government at all levels linked to the paradigm
oI the communicative generation oI powera veritable "great republic."
59
She is conscious oI this idea's link to the ancient model oI politike koinonia; reIerring to
colonial America, she speaks oI the selI-constitution oI "civil bodies politic'' that were "political societies" open to Iederalism but hostile to the depoliticization that
would accompany a centralized state with uniIied sovereignty.
60
As against the contemporary constituted bodies oI the European old regime ("diets and parliaments,
orders and estates"), the American political societies were not tied to privilege, birth, or occupation and opted Irom the outset Ior status in public rather than private
law.
61
Thus, rather than resembling the political societies oI the age oI absolutism or even the Stàndestaat, the small American republics consciously returned to the
ancient model oI an incorporated citizen society, a genuine res publica. It is emphatically this conception that remains

Page 192
normative in Arendt's political project. And yet, the very egalitarian universalism that diIIerentiates these two conceptions oI political society derived Irom something
new, namely, the constitutive principles oI civil society, as Tocqueville knew well.
While insisting on the continuous history oI her model oI political society Irom the age oI revolutions to our own day, Arendt is Iorced to concede the repeated Iailure
oI permanent institutionalization. Unlike Tocqueville, though, she does not seem to be aware oI the contradictions oI what she knows to be an aristocratic model oI the
political selI-selection oI elites in a "democratic age."
62
But she does note three areas as sources oI this repeated Iailure: (1) internal limits oI the historical attempts to
build council governments; (2) the diIIiculty oI stabilizing an instituted or constituted power in a model based on the act oI instituting or constituting; and (3) the clash
between republicanism and liberalism, between models oI public and private happiness.
The Iirst problem area revolves around the encounter between the council model and the modern economy and the modern state. Arendt repeatedly laments that all
council experiments aIter the American revolution became mired in the social question (e.g., the Parisian societies) or in impossible attempts to democratize the world
oI work (e.g., workers councils Irom St. Petersburg to Budapest). Arendt's dismissal oI any sort oI industrial democracy Ilows Irom the dogmatics oI her conception,
Irom the automatic institutional translation oI her separation between action and work. Her caution about deriving utopian models oI workers' control Irom a monolithic
concept oI democracy is well Iounded, oI course, although it is certainly wrong to pose the question oI industrial democracy as an all-or-nothing proposition.
Moreover, her notion that revolutionary councils should have Iocused exclusively on the questions oI establishing and preserving the new political regime is quite
unrealistic, even iI one accepts her thesis oI the primacy oI the political rather than the social moment in modern revolutions. In this context, her hard-headedness
concerning the constraints implied by the modern state is surprising, iI welcome. She admits the need Ior a modern administration in a modern society and rightly
criticizes the inability oI the council experiments to come to terms with the "enormous extent to which

Page 193
the government machinery in modern societies must indeed perIorm the Iunctions oI administration."
63
Thus, she is Iorced to return to precisely the model oI
diIIerentiation that she rejected in her discussion oI the two contracts. UnIortunately, however, she turns once again to an overly rigid version oI this model, conIusing
diIIerentiation in principle with that oI the concerns oI actual institutions. As a result, the iron-tight division oI Iunctions she suggests between the political action oI
councils and the administrative work oI a civil service represents no solution whatsoever to the task indicated.
Her attempt to deal with the second problem is also only partially successIul. Arendt is Iully conscious oI the diIIiculty or even selI-contradiction inherent in a project
aiming at the embodiment oI the revolutionary spirit in enduring institutions.
64
Without hesitation she renounces the politics oI any kind oI permanent revolution based
on the continuous Iunctioning oI a pouvoir constituant that inevitably produces its own tyrannical opposite.
65
But how, then, can the revolutionary spirit be embodied
at all? The aim oI revolution, according to Arendt, must be the creation oI Ioundations Ior a new political order, a new constitution. She maintains that such a
constitution, as against any liberal or even "constitutionalist" interpretation, must establish power rather than limit it. The juxtaposition is misleading, however, because
the establishment oI an unlimited power, inevitably returning us to a model oI permanent revolution, could not yield any institutionalization oI stable political Ioundations.
And indeed Arendt attributes this dimension oI institutionalization to the rule oI law rather than the exercise oI power.
66
But what is the source oI a law that could lend
stability to a constitution iI our positive laws are Iounded in the constitution itselI? How are we to escape the vicious circle inherent in constitutional lawmaking itselI?
What is the source oI the legitimacy oI a constituent assembly, and iI it is legitimate, what can justiIy its selI-dissolution? Arendt does not believe that any version oI a
return to the eighteenth-century theory oI an absolute natural law, prior to and above constitutions, can supply the answer to these questions today.
67
As a result, she
has a great deal oI diIIiculty in distinguishing between the source oI law and that oI power, precisely the dilemma that, in her analysis, leads to the radical instability oI

Page 194
constitutions. Her only answer, in the spirit oI Roman rather than Greek antiquity, is that the constitution as an act oI Ioundation can replace the absolute source oI law
iI it develops, as it did in the United States, into a tradition oI a new type. In such a context, the vicious circle oI constitutional law will apply only to the Ioundational
moment. Subsequently, the constitutional tradition itselI, authoritatively interpreted by a body without power, will supply the sanction Ior a law capable oI stabilizing the
Iramework oI the exercise oI power. But is the making oI constitutional authority a matter oI tradition compatible with the pursuit oI public Ireedom as our highest end?
Can a Ireedom whose vehicle is public communication and discourse stop at the limits constituted by a supposedly sacred Ioundation? The deep tension between civil
religion and public Ireedom is built into this model Irom the outset, a tension only exacerbated when the concept oI public Ireedom is replaced, as in the actual
historical trend, by that oI the private.
The third reason Ior the diIIiculty oI establishing institutions oI public Ireedom in Arendt's analysis is represented by the clash oI ancient republican and modern liberal
principles, by the subversive implications oI the goal oI private happiness Ior public Ireedom. In this context, Arendt Iinds that she cannot give a selI-contained
republican answer to the challenge oI the liberal model oI civil society based on the separation oI a valued society and a state without norms. This comes about
primarily because she holds the public and private spheres to be, in their diIIerentiation, constitutive Ior one another. She does maintain with approval that the "actual
content oI the |U.S.| constitution was by no means the saIeguard oI civil liberties but the establishment oI an entirely new system oI power."
68
But she also makes
repeatedly clear that, without the saIeguard oI civil liberties, at least in the modern world, public political liIe cannot be maintained. She is leIt in the end with the
precarious position that, while the establishment oI civil liberties represents a very real, though unIortunately all too exceptional, gain in revolutions, too great a Iocus on
rights and the private happiness they can secure tends to devalue public happiness and Ireedom.
Arendt is well aware oI the origins oI civil rights in the modern sense. On one side, the modern sovereign state represented an entirely new type oI threat to individual
autonomy; on the other,

Page 195
the more or less contemporary erosion oI the traditional, religious, and corporate Iorms oI protection made individuals increasingly deIenseless. The paradox oI
"human rights" Irom Arendt's point oI view is that while protection is needed in the Iace oI the modern state, only within the Iramework oI a state is such protection
plausible at all. Outside the body politic, the most Iundamental right, namely, the right to have rights based on the ability to assert and deIend rights publicly, cannot be
secure. Thus, modern rights should be understood as citi:en rights guaranteed by constitutions.
69
This early strain oI her argument seems to make civil rights Iunctions
oI a public sphere to be established in the Iace oI the modern state.
Arendt soon came to understand that rights, even iI they could be stabilized only as rights oI citizens, must be deIended even in Iace oI a citizen body, iI need be. Given
the tendency, always present, oI public power to absorb private interests, and given the modern erosion oI a Iorm oI property capable oI carving out a private space
oI protection Ior citizens, civil rights are needed to stabilize the private sphere. It is at this point that Arendt most clearly concedes the Iundamental liberal claim that in
a modern society, Ireedom is not possible unless civil society and state are diIIerentiated by mechanisms oI civil rights.
Arendt then immediately moves Irom the liberal thesis to one inspired by Marx. While civil rights can indeed protect the private sphere Irom penetration by the modern
state, they cannot do so in the Iace oI the modern economy.
70
Arendt does not in this context consider the possibility that an expanded and reorganized catalogue oI
rights could actually have an analogous relationship to both state and economy. Whatever hope she has concerning the restriction and control oI economic Iorces and
growth thereIore depends on the existence oI a public realm rediIIerentiated Irom the social one, independent oI "political economy." Yet here, too, civil rights must
play a role to the extent that a diIIerentiated private sphere remains the sine qua non oI the emergence oI personalities capable oI participating in the public sphere
itselI. Whatever success the American revolution had in establishing republican institutions is related to the preservation oI civil rights, while the Iailure oI all other great
revolutions in this respect is linked to their systematic violations oI rights.
71


Page 196
Thus, in Arendt's conception, the diIIerentiation oI civil liberty protected by rights and public Ireedom secured by the exercise oI political power helps to establish
each. But at the same time, the erosion oI the line between them tends to destroy both. This restatement oI her conception oI public and private in terms oI democracy
and rights is not, however, the end oI Arendt's consideration oI the problem oI civil liberties. She also maintains that, Irom the moment oI its establishment, the model
oI Ireedom based on rights remains a threat to the model based on power.
According to Arendt's interpretation, the U.S. Bill oI Rights sought only to control and limit republican power instead oI trying to replace its aims by nonpolitical ones,
as was the case with the French Declaration of the Rights of Man and Citi:en.
72
And yet even here, a reversal was to occur in which public Ireedom came to be
subordinated to civil liberties, the citizen to the private individual. At stake are two diIIerent models oI happiness leading to two diIIerent understandings oI
"constitutionalism." In Arendt's over-all diagnosis, the shiIt Irom the values oI public happiness, Ireedom, and civic spirit to private happiness and the corresponding
negative model oI Ireedom tend to be ascribed Iundamentally to the rise oI society. But since her analysis oI the rise oI society, especially in the American context, is
never really adequate, she is also at times tempted to reverse the causal nexus. She maintains, in other words, that a liberal component stressing private happiness (the
cultivation and enjoyment oI one's private concerns) as the highest end oI liIe tended Irom the beginning to undermine, in the philosophical selI-understanding iI not the
practice oI the American revolution, the republican component linked to the idea that public happiness based on political participation is the highest good.
73
Thus, the
model Arendt deIends not only Iully diIIerentiates public and private but also asserts the motivational primacy oI the the Iormer. With private happiness achieving
primacy, Ireedom was redeIined: Instead oI meaning the positive Ireedom to act, it came to mean negative Ireedom Irom the action oI others. Even more decisively,
the aim oI a constitution"constitutionalism"shiIted Irom the establishment oI a new Iorm oI genuinely public power to the protection oI individuals Irom the exercise
oI power. Political Ireedom came to be understood not as a Iunction oI an

Page 197
increase oI power, but as one oI a limitation oI power. This liberal ideal oI constitutionalism came to mean a distrust regarding all Iorms oI power and an increasing
indiIIerence toward the Iorm in which power was exercised, so long as civil liberties (themselves not powers) as the bulwarks oI private happiness were protected.
74

This version oI constitutionalism, however, proved to be quite compatible with a bureaucratic Iorm oI government expressing the logic oI the "rise oI the social." In the
end, a rights-centered politics could not deIend even the private sphere itselI against the destructive trends oI the modern state and society.
Arendt's understanding oI rights itselI suIIers Irom a lack oI diIIerentiation reIlecting her unwillingness to take seriously the idea oI mediation between private and
public. The ambiguous status oI the rights oI assembly and association in her work proves this point. On the one hand, these rights are classed with negative liberties,
that is, Ireedom Irom unjustiIied restraint. Even in the American Bill oI Rights, the right oI assembly was, according to her, only "the right to assemble in order to
petition." What the individual gains Irom such a right is "liberation" rather than "Ireedom"; at most, the ability to petition collectively may lead to some restraining
inIluence over, but never participation in, a government.
75
On the other hand, the rights oI assembly, association, and speech are also reIerred to as the most important
truly political Ireedoms, as contrasted with apolitical Ireedoms such as that oI enterprise.
76
While she argues that this status was reached through a development
beyond the limits oI the Bill oI Rights, Ior example,
77
she does not clariIy whether and how this supposed development produced a new status Ior what remained
juridically a constitutional liberty. In any case, even the discussion admitting the status oI the right oI assembly as a political Ireedom culminates with a declaration that
"political Ireedom, generally speaking, means the right 'to be a participator in government' or it means nothing.''
78
This declaration sets up standards that can rarely be
satisIied by what the right oI assembly actually guarantees in even its most developed versions.
The issue goes deep in Arendt's conception oI rights and reIlects her ambivalence concerning the ultimate Ioundation oI rights. Indeed, she has two conceptions
concerning the core oI a system oI

Page 198
rights. One links "the right to have rights" to access to a public sphere in which rights can be asserted and deIended.
79
The other isolates the right to carve out a
private sphere linked to private property as the basic model oI all rights. Accordingly, the right oI assembly is interpreted in each oI these ways, as a dimension oI
political participation and also as part oI the private to be protected Irom the public. The Iirst conception would make the rights oI assembly and speech the most
Iundamental rights. The second conception, however, tends to assimilate these "rights oI communication" to the model oI property right, depriving them oI any special
importance in the catalogue oI rights. This ambiguity does in Iact reveal something about the peculiarly double nature oI the rights oI communication. But,
characteristically enough, what does not arise at all in Arendt's conception is that the right oI assembly is both civil and political, both private and public. In other
words, Arendt has no room Ior the concept oI a right oI juridically private persons who can thereby attain publiclaw status and even exercise an important public
role, thus mediating between private and public spheres.
The absence oI the very possibility oI mediation between public and private in Arendt's work is all the more serious when the exercise oI the rights oI assembly and
association turns explicitly political, in particular in the case oI social movements. Indeed, social movements could have played a constitutive role in Arendt's theory in
the context oI a problem she could not adequately deal with. Since movements have empirically demonstrable liIe cycles, she could have cast them as embodiments oI
revolutionary spirit that do not imply a permanent revolution. Indeed, she could have interpreted them as extrainstitutional instances oI the generation oI power that in
the long run presuppose and promote rather than interIere with institutionalization.
80
Arendt is, oI course, aware oI the role oI movements in the emergence oI council-
Yrepublican experiments; above all, she examines the workers' movement, which "has written one oI the most glorious and probably most promising chapters oI
recent history."
81
Working outside the economically oriented labor unions and the socially oriented "political" parties, the movement oI the industrial working class
repeatedly reinvented the genuinely political project oI constructing new,

Page 199
republican institutions. According to Arendt, this was possible whenever the class oI labor, not yet admitted to society ("class society"), suddenly appeared on the
public political stage.
82
Her argument is entirely Iictitious, though, since the movements Irom 1848 to 1956 to which she reIers cannot be represented as having no
social and economic interests and demands, and even less as not playing a major part in the economic reproduction oI society. And indeed Arendt herselI is Iorced to
admit this implicitly when she argues that interest in workers' control oI industry, ever present in the movements under consideration, was a major reason Ior the
downIall oI council experiments.
83
Arendt's linking oI the movement Iorm with republican experiments is, in her own view, relevant to a phase oI modern history already past. With the emergence oI
"mass society," a Iorm oI "society" capable oI absorbing all classes including that oI labor, no movement can hope to claim a status that is exclusively political, rather
than social or economic. Now the labor movement becomes a pressure group like any other.
84
Since interest articulation and interest representation (parties) are, at
best, the "politics" oI civil society, they substitute Ior real political participation, replacing political society and discursive opinion Iormation with bargaining and deals.
Because interest groups and party politics destroy the parliamentary public space, they are in Iact inIerior to the administrative processing oI interest claims.
85
How does this square with the persistence oI the movement Iorm in our time? Could it be that the despised terrain oI the social could aIter all become the scene oI
repoliticization in the context oI movements that constitute a new public sphere and thereby mediate between the private and the public? Arendt certainly argues that
the movement Iorm itselI does not disappear along with the classical workers' movement. Indeed, she assumes that their terrain is the social realm between what is leIt
oI the private and the public. Adopting a radicalized version oI the pluralist critique oI mass society that uses totalitarian movements as its paradigm, however, Arendt
is convinced that social movements accelerate and complete the social realm's destruction oI the public and the private. That is, social movements proper Ieed oII and
help to create and perpetuate the atomization and depoliticization charac-

Page 200
teristic oI mass society. Under conditions oI the modern party system and the deep distrust engendered by it, Arendt sees Iertile soil Ior the emergence oI
extraparliamentary and extraparty movements. The more glaring the Iailures oI the party system, the easier it is Ior movements to arise and to appeal to wide
constituencies. But in the absence oI genuine public institutions, movements either organize masses or turn those they organize into masses. Social movements are mass
movements, and mass movements carry on the work oI the social principle by invading and leveling all hitherto private domains oI liIe, including Iamily, education, and
culture.
86
Thus, social movements are proto-totalitarian, and the totalitarian completion oI the rise oI society is not possible without them.
Given their starting point in society and their mobilization oI social needs and motivations, movements cannot reinvent Iorms oI public liIe. This thesis, we must note,
coincides with the conception oI social movements dominant in the early postWorld War II paradigms that studied social movements under the names oI collective
behavior and mass society.
87
Arendt's radical democratic political philosophy distinguished her work Irom these paradigms. But by partially buying into them,
probably under the impact oI her own experience with totalitarian movements, she deprived her political philosophy oI any possible politics.
88
II movements today,
because oI the inevitably social terrain oI their emergence and existence, cannot reinvent or extend the public sphere, and iI rights-oriented collective action is a threat
to the love oI public Ireedom, then it is not at all clear that in our epoch the experiments oI the working-class movement in creating political institutions can have any
continuation whatsoever. II Arendt is right about social movements as such, her dream oI the revival oI republicanism should be pronounced Iinally dead.

Page 201
5-
The Historicist Critique:
Carl Schmitt, Reinhart Koselleck, and 1ürgen Habermas
The Origins of the Liberal Public Sphere:
Carl Schmitt and Reinhart Koselleck
Hannah Arendt Iailed to demonstrate that her normative ideal oI the public sphere is compatible with modernity. We have argued that this Iailure was strongly linked to
her uncompromising critique oI the social sphere oI mediation, which she had identiIied as the speciIically modern dimension oI institutional liIe. Thus, it is oI great
importance that there is an alternative tradition oI interpretation Iocusing on the problem oI the public sphere. The approach oI Jürgen Habermas and his Iollowers
counterposes a sociallv rooted Iorm oI the public sphere to the ancient model identiIied with the state.
1
Remarkably, this second tradition goes back to Carl Schmitt,
who sought to deIend a conception oI "the political" based on a model oI war against what he took to be an apolitical conception based on public discussion, a model
that was to deIine the deepest impulses oI both Arendt and Habermas.
2
According to Schmitt, one oI the best ways to understand modern liberalism is by Iocusing on its "political" expression, namely, parliamentarism. The principle oI the
latter is open public discussion or deliberation.
3
Beyond mere negotiation and bargaining, what Schmitt has in mind is a model oI discussion in the sense oI
an exchange oI opinion that is governed by the purpose oI persuading one's opponent oI the truth or justice oI something, or allowing oneselI to be persuaded oI something as
true and just. . . . To discussion belong shared convictions as premises, the willingness to be persuaded, indepen-

Page 202
dence Irom party ties, Ireedom Irom selIish interests. . . . |T| he essence oI parliament is thereIore public deliberation oI argument and counterargument, public debate and public
discussion.
4
Thus, a common political will results Irom the process oI the genuine and open conIrontation oI diIIerent opinions. This process is supposed to be public in two senses:
by reIerring to the work oI an autonomous public body Iree to deliberate without any external compulsion imposed upon its members, and by being genuinely open to
the outside. In both oI these senses, modern parliamentarism is deIinitively contrasted with its Iorerunners, the estate assemblies, which were based on the imperative
mandate and closed sessions. Under modern parliamentarism, instead oI the direct pressure oI constituencies or any Iorm oI bound or mandated representation, public
opinion is supposed to "inIluence" the parliamentary public only through argumentation and persuasion that presupposes rather than suspends the independence oI the
representatives.
Schmitt anticipates and ascribes to liberal parliamentarism both the Arendtian deIense oI opinion against interest and the Habermasian model oI genuine argumentation
as distinct Irom strategic and rhetorical uses oI political speech. Unlike both oI them, however, he treats the discussion model as deeply apolitical, linking it to the
Iundamental liberal Iaith that unrestricted competition, which takes the Iorm oI discussion in the intellectual realm, produces harmony.
5
According to Schmitt, this
liberal model oI the parliamentary public sphere is taken over Irom moral and intellectual discourse on the one side and Irom economics on the other. It turns a
"politically united people" into a culturally interested public or an industrial concern operating in a market, in the process depoliticizing and demilitarizing the political
sphere, turning the state into society.
6
Schmitt is keenly aware that the state and politics in his sense do not thereby disappear in liberal society. The principle does not, need not, and cannot Iully correspond
to the actual practice. As he puts it in somewhat obscure language, "there is heterogeneity oI purposes . . . but there is no heterogeneity oI principles."
7
The principle oI
open public discussion is actually a principle oI legitimation, a normative and even metanormative principle. As such, its

Page 203
immediate importance is that it is the basis Ior the validity oI other norms. Schmitt particularly stresses that the norms oI the independence oI representatives, their
Ireedom oI speech and immunity, and the openness oI proceedings all receive their validity Irom the principle oI public discussion as the only legitimate method Ior
attaining a collective will.
8
Even the twentieth-century claim that parliament is the "best" method Ior the selection oI elites draws its legitimacy Irom the discussion
model (or what is leIt Irom it in a Iramework oI increasingly rhetorical interaction), since the valid testing oI leaders is identiIied with perIormance in debate and with
having the ability to persuade others successIully.
9
Schmitt is well aware that the principle oI publicity was capable oI operating only in a world diIIerent Irom that oI its own assumptions, involving a reduction oI all
politics to discussion. While the deepest striving oI liberalism, in theory at least, was to reduce the state to society in either the economic or the cultural sense, in Iact
liberalism presupposed and could not survive without a state, or without the dualistic coexistence oI state and society. Moreover, and this is the important point,
Schmitt, unlike Arendt, realizes that the principle oI discussion belongs to the level oI society rather than that oI the state. Quite in the spirit oI Hegel's
Rechtsphilosophie, parliament is thus seen as the penetration oI society into the state, reproducing in eIIect the society-state dualism in the state sphere itselI, thereby
"mediating" the split between the poles oI the duality.
Schmitt's modernized reconstruction oI the Hegelian Iramework is much cruder than that oI the master whose conception oI the "estate assembly" he cites.
10
In
particular, he does not distinguish between the system oI needs and the other levels oI civil society, nor does he recognize any mediation other than that oI parliament
between society and state. For him, all the Iundamental political polarities oI the epoch oI constitutional monarchies (prince vs. people, government vs. popular
representation, administration vs. selI-administration), under which he (inconsistently) subsumes classical liberalism, express one Iundamental dualism: society vs. the
state.
11
This dualism is, in turn, a Iunction oI the "polemical" attitude oI social Iorces (economic, intellectual, and religious) toward the bureaucratically uniIied military-
administrative state inherited Irom the epoch oI absolutism.
12


Page 204
But it was also a Iunction oI this state to remain independent and strong enough to stand above the other social Iorces: to be a strong enough threat to motivate the
relativization oI other Iorms oI social (economic, conIessional, cultural) opposition and conIlict and also the resulting selI-constitution oI a more or less uniIied "society."
At the same time, this state had to be selI-suIIicient enough to undertake and survive (and perhaps also be strengthened by) a policy oI nonintervention and selI-
neutralization vis-a-vis the societal spheres, allowing these spheres (economy, culture) to unIold their autonomous logics.
The stability and the equilibrium oI the resulting duality is achieved by the mediation oI parliament. "Popular representation, parliament, the lawmaking body is
conceived as the stage (Schauplat:) where society appears in the Iace oI the state."
13
On this stage, state and society are "integrated into" one another. In terms oI
Iorm, the result is dualistic, comprising a "legislative state" and an "executive state," with the Iormer, the Geset:gebungsstaat, gradually achieving primacy as the
nineteenth century proceeds. This development corresponds to the ideology oI parliamentarism already discussed, according to which only decisions achieved through
"discussion and the conIlict oI opinions'' are legitimate. The idea only apparently contradicts Schmitt's notion that the principle oI discussion is social and indeed
apolitical. The metaphor oI a stage seems to indicate that what actually occurs here is a mere play or show, necessary Ior integrating social Iorces and legitimating the
real decisions that are taken elsewhere and in another manner.
The polemical attitude oI society against the state implies that such a state oI aIIairs cannot be accepted. This is especially the case when the idea oI the selI-
organization oI society is democratized. For democratic Iorces that identiIy with their parliamentary representation, the residual nonparliamentary decision-making
power oI the executive, which bypasses the plurality oI social opinions instead oI integrating them, must seem illegitimate. The goal oI a completed legislative state
cannot be achieved, though. What is at issue here is not that a pure parliamentary state cannot be Iound in reality, any more than can other pure state types. Rather, the
parliamentary state, unlike other Iorms, represents the ideal oI the state as the selI-organization oI society, as the organization oI the

Page 205
state according to the societal principle oI discussion. According to Schmitt, as this ideal nears realization under the impact oI democratization, paradoxically the
parliamentary principle oI integration loses its Ioundations, and the state itselI, deprived oI another principle oI unity, is threatened with disintegration.
In Schmitt's conception, the ability oI those outside parliament to identiIy with their representatives rests on a polemical attitude to the state that guarantees the
uniIication oI a society otherwise potentially divided by conIlicts oI both opinion and interest. But this is not the whole story. The Iorms oI the selI-constitution and selI-
protection oI parliament vis-a-vis the executive actually turn out to be identical to the mechanisms diIIerentiating society and state. Clearly, parliamentary discussion
would be meaningless without the Ireedoms oI opinion and speech as well as the immunity oI representatives; these are presuppositions oI the constitution oI a genuine
public body. But Schmitt also indicates that a parliamentary public sphere implies the Ireedom oI public liIe outside oI parliament.
14
Interpreting Guizot, he asserts that
the openness oI parliamentary procedures would be meaningless without general Ireedoms oI opinion, speech, and the press. Without these Ireedoms, all Iorms oI
social control over parliament, which are required Ior the parliamentary representation oI society in the Iace oI the state, would disappear. Since Schmitt's model
presupposes and requires the ability oI private individuals to acquire and communicate their opinions Ireely, it seems that some other Ireedoms, such as those oI
assembly and association in their extraparliamentary Iorms, also represent "liIe-and-death questions Ior liberalism."
15
But Schmitt pays no attention to the social
consequences oI these latter Ireedoms, which provided Ior Hegel the possibility oI mediations other than parliamentary between individual and state. Finally (and
consistently), Schmitt makes no mention oI any Iundamental rights that cannot be derived Irom the principle oI parliamentary publicity, whatever their importance may
be Ior the liberal epoch (e.g., property). This consistency, however, only permits him the preposterous Iormulation that, with the decline oI parliamentarism, "the whole
system oI Ireedom oI speech, assembly, and the press, oI public meetings, parliamentary immunities, and privileges loses its rationale," which is based on the

Page 206
belieI that "Just laws and right politics can be achieved through newspaper articles, speeches at demonstrations, and parliamentary debates."
16
Schmitt's analysis leads to this conclusion, irrespective oI his political predilections, because his perceptive recognition oI the social Ioundations oI the model oI
discussion is coupled with a conception asserting more than the claim that the existence oI parliaments in the modern sense presupposes the diIIerentiation oI society
and state. He also aIIirms the converse, namely, that the unity and diIIerentiation oI society is structurally dependent (at least in the long term) on the existence oI a
parliamentary representation in the Iace oI the state, to which he, unlike Hegel, reduces the whole problem oI mediation. Yet he notes in passing that there are not
many people "who want to renounce the old liberal Ireedoms, particularly Ireedom oI speech and press," even when their political eIIicacy has become doubtIul.
17
In
Schmitt's entirely political analysis and critique oI liberalism, however, it is quite unclear why, with their political eIIicacy gone, anyone should cling to these norms.
There are, to be sure, hints in his analysis that the society-state opposition and even the constitution oI a public sphere are not identical to the issue oI parliamentarism,
indeed, that they actually predated it historically. He writes:
public opinion attained this absolute character Iirst in the eighteenth century, during the Enlightenment. The light oI the public is the light oI the Enlightenment, a liberation Irom
superstition, Ianaticism, and ambitious intrigue. In every system oI enlightened despotism, public opinion plays the role oI an absolute corrective.
18
This thesis, relatively unimportant in Schmitt's own work, was powerIully expanded by a historian he strongly inIluenced, Reinhart Koselleck, in his Kritik und Krise
(1959).
19
According to Koselleck, the absolutist state on the European continent, Iormed as a response to religious civil war, created the Ioundations Ior a political
dualism by Ireeing itselI Irom all norms in line with the doctrine oI raison d´etat.
20
The resulting separation oI politics and morals, as well as the increasing disinterest oI
the state (anticipated by Hobbes) in controlling private, individual conscience, created a

Page 207
possible Ioothold Ior the constitution oI a new Iormation, "society," Iirst apart Irom and later against the state. The old regime, oI course, never created a completely
monistic, statized society: The older estates, now depoliticized, preserved their corporate existence. There were, moreover, new organizational Iorms oI an emerging,
bourgeois class composed oI the beneIiciaries oI the Iirst truly national economic policy in European history. Out oI these two strata, combining with elements oI
intellectual and judicial elites, came the social bases oI the enlightenment, one possessing money, social recognition, and intellectual inIluence, but not political power.
Nevertheless, the "society" oI the enlightenment was organized, with the private salon, the caIe, the club, the library, the Masonic lodge, and later the secret society as
its major Iorms. According to Koselleck, many oI these unpolitical Iorms oI assembly and association were, in Iact, protected by oIIicials oI the absolutist state.
21
In
spite oI such protection, they would take an antistate turn as the eighteenth century progressed.
The support oI enlightened state oIIicials is relatively easy to explain, since the new Iormation "society," as typiIied by the Masonic ideology, was egalitarian in its ethos
and opposed to the privileged society oI aristocratic and ecclesiastical orders, itselI the main enemy oI "enlightened absolutism." Moreover, society was not supposed
to be a threat to the state because its selI-understanding was moral rather than political. Precisely on the ground oI the absolutist understanding oI politics as raison
d´etat, moral virtue was deIined as Ireedom Irom politics. This expansion oI the Hobbesian Ireedom oI private, individual conscience was, however, no longer
compatible with the internal logic oI absolutist depoliticization. Following Schmitt, Koselleck implies that the unity oI the heterogeneous elements oI "society" could be
maintained only in opposition to the state. Indeed, as his analysis oI the Masonic movement shows, such an opposition was made possible by the Iact that weapons oI
the established powers were utilized, at least initially, Ior the selI-organization oI society: The secrecy oI the absolutist regime and the hierarchical organization oI the
social orders were the glue behind the ideology oI Iraternity and solidarity.
OI course, the enlightenment became both more public and more egalitarian as it became a broad movement. According to

Page 208
Koselleck, such a transIormation, leading to a polarization between society and state, was already implicit in the rigid juxtaposition oI morality and politics. The very
constitution oI a "society" based on morals represented a judgment over and a rejection oI absolutist sovereignty, without any visible attack on state institutions. The
rejection oI politics was at the same time the establishment oI a moral vantage point Ior criticizing and judging politics. The moral pressure emanating Irom "society,"
creating a whole system oI values alternative to the established ones, could not avoid being a source oI inIluence over action and thereIore becoming an indirect Iorm
oI political power. Morality was directly unpolitical, but exactly Ior this reason it could put an amoral state into question and thus become, aIter all, political, iI indirectly
so.
22
The radicalization oI the program oI society against the state postponed the appearance oI an entirely unpolitical program. In Koselleck's presentation, this program
went through the stages oI taking a distance Irom politics, critique, judgment, and execution. Since the absolutist state could not be eliminated, selI-limitation had to be
practiced. Initially this selI-limitation contained a component that was, because oI the overwhelming disparity oI power, merely strategic. But it also had a normatively
validated antipolitical component based on principles. The latter, however, was selI-negating to the extent that even an antipolitical morality had diIIiculty reconciling
itselI with immorality in the world oI politics. In the radical enlightenment, then, the moral sphere constituted itselI, in secret, as another, alternative, political one. The
aim oI this political society was no longer coexistence with the state but rather its dissolution and replacement. The methods oI education, schooling, propaganda, and
enlightenment were no longer adequate Ior the new purpose, and this implied that even strategic selI-limitation had to be seen as merely temporary.
In this way, Koselleck convincingly revives the idea oI an intrinsic connection between enlightenment and the crisis oI the old regime, and between this crisis and the
coming revolution. It is in this context that he seeks to locate the Schmittian topos oI the emergence oI the liberal public sphere, here representing the political turn oI
society in opposition to the state. Pierre Bayle's idea oI a republic oI letters, according to Koselleck the model Ior Rousseauian

Page 209
radical democracy, indicates what is at stake. On the one side, this "republic" is still to be based on the contrast between a powerless moral law and an amoral power.
On the other, this contrast is interpreted as the conIrontation oI the regne de la critique with the rule oI the state, indicating that critique, the weapon par excellence oI
the public sphere, has turned political.
This transIormation carried risks. Taking the point oI view oI the state, Koselleck argues that the idea oI critique, turning inward to society itselI, must Iail as a means oI
social integration and must ultimately lead to a reappearance in the private sphere oI the civil war suppressed by absolutism. Here the deeply apolitical potential oI the
liberal idea oI the public sphere, as in Schmitt's doctrine, shows itselI. At the same time, as long as the state as "enemy" exists, the critical, polemical contestation oI its
legitimacy provides the cohesion oI the "Iriend" component oI the polarity, the alternative political society. This contestation is carried out in the medium oI public
criticism. In the public realm, critique becomes the means oI ampliIying public opinion, exposing everything, destroying all taboos, and depriving its political enemies,
organized around the state, oI legitimacy and means oI cohesion.
23
EIIectively built to counter the criticism oI weapons, the absolutist state Iails against the weapon oI
criticism, which, because oI its supposedly unpolitical nature, disempowers a properly militant political response.
Because it is concerned with the rise oI the dichotomy oI state and society, Koselleck's analysis stresses the political dimension oI the liberal public sphere rather than
the potentially apolitical implications that, in Schmitt's conception, characterize the triumph oI society over the bureaucratic-military state. Nevertheless, these apolitical
potentialities appear in Koselleck's picture in the tendency oI the agents to hide the political dimension oI their actions not only Irom the state but also Irom themselves.
Paradoxically, it is this reIusal oI politics by political agents that leads not only to the dissolution oI the absolutist state but also to an inability to establish a new model
oI the political. Even beIore the collapse oI the old regime, by insisting on recognizing only its own moral motivation, critique Ialls prey to hypocrisy.
Koselleck's conception oI the hypocrisy oI enlightenment antipolitics adopts the point oI view oI the state itselI. The critique

Page 210
oI power and the attempt to limit it are unhesitatingly qualiIied as hypocritical, although the author does not make up his mind whether he seeks to indict the will to
power oI critical reason or its implicit drive toward civil war. This ambiguity can also be Iound in Schmitt. While Koselleck goes beyond Schmitt in discovering the
enlightenment roots oI liberal parliamentarism, in his own Schmittian analysis all we get is an anticipation oI the rise and decline oI the political public sphere in the logic
that leads to revolution. Indeed, it is diIIicult to connect this prehistory in France, where the collapse oI the old regime did not initially lead to a stable parliamentary
outcome, to the history oI parliamentarism as analyzed by Schmitt. The connection can be made only when one recognizes that the enlightenment dualism, with the
public sphere as its central mediation, was not merely a strategy Ior the disempowering oI the state by politically weak competitors with a relentless power drive, but
could also be institutionalized as a new political alternative.
24
Koselleck comes close to such a thesis only when he uncharacteristically uses Marxian arguments to bolster an essentially Schmittian position. For example, he argues
that the bourgeoisie constituted itselI as a new elite precisely through the dualistic Iigure oI thought. Yet even here the argument is that the dualistic conception, as
preparation Ior the taking oI power, served only to eliminate all dualisms. UnIortunately Ior Koselleck, neither the normative achievement oI the liberal public sphere
nor even its possible and eventual institutionalization can be thematized in such an argument. Both are, however, insisted upon by Jürgen Habermas, in an analysis in
many ways indebted to, yet quite distinct Irom, Koselleck's.
From a Literary to a Political Public Sphere:
1ürgen Habermas
The Schmittian thesis concerning the Ioundation oI parliamentarism in the diIIerentiation oI society and the state can be seen as a narrow version oI the Hegelian
conception. In particular, the problem oI mediation is reduced to a single component, the political public sphere, which is in turn presented in a normatively aggressive
Iashion entirely disinterested in public discussion as an end in itselI. Habermas's conception, on the other hand,

Page 211
attempts to go beyond this reduction in two respects: Iirst, by recapturing a richer set oI mediations between civil society and state, and second, by reemphasizing and
revalorizing the normative claims oI the public sphere. Habermas's analysis also takes up the Hegelian project oI bringing together the normative achievements oI both
the ancients and the moderns (and does so more successIully that Hannah Arendt's).
Habermas's original theory oI the public sphere, worked out in the intellectual milieu oI the older FrankIurt school, represents a species oI Jerfallsgeschichte, a
history oI decline. This similarity to Arendt's conception tends to disguise the entirely diIIerent relation oI the two schemes to history. As we have seen, Arendt's public
sphere, modeled on an idealized conception oI Greek or Athenian politics, is paradoxically said to decline with the rise oI modern society, state, and economy, even
though she admits that the original model had long since disappeared. Moreover, Arendt is not at all inhibited by her theory oI decline Irom postulating the repeated,
but always temporary, reemergence oI experiments in public Ireedom during modern revolutions. It is as iI Ireedom and unIreedom moved in two separate and only
occasionally connected temporalities; Ireedom, in other words, is always (but also only) possible whenever the dialectic oI history stands still.
25
Habermas, on the contrary, inserts the emergence and decline oI a new type oI public sphere into the history oI modern society. While Arendt associated only the
decline oI the public with the rise oI modern state and economy, in Habermas's conception the rise, contradictory institutionalization, and subsequent decline oI this
sphere are all related to this event. The new public sphere is accordingly seen as bourgeois, because in it independent owners oI property, divided in their competitive,
egoistic economic activities that have grown vastly beyond the limits oI the household, are capable oI generating, at least in principle, a collective will through the
medium oI rational, unconstrained communication. But it is also liberal, in that the sets oI rights deemed necessary to secure the autonomy oI this sphere (Ireedoms oI
speech, press, assembly, and communication), together with those dimensions oI individual autonomy that it presupposes ("privacy rights"), simultaneously constitute
the public and private domains oI civil society and serve

Page 212
as limits to the reach oI state power. Indeed, the new public sphere is also democratic in principle: The emergence oI a new Iorm oI uniIied, depersonalized,
bureaucratic public authority, the modern state, is to be checked, supervised, and even controlled not only by the rule oI law but also by a second political public
sphere (emerging within society and penetrating the state in the Iorm oI parliaments) that challenges raison d´etat as well as arcana imperii. The tendency oI the
modern state to level and dismantle all corporate and estate organizations oI a Iormerly divided sovereignty is countered by the emergence oI a diIIerent, normatively
grounded reason operating in the Iull view oI all concerned, within new societal institutions that come to penetrate the domain oI politics itselI.
26
Habermas's depiction oI the emergence oI the institutions oI a new type oI public liIe, polemically juxtaposed to both the absolutist state and the privileged society oI
orders, draws heavily on Koselleck's picture oI the organization oI the enlightenment. However, three dimensions oI Habermas's conception diIIer Irom his
predecessor's:
First, Habermas believes that the peculiar logic oI the new public is continuous with, and constitutes a projection oI, the Iorm oI interaction oI the new intimate sphere
oI the bourgeois Iamily, a sphere that Arendt considered to be the most characteristic product oI modernity.
Second, he distinguishes not only between the literary and political public spheresa distinction played down by Koselleck, who suspects hypocrisy in every
antipolitical claimbut also between the small-group interaction represented by the salon, the caIe, the table society, and the lodge and the extension and
generalization oI public discourse through the media oI communication, above all the press.
Finally, Habermas distinguishes among at least three national variants (English, French, and German) oI the institutionalization oI the political public sphere, in the
process showing the development oI common norms in the context oI a heterogeneous set oI political projects diIIicult to reduce to a single one, especially to the will
to power oI the weak.
We shall start with these three points and then turn to a more systematic analysis oI Habermas's conception.

Page 213
1. Habermas's depiction oI the Iorce Iield between individual and state, unlike that oI Schmitt, involves at least three levels oI mediation: Iamily, literary public, and
political public spheres. These levels are not identical to the corresponding Hegelian categories, and the choice changes the theoretical role oI "mediation." The
category oI Iamily has great importance in this context. In Hegel's scheme, the Iamily is the precondition oI bourgeois individuality, and as such it is prior to and outside
civil society Ior primarily logical reasons that are sociologically nonsensical under conditions oI modernity.
27
For Habermas, the early modern, small-scale, bourgeois,
patriarchal Iamily is not only (as Ior Hegel) the place oI origin oI bùrgerliche Gesellschaft. Nor is it even what it could be in a sociologically extended orthodox
Hegelian conception, namely, one oI the levels oI integration oI egoistic individuals into the culture oI the state. In Habermas's version, the intimate sphere oI the small-
scale bourgeois Iamily also represents the establishment oI a principle counterposed to those oI both the modern economy and the state. It is not that he neglects the
Hegelian idea that the Iamily represents the background oI socialization that is the condition oI possibility oI the existence oI individuals oI civil society; rather (and
more in the Arendtian sense), he converts this background Irom a point oI origin to an institution that continues to participate in social liIe and to which individuals can
continually return as their home. For this reason, the Iamily prevents the dissolution oI individuality on the various levels oI collectivity. Thus, as in Arendt's theory, it
represents a private sphere without which a public sphere based on autonomous individuals would not be possible. But whereas Arendt sees the complementarity oI
private and public as possible only because oI their radically diIIerent principles, conceived along the lines oI the ancient duality oI polis and oikos, Habermas uses the
Arendtian notion oI the intimate to generate a single principle Ior both, one that is normatively adequate to the modern ideal (though not the reality) oI the Iamily:
interaction Iree oI domination and oI external social constraint. This ideal, leading to a new conception oI humanity, is Iurther analyzed
28
into the components oI
voluntariness, emotional community, and cultivation: "It appears that the Iamily is established and maintained voluntarily by Iree individuals without

Page 214
constraint; that it is based on the lasting emotional community oI the partners; that it guarantees the development oI all capacities that signiIy a cultivated person as ends
in themselves.''
29
It is not hard to recognize speciIic versions oI the ideas oI liberty, solidarity, mutual recognition, and equality in such a conception oI humanity.
In line with the classical Marxian critique, Habermas is quick to point out the counterIactual character, and even more the legitimating Iunction, oI the ideal he depicts.
He stresses its clash with the real economic Iunctions oI the new Iamily type, as well as with its patriarchal Iorms oI subordination, both oI which also penetrate the
intellectual elaboration oI bourgeois utopias.
30
All the same, Iollowing a Iamous analysis by Horkheimer Irom 1936, Habermas maintains that the ideal is not mere
ideology. The new solidaristic norms that play a role in legitimating the arrangements oI a competitive and nonsolidary private economy are always in tension with what
is established, promising a this-worldly transcendence oI all states oI aIIairs incompatible with Ireedom, solidarity, and cultivation. Thus, these norms represent both
ideology and the Ioundations oI the critique oI ideology.
31
Moreover, the Iamily, although incapable oI eliminating the constraints oI the economic world or even oI
Ireeing itselI Irom its own patriarchal heritage, nevertheless deIends the intimate subjective experience and intersubjective ties oI its members, qua human beings, in the
Iace oI external powers. Equally important, it is the living source oI experiences oI passionate selI-examination and rational searching Ior mutual understanding that are
capable oI Iinding other Iorms oI institutionalization than the Iamily itselI.
32
Habermas argues Ior an empirical connection between the private world oI the bourgeois Iamily and the primordial Iorms oI the literary public sphere. While the salon
admittedly originates in aristocratic society, the bourgeois salon loses its representational and ritualistic Iunctions: Its Iorm oI communication is no longer dramaturgical
and rhetorical; its social structure no longer reIlects the hierarchy oI a society oI orders.
33
Architecturally and socially linked to the private living quarters oI the Iamily,
the new salon extends and enlarges the original principle oI intimacy by revealing the subjectivity oI each individual in the presence oI the other, thus

Page 215
linking privacy to publicity. The ideal oI seeking understanding through open-ended reasoning and mutual persuasion, without regard Ior prestige and status, is
maintained. Somewhat more distantly, Habermas sees the institutions oI club, caIe, and lodge as extensions oI the same principle. He does explicitly note, however,
the exclusion oI women Irom these latter institutions oI enlightenment, linking this exclusion to the discussion oI political and economic rather than primarily literary and
artistic matters.
34
Yet the connection oI the Iirst institutions oI an audience Ior works oI art, and especially oI literary and reading circles, to salons dominated by
women remains close, and it is through these agencies that the reasoning public modeled on the intimate Iamily Iirst begins to approach universal signiIicance.
35
This
connection to the reception oI art also develops a dimension oI the literary public that is present in the new intimate sphere only in the Iorm oI selI-reIlection and selI-
examination: the critique oI all received ideas and meanings.
2. While Koselleck tends to Iocus on those enlightenment institutions, Irom the lodge to the secret society, that paradoxically seek to establish the principle oI publicity
by negating it, and Ior which critique eventually became a means rather than an end in itselI, Habermas's own stress is on institutions whose road to politics, slower and
less complete, implies neither a compromise oI Iundamental principles nor a merely hypocritical renunciation oI power. The public sphere in his conception comes into
being not through the politicization oI small-scale Iace-to-Iace intimate interaction but through the establishment oI a critical audience Ior literary works by means oI
newspapers, journals, and public perIormances. Only this road allows the conversion oI the principles oI intimacy into those oI a critical publicity. But even on this
longer road, the literary public grows into politics, into a political public sphere with a structure diIIerent Irom those oI political organizations dedicated to the pursuit oI
power. Even iI both roads were in Iact divorced Irom the more Ieminine world oI the salon, the political public sphere maintained something oI its spirit in the idea oI
critique as an end in itselI.
It is Habermas's thesis that the emergence oI a political public sphere Irom the critical literary one preserves the principle oI

Page 216
unconstrained communication originally established in the intimate sphere oI the new Iamily type. Unlike Koselleck, who points to a project oI a counterpower
hypocritically aiming at destroying and replacing established power, he insists that what is at stake is the transIormation oI the principle according to which power, old
or new, is to operate.
36
Critique in this model attempts its own institutionalization rather than a conversion into a new Iorm oI power that would potentially Ieel itselI
endangered by critical reason. Even in Habermas's analysis, the modern state, in its originally absolutist Iorm, represents the challenge motivating the establishment oI a
veritable countersociety, a society against the state. But this society, even when it turns political, aims neither at the utopian destruction oI the state nor at becoming a
new state, nor even at the uniIication oI these aims as in the Reign oI Terror, but rather at a new Iorm oI political dualism in which a political public sphere would
control the public authority oI the modern state.
The argument goes against the grain not only oI Koselleck's Schmittian analysis but also oI the Marxian conception oI the bourgeois revolution. Nevertheless,
Habermas hopes to save something oI the latter by insisting that the bourgeoisie, whose power is by deIinition private, cannot rule and yet cannot accept a Iorm oI
state that is potentially arbitrary and uncontrolled. A Iurther complication: This same class, unlike the aristocratic opponents oI absolutism, needs and wants a Iorm oI
uniIied sovereign power capable oI guaranteeing the political and legal preconditions oI a private capitalist market economy within and even beyond a national
territorial setting. The historical solution was to preserve the modern state created by absolutism, but to Iormalize and rationalize its operation in terms oI the rule oI
law, to Iorce it to establish Iorms oI selI-restraint as deIined by Iundamental rights, and to bring it under social scrutiny and control through the establishment oI the
political public sphere, itselI rooted in the rights oI communication and Iranchise. It is these normative limitations that Habermas has in mind when he reIers to changing
the principle oI the operation oI power.
3. It is not clear that the suggestive ideal type can save the thesis oI a bourgeois revolution. In France, where a revolution did occur,

Page 217
one that was hardly bourgeois,
37
the pattern outlined by Habermas was originally established only transitionally, during the constitutional monarchy. Moreover, given
the outcome, it is not diIIicult to argue that the Iorms oI public liIe Habermas describes as extending Koselleck's analysis into the revolutionary period (journals,
pamphlets, clubs, popular assemblies) represented the projects oI counterelites hoping to replace the existing elite (and, soon enough, one another). To show that an
alternative principle could have been established, Habermas is Iorced to shiIt his emphasis Irom Koselleck's terrain oI French politics, culminating in revolution and
terror (the context oI choice oI conservative opponents oI the liberal idea oI politics), to the English context oI the evolutionarv transIormation oI parliamentary
absolutism. This "model" is in turn used as the standard Ior evaluating the constitutional monarchies oI the early liberal epoch. From its point oI view, French
developments in the period oI high absolutism appear incredibly retarded. They seem to Iollow a slower but Iundamentally English path during most oI the eighteenth
century, when pole##ics against the absolutist regime Irom the points oI view oI the traditional estates and the new public Iorms were not always easily
distinguishable.
38
The revolutionary period enIorced this distinction dramatically in a process oI tremendously accelerated creation oI public political Iorms (the
transIormation oI the estates assembly into a modern parliament, the creation oI journals, clubs, associations, and assemblies, and, above all, the institution oI Iormal
constitutional guarantees Ior all oI these). The revolutionary dictatorship and Napoleon destroyed the institutions oI the political public sphere, however, and France
paradoxically (and still inconsistently and with many reversals) reentered the basic model oI liberal development only with the Restoration. Accordingly, in this
depiction oI French developments Irom the point oI view oI the liberal public sphere, the accelerating revolution turns out to be a parenthesis. From the same
perspective, developments in the Germanies through various models oI the authoritarian Rechtsstaat appear simply as slower and perhaps never entirely completed
versions oI the English model.
The choice oI England to outline an actual historical path that is somehow adequate Irom the point oI view oI the normative con-

Page 218
struct oI the liberal public sphere helps to dispel the doubt raised by Schmitt that the parliamentary state as a Iorm oI selI-organization oI society breaks down at the
moment oI its realization. Against this objection, Habermas is able to show the institutionalization oI dualism in terms oI parliament and a political public sphere. The
same choice, however, is still potentially exposed to Koselleck's critique, which might Iocus in the English context on the hypocritically bourgeois rather than the
hypocritically statist character oI the liberal public sphere. In other words, in the English case the project Ior liberal publicity seems to have been a cover Ior the will to
power oI the propertied classes. The charge is not as strong as it might at Iirst seem, though, because the parliamentary absolutism that emerged Irom the Glorious
Revolution was already Iully compatible with the economic interests and political representation oI the propertied classes. The struggle Ior a political public sphere and
Ior the rights oI speech, press, assembly, association, and the Iranchise that would sustain it was not restricted to the owners oI bourgeois property, nor did it stop with
the Iull political victory oI their program in the New Poor Law. While it is possible to argue that the outcome oI these struggles helped make parliamentary rule
legitimate and thus stabilized bourgeois domination, this legitimacy was nevertheless a Iunction oI new Iorms oI protection, selI-organization, and public oversight
achieved by social strata whose traditional Iorms oI liIe were undermined by the transition Irom a paternalistic, moral economy to the selI-regulating system oI liberal
markets.
39
English absolutism does not end in Habermas's picture with the demotion oI the monarch to "King in Parliament" but with the new relation between public sphere and
state expressed in the Iull publicity oI the proceedings oI Parliament.
40
However, when publicity, originally a weapon, becomes a principle linked to the normative
experience oI everyone capable oI reasoning, it cannot be restricted either institutionally (to the press and the parties) or socially (to the middle classes).
41
The
growing public thematization oI Iundamental political questions leads to the organization oI political meetings, clubs, associations, and committees
42
that in turn provide
Iorms Ior the selI-organization oI strata that are not Iormally included in the political system until the end oI the

Page 219
century. Democratization does not in itselI, as Habermas elsewhere unIortunately suggests,
43
lead to the decline oI the critical capacities oI the public: It is, in Iact,
aIter the First ReIorm Bill, when the parties must appeal to a socially much more heterogeneous electoral public than beIore, that they are Iorced to publicize their
electoral programs and to discuss them in terms oI arguments and principles rather than slogans, personalities, or even narrow sectoral interests.
44
Habermas's linking oI his study oI the development oI the liberal-bourgeois public sphere to a speciIic historical pattern oI development should not lead us to neglect
his theoretical model oI this sphere, however ideal-typical or even composite it may appear. This is all the more important because he insists that it is this abstract
model, rather than any particular historical version, that attained normative and even utopian status Ior modern society. Broadly speaking (in the tradition oI Hegel),
Habermas not only diIIerentiates between civil society and state but also relativizes the traditional distinction oI private and public with which the liberals and Marx
identiIied the new polarity. He does this by dividing each sphere, public and private, into two:
private: intimate sphere (Iamily) private economy
public: public sphere public authority (state)
We expect one speciIic role to correspond to each oI the spheres, though Habermas makes the point clear only in the case oI the private sphere:
45
private: human being ("homme") bourgeois
public: |citizen| |subject|
Habermas recognizes that the relationship oI this IourIold categorial Iramework to the concept oI civil society, or bùrgerliche Gesellschaft, is ambiguous.
46
In the
narrow sense (that oI Marx), bùrgerliche society reIers to the sphere oI the private, bourgeois economy. When used in this sense, the public sphere is to be
understood as a mediation between society and state. However, in the broader sense (that oI Hegel), the term civil societv means all the spheres oI society
juxtaposed to the state.
47
In that case, it will include the

Page 220
public sphere as well as the domestic one, and thus it will have three Iundamental roles (oI which Habermas stresses only the Iirst two): human being, bourgeois, and
citizen.
II Habermas does not consistently adopt this second, more Hegelian, usage, it is because he seems to be sensitive to a Iictional identiIication criticized by the young
Marx: that between "l´homme" and "bourgeois."
48
He, too, considers this identiIication to mask the bourgeois character oI the new public sphere and an ideology
that subordinates the sphere oI the citizen to the imperatives oI the private economy. As a result, and in order to provide an analytical contrast to liberal ideology,
Habermas reIuses to make the category oI the public sphere simply an internal selI-determination or mediation oI civil society.
He does not thereby manage to Iind an adequate locus, even in principle, Ior the activity oI the citizen. His desire to diIIerentiate spheres stops short exactly at this
category. He is, however, on his way to doing this when he points to a second Iiction in liberal ideology: the identiIication oI literary and political publics as a uniIied
public opinion. UnIortunately he tends to regard this identiIication only as the vehicle by which the Iirst Iiction, the identity between man and bourgeois, claims
normative superiority over the citizen. Thus, he seems not to see the necessity in this case Ior yet another analytical diIIerentiation oI what ideology misleadingly
identiIies: man and citizen. This omission seems to concede the liberal point that subordinates the normative source oI the status oI the citizen in the modern world to
the norm oI the new conception oI humanity, even iI not in its bourgeois version.
The basic model is at times diIIerentiated as iI Habermas wanted to avoid both Iictional identiIications:
49
private: intimate

private economy
public: literary political state
This scheme corresponds to the historical development oI the political public sphere, which may have emerged Irom the literary public sphere but can Iully replace or
subsume it only at its peril. "The humanity oI the literary public," he says obliquely enough, "serves as a mediation Ior the eIIectiveness oI the political public."
50
On the
other side, however, the argument presupposes that a

Page 221
literary-cultural public sphere cannot itselI control or directly inIluence the modern state. Habermas stresses the diIIerentiation oI the two publics in terms oI two
audiences that draw on diIIerent sources Ior their members, one mainly women, the other exclusively men.
51
All this would seem to point to a diIIerentiation in the
tradition oI Tocqueville between civil and political society, corresponding to Habermas's own diIIerentiation oI two publics (literary and political) and two roles (human
being and citizen). It is just this diIIerentiation, however, involving sharper boundaries between political and prepolitical public spheres, that Habermas wants to avoid.
To the extent that the two publics have important continuities and even Iormal similarities, Habermas is right. But another motive is at work here as well, one that
produces a certain overreaction. In order to preserve the modernity oI his conception as against Arendt's stylization oI the ancient notion oI citizenship, Habermas
wants to break deIinitively with the old meaning oI societas civilis that contained the level oI political society. Instead oI choosing a strategy oI diIIerentiation,
however, he abandons the latter notion. In his conception, all that is leIt oI political society is the political public sphere as a projection oI the literary public into areas
dealing with questions oI economic policy.
Habermas quite deliberately constructed his model oI the public sphere in the structural position that Arendt considered the very negation oI public liIe, the
intermediary or mixed realm between private sphere and state, which she called "society."
52
Though he admits that the ideological inspiration oI the Greek model
continues into our own time, Habermas consistently disputes its institutional relevance. Unlike Arendt, he has no use Ior a concept oI political society, admittedly still a
component oI the eighteenth-century conception oI societe civile or Zivilso:ietàt, that would somehow preserve what is essential about the ancient republican idea oI
citizenship. This idea Habermas understands as membership in an incorporated, genuinely political body, the res publica, that collectively acted to guarantee justice
and military security. The "political" task oI the bourgeois public sphere is, on the contrary, the regulation oI bùrgerliche Gesellschaft in the sense oI securing the
exchange oI commodities in the market.
54
Thus, Habermas seems to make the assumption oI the tasks oI the oikos the Iunctional deIinition oI the new bourgeois public sphere;

Page 222
this is what Arendt considered the basis oI the decline oI publicity as such. But it is the liberal as well as the bourgeois dimension oI the modern public sphere that sets
it apart Irom the ancient notion oI citizenship. Contrary to the Greek model, the modern public sphere is juridically private. Legally separated Irom the state, this sphere
and its members have a polemical, critical, argumentative relation to the state rather than a participatory one. They can supervise, inIluence, and perhaps somehow
"control" power, but they cannot themselves possess a part oI state power.
In spite oI some serious inconsistencies, Habermas's model oI the political public sphere does not reIer primarily, as does Schmitt's, to the parliamentary deliberative
body itselI, whose members do in Iact have public law status. The importance oI parliamentary deliberations is established only with their publicity, and this is what
makes this Iorm oI rule uniquely permeable to the gaze oI a public composed oI private individuals. II the parliamentary deputies are part oI the political public sphere,
this is because oI their continuity with the society oI private, reasoning individuals who compose that sphere. The point is somewhat lost when Habermas argues that
public opinion came to regard itselI as the only legitimate source oI law.
55
But he interprets this claim in terms oI the contrast between the rule oI law and rule by men,
with society supposedly achieving a condition beyond all domination through a transIormation oI the Iorm oI law (generality) and the Iorm oI lawmaking (publicity).
Thus, Habermas argues that the political public sphere "puts pouvoir as such up Ior debate."
56
This argument seems to conIlict with the dualistic conception according to which the public sphere is to coexist with the modern state, whose principle oI operation,
but not whose existence, is to be placed in question. Habermas is, oI course, well aware oI the resistance oI "public" administration and other organs oI executive
power to the principle oI publicity.
57
But he Iollows the internal logic oI the liberal conception oI the public sphere to such a point that the only Iorm oI eIIective social
control oI the state that seems to be logically possible is its abolition. Rightly rejecting the ancient notion oI citizenship proposed by Arendt, Habermas was not able to
point, at least within the tradition he reconstructed, to a modern, intermediary model. In short, the liberal model oI the literary public sphere, with its overarching norms
oI humanity and critical

Page 223
reason, tends, once "politicized," to point not to participation within, but to the abolition oI, state power, indeed, oI power tout court, and its replacement by a closed
system oI legal norms.
Interestingly enough, in view oI Habermas's deep analysis oI Hegel, he does not use the latter's conception oI a plurality oI associations within the private sphere that
might prepare the participation oI citizens. In his critique oI Tocqueville, too, there is little interest in or sensitivity to the prepolitical dimension oI small-scale selI-
organization required Ior the eIIective and democratic limitation oI democratic sovereignty.
58
Undoubtedly, these levels oI analysis stressing the need Ior intermediate
powers did not appear to complement his own analysis oI mediation through the public sphere. Probably they seemed to point to irrelevant atavisms or to anticipate
the corpora Iist deIormation oI publicity itselI. But it remains the case that his identiIication oI the prepolitical dimension oI the public sphere with a literary public,
although essential as a legitimating background, involving a certain reduction vis-a-vis Hegel's classical model, renders the political public sphere much too weak in the
Iace oI state power. Habermas is aware oI this weakness but not oI all the causes or oI the available alternatives. Thus, he is Iorced to register rather passively that the
"person" oI the political public sphere turns out aIter all to be the "homme" oI the literary extension oI the intimate sphere; he is able to propose no concept oI the
political to counteract the "characteristic erosion oI the boundaries oI the two publics"
59
that was the very object oI Schmitt's savage criticism oI liberal ideals.
60
Habermas considers the diIIiculty to be a Iunction not oI the norma Iive project but oI the contradictory institutionalization oI the public sphere. It is thus the speciIic
Iorm oI the institutionalization oI the new norm oI "humanity" that proves to be powerless to block the triumph oI the bourgeois and the oIIicial. From this critical
juxtaposition oI norm and institution, Habermas cannot, however, derive the philosophical Ioundations Ior an alternative institutionalization. In relation to the capitalist
economy and the modern state, the value oI humanity, unlike that oI citizenship, is bound to remain in question.
The contradictory institutionalization oI the public sphere is already apparent at the level oI its original model, the intimate sphere. Habermas describes it in terms oI the
ambivalence oI the

Page 224
Iamily, which is ''the representative oI society, and yet is in a certain way emancipation Irom society and against society, held together by patriarchal domination on the
one side and human intimacy on the other."
61
More precisely, the compulsion Iaced by the bourgeois Iamily is a Iunction oI its speciIic role in the process oI the
'valorization" oI capital and oI the transmission oI legal-political constraints through socialization. Habermas, still presupposing the doctrine oI state and law as
superstructure, unIortunately treats these two dimensions as Iunctionally identical. In this conception, patriarchal authority, expressed in the subordination oI women
and children, is a transmission belt Ior economic and political powers that then deIorm the components oI humanity: Autonomy, emotional community, and cultivation
are subordinated to money through the instrumentalities oI power.
It is an open question whether the ideals oI the liberal-bourgeois public sphere are themselves deIormed by patriarchal authority, or whether the deIormation occurs
when the state and the capitalist economy manage to impose their logic on the political public sphere. Habermas seems to choose the second oI these options, though
at times he also says that the ideology reIlects the ambivalence. This choice may be a signiIicant mistake, however, since the notions oI "homme," emerging Irom the
Iemale-dominated salons emphasized by Habermas, and "citoven," Iorged in the male-dominated secret societies emphasized by Koselleck, seem to represent
opposite sides oI the same deIormation in the political realm: the powerless human being and the inhuman citizen.
The contradictory institutionalization oI the public sphere, and in particular oI its political dimension, parallels the ambivalence oI the intimate sphere. Habermas
explores the contradiction Irom the point oI view oI the bourgeois Iunction and then Irom that oI the liberal structure oI the political public. The Iormer is linked to the
restricted or narrow concept oI civil society inherited Irom Marx, representing the market-oriented interaction oI private economic subjects Ireed (in two stages oI
developmentabsolutist and liberal) Irom estate hierarchy and state paternalism. In this materialist-Iunctionalist train oI argument, the task oI the political public sphere
is to mediate between civil, or rather bourgeois, society and "the state power corresponding to its needs." First and

Page 225
Ioremost, the task oI this state is to work out, administer, and protect a system oI private law establishing, through the laws oI property, contract, employment, and
inheritance, a private sphere in the strict sense.
62
Paradoxically, then, the task oI state intervention is to Iree civil society oI this intervention, to diIIerentiate and maintain the diIIerentiation oI state and civil society. This
paradox appears on the level oI the laws establishing the mediating institutions oI the public sphere. The linkage oI state action in the Rechtsstaat or state governed by
law to general norms and the publicity oI the making and application oI law provide not only Ior the selI-limitation oI sovereign power but also Ior the illusion oI its
disappearance. This illusion, in the present argument, is traced back to the interaction oI small, relatively equal owners oI property who imagine that the rules oI the
sphere oI competition make impossible the ascendancy oI one owner over another. These agents desire no political rule in their aIIairs, exercised by a state or even by
themselves, yet they require legislative provisions Ior their activity. The political public sphere was to be the solution oI the diIIiculty, implying the production oI
measures rooted exclusively in reason rather than will.
Aside Irom the conIlicts with arbitrary power, involving the exertion oI will rather than rational persuasion and surviving in the resistance oI the executive and its
administration to supervision by the public sphere, the division between will and reason in the concept oI law could not be removed Irom the political public sphere
itselI. On the one hand, this institution could be regarded as the Ioundation oI the rationality oI law, since it links legislatures to the ongoing critical discussion oI a
reasoning public. On the other hand, the laws emerging Irom such processes oI communication had to maintain their coercive aspect in relation to those to whom they
were applied.
63
The rule oI law thus turns out to involve not the abolition oI rule as such but the institution oI rule by the legislature. The liberal bourgeois idea oI
abolishing the state, replacing it as the agency oI rule by a system oI gapless norms validated by the public sphere alone, turned out to be incoherent and impossible to
realize.
Formally speaking, the liberal idea oI the public sphere reIers not to bourgeois society but to a wider conception oI civil society that

Page 226
establishes, on the level oI constitutional rights, not merely an economic society but the public sphere itselI Ireed Irom arbitrary state intervention. Habermas presents a
classical catalogue oI Iundamental rights to indicate the centrality oI the deIense oI the public sphere (Ireedoms oI speech, opinion, press, assembly, association, etc.)
and the intimate sphere (inviolability oI person and residence, etc.). Constitutions also guarantee the rights oI individuals to engage in political activity in the public
sphere (rights oI petition and suIIrage, etc.) and economic activity in the private sphere (equality beIore the law, right oI property, etc.).
64
Finally, by establishing the
centrality oI the public sphere in political processes, constitutions go beyond the level oI the rights oI private individuals; in particular, constitutional guarantees oI the
publicity oI procedures are meant to establish the "inIluence" oI the public over parliamentary discussions and the "supervision" by the public oI the courts.
According to Habermas, the model oI civil society implied by this classical version oI constitutionalism "does not correspond at all to the reality oI civil society."
65

There are two reasons Ior this. First, the number oI private individuals who possess the autonomy secured by property and the cultivation guaranteed by education is
small. Indeed, a second minority, the traditional classes rooted in land ownership, the army, and the administration, still holds signiIicant power. Second, bourgeois-
liberal constitutions do not provide Ior those who do not possess the resources Ior participating in the literary and political public spheres, nor do they guard against
those who can generate and utilize power in secret. Again, the dimension oI domination reappears: that oI the public sphere over those excluded Irom the practice oI
rights and that oI those capable oI excluding themselves Irom the duties required oI the rest oI society.
All the same, it is not Habermas's intention to interpret the liberal dimension oI the public sphere as merely an instrument Ior exclusion. "The bourgeois public sphere
stands or Ialls with the principle oI general accessibility. A public sphere Irom which deIinable political groups are eo ipso excluded is not only imperIect but is not
public at all."
66
Habermas does not maintain that the bourgeois public sphere was mere deception. Though it has class

Page 227
interest at its Ioundations, there is also some overlap with general interests.
67
Leaving aside this dogmatic, traditional Iormulation, the point seems to be that the
boundaries oI exclusion could not be Iixed because oI the norm oI publicity itselI. In other words, this norm, established through constitutional and legal guarantees and
practiced in processes oI critical discourse, made the boundaries oI the public sphere permeable to themes and persons representing the interests oI those excluded.
The public sphere was an ideology, but because it contained a utopian promise, it was more than mere ideology.
68
This point is then reIormulated in two ways. First,
the idea oI publicity, "in principle opposed to all domination, helped to Iound a political order whose social bases did not make domination aIter all superIluous." This
Iormulation juxtaposes an idea linked to liberation to institutions establishing a new Iorm oI domination. Second, the ideology led, on the basis oI the domination oI one
class over the other, nevertheless to the development oI institutions "Which contained, as their objective meaning, the ideal oI their own abolition." This second
Iormulation implies that something oI the liberating ideal oI publicity was indeed institutionalized in the bourgeois public sphere.
The notion oI the contradictory institutionalization oI the liberal public sphere points to a direction consistent with the second reading. But the idea that the contradiction
is to be resolved, in accordance with normative requirements, by abolishing the whole institutional complex supports the Iirst. In Iact, several points remain unclear in
the analysis. First, as we asked beIore, are the normative expressions oI the principle oI publicity Iree oI the contradiction oI its institutionalization? Second, what
would be the Iorm oI a noncontradictory institutionalization oI either the original ideal or its reconstructed version?
The diIIiculties Habermas encounters in answering these questions have to do with the inIluence oI both Marxian and liberal utopias on his construction. He attempts to
hold the two strands together through the notion oI immanent criticism. Accordingly, he claims that Marx not only unmasked public opinion as Ialse consciousness but
did this in the name oI a staunchly held ideal oI a liberal public sphere.
69
Habermas's argument cannot succeed, however, to the extent that the Marxian critique
always involves

Page 228
both immanent and transcendent elements. II Marx does indeed want to maintain in a radicalized version the ideal oI politics based on democratic communication and
decision making, he nonetheless rejects the ideal oI diIIerentiation between public and private, between state and civil society, that this politics presupposes.
70
One
obviously cannot deIend the ideal oI a liberal public sphere without the model oI diIIerentiation, which has normative implications oI its own expressed in catalogues oI
Iundamental rights. Marx, however, considers diIIerentiation to be the secret oI deIormation to the extent that a diIIerentiated civil society, in the sense oI the private
economy, avoids thereby the possibility oI public control and oversight, a process that inevitably turns the modern citoven into the instrument oI the bourgeois who
disguises himselI as homme. This line oI analysis accordingly leads to the establishment oI a dediIIerentiated state-society by a revolutionary class that has no interest in
diIIerentiation. The strategy points to a new normative model oI individuality as well: Instead oI the Iictional identity oI man and bourgeois, Marx, according to
Habermas, posits the real identity oI man and citizen.
71
This goal seems to be accepted by Habermas himselI.
72
The transcendent Ieatures oI the Marxian critique to the contrary, however, Habermas himselI staunchly deIended the liberal idea oI the public sphere. Thus, while he
did not reject the Marxian project oI dediIIerentiation, he put another, one oI rediIIerentiation, by its side. This he achieved through an immanent critique oI his own.
From the point oI view oI the model oI diIIerentiation, he implicitly charges the bourgeois public sphere with being insuIIiciently diIIerentiated. In particular, the Iictional
identity oI bourgeois and man expresses the very real penetration oI the intimate sphere by the processes oI the private economy. Hence, the true aim oI the publicly
controlled state-society-economy is to Iree the intimate sphere Irom economic constraint and social intervention.
73
This argument, attributed to Engels, in Habermas's
version amounts to a project to establish a new Iorm oI private autonomy.
74
What Habermas does not tell us is how such private autonomy could be institutionalized without rights, though it is certainly possible that he simply presupposes some
version oI the classical catalogue. But iI we are to return to such a catalogue oI rights, how

Page 229
are we to avoid reaIIirming the normative model oI comprehensive diIIerentiation that these rights guarantee through their very Iorm? Habermas could have perhaps
countered this argument by reIerring to the need Ior redeIining the inherited catalogues oI rights, and especially their internal hierarchy. His notion, ascribed to Marx,
that autonomy in the new model would be based on the public sphere rather than private property points in this direction.
75
But here the dangers oI an overall model
based on uniIication rather than diIIerentiation show themselves; what would have been an important insight in the context oI a theory oI rights, undeveloped here,
becomes a perilous one within the actually aIIirmed project oI nonliberal democracy:
private autonomy is a product oI an original autonomy, which is brought about by the collectivity oI social-citizens exercising the Iunctions oI the public sphere expanded in a
socialist manner. It is private individuals who are regarded as the private individuals of the public, rather than the public as the public of private individuals. In the place oI
the identity oI bourgeois and homme . . . steps the identity oI citoven and homme. The Ireedom oI the private individual will be deIined according to the role oI human beings as
social citizens (Gesellschaftsbùrger); the Ireedom oI human beings as property owners will no longer deIine the role oI the citizen oI the state (Staatsbùrger).
76
It is obvious Irom this text that Habermas represents a position, without the slightest trace oI criticism, that explicitly breaks with the bourgeois-liberal ideal oI the
public sphere. The point is not only that one Iunctionalization oI the intimate sphere is replaced by the project oI another. More generally, the model replaces bourgeois
dediIIerentiation, which violates the constitutional norms oI the liberal public sphere in Habermas's own argument, with a scheme oI reverse dediIIerentiation that would
be equally incompatible with these norms iI they were maintained or reestablished. Although a case could be made that the project here outlined continues the
democratic dimension oI the normative model oI the public sphere, it most assuredly breaks with its equally important and constitutive liberal dimension. That
Habermas was, in 1962 at least, insensitive to such an outcome is shown by his treatment oI the "liberal" thinkers J. S. Mill and Tocqueville.
77


Page 230
Habermas is certainly right in using Marx to criticize the model oI the bourgeois public sphere, its tension between norm and institutionalization. Much more
questionable is his obvious preIerence Ior Marx over Mill and Tocqueville in the Iurther development oI the normative model. Arguing Irom the point oI view oI
Marxian radical democracy, Ior example, he has no use Ior Mill's concern, consistently on the ground oI diIIerentiation, to deIend private autonomy and the Ireedom oI
minorities Irom the greatest democratic power, the power oI public opinion. Inexplicably, he takes this idea, in reality a precondition Ior the rationality oI public
deliberation, to be a diminution oI the public sphere itselI.
78
Moreover, he does not seem to understand that the idea oI the public as the abolition oI political power
involves a renunciation oI the need to limit all power through the only means possible, the establishment oI counterpowers and organizations, and he is thereIore
powerless against the increasing power oI the modern bureaucratic state. From the point oI view oI a strategy oI democratic dediIIerentiation, Iinally, Habermas has no
sympathy Ior Tocqueville's stress on voluntary associations as the intermediary bodies required Ior the stabilization oI diIIerentiation and the establishment oI
democratic mediation. He does not realize that this model, required Ior the preparation oI citizenship on levels where participation is still possible in modern societies,
involves a potential relationship between homme and citoven that escapes the invidious alternative oI powerless human being and inhuman citizen. The associations oI
civil society in Tocqueville's theory prepare private individuals Ior the exercise oI public power, a task that the literary public sphere is, on its own, incapable oI
perIorming. At the same time, these associations preserve the connection oI citizens to the prepolitical social networks that serve as their background.
79
In place oI the
Marxian identity oI man and citizen, Tocqueville thus proposes a diIIerentiated and interdependent model oI social being and citizen.
Admittedly, Mill and Tocqueville are only partially concerned with the dediIIerentiating implications oI the link oI man and bourgeois. Habermas is right to appeal to
Marx when he seeks to expand processes oI public criticism and supervision to the economic sphere.
80
It is not clear, however, whether the ideal proposed

Page 231
involves an abolition oI the economy in the way that the liberal utopia (judged incoherent and impossible by Habermas himselI) seeks to abolish political power as
such, replacing it with public discussion. An alternative would have been to aIIirm the diIIerentiation oI the economic realm and oI its speciIic roles and to postulate
new Iorms oI complementarity and interdependence between economic actors, private individuals, members oI associations, and participants in the public sphere.
OI course,Habermas would have considered irrelevant the combination oI Marxian critique and democratic liberal norms proposed here on the basis oI some strains
oI his early work, because he believed that neither the Marxian nor the liberal utopias were adequate guides Ior exploring what occurred in the liberal public sphere. In
his analysis, neither option depicted hereMarxian, liberal, or even their combinationwas actualized. Instead, the liberal-bourgeois public underwent a change oI
structure entirely incompatible with its original normative project. Tocqueville and Marx would both have understood the cause Ior this Iundamental change, namely the
dramatic expansion oI the extent and power oI the modern administrative state, which has continuously resisted invasion by public processes and procedures. What
neither Tocqueville nor Marx could have imagined was that, apart Irom the socialist state-society that one Ieared and the other Iervently desired, a comprehensive
repoliticization oI society could occur, supposedly removing the Iield oI Iorce in which the bourgeois public sphere was constituted and apparently abolishing the
diIIerentiation oI civil society and state Ior which publicity served as a stabilizing mediation. It was Carl Schmitt who was the Iirst to work out a comprehensive theory
oI the decline oI the public sphere in terms oI the alleged Iusion oI society and state.
The Fusion of Civil Society and State:
Carl Schmitt
The shiIt oI the locus oI genuine publicity Irom the state (the model oI antiquity) to an independently organized and juridically private societal sphere does not in itselI
avoid the thesis oI Iusion and decline. As already indicated, Carl Schmitt developed his interpretation oI parliamentarism around this transmutation oI the con-

Page 232
cept oI publicity. It is thus all the more striking that he was the Iirst important thinker to link the end oI the liberal era to the reIusion oI society and the statea process
that supposedly eliminated the only sphere capable oI sustaining the claims oI publicity under modern conditions. Accordingly, parliamentary discussion, and with it the
"whole system" oI the protection oI social communication, has become today an empty Iormality.
81
Parliament is now nothing but an antechamber to the real loci oI
power: the bureaus or committees oI invisible rulers.
82
The parliamentary stage has been transIormed Irom a Schauplat: Ior "the Iree deliberation oI independent
representatives seeking unity" into an arena where the "plurality oI divided yet highly organized social Iorces" meet and clash.
83
In the process, all the old claims Ior
publicity have collapsed.
For a complex set oI reasons, it is democracy, or rather democratization, which Schmitt takes to be the Iundamental tendency oI the modern era, that is responsible
Ior the crisis oI parliament and its legitimacy. To begin with, he argues that democracy and liberal parliamentarism have entirely diIIerent principles. Democracy is a
Iorm oI rule resting on social (in modernity: national) homogeneity and "iI the need arises the elimination and eradication oI heterogeneity." Given the actual and
structural diIIerence between rulers and ruled, democracy is possible only when, on the ground oI homogeneity, the ruled can "identiIy" with the rulers. Starting with the
Rousseauian idea according to which democracy is the actual identity oI those who command and those who obey,
84
Schmitt winds up reducing this to a string oI
identiIications that rest on no "palpable reality . . . something actually equal legally, politically, sociologically" but only the ''recognition oI identity."
85
Moreover, given
suIIicient identiIication, dictatorship, especially iI supported by pedagogic claims, is compatible with democracy in this view; indeed, Schmitt believes that radical
democracy must lead to dictatorship because oI the inevitable lack oI preparation oI the masses Ior selI-rule.
Schmitt argues that liberalism is quite diIIerent Irom democracy. Above all, it is a deeply unpolitical model in that it rests on discussion rather than identiIication,
presupposing a corresponding plurality oI opinions rather than their homogeneity. Schmitt

Page 233
does not consider Ior a moment the possibility that the structural connection oI public opinion and parliamentary publicity establishes a medium oI genuine iI incomplete
identity between rulers and ruled. Democracy Ior him is based not on actual though incomplete institutional identity but on complete though necessarily mythological
identiIication. Thus, the two principles, liberal parliamentarism and democracy, are contrary and incompatible.
There is one historical context in which liberalism and democracy appeared as allies. In Schmitt's diIIicult and impressionistic line oI argument, what was required Ior
this alliance was the "identiIication" oI the extraparliamentary "people" with the parliamentary public as its representative. Given the very real diIIerences oI
parliamentary notables and outside constituencies, and oI the latter among themselves, the illusion oI necessary homogeneity and unity within society and oI society
with parliament could arise only in the Iace oI an enemy: untamed state power. It is this Iriend-Ioe relation, rather than any Hegelian integrating activity oI the state, that
achieved the temporary unity oI society responsible Ior the illusory identity oI liberalism and democracy.
The problem, however, is not that this identiIication is illusory but that it is temporary. Although the existence oI an undemocratic and illiberal state is necessary Ior the
alliance oI liberalism and democracy, both ideologies, albeit Ior diIIerent reasons, push toward its abolition or its transIormation into a state as the selI-organization oI
society. The military-administrative state is unacceptable to liberal principles, Ior these recognize the legitimacy oI decisions only iI they have been arrived at through
the apolitical principle oI discussion. To be sure, liberalism is skeptical toward any state and seeks a reduced "night watchman" variant. It does not attempt to abolish
Iully or to replace the military-administrative state. The latter, however, insoIar as it is the remnant oI a hierarchical and authoritarian era, is Iar more unacceptable to
democracy. Moreover, once democratic Iorces identiIy with the liberal parliament, they, unlike liberal Iorces, cannot tolerate the Iact that the state is not identical with
this parliament.
Paradoxically, as the triumph oI the alliance oI liberalism and democracy nears its goal, with the creation oI a state that represents the selI-organization oI society
(through extension oI the Iran-

Page 234
chise, which is a precondition oI the alliance), a state toward which a polemical attitude is necessary and possible is no longer a possibility. Along with its (supposed)
disappearance, the conditions oI social unity also disappear, putting liberalism, democracy, and the state itselI into crisis.
86
Schmitt explores the nature oI this crisis by
analyzing two developments linked to the process oI democratization: the emergence oI a new type oI mass-bureaucratic party, and the advent oI state interventionism.
The Iirst leads to a Iundamental transIormation oI the institutions and processes that the liberal model oI discussion presupposed, even iI counterIactually. The second
represents a change with even more radical ramiIications: the "Iunctional dediIIerentiation" oI society and state. This "Iusion" oI the political and the social eliminates the
space Ior a public discursive Iorm oI intermediation, transIormingindeed dissolving, as it werethe public spaces in both society and state.
According to Schmitt (who obviously takes England as his model), the liberal party system was originally based on Iree competition, through the means oI discussion
and persuasion, Ior the votes oI an educated and independent (elite) public. Indeed, liberal parties were to take shape in the sphere oI public opinion, that is, in
parliament. This principle Iound its sociological correlate in relatively small, collegial parties oI notables. Because oI a lack oI attachment to both Iixed interests and
organizational structures, the representatives elected by parties were supposedly capable oI Ireedom oI action and deliberation in parliament; hence the assumption
that they were, as a body, in the position to generate a uniIied will oI the state through discussion and mutual persuasion.
87
Democratization, however, has led to the
emergence oI an entirely new type oI competitive party based on mass membership, sociologically linked to a speciIic constellation oI interests, and heavily
bureaucratized with numerous paid Iunctionaries.
88
Such a party does not value neutrality vis-a-vis its members and tends to be deeply involved in the social,
economic, and cultural liIe oI its "clientele" in all stages oI the human liIe cycle. Nor is it tolerant oI the Iorms oI liIe represented by its competitors. Each "democratic"
party is tendentially totalistic insoIar as it seeks Iull possession oI the state apparatus, which is seen as the instrument Ior carrying out its

Page 235
social goals. The multiplicity oI such parties does keep each one in check; together, they constitute a pluralistic party state (as against a single-party state), a "labile
coalition state." Schmitt pointedly maintains that this type oI state has itselI attained a total character with respect to its predecessor, representing in eIIect a Iragmented
or parceled-out totality in which every organized complex oI power seeks to actualize a totality "in itselI and Ior itselI" (in sich selbst und fùr sich selbst).
Schmitt's explanation Ior the changing character oI political parties in the context oI political democratization diIIers Irom conservative and socialist analyses oI these
phenomena. While conservatives stressed the supposedly inevitable bureaucratization oI politics, given the problems oI organizing uneducated and atomized "masses,"
socialists Iocused on the tendency to create new mechanisms oI exclusion and depoliticization reconciling the "participation" oI the exploited with the imperatives oI
maintaining the existing, exploitative socioeconomic system. Schmitt, in spite oI his bizarre set oI aIIinities with strains within conservatism as well as with authoritarian
versions oI Marxism, bypasses both oI these explanations, Iocusing instead on the end oI the polemical relation oI state and society under the impact oI democratic
parliamentarism. The unity oI the diverse sociological Iormations oI a depoliticized society depended on the survival oI the authoritarian state Iorm. The emergence oI
the state as the selI-organization oI society and the weakening oI the executive Iragment society along lines oI a plurality oI interests and belieIs. Political appeal across
the sociological dividing lines becomes impossible, and political parties must now organize within rigidiIied categories. Furthermore, successIul electoral appeal now
depends on satisIying sectoral economic, cultural, and ideological demands. Accordingly, the parliamentary Iield once again mirrors society at large. This time,
however, the society it mirrors is pluralistically organized, and each segment demands speciIic perIormances in economic, social, and cultural policy. Just as the state
becomes the parliamentarv state, parliament itselI becomes the expression oI mutually hostile societal pluralities capable oI strategic compromise but not genuine
agreement.

Page 236
Moreover, compromise can no longer be achieved through discussion oI the truth and justice oI a given policy, nor can it be openly and publicly arrived at, Ior
compromise and open discussion violate the principles oI the new type oI totalizing political party. The parliamentary discussion that does take place is an empty
Iormality, a mere Iacade, located" in a gigantic antechamber in Iront oI the bureaus or committees oI invisible rulers. . . . Small and exclusive committees oI parties and
party coalitions make their decisions behind closed doors, and what representatives oI the big capitalist interest groups agree to in the small committees is more
important Ior the Iate oI millions oI people, perhaps, than any political decision."
89
Schmitt's concern, unlike that oI Marxian critics oI pluralism, is not that the same interests always dominate through extraparliamentary pressure and deal making.
Because party committees must work through an elected parliament, rule by them creates inconsistent outcomes, depending on results oI elections and coalitions that
strengthen one or another Iaction. The real danger he Iears is not oligarchy but what was later called "ungovernability," since he is convinced that the pluralistic party
state Iragments the two conceivable sources oI unity: state and society.
This Iragmentation is in Iact simultaneous as state and society become one. However, Schmitt's Iusion thesis is not based simply on the actualization oI the program oI
the state as the selI-organization oI society. Indeed this idea, based on Iacile generalization Irom the case oI Weimar, is not convincing despite the dialectical virtuosity
involved in the reversal oI the Hegelian argument. The reality oI the modern state does not in Iact disappear when the democratic transIormation oI parliamentary
democracy is complete. This is certainly not the case in presidential systems, but even in parliamentary systems a growth oI the power oI the executive has historically
accompanied democratization. This growth oI the executive is both a condition oI the constitution oI civil society and a threat to its independence and diIIerentiation.
90

Thus, iI the Iusion oI state and society is the presupposition Ior the decline oI the parliamentary public sphere, this Iusion must have Ioundations in addition to Iormal
democratization processes, ones linked to the expansion rather than the weakening oI the modern state.

Page 237
Schmitt does provide a second train oI argument Ior the Iusion oI state and society, one whose consequences are Iar more general with respect to the roots oI
independent social liIe. This argument, Iocusing on the mutual interpenetration oI state and society, is diIIicult to disentangle Irom the primary emphasis on the
socialization oI the state, but upon closer examination it turns out that here the issue is speciIically that oI a two-directional Iunctional dediIIerentiation. Accordingly, the
nineteenth-century liberal state was diIIerentiated Irom society not only in the sense oI being independent oI segmental constellations oI Iixed social interests but also in
the sense oI being neutral with respect to the great functional spheres oI society that are thereby depoliticized: religion, culture, economics, law, science.
91
Here
Schmitt's model is above all that oI the laissez-Iaire economic order and a state that intervenes at most to restore the disturbed conditions oI economic competition.
From this point oI view, we get an altered catalogue oI liberal Iundamental rights and Ireedoms (personal Ireedom, Ireedoms oI expression oI opinion, oI contract, oI
enterprise, oI property) that does not even include the key Ireedoms oI communication (assembly and association).
92
Here the Iunction oI rights is to maintain
diIIerentiation and depoliticization, rather than to guarantee the preconditions oI public communication.
According to Schmitt, the liberal model oI Iunctional diIIerentiation is assailed Irom two directions. The postliberal state is a "total state which potentially embraces
every domain."
93
This statement has a double meaning. First, the new type oI state is no longer neutral with respect to the various spheres oI society and becomes in
eIIect an economic, welIare, cultural, educational, scientiIic, even "religious" statein a word that Schmitt does not seem to use in this context, it is a So:ialstaat or
social state.
94
Second, the new type oI state intervenes in and politicizes all spheres oI society. The implication here is that the society-state distinction is abolished
with such radicality that the private sphere itselI, stabilized by rights modeled on that oI property, is penetrated, politicized, and abolished as an independent sphere.
While the model oI the rise oI pluralist segmental diIIerentiation seems to make only some rightsthe ones linked to communicationpolitically irrelevant, the model
oI Iunctional dediIIerentiation actually supports Schmitt's statement that liberal rights as such have become obsolete.

Page 238
The relationship oI these two models in Schmitt's argument is complex. The only "explanation" he provides Ior Iunctional dediIIerentiation is, once again,
democratization, which Ior rather unclear reasons "must do away with . . . the Iorms oI depoliticization characteristic oI the liberal nineteenth century."
95
In Iact, the
argument again seems to rest on the extent to which the program oI "liberal democracy" can establish the state as the selI-organization oI society. Here Schmitt would
have us think that the idea oI a So:ialstaat, in the sense oI an economic, welIare, cultural, etc., state, and that oI society becoming the state (:um Staat gewordene
Gesellschaft) are the same. But in his own argument the culmination oI the state as the selI-organization oI society only leads to the Iragmentationthat is, the
segmentation along lines oI interest and ideologyoI the society that takes over the state. The outcome, as we have shown, is a Iragmented pluralistic party state
whose sovereignty is parceled out among the units. Schmitt's never clariIied argument seems to hinge again on the type oI the democratic-mass-ideological party that
involves itselI in all aspects oI the social liIe oI its members. Such a party would presumably seek a state modeled on itselI, intervening in society on behalI oI the
economic, cultural, and other interests it represents. Unless Schmitt has in mind the speciIic examples oI the relationship oI the Social Democrats to the economy, or oI
the Catholic Center party to religion, it is entirely unclear why the new party system should lead to a wholesale process oI Iunctional dediIIerentiation oI state and
society. Indeed, even Mussolini's single-party state could coexist Ior a time with a liberal economic order.
Once again, we believe, the source oI the conIusion is Schmitt's unwillingness to concede that, whereas in the case oI segmentation the source lies in social complexes
seeking to capture or at least parcel out the state, in the case oI Iunctional dediIIerentiation we are dealing with a powerIul administrative-bureaucratic state seeking to
penetrate society. From his highly committed Weimar perspective, Schmitt saw "social-democratization" but not state interventionism as a dynamic Iorce leading to
political crisis. Nevertheless, he is aware oI two possible outcomes that are in line with the two tendencies we were Iorced to separate in his thought. OI the two
versions oI the "total" state he outlines, the Iragmented

Page 239
pluralist variety is the product oI the tendency toward segmentation; the authoritarian variety is the product oI Iunctional dediIIerentiation driven by the logic oI the state
itselI.
96
Schmitt seems to have some idea that the two versions oI the total state Ilow Irom diIIerent meanings oI the term "social-state" or "state-society," one implying the
primacy oI the social, the other that oI the political. He states that the "pluralistic party state becomes 'total' not out oI strength but weakness; it intervenes in all areas oI
liIe because it must satisIy the claims oI all those interested."
97
Nevertheless, he also believes that the Iragmented variety oI the total state is not so much an alternative
outcome oI the repoliticization oI society as an artiIicial product, by deIinition almost always in crisis, a result oI the survival oI obsolete legal and parliamentary
institutions. In particular, he believes that the culmination oI the trend against the liberal neutralization oI the state and the depoliticization oI society has already
produced the Ioundations oI another authoritarian Iorm oI power resting on democratic-plebiscitary legitimacy. Indeed, it is an unstated consequence oI his argument
that such an outcome could even converge with the logical selI-abolition oI the party system, with the rule oI many parties being replaced by that oI a single,
monopolistic party. Thus, the two tendencies toward Iusion, segmentation and Iunctional dediIIerentiation, could converge in a new type oI "democratic'' dictatorship.
On the basis oI his interpretation oI the experience oI Weimar, however, Schmitt is convinced that the operation oI parliamentary legality, even iI it is no longer in a
position to produce a legislative state, is nevertheless capable oI checkmating the emergence oI a genuinely politicalthat is, authoritarianstate Iorm.
98
A parliament
guaranteeing the political rights oI a plurality oI parties is capable oI checkmating decisions oI the executive that arise outside the given conditions oI coalition
Iormation. And one might add that the survival oI the liberal Iramework oI legal protection outside oI parliament makes the replacement oI a system oI a plurality oI
parties by that oI a single party nearly impossible.
99
According to Schmitt, the alliance oI liberalism and democracy is (Ior the present) beyond repair. The instrument oI majority rule in parliament loses its chances oI
popular acceptance when highly

Page 240
organized political groupings predetermine all possible outcomes, establish an irreversible advantage oI incumbency, and rigidiIy a given structure oI majorities and
minorities and even oI complete political exclusion. Thus, each partner oI the Iormer marriage oI liberalism and democracy is now in crisis: democratic legitimacy along
with the parliamentary principle. Their crises produce a third one, that oI the state itselI, to the extent that solutions beyond liberalism and the existing Iorm oI
democracy are successIully blocked and the possibilities oI decision are continually stymied. To Schmitt, there seem to be two choices inherent in this situation: the
continuation oI an antipolitical, pluralistic party-state in permanent crisis but protected and disguised by liberal principles, or the creation oI a genuinely political, no
longer pluralistic, authoritarian state legitimated by a new, plebiscitary version oI "democracy." It is useless to deny that it is this second option that Schmitt chooses.
Indeed, it was this choice that allowed Schmitt to be enthusiastic about Italian Fascism and that made his turn to National Socialism intellectually authentic, iI not
inevitable. For Schmitt, no return to a conservative, nonplebiscitary authoritarian regime could provide a solution Ior the crisis oI the state, since such an alternative,
by reconstituting their earlier polemical adversary, would lead to a reconstitution oI the alliance oI liberalism and democracy and again undermine the state. Like the
leItists and rightists he admired, Schmitt proposed an alternative marriage: that oI democracy and authoritarianism.
In any case, with the coming oI the total state, neither oI the two options (pluralistic or authoritarian) is consistent with a dualism oI state and society, or with the
operation oI a parliamentary mediation between them. What does not occur to Schmitt is the possibility, so obvious in the American context, that the two principles,
statist and pluralistic, stabilized in a Iramework oI liberal rights, could combine to constitute a new version oI the state-civil society dualism. Three Ieatures oI his
thought were responsible Ior this myopia: an unwillingness to recognize the continued existence oI a tendentially authoritarian state in the pluralistic era; an inability to
see the whole gamut oI reasons, including especially the economic ones, Ior state interventionism in society; and a Iailure to note the emergence oI yet another new
type oI political party, the

Page 241
catch-all party, based on a mixed constituency, interested neither in totally dominating nor in parceling out the political system, capable oI greater Iluidity in the
parliamentary arena and oI more than merely strategic compromise with its adversaries.
The disappearance oI the state in Schmitt's picture oI liberal democracy was hardly innocent: He sought to reinIorce an authoritarian administration that he presented
as weakened, and to do this he had to disguise its role in the crisis oI the political order oI Weimar. The pretense that the authoritarian element oI the state was
moribund, despite the power oI the army, the administration, and the legal system allied to the administration, not to speak oI the presidential prerogatives oI the
constitutional system, helped him to attack the pluralistic party system that produced new links as well as tensions between democracy and liberalism.
The Fusion Argument in Habermas's Strukturwandel
Given its thinly disguised authoritarian intentions, it is all the more striking that the Iusion argument Schmitt worked out was adopted, and indeed dramatically
reIurbished, by the writers oI the FrankIurt school. Their attitude to liberalism, democracy, and authoritarianism was the opposite oI Schmitt's, yet the Iusion argument
became Ior all oI them a signiIicant Ieature oI the "critique oI the authoritarian state."
Consistently enough, neither the alliance oI liberalism and democracy, nor the supposed decline oI their adversary, authoritarian executive power, plays a role in the
FrankIurt analyses. This structure oI the argument is replaced by a new one: the great transIormation oI the capitalist economic order Irom liberal to monopoly and
Iinally to state-organized capitalism. The argument, although Iirst developed in relation to the rise oI authoritarian states, also proved applicable in the postwar period,
when liberal democracy was reconstructed.
100
Habermas's theory oI the decline oI the public sphere, however much inIluenced by the earlier theses oI Schmitt and
Arendt, derives Iirst and Ioremost Irom the various strands oI the FrankIurt school analysis oI the 1930s. Indeed, Habermas ultimately managed to recast almost all oI
these strands in a new theoretical Iramework, where they became quite

Page 242
useIul Ior a democratic theory oriented to practice. But in 1962, at the time oI writing oI Strukturwandel der Offentlichkeit, Habermas had not yet achieved this
position. As a result, unIortunately, he linked the notion oI the transIormation oI the public sphere to the negative philosophy oI history oI Adorno and Horkheimer, and
consequently he was unable to see much beyond a thesis oI decline, except to the limited extent that he, unlike his teachers, still harbored some classical Marxian
assumptions. The application oI the theory oI the public sphere to contemporary politics had to wait.
Here we need only summarize Habermas's multidimensional synthesis. The argument is composed oI six levels:
1. The thesis oI state interventionism in the capitalist economy. This argument, almost entirely missing in Schmitt, involves something qualitatively diIIerent Irom the
expansion oI state administration and political bureaucracy during the absolutist and even liberal epochs stressed by Marx, Tocqueville, and, in her own way, Arendt.
The modern state intervenes in the liberal capitalist economy, at the price oI its liberal character, to protect the capitalist structure endangered by endogenous crisis
tendencies and processes oI impaired selI-regulation. The state seeks to correct disequilibria produced both by selI- regulating market processes and by phenomena
oI imperIect, oligopolistic competition (Iiscal and monetary regulation oI the business cycle), to underwrite processes oI investment, accumulation, and technical
innovation, and to support aggregate demand through welIare-state expenditures. This thesis, rather undeveloped in Strukturwandel, was Iully integrated in the
FrankIurt tradition by F. Pollock and his colleagues (19321941) in relation to "the authoritarian state." It was powerIully extended in the writings oI Habermas and
Claus OIIe aIter 1968 in the Iorm oI a critique oI welIare-state crisis management.
101
2. The thesis oI the assumption oI public powers by private associations (new corporatism). This thesis, Iirst introduced into FrankIurt discussion by O.
Kirchheimer,
102
derives Irom Schmitt's critique oI Weimar pluralism. In Habermas's 1962 argument, the critique is extended to the prepolitical level. In processes oI
oligopolistic competition, private organizations are capable, as

Page 243
against liberal capitalism, oI Iormulating what is in eIIect public economic policy.
103
The collective agreements among private associationsin particular, employment
associations and unionslose their private law status in Iavor oI a Iorm oI rule creation that was previously reserved Ior public law entities. While important areas oI
administration now Iall to private law entities, the state itselI increasingly uses private legal contractual devices to regulate its relations with its social partners. This
argument, albeit deemphasized by Habermas in his subsequent work, was powerIully extended by OIIe in the 1980s.
104
It is worth stressing, however, that he did so
not only to indicate a component oI overall welIare state structure (in any case, a component not equally important under all welIare states) but, under the impact oI the
neoconservative challenge, to stress one potential, albeit internally problematic (and normatively unattractive!), avenue Ior reducing the administrative and legitimating
burdens oI the interventionist state.
3. The thesis oI the decline oI the intimate sphere oI the Iamily. This thesis, an important component oI Arendt's analysis whose very Iormulation is taken over by
Habermas ("the polarization oI the social and intimate spheres"), was a key contribution oI Horkheimer and his colleagues in the 1930s to social theory. Habermas's
1962 analysis, drawing on new literature, stresses the destruction oI the private shell oI bourgeois property around intimacy, caused by the loss oI the Iamily's
economic Iunctions and the growth oI client relations to a state in its capacity oI providing social insurance. The Iamily increasingly loses its Iunctions oI "education,
deIense, caring and direction, and even oI providing traditions and orientations . . . its conduct Iorming power in areas that counted as the most internal spheres oI the
members oI bourgeois Iamilies."
105
The decline oI the authority oI the Iather is, Irom this point oI view, ambiguous: The Iamily loses not only its repressive but also its
deIensive Iunctions. The new Iorms oI even more intensive intimacy are seen as hopelessly deIensive, in the manner oI Arendt; private liIe becomes more and more
open to the gaze oI outsiders, down to the very level oI architecture. The Iake intimacy oI public communication, stressed by both Adorno and Arendt, represents to
Habermas both a Iorm oI the subsumption oI the intimate sphere and the degradation oI public into mass.
106


Page 244
4. The thesis oI the decline oI the literary public sphere and the rise oI mass culture. This complex oI arguments represents the most successIul and best-known
dimension oI the theory oI the early FrankIurt school, above all oI Adorno. The stress in Habermas's version is on the growth oI the literary public into the sphere oI
consumption and manipulated leisure. This is linked to the decline oI Iamily-based institutions oI cultural reception and criticism as well as to the industrial-commercial
transIormation oI the media oI communication. A market is no longer the precondition oI autonomous art; marketability becomes a principle oI the industrial
production oI art. The "democratization" oI culture is a pseudo-democratization; what is democratized is no longer culture. The dramatic expansion oI the literary
public sphere is simultaneous with the decline oI its critical character.
107
The new media Ioster a merely passive Iorm oI participation. The survival oI avant-garde art
and culture only splits the classical literary public sphere into "a minority oI reasoning, no longer public, experts and the great mass oI public consumers."
108
The erosion oI the intimate sphere and oI a genuine literary public leads to the loss oI the tension between homme and bourgeois, abolishing the private Ioundation oI
autonomy without providing a new public one. Here the theses oI the decline oI the Iamily and the rise oI mass culture are linked to the classical FrankIurt thesis oI the
decline oI the individual.
5. The thesis oI the transIormation oI the political public sphere represents a selective extension oI arguments developed in relation to the prepolitical dimensions oI
publicity. Interestingly enough, statist-bureaucratic intervention into the economy, Ior Arendt the replacement par excellence oI egalitarian public interaction by
paternalism, is somewhat deemphasized in the analysis, although Habermas does mention the growth and increasing independence oI an administration that successIully
resisted, even in the liberal era, the demands oI publicity. All the more important is the Schmittian argument, stressed also by Kirchheimer, according to which the
assumption oI public powers by private associations leads to the emergence oI corporatist processes oI negotiation, bargaining, and compromise that bypass public
processes oI scrutiny
109
and reduce parliamentary discussion and debate to a post hoc process oI

Page 245
legitimating decisions arrived at under the protection oI a new "arcanum." No longer attempts by representatives to convince one another, speeches in parliament now
seek to mobilize a plebiscitary opinion outside parliament. As Schmitt argued, representatives bound by party discipline lose their independence as something
resembling the bound mandate is revived. Habermas recognizes that Schmitt's conception oI the transIormation oI the party system Irom loose collegial groupings
bound by common opinion to parties as rigid sociological groupings no longer corresponds to reality. The new type oI "catch-all party" stressed by Kirchheimer
among others, a Iurther stage in the "democratization'' and "massiIication" oI the political system, only increases depoliticization by Iurther reducing the level oI political
discourse and argument.
110
OI course, the new type oI party is no longer associated here with the parceling out oI sovereign power. Its most important result, the
"vanishing oI the political opposition," to use Kirchheimer's phrase, has the eIIect oI reducing public controls upon the administration, as stressed by Max Weber,
thereby strengthening authoritarian power without authoritarian means.
6. Habermas extended Schmitt's thesis that the role oI parliament as sphere oI mediation between a strengthened bureaucracy and private associations must decline.
Equally important, however, was his use oI the FrankIurt school thesis on mass culture to demystiIy the allegedly "democratic" character oI the plebiscitary
components oI the new situation, stressed by Schmitt. In the tradition oI Adorno and Lowenthal, who emphasized the authoritarian political potential oI the new mass
culture and its media, Habermas points to the place oI propaganda in contemporary political discourse. Modern political manipulation presupposes the Iorms oI
commercial advertisement that become dominant as price competition ceases to be the mechanism coordinating oligopolistic groupings in their struggle Ior market
shares. As Adorno and his colleagues well knew, propagandathe advertising and selling oI political leaders, parties, and policiespresupposes already Iormed,
passive, uncritical, yet mobilizable audiences. While advertising as such turns to individuals in their private capacities, thus helping to decompose the intimate sphere,
the intermediary Iorm oI "public relations" turns to and deIorms

Page 246
"public opinion" through "the engineering oI consent."
111
It is this task that becomes central Ior political parties oI the contemporary type, in parliament and especially
in the electoral process. Such parties do not need continuous mass membership as much as an apparatus capable at periodic intervals oI mobilizing electoral support in
the manner oI an advertising agency. Although the reconstitution oI some kind oI political public sphere in electoral campaigns is unavoidable,
112
the preIerred targets
oI parties are those individuals, generally not members oI associations or higher status groups, who have no access to what are depicted here as residual Iorms oI a
reasoning public. The targeted voters are approached not through enlightenment but through appeals to consumerist behavior, and not by agitators or even
"propagandists" oI the old type but by advertising experts.
113
To be successIul, ''the organizers oI elections must not only recognize the disappearance oI a genuine
political public sphere, but Iully consciously must help produce this outcome."
114
The result is not understanding or agreement with policies but a "symbolic
identiIication" with leaders that is measurable, and Iurther opened to manipulation, through popularity indices and "public" opinion polls that reIer exclusively to
nonpublic and atomized opinion. Even iI parties and governments were actually responsive to "nonpublic opinion," the result would still be more like enlightened
absolutism than a genuine democratic will Iormation based on the transIormation oI personal opinion through processes oI rational deliberation into a genuine public
opinion.
115
Habermas's aim on all these levels oI analysis is not only to demonstrate the deIormation and deterioration oI the principle oI Iree public communication. Even more
important Ior us is his complementary thesis: that oI the destruction oI the model oI diIIerentiation between civil society and state through a Iusion oI levels. II the
deIormation oI mediating institutions itselI promotes dediIIerentiation, it can be also be argued that the tendencies toward Iusion oI state and society remove the social
space in which the liberal public sphere could Iunction. On one level, the diIIerence between the two processes is only one oI emphasis: Habermas is interested in the
decline and revival oI the public sphere, which in 1962 he still imagined to be possible without a model that

Page 247
diIIerentiated between state and civil society. We, on the other hand, are interested in reconstructing the diIIerentiated model, which we do not think possible or
normatively desirable without a renewal oI the liberal and democratic project oI the public sphere.
But there is also a systematic diIIerence between our two approaches, to the extent that the model oI the decline oI the public sphere reIers to a much more complete
process oI Iusion and even "one-dimensionality" than that oI the new relation oI state and society. This can be seen in the structure oI Habermas's argument. He tells
us, rightly, that the model oI the repoliticization oI society through state intervention in the economy cannot, on its own, establish a Iusion argument, since private
economic activity could be limited in important ways without such intervention aIIecting the private nature oI large areas oI personal interaction. But he is wrong to
suggest that the case can be completed by reIerring to the complementary assumption oI public powers by private associations. Even iI the two processes do produce
an intermediary sphere to which the distinctions oI private and public, society and state, no longer apply, they do not in themselves make the distinction disappear, as
the terms "statizing society" and "socializing the state" seem to imply. In particular, the spheres oI intimacy and publicity proper are not directly decomposed by the two
processes; Ior this to happen, a reiIication and instrumentalization oI these two ultimately cultural spheres is necessary. II the two complementary processes leading to
Iusion are to reach their goal, the reiIication oI the space between them, that oI culture, must become more or less total. The thesis Iirst introduced by Schmitt can be
saved only with the help oI the cultural theory oI the FrankIurt school, especially in the version oI Adorno. But this choice would lead, in the case oI Habermas's own
thesis as well, to a manipulated public whose agents are entirely passive and whose present dynamic could in no way point to the revival oI its original promise.
116
Such a constellation could still be open to revolutionary ruptures, in Arendt's sense. And it is in Iact Iair to ask at the end oI Habermas's book to what extent he has
escaped the ancient republican model oI the public sphere he criticized in Arendt's work. In working out the consequences oI the Iusion argument, he suddenly tells us
that "the model oI the bourgeois public was based

Page 248
on the rigid separation oI public and private spheres, since the public sphere oI private individuals organized as publics counted as private."
117
While juridically
correct, this argument breaks with Habermas's earlier, more Hegelian, argument according to which it was precisely the rigid distinction oI public and private that was
relativized by the various levels oI mediation.
We should note, moreover, that the utopia Habermas derived Irom Marx, involving the duality oI a public state-society and the intimate sphere, along with the primacy
oI the Iormer, coincides with Arendt's republican model. Habermas's theory oI decline also postulates the emergence oI a mixed realm that is neither public nor
private, leading to the collapse oI genuine publicity. For Arendt, the model oI the liberal public sphere worked out by Habermas, concerned as it was with mediating
between state and society through regulation oI the preconditions oI the market economy, alreadv was this mixed sphere and could involve no genuine public liIe and
action. The idea oI a public realm controlling and inIluencing the state without sharing in power would have appeared senseless to her. All the same, it could be argued
that the only real diIIerence between their analyses is that Habermas gives Arendt's model oI decline a historically distinct starting point Irom which a decline could take
place. And indeed, however inconsistently, the emergence oI a mixed realm also seemed to Habermas to deprive "the public sphere oI its old basis without giving it a
new one."
118
OI course, this was a Iunction not oI the rise oI the modern state as such but oI the postliberal relations oI state and economy. Obviously, Habermas and
Arendt share an interest in working out anew such a basis. In this context, however, we should also recall that Habermas repeatedly asserts that he seeks to
reinstitutionalize the liberal rather than the ancient model oI the public sphere.
The ideal oI the liberal public sphere contains Ior Habermas that oI democratization. Paradoxically, the historical processes oI democratization, whether oI politics as in
the party system or oI culture as in mass culture, contributed to the decline oI the institutions that Sustained this ideal, in however contradictory a Iashion, reducing it to
an abstract principle oI legitimation. The decline oI liberal institutions could, however, be seen Irom two points oI view: that oI diIIerentiation oI state and civil society
as

Page 249
expressed by the principle oI rights, and that oI the public sphere as expressed by the principle oI rational communication. It is thereIore ambiguous to argue Ior a
reinstitutionalization oI liberal principles unless one speciIically reIers to both or only one oI these. Habermas's obvious inclination was to deIend the principle oI
communication primarily. To be sure, the classical catalogues oI rights posited this principle in terms oI a series oI well-known rights (Ireedoms oI speech, assembly,
suIIrage, etc.). But the very meaning oI "rights" in this case, as in others, involved something more: Rights as liberties diIIerentiated between the private sphere and
public authority, and they implied the protection not only oI the mediating public sphere Irom state power but also oI the private sphere Irom both publics.
Since Habermas does not want to abandon such catalogues, he argues Ior their redeIinition and reconstruction. In this context, he maintains not only that the actual
trend oI welIare-state jurisprudence is in a direction that transIorms the merely deIensive, negative structure oI inherited constitutional rights but also that this
development represents what is in eIIect the only immanent tendency in our societies toward a reinstitutionalization oI the public sphere.
119
Not only, then, does he
speak about the survival oI the principle oI the liberal public sphere on the normative level, but he also claims that both the letter and the spirit oI constitutional norms
seeking to regulate the transition Irom liberal Rechtsstaat to the welIare state anticipate the new Iorms oI reinstitutionalization oI this principle, thereby contradicting the
institutional practices oI existing welIare states.
120
It is at this point that an argument that previously treated the modern public and intimate spheres as passive objects
Ior economic and political processes leading to their disorganization suddenly discovers that the norms originating in these spheres are possible points oI orientation Ior
an alternative strategy. Accordingly, Habermas proposes a model oI reconstruction. It should not come as a great surprise that we get a new version oI the antinomy
we Iound in both Hegel and Gramsci, involving two opposed orientations, one statist and one oriented toward civil society.
We should note that the argument dealing with legal developments in the welIare state suddenly breaks with the general trend

Page 250
oI Habermas's analysis that aligns it with the negative philosophy oI history and social theory oI the late FrankIurt school, and with the school's legal theory as well.
Habermas does reIer to Franz Neumann in suggesting that, with the Iusion oI state and society, the generality oI legal norms cannot be maintained; rather, law and
administration are increasingly dediIIerentiated.
121
Neumann would argue, however, that without the generality oI norms it would be impossible to sustain the principle
oI Iundamental rights, which would be incoherent without any limitations and impossible iI the limitations were not deIined according to rigorously general standards.
Habermas, on the other hand, maintains that only the negative and deIensive aspects oI rights vis-a-vis the state are challenged under welIare-state
constitutionalism.
122
The motivations oI the state in this context are clear: With its intervention into society, selI-limitation with respect to social autonomy may seem obsolete, and, more
important, new justiIications are needed that can validate the new Iorms oI state action as just. Given the survival oI liberal norms as legitimations, such validation can
be developed by relying on the internal logic oI liberal rights. And given the decline oI the competitive economic system in the context oI an interventionist and
redistributive state, the "positive IulIillment" oI the negative, deIensive rights in terms oI an actual ability to practice the Ireedoms oI speech, assembly, and association
as well as those oI political participation no longer Iollows more or less automatically. The state must thereIore provide the positive and, indeed, the material
guarantees Ior participation in terms oI new social rights. From the point oI view oI liberal rights themselves, iI these are "to remain IaithIul to their original intentions,"
their "normative interpretation must be changed." While negative rights as "liberties'' (Freiheitsrechte) are preserved in welIare-state constitutions, they must now be
seen as rights oI participation (Teilnehmerrechte), which will be interpreted in terms oI positive social rights (So:ialrechte) to state activities rather than Iorms oI selI-
deIense and selI-diIIerentiation with respect to the state.
123
To be sure, there is sleight oI hand involved here. Even the constitutions Habermas considers the most advanced contain, he admits, negative rights, rights oI
participation, and social rights

Page 251
alongside one another. This raises the question oI whether Habermas himselI means to aIIirm the need Ior both negative and positive rights, or whether he is arguing Ior
a transition Irom the Iirst to the second. While the issue is ambiguous in his presentation, he seems to consider the survival oI negative rights to be a mark oI an
insuIIicient overcoming oI the bourgeois "tax-state" (Steuerstaat) character oI the welIare state, the incompleteness oI its realization oI the goal oI a uniIied state-
society subordinating economic processes to its direction.
124
In light oI this goal, even rights oI the intimate sphere, no longer protected by the outer shell oI property,
need to be redeIined, according to him, as Iunctions oI or derivations Irom the public processes oI democratic participation.
125
In this context, Habermas seems to
Iully aIIirm W. Abendroth's claim that the supposedly authoritarian implications oI such a model actually involve Ior most individuals only a transition Irom dependence
on the private power oI particular interests to dependence on processes oI collective control "whose highest unit oI decision is the state itselI." The only thing
Habermas adds to this clearly statist and authoritarian model is the desideratum that the state as the uniIied planning and control organ oI all social processes is itselI to
be subordinated in the uniIied state-society to processes oI the ''public opinion and will Iormation oI the citizens."
126
This democratic statism is then supposed to make
negative rights oI individuals and groups superIluous.
Habermas does also note and aIIirm a competing model within jurisprudence oriented toward the welIare state. In this model, the Iunction oI mediation between social
interests and state decisions does not disappear in the welIare state; only its public character is abandoned. The private-public organizations that assume this role,
arising in part Irom the private sphere (social associations and organizations) and in part Irom the public sphere (parties), cooperate with the administration oI the state
and attempt to secure "public" acceptance through manipulative, hierarchical procedures.
127
What is leIt oI a political public sphere is dominated by these entities, one
oI whose tasks is to inIluence the redistributive activities that represent the positive guarantees behind "social rights." The real bargaining processes in which this occurs
are not public, and the demands oI publicity aimed at state agencies bypass

Page 252
the juridically private structure oI the negotiations. In this context, Habermas stresses the trend in welIare-state constitutionalism toward extending the demands oI
publicity Irom the state to the relevant social associations and political parties, and to the processes oI their interaction with the state. Only such legislation could revive
public discussion in the sphere that really matters, by substituting "in place oI the no longer intact public oI private persons interacting only as individuals, a public oI
organized private individuals." It is this trend that Habermas considers identical to the project oI establishing a critical public sphere under contemporary conditions, in
deep and not yet decided conIlict with the now apparently dominant trend oI the manipulation oI publicity.
128
Habermas does not seem to realize that this pluralistic model oI the critical public sphere is in conIlict with the ideal oI a uniIied state-society as well. Undoubtedly, he
identiIied both the agent (state legislative activity) and the end result (a Iully public process oI decision concerning all socially relevant questions) oI the two processes.
All the same, the project oI establishing a uniIied state-society, expressed in the transition Irom negative rights restricting the state to positive rights implying state
action, points toward a monolithic democratic society with a single collective actor, promoting the participation oI individuals in a single, uniIied societal public sphere.
In such a context, minorities as groups and even associations with particular interests and identities would not be protected; only their individual members would be
protected as citizens oI the whole. Even iI such a model does not become the mask oI statist authoritarian rule, it has no saIeguards against a totalitarian democracy.
In comparison, the project oI democratizing existing associations and parties is pluralist rather than collectivist. While its aim is to reestablish the public sphere, this is to
be done in terms oI establishing small publics in each association, linked together in terms oI more general and, again, public processes oI interaction. Even iI state
legislation is to play a role in establishing this model, the old polemical attitude to the authoritarian dimensions oI state administration would inevitably return, and the
state would be pressed not only to guarantee the new publics materially but to limit itselI as well. Unless we believe that the administration oI the

Page 253
state could altogether disappear, this double relation oI publics to state would have to be institutionalized, a requirement reIlected precisely in the ambiguity oI the
overall structure oI rights Iound in modern constitutions. The new Iorms oI publicity obviously require not only material inputs Irom the state but also Iorms oI
protection Irom state interIerence. The small publics oI associations and parties, which must be autonomous even vis-a-vis the larger public process regulating their
interaction, cannot do without both negative and positive rights. However, this requirement reestablishes the two normative Ioundations oI the liberal public sphere:
diIIerentiation and communication. The point does not apply only to the rights oI communication, though. The members oI democratized associations need the same
double protection. To be able to participate at all, they need positive supports and guarantees; to be able to Iunction Ireely, they need negative rights and liberties.
129
Habermas undoubtedly believed that his two models were competitive only to the extent that they aimed at democratizing the two separate processes leading to Iusion:
"the statization oI society" (state interventionism) and "the socialization oI the state" (neocorporatism). In assuming ultimate Iusion, he assumed the convergence oI the
two democratizing processes as well. What he does not realize is that his Iirst democratizing process produces only the social conditions necessary Ior the exercise oI
public Ireedom, in the Iorm oI "social rights," which in themselves are quite compatible with an enlightened and paternalistic absolutism. Only the second process
revitalizes the constitutive interaction oI the public sphere itselI, in the Iorm oI genuine "rights oI participation." The two processes do not Iully converge, and in Iact they
reproduce the diIIerentiation that state interventionism and corporatism together endangered. Moreover, they Ilow Irom two distinct theoretical traditions: the Marxian
utopia oI state-society, and the Tocquevillian project oI reestablishing the intermediary associations oI civil and political society in a democratic Iorm.
The second reason why the two models may seem to converge is the common process by which they are to be instituted: welIare-state legislative action. Habermas
does postulate the survival oI the liberal value oI publicity, which serves as the normative background

Page 254
Ior state actors seeking legitimacy in the context oI increasing interventionism. But the norms are not linked to other actors on their behalI, Ior this is next to impossible
in the context oI the deIormed and manipulated public sphere.
Logically, at least, state action can aim at its own selI-limitation. All the same, there is reason to believe that models oI diIIerentiation based on rights have never been
established without actors outside oI and even antagonistic to the state. The model oI the deIormed public sphere, however, implies a society without opposition and
the passivity oI potential social actors. Habermas's choice Iollows Irom his analysis. The implied identiIication oI the two models Ior restoring public liIe is a result not
only oI his socialist convictions but also oI his diagnosis oI an irreversible statist turn in the organization oI modern societies. Thus, the choice between two models oI
the statization oI societyone public-democratic and one manipulative-democraticturns out to be no choice at all. Paradoxically, the analyst who has done most to
identiIy the norma Iive ideal oI the modern public sphere with the diIIerentiation oI state and civil society came to the conclusion that this ideal could be saved only by
accepting what has already occurred: dediIIerentiation and the abolition oI an independent civil society.

Page 255
6-
The Genealogical Critique:
Michel Foucault
One could interpret Foucault's work as another critique oI the welIare state parallel to that oI Arendt, Schmitt, and Habermas, albeit one that derives Irom a diIIerent
theoretical tradition and uses diIIerent means. OI greater importance Ior us is the Iact that Foucault presents a Iar more relentless critique oI modern civil society than
any oI his predecessors or contemporaries. While he shares with Arendt her suspicions regarding the genesis and Iunctions oI the social, while his genealogical account
oI modern power relations has the same target as Schmitt's historicist critique (the liberal-democratic model oI law and the normative conception oI civil society), the
thrust oI his analysis is neither anti- nor prostatist. Its target comprises, rather, the categories oI civil society. These move to center stage and play key parts in the story
oI the birth, growth, and dynamics oI modern power relations. To be sure, the contemporary welIare state plays a role in the globalization and deepening oI modern
Iorms oI domination, but it is neither their source nor the main actor in the drama.
Indeed, while Foucault would certainly agree with Habermas's account oI the ways in which the deIormed public sphere Iunctions, as well as with the thesis oI an
interpenetration oI societal and state power relations, he would reject the very notion oI deIormity to which Habermas counterposed the continued relevance oI the
norms oI civil society. In this respect, at least, Foucault's analysis parallels that oI Niklas Luhmann. Both argue that the normative conception oI legitimation, law,
publicity, and rights is an obsolete remnant oI the aristocratic-monarchic system. Although both are

Page 256
aware that these concepts (along with democracy) were taken up by reIormers and revolutionaries in the late eighteenth and early nineteenth centuries, they insist that
they are irrelevant to modern decentered societies. However, Luhmann and Foucault give rather diIIerent reasons Ior this thesis. As we shall see in chapter 7,
Luhmann locates his explanation in the modiIication oI the primary principle oI societal diIIerentiationi.e., in the reorganization oI the social system oI stratiIication
into Iunctional diIIerentiation. In modern, diIIerentiated social systems, it is no longer possible to represent the unity oI society; representation and the normative
categories oI civil society have become hopelessly romantic. For Foucault, however, it is not Iunctional diIIerentiation but the emergence oI a new Iorm oI stratiIication
and new power relations that renders the normative juridical model anachronistic. While the problem oI domination recedes into the background in Luhmann's work, it
is central to Foucault's. Accordingly, and in contradistinction to Habermas, Foucault's version oI the rise and development oI modern civil society is unambiguously
negative Irom the start. Moreover, since they are conceived as the product oI modern technologies oI power, none oI the categories oI civil society can provide a
reIerence point Ior any project to the structures oI domination pervading our societies. It is to this rather alarming conception oI civil society that we now
turn.
Marx, Generalized
In many respects, the most important touchstone Ior understanding Foucault's critique oI civil society is the work oI Karl Marx, rather than that oI his own
contemporaries. II Marx was the peerless nineteenth-century critic oI modern civil society,
1
surely Michel Foucault deserves to inherit that title Ior the twentieth. Like
Marx, his purpose is to analyze the Iorms and techniques oI a modality oI power that is uniquely modern. His analysis, again reminiscent oI Marx, takes up the core
categories oI civil societylaw, rights, autonomy, subjectivity, publicity, plurality, the socialin order to show that, Iar Irom articulating the limits to domination, they
are instead its supports. Although we intend to show that this

Page 257
analysis is one-sided, indeed, that Foucault is caught up in the very standpoint oI the modality oI power that he analyzes (strategic reason), it is nevertheless clear that
no theory oI civil society can ignore his contribution iI it is to avoid apology.
Despite important diIIerences, Foucault's analysis oI the speciIicity oI modern society builds on a core insight oI Marx: Modernity involves the emergence oI a new and
pervasive Iorm oI domination and stratiIication. This is not to suggest that Foucault operates within the Marxist universe oI discourse; indeed, the dialectic, economic
determinism, historical materialism, the base/superstructure model, the concern with ideology, the strategy oI immanent critique, and the Iocus on class struggle are all
absent Irom his work.
2
He explicitly abandons this discourse Ior several reasons. First, the Marxian Iocus on the economy yields an inadequate account oI power
relationsneither the Iorms, the strategies, nor the actual Iunctioning oI power can be located in the economy or placed in a subordinate position relative to it.
3

Second, the dialectical theory oI history that postulates the emancipatory potential oI a macrosubject capable oI totalizing local resistances into a revolutionary political
movement that could end societal domination once and Ior all is deeply misguided and dangerously utopian.
4
Moreover, totalizing theory in any oI its guises is both a
hindrance to research and politically disadvantageous. According to Foucault, global, systematizing theory tends to gloss over the details, local Iorms, and speciIicity oI
the mechanisms oI power, while at the same time holding everything in its place instead oI loosening the Iight grip oI unitary discourses on our thinking.
5
Foucault does
not reject Marxism Ior the sake oI emphasizing the positive achievements oI modern civil society. On the contrary, he does so in order to provide a superior account
oI the new kinds oI power relations that pervade social liIe Iar more thoroughly and extensively than Marx imagined possible.
Foucault does not use the term "civil society," but he does presuppose the diIIerentiation between state and society that, according to Marx, was the hallmark oI
modernity.
6
Moreover, like Marx, he argues that the locus oI modern power relations is society, independent oI and distinct Irom the sovereign state. Foucault does
not reduce society to its economic substructure, nor does he

Page 258
see class relations as the paradigmatic Iorm oI power relations or struggle in modern society. Instead, he takes the Marxian insight into the "anatomy" oI civil society a
step Iurther;
7
just as Marx discovered power relations in the Iactory, constituted and concealed by the juridical niceties oI the labor contract, Foucault uncovers
asymmetric relations oI power in the other key institutions oI modern society: hospitals, schools, prisons, asylums, armies, the Iamily, and so on. Indeed, what Marx
claims regarding exchange relations and contract law is, according to Foucault, true oI all juridical Iorms and all the major institutions oI modern society: Norm, legality,
and rights go together with discipline, power relations, and subjugation:
Historically the process by which the bourgeoisie became in the course oI the eighteenth century the politically dominant class was masked by the establishment oI an explicit,
coded, and Iormally egalitarian juridical Iramework, made possible by the organization oI a parliamentary, representative regime. But the development and generalization oI
disciplinary mechanisms constituted the other, dark side oI these processes. The general juridical Iorm that guaranteed a system oI rights that were egalitarian in principle was
supported by these tiny, everyday, physical mechanisms, by all those systems oI micropower that are essentially nonegalitarian, and asymmetric that we call the disciplines. And
although . . . the representative regime makes it possible . . . Ior the will oI all to Iorm the Iundamental authority oI sovereignty, the disciplines provide, at the base, a guarantee oI
the submission oI Iorces and bodies. The real, corporeal disciplines constituted the foundation of the formal, furidical liberties.
8
(Our emphasis)
Thus, Foucault also looks behind the juridical relations oI liberal democratic regimes and an apparently egalitarian market society to the systematic (nonaccidental)
Iorms oI domination within society. Indeed, a central concern oI Foucault's project is to dispose once and Ior all with what he calls "the juridical model oI power" that
still dominates our thinking, in order to direct our attention (and resistance) to the subtle yet pervasive Iorm oI power typical oI modern societies that escapes
articulation in juridical terms.
9
Since the Iate oI the categories oI civil society is bound up with the contrast he sets up between the two models oI power, it is well
worth looking into them.

Page 259
According to Foucault, the juridical model oI power and the legal ediIice oI our own society are inherited Irom the ancien regime. The revitalization oI Roman law
begun in the twelIth century, together with the discourses oI sovereignty, legitimacy, and rights, played a constitutive role in establishing the absolute power and
authority oI the monarchy. Right, according to Foucault, is, in the West, the king's right. Even when the juridical discourse turns against the monarch's control (in the
name, Ior example, oI preserving Ieudal rights or oI establishing individual rights against the state), it is always the limits oI this sovereign power that are put in question,
its prerogatives that are challenged. Whether the juridical discourse oI right was aimed at limiting or assuring the absolute character oI the king's power, its aim was to
constitute power as his right. "The essential role oI the theory oI right, Irom medieval times onwards, was to Iix the legitimacy oI power; that is the major problem
around which the whole theory oI right and sovereignty is organized."
10
Sovereignty, in short, is deIined in juridical terms, while law constitutes power as the legitimate
right oI sovereignty.
In part, oI course, this juridical construction served to eIIace the domination intrinsic to power, making the latter appear as the legitimate right oI the sovereign and
involving the legal obligation to obey it. In part, it also served as the instrument and justiIication Ior constructing large-scale administrative monarchies. Accordingly, the
juridical does articulate the Iorm in which power was exercised under absolute monarchies, that is, the relationship between sovereign and subject.
11
Indeed, the
juridical model articulates a speciIic conception oI the ways in which power is exercised: It is based on a model oI power that operates through the mechanisms oI law,
taboo and censorship, limits, obedience and transgression.
Whether one attributes to it the Iorm oI the prince who Iormulates rights, oI the Iather who Iorbids, oI the censor who enIorces silence, or oI the master who states the law, in any
case one schematizes power in a juridical Iorm, and one deIines its eIIects as obedience. ConIronted by a power that is law, the subject who is constituted as subjectwho
is"subjected"is he who obeys. . . . A legislative power on the one side, an obedient subject on the other.
12


Page 260
In short, the model oI power corresponding to the juridical is repressive. Accordingly, power appears to be "strangely restrictive": It is poor in resources, sparing oI its
methods, monotonous in the tactics it utilizes. The only Iorce it has is the Iorce oI the negative, a power to say noit posits limits, it does not produce. This power is
incapable oI doing anything except preventing what it dominates Irom doing anything but what it is permitted to do. As such, sovereign power is indeed limited, insoIar
as it involves the right over liIe and death only vis-a-vis the exercise oI the right to kill or reIrain Irom killing, to let live or to take liIe. It is no accident that the symbol oI
such power is the sword, Ior the juridico-political model oI power was indeed exercised as a means oI deduction, a subtraction mechanism, as a right to appropriate a
portion oI wealth, a tax on products, goods, services, labor, and blood levied on the subjects. Such a Iorm oI power silences, represses, Iorbids, takes, seizes, but that
is all.
13
Needless to say, it is Foucault's central thesis that the new type oI power that began to develop in the seventeenth and eighteenth centuries and became globalized and
perIected in the nineteenth and twentieth is incompatible with the relations oI sovereignty and is in every aspect the antithesis oI the mechanism oI power described by
the theory oI sovereignty. The new type oI disciplinary power, one oI the great inventions oI bourgeois society (sic),
14
is irreducible to the representation oI law: The
juridical cannot serve as its system oI representation.
15
Nor can the model oI repression account Ior the mode, techniques, or exercise oI this Iorm oI power.
Nevertheless, this model continues to hold sway today, in part as an ideology oI right, in part as the organizing principle oI the legal codes Europe acquired in the
nineteenth century.
16
It remains hegemonic in the Iield oI political theory, inIorming both the liberal and the radical democratic versions oI contractarianism.
Indeed, in a way quite reminiscent oI Marx (and Carl Schmitt), Foucault scoIIs at liberal political theory that sees in the universal legalisms oI society (in Iormal
equality, rights, and parliamentary democracy) limits imposed by a Iree societal community (composed oI sovereign individuals) on the exercise oI power. The
contractarian illusion that power can be made visible, localized, and restricted to the political state whose boundaries are clearly

Page 261
delimited by the rights oI juridical subject, oI course had a role to play in the construction oI the model oI parliamentary democracy in opposition to the administrative,
authoritarian, absolutist monarchies. But it remains prisoner oI the juridical model oI power Iirst erected by these monarchies: Eighteenth-century contractarian criticism
oI the monarchy was not aimed against the juridical system but rather spoke in the name oI a purer and more rigorous legality to which all the mechanisms oI power
would conIorm. ''Political criticism availed itselI, thereIore, oI all the juridical thinking that had accompanied the development oI the monarchy, in order to condemn the
latter; but it did not challenge the principle which held that law had to be the very Iorm oI power, and that power always had to be exercised in the Iorm oI law."
17

Neither the Rousseauian, radical democratic transposition oI sovereignty Irom the king to the people nor the liberal idea oI rights antecedent to government transcends
the juridical conception oI power, the doctrine oI sovereignty, or the concern with legitimacyboth assume that the rule oI law and the codiIication oI rights render
power legitimate and controllable. Both discuss power in terms oI the state, sovereignty, consent, contract, and rights, implying that power is visible, localizable in one
place, limitable, and to be exercised in accordance with a Iundamental lawIulness.
The very idea oI a contract among individuals that establishes legitimate power by limiting it through law and rights construes power as an original right oI sovereignty
that is given up, when political society is established, to the artiIicial sovereign. This model construes oppression as the transgression oI the limits oI the terms oI the
contract. The right to rebel against power that has transgressed its limits, thereby violating the rights oI another, is the right to reestablish legitimate, juridically bound
power. Accordingly,
the representation oI power has remained under the spell oI monarchy. In political thought and analysis, we still have not cut oII the head oI the king. Hence the importance that
the theory oI power gives to the problem oI right and violence, law and illegality, Ireedom and will, and especially the state and sovereignty (even iI the latter is questioned insoIar
as it is personiIied in a collective being and no longer a sovereign individual). To conceive oI power on the basis oI these problems is to conceive oI it in terms oI a historical Iorm
that is characteristic oI our societies: the

Page 262
juridical monarchy. Characteristic yet transitory. For while many oI its Iorms have persisted to the present, it has gradually been penetrated by quite new mechanisms oI power
that are probably irreducible to the representation oI law.
18
Foucault's point is, oI course, that this model oI power is anachronistic. But why is it still accepted? Apart Irom the historical reasons mentioned above, Foucault
mentions three other roles that the juridical plays in modern society. The Iirst is clearly ideological, despite Foucault's rejection oI the notion oI ideology. For he states
many times that the discourse oI law and rights masks the operations oI power by diverting us Irom attending to the newly emerging discourses oI the disciplines
themselves, and by concealing the mechanisms oI disciplinary power that operate outside, underneath, and through the law. It orients us, in other words, to questions
oI legitimacy and illegitimacy rather than issues oI struggle and submission, to relations oI sovereignty rather than domination:
The theory oI sovereignty, and the organization oI a legal code centered upon it, have allowed a system oI right to be superimposed upon the mechanisms oI discipline in such
away as to conceal its actual procedures, the element oI domination inherent in its techniques, and to guarantee to everyone, by virtue oI the sovereignty oI the State, the exercise
oI his proper sovereign rights.
19
In reality, the disciplines have their own discourse, which is not that oI norms but oI normalization. The discourse oI rights conceals the Iar more important disciplinary
discursivity. Here, the relation oI the discourse oI right to actual power relations is one oI Iorm and content. Modern society, then, Irom the nineteenth century up to
our own day, has been characterized, on the one hand, by a legislation, a discourse, and an organization based on public right, whose principle oI articulation is the
social body and the delegative status oI each citizen; and, on the other hand, by a closely linked grid oI disciplinary coercions whose purpose is in Iact to assure the
cohesion oI this same social body. The Iormer anachronistic yet useIul normative discourse oI right and sovereignty disguises the new power relations oI modernity.
20


Page 263
Foucault does, oI course, discuss a new, modern development oI the discourse and organization oI law and right. But, as Habermas has pointed out, the
reorganization oI right that Foucault stresses has nothing to do with normative developments internal to law since the eighteenth century or with the explosion oI civil
rights in our century.
21
Not only does Foucault entirely neglect the development oI normative structures in connection with the modern Iormation oI power, but his
discussion oI "the juridical" as integral to "Ieudal monarchic" power misses the diIIerences between the old conception oI privileges and the modern conception oI
rights. Indeed, he seems to believe that the modern structures oI right that are constitutive oI the various domains oI civil society and oI the new relation between
citizens and the public sphere are essentially the same as under absolutist regimes. Apparently, we are to conclude Irom his analysis that concern with the procedural
principles oI democratic legitimacywith civil, political, and social rightsin short, with constitutionalism, is a relic Irom the period oI absolutism:
22
We have entered a phase oI juridical regression in comparison with the pre-seventeenth century societies we are acquainted with; we should not be deceived by all the
Constitutions Iramed throughout the world since the French Revolution, the Codes written and revised, a whole continual and clamorous legislative activity: These were the Iorms
that made an essentially normalizing power acceptable.
23
II right serves solely to establish the legitimacy oI sovereign power, simultaneously concealing domination, then Foucault's strategy seems to be the demystiIication oI
Iormer in order to make the latter visible.
But the discourse oI rights and the juridical conception oI power have another Iunction. It is not only the ideological cover Ior a new Iorm oI domination, but is
constitutive oI the latter: "The system oI right, the domain oI the law, are permanent agents oI these relations oI domination, these polymorphous techniques oI
subjugation. Right should be viewed, I believe, not in terms oI a legitimacy to be established, but in terms oI the methods oI subjugation that it instigates."
24
As
indicated earlier, the real, corporeal disciplines constitute the Ioundation oI the Iormal, juridical liberties.

Page 264
Indeed, what the new, nonanachronistic (nonnormative) development oI the juridical discourse and Iorm entails is its "colonization" by the procedures oI normalization,
by the empirical disciplines Irom sociology and medicine to psychology. Individual rights, individualizing law, and the penetration oI the old normative structure oI law
by the disciplines turn law itselI into an eIIective medium oI, and a partner in, the disciplining, normalizing techniques oI domination, despite the ultimate heterogeneity oI
the levels oI discipline and sovereignty.
25
Moreover, it is precisely these nonnormative developments within law and legal discourses that implicate it in the modern
structure oI power. The use oI medical, psychological, sociological expertise, oI statistical data, in short, oI empirical inIormation and nonlegal languages within legal
discourse to make one's case, is prooI that the disciplines have penetrated the juridical structures and rendered them positive, empirical, Iunctional, and quasi-
disciplinary themselves. Thus, law does not necessarily Iade into the background in the nineteenth and twentieth centuries, but it now operates more and more in the
service of normali:ation as the juridical institution is incorporated into a continuum oI apparatuses (medical, administrative, etc.) whose Iunctions are Ior the most
part regulatorv.
26
The idea oI the constitutive role oI law vis-a-vis subjugation is evocative oI the old Marxian-Iunctionalist critique oI rights and juridical Iorms. Here, too, the juridical
structures are constitutive oI the modern modality oI power, and the juridical subject appears not as the limit to, but as the eIIect oI, power. The analogy with the labor
contract as the legal Iorm that encodes, conceals, and constitutes the asymmetric power relations in the sphere oI production is strong indeed. However, Ior Foucault,
the modern Iorms oI power do not contradict or violate the egalitarian norms oI civil society but are, rather, their Ioundation. This normalizing Iunction oI an
increasingly positivist and empirical conception oI colonized law is quite absent in Marx. Hence, unlike some versions oI Marxism, Foucault argues that the normative
principles oI civil society cannot serve as the reIerent Ior a critique oI domination or provide valid orientations Ior social movements that might seek to realize them
more Iully. InsoIar as they remain normative, the principles oI right, the rule oI law, legitimacy, etc., are anachronis-

Page 265
tic; insoIar as law becomes colonized by the disciplines and, as it were, empirical, it serves domination. In short, Foucault explicitly rejects the path oI immanent
critique.
The third reason Ior the persistence oI the juridical model oI power given by Foucault is quasi-psychological. The contract model construes power as a mere limit on
one's desires or Ireedomlegitimate power is itselI limited vis-a-vis rights and Ireedoms that are reserved to the people. On this model, we remain Iree to do what the
law does not proscribe. Power as a pure limit on Ireedom implies that a measure oI Ireedom (negative liberty) remains intact. Indeed, this is the general Iorm oI its
acceptability in our society. Thus, the social-psychological explanation oI the seductiveness oI the juridical model oI power is predicated on the Iact that "power is
tolerable only on condition that it mask a substantial part oI itselI. Its success is proportional to its ability to hide its own mechanisms. Would power be accepted iI it
were entirely cynical?"
27
While this explanation sounds suspiciously like a theory oI legitimation, Foucault would reject such an interpretation. The juridical model oI power is not the legitimating
discourse oI disciplinary power but a diversionary tactic; the discourses oI the disciplines are quite diIIerent. While we shall show that, despite his disclaimers, Foucault
needs a theory oI legitimation and in statements like the one above does indeed bring the concept back into his Iramework, this is hardly the way he would wish to be
interpreted. Far Irom leading us to analyze questions oI legitimacy, consent, sovereignty, and obedience, he wants to steer us in the opposite direction, to make us look
directly at domination/subjugation in its material instances, in its positive real Iorms and techniques. Indeed, the entire preoccupation with the normative distinction
between legitimate and illegitimate power, questions oI justice, the discourse oI rights, and so on, must be abandoned and replaced with a reverse mode oI analysis,
one that starts with the microtechniques oI domination in the local disparate regions oI society rather than a conception oI sovereign power, the state, and legitimacy.
28
For this, however, a diIIerent concept oI power is needed. II the juridical model was useIul Ior representing a power centered around deduction and death, it is "utterly
incongruous with the

Page 266
new methods oI power whose operation is not ensured by right but by technique, not by law but by normalization, not by punishment but by control, methods that are
employed on all levels and in Iorms that go beyond the state and its apparatus."
29
We are oIIered an analysis oI this new modern type oI power in Discipline and
Punish, then in the series oI essays collected in Power/Knowledge, and Iinally in the Iirst volume oI The Historv of Sexualitv. Unlike the juridical model, which
conceives oI power as something that is possessed by an individual or a group, that is exchangeable and recoverable, subject to legal limits and dissolved by
knowledge, truth, and authentic discourse, this disciplinary, normalizing power is conceived oI above all as a relation oI Iorces: It is exercised, not exchanged, and it
operates through an intimate association with discourses oI truth and the production oI truth. Accordingly,
Power must be analyzed as something which circulates, or rather as something which only Iunctions in the Iorm oI a chain. It is never localized here or there, never in anybody's
hands, never appropriated as a commodity or a piece oI wealth. Power is employed and exercised through a net-like organization. And not only do individuals circulate between its
threads; they are always in the position oI simultaneously undergoing and exercising this power.
30
Moreover, Iar Irom being localized in one macroinstitution such as the state, power is coextensive with the social bodythere are no spaces oI primal liberty between
its meshes. Rather, the relations oI power are interwoven with other kinds oI relations, including production, kinship, Iamily, knowledge relations, sexuality, and the
like. Power relations are, as it were, the immediate eIIects oI the divisions, inequalities, and disequilibria that occur in the latter, and, conversely, they are the internal
conditions oI these diIIerentiations. While power relations are sui generis,
31
emerging in dispersed, heterogeneous, localized arenas and exercised through a range oI
"microtechniques," they can be integrated into more global strategies and serve, Ior example, economic or state goals.
In short, Foucault replaces the juridical conception oI power with a strategic model oI a hostile asymmetric relation oI Iorces.
32
Power is everywhere, not because it
embraces everything, but because it comes Irom everywhere.
33
In addition, modern power is not exer-

Page 267
cised through prohibition and negation. Rather, it operates through a multiplicity oI technologies oI control, sorting, surveillance, and interrogation that are
productiveoI new discourses, knowledge, and truths, oI new kinds oI individuals or subjects, oI required behaviors and Iunctional results. Power relations are both
intentional and nonsubjective, based on calculation and clearly decipherable logic and aims that are nonetheless anonymous.
34
Finally, there are no relations oI power
without resistances Iormed at the point where relations oI power are exercised.
While this conception is certainly more proIound than that oI the liberal-legalistic model, Foucault is not the only one to view power in this way. One can Iind a not
dissimilar positive-sum conception oI power in the work oI both Talcott Parsons and Niklas Luhmann.
35
However, Foucault does provide a compelling and detailed
analysis oI the two main Iorms in which this model oI power came to be exercised, as well as a unique thesis regarding the relationship among knowledge, power, and
truth that these entail. Disciplinary-normalizing power, geared to the subjugation oI bodies and exercised through an "anatomo-politics oI the human body," is analyzed
in depth in Discipline and Punish. The Historv of Sexualitv, on the other hand, Iocuses on regulatory-productive biopower, oriented to the control oI populations
their health, liIe expectancy, and longevityexercised through a "biopolitics oI the population." While not identical, these two Iorms oI power, emerging in the
seventeenth and eighteenth centuries respectively, constituted the two poles around which the productive organization oI power over liIe was deployed.
36
Each one
developed a speciIic range oI techniques, type oI discursivity, and knowledge, and each resulted in a speciIic product: the soul, the docile body, and man in the Iirst
case; the desiring individual and sexuality in the second.
The new human sciences oI criminology, medicine, psychology, education, sociology, education, etc., come together with the new techniques oI surveillance,
examination, sorting, individualizing, and normalizing to constitute discipline. It is through disciplines/disciplining that the body is diminished as a political Iorce at the
least cost and maximized as a useIul Iorce.
37
These Iorms oI knowledge and power techniques also constitute the soul as the product oI the judgmental gaze oI
teachers, doctors, educators,

Page 268
prison guards, and social workers. The eIIect oI disciplinary power/knowledge is thus man as knowable, calculable, normal, useIul.
Biopower also operates through discursivity, produces new types oI individuals, and results in knowledge that is linked to a regime oI power. The discursive explosion
with regard to sexuality that erupted in the eighteenth century and constituted individuals as desiring subjects also made use oI techniques that emerged in disparate
settings. Foucault cites the conIessional techniques developed within monasteries and perIected by psychology and the gathering oI statistical inIormation by the police
on the wealth, manpower, productive capacity, and health oI the population. He also discusses the corresponding human sciencesespecially demography, medicine,
biology, psychiatry, psychology, ethics, pedagogy, and urbanologythat Iocused on birth and death rates, liIe expectancy, Iertility, and patterns oI diet and habitation
and instigated a ceaseless discussion about the details oI sexual behavior. These new Iorms oI knowledge constitute the people as a population to be regulated and
controlled in the name oI increasing its liIe, productivity, wealth, and utility. They also constitute the individual as a desiring sexual being whose secret longings must be
Ierreted out, made to speak, and channeled in the proper (useIul) direction through processes oI selI-interrogation aided, oI course, by experts. Thus, sex stands at the
center oI the new techniques oI liIe. Here, too, what is at issue "is the type oI power it brought to bear on the body and sex. In point oI Iact, this power had neither the
Iorm oI the law nor the eIIects oI the taboo. On the contrary, it acted by multiplication oI singular sexualities . . . it extended the various Iorms oI sexuality."
38
The new
sexualities that appearinIantile sexuality, the perversions, the hysterical womanand that haunt the spaces oI the home, the school, the prison, "all Iorm the correlate
oI exact procedures oI power."
39
In the process sex itselI becomes constituted as a problem Ior truth and the target oI an immense (medical/psychological) apparatus
Ior producing the truth about ourselves.
These analyses oI the Iorms oI modern power relations are both instructive and compelling. What is questionable, however, is the theoretical presuppositions oI the
genealogical method oI analyzing power and its implications Ior a theory oI modernization and

Page 269
oI modern civil society. Since our main interest is the latter, we shall only brieIly touch on the Iormer.
The Genealogy of Modern Civil Society
1he Philosophical and Aormative Ambiguities of Cenealogy
The philosophical presuppositions oI what Foucault calls humanism serve as the main contrast to his own genealogical approach. The idea that there is a human soul or
selI, subjectivity, an inner human nature (either as a desiring sexual being or as an autonomous sovereign subject), or an essence oI man that is universal, that can serve
as the ground oI the basic values oI autonomy, equality, Ireedom, and liIe, and that disinterested knowledge can express and liberate is rejected by Foucault in his
masterIul critique oI the very concept oI man in The Order of Things. Both the subject/object duality and Ioundationalist assumptions at the core oI humanism lead to
unresolvable antinomies. But this is not all. The genealogy oI the modern soul presented in Discipline and Punish goes beyond the philosophical critique to reveal that
the very notions oI subjectivity, the soul, the selI, autonomy, and normativity (always interpreted as normalization) are the products oI disciplinary power/ knowledge.
40
Accordingly, Foucault warns us against the misconception that knowledge can exist independently oI the interests oI power, or only where power relations are
suspended. There is, on the genealogical analysis, no knowledge that does not presuppose and at the same time constitute power relations.
41
The human sciences, the
disciplines, yield the objective knowledge oI man, the soul, the subject, and the individual required by disciplinary power.
The same holds true Ior the subject side oI the equation and Ior discursivity. The idea that intensive selI-interrogation and speaking or communicating the truth that we
have discovered about ourselves to others is the road to selI-mastery, authenticity, and liberation Irom repression is as naive as the idea oI disinterested objective
knowledge. Far Irom dissolving the eIIects oI power, the authentic individual who speaks the truth about herselI, her desires, needs, identity, innermost concerns, is the
product oI conIessional power techniques. The genealogical account oI sexuality aims to show

Page 270
that the hermeneutic subject is the historical product oI a power/ knowledge regime that Iunctions in and through discourse. Genealogy analyzes the discursive
techniques oI the constitution oI subject-selves who probe their own depth (through selI-interrogation) and speak/conIess the truths discovered thereby. The rituals oI
conIessional discourse involve the actual or virtual presence oI a partner who stands as the authority prescribing and appreciating the conIession, judging, punishing or
Iorgiving, and consoling the person who articulates it.
42
Thus, not only the objectiIying discourses oI the social sciences, but also the subjectiIying discourses by and
about ourselves, are, on the genealogical analysis, deeply connected to power. Far Irom having an aIIinity with Ireedom, or with universality, reason, and truth, they are
imbued with relations oI power and are always historically speciIic. Accordingly, Foucault sees his genealogical investigations as part oI a "political history oI truth."
43

Knowledge, truth, reason, and power are intertwined and context-relative; genealogical investigations into Iields oI knowledge, types oI normativity, Iorms oI
subjectivity, individual and collective identities reveal the technologies oI power whereby truth, knowledge, and identity are produced.
The problems with this Nietzschean stance vis-a-vis norms, reason, and truth have been pointed out many times. We shall mention only a Iew oI the most Irequent
objections that bear on the normative dimensions oI our concept oI civil society.
First, there is the problem oI the normative ambiguity oI Foucault's genealogical account oI normativity. Are we to take this to be one among many critiques oI the
Ioundationalist metainterpretation oI humanist values, or is it aimed at the substantive core oI these values themselves? II the latter is the case, and the other aspects oI
genealogy to be explored below indicate that it is, then Foucault is leIt in the paradoxical position oI having to deny any normative status Ior his own critical analyses or
oI being unable to justiIy the normative political implications oI his work.
44
Second, iI one takes Foucault at his word regarding the powerrelatedness oI truth, then the obvious question arises: What is the status oI the "truths" revealed by
Foucault's own genealogical investigations? Which interests, what strategies, what Iorm oI power relations does Foucault stand Ior?
45


Page 271
Third, isn't the claim that all knowledge and rationality itselI derive Irom the practices oI power based on an undiIIerentiated concept oI power? Are all power relations
the same? What exactly is the diIIerence between power and domination, iI there is any?
46
The obIuscation in Foucault's concept oI power lies, according to
Habermas, in its concealed derivation Irom the concept oI the will to knowledge.
47
And this, in turn, rests on an ambiguous use oI the category ''power." As
Habermas points out, Foucault's use oI the concept oI power reproduces the "transcendental-empirical ambiguity" that he relentlessly uncovers in the humanist
conception oI man: On the one hand, it is used descriptively in the empirical analyses oI power technologies; on the other, it is a basic concept within a theory oI
constitution.
48
The Iirst explains the Iunctional social context oI the sciences oI man; the second, the condition oI possibility oI scientiIic discourse about man. But
doesn't the genealogical approach that claims to do both oI these at once simply replace the objectivism oI the human sciences with a radically historicist subjectivism?
49
And doesn't the transcendental-empirical ambiguity in his concept oI power lead Foucault to overgeneralize and even ontologize power relations?
50
Moreover,
doesn't the equation oI reason, knowledge, and discourse with the rationality oI domination derive Irom this ontologizing oI power and involve a reductionist, one-
sided, strategic-instrumental conception oI reason itselI?
51
A Iourth objection to genealogical assumptions is that the concept oI the relativity oI truth to a (power) regime is ultimately incoherent. On the relativity thesis, the
transIormation Irom one regime to another cannot yield a gain in truth, nor can there be liberating transIormations within a regime. There is no such thing as truth
independent oI its regime, since each regime produces its own truth. But what, then, is the meaning oI Foucault's claim that the truth manuIactured by power is its
mask, disguise, that is, untruth?
52
Does one untruth simply cover over another? Or is the discourse oI the disciplines truer than the juridical discourse?
FiIth, and Iinally, doesn't the very notion oI power relations as Foucault uses it, namely, always with the qualiIication "inegalitarian," imply domination, and isn't this
concept meaningless without its opposite, Ireedom?
53
Moreover, even iI we grant the idea oI

Page 272
power without a (global) subject, even iI we recognize that there is always a strategic context in which power relations are embedded and which is not under the
control oI the actors, does it make sense to speak oI strategies oI power without projects, or oI society in terms oI anonymous relations oI Iorces?
54
Foucault's
insistence that power relations are inegalitarian and intentional, that there is no power without resistance, implies at least that there are speciIic interests involved in
exercising and maintaining power, and speciIic victims whose interests lie in overturning power relations. But whose interests are involved in the development and
maintenance oI disciplinary-regulatory modern power relations? Once these are in place, how is resistance possible in the carceral civil society, and in the name oI
what does one resist? It is time to turn to these questions.
1he Cenealogical Account of Modernization
According to Foucault, the historical processes that constituted the social sphere in which the modern individual lives have deprived the ideal oI the autonomous
sovereign subject oI any progressive content
55
and have denuded social institutions oI any autonomous solidarity or horizontal relations. Neither the concept oI the
individual nor the norms, structure, or dynamics oI civil society can be understood as a gain in Ireedom or serve as a reIerent Ior emancipatory politics. We shall return
to this theory oI modern individuality and sociality, but Iirst it is worth looking brieIly at the historical "genealogy" oI modern society that is clearly meant to replace the
materialist theory oI history and deprive critics oI its reassuring dialectic.
Discipline and Punish, the Iirst book in which Foucault presents his theory oI power, also provides the clearest statement oI his genealogical theory oI modernization,
that is, oI the transIormation involved in shaping our contemporary "carceral," "disciplinary" society.
56
Although the book Iocuses on the genealogy oI the modern
prison, it is clearly meant to be taken as exemplary Ior a wide range oI homologous changes that characterize the transition Irom the "classical age" (the age oI
absolutism or, more generally, the ancien regime) to modern society (late eighteenth century to

Page 273
the present).
57
For it is Foucault's thesis that the asymmetric power relations and the techniques oI learning about and disciplining bodies that were perIected in the
prison now pervade an ever broader range oI contemporary societal institutions and aIIect everyone. Indeed, "the carceral archipelago transported this technique Irom
the penal institution to the entire social body."
58
Thus, the genealogy oI the modern prison reveals a modality oI power that is all-pervasive in modern civil society.
The innovations in Foucault's genealogical account oI modernity do not lie in the speciIic epochs outlined in the historical trajectory he traces.
59
These epochs are quite
standard in modernization theory. We are, in short, presented with descriptions oI two societal types and a transitional period between them: traditional society or the
"ancien regime," composed oI the society oI orders and the absolutist state
60
(seventeenth to nineteenth centuries), and modern society, emerging in the eighteenth
century and developed throughout the twentieth. The transitional period is dealt with by analyzing the theories oI the Enlightenment and the reIormers' discourse
preceding and during the French revolution.
Nor is Foucault's assessment oI these changes in terms oI a replacement oI one Iorm oI domination with another particularly new or shocking, despite the challenge it
poses to standard liberal accounts (contractarian or enlightenment theories). Indeed, at Iirst sight, the similarities oI Foucault's approach with at least one important
stream within sociological theories oI modernization are striking.
61
The red thread oI Foucault's text is the theme oI the emergence oI the modern individual as the
story oI a new and pervasive Iorm oI domination evolving through two interrelated processes: the destruction oI traditional group solidarities and the Iragmentation or
leveling oI peoples, orders, and coherent social groups; and the consolidation oI disciplinary techniques oI surveillance and control oI bodies that Iabricate a new Iorm
oI individuality whose illusion oI sovereignty is the counterpart to the absence oI any autonomous group liIe or group identity, meaningIul traditions, Iorms oI
association, or power resources. The only serious diIIerence on the level oI content between this version oI modernization and that oI Tocqueville or Nisbet, Ior
example, is that the latter attribute the leveling, individualizing Iorm oI power prima-

Page 274
rily to the emergence oI the modern state, while Foucault sees it as the result oI a multiplicity oI institutional Iorces or developments in society, economy, and polity.
For both (as well as Ior Marx), however, modern civil society is nevertheless that colonized terrain where solidarity, association, group autonomy, and spontaneity
have been replaced by a new Iorm oI social control.
To be sure, Foucault does not describe the contrast between the old regime and modern society in order to idealize the intermediary political bodies oI the
Stàndestaat, which represented, Ior Tocqueville, at least, the crucial loci oI political liIe that limited the administrative power oI the state.
62
Indeed, there is no
systematic distinction in Foucault's work between the type oI political action within the Iramework oI assemblies and the state action typical oI administrative power
relations.
63
It was precisely this sort oI distinction, however, that led Tocqueville to seek modern equivalents Ior the old Iorms oI association, autonomy, and
counterpowera search that is doomed to Iailure on Foucault's theory.
64
Nor does Foucault assign positive value to the cultural traditions or the integrative Iunctions ensured by the old intermediary bodies (as Nisbet did).
65
On the contrary,
it is the opportunities Ior disorder in the interstices oI the society oI orders, during the absolutist period, Ior which Foucault seems to be nostalgic. Thus, what is
pinpointed (and somewhat idealized) in Foucault's contrast between traditional and modern society is neither the political liIe oI the aristocracy nor the richly textured
and communally integrated traditions oI the social orders or semiautonomous regions, but the incomplete control, regulation, organization, and disciplining oI society in
the premodern period and the spaces Ior solidarity and spontaneous rebellion that this created. It is this relative absence oI eIIicient control that contrasts so sharply
with the inexorable organization, discipline, and surveillance techniques oI modernity. And it is here that the originality oI Foucault's treatment lies.
66
Foucault's thesis is that the speciIic nature oI the exercise and modality oI absolutist power encouraged the emergence oI popular revolts. This thesis is demonstrated
through an analysis oI the Iorm and meaning oI punishment in absolutist regimes. On the one hand, the "supplice" or public torture and execution oI the

Page 275
criminal symbolize the absolute power oI the sovereign to codiIy the lack oI power oI his subjects.
67
Publicity, visibility, and the light oI appearance are all the
exclusive attributes oI the sovereignthe means to express and represent his personal power and his monopolization and control oI the public space. Sovereign
power, as indicated above, is a mixture oI repression and juridical controlthe sovereign is he who makes, and thereIore is above, the law.
68
The discourse oI rights
here is the discourse oI this poweroI jurisdiction and immunity.
69
The power oI the sovereign is the power to silence, banish, punish, and annihilate those who
transgress his law. Crime is seen as an attack on the will and body oI the omnipresent sovereign, that is, as an act oI war or treason.
70
Punishment, as the ceremony oI
sovereign power that marks the body oI the oIIender, restores and reconstitutes sovereignty. It reveals the Iorce, terror, and vengeance oI a power that is personal and
arbitrary, that is made public through its periodic expenditure, yetand this is the keyis discontinuous in time and space.
Discontinuous in two senses. First, within the Iramework oI the society oI orders, the phenomena oI rights and immunities (in Foucault's terminology, illegalities)
constitute a source oI counterpower and autonomous group solidarity Ior the privileged, signiIying the nonpervasiveness and incompleteness oI sovereign power. But
Foucault is Iar more interested in another type oI discontinuity or "illegality," namely, that oI the least Iavored stratumthe people. The lower orders had no positive
privileges, but they beneIited Irom a space oI toleration gained "by Iorce or cunning" in which illegality, or the possibility oI acting outside oI, or oI ignoring, law and
custom was regularly practiced: "Roughly speaking, one might say that, under the ancien regime, each oI the diIIerent social strata had its margin oI tolerated illegality:
the nonapplication oI the rule, the nonobservance oI the innumerable edicts or ordinanceswere a condition oI the political and economic Iunctioning oI society."
71
This
tolerance oI illegality was a sign not oI sovereign beneIicence but oI the discontinuity oI monarchic power. It was tied to the relatively weak penetration oI the social
body by this power and, correspondingly, to the existence oI spaces within society Ior the emergence oI autonomous solidarities and revolts.

Page 276
Indeed, next to the monopoly oI publicity and action by the only real individual, the sovereign, there was another Iorm oI action and publicity available to the people,
namely, the riot and the revolt. This is the other side oI the supplice. The necessary presence oI the people at public executions provided the occasion Ior constituting
centers oI illegality in the very exercise oI sovereign vengeance. The spectator, the guarantor oI punishment, could, in other words, turn rebel and challenge punitive
power.
72
It is here, in the carnivallike inversion oI rules, in the mockery oI authority, and in the transIormation oI the criminal into a hero,
73
that Foucault situates the
link between illegality, the spontaneous solidarity oI a whole segment oI the population (vagrants, the poor, beggars, etc.), and revolt. This spontaneity oI the
assembled populace is the uncontrolled and unmastered reIerent oI the very exercise oI sovereign power. Their resistance to central control is indicative oI still intact
local autonomy, cultural traditions, and moral resources Ior constituting collective identities and solidarities opposed to the sovereign's project oI monopolizing power.
These popular solidarities were gloriIied in the broadsheets and pamphlets meant to degrade them;
74
as spaces Ior popular illegalities leIt open by the discontinuous
Iorm oI sovereign power, they became the target oI the new, modern modality oI discipline and surveillance.
Foucault's description oI sovereign power is strikingly similar to Habermas's analysis oI prebourgeois repràsentative Offentlichkeit. Both Iocus on the public display
oI magniIicence and might, on the demonstrative dimension oI the excesses oI sovereignty, on the show oI Iorce as representative oI power, and on the codiIication oI
its monopoly by the sovereign. But an analysis oI the other side oI "public power" in the old regime, oI the "illegalities" and broadsheets oI the popular classes, oI the
interrelation between representative publicity and the publicity available to the people, is not to be Iound in Habermas's study. This is a major omission. Habermas,
conversely, analyzes two additional dimensions oI publicity within absolutist society that are strangely underemphasized in Foucault's account: the emerging
administrative apparatus oI the state characterized by the term "public oIIice," and the development oI the "bourgeois public sphere" in the caIIes, salons, literary clubs,
newspapers, and so on, oI the eighteenth century.
75
In Habermas's

Page 277
study, as we have seen, these preIigure important dimensions oI public Ireedom in modern society insoIar as the modern principle oI democratic legitimacy and the
conception oI public oIIice as public service, implying accountability, have their origins here.
Foucault is certainly aware oI the state-making processes under the old regime, but his emphasis is quite diIIerent Irom Habermas's.
76
Foucault points out that it was
the emerging centralized apparatus oI public administration that began gathering "useIul" inIormationdemographic data on births, deaths, health, crime, poverty,
welIare and so onon an increasingly leveled (Irom the state's point oI view) population, turning the sovereign's subjects into objects oI knowledge and power. This
knowledge was intimately connected with a new Iorm oI disciplinary power ("biopower") emerging within the administrative agencies oI the state alongside the juridical
discourses oI sovereignty and legitimacy.
Nevertheless, Foucault insists repeatedly that the new technologies oI power cannot be comprehended either through juridical concepts, as a relation between
sovereign and subjects, or in terms oI the opposition between state and society. For the state is not their sole or even primary source; rather, they emerged slowly in a
wide range oI institutions (the convent, the army, the clinic, the school, the Iactory, the prison) alongside the visible play oI sovereignties in the absolutist period. These
processes constitute Ior Foucault the birth oI the modern within the womb oI the old society. Accordingly, there is no need to emphasize the new Iorm oI the state as a
hierarchy oI public oIIices, nor to mention its counterpartthe new Iorms oI bourgeois publicity that emerge within civil society, with their speciIic projects oI
liberalization and democratization. The public, impersonal, rule-bound character oI state bureaucracies does nothing to diminish or restrict the reach or scope oI
administrative power; on the contrary, it makes it more eIIicient. And presumably the claims made Ior the bourgeois public sphere are suIIiciently dealt with as part oI
the reIormer's discourse. In our view, this is an error Iraught with consequences, Ior it is precisely the new Iorms oI publicity, association, and rights emerging on the
terrain oI modern civil society that will become the key weapons in the hands oI collective actors seeking to limit the reach oI state and other societal Iorms oI
disciplinary power.

Page 278
As a consequence oI the theoretical decision to restrict the concept oI sovereignty as a Iorm oI power to the old regime, Foucault in eIIect agrees with those reIormers
who Iocus exclusively on its juridical dimensions (on the proper locus oI power and on its legality and legitimacy), only to declare the entire discourse to be
anachronistic. The discussion oI rights, contract, popular sovereignty, and similar topics is, accordingly, nothing more than a Iacile inversion or imputation oI the king's
attributes to "the people." Instead oI the representative publicity oI the king's power, publicity, as claimed by reIormers Ior the people, is to be the expression oI their
newly acquired sovereignty and their mode oI limiting state power (i.e., the law). This discourse is epiphenomenal, however, insoIar as it occurs above the real loci oI
modern power relations.
77
The reIormers' discussion oI power in terms oI the state, sovereignty, consent, contract, and rights implies that power is still public,
localizable in one place, and limitable. The liberal juridical concept oI power, in other words, misses the essence oI the new mode oI domination. To Iocus on the
ediIice oI rights and publicity embodied in constitutions and parliaments, to stress the development and democratization oI the state, is to be deceived regarding the
real dynamics oI power in modern societies.
Accordingly, Foucault argues that the discourse oI reIorm in the transitional periodthe conception oI a transparent power that Iinds its legal limit in the notion oI
human dignity, that punishes humanely with a view toward restoring rather than destroying the integrity oI the criminal, together with the themes oI sovereignty, consent,
and legitimacyconstitutes a utopian model oI society that is never, nor could ever be, institutionalized. This discourse has, nonetheless, certain not so unintended
consequences. The most important oI these is the shiIt in the "right" to punish Irom the sovereign monarch to "society." Leniency oI punishment is indeed accomplished,
but with the corollary that crime is seen no longer as an attack on the sovereignty oI the monarch (i.e., the other) but rather as an attack on society as a whole (us),
turning the oIIender into a "public enemy" or monster who must be rehabilitated in order to reemerge as a juridical and moral subject. "The society that has
rediscovered its laws has lost the citizen who violated them."
78
In other words, once crime is seen as the violation oI society's own

Page 279
laws, the solidarity between the popular illegalities and the criminal is severed. Indeed, the destruction oI solidarity between the oIIender, the rebel, the criminal who
reIuses the law, and the population turns out to be the real target oI the reIormers' projects.
79
"The true objective oI the reIorm movement, even in its most general
Iormulations, was not so much to establish a new right to punish based on equitable principles, as to set up a new economy oI the power to punish . . . so that it could
be distributed in homogeneous circuits capable oI operating everywhere, in a continuous way, down to the Iinest grain oI the social body."
80
Inserting the power to punish more deeply within the social body could accomplish two things: control oI the popular illegalities, which had become too costly, and
development oI a more eIIicient economy oI power. The dimension oI the reIormers' projects that suited this goal most admirably was, oI course, the discovery oI the
advantages oI disciplinary technologies. Bentham's panopticon speaks more loudly Ior Foucault than all the theories oI legality, popular sovereignty, rights, and
legitimacy. "The 'Enlightenment' which discovered the liberties also invented the disciplines."
81
Thus, on the one side, the reIormers' discourse operates with
representation, visibility, publicity (oI trial and sentencing), and transparency oI the power to punish and oI the laws that deIine crimes and punishments appropriate to
them. On the other side, a disciplinary technology is discovered that involves secret, continuous, and autonomous punishment processesin short, a power that
operates on the other side oI legality, isolated Irom both the social body and juridical power. The juridical model reintegrates the juridical subject into society; the
technological practice creates obedient subjects and docile bodies. For Foucault, then, it is not to a new Iorm oI publicity, legislation, and legality that we must look to
Iind the seed oI the modern in the transition Irom the old to the new regime. Rather, we should look to the new technologies oI power developing in societal institutions
and articulated in reIorm projects. The discourses worth attending to are those oI the human sciences, which, together with the new disciplinary techniques, provide the
means Ior constituting, learning about, and controlling the modern individual. There is one important new Iorm oI "publicity" worth noting, but it is not that oI elections,
legislation,

Page 280
rights, courts, and the like. Rather, it is the visibility oI subjugated, individuated individuals beIore the eye oI a now invisible powera visibility at Iirst oI the inmate to
the supervisors oI closed institutions, but ultimately oI the deviant beIore society at large.
The genealogical approach to "modernization" thus discounts as hopelessly naive any interpretation oI the principles oI civil societylegality, rights, plurality,
publicityas a basis Ior the emergence oI spaces within modern society Ior new Iorms oI autonomous association and solidarity. Foucault's vision oI modern
disciplinary power as complete and continuous and his (at times Iunctionalist, at times constitutive) interpretation oI rights as the anchors oI this power keep him Irom
recognizing that, like the immunities in an earlier period, modern civil and political liberties also secure spaces Ior autonomy, Ior association, Ior solidarities, and Ior the
selI-constitution oI group liIe, new identities, and the development oI counterpowersthe sine qua non Ior the resistance to biopower that he nevertheless believes is
still possible. Moreover, Foucault's positivist attitude and his emphasis on the strategic dimension oI the reIormers' projects predispose him to view the new
disciplinary technologies as the "real" innovation next to which the normative and symbolic principles oI modern civil society appear as secondaryat best Iunctional
to, but ultimately irrelevant appurtenances oI, disciplinary power.
82
The above notwithstanding, Foucault's analysis oI the emergence oI modern society does not quite negate the thesis oI diIIerentiation as a key element oI
modernization. Indeed, his discussion oI the genesis oI the technologies oI power and their globalization within contemporary society presupposes diIIerentiation. As is
well known, Foucault insists that a multiplicity oI projects and interests came together to produce a new political economy oI punishment, discipline, and control. He
argues that Enlightenment philosophers and associated social groups contributed to this transIormation but that "it was not they alone; in this overall project oI a new
distribution oI the power to punish, and oI a new distribution oI its eIIects, many diIIerent interests came together."
83
Following Weber, Foucault argues that the
speciIic disciplinary techniques were discovered independently and locally in distinct institutions such as the monastery, the army, the Iactory,

Page 281
and the prison. OI course, multiplicity is not the same as diIIerentiation: Some oI these are state institutions, others are societal. However, Foucault does diIIerentiate
between state and society when it comes to identiIying the interests behind the globalization, iI not the genesis, oI the modern techniques oI power. Indeed, despite
disclaimers with regard to Marxian class theory and theories oI state power, the two sets oI interests involved in the globalization oI disciplinary-regulatory power turn
out to be those oI the bourgeoisie and the administrative state. Let us consider each oI these in turn.
Within the modernizing society oI the old regime, there is one major set oI interests behind the struggle against the arbitrary monarchic power and the society oI orders:
the concern oI the bourgeoisie to abolish popular illegalities, especially vis-ß-vis property rights.
84
According to Foucault, it was the need to protect accumulations oI
mercantile and industrial capital more than anything else that necessitated a severe repression oI popular illegality.
85
There emerged the need Ior a constant policing concerned essentially with this illegality oI property. It became necessary to get rid oI the old economy oI the power to punish,
based on the conIused and inadequate multiplicity oI authorities. . . . It became necessary to deIine a strategy and techniques oI punishment in which an economy oI continuity
and permanence would replace that oI expenditure and excess.
86
In short, penal reIorm was essential Ior a capitalist market economy to emerge and to Iunction; hence the struggle against the ''superpower" oI the sovereign, with its
incalculabilities, and against the "inIrapower" oI acquired privileges and tolerated illegalities. Accordingly, within the conIused sets oI interests and goals involved in the
transition Irom absolutism to modernity, Foucault stresses the importance oI that new, diIIerentiated structure, the capitalist market system, and its speciIic
requirements.
The class interests oI the bourgeoisie are also at stake in the development oI the second dimension oI the modern Iorm oI power: regulatory biopower. Here Foucault
explicitly rejects the neo-Marxian thesis that the sexuality oI the middle and especially lower classes had to be repressed because it was incompatible with

Page 282
a general and intensive work ethic.
87
"The primary concern was not repression oI the sex oI the classes to be exploited, but rather the body, vigor, longevity,
progeniture, and descent oI the classes that 'ruled.' "
88
SelI-aIIirmation and the need to diIIerentiate itselI as a class Irom the unhealthy lower orders and the degenerate
nobility are the interests at work in the investment oI its own sex with a technology oI power and knowledge that the bourgeoisie had itselI invented. In part, this
involved a transposition oI caste manners oI the nobility, based on blood, to the bourgeoisie in the guise oI biological, medical, or eugenic precepts Iocused on bodily
health, the indeIinite extension oI strength, vigor, and so on. Only later, in the second halI oI the nineteenth century, were the techniques oI regulatory biopower
generalized to the rest oI the population. That is, only aIter the need developed Ior a stable and competent labor Iorce and a secure technology oI control was in place
(through schooling, the politics oI housing, public hygiene, institutions oI relieI and insurance, the general medicalization oI the population), was the proletariat granted a
body and a sexuality and were middle-class values imposed upon them. This does not challenge the main claim, however, "that sexualitv is originallv, historicallv
bourgeois.
89
The place oI the state and its interests is somewhat more ambiguous in Foucault's analysis. On the one hand, the critique oI the sovereignty model was meant to steer
us away Irom the state as a central locus oI power or the key Iorce in creating disciplinary techniques. On the other hand, most oI the loci in which the technologies oI
disciplinary power did develop were (in France, Foucault's reIerent) state institutions: armies, schools, clinics, prisons, etc. Moreover, Foucault grants the immense
importance oI the development oI a centralized organization oI the police, "the most direct expression oI royal absolutism."
90
For it is the state police who take over
the previously Iragmented Iunctions oI surveillance oI criminality and economic and political supervision and uniIy these into a single administrative machine, assuring
continuity oI control. And this dimension oI state sovereignty, at least, endures with the transition to modernity. While Foucault insists that the state is not the sole origin
oI disciplinary power, he grants that "the organization oI the police apparatus in the eighteenth

Page 283
century sanctioned a generalization oI the disciplines that became coextensive with the state itselI."
91
In Iact, the major Iunction oI the state apparatus was to ensure
that discipline reigns over society as a whole.
92
This breadth oI discipline along with the continuity oI its exercise, potential or actual, are speciIic to modern
domination. What has been said oI the bourgeoisie can thus also be said oI the state: The newly diIIerentiated, centralized administrative state apparatuses also had an
interest in abolishing the old, incalculable, and expensive personal Iorms oI power and substituting its new techniques Ior them. The state, then, as a key actor in
generalizing disciplinary power, does play a major role in Foucault's account oI the transition to modernity.
The state's interests also play a central part in the globalization oI biopower. The beginning oI the eighteenth century saw a demographic upswing accompanied by an
increase in wealth and an end to the great ravages oI plagues and starvation; as a result, the societal preoccupation with death is replaced by a concern with managing
liIe and accumulating people. Accordingly, the state becomes interested in gathering inIormation about and controlling the health, wealth, manpower, resources,
reproduction, and welIare oI that new entity, "the population," as a means to increasing state power. InIormation gathering and supervision, involving a maximizing oI
collective and individual Iorces rather than a repression oI disorder, was, it now turned out, a natural Iunction Ior the police:
We must consolidate and augment, through the wisdom oI its regulations, the internal power oI the state; and since this power consists not only in the Republic in general, and in
each oI the members who constitute it, but also in the Iaculties and talents oI those belonging to it, it Iollows that the police must concern themselves with these means and make
them serve the public welIare. And they can only obtain this result through the knowledge they have oI those diIIerent assets.
93
The state's interest in the power-knowledge generated by the emerging disciplines Ior the purpose oI administering and optimizing the liIe and utility oI the populations
under its control is thus paramount in the globalization oI biopower.
94
Sex was at the heart oI this political economy oI population: "It was essential that the

Page 284
state know what was happening with its citizens' sex, and the use they made oI it,"
95
because power is situated and exercised at the level oI liIe, the species, the race,
and the large-scale phenomena oI population. Indeed, it is the new concern on the part oI the state with liIe and population that marks a society's "threshold oI
modernity," according to Foucault.
96
It seems to be Foucault's thesis, moreover, that by the late nineteenth century, the two Iorms oI powerdiscipline and the regulation oI populationsand the two
great interests behind their globalization came together. These techniques came "to reveal their political useIulness and to lend themselves to economic proIit. . . . All oI
a sudden, they came to be colonized and maintained by global mechanisms and the entire State system."
97
The apogee oI this development is, obviously, the
contemporary welIare state. Through its regulatory controls, the welIare state constitutes the social as a distinct object-domain oI great "public" interest while
simultaneously making use oI the disciplinary, conIessional techniques already perIected by societal disciplines and institutions to control it. On Foucault's account,
however, it is not the logic oI the economy or the state that penetrates and colonizes civil society. Functional reason, Ior Foucault, works the other way around: The
institutions and practices oI civil society generate the technologies oI power that are then taken up and globalized by the state and the bourgeoisie.
This should provide a clue to resolving the ambiguity we have noted concerning the place oI the state in Foucault's analysis oI power relations. Because he insists on
the decentralization and deinstitutionalization oI power, yet identiIies state apparatuses as key loci oI disciplinary-regulatory power, commentators have come up with
diametrically opposed interpretations oI the place oI the state in his overall analysis. Axel Honneth, among others, accuses Foucault oI ignoring the state altogether by
virtue oI his decentralized concept oI power.
98
Peter Dewes, however, asserts that, in analyzing the various disciplinary institutions oI the asylum, clinic, and prison,
"Foucault wishes to show that Irom the beginning intervention and administrative control have deIined the modern state."
99
According to Dewes, Foucault is
concerned to show that intervention in a societal domain by state agencies is a more

Page 285
Iundamental characteristic oI modern societies than an economy released Irom directly political relations oI domination.
100
For one interpreter, the state plays no role
at all in modern power relations; Ior the other, it is everything.
Foucault was questioned directly about this ambiguity. His response implied that the state, the economy, and society are three distinct elements within modern social
systems, each oI which has its power relations, disciplinary technologies, and modes oI Iunctioning.
101
Although the state (governmental administration)
102
does
become a coordinating center Ior societal disciplinary power, although its administrative agencies do penetrate social institutions, these nevertheless retain speciIic
internal power relations that have their own conIiguration and "relative autonomy."
103
The state, not qua sovereign but qua government,
104
does penetrate society, yet
"it would be wrong to believe that the disciplinary Iunctions were conIiscated and absorbed once and Ior all by a state apparatus."
105
In short, Foucault maintains that
the state cannot occupy the whole Iield oI power relations and can operate only on the basis oI already existing power relations connecting the Iamily, knowledge,
technology, the Iactory, sexuality, etc., to which the state relates as a superstructure. The state is one locus oI disciplinary technology among many.
We might note that, like the modality oI power he describes, Foucault's goal is to make visible not the state but society. And oI course he is right in insisting that power
relations are not exclusively located in, nor do they emanate Irom, any one place in modern society. Nevertheless, despite the elegance oI some oI his Iormulations, he
does not resolve the dilemma articulated by his interpreters; he seems, rather, to validate both antinomic positions. But iI the state is simply one locus oI disciplinary
power among others, then the very meaning oI the modern state is lost, Ior the term reIers to the diIIerentiated entity that succeeds in monopolizing the (legitimate)
means oI war and violence and, in nonIederal polities, oI administration as well. Such an "order" is hardly "one" among many. By using the thesis oI decentralized
power to deny state sovereignty, Foucault reproduces the position oI the philosophical pluralists (although Ior opposite reasons) and opens himselI to Carl Schmitt's
objection that a state that is like any

Page 286
other association or organization oI power in society is no state at all. II, conversely, the state is the coordinating mechanism oI disciplinary power, iI social institutions
are the necessary supports and complements oI state administration, iI within societal institutions one Iinds homologous Iorms oI domination, iI, in short, "society" is
equivalent to the Iield where administrative apparatuses have their play, then indeed the state, or at least its "logic" or modus operandi, is everywhere. But this is a
convincing idea only with respect to the symbolic meaning oI ''totalitarian" regimes.
106
Foucault is able to hold both positions because he sees state and society only Irom the point oI view oI strategic power relations.
107
Indeed, state, society, and
economy are presented as three strategic Iields with essentially the same internal dynamics and, as stated above, homologous technologies oI power. Modernity is not
characterized by a state that penetrates society or by socioeconomic powers that penetrate and control the state. Rather, it is constructed in terms oI the penetration oI
each distinct realm by disciplinary technologies oI power and strategic power relations. What this means is that state, economy, and society are diIIerentiated Irom one
another not in terms oI any speciIic rationality oI action, mode oI integration, or Iorms oI interaction but only, somehow, as separate sites oI power. This is a
diIIerentiation that seems to make no diIIerence.
108
The Negativity of Civil Society and the Loss of the Social
Foucault presents us with a deeply disturbing analysis oI the dark side oI modern civil society. As indicated above, Iar Irom constituting an "increment in
Ireedom" (Marx), the development oI the components oI civil society in modernitya new Iorm oI individuality, subjectivity, rights, plurality, publicity, legality, and
socialitynow appears as nothing but an eIIect oI power relations. Civil society, in short, is equivalent to its negativity.
What is lost in this conception is a distinct concept oI the social,
109
This is the real reason why Foucault gives us such an exceedingly one-dimensional discussion oI
rights and democracy.
110
We need to look again at Foucault's assessment oI each oI these key components oI modern civil society to make our point.

Page 287
We have already seen that Ior Foucault, the juridical subject is merely the support oI disciplinary power. The modern legal person endowed with rights is a dimension
oI modern individuality that, Iar Irom indicating autonomy, is Iunctional to, even the product oI, disciplinary control. Through observation, continual surveillance,
sorting, partitioning, ranking, examining, training, and judging, discipline creates the material counterpart oI the juridical subject by investing the body with power
relations.
But Foucault's genealogy oI the modern individual does not restrict itselI to revealing the underside oI the "legal Iiction" oI the juridical subject; it extends to an attack
on modern selI-reIlexive subjectivity as such.
111
Disciplinary practices objectiIy the subject and create sets oI dichotomies, each side oI which is an eIIect oI power:
mad/sane, sick/healthy, criminal/good citizen, abnormal/normal. What Hegel saw as the two key achievements oI modern civil societythe abstract right oI the legal
person and the principle oI subjective Ireedom oI the moral subject whose intentions and will must be considered in any judgment oI an act
112
become in Foucault's
hands the products oI power relations. The moral subject is the result oI the normalizing judgment that is exercised through surveillance, examination, and with the help
oI the objectiIying sciences oI man: criminology, sociology, medicine, psychology, psychiatry, statistics, demography, etc. Moreover, it is not through the
"internalization" oI values and norms that the "Ialse consciousness" oI the moral subject is created, nor can this subject be emancipated through the development oI a
"true'' consciousness. Power does not stop where knowledge and selI-reIlection begin. Rather, knowledge, truth, subjectivity, and reIlective consciousness are the
coproducer and product oI the objectiIying disciplines. They constitute, together with the normalizing gaze oI the guard, the doctor, and the teacher, a subject
(subjected) object oI power/knowledge.
The same holds true, oI course, Ior the soul or psyche. These are not the products oI an emancipatory process oI selI-understanding but oI a "pastoral power" whose
techniques oI selI-surveillance, selI-interrogation, conIession, and thereby selI-constitution and selI-discipline, initiated by the church, have become secularized and
generalized in modern culture and society. Thus, the political axis

Page 288
oI individualization has been reversed with the shiIt Irom Ieudal to modern society. Ascending individualization reIlecting the power, privilege, and status oI a Iamily or
group is replaced by a descending individualization that increases the visibility and singularity oI those subjugated by and subjected to disciplinary techniques. In other
words, as power becomes more anonymous and more Iunctional, those on whom it is exercised tend to be more strongly individualized and made visible.
113
The
modern individual is the combined eIIect oI disciplinary and pastoral powera selI-monitoring subject who Iunctions as his own soldier-priest.
This theory oI individualization has clear consequences Ior the meaning and role oI the new Iorm oI publicity speciIic to modern civil society. As the disciplines become
deinstitutionalized and circulate Ireely in society,
114
"it is the dust oI events, actions, behavior, opinions'everything that happens'" that becomes visible, public to
omnipresent surveillance by the Iaceless gaze oI power.
115
Indeed, like the process oI individualization, relations oI public and private become inverted with the
development oI modern society. Instead oI the spectacle oI public representation oI sovereign power, it is now the population who become visible to the "public" gaze,
while power recedes into the background. This is, oI course, the point oI the panoptic metaphor. The shiIt in publicity Irom the punishment to the trial does not mean
that the principles oI dignity and moral Ireedom are respected, but rather that justice no longer takes public responsibility Ior the violence bound up with its practice. It
also means that all oI society, vicariously (through publicity) or directly, takes on the role oI judge and engages in normalizing judgments. Even aIter the prison and
punishment have been open to public scrutiny, the public remains complicitous with a technology oI punishment that by deIinition yields visibility and control to the
observer. For "discipline makes possible the operation oI a relational power that sustains itselI by its own mechanism and which Ior the spectacle oI public events,
substitutes the uninterrupted play oI calculated gazes."
116
In place oI the sovereign who displays his power, we have the carceral society displaying its disciplined
subjects to the anonymous viewer. Thus, iI the individual produced by disciplinary pastoral power approximates the soldier-priest, the public beIore

Page 289
whom the exercise oI power is made visible is hardly distinguishable Irom the police.
117
Accordingly, democratization, or the control by the public oI administrative Iunctioning, in no way limits power, as the liberal would have it, or generates a kind oI
power diIIerent Irom administrative control: It simply ensures its proper Iunctioning. Democratic "control" oI the disciplinary mechanisms through publicity entails
accessibility to the great tribunal committees oI the world. For Foucault, this simply means that anyone can come and see with her own eyes how schools, hospitals,
Iactories, and prisons Iunction.
118
Modern publicity provides no alternative, limit, or challenge to disciplinary and pastoral power.
Pluralitv, the third element in modern civil society hailed by its partisans, Iares no better in Foucault's hands. It is simply reduced to the many loci oI power relations
and strategies, and the multiplicity oI atomized individuals who are already products oI knowl-edge-power relations. The discourse oI these individuals, their
"consensus," is as much an instrument oI power relations as is the discourse oI the modern sciences: It normalizes, and normatizes, while maintaining the object oI
power in subjection and as a potential actor only in the purely strategic sense. Thus, neither publicity nor plurality constitutes a check to power.
But what about the Iinal term in our equation, the social? We said earlier that Foucault loses the concept oI the social in his analysis oI modern society. This is not,
strictly speaking, correct. Rather, he presents us with a concept oI the social that is identical with the network oI strategic power relations described above. As already
indicated, society is the terrain oI apparatuses and institutions with multiple Iorms oI subjugation. For Foucault, its "normative" dimension, so crucial to Durkheim's and
Parsons's understanding oI social integration, is, as we know, simply normalization. The social bond, Iar Irom being a moral commitment or a normative consensus
constructed through the medium oI language, tradition, and/or a reIlective, discursive relation to parts oI tradition, is the network oI interwoven and mutually reinIorcing
strategies.
119
Indeed, Foucault is able to view plurality, publicity, and individuality in purely strategic and Iunctional terms because his very concept oI modern society
is that oI a strategic Iield pervaded by administra-

Page 290
tive technologies. These technologies level, individualize, and normalize, but they also rank and sort individuals and populations in a hierarchical manner that permits
communication only through a third elementunequal power relations. This is the new mode oI stratiIication that substitutes Ior horizontal and autonomous social
interaction.
We have already seen that popular solidarities were the target oI disciplinary power. The modern society that succeeds in destroying them is one "in which the
principal elements are no longer the community and public liIe, but, on the one hand, private individuals, and on the other, the state."
120
Such an image oI modern
society precludes any meaning oI sociality other than coordination "Irom above" (through administrative techniques) and/or strategic interaction. It also denies the
existence oI any spaces within modernity Ior the emergence oI new Iorms oI solidarity and association. Indeed, since Foucault maintains that disciplinary/pastoral
power extends beyond the enclosed institution to become complete, consistent, and total, Ior the purpose oI eIIicient and economical production oI wealth,
knowledge, and useIul individuals, nothing else seems to be possible. The disciplinary organization oI societal space multiplies communications and contacts, but only
within the Irame oI strategies and apparatuses that have already reconnoitered and controlled the terrain. Reminiscent oI Marx's notion oI cooperation within a
capitalist Iactory, Foucault's modern society is preschematized by the strategist's gaze: "The classical age saw the birth oI the great political and military strategy by
which nations conIronted each other's economic and demographic Iorces; but it also saw the birth oI meticulous military and political tactics by which the control oI
bodies and individual Iorces was exercised within states."
121
Accordingly, modern civil society is composed only oI individualized strategists engaged in a struggle oI
each against all, pervaded by power and politics understood as war carried on by other means.
122
As we stated at the outset, an analysis oI the negative side oI civil society and oI the speciIically modern Iorms oI domination and stratiIication is an important
component oI any critical theory. One might, in Iact, argue that this is all that Foucault intended to do and that it is unIair to accuse him oI presenting a general model oI

Page 291
society. One could, in short, claim that he has analyzed the logic and project oI contemporary Iorms oI power relationsthe negative side, not the whole, oI civil
society. Perhaps. Yet it remains the case that Foucault's critique is itselI caught in the strategic reason he exposes.
123
For, on the basis oI his theoretical Iramework,
he cannot point to any other category oI action, any other mode oI integration and interaction, that would be a basis Ior analyzing the struggles against disciplinary
power, or the "positive side" oI modernity, iI there is one.
Foucault does insist that "there are no relations oI power without resistances; the latter are all the more real and eIIective because they are Iormed right at the point
where relations oI power are exercised."
124
But, having equated legality and normativity with normalization, subjectivity with subjugation, selI-reIlection, morality, selI-
consciousness, and the soul with the products oI disciplinary pastoral power, discourse and truth with administrative strategies oI control, and the human sciences with
the disciplines that serve or, rather, are part oI power, Foucault is leIt with no conceptual means Ior describing resistances as anything other than counter-strategies oI
power. We are accordingly leIt in the dark regarding the practical thrust oI the genealogical strategy oI analysis, which Foucault nevertheless posits as a Iorm oI
political engagement.
One thing, however, is clear: Foucault is not a partisan oI a simplistic reversal oI values. Genealogical analyses reveal the power strategies involved in constituting new
objects and identities (the homosexual, the hysterical woman, the pervert, delinquency, insanity, sexuality) and the pejorative connotations attached to them. But the
purpose oI such analyses is not to encourage a revaluation in which homosexuality, the perversions, crime, insanity, sexuality are liberated, deemed natural, Ireed to
speak out in their own voice. Such a strategy would do nothing to question the categorization in the Iirst place or to undermine the agencies and mechanisms that
perpetuate the grips oI power on bodies, pleasures, and Iorms oI knowledge. Instead, genealogy is meant to challenge not only the moral valuations oI the normal and
the perverse, Ior example, but also the very normalizing tendency associated with the demand that we understand ourselves through our sexuality, as iI this says who
we are.

Page 292
Perhaps the critical thrust oI genealogy is simply to uncover power strategies involved in the genesis oI power/knowledge regimes in order to disturb the unitary, global
Iorm these take and to reveal their historical and hence contingent character. Such a project would place Foucault close to the critical theory oI the FrankIurt school.
125
Presumably, this strategy would reveal the battle lines and create the possibility Ior a counteroIIensive. Indeed, one could even interpret the Iocus on the societal
genesis and multiple loci oI domination as an attempt to invoke a civil-society-oriented strategy oI resistance against both the local power structures within civil society
and their globalization/colonization by the state.
Such an interpretation, however, does not resolve the diIIiculties created by Foucault's relentless critique oI power, Ior he is still unable to articulate "otherness" or the
Iorms oI action that escape the logic oIinegalitarian strategic power relations. On the one side, he disempowers critique, including his own, by an analysis that equates
discourse, reIlection, and truth with power strategies. On the other side, he cannot speak Ior the victim, as Walter Benjamin did, or oIIer a naturalistic notion oI what is
repressed by disciplinary power, as Herbert Marcuse did,
126
because the victim as well as her psyche are already products oI power and because Foucault has
rejected the "repressive" thesis regarding power relations. Indeed, iI resistance is just the counterstrategy oI that very product oI power, the modern individual, then
why support it? Why is it even interesting? What diIIerence would it make?
127
Apparently, all that successIul resistance can produce is a substitution oI one strategy oI
power Ior another.
There is, in short, no basis within Foucault's work Ior distinguishing resistance Irom other strategic Iorms oI action or strategies oI control. He cannot appeal to the
norms articulated by collective actors, Ior any appeal to norms either reproduces the discourse oI power (and locks the resisters into normalization) or constitutes
simply another strategy oI power. Indeed, Foucault sees the coordination oI action through norms as, in essence, strategic. Nor can he Iollow the path taken by
Habermas, identiIying communicative interaction as the core oI an emancipatory practice that involves a reIlection on and challenge to norms, institutions, and practices
in

Page 293
the name oI alternative (more just, more democratic, more liberal) norms and institutions, because, Ior Foucault, communication is only a means oI transmitting
inIormation and (through the making oI truth claims) controlling and disempowering opponents. The theoretical strategy oI opening norms to reIlection is closed to the
theorist who views reIlection as mere strategy. In other words, on the basis oI Foucault's categorial Iramework, it is utterly unclear what goals or principles those who
resist disciplinary power might invoke that could have a claim on our solidarity. The only clue he gives us is a Iew elliptical statements to the eIIect that "The rallying
point Ior the counterattack against the deployment oI sexuality ought not to be sex-desire, but bodies and pleasures."
128
However, as Foucault himselI showed in the
second and third volumes oI his history oI sexuality, neither bodies nor their pleasures are matters oI sheer Iacticity: Both are constructed symbolically, as objects oI
knowledge and identity, albeit in diIIerent ways in diIIerent types oI societies. Thus, to evoke the body and its pleasures as away to break with the sex-desire regime is
ambiguous, to say the least. Without this reIerent, however, Foucault is leIt with the simple Iact oI resistance to power, but this simple Iact has no normative weight, Ior
it would also be prey to the genealogist's cynical gaze and be revealed as another strategy Ior power.
But there is a prior question to that raised above regarding the reasons Ior partisanship with resistance. How, on the basis oI Foucault's analysis, is resistance on the
collective level even possible? Such resistance would have to be understood either as the deIensive action oI groups whose identities and solidarities have not vet been
penetrated by disciplinary apparatuses, or as the counterstrategies on local levels oI individuals who are alreadv their products and hence are selI-monitoring, purely
strategic actors. In the Iirst case, we would be seeing premodern solidarities in a purely deIensive posture; in the second, modern rebels without any norms, institutions,
principles, or discourses to appeal to, Ior these are already mechanisms oI cooptation. The latter could only appeal to or gesture toward abstract otherness or
diIIerence per se. Indeed, it is unclear how, on the basis oI Foucault's theory, individuals who wish to resist could come together to Iorm the solidary and autonomous
groups, associations, and collective iden-

Page 294
tities that are the sine qua non Ior collective action in the Iirst place. The body might put up some resistance to the drill, the "subject" might shy away Irom surveillance,
and the individual might struggle against the manipulations oI biopower, but even iI Foucault were willing to postulate the other oI reason and discourse in its primordial
vitality (which he explicitly is unwilling to do), this would hardly suIIice Ior an explanation oI the emergence, solidarity, resources, collective identities, and projects oI
collective actors who challenge modern Iorms oI domination. Foucault's analysis has deprived the modern rebel oI any institutional, normative, or personal resources
Ior constituting herselI in terms other than those made available by the Iorces that already control her. The traditions, solidarities, and spaces Ior autonomous action leIt
open by the ineIIicient, discontinuous modality oI power in the ancien regime Iind, in Foucault's work, no modern equivalents. This is not because he was intent on
analyzing something else but, rather, because the genealogical account oI modern power relations turns the very concept oI autonomous voluntary association into an
anachronism in the carceral society. Autonomv is the illusion oI the philosophy oI the subject, voluntarv consent is part oI the deceptive juridical discourse,
association (in our view, the truly modern dimension oI sociality) is simply impossible in a society conceived oI as a strategic Iield constituted by a kind oI
Gleichschaltung oI all organizations by disciplinary administrative apparatuses. We are thus leIt with a critique oI power that insists that resistance exists but cannot
tells how it is possible, what it is Ior, or why it merits our support.
But isn't it obvious that disciplinary power in modern society is aimed against the new solidarities, associations, and movements that emerge on the terrain oI modern
civil society itselI? And isn't it clear that collective actors must articulate distinct projects, new collective identities, and speak in the name oI speciIic values and norms
iI they are to become collective actors and act at all? Moreover, in so doing, they appeal to precisely those new traditions (or discourses), norms, and institutions,
stemming Irom the democratic revolutions oI the eighteenth and nineteenth centuries, that Foucault has so cleverly disempowered: Ireedom,justice, solidarity,
democracy, and, more concretely, parliaments, elections, associa-

Page 295
tions, rights, and so on. Without an analysis oI the two-sidedness oI these institutions,
129
not to mention modern Iorms oI individuation and selI-reIlection, the idea
that modern social movements continuously emerge and challenge disciplinary power would be incomprehensible. Charles Taylor makes a similar point with respect to
the tradition oI civic humanism, the movements inspired by it, and the Iree institutions created in its name. He correctly points out that collective disciplines can Iunction
in two diIIerent ways: as structures oI domination, and as bases Ior equal collective action. Such disciplines can, oI course, undergo a change in Iunction, sliding as it
were Irom Iounding egalitarian politics into serving domination. But Foucault's analysis oI modern power blurs over these processes, revealing only the negative side oI
modernity. He thus serves as a "terrible simpliIicateur."
130
It is our thesis that the condition oI possibility Ior the emergence oI modern social movements, with their autonomous solidarities, newly created identities, and strategic
resources, is precisely the diIIerentiated structure oI modern civil society:
131
legality, publicity, rights (to assemble, associate, and communicate Iree Irom external
regulation), and the principles oI democratic legitimacy. Indeed, we contend that the modern conception oI Iundamental rights is at least as important, in this regard, as
the tradition oI civic humanism cited by Taylor. How else can one account Ior the workers' movement, civil rights movements, the women's movement, the ecology
movement, regionalist struggles Ior autonomy, or any modern social movement or, Ior that matter, the Iorces arrayed against them? Unless one sees at least the
doubleness oI rights and oI legality, one would be Iorced to conclude that collective actors who do appeal to rights, and who reinterpret the key norms oI modern civil
society with their demands Ior more autonomy, more democracy, Ior public recognition as individuals and as group members diIIerent Irom one another yet meriting
equal concern and respect, are somehow all mistaken, somehow articulating irrelevant, anachronistic principles and ridiculous projects.
132
Since Foucault rejects the
only conceivable alternative, the project oI total revolution, he has worked himselI into a vicious circle: Either the norms and projects articulated by social movements
are strategies oI counterpower and as such have no greater

Page 296
normative claim than those oI other power seekers, or they simply reproduce the existing discourses oI power. For a critical theory with a partisan intent, as Foucault's
surely is, this is indeed a serious Ilaw.
It is telling that Foucault cannot consistently maintain this stance, at least with respect to norms and rights. Although he reduces normativity to normalization, he
nonetheless always speaks oI modern power relations as inegalitarian, implying that egalitarian relations would be preIerable. He always describes the latter with the
imagery oI a leveled strategic Iield oI power, but it is obvious that his entire analysis is parasitic on the norm oI equality, however much he may disparage norms.
Similarly, he insists that rights should be viewed not in terms oI legitimacies to be established but in terms oI the methods oI subjugation they instigate.
133
In part, this
is because rights have been reorganized in our time, insoIar as they have been invaded by the procedures oI normalization that colonize law, thus making the legitimacy
question irrelevant.
134
He even notes the tendency, on the part oI those seeking to resist the disciplines and all the eIIects oI power and knowledge that are linked to
them, to resurrect the discourse oI rights and legitimacy. But he sees this as a blind alley, Ior "it is not through recourse to sovereignty against discipline that the eIIects
oI disciplinary power can be limited."
135
The new twentieth-century discourses oI social rights operate on the terrain oI normalized, colonized law, while the older
discourses oI civil and political rights are anachronistic.
Nevertheless, even Foucault is Iorced to return to the language oI rights when he tries to articulate struggles against disciplinary power:
II one wants to look Ior a nondisciplinary Iorm oI power or, rather, to struggle against disciplines and disciplinary power, it is not toward the ancient right oI sovereignty that one
should turn, but toward the possibility oI a new form of right, one that must indeed be antidisciplinarian but at the same time liberated Irom the principle oI sovereignty.
136
(Our
emphasis)
That this is where he leaves the matter is not surprising. Foucault can say nothing positive about this "new Iorm oI right" because he has denuded the very category oI
rights and/or law oI its multidi-

Page 297
mensionality. Certainly law can Iunction as a medium oI domination and control, and some rights do seem to disempower the rights-bearers. But surely this is not the
whole story, or even the main part oI it. As we noted earlier, Foucault misses the normative and empowering dimensions oI law and rights because he, like Marx,
takes the liberal ideology oI rights at Iace value, only to reject it. On this account, the discourse oI rights means the discourse oI sovereignty-contract-legitimacy
transposed Irom the king to the people, and construed this time as the polar opposite oI the political, the state, and power. This Iorm oI the discourse oI rights is, oI
course, ideological and unacceptable. But there is another meaning to the conception and eIIect oI rights claims: In modern civil society, rights are not only moral
oughts, they also empower. Rights do not only individualize, they are also a medium oI communication, association, and solidarity. They do not necessarily depoliticize;
they can also constitute a vital connection between private individuals and the new public and political spheres in society and state. Nor is it the case that questions oI
justice and legitimacy are somehow anachronistic in modern disciplinary society: These remain important to any society, no matter what Iorm power takes.
Foucault is right in arguing that modern civil society is not equivalent to its principles oI Ireedom, equality, democracy, justice, rights, autonomy, and solidarity. But it is
also not equivalent to its strategies oI domination and control. Dr. Mengele is not the truth oI medical knowledge and practice but only their perversion; the use oI
mental institutions to punish political dissenters is not the truth oI psychiatry or psychoanalysis but its abuse. Institution-alized norms (in the Iorm oI law, rights, and
customs) do not only normalize, they also empower and provide a standpoint and a space Ior criticizing and challenging speciIic institutional arrangements and creating
new collective and individual identities. Indeed, the symbolic dimension oI discourse cannot be reduced to its ''real" Iunctions. The institutional articulation oI civil
society provides Ior a modern Iorm oI the social that is more than and other than the disciplinary apparatuses analyzed by Foucault. The two go together; both are
modern, but they are neither identical nor oI the same cloth. Only an analytical Iramework broad enough to encom-

Page 298
pass the dark and light sides oI modernity can account Ior the conditions oI possibility oI the numerous and important social movements or "resistances" that animate
and dynamize modern civil society. And only within such a Iramework can one place the IruitIul yet dangerously one-sided work oI Foucault in its proper perspective.

Page 299
7-
The Systems-Theoretic Critique:
Niklas Luhmann
We inherit the concept oI civil society Irom two sources: the history oI concepts and theories, and the selI-understanding oI social movements. The ideologists oI social
movements seem to conIirm that a rich tradition oI interpretation has not been exhausted, that it remains an adequate basis Ior the symbolic orientation oI
contemporary social actors. This argument could easily be mobilized against the historicist theses oI Riedel, Koselleck, Arendt, and the early Habermas, according to
which the relevance oI the early modern concept oI civil society, Ior better or worse, is to be conIined to its eighteenth- and nineteenth-century origins. Indeed, their
own intense interest helps to negate their claims and has itselI contributed to the revival oI the concept. And yet the case oI the critics cannot be so easily disposed oI,
Ior their claim that the very concept oI civil society is anachronistic is linked to an analysis oI contemporary society as involving a Iusion oI realmsin particular, those
oI state and societythat were diIIerentiated in the earlier liberal epoch. To respond to them one must go beyond the eIIort oI hermeneutic recovery.
Deeply convinced oI the limitations oI even a critical hermeneutics,
1
we believe that it is essential to examine the concept oI civil society also in light oI a social-
scientiIically elaborated theory that at the very least incorporates an objectivating perspective. The link between the history oI concepts and the selI-understanding oI
movements may be based on a questionable double projection: The very same categories that inIorm the selI-understanding oI contemporary social actors may be
projected backward by histori-

Page 300
ans, who are never Iree oI contemporary concerns, and then projected Iorward by movement ideologists to prove the depth and historicity oI their projects.
2
While
social theory also has internalized structures oI interpretation and commitment, on the whole these include precisely that objectiIication oI modern, global societal
contexts that neither historians nor movement theorists are willing and able to accomplish. Thus, identity-Iorming narratives can be conIronted with descriptive and
explanatory materials.
Even more important, since modern social science has adopted a polemical attitude toward the categories oI traditional political philosophy, it is in this context that we
Iind some oI the best arguments against contemporary applications oI the concept oI civil society. Thus, a conIrontation with the results oI social science represents an
important test Ior those seeking to save or revive the classical concept. It is our belieI that this test can be sustained only iI the conIrontation involves a theoretical
reconstruction in light oI contemporary developments addressed by systematic social theory.
Because oI the normatively marked heritage oI the concept, it is diIIicult to Iind systematic social theorists who take up the issue oI civil society. In Max Weber's many
great works, Ior example, there is hardly a mention oI the term or oI any obvious substitute. Talcott Parsons and Niklas Luhmann represent important exceptions to
this trend.
3
We have already presented Parsons's concept oI societal community as an attempt to translate the Hegelian category oI civil society, enriched by
Durkheim's concept oI "the social," into contemporary terms. Luhmann, however, is right to note that this move by Parsons involves a break with the systems-theoretic
assumptions oI his own work, without any general theoretical justiIication. Here is a clue to Luhmann's surprising preoccupation with the problem oI civil society.
4

Undoubtedly, his interest stems Irom a conviction that sociologists such as Durkheim, Parsons (his major Iorerunner),
5
and Habermas (his most important rival) are still
under the sway oI this major concept oI "old European" practical philosophy. Luhmann's strategy against the concept oI civil society and its social-scientiIic
precipitates is to identiIy them with the traditional societas civilis and show the resulting inadequacies Ior the study oI modern conditions.
Paradoxically, Luhmann's own sophisticated theory oI diIIerentiation, developed in an entirely diIIerent context, replaces Carl

Page 301
Schmitt's notion oIIusion between previously diIIerentiated spheres with one oI increasingly complex input-output relations among them. In this respect, he is highly
important Ior us because he potentially resuscitates one aspect oI the concept oI civil society. However, he emphatically rejects the notion that one oI the diIIerentiated
spheres should be understood as any kind oI replacement Ior civil society, or the social, or normative integration. Not even law, the last signiIicant repository oI a
"normative style oI expectation," plays such a role in his theory. Society, in his analysis, stands only Ior the whole, and in some versions even Ior a "world society."
6
OI course, Luhmann "reconstructs" many oI the early modern subcategories oI civil society on the terrain oI systems theory. In each case, however, the reconstruction
involves a decisive break with early modern intentions: Positive law is seen as normless in its deepest Ioundations, association is understood as bureaucratic
organization, and public opinion is reduced to the manipulation oI the themes oI communication. It is characteristic that democracy is identiIied with the general social-
cybernetic Iunction oI "meaning," that is, with the maintenance oI reduced complexity. On the basis oI systems theory, all that remains oI the modern concept oI civil
society is the bare Iact oI diIIerentiation itselI.
7
Thus, Luhmann is also important Ior us, because, on the level oI systematic social science, he works out a highly
comprehensive challenge to the whole tradition oI the concept oI civil society.
Luhmann's preoccupation with the problem oI civil society is indeed surprising, given his own theoretical assumptions and interests. His exercises in a sociological
version oI conceptual history rank with the best in this Iield. According to him, politike koinonia, translated as "political society," was Iirst used as a concept to
describe and elaborate upon the emergence oI an evolutionary stage oI human development, namely, the constitution oI political rule that suppressed or greatly
reduced the importance oI archaic, kinship-based associations and the power oI religion in the immediate relations oI sub- and superordination.
8
The institutions oI
political oIIice and political procedure were the means by which the reordering oI society was accomplished, the major result being "the possibility oI resolving conIlicts
through binding decisions."

Page 302
Political rule, to be sure, meant the emancipation oI human beings qua individuals. But it also meant their seamless integration into a politically deIined societal
Iramework.
Luhmann is somewhat unclear on why the "selI-thematization" oI this development occurred only in the Greek city-state, in particular the democratic polis oI Athens.
9

More revealing than his actual explanation is his somewhat underemphasized addition oI the principle oI citizenship when speaking oI the polis as a version oI
politically constituted rule.
10
He does not notice that political rule needs to be and can be thematized as such only when the wielders oI the instruments oI domination
(here the oikos patriarch-despots) constitute a public. His own stress is on the dimension oI domination rather than on public action. The actual rule in any political
society is that oI a part (in the Greek republics, citizens) over the whole. To Luhmann, several logical paradoxes associated with the concept oI politike koinonia are
to be traced to this state oI aIIairs. Through its linguistic Iorm and its opposition to the oikos, politike koinonia is understandable as only one type oI koinonia among
others. Yet it is also the all-encompassing social system, the polis. Thus, it is a whole that is paradoxically conceived as its own part.
11
Or: it is a whole that has parts
outside itselI, in particular the oikos.
12
The lesson is clear to Luhmann: The society that thematized itselI as political society misunderstood itselI. It was only a social
system in which a newly diIIerentiated political subsystem had Iunctional primacy.
13
For Luhmann, a second, related diIIiculty oI the classical conception oI politike koinonia lies in the attempt to view society as action. This was possible, according to
him, because the political system, supposedly oriented toward right, just, and virtuous action was identiIied with the whole oI society. Equally important was the
understanding oI political society as a body, as a corporate unity capable oI action.
14
In this context, the relatively exceptional existence oI diIIerentiated, specialized
organizations and their slight impact on society permitted a conception oI political society as a whole as itselI an organization, an organized body. OI course, the action
and the goals oI this supposed body were actually the actions and the goals oI its ruling part; only this part constituted an organization.

Page 303
According to Luhmann, the concepts oI politike koinonia and, later, civil society in all its variants thematized the integration oI this organization oI rulers and the
orientation oI its individual actors in terms oI the norma Iive categories oI morality and law (in the latter case, moralized law). Political society was stabilized through the
institutionalization oI "relatively universal . . . rules Ior interpersonal respect and mutual esteem."
15
In other words, the "generalized morality" oI political societies thus
served as the basic legitimation oI political authority. Nevertheless, Luhmann claims that it was only Iunctionally (and not logically) necessary to understand political
society in normative terms.
16
Perhaps what he has in mind is that, although the medium oI power has already replaced ordinary language communication as a means oI
transIerring decisions, its lack oI Iull development or the absence oI other "media oI generalized communication" has made a continued reliance on earlier Iorms oI
direct, linguistic models oI command and obedience unavoidable. The latter, however, cannot operate without normative Iorms oI justiIication. More likely, the point
may be linked to his notion that the medium oI power requires a normatively constructed linguistic code Ior its operation.
17
The binary code oI right and wrong,
allowing in principle the schematization oI all decisions does not represent power as it actually operates; hence, normative language is actually not indispensable to the
description oI politically organized society. Nor would its actors need it to orient themselves within a system oI power. But as long as law is not yet made positive, this
moralistic-legal language is required to represent the operation oI power and the workings oI political society to its social environment, which is not yet linked to the
political subsystem by other, Iunctionally interchangeable, media.
Thus, in Luhmann's terminology, the institutionalization oI the medium oI power allows an important but incomplete replacement oI normative styles oI expectation by
cognitive ones. Nevertheless, while on the level oI social selI-reIlection a secular morality has now taken on the central role in social integration, in reality the
emergence oI power as the Iirst "symbolically generalized medium oI communication"
18
granted immense importance, and indeed Iunctional primacy, Ior the Iirst time
to a subsystem relying on a

Page 304
cognitive rather than a normative attitude to social norms themselves. This subsystem remains linked to a structure oI rules that, even iI no longer tied to immediate
interaction, is capable oI reducing the contingency oI action only by generalizing, universalistic orientations and mutual expectations that remain normative in the sense
oI being "counterIactually" maintained even in the Iace oI empirical "disappointments." This linkage is Iunctionally necessary, at least until it is replaced by Iunctional
equivalents, in order to unburden the power system oI some oI the needs oI integration and thereby protect it (and society?) Irom its potentially vast overextension or
"inIlation."
Ancient practical philosophy, our Iirst source Ior the concept oI civil society, was in this context the theoretical thematization oI both the primacy oI the political and the
moralization oI politics. According to Luhmann, its errors involved a conIusion oI the part (politics) with the whole (society), oI action with system, oI power (as a
medium) with morality (tied to ordinary language interaction), and oI morality as a social reality with the morality oI the moralists.
19
The theory oI bourgeois society is convicted oI analogous iI Iewer errors. Bùrgerliche Gesellschaft represents to Luhmann only superIicially a revamping oI the old
societas civilis, in spite oI the etymological derivation oI the Iirst category Irom the second. Actually, as the suggested alternative term, "economic society," indicates,
bùrgerliche Gesellschaft reIers to a topos that is not identical with but parallel to "political society." The two also turn out to be structurally diIIerent.
20
Again,
Luhmann begins with the selI-thematization oI economic society, which is classically represented by Marxian social theory. Here economic society is understood as a
new type oI society in which production, and even more "a metabolically Iounded system oI needs," replaces politics as the central social process.
21
From a diIIerent
point oI view, also characteristic oI Marxism, bourgeois society means that a politically deIined ruling ''part" (e.g., Bùrger in the sense oI citoven) is now replaced as
the dominant stratum by the owners oI property (Bürgerin the sense oI bourgeois). Luhmann's reservations concerning the Marxian (as well as bourgeois) theory oI
economic society parallel his criticisms oI Aristotelian political philosophy as a theory oI political society. Both make

Page 305
the understandable error oI taking the part Ior the whole, oI identiIying a societal subsystem with the whole oI society. The error is understandable because oI the
dramatic nature oI the emergence oI each oI the subsystems and their Iunctional primacy (Ior a time) in relation to the other spheres oI society.
22
Nevertheless, only
this Iunctional primacy should have been asserted in the case oI the economy, and not the reduction oI all spheres oI liIe to economics. Only the notion oI the Iunctional
primacy oI the economy is compatible with the empirical Iact that the extent and internal complexity oI the political subsystem continued to grow in the whole capitalist
epoch.
23
For Iunctional primacy need only imply that the leading subsystem has the greatest internal complexity and that the new developmental stage oI society is
characterized by tasks and problems that originate primarily in this sphere.
Thus, "political" and "economic" society represent not only parallel processes oI diIIerentiation, along with parallel Iorms oI selI-thematization, but also successive
evolutionary stages. DiIIerent levels oI complexity indicate Ior Luhmann three structural diIIerences between the earlier political society and the later, more complex,
economic society: (1) transIormation oI the meaning oI primacy; (2) replacement oI a (mainly or partially) normative by a cognitive style oI expectation; and (3) loss oI
the capacity Ior action on the part oI the leading subsystem as a whole (not to mention the social system). Let us take each oI these in turn.
First, in discussing the relation oI the economic to the other subsystems, primacy can no longer be even approximately represented in terms oI authority or domination,
but only by the preeminence oI the problem the economy deals with. The diIIerence Ilows Irom Luhmann's distinction between the structures oI power and money as
communications media, with money being the medium around which the diIIerentiated economic subsystem is organized. In the case oI power, a selective decision is
made Ior someone else who is motivated to accept or "to make" this speciIic decision through a particular code, and in view oI negative sanctions. In the case oI
money, a decision is made Ior oneselI, and the other is motivated to carry out his own complementary but generally diIIerent decision in view oI possible rewards, or
positive sanctions.
24
In the Iirst case, decisions are transIerred; in the

Page 306
second, only problems that must be dealt with. For this reason, the level oI social diIIerentiation allowed by the Iunctional primacy oI the economic subsystem is Iar
greater than the level possible in "political society." This capacity is Iirst thematized in terms oI the "Ialse dichotomy" oI state and society, an issue to which we shall
return.
Second, the primacy oI the economic subsystem no longer requires a generalized morality Ior the integration oI society. "It seems that the constancy oI morality over
time, which is supported by all oI society, can be replaced by the constancy over time oI purely economic opportunities."
25
Whereas politics still required (in the
epoch oI its primacy only?) "a kind oI moral 'cover' or legitimation,"
26
the economic subsystem requires it neither "Iunctionally'' nor "logically," neither on the level oI its
representation nor Ior its operation. This is true because the emergence oI the economic subsystem implies "a switch Irom a normative to a cognitive attitude.
Expectations that are normativei.e., counterIactual and incapable oI adapting to changed conditionsare replaced by expectations that can learn and adapt to
change."
27
The moral integration oI economic liIe and the need oI society in general Ior this type oI integration recede with the diIIerentiation oI the economic
subsystem. The society in which this subsystem has become primary can thereIore (contrary to the opinion oI Durkheim and Parsons) gradually dispense with
normativity or conIine it to the single subsystem oI law, whose own Ioundations also become cognitive.
Finally, the disappearance oI a generalized morality as a Iorm oI social integration signals (and is in part caused by) the loss oI society's capacity Ior action. With the
dominance oI the market economy, it is impossible to understand the social whole as a body. "No one can claim to be the plenipotentiary representative oI the
economy."
28
The economic system is not a collectivity. Nor can one so represent the society in which it is primary. Any attempt to discover uniIied agency or
subjectivity representing this society is merely an illegitimate transposition oI a partially genuine possibility oI political society and leads inevitably to conceptual
mythology. Equally important is the stimulus given in economic society to the diIIerentiation oI organizations Irom the rest oI society and Irom

Page 307
one another. The result oI this process is that society itselI can no longer even appear to IulIill the requirements oI an organization, oI an organized body. A plurality oI
the organizations in society are integrated not by means oI a superordinate organization but instead by the workings oI the systemic media oI power and money. Thus
(leaving aside the notion oI interaction or intersubjectivity), the transition to the Iunctional primacy oI the economic subsystem means, Ior Luhmann, the necessary
replacement oI social integration by system integration, oI action as a theoretical paradigm by system. Concepts such as "civil society" and "societal community" are the
obvious theoretical victims oI this shiIt. To this issue, too, we shall return.
Luhmann sees as obsolete not only the concept oI political or civil society but also the one that replaced it. Economic society, or even the primacy oI the economic
subsystem, is now a thing oI the past. This primacy has led to dysIunctional side eIIects Ior its various "environments," which may not have strictly economic solutions.
29
In one version oI his argument, when the primacy oI the economy is at an end, no subsystem is capable oI dominating or even representing the whole. In an earlier
version, the possible subordination oI the economy and oI politics to conscious, scientiIic control or coordination is leIt open. But such subordination could represent a
developmental stage only iI the integrity oI the economic subsystem were preserved, as earlier that oI the political subsystem was preserved, and iI, along with this, the
diIIerentiation oI society were increased. The primacy at this stage would belong to the subsystem oI science
30
and not that oI politics, as in Soviet-type societies. For
such a society, a conception oI societas scientifica would represent an appropriate Iorm oI Ialse consciousness, although the level oI reIlection characteristic oI the
subsystem oI science can also lead to a more appropriate (i.e., systems-theoretic) thematization oI the new Iorm oI Iunctional primacy, this time avoiding the Iallacious
hypostatization oI pars pro toto.
31
Whichever version we choose (and the recent conception oI autopoietic systems clearly indicates the Iirst), the three consequences oI the primacy oI the economic
subsystem will continue to apply to Luhmann's understanding oI modern society. For him, greater diIIerentiation, the decline oI normative integration, and

Page 308
the end oI the capacity oI society (or even a representative part) Ior action preclude any justiIiable conception oI modern society, or even one oI its diIIerentiated
subsystems, as political or civil society. However, Irom the point oI view oI the concept developed in this book, which is not linked to any utopia oI society as a uniIied
agent, subject, or organization, it seems that, with respect to the Iusion argument developed by thinkers Irom Schmitt and Arendt to Habermas and OIIe, the
Luhmannian theory oI diIIerentiation reIurbishes an important aspect oI the concept that is under threat. And yet, while on the most abstract level he does oIIer an
alternative to the Iusion thesis, this cannot beneIit any conception oI civil society, at least in his model. The reasons Ior this seem to be that he considers the state-civil
society dichotomy to be Ialse, and he replaces it by a model that draws the lines oI diIIerentiation quite diIIerently, and that, even in an expanded model oI
diIIerentiation, he sees no need to include a sphere whose Iocus is social integration through both norms and participation in associations.
The argument Ior Iusion oI state and society has always been plagued by a key contradiction: Many (especially the neo-Marxist) proponents oI this thesis invoke it
when they alternately depict the same epoch as that oI the repoliticization oI economy and society and that oI the transition oI the state Irom Iull dependence on or
"positive subordination" to the (capitalist) economy to "relative autonomy" and "negative subordination."
32
Thus, they must assert dediIIerentiation and diIIerentiation
at the same time. This puzzle disappears in the earlier, technocratic version oI Luhmann's argument as well as in the later, liberal version. In the one case, he would
speak oI a movement Irom one Iunctional primacy to another, Irom that oI the economy to that oI scientiIic planning, expanding the diIIerentiation among spheres or,
rather, subsystems. In the other, he would speak oI increasing diIIerentiation, permitting and permitted by increasingly complex subsystems whose network oI mutual
input-output relations could grow correspondingly more dense, giving the appearance oI Iusion. As he notes, the autonomy oI the political system never meant its
isolation. Events in the economy, Ior example, can help constitute problems and motivations in politics, although an autonomous political system will have to produce
relevant decisions according to its own criteria. Thus,

Page 309
intersystemic communication is intensiIied, not reduced. "With the independence oI politics, its dependence on society also increases."
33
Instead oI Iusion, Luhmann
provides us with a persuasive model oI the growth oI both diIIerentiation and interdependence, oI both systemic selI-closure and openness to other systems.
"Reciprocal dependencies and independencies among subsystems increase simultaneously. In principle, this is possible because there is an increase oI circumstances in
which one can be dependent and independent."
34
According to Luhmann, the whole discussion concerning the separation oI state and society has misunderstood this phenomenon oI increasing diIIerentiation and
interdependence. In the spirit oI his thesis, one might say that the Iusion argument amounts only to a partially Ialse selI-thematization oI greater intersocial complexity
characterizing the evolutionary stage succeeding that oI the primacy oI the economic subsystem. UnIortunately Ior the standard dichotomous conception oI the
opposition oI state and civil society, however, this criticism oI the Iusion argument cannot alter Luhmann's view that it, too, represented a Iorm oI "Ialse consciousness,"
this time oI the historically new level oI diIIerentiation characteristic oI economic society.
35
The critique oI one Iorm oI Ialse consciousness cannot reIurbish an earlier
Iorm.
But what is Luhmann's case Ior claiming that the dichotomy oI state and society is Ialse? First, and least importantly perhaps, he thinks that the category oI the state is
too diIIuse: It means everything Irom government to bureaucracy, Irom a part oI the political system to its whole.
36
It is not obvious, however, how this criticism
applies to such relatively rigorous deIinitions oI the state as Max Weber's,
37
which could be and have oIten been used in political science to reIormulate the opposition
oI state and society. Perhaps Luhmann would answer that the all-inclusive concept oI the state as a political organization that monopolizes, through its administrative
staII, the legitimate use oI the instruments oI violence in a given territory violates the internal and organizational diIIerentiation oI the political system, or reduces the
political system to merely one oI its aspects.
38
Second, and next in order oI importance, Luhmann rejects the supposed implication oI the dichotomythat state and society (or

Page 310
civil society) each consists oI sets oI concrete human individuals separated Irom one another in terms oI their whole lives.
39
While this objection applies to many
versions (prevalent especially in movements) oI the polemical juxtaposition oI society and state, even a cursory study oI the more sophisticated conceptions treated
here should dispose oI it. For Hegel, Ior example, members oI "estates" and "civil servants" are to be Iound in both civil society and the state, albeit in diIIerent "roles"
and "Iunctional" relationships. Luhmann might respond, however, that diIIerentiated political roles should be contrasted with the multiplicity oI social roles, a contrast
that is still occluded by the splitting oI human beings into merely two roles, whether public and private, citoven and bourgeois, or citizen and man.
This argument is based on Luhmann's Iinal and most important objection. He points to a characteristic diIIuseness in the concept oI society when juxtaposed to that oI
the state. Assuming that we know what "state" means (and, at best, Ior Luhmann it means "political system"!), the term "society" is a loose one describing its whole
environment.
40
While ancient political society, understanding itselI as the whole, did not recognize its environment at all, the notion oI the state expresses the point oI
view oI the political system when it is capable oI seeing itselI as part oI a diIIerentiated whole, a development that presupposes the political neutralization oI religious,
cultural, and kinship roles and meaning complexes.
41
This level oI selI-thematization in turn presupposes, at least in the main version oI the argument, an
institutionalization oI the Iunctional primacy oI the economic, allowing a new level oI societal diIIerentiation. Nevertheless, even the economic subsystem does not
represent the whole social environment oI the political subsystem. Indeed, the diIIerentiation oI a legal subsystem allowed the diIIerentiation oI the "state" Irom religion
(through constitutional law) and the economy (through private law).
42
Only slightly less important Ior the diIIerentiation oI the political system, one could speak oI an
institutionalization oI the subsystems oI Iamily, science, and culture or art in the same historical context. All these subsystems, which cannot be reduced to a single
"organization'' or "collectivity" or "sphere" or "logic" or, least oI all, "system," constitute the internally dynamic and diIIerentiated social environment

Page 311
oI the political system, which has separate input-output relations with each. Moreover, they have input-output relations with each other. They do not constitute any
coherent entity (Ior Luhmann, a system) in relation to the political system. The notion oI civil society is thus decomposed rather than saved by the model oI
diIIerentiation.
43
But which conception oI civil society is thereby decomposed? Certainly the liberal or Marxist dichotomous models do not stand up to Luhmann's criticism. The
Hegelian theory, on the other hand, while it does not include art or science or Iamily, was highly diIIerentiated internally. To the objection that this model no more
diIIerentiated the economic subsystem Irom law, associations, etc., than did its Marxian heir, Gramsci's response, diIIerentiating economy and civil society, might seem
suIIicient. The potential outcome becomes clearest in Parsons, who diIIerentiates the cultural and economic systems Irom both the political system and the societal
community, the latter understood as the integration subsystem oI society. It is this last sphere oI society, composed oI normative-legal and associational components,
that we consider the most advanced reconstruction oI the concept oI civil society within academic social science. Luhmann shares this interpretation oI Parsons but
does his best to eliminate any such a sphere in whatever guise Irom the systems theory oI society.
Here Luhmann's strategy is twoIold. First, he draws the line deIining the polity in such a way as to include within it all politically relevant associations and publics.
Accordingly, institutions that other theorists rooted in civil society and that served as mediations with the state are now located within the political system proper. In the
process, however, Luhmann severs the connection oI these institutions with rational communication and even with the Parsonian "medium" oI inIluence that is
dependent on these processes. Second, he interprets the Iunction oI law and rights in the diIIerentiation oI society as pertaining only to the (selI-) limitation oI the
political system, not to the institutionalization oI any speciIic sphere in need oI protection Irom administrative penetration. Explicitly rejecting the idea that rights might
also protect against economic tendencies toward dediIIerentiation, he emphasizes the standard liberal notion oI protecting private spheres

Page 312
Irom the state. This model oI law does apparently create a reservation Ior normativity. UnIortunately, within the terms oI Luhmann's theory, the boundaries oI a legal
subsystem that is not stabilized by a medium on the model oI power or money cannot be easily maintained against either a cognitive style oI expectation or, more
concretely, the administrative subsystem oI the political system. In the rest oI this chapter, we shall address in more detail Luhmann's analyses oI the relations between,
Iirst, the political system and civil society and, second, the legal system and civil society.
1. The diIIerentiation oI the political system into administration, parties, and publics seems to be common to Parsons and Luhmann. Actually, Parsons's conception is
quite diIIerent Irom what Luhmann makes oI it. For Parsons, parties and publics as institutions can play a role in the "support system" oI politics because they are
rooted in the societal community. What seems to be an ambiguity among various Parsonian texts concerning the primary location oI these institutions in the polity or the
societal community is, rather, an example oI a quasi-Hegelian theoretical move Iocusing on mediation, in the sense oI providing both diIIerentiation and the
interpenetration needed to stabilize diIIerentiation.
44
II, Irom the point oI view oI the political system, the Iunction oI publics and parties operating in the public sphere
is to generate consent and loyalty Ior binding decisions, Irom the point oI view oI the societal community their role is primarily social integration and, secondarily, to
establish elements oI social control over the state. Located Iirst and Ioremost in the societal community,
45
the public is capable oI generating support Ior the political
system only to the extent oI being able to draw on the resources oI solidarity generated by autonomous rather than bureaucratic associations in civil society.
46
While
Parsons recognizes the possibility oI manipulation and opinion creation by the mass media, he believes that even stronger trends toward autonomous expression and
discussion counteract this possibility.
47
In Parsons's conception, the internal diIIerentiation oI the political system into leadership, administrative, integrative, and
legitimating subsystems
48
(or government, bureaucracy, legislature and parties, and the judiciary) gives the latter two the role oI generating legitimacy and motivational
commitment Ior

Page 313
decisions produced and executed by the Iirst two. But he does not share what he sees as the illusion oI the elite theory oI democracy, namely, that these resources
can be generated entirely Irom above. Nor does he accept the view oI legal positivism that legislative enactment is the only source oI law, or even oI its validity. The
very idea oI seeing the "support system" in terms oI a double interchange between polity and societal community presupposes important trade-oIIs: Political power is
increased and exposed to genuine social control through the same institutions.
49
Thus, it is Iair to conclude that, like Hegel but less consistently, Parsons presents the institutions oI political association and publicity in terms oI a double location that
both diIIerentiates and interconnects state and civil society. To Luhmann, however, the supposed double role oI political institutions on which Parsons bases his
dualistic topological conception reIlects only the diIIerence between the oIIicial, textbook version oI politics and the reality accessible to social science. Moreover, an
internal diIIerentiation oI the political system reIlecting term by term the diIIerentiation oI its environment (Parsons, Hegel) would seriously endanger the autonomy oI
this system.
50
In order to be autonomous, the political system must have time, which in turn presupposes an internal structure that need not immediately react to inputs
Irom its various environments. But this could not be avoided iI the structures oI the environment were reproduced within the political system, or even directly linked as
constituencies to the subsystems oI the polity. "II all subsystems would have their legitimate spokesmen in the political system, politics would be continually conIronted
with an overproduction oI the possible."
51
This is Luhmann's shorthand Ior the Schmittian topos oI a decisionless, ungovernable Iorm oI democracy. In his conception,
however, this is not necessarily the implication oI contemporary political party and parliamentary institutions. On the contrary, when they Iunction properly, they
operate neither in terms oI the traditional bridging Iunction between society and state nor in terms oI the Iusion oI these two domains, but as autonomous Iorms within a
political system uncoupled Irom just those types oI inputs that lead to problems oI governability.
The autonomy oI the political system also depends on its "acceptance" by its various environments. This acceptance, however, is

Page 314
Iavored by the diIIerentiation oI the environment, which Iragments the various possible sources oI demands. ThereIore, it can be a Iunction primarily oI the internal
processes oI the political system, and only secondarily oI exchanges with the various environments. Indeed, the internal diIIerentiation oI the political system into public,
politics, and administration Iavors the crystallization oI certain roles whose Iunction is to link the environments in a desirable way but also to limit this linkage to Iorms
that are separated Irom other roles and are internally Iragmented. Thus, the client, the voter, and the participant oI the public are divorced Irom the Iamily member, the
worker, and the proIessional, on the one side, and do not add up to a comprehensive citizen role, on the other. It is above all this specialization into separate political
roles that produces a Iorm oI acceptance oI political decisions that Luhmann repeatedly describes as quasi-automatic and almost without motivation. This thesis
requires a redeIinition oI the meanings oI publicity, party politics, elections, and parliamentary representation (here a part oI administration), all oI which were once
linked to the category oI civil society but are now placed within the political system. Is this a rediIIerentiation without a diIIerence? Luhmann's redeIinition oI
democracy is our Iirst sign that it is not.
According to Luhmann (here squarely in the tradition oI Schumpeter), any normative deIinition oI democracywhether based on participation, representation, or
pluralistic competitionshould be abandoned. One reason is that each one seeks to make sense oI the idea oI popular selI-government or selI-rule, which is in Iact
incompatible with the logic oI an autonomous political system diIIerentiated Irom the other spheres oI society. Moreover, any scheme to extend participation in the
business oI rule, in terms oI either a direct role in the production oI decisions or a control and monitoring oI those who actually decide, can only raise perpetual
Irustration to a principle because oI the scarcity oI time to participate in relation to the quantity and complexity oI what must be decided.
52
The second reason is even more revealing. Any normative deIinition threatens to prejudice one's own political system (in this case, Western multiparty systems) against
"Iunctional equivalents" (in particular, single-party regimes oI the Soviet type). For Luhmann,

Page 315
even the Schumpeterian residue oI democratic theorynamely, the existence oI competitive parties and contested electionsrepresents a merely secondary
consideration in analyzing the democratic character oI a society. One ought, instead, to turn to more abstract matters and develop a concept oI democracy that can
apply to a variety oI systems, as long as they are suIIiciently complex.
53
Luhmann produces such a deIinition. As processes oI decision imply the reduction oI
complexity, a selection oI a relatively small segment Irom the realm oI possible events and the elimination oI the rest, "democracy means the maintenance oI complexity
in spite oI the ongoing work oI decision, the maintenance again and again oI a sphere oI selectivity as wide as possible Ior Iuture and diIIerent decisions."
54
Luhmann realizes that this deIinition associates democracy with his differentia specifica oI social systems as such, namely, "meaning" itselI, understood as a Iorm oI
reduction oI complexity that maintains the eliminated options within the horizon oI possibilities.
55
He does not notice, however, that this move tends toward the
deIinition oI all societies as democratic; at most, there can be diIIerences oI degree that seem to correspond primarily to the level oI complexity. Indeed, Soviet-type
single-party, ideologically steered societies are repeatedly pronounced democratic, indeed as democratic as multiparty systems as long as ideology is "preserved Irom
dogmatism and is practiced opportunistically," which means the continual possibility oI changing relations oI priority among a high number oI core values.
56
Luhmann
does recognize that single-party rule threatens to restrict consequential social communication to a small politocratic group and to turn other spheres oI society to
secondary Iunctions oI the political system that instrumentalize them. This trend is one oI dediIIerentiation and is contrary to the increase oI complexity.
Characteristically Ior that time (1968), Luhmann proposes that the recovery oI the primacy oI a diIIerentiated economy represents, in context oI a single-party regime,
the major dimension oI the work oI democratization.
57
Indeed, he considers the Ireeing oI social expectations and demands as well as "public opinion" Irom ideology
and the radical expansion oI elements oI dependent pluralism incompatible with the nature oI such a system.
58
While representing some ideal limit on the level oI

Page 316
attainable complexity, these restrictions are not such as to place Soviet-type societies out oI the range that deIines them as democratic. It is in this sense, too, that they
are the Iunctional equivalents oI today's liberal democracies. The reader has a hard time avoiding the suspicion that this is the case only because Luhmann has adopted
the most "disenchanted" and "realistic" view possible oI Western multiparty democracies.
59
Luhmann's realism is in many respects welcome. It is helpIul, Ior example, to see that, Irom the point oI view oI maintaining structurally permissible complexity, the
assimilation oI party programs to one another and the systematic elimination oI many intelligent options Irom political discussion diminish the range oI democratic
options. It is even more important to admit the tension between the open horizons oI possibility Ior action and experience and the realistic recognition by individuals
that they can actually "change nothing."
60
It is, however, both premature and dogmatic to define this paradox as democracy and to declare the goal oI institutionalizing
the ability to change something as by deIinition irrelevant and obsolete. It is, moreover, unconvincing to dismiss all reIorm attempts based on the extension oI politically
consequential communication with a simple reIerence to the scarcity oI time. Once this is done, one gets the strongest impression that, in Luhmann's view, both Soviet-
type societies, at least those with reIormed economies, and Western multiparty regimes in their present Iorms are in principle impervious to attempts at structural
transIormation oI their political systems, in the sense oI democratization.
61
Thus, in the case oI Western societies, Iorms oI social-political interaction that others have
strongly criticizedin particular, a public sphere assimilated to mass culture, depoliticized parties, plebiscitary elections, and parliamentary theatricsturn out to be
elements oI the mature organization oI a genuinely autonomous, diIIerentiated political system.
Luhmann, like Habermas, presents the liberal model oI the public sphere as historically conIined to a single epoch, as indicated by its linkage to the polemical,
enlightenment notion oI "society"yet another version oI the pre-systems-theoretic Iallacy oI pars pro toto. All publicsancient, liberal, and modernrepresent,
according to Luhmann, a neutralization oI role demands

Page 317
Irom diIIerentiated social spheres. The liberal version involved the diIIerentiation oI a sphere oI small circles oI communication integrated through public discussion
Irom already modern, Iunctionally diIIerentiated subsystems oI society: economics, politics, science, religion, Iamily. The internal diIIerentiation oI this new public
sphere was anachronistically segmental; externally, it was a diIIerentiated sphere without a speciIic Iunction. Without a Iunction, the new public (a part) could (mis)
understand itselI as society (the whole), but only Ior a transitional moment because oI its built-in instability. Its role structure not only was not in a position to control
the other spheres oI society, but was completely at the mercy oI Iunctional roles, with their access to money, power, etc.
62
Against Habermas, Luhmann thereIore
denies that a structure oI rational communication, inherited Irom a Iunctionally undiIIerentiated public, could today be revived (as part oI a program oI democratization)
within Iunctionally diIIerentiated organizations themselves necessarily based on the "parcelization oI consciousness." Thus, he asserts not only the structural
transIormation oI the public sphere but the obsolescence oI its normative assumptions as well.
Luhmann does seek to save something oI the liberal notion, but only in the context oI transposing the public sphere into the political system as one oI its subsystems.
Now neutralization becomes the speciIic integration Iunction oI the political system as a whole; its role is to establish a Iorm oI communication not determined by the
nonpolitical roles oI society (Iamilial, commercial, scientiIic, religious) or even by partial political interests (party political or bureaucratic).
63
This may sound like a
repackaging oI the liberal norm in a Iunctionalist wrapping, but there are two major diIIerences. First, the purpose oI neutralization is now the uncoupling oI politics,
and in particular the processes oI decision, Irom society, not the creation oI a new Iorm oI social control over the state. Second, the process oI neutralization lies not
on the level oI the open interaction oI participants but on that oI the Iormation oI the implicit themes oI their various Iorms oI political communication.
Indeed, public opinion is here deIined not in terms oI the "unattainable publicity" oI all political communication but as the structuring oI even nonpublic communication
by institutionalized

Page 318
themes. It is the themes, deIined phenomenologically as "pre-understandings hardened during the course oI communication into more or less Iirm systemic boundaries
in a commonly accepted liIeworld, presupposed in an inarticulate manner" that structure political communication, not the opinions articulated and expressed.
64
Public
opinion thus not only reIers to but also derives its relative unity Irom institutionalized themes, that is, subtexts oI communication, rather than Irom the generalization oI
articulated opinions. These themes contribute to decision making by limiting the arbitrary nature oI what is politically possible. But they also contribute to democracy as
deIined here by keeping alive possibilities according to a diIIerent logic than that oI decision making itselI. They are not parts oI the mechanism oI democracy in any
other deIinition, however; public opinion "takes over the Iunction oI a steering mechanism that, while not determining the exercise oI rule and the generation oI opinion,
lays down the boundaries oI the possible at any given time."
65
ReIerring to such subjects as priorities among various values, the meaning and perception oI crisis, the status oI various individuals who play important communication
roles, the (relative) newness oI events, and the deIinition oI socially relevant pain or pain substitutes (threat, stress, loss), the key themes oI public opinion are ultimately
understood as rules that determine, in the context oI the scarcity oI the resources oI attention, that to which attention can and even must be paid at a given time. These
themes or attention rules are seen as contingent and variable, in line with the steering requirements oI complex systems. Their origin and logic oI development are leIt in
some doubt. On the one side, the institutionalization oI themes is said to depend on the structure oI the political system, which regulates public opinion without rigidly
determining it.
66
This view, consistent with the aim oI presenting the political system as Iully autonomous, seems to imply mainly that the structure oI the political system
determines which institutionalization oI themes is possible, not what is actually institutionalized. Given the stated Iunction oI public opinion, however, this ultimately
means that the structure oI the political system determines what themes are possible, which in turn determines what decisions are possible. In eIIect, then, the structure
oI the political system determines what

Page 319
is politically possible, with public opinion representing only the dependent process by which this is accomplished.
On the other hand, Luhmann also wants to suggest that public opinion has important reciprocal eIIects or Ieedbacks (Rùckwirkungen) on the structure oI the political
system. But this takes the particular Iorm oI developing modes oI organization and processes that would not be aIIected by the variability oI themesin particular,
proceduralism and neutrality toward values. In otherwords, the response to public opinion is to generate and maintain Iorms that allow the political system not to
respond to public opinion.
Such a revealing way oI speaking is also important in the present context, because it implies that shielding the political system Irom publicity is part oI preserving its
autonomy, as iI public opinion had, aIter all, something to do with the nonpolitical environment oI the political. And Luhmann does in Iact call it an overhasty judgment
that public opinion has now been reduced to the inner medium oI the political system without any overall social Iunction, the language only oI the interaction oI
politicians within a political system totally diIIerentiated Irom the social, everyday, diIIuse liIeworld.
67
In this context, he is Iorced to restate and, in eIIect, partially
abandon his hypothesis on neutralization. II it is still true that unpolitical roles are neutralized in the political system by the public sphere, the same is not true Ior political
communication outside the political system.
68
But can there be political communication at all outside the political system, which is itselI deIined in terms oI speciIic communication processes? Luhmann insists that
diIIerentiation does not represent a tearing out oI the social Iabric oI communication and the establishment oI selI-reIerentially closed subsystems. Thus, the
communication oI public opinion cannot be exclusively assigned to the political subsystem; its themes have a relatively context-Iree character that can structure
communication in contexts whose nonpolitical nature is selI-conscious.
69
But now neutralization oI nonpolitical inputs cannot be deIined as the Iunction oI the public
sphere. Instead, and rather surprisingly, Luhmann returns to the classical Iunction oI "mediation" (Jermittlung), deIined in terms oI both diIIerentiation and integration
between political and unpolitical contexts. The presentation oI mediation is, however,

Page 320
astonishingly impoverished: The possibility oI transposing themes Irom a political to a nonpolitical context and the activation oI diIIerent roles oI the same person,
political and nonpolitical, are said to help stabilize the diIIerence between the political and the nonpolitical. The aim remains the diIIerentiation and autonomy oI the
political system; mediation accomplishes this not by neutralization but by Iorcing the processes oI intersystemic communication into narrow and politically manageable
channels.
70
Despite these eIIorts, Luhmann does not manage to present a concept oI the public sphere that completely shields the political Irom the nonpolitical. His second, liIe-
cycle, model oI the origin and logic oI public opinion is a clue to this Iailure. According to the liIe-cycle model, themes that can be articulated in their ''latent phase" by
anyone become political themes only when they get into the hands oI those who make politics with changing themes, namely, the politicians. But whether they do so
(and with what Iorce) depends on the energy oI their generally nonpolitical suppliers and on the success oI these suppliers in making a theme "popular" and
"Iashionable." AIter this happens, powerholders are no longer in a position to censor themes. Now, politicians can compete only in getting themes into the decision
processes oI the administration or in delaying this as long as possible. Either way, the importance oI themes tied to their novelty will diminish, and new ones will take
their place.
71
This whole train oI argument indicates that Luhmann's linkage oI the model oI public opinion to a prepolitical setting does not restore the liberal meaning
behind what is in eIIect a "liberal" topos, but rather ties the nonpolitical dimensions oI publicity to the mechanisms oI commercial, indeed manipulated, communication.
Here, too, he is in the Schumpeterian tradition.
Luhmann seems to deny the necessary role oI manipulation, deIined in contrast to interaction as a Iorm oI unanswerable communication.
72
But when admitting the
possibility oI going around public opinion or using it tactically, his analysis is Iar more detailed and convincing than that oI "mediation."
73
Technically, only methods oI
going around public opinion are manipulative according to his deIinition. Moreover, both these Iorms and those oI instrumentalizing public opinion are presented as
ways oI regulating the internal processes oI the political system. Neverthe-

Page 321
less, the techniques he mentions, such as the production oI pseudo-crises, pseudo-novelties, or pseudo-expressions oI the will oI the electorate, represent direct
utilizations in the political system oI methods oI manipulative, commercial advertisement that, in eIIect, dediIIerentiate the political system by turning one oI its
subsystems into commercialized entertainment.
74
Undoubtedly, Luhmann does not believe that manipulative mechanisms oI either type exhaust the possibilities oI public opinion Iormation. Nevertheless, it is precisely
in this context that he draws the Iollowing conclusion: "Under the described conditions, in the realm oI politics we can count on the multiplication oI the possibilities oI
behavior and at the same time on the restriction oI the possibilities oI active participation." Because oI the specialized technical skills required Ior the tactical use oI
public opinion, what starts out as "management by participation" invariably winds up as "participation by management.''
75
The model oI diIIerentiation and manipulative linkage runs through Luhmann's discussion oI elections and legislatures, moving the analysis to the interior oI the political
system, whose relation to its public subsystem duplicates the latter's relation to the nonpolitical spheres oI society. More exactly, electoral politics and political party
structures are understood to constitute the "political" subsystem proper oI the political system, while legislatures are put within the administrative subsystem. The
Iunction oI the Iormer is to build political support, to provide a mechanism Ior recruiting oIIicials, and to manage and absorb conIlict and protest. Only the latter is to
have any role in decision making, which is understood as a particular combination, uncoupling, and reconnecting oI actual processes oI decision making with that oI the
"presentation" oI its production. By putting the legislature into the realm oI administration "broadly understood," Luhmann makes a shiIt within the political system that
parallels his shiIting the public into the political system. In each case, he moves a structure classically understood as an element oI the public mediation between society
and state closer to the interior oI the political system itselI, understood as administrative decision making.
It is striking that in these shiIts Luhmann cannot Iully eliminate the element oI publicness that seems to be attached to elections and parliaments. The speciIically political
role oI the voter is linked

Page 322
to participation in the public
76
up to the point oI actually voting; the working out oI themes capable oI consensus is said to be among the tasks oI party politics;
77
the
maintenance oI the image oI politicians is among the tasks oI parliament; and Iinally, the public presentation oI grounds and arguments in parliamentary sessions is said
to seriously reduce the choice oI representable positions.
78
In all this, more seems to be involved than merely the useIulness oI being able to represent the process oI
decision making by two stories: an "oIIicial," civics textbook story that is important Ior building support and shielding the actual, nonpublic process oI decision making,
and a "realistic" story (Luhmann's own) that is important Ior the selI-reIlection (or at least the proper scientiIic understanding) oI the political system. Characteristically,
the democratic Iunction oI maintaining reduced complexity in the realm oI the possible is assigned not only to the public but to politics
79
and parliament
80
as well,
linked in particular to the institution oI the opposition, whose alternatives survive even electoral or parliamentary deIeat.
With this said, the essential Iunction oI both politics and legislature remains, Irom the point oI view oI the social system as a whole, diIIerentiating the political system
and ensuring its autonomy by uncoupling political decision making Irom social inputs. This problem is solved not through total separation but by processes oI Iiltering
and selection that manage society and build political support (a "permanent problem" with the passing oI premodern Iorms oI legitimation) at the same time. Electoral
procedures convert the problem oI support Irom relying on the nonpolitical roles oI the (premodern) ruler to drawing on the strictly diIIerentiated political roles oI
voters.
81
In their roles as voters, individuals are guaranteed access to the political system independently oI other social roles or statuses (universal suIIrage, equality oI
votes), and the inIluence oI social ties and pressures is minimized (secret ballot).
82
Indeed the particular, atomized choice oI the voter, having almost no consequences
Ior other aspects oI the individual's liIe, including other politically relevant roles, involves no social responsibility and cannot be the source oI any social conIlicts.
83
This
point has several consequences, all strengthening the autonomy oI the political system. Not being open to "social" inIluence, the voter is all the more exposed to
immanent political

Page 323
inIluence, presumably by the mechanisms oI public opinion. In seeking to inIluence political processes, the voter has a choice oI a small degree oI inIluence at minimum
cost (voting) or more inIluence at great cost (voluntary associations, petitions, letters to newspapers, etc.). Given the separation oI either Iorm oI inIluence Irom
decision making, Luhmann has no doubt that the Iirst option will be chosen, although the continued presence oI the second contributes to democracy, at least in the
sense oI "everything is possible, but I can do nothing." But even the restricted and minimized inIluence oI the voter role distinguishes the individual Irom a subject
(Untertan) oI rule who receives but never sends political communications, thereby contributing to legitimation through procedure.
84
The situation is analogous Ior conIlict-oriented collective actors with speciIic interests. Luhmann adheres to the view that elections are not suited to the expression oI
particular interests. Because those elected receive generalized support and are not bound to any constellation oI interests, electoral processes cannot easily produce
decisions Ior concrete conIlicts. Nevertheless, they do allow the political system not to suppress conIlicts but to channel them, including even radical protests, into the
interior oI the political party subsystem in a manageable Iorm. Here the advantage oI competitive elections over single-party uncontested elections shows itselI.
UnIortunately, multiparty systems with conIlicting lists do not solve the problem automatically because oI the tendency to undiIIerentiated programs. The continuing
dilemma oI political party subsystems is to avoid both reproducing too much social conIlict (which would threaten the diIIerentiation and stability oI the political system)
and absorbing too much conIlict (which could mean the reappearance oI unmanageable conIlict outside the political system).
85
Characteristically, Luhmann tells us next to nothing about what happens in the context oI being caught on either oI the horns oI the dilemma oI too much or too little
conIlict in the political system. It would appear that the legislature plays a role in the resolution oI too much political conIlict. Here, Luhmann runs into Carl Schmitt's
thesis oI the Iragmentation oI sovereignty and the reduction oI parliament to mere show. To Luhmann, the thesis is based on the

Page 324
Ialse assumption that open sessions oI parliament ever were or should be at the center oI actual decision making. Parliament, especially its plenary session, is and
should be "mere show," in the sense oI symbolically presenting the production oI decisions in accordance with our oIIicial script oI politics. Such a show (with its
important Iunctions Ior democracy, in Luhmann's sense) can have relatively large room Ior plural interests, open conIlict, and the selI-presentation oI political
personalities.
86
It is, however, the inIormal mechanisms, shielded and veiled by the Iormal procedure, that are the stuII oI the realistic script oI decision making. While
the parliamentary process as awhole, as even the classical theory oI Iree representation realized, should not mirror social conIlicts, the appearance oI the party system,
at least in the version analyzed by Schmitt, threatens to do just this. Luhmann implicitly accepts here the decline oI the classical principle oI representation, and he
admits some dangers Ior the autonomy oI the process oI decision making. In his own terms, there is a danger oI a bottleneck between the political and the
administrative subsystems oI politics.
87
Models oI endless discussion or conIlict indicate only the problem rather than the solution in this context. Instead, the
separation oI decision making Irom Iormal parliamentary procedure implied by reliance on inIormal and even deviant mechanisms
88
clears the potential bottleneck and
reduces the inIluence oI politics to its proper measure. The real decision making occurs elsewhere than in parliamentary procedure, although the conversion oI political
power into a zero-sum game by the Iormal mechanisms oI majority rule considerably simpliIies the interactions and bargaining processes oI those who actually decide.
Thus, the old thesis concerning the crisis oI parliamentarism is resolved by Luhmann in a way that points to something like the neocorporatist duality between public
and secret, Iormal and inIormal, parliamentary and Iunctionalist processes oI interest aggregation. He is, however, perceptive enough to realize that today there is a
new threat to parliamentarism. A crisis oI parliamentary legitimacy can result not only Irom too many societal inputs and too much party conIlict, but also Irom too
much social apathy and too much absorption oI conIlict. The method oI shielding the mechanisms oI decision making can be overly suc-

Page 325
cessIul; the number oI logically possible social alternatives loses its link to the actually possible iI the Ieeling that "I can do nothing" becomes generally and publicly
thematized.
In this context, Luhmann passes up the chance to build on the one element oI genuine democratic legitimacy that appears in his presentation. In his conception, it is
above all the dramaturgical elements oI elections and parliaments that have the Iunction oI "inIorming" the uninIormed, oI energizing the apathetic, oI symbolizing
democracy as an open horizon oI possibility somehow present and charged with meaning, even iI detached Irom the possibilities oI action. But this argument, as he
elsewhere notes, threatens to dediIIerentiate politics, this time in relation to art or mass culture and entertainment. The citizen is said to participate in politics to the
extent oI being able to identiIy with some oI the actors oI the drama, becoming part oI the public in the sense oI audience (Publikum).
89
It is hard, though, to keep
the show good or even entertaining when people begin to notice that there is nothing at stake. This train oI argument soon leads us back toward Luhmann's concept oI
public opinion, involving the compulsion to produce novelty in the Iace oI the predictable obsolescence oI Iashionable themes and even the manipulative use oI this
opinion to produce pseudo-events, pseudo-crises, and pseudo-personalities.
At one point, however, Luhmann points to another type oI phenomenon, and implicitly to a model oI the public elsewhere denounced as obsolete. During plenary
sessions oI parliament, "one's grounds, unlike motivations and backers, have to be publicly presented and exposed to the criticism oI opponents. This restricts the
choice oI representable positions."
90
Luhmann does not, and we believe cannot, tell us where this compulsion to deIend positions "with the help oI arguments and
reasons Ior decisions (Argumenten und Entscheidungsgrùnden)" originates. Some candidates Ior a possible answer, such as a political culture with built-in
standards oI rationality, or a liIeworld that has undergone normative as well as cognitive learning, or a public sphere organized according to the possibility oI rational
discourse rather than merely dramaturgically or as an organ oI mass culture, are in principle excluded Irom his theory.

Page 326
Our point is not to deny the empirical importance oI Luhmann's description oI the political system, based on the primacy oI a core administrative system capable oI
shielding its autonomy and internal selection processes by the outer rings oI politics and publics. Rather, it is to register an uneasy relationship between the two
scenarios Luhmann associates with the political system, the "realistic" one and the "oIIicial" one. The latter, in order to play its role, cannot conIine itselI to a
dramaturgical status. But to eliminate its discursive or rational components, which represent, in Luhmann's oIt-stated opinion, undesirable (societal) restrictions on the
Ireedom, variability, and pragmatic case-by-case nature oI decisions, would put the procedural legitimacy oI the political order into jeopardy.
91
2. Luhmann is Iully aware that an overextension oI the logic oI the political system would be harmIul Ior this system itselI. His theory oI the autonomy oI the political
system Irom societal inputs is not automatically a theory oI the Ireedom oI the various social spheres Irom political penetration. A diIIerentiated political system is
indeed Iar more powerIul than its predecessors and has both greater possibilities and greater interest in intervention. Luhmann undoubtedly accepts Schumpeter's
insight that iI the realistic model oI democracy is to work at all, care must be taken that political mechanisms not be extended to too much oI society.
92
He also agrees
that such limitation must be primarily a selI-limitation oI the political system. In contrast to Schumpeter's version oI legal positivism, however, he claims that the
mechanism can be that oI legal enactment, which is necessarily the product oI political decision. In Iact, he develops a Iunctional theory oI Iundamental rights as Iorms
oI protection against the overextension oI the political. Such a move, iI justiIiable, could help to diminish qualms such as Schumpeter's that positive enactment is not in
itselI suIIicient to limit political power. Unlike Parsons, however, Luhmann does not locate a societal center oI normative integration and associational liIe as the core oI
what is to be protected by the selI-limitation oI the political system.
It is instructive to compare the conceptions oI Iundamental rights in Parsons and Luhmann. Derived Irom equality, one oI the

Page 327
core values oI the "democratic revolution," rights in Parsons's theory seem to have more to do with the internal structure oI the "societal community" than with its
diIIerentiation Irom polity, economy, or culture. Following a Iamous and inIluential text oI T. H. Marshall,
93
Parsons decomposes citizenship into civil and political
rights and their social prerequisites.
94
Equal participation in these three components deIines Iull admission to or membership inthat is, citizenship inthe modern,
democratic societal community.
95
OI course, Parsons understands the democratic revolution and especially its other core values, liberty and Iraternity, in terms oI a
large-scale process oI diIIerentiation between societal community and polity. Moreover, the prehistory oI the democratic revolution, especially English legal
developments, already involved a transIormation oI law Irom an ''instrument oI government" to a "mediating interIace" between state and society. In particular, the
establishment oI "the rights oI Englishmen" (such as habeas corpus, Iair trial, and protection against arbitrary searches) is said to play an important role in this
development.
96
Thus, while Parsons never brought together the strands oI his argument about rights, it is Iair to say that, aside Irom the Iundamental problem oI
inclusion to which he links his whole citizenship complex, his conception stresses both diIIerentiation and integration, with civil rights playing a more obvious role in
diIIerentiation and political rights providing Ior new Iorms oI integration ("mediation") between the spheres oI state and society (polity and societal community).
It is striking that Luhmann makes a determined attempt to reduce the Iunction oI Iundamental rights to the single dimension oI diIIerentiation.
97
His stark, "realistic"
conception oI the modern political system and oI political power as potentially "totalitarian," aiming at the politicization oI all spheres oI liIe, underlies this thesis.
98
And
yet the modern political system is born oI social diIIerentiation. Its modernity presupposes diIIerentiation, and its perIormance Ior other societal subsystems requires
economy oI power resources.
99
The establishment and selI-establishment oI limits to state power is thus a positive-sum game. Whatever the actual historical origins oI
basic rights,
100
neither the state nor a purely social sphere produces them alone; they represent gains in the autonomy oI the nonpolitical and the power oI the
political.
101


Page 328
The logical paradox oI legal positivism with respect to rightsthe supposed impossibility oI the selI-limitation oI political power through political enactmentis thus
sociologically resolvable. Fundamental or constitutional rights are not rooted in an extrapolitical or extralegal order but are presuppositions and products oI the
diIIerentiation oI society. While they are not the only institutions that stabilize this diIIerentiation, today at least they are indispensable Ior this purpose.
102
Accordingly, the structure oI rights cannot be deduced Irom a single principle such as "individual Ireedom" or "society against the state." Nor can they be arranged
according to a hierarchy.
103
The reason is that Iundamental rights consist oI several complexes, each oI which regulates the relationship oI the political system to one
or another subsystem according to diIIerent and unique structural requirements. To begin with, liberties or Ireedoms (Freiheitsrechte) have to do not with the
autonomy oI the individual in the strong sense, but with the protection oI the individual personality (itselI a subsystem presupposed by the other subsystems), which in
turn is highly dependent on the maintenance oI conditions Ior adequate selI-presentation. These depend on the Ireedom oI the actor Irom visible and open constraint, in
particular Irom binding decisions, and on a basic consistency oI selI-presentation, here deIined as the essence oI dignity. Within what are ordinarily considered
Ireedoms, Luhmann distinguishes between rights oI Ireedom and oI dignity, respectively related to the external and internal preconditions oI the presentation oI the
selI.
104
As goods that exist prior to the state, they are not products oI rights and are only protected by them with respect to the political system. Rights oI Ireedom,
strictly speaking, protect the space oI individual action and expression; Ireedom oI speech in all its Iorms seems central in this context. Luhmann considers the "rights oI
dignity" to be more diIIicult to deIine and to establish, and he notes a certain tendency in many (especially liberal) legal systems to subsume them under
Freiheitsrechte. Nevertheless, he considers them in principle quite distinct, to be connected with the protection oI an intimate sphere that should be separated Irom
that oI public action.
105
So-called Ireedom oI conscience is the best contemporary illustration oI this requirement.
106
Without it, the indi-

Page 329
vidual loses the responsibility to work out Ior herselI a consistent and convincing selI-presentation.
As in the case oI Ireedom, Luhmann considers the protection oI dignity by Iundamental rights to be relevant only when the threat comes Irom the state.
107
Yet he
believes that the Ialse dichotomization oI state and society leads only to the mistaken liberal attempt to derive all Iundamental rights Irom Ireedoms.
108
He nevertheless
Ieels compelled to note the importance oI Freiheitsrechte in stabilizing the other complexes oI rights, relevant to other spheres oI society, all oI which presuppose the
possibility oI the Iree selI-presentation oI individual personality. This seems to be the case especially Ior the so-called Ireedoms oI communication. Let us note, in
passing, that Luhmann also considers the rights oI personality to be linked to a type oI communication, namely, selI-expression in a Iorm recognizable to others as Iree
and digniIied.
In the case oI the rights oI assembly, association, press, and opinion, however, the context changes Irom personality to culture, Irom subjectivity to intersubjectivity
and its presuppositions. As beIore, Luhmann considers Iundamental rights to be relevant protections oI communication only as long these are potentially threatened by
the state.
109
He is not particularly successIul in connecting in a clear-cut way a set oI communication Iunctions (culture and its internalization, the speciIication oI the
need Ior consensus, the mobility oI contacts, and the determination oI the themes oI public opinion) with a series oI rights (oI religion and belieI, oI association and
assembly, oI the press, oI art, oI scientiIic research and teaching, and many others in an eclectic list). The point is nevertheless clear enough: In diIIerent ways, the
modern state needs, yet potentially threatens, a many-leveled Iramework oI societal communication that can be stabilized, in part, through Iundamental rights.
The threat is statization and not politicization as such. For Luhmann, the state/society dichotomy is a misleading basis on which to construe the rights oI communication,
because it supposedly implies the political neutralization oI nonstate spheres. Political problems and political power arise not only in the political system but also in
Irameworks oI protected social communication. This societal power should be absorbed and processed by the

Page 330
political system rather than eliminated through statization. The unburdening oI the state is essential, even at the cost oI political threats arising Irom other social spheres.
110
As in the case oI rights relevant to personality, here, too, Luhmann claims a preeminent status Ior rights oI communication. All social systems and identities, including
personality, presuppose processes oI social communication and require their protection vis-a-vis a dynamic modern political system. Economic rights do not seem to
have the same Iundamental importance in this presentation. While they themselves presuppose Iree personality and communication, the reverse is not argued (in
contradistinction to liberal and neoliberal claims). To be sure, Luhmann does also oppose the derivation oI the right oI property and the "Ireedom" oI proIession Irom
Freiheitsrechte.
111
It is not persons but roles and Iunctions that must be protected in the case oI the economy. Once again, despite the possibility oI other social
spheres (Iamily, religion, science, etc.) inhibiting economic processes, Luhmann maintains that Iundamental rights are relevant only when the state is the source oI the
threat. While the modern state and a diIIerentiated economic order have long been presuppositions Ior each other,
112
the state as the source oI binding decisions
nevertheless has a tendency to intervene directly in economic processes. Rights oI property and the Ireedoms oI contract and proIession protect the diIIerentiation oI
economic processes and roles. They block some interventions not in the name oI justice and injustice but in order to protect the economy Irom uncertainty and
disorganization.
113
For this reason, these rights can be and are generally made compatible with Iorms oI intervention that increase interdependence without
dediIIerentiation and with interventions that increase economic eIIiciency.
114
Luhmann stands apart Irom the classical liberal and neoliberal notion oI rights, based on a polemical rejection oI state intervention in society, yet he stays within this
tradition to the extent that he repeatedly claims that Iundamental rights by their nature, and not only historically, represent Iorms oI protection in the Iace oI the state or,
in other words, Iorms oI selI-limitation oI the state. One reason Ior this preIerence lies in his deIinition oI rights as Iorms oI selI-limitation by means oI legal enactment.
For the legal positivist,

Page 331
the only source oI such enactment is the state. In the present context, however, this position leads to the paradoxical consequence that, although the primacy oI the
economy has replaced that oI the political system,
115
and in principle the highly precarious economization oI other spheres oI society, including politics, is seen as a
genuine danger,
116
selI-limitation in the Iorm oI economic "constitutionalism" cannot and should not be introduced.
117
In Luhmann's Iramework, there are no rights
against the economy. This prejudice leads him to rely even more on interventions Irom the political system to manage the risks oI a highly dynamic economic
subsystem, a position not really compatible with his intention to limit political intervention to acts designed to improve internal Iunctioning. Indeed, as we now know,
political interventions oI the type he readily aIIirmed as late as the early 1970s could become dysIunctional Irom the long-term economic point oI view, in the process
producing additional negative side eIIects.
This partially selI-contradictory outcome is all the more paradoxical because Luhmann cannot consistently restrict the notion oI Iundamental rights to selI-limitations oI
the state in contexts where the political system represents the major source oI risks Ior other subsystems. A case in point is political rights, which Ior Parsons
represented mediating and integrating principles primarily. Eschewing this interpretation, Luhmann saves his general conception based on diIIerentiation by reversing his
perspective. Political rights such as suIIrage, the secret ballot, as well as the rights oI political associations (parties) and oI elected oIIicials represent Ior Luhmann,
however paradoxically, Iorms oI protection oI the political subsystem against external (including economic!) pressures. They are ultimately mechanisms oI selectivity
uncoupling and insulating the highest instance oI producing binding decisions, namely, the administration.
118
We have already seen this train oI thought in Luhmann's
political sociology. His stress is on preserving elections as the narrowest possible channel through which societal conIlict, communication, and inIluence can enter the
political system Irom outside and enter the administrative subsystem Irom the public and political subsystems oI the political system. While, in comparison to Soviet-
type systems,
119
he seems to note the role oI political rights in protecting society Irom over-

Page 332
politicization and the political subsystem Irom overbureaucratization, his emphasis in relation to the Western liberal democracies is entirely on shielding the political and
the administrative. Indeed, in this context the protection oI electoral and public-political procedure has its point only in the legitimation oI the decisions oI an
administration, which are arrived at through wholly internal and uncontrolled procedures.
120
Luhmann's aIIirmation oI political rights as the selI-protection oI the political sphere rather than its selI-limitation with respect to other social spheres is not only
inconsistent with his conception as a whole but also a statist aspect oI his doctrine oI rights. This stress is only partially explained by the legal-positivist search to Iind
adequate political motivation Ior the selI-limitation oI the political system through legal (including constitutional) enactments. The idea that political rights are the selI-
protection oI the political once again helps Luhmann demonstrate the inadequacy oI a model oI rights derived Irom the idea oI deIending society against the state. For
each complex oI rights, he has used both the idea oI the diIIerentiation oI the spheres to be protected and the idea oI "interdependent independence" to criticize the
rigidly dualistic model oI society and state. DiIIerentiation in his model oI rights works initially through political-legal enactment, itselI a Iorm oI interdependence. Nor
does diIIerentiation, as we have seen in the case oI economic rights, exclude the possibility oI new interrelationships. Here, however, these considerations do not lead
Luhmann to claim the complete obsolescence oI the state/society dichotomy. Instead, he argues Ior its preservation through generalization in a conception oI systems
communicating with one another.
121
This new model is not designed to save the conception oI civil society. On the contrary. Luhmann seeks in particular to decompose the idea oI a sphere in which
mutually reinIorcing and stabilizing normative structures, Iorms oI association, and public communication conIront the modern state and the modern economy. To be
sure, his hint that the rights oI personality and oI communication represent each other's presuppositions on the deepest level cuts through his Iramework oI rigid
diIIerentiation. Personality and communication are presented in some contexts (even iI vaguely) as each other's Ioundation, not as logically

Page 333
separate albeit interdependent systems. But Luhmann does not develop this insight, although it could have served as the Ioundation Ior a deeper theory oI rights. For
him, Iundamental rights diIIerentiate and protect diIIerentiated systems; they do not have their ground and justiIication in a single uniIied Iramework that, together, they
help to establish as well as diIIerentiate.
An exception, perhaps, is the legal system itselI. Whatever else rights help to diIIerentiate, their ability to Iunction at all seems to depend on the diIIerentiation oI a
system oI procedures in which they can be autonomously interpreted and applied, and perhaps even enacted.
122
II rights are to deIend diIIerentiation Irom the
political system, it would seem that they themselves must be diIIerentiated Irom this system. And Luhmann does in Iact attempt to treat (increasingly, as his legal
sociology is developed) the legal system as a diIIerentiated subsystem oI society. Rights, which Ior him are legal institutions like any other, albeit with speciIic Iunctions,
belong to this subsystem. Since Luhmann considers law to be Iundamentally linked to a normative style oI expectation, we might assume that the legal subsystem itselI
represents a diIIerentiated residue oI the conception oI civil society constructed in part around shared Iundamental normative structures. In our view, however, and
probably in his own, Luhmann's intended break with the concept oI civil society is too radical to allow such an interpretation. The question is whether he can work out
an adequate and consistent theory oI the legal system, as diIIerentiated Irom the political, in the context oI his radical campaign against civil society.
The reevaluation oI the problem oI norms in Luhmann's legal sociology, and the restoration oI a central place Ior norms in his sociological analysis, is striking, given his
previous polemic against the theory oI normative integration in Durkheim and Parsons. This polemic is now only partially mitigated. He argues that norms are important
in the social structure, but to construe them as identical to that structure is to misunderstand their place.
123
Nor should norm and institution be considered synonymous:
Not all institutions embody norms, and not all norms are institutionalized. Finally, it is wrong to assume that the normative integration oI society is based on common
and shared norms. In all diIIerentiated societies, norms are contested and represent important stakes oI

Page 334
conIlict.
124
In this theory, legal norms, representing only a small portion oI normative phenomena,
125
play a crucial role in managing and stabilizing normative conIlict
rather than expressing, symbolizing, and reaIIirming normative order.
According to Luhmann, norms are "counterIactually stabilized behavioral expectations."
126
Laws are institutionali:ed norms, stabilized in terms oI procedures,
whose structure oI expectations is guarded Irom and restored aIter disappointments by sanctions.
127
These deIinitions rest on detailed theoretical considerations that
can only be outlined here. In the context oI complexity and contingency, social action can be coordinated only through structures oI complementary expectations and
mechanisms capable oI dealing with disappointment.
128
For Luhmann, "internal" expectations by individuals about the actions oI others generally replace coordination
through actual communication, which is understood as a time-intensive and thereIore scarce resource, one best reserved Ior a Iew open, unsettled contexts, generally
conIlicts.
129
Expectation, however, as a response to the contingency oI the other's actions, is put at risk by the Iact that the other is the same as the selI and has its
own expectations. This leads potentially to double contingency: Each can be disappointed by the other. The coordination oI social action is thereIore possible only iI
the expectations of expectations are stabilized.
130
In Luhmann's largely silent society, there are two and only two basic styles oI expectation: the cognitive style, which is capable oI learning and altering expectations in
the Iace oI disappointments, and the normative style, which involves inability or, rather, unwillingness to learn. What is typically considered to be a necessary yet very
precarious Iorm Ior projecting selI-identity in the case oI the individual psyche (not learning as involving immune reactions bordering on the pathological) becomes, in
the case oI normative expectations, a socially stabilized and guaranteed structure.
131
For both psychological projection and social norm, the main goal is to stabilize an
identity-related structure oI expectation rather than to secure empirical compliance. But while the origin and operation oI psychical projection can be entirely internal to
the individual, Luhmann is able to point to genuinely external, social mechanisms Ior stabilizing and reproducing norms.

Page 335
Luhmann's treatment oI the problem oI origins is inadequate. The only social process oI norm creation to which he can pointactual communication and coming to an
understanding to create or alter rules and deIine deviancehe considers exceptional, characteristic only oI small-scale social systems. Indeed, the validity oI norms
supposedly depends on the impossibility oI actual communication concerning them, or at least concerning all oI them within the same time horizon.
132
The diIIerentiation oI society involves an increasing diIIerentiation oI the normative and cognitive styles oI expectation. In their pure Iorm, each is open to new risks: the
risk oI a hardening oI social identities in one case, those oI a completely contingent, and thereIore unbearable, Iuture in the other. The main response in modern society
is not dediIIerentiation but combinations involving "contrary ordering," as allowed by the reIlexive structure oI the expectation oI expectations. In particular, one can
cognitively expect a normative expectation and normatively expect a cognitive expectation.
133
The Iormer combination, the cognitive expectation oI the normative, has
key importance Ior Luhmann's legal theory.
Norms become laws only iIinstitutionalized in terms oI sanctions and procedures. Institution building plays a crucial role in managing normative conIlict. Luhmann
deIines institutionalization by the possibility oI basing expectations on "the presupposed expectations oI expectations on the part oI a third party."
134
DiIIerent Irom
external observers, third parties are potentially coexperiencing and coexpecting, albeit unknown and anonymous, members oI the same Iabric oI interaction. The role
oI the judge crystallizes historically around the Iigure oI the third party. For Luhmann, institutions, like norms, do not depend on actual communication or consensus.
Actual consensus being scarce, institutionalization uses it economically. Instead oI creating or presupposing consensus, institutions involve a better use oI the small
amount available, distributing it to relevant areas. For their own Iunctioning, institutions only need an anticipation oI consensus, with relevant third parties, in the
expectation oI expectations, a presumption that is rarely tested.
135
While empirically there is little to object to in this conception, we again note Luhmann's repeated
inability to link the mechanisms oI real communication and consensus building, which

Page 336
he cannot totally neglect, to his other mechanisms oI stabilization, or even to assign any other reason Ior their existence than the implicit one that some actual consensus
is needed to make plausible the anticipation, or "successIul overestimation," oI consensus.
In the case oI legal norms as institutions, the actual mechanisms needed to stabilize expectations are sanctions and procedures. The importance oI sanctions lies not in
their secondary task oI motivating compliance but in the possibility oI relieI Irom disappointment through a symbolic restitution oI the norm. In developed societies,
according to Luhmann, sanctions are the only way oI demonstrating "the presumed consensus oI third parties." Occasional coercion thus symbolizes anticipated
consensus and can thereIore replace Iactual consensus in Luhmann's model oI law. In this model, though, the continuous Iunctioning oI law is not based primarily on
coercion, an instrument that would become blunted by its very use. There is a need to represent continuity through a mechanism that is entirely present and yet can be
presumed to exist beyond the current community oI participants. DiIIerentiated procedures play this role and thus have priority in the institutionalization oI law.
136

Procedures are better than sanctions Ior symbolizing continuity because they can reIocus concern Irom (increasingly less likely) agreements about outcomes to mutual,
iI only implicit, acceptance oI an abstract Iramework Ior determining possible outcomes.
137
Procedures are the central presupposition Ior the emergence oI positive law. Not only are they the only mechanism available Ior the operation oI the new level oI
reIlexivity involved in the "normative regulation oI the creation oI norms,"
138
they are the (quasi-) medium
139
around which the diIIerentiation oI law Irom religion,
morality, and scientiIic truth becomes possible. According to Luhmann, the central premise oI positive law is production and alterability through enactment, that is,
through procedurally correct decision. This can be put two waysone legal, the other politicalthat indicate reIlexivity: Norms regulate the making oI norms, and
decisions regulate the making oI decisions. The norms that guide the making oI norms, such as the constitution, are a set oI norms like any other. So are the decisions
that regulate decision making. Positive law means rejecting the possibility oI extralegal

Page 337
sources oI law and even oI a hierarchy oI legal levels. Nevertheless, it would be a mistake to interpret the positivity oI law as meaning that normatively valid decisions
are the only source oI law. Norms, even potentially legal norms, emerge Irom all spheres oI society. Legislation involves a process oI making a selection Irom what is
projected Irom elsewhere as potential law and then validating that selection as law. Only what passes through the procedural Iilter oI legislation becomes valid law in
this model.
140
Luhmann's treatment, unlike some other versions oI legal positivism, leaves room Ior sources oI law creation other than legislative enactment. While he thereby
prepares the way Ior reconciling historical and positive jurisprudence, he does so in an undiIIerentiated way at both poles. First, he does not distinguish between active
and passive societal sources oI law creation. This is connected with his Iocus on isolated subsystems and a theoretically anchorless, nondescript everyday liIe rather
than, in the manner oI Parsons, an organized social sphere in which culture and associations intersect. While he notes that patterns or institutions in any sphere oI
society can be turned into legal norms, he does not note the diIIerence between social norms and social Iacts raised to the level oI legal validity. Thus, he bypasses the
question oI whether normative as against legal validitv can be produced apart Irom legislation and whether, thereIore, the legislative process in the case oI valid norms
is a source oI a higher validity or only oI a Iorm oI binding and possibly universalization. Most importantly, he does not raise the question oI whether or not a special
role is played, as a source oI norms Ior the legal system, by the processes oI norm creation through coming to an understanding that he has described.
Second, his Iramework expresses uncertainty, similar to that oI the tradition oI legal positivism taken as a whole, concerning the legal as against the political character
oI positive law. The issue is whether or not the making and operation oI positive law are Iunctions oI the political subsystem, in a way recalling the incorporation oI
other dimensions and mediations oI civil society into this subsystem. In early writings on this topic (1967), Luhmann tended simply to aIIirm that the political subsystem
supports and administers the mechanisms oI positive law.
141
Later (1976), with the diIIerentiation and autonomy oI the legal subsystem already

Page 338
aIIirmed, Luhmann was still constrained to point to the overlap oI the institutions and events oI the two subsystems and to note the diIIiculties Ior law making inherent in
legislation by a political body, the parliament.
142
Indeed, this overlap goes so Iar that the institutions oI making, applying, and executing law turn out to be the three
branches (legislative, executive, and judicial) oI the central, administrative, decision-making subsystem oI politics.
143
Thus, his assertion oI the autonomy oI the legal
system has some diIIiculty in overcoming his other, earlier depiction according to which positive law is ''state" law whose "destiny is bound up with that oI the political
system oI society."
144
Luhmann does speak oI diIIerent selectivity
145
and, later, oI diIIerent connections, linkages, and exclusions
146
oI the two systems, legal and political, even in the case
oI shared institutions and events. One might argue, although he does not, that legislative decision making selects norms Ior legalization, whereas legislative procedures
endow laws with the structure oI validity. Finally, as in Luhmann's recent conception oI law as an autopoietic system, one might consider law to be normatively closed
while cognitively open. The Iirst oI these dimensions would yield legal autonomy and selI-reproduction, while the second would provide Ior openness to the political
system in which learning takes place.
147
The reason why none oI these strategies will succeed in providing Ior the diIIerentiation and autonomy oI the legal system lies
deep in Luhmann's conception oI positive law and the shiIt contained therein toward a cognitive style oI expectation.
Positive law is here understood as a system oI norms that comes into being through decision and can be changed through decision. We Iind in the processes oI law positing
decisions primarily a cognitive learning, determined by goals, one hardly structured by norms. . . . Correspondingly those aIIected by law must constantly learn the changes in the
law, whether or not they are disappointed. They will have to take a primarily cognitive attitude to law.
148
What is involved here is not a simple shiIt Irom a normative to a cognitive style oI expectation but a shiIt to a combination ("contrary ordering") in which we cognitively
expect a normative style oI expectation. Positive law can adopt this structure because oI the diIIerentiation oI legal procedures and roles. In the context oI the

Page 339
general alterability oI all legal norms, including constitutional ones, the natural attitude is that oI learning. But positive law, in order to remain law at all, must preserve its
normative Iunction within alterability. In principle, this is possible as long as structures are not problematized in the situations they structure and as long as these
situations are diIIerentiated Irom others in which the same structures are questioned and perhaps changed.
149
It is judicial procedure and the role oI the judge that
institutionalizes a normative attitude to structures within a system oI positive law. OI course, given the obvious alterability oI law, even judges must "learn not to learn."
While it is the task oI the legislator to process disappointments, to correct expectations, and to take responsibility Ior Iailure to learn, the judge is ordinarily determined
not to learn Irom the lawbreaker and learns how not to learn in the Iace oI inIringed norms.
150
One mechanism Ior this in the courtroom, paradoxically, is the
technique oI converting conIlicts about norms into conIlicts about Iacts, normative stakes into cognitive ones. In this way, judges need never expose their own norms to
critical questioning and need not learn Irom those who have disappointed their expectations because oI alternative normative expectations.
151
Can a system that combines normative and cognitive expectations still be described as primarily normative? Luhmann does not ask this question, but he does indirectly
answer it. He introduces the problem oI legitimacy as a way oI dealing with the binding character oI the legal system as a whole. Here, too, the issue is the combination
oI learning and nonlearning, oI cognitive and normative expectation. Both those who make and those who are aIIected by decisions avoid learning in the context oI
legal contingency only at their peril. Legitimacy in this context is deIined by the possibility oI assuming "that any third parties expect normatively that the directly
aIIected persons cognitively prepare themselves Ior what the decision makers communicate as normative expectations."
152
An assumption is a cognitive expectation.
Legitimacy is a circle oI cognitive expectations in which only third partiesjudgesare expected to expect normatively, and even their normative expectation oI
others is only that they will cognitively adapt to the judges' normative expectation. No wonder that Luhmann, almost uniquely in the sociological literature, considers
physical Iorce to be an

Page 340
essential legitimating Iactor,
153
most likely because it is the basis oI the judges' oIten mistaken but not thereIore abandoned expectation that potential lawbreakers will
cognitively adapt. The structure oI the law in this conception rests only on attitudes oI cognitive expectation and the learned Ialse consciousness oI judges.
With all this said, the idea oI law as an autopoietic, normatively closed, and cognitively open system seems to be merely a verbal solution to the problem, or at best a
normative desideratum Ior the reconstruction oI law. It is hard to see how Luhmann actually conjoins the two premises that "there may be political control oI
legislation, but only the law can change law."
154
Even iI it is insertion into the legal system, with its own internal requirements, that turns legislation into law, a normative
attitude oI expectation would be saved only as a characteristic oI a purely intellectual system incapable oI perIorming Iunctions Ior the rest oI society, aside Irom
motivating the Iunctionally necessary Ialse consciousness oI judges. Because he has never Iound an integrating medium Ior law comparable to money or power,
Luhmann's claims Ior boundary-maintaining, selI-productive autonomy sound rather hollow. Thus, while the cognitively open dimension oI law would remain rooted in
the political system, which in turn is not opened up to the normative inputs oI law, its normatively closed dimension would be suspended without social Ioundations, or
at best become one oI the closed rule systems established and institutionalized in the cultural sphere alone. Such are the consequences oI abandoning the constitutive
links oI law to a Iramework oI societal action, association, and communication and oI the one-sided acceptance oI the privileged relation oI law to the political system,
a deIiciency balanced only partially by the aIIirmation oI the heterogeneous societal sources oI norm creation.
Despite Luhmann's selI-understanding, the idea oI law as an autopoietic system may be a normative desideratum born in a context characterized by increasing doubts
about welIare-state interventions into society, which seems to involve a loss oI legal Iormality and autonomy. But even as a project oI reconstruction, the idea oI the
autonomy oI law Irom politics requires an independent institutional context on which law can rely, without the dangers oI an alternative (e.g., economic)
instrumentalization.

Page 341
This insight calls not only Ior a notion oI civil society but also Ior its reconstruction in terms other than that oI a subsystem oI society in the manner oI the Parsonian
societal community. It is in the context oI such a reconstruction that Luhmann's notion oI autopoiesis Iirst becomes serviceable Ior a postinterventionist model oI the
relations oI the political system to the other spheres oI society.

Page 343
III-
THE RECONSTRUCTION OF CIVIL SOCIETY

Page 345
8-
Discourse Ethics and Civil Society
We have beIore us two theoretical topoi: modern civil society and discourse ethics. The Iirst evokes the theme oI classical liberalism: The term "civil society" today
calls to mind rights to privacy, property, publicity (Iree speech and association), and equality beIore the law. The second, with its emphasis on the equal participation
oI everyone concerned in public discussions oI contested political norms, obviously reIers to the principles oI democracy. The current vogue in political theory is (once
again) to view liberalism and democracy as Iundamentally antithetical. DeIenders oI the core tenets oI classical liberalism tend to see democracy, with its emphasis on
majority rule and participation, as either illusory or, even worse, dangerous to existing liberties, unless suitably controlled or restricted.
1
Advocates oI direct or radical
democracy, on the other hand, have come to stigmatize the liberal tradition itselI as the main impediment to achieving a participatory democratic society.
2

Nevertheless, we contend that the plausibility oI each depends on its intimate conceptual and normative relation with the other. Even more, we assume that the deIense
and expansion oI acquired liberties rests on the Iurther democratization oI the institutions oI modern civil society and on their achievement oI greater inIluence over the
polity. We shall demonstrate this thesis by exploring the concepts oI democratic legitimacy and basic rights in the Iramework oI the theory oI discourse ethics and by
establishing the connection oI both to a coherent conception oI a modern, and potentially democratic, civil society.

Page 346
Fortunately, we are not speaking in a void. The connection between the two principles has been made by many contemporary collective actors in the West and in the
East who have put the project oI the deIense and/or democratization oI civil society on their political agendas.
3
By "civil society," these actors have in mind a
normative model oI a societal realm diIIerent Irom the state and the economy and having the Iollowing components: (1) Pluralitv: Iamilies, inIormal groups, and
voluntary associations whose plurality and autonomy allow Ior a variety oI Iorms oI liIe; (2) Publicitv: institutions oI culture and communication; (3) Privacv: a domain
oI individual selI-development and moral choice; and (4) Legalitv: structures oI general laws and basic rights needed to demarcate plurality, privacy, and publicity
Irom at least the state and, tendentially, the economy. Together, these structures secure the institutional existence oI a modern diIIerentiated civil society.
The rediscovery oI the key components oI civil society by contemporary collective actors, however, does not in itselI imply its normative justiIication. The projects oI
social movements are hardly selI-validating. Furthermore, the normative ideals oI civil society are not without their critics. As we have seen, Hannah Arendt and
Michel Foucault have each made powerIul arguments attacking these claims.
4
For Arendt, the diIIerentiation oI a social realm distinct Irom the state was the beginning
oI a IateIul depoliticization oI society, leading to the collapse oI the boundary between public and private and the emergence oI both mass society and totalitarianism.
For Foucault, the very norms oI civil society constituted only the visible support oI less obvious social disciplines and microtechnologies that combine into a new and
seamless system oI bondage. We should remember, too, that the young Marx, the Iorerunner oI these views, produced powerIul arguments Ior equating civil with
bourgeois society and the separation oI state and society with political alienation.
5
II these and other critics oI the norms oI civil society are to be answered, it must be
on the basis oI a new, comprehensive, and justiIiable practical political philosophy. It is our contention that discourse ethics, suitably reinterpreted, is the best candidate
to accomplish this task.
Admittedly, the theory oI discourse ethics also has its diIIiculties. First, there is some question whether its domain oI application is

Page 347
morality, politics, or both. Second, it has been argued that the theory has authoritarian implications. Third, it is uncertain whether discourse ethics can make genuine
universal claims without prescribing a particular Iorm oI liIe. Finally, the relation oI discourse ethics to democratic and liberal institutions has never been satisIactorily
elaborated.
We hope to show that it is possible to articulate plausible responses to all oI these issues. We shall do so in Iive steps. Beginning with a discussion oI Habermas's most
detailed version oI discourse ethics, we Iirst consider the proper object domain oI the theory. We then argue that, when suitably reinterpreted, discourse ethics avoids
authoritarian implications. In order to make this point, we replace "generalizable interests" with "rational collective identity" as the legitimate substantive reIerent oI
Iormal discursive procedures. Next, we Iocus on the relationship oI discourse ethics to concrete Iorms oI liIe (Sittlichkeit). We go on to argue that while no single
model oI the good liIe Iollows Irom discourse ethics, this need not mean an insoluble institutional deIect Ior the theory. It is in this context that the category oI civil
society allows us to bring together a plurality oI Iorms oI liIe with a political model that implies the institutionalization oI discourses. SpeciIically, we link discourse ethics
and modern civil society through the categories oI democratic legitimacy and basic rights. Finally, we shall try to show that our reinterpretation oI discourse ethics has
the utopian horizon oI what we shall call a "plurality oI democracies."
The Object Domain of Discourse Ethics
The basic Iramework oI discourse ethics consists oI two dimensions.
6
The Iirst speciIies the conditions oI possibility Ior coming to legitimate rational agreement; the
second speciIies the possible contents (on a Iormal level) oI such an agreement.
7
A legitimate or rational procedure Ior coming to an agreement has been deIined by
Habermas as the metanorm that prescribes the only valid procedure Ior grounding or justiIying norms oI action.
8
No norm is assumed Irom the outset to be valid
only the procedure Ior validating norms can make such a claim legitimate. According to Habermas, a norm oI action has validity only iI all those possibly aIIected by it

Page 348
(and by the side eIIects oI its application) would, as participants in a practical discourse, arrive at a (rationally motivated) agreement that such a norm should come into
or remain in Iorce.
9
What is to be understood as rationally motivated agreement, however, has rather demanding preconditions. In order that all those aIIected have
an "eIIective equality oI chances to assume dialogue roles," there must be a mutual and reciprocal recognition, without constraint, oI each by all as autonomous, rational
subjects whose claims will be acknowledged iI supported by valid arguments.
10
But, in order that the dialogue be capable oI producing valid results, it must be a Iully
public communicative process unconstrained by political or economic Iorce. It must also be public in terms oI access: Anyone capable oI speech and action, who is
potentially aIIected by the norms under dispute, must be able to participate in the discussion on equal terms. Furthermore, the participants must be capable oI altering
the level oI discourse in order to be in a position to challenge traditional norms that may be tacitly presupposed.
11
In other words, nothing can or should be taboo Ior
rational discoursenot the preserves oI power, wealth, tradition, or authority. In short, the procedural principles underlying the possibility oI arriving at a rational
consensus on the validity oI a norm involve svmmetrv, reciprocitv, and eflexivitv.
12
These Ieatures constitute an "ideal speech situation," in which the validity claims implicitly raised in any act oI communication can be discursively redeemed. It should
be stated at the outset, however, that a theory oI legitimacy should not be conIused with a theory oI organization. II we view the much-disputed concept oI "ideal
speech situation" as a set oI criteria (metanorms) that enable one to distinguish between legitimate and illegitimate norms, we can avoid the conIusion caused by
interpretations that identiIy the Iormal rules oI argumentive speech or discourse as a concrete utopia. The "ideal speech situation" reIers solely to the rules participants
would have to Iollow iI they were to strive Ior an agreement motivated by the Iorce oI the better argument alone. II these conditions are not metiI, Ior example,
actors in a debate do not have equal chances to speak or to challenge assumptions; iI they are subject to Iorce or manipulationthen participants are not taking all
other arguments seriously as arguments and hence they are not really engaging in argumentative speech.

Page 349
Clearly, not all processes oI coming to an agreement satisIy such conditions. Habermas (and Karl-Otto Apel) repeatedly distinguish between "rational" and "empirical"
consensus. Most processes oI consensus Iormation are "only empirical."
13
The norms oI discourse that are the source oI validity are not produced by agreements;
rather, they are the conditions oI possibility oI valid agreements. The results oI actual agreements carry normative validity only to the extent that they are consistent with
the metanorms. On the other hand, and some think paradoxically, Habermas insists on an actual rather than a virtual dialogue because only an actually carried out
discourse allows an exchange oI roles oI each with every actor and hence a genuine universalization oI perspective that excludes no one.
14
He thereby distinguishes
himselI Irom all approaches that assume agreement to Iollow Irom monologically attainable truth as well as Irom most traditions oI contract theory (which postulate a
discursive model only in terms oI a myth oI origins). Only an actual, practical discourse cooperatively engaged in by all participants potentially aIIected by the norm
under discussion can lead to a rational consensus on its validity, Ior only under such conditions can we know that we, together, and not privately, are convinced about
something. The metanorms oI discourse ethics are thus peculiar in the sense that their normative implications are available only in contexts oI actual dialogue.
Accordingly, Habermas has reIormulated the Kantian categorical imperative along lines compatible with the procedural rules oI argument: "Instead oI prescribing to all
others as valid a maxim that I will to be general law, I have to oIIer my maxim to everyone with the aim oI discursively testing its claim to universalizability. The
emphasis shiIts Irom what each can will without contradiction to be a general law, to what all can will in agreement to be a universal norm."
15
The idea oI a rational consensus, however, involves more than the actual participation oI every aIIected person in the relevant discussion. In addition to a process oI
consensual will Iormation, our assertion that a norm is legitimate means that we hold it to be right and not merely conIormable to our collective will. Habermas insists
that discourse ethics, like all cognitivist ethics, assumes that claims to normative validity have cognitive meaning and can be handled,

Page 350
with certain adjustments, like cognitive truth-claims.
16
The Iactual recognition oI a norm by a community merely indicates that the norm could be valid. Its validity can
be ascertained only iI we make use oI a ''bridge principle" that establishes a connection between the process oI will Iormation and the criteria Ior judging the
acceptability oI a particular norm. The Unparteilichkeit oI judgment must complement the Unbeinflussbarkeit oI collective will Iormation.
17
In elaborating this second aspect oI discourse ethics, Habermas addresses the dimension oI content alluded to above. This brings us to the second aspect oI discourse
ethics: the Iormal content oI agreements. Habermas maintains that in order to be objective (unparteilich), rational, and legitimate, norms oI action upon which we
agree must express a generalizable interest:
18
Every valid norm must satisIy the Iollowing condition; "All aIIected can accept the consequences and the side eIIects that
its general observance can be anticipated to have Ior the satisIaction oI evervone´s interests (and these consequences are preIerred to those oI alternative possibilities
Ior regulation)."
19
This "principle oI universalization" requires actual discourses iI those aIIected are to be able to discern what all can agree to recognize as a universal
norm.
Thus Iar we have simply summarized Habermas's Iormulation oI discourse ethics. As several critics have recently pointed out, however, the status or object domain oI
the theory is unclear.
20
On the one hand, Habermas clearly considers it to be a universalistic moral theory in the Kantian tradition. On the other hand, he also presents
discourse ethics as the heart oI a theory oI democratic legitimacy and the core oI a universalist conception oI human rights that provide alternatives to traditional and
neocontractarian theories. To make matters even more complicated, Habermas has argued that, as a principle oI legitimacy, discourse ethics can resolve the apparent
riIt between legality and morality by revealing the political ethic underlying law.
21
His purpose is to account Ior the link between morality and legality in a way that
unlike Marxist approaches, which seek to abolish the distinction between the twopresupposes their diIIerence but nevertheless adjusts Iormal law to moral
principles. The Iirst question to address, then, is what exactly is the object domain oI discourse ethics? Is it a theory oI morality, or a theory oI political legitimacy? Can
it be both?

Page 351
We intend to deIend discourse ethics as a political ethics and as a theorv of democratic legitimacv and basic rights. We hold that it provides a standard with
which we can test the legitimacy oI sociopolitical norms. Terms such as "public dialogue," "general interests," ''all those aIIected," and "social norms" do in Iact evoke
the categories oI political philosophy. The theory becomes unnecessarily overburdened when presented as more than this. Indeed, the two most signiIicant objections
that have been raised against the ability oI discourse ethics to serve as a moral theory Iocus on those dimensions that make it a plausible candidate Ior a theory oI
democratic legitimacy, namely, the reIormulated principle oI universalization and the requirement oI an actual dialogue.
22
We want to bracket the question oI which
general theory works best in the realm oI autonomous moral judgment. We believe it is nonetheless possible to deIend discourse ethics as a political ethic without
committing oneselI to a speciIic moral philosophy.
This means that we construe the project oI discourse ethics as an attempt to employ the insights oI deontological ethical theory primarily against legal positivism and
legal realism as well as against systems theory oI the Luhmann type. The task, in short, is to show that there is a normative and rationally deIensible component oI
legality and politics that, independent oI sanctions or empirical motives, accounts Ior the obligatory dimension oI legal norms and the legitimacy oI a sociopolitical
system.
DiIIerentiating between a general moral theory and a theory oI political legitimacy, however, leaves us with a key question: How does one draw the boundary between
the two? It is not suIIicient to state that morality entails the individual reIlections oI a moral conscience whereas justice concerns social norms and requires a real
dialogue, Ior both morality and legality relate to societal norms, and the issue at hand is precisely the reach oI law with respect to these norms. Nor are those attempts
convincing that try to draw the boundary between the two by designating certain spheres oI liIe as private by deIinition and oII-limits to law and others as public and
thus open to legal-normative regulation. In our view, this approach cannot work, Ior a society's understanding oI the institutional arrangements and relations that should
be set beyond justice and leIt to individual judgment changes over time. Moreover, the designation "private" with regard to institutions and

Page 352
relations does not exempt them Irom satisIying the demands oI justice but, rather, implies a diIIerent Iorm oI legal-normative regulation.
23
One cannot reason Irom a
spatial metaphor or division among institutions to designate the boundary between private and public, between what should be leIt to the moral choice or personal
judgment oI individuals and what should be legally regulated. Instead, we must start Irom the assumption that privacy attaches to the individual in certain capacities (as
an autonomous moral subject), regarding certain choices (those impinging on identity needs), and within the Iramework oI certain relations (Iriendship, intimacy) that
we must be ready to analyze and give arguments Ior. Indeed, the private and even the intimate "spheres" have always been constituted and regulated by law, even iI
what is constituted includes a domain oI autonomous judgment that can come into conIlict with law. Thus, we insist on retaining the analytic distinction between
between a domain oI autonomous moral reIlection or judgment and a domain oI legal norms, but we reject any attempt to set up a one-to-one correspondence
between this distinction and spheres oI liIe or sets oI institutions. Rather, law must be selI-limiting with respect to the autonomous judgment oI individuals, provided that
this does not entail the violation oI basic principles oI justice. Privacy rights operate precisely in this manner, although just what content Ialls within the protection oI a
right to privacy is, oI course, open to debate and revision.
From the other side, our interpretation oI discourse ethics as a theory oI democratic legitimacy and basic rights presupposes the sociological insight concerning the
positivization oI law and the corresponding separation between the spheres oI legality and morality. Yet our version oI the theory rejects the view that the total
denormatization oI politics or law and the depoliticization oI morality are the inevitable consequences oI this process. How can this apparent paradox be resolved?
It is clearly Habermas's view (and ours) that the development oI autonomous universalist morality as well as the emergence oI a Iormal, diIIerentiated system oI
positive law must be seen as immense historical achievements. These developments, moreover, are linked to the emergence oI speciIically modern conceptions oI
democracy and rights, representing the constitutive conditions oI

Page 353
a modern version oI civil society. There is, however, another side to this process: The uncoupling oI positive legal norms Irom the body oI private morality based on
principles that accompanied the emergence oI constitutional states and capitalist market economies entails a potential conIlict between the loyalty oI the citizen to the
abstract rules oI the legal system (which are valid only Ior the area "paciIied" by a particular state) and the "cosmopolitanism oI the human being" whose personal
morality makes general claims.
24
Even more important, ever since the decline oI modern natural law theories and the rise oI legal positivism, the claim that laws have
normative content beyond the correctness oI the appropriate legislative and legal procedures, that they are binding independently oI relevant sanctions, has been
repeatedly disputed. The diIIerentiation between legality and morality has involved both the separation oI politics Irom the everyday liIe oI citizens and the
denormatization oI legality itselI, at least according to a good deal oI legal theory since the nineteenth century.
25
Moreover, when law is understood as the will or command oI the sovereign (Hobbes, Austin), and when constitutions and Iundamental rights are declared to be only
special instances oI positive law, the results go beyond the separation oI morality and law. In eIIect, legal positivism announces the denormatization oI law, its
transIormation into a class oI empirical Iacts. Obligation is turned into prudent behavior in the Iace oI possible sanctions. Even within legal positivism (H. L. A. Hart),
such extreme results are oIten rejected and the ideas oI law as command and obligation as calculation have been decisively reIuted. Nevertheless, it is hard to see how
a conception oI law as a system whose purely legal terms need be related only to one another and satisIy only demands oI consistency (Kelsen) or validity in terms oI
a "secondary" legal order (Hart) can lead to anything like a genuine political ethics capable oI grounding legal or political legitimacy. This is even harder to see in the
case oI the view that reduces law to sociological predictions about what courts, legislatures, communities, and political oIIicials or other holders oI power will wish to
enIorce with sanctions (legal realism; some versions oI critical legal studies).
In his debate with Weber and Luhmann concerning the Ioundation oI legal-rational domination, Habermas has repeatedly

Page 354
pointed to the impossibility oI deriving the legitimacy oI a modern legal system as a whole solely Irom the Iormality and systematic nature oI legal procedures.
26
Law
as legitimate authority rests on extralegal sources oI justiIication. ReIerences to constitutions as the ultimate source oI authority, at least on the part oI Iormally
democratic states, implies that the legitimacy oI law is ultimately parasiticon the principles oI democracy and basic rightsembodied in constitutions and in the
democratic process allegedly behind the development oI constitutions. The principles oI democratic legitimacy and basic rights underlie the authority oI law. These
principles, however, can no longer be deIended as sacred, "selI-evident truths," as they were both in theories oI natural law and in republican theories oI civic virtue.
The task oI discourse ethics is to provide a contemporary equivalent oI such theories while avoiding their presuppositions. Thus, the principles oI democracy it justiIies
must not be seen as given, once and Ior all, but as the outcomes oI an original and repeatable communicative process that ascertains the generality oI admissible norms
and the discursive redeemability oI the validity claims with which they appear.
We propose to deIine legality in terms oI the old reIerence to Iormal sanctions that potentially invoke the executive and judicial powers oI the modern state on behalI
oI valid norms. Moral rules cannot call upon such enIorcement. Accordingly, discourse ethics as we see it would apply to the legal and political system as a whole, as
well as to particular complexes oI legal norms that depend both on sanctions and on the interpretation and compliance oI those concerned. In the Iirst instance, we
reinterpret discourse ethics as a principle oI democratic legitimacy; in the second, as part oI a theory oI basic rights that can be institutionalized. As we shall show,
these two dimensions oI discourse ethics imply a province oI autonomous moral judgment that is beyond its reach but nonetheless is its presupposition and must be
guaranteed by basic rights. Let us Iirst address this latter issue.
We are assuming that discourse ethics pertains to the sphere oI legality in two interdependent yet distinct dimensions: democratic legitimacy and basic rights. Each oI
these dimensions touches on morality. However, even iI we can say empirically where legality begins and autonomous judgment ends by reIerring to Iormal

Page 355
sanctions, we have not yet touched on the normative question oI where these boundaries ought to be. OI course, all modern civil societies draw a boundary between a
realm oI autonomous judgment and what can be legally regulated, but they draw these boundaries at diIIerent places. In case oI disputes, the issue is inevitably whether
the boundaries should be drawn Irom the point oI view oI legality or individual judgment, public discourse or private moral reIlection. In our opinion, in such cases
discourse ethics must be considered superior to any monologically attained moral standpoint, at least in the Iirst instance. This is so because only in an actual discussion
with everyone potentially aIIected by a legal norm can we Iind out what, iI anything, is common to us all, what should be the domain oI legal regulation, what Iorms oI
political decision making are legitimate, what should be leIt to the autonomous subject's personal judgment, and what must be compromised with. In other words, it is
only aIter debatable issues have been publicly discussed that we can decide which must be considered "private," that is, leIt to the autonomous judgment oI the
individual to determine with respect to a personal ideal oI the good liIe.
27
Discourse ethics thus has a double status: Its speciIic object domain comprises institutionalized social relations, the legal and political system as a whole, and particular
laws and rights. It also provides a way to decide the boundary question between autonomous individual judgment and justice. To be sure, the boundaries drawn Irom
the point oI view oI actual discursive processes may not be acceptable Irom the standpoint oI the moral convictions or identity needs oI individuals or groups. A
majority might seek to legally regulate areas oI decision making that have previously been construed as private and that a minority does not want to deliver to such
regulation. Conscientious objection and civil disobedience are legitimate options Irom the moral point oI view. They should be respected as eIIorts to acknowledge
publicly drawn boundaries while attempting to circumvent or change them Irom the point oI view oI an unusually intense concern. However, in these cases, the claims
oI justice have priority in the Iollowing sense: One cannot be compelled to renounce one's way oI liIe, identity, or moral convictions, and yet the moral consciousness
that doesn't want to be

Page 356
unjust must be selI-limiting in that it must accept the principle oI democratic legitimacy and basic rights provided that these are selI-limiting in turn. In other words, they
must protect the space Ior articulating diIIerence. This means that, in the case oI conIlict between conceptions oI the good liIe and legality, it should not be deemed
unethical Ior the individual to Iollow his or her moral conscience or judgment and to act accordingly. But one must, nonetheless, act under the dictates oI selI-limitation.
Within the Iramework oI a democratic constitutional polity, a morally legitimate violation oI law presupposes the acknowledgment oI constitutional principles,
acceptance oI the democratic order, and a symbolic orientation oI the action toward inIluencing public opinion and developing a new normative consensus.
28
The
legal response to such action ought to be able to distinguish it Irom common criminality and thus avoid being overly harsh.
29
All principled acts oI disobedience, Irom
individual acts oI conscience to tactics oI social movements, rest on these ideas.
Thus, our interpretation does not collapse the boundary between morality and legality. On the contrary, it preserves a realm oI autonomous judgment Ior the individual.
At the same time, it protects positive law Irom the potentially incapacitating interIerence oI absolute moral judgments without thereby delivering it into the hands oI legal
positivists. Indeed, once we restrict the relevance oI discourse ethics to questions oI democratic legitimacy and rights, it leaves room Ior a variety oI moral principles,
cultural values, and ways oI liIe. Without having to judge the internal adequacy oI any oI these, discourse ethics adjudicates between them only in cases oI conIlict over
general societal norms. Thus, the autonomy oI conscience and the plurality oI ways oI liIe are respected by the principles oI democratic legitimacy and basic rights,
even though the latter bring principles to bear on the domains oI law and politics. Although in this case, too, processes oI discursive will Iormation decide the boundary
between "private" and "public," they cannot entirely abolish the private (understood here as the domain oI autonomous individual moral choice or judgment).
The metanorms oI discourse themselves presuppose, even iI they cannot ground, the autonomy oI the individual moral conscience.

Page 357
II all persons aIIected have the chance to assume dialogue roles, iI the dialogue must be unconstrained, iI each individual can shiIt the level oI discourse, and iI
everyone can articulate their need interpretations, then practical discourse presupposes autonomous individuals with the capacity not only to be selI-reIlective regarding
their own values but also to challenge any given norm Irom a principled standpoint. The processes necessary Ior the requisite socialization oI individuals would be
impossible without institutionalizing moral autonomy and the mutual recognition oI diIIerence secured by rights.
Thus, the very rules that underlie argument and the cooperative search Ior consensus imply the distinction between morality and legality. By articulating the metanorms
oI the principle oI democratic legitimacy and some key rights, discourse ethics presupposes the justiIication Ior the autonomy oI the moral sphere and, as it were, its
own selI-limitation. There is yet another reason Ior this. No consensus, no matter how unanimous or long-lasting, can know itselI to be permanent, Ior there is no
automatic coincidence between the just and the moral, between what is deemed normatively right at any given time Ior a solidary community and what is always
morally acceptable to each individual. Even iI the legal norm has survived the most ideal process oI discursive testing, it may yet conIlict with the particular values or
identity requirements oI an individual. Neither moral autonomy nor individual identity can be sacriIiced to the collective identity or consensus oI a group, because this
would violate the very raison d'être oI discourse ethics: to provide a Iormal principle Ior the legitimacy oI norms in a society that is plural and composed oI individuals
with distinct and diIIerent conceptions oI the good liIe. Even in a situation that closely approximates the requirements oI symmetric reciprocity, there is no basis Ior
assuming either the absence oI diIIerence or the absence oI change. Every consensus is, aIter all, only empirical and must be open to challenge and revision.
30
From
the standpoint oI justice, we cannot know that today's value change on the part oI a minority oI individuals might not become tomorrow's general will. Thus, individual
judgment, diIIering ways oI liIe, and experiments with new ways must be granted autonomy Irom the current consensus on what is just.

Page 358
One might nonetheless object, Irom the point oI view oI a moral consciousness, that a separate theory oI ethics Ior the realm oI law and politics is unnecessary. As a
moral subject, I obey the law because it is right, and when it becomes morally wrong to do so, I would have to disobey the law, whatever the consequences to myselI.
Morality is certainly wider than legalityIrom both objective and subjective points oI view. Formal law cannot regulate every domain oI action, whereas, Irom the
subjective point oI view, morality ought to. The moral consciousness could grant the necessity oI law and sanctions because we are not gods, not always moral, and
thus need external constraint in certain cases. But iI the moral component oI law is equivalent to what the moral reIlections oI an individual actor could arrive at, then
there is no need Ior a separate ethical theory Ior politics. Why develop a discourse ethics at all?
There are, in the modern context, two reasons why we cannot move directly Irom morality to legality or resynthesize them, as it were. First, as is well known, we
moderns live in a plural moral universethe plurality oI value systems, modes oI liIe, and identities would be violated iI laws or political decisions were made Irom the
point oI view oI any oI them. Every good liberal thus can argue against making any single moral standpoint absolute Ior the whole oI society. To do so would lead
either to subjugating individual dignity and rights to the concerns oI general welIare or to violating the integrity oI those who do not share the particular concept oI the
good liIe that has become dominant. Not all action, not even all moral action, can or ought to be institutionally regulated.
The second and more compelling reason why we cannot equate the obligatory dimension oI social/political norms with what motivates even the postconventional moral
actor is that the genesis oI legality, unlike morality, can and must involve in principle actual discourse. To be sure, Habermas himselI tends to conIlate morality and
legality because he rightly sees that moral testing involves an inner dialogue to which the rules oI argument apply.
31
It would thus seem possible, iI one Iollowed these
rules and considered the potential side eIIects oI a maxim on all others, that one could arrive at the same judgment that an actual discourse would yield. The core
diIIerence between virtual and actual dialogue would, nonetheless, remain: Only an actual dialogue in which all concerned can

Page 359
participate on equal terms oI mutual recognition would involve a reversal oI perspectives and yield or reaIIirm a we, a solidary collectivity, having a collective identity
and the capacity Ior articulating a general or common interest. As Hannah Arendt pointed out long ago, only in a public space can a public opinion merge. Even iI one
imagined an ideal moral subject, able to consider all the possible arguments oI everyone involved, the outcome would not automatically converge with the political
judgment oI a duly constituted public, because the relevant, emergent collective identity would be missing. At best, an idealized, selI-reIlective moral judgment could
imply tolerance oI others and oI diIIerent arguments, but it could not yield or reaIIirm the solidarity oI a collectivity or an understanding oI what our collective identity is
and, Ilowing Irom this, what our general interests might be. This, however, is the object domain oI institutionalized norms. Nor would it yield insight into perspectives
entirely diIIerent Irom our own, and thus the possibility oI solidarity with diIIerenceand the limits this implies on normative regulationwould be missing. Indeed, it is
quite possible that a judgment could be moral and yet not be just. On our interpretation, discourse ethics implies that the justice oI justice, the legitimacy and normative
Iorce oI law, derives in principle Irom democratic will Iormation and the articulation oI a general interest in the norm. From the point oI view oI morality, a law imposed
by an enlightened despot might be moral according to everyone's personal point oI view, and it might even articulate a general interest (the common good). Yetand
this is the limit to the standpoint oI the moral consciousnesseven iI it were moral, even iI it were to coincide with what a community would have agreed upon as its
interest, it would not be just, Ior justice requires that those aIIected determine this Ior themselves, in a discursive process oI collective will Iormation.
Let us summarize the argument thus Iar. (1) The division between morality and legality is a major and characteristic achievement oI modernity. (2) Discourse ethics
provides the core oI a normative theory oI political legitimacy and oI a theory oI rights, but it cannot serve as a moral theory inIorming the choices oI individuals in all
areas oI liIe. (3) We interpret the meaning oI justice along the lines oI a concept oI democratic legitimacy and basic rights. Accordingly,

Page 360
the object domain oI discourse ethics comprises institutionalized norms with legal sanctions attached. (4) Discourse ethics grants autonomy to other modes oI moral
reasoning. (5) Based on the theory oI communicative action, discourse ethics is able to account Ior the obligatory aspect oI social norms that is distinct Irom the
attached sanctions. (6) Political and legal institutions can be made responsible to moral insight without involving the collapse oI legality and morality. Indeed, in
constitutional democracies with civil societies, the principles oI democratic legitimacy and basic rights already are the ultimate source Ior justiIying political norms and
processes.
The Charge of Authoritarianism
The charge oI authoritarianism, leveled speciIically against Habermas's version oI discourse ethics, contends that the Iocus on rational consensus implies a Jacobin-
Bolshevik suppression oI independent ways oI liIe and, hence, oI civil society. We shall start by reIuting this charge and developing a version oI discourse ethics
immune to it. Our next step will be to deny an intrinsic connection between discourse ethics and any speciIic concrete ethos or Sittlichkeit, while demonstrating that
this does not leave it merely Iormalistic or empty. Indeed, we shall argue that discourse ethics has an elective aIIinity Ior a societal arrangement that permits a plurality
oI ways oI liIe to coexist. In this way we hope to show that, among the versions oI civil society, only the modern ones are relevant to discourse ethics.
It seems that two apparently contradictory charges have been made against Habermas's discourse ethics: authoritarianism on the one side, and excessive Iormalism on
the other. Presumably the two charges could be combined: Either discourse ethics is so Iormalistic as to have no institutional consequences, or, iI it has, they inevitably
have authoritarian implications. We preIer to deal with these charges separately, since the issues involved are completely diIIerent.
The charge oI authoritarianism has several variants. The Iirst involves a blanket application oI Hegel's critique oI Kant, linking abstract morality and terror to discourse
ethics as a whole. On this

Page 361
level, the objection has been successIully dispatched by Albrecht Wellmer.
32
More speciIic objections discover an authoritarian potential in two particular sets oI
distinctions made by Habermas: (1) between ''empirical" and "rational" consensus, and (2) between "particular" and "universal" or "general" interests. According to
Robert Spaeman, Ior example, with these distinctions "the utopian goal Ior the abolition oI domination serves precisely the legitimation oI the domination oI selI-
appointed enlighteners."
33
1. It is certainly very wrong to apply this objection to Habermas, as iI he merely belonged to the romantic, anticapitalist generation oI Marxists oI the early twentieth
century. Yet the distinction between "empirical" and "rational" consensus, iI interpreted as calling Ior the abolition oI one Ior the sake oI the other, indeed recalls the
classical Jacobin-Bolshevik contempt Ior the merely empirical people or working class.
Habermas, however, has been careIul to avoid this implication. Even aIter rejecting the Iull applicability oI a psychoanalytic model oI reIlection to a critique oI society,
he holds on to the assumption that "only the techniques oI discourse (should be used) to establish the conditions Ior beginning possible discourses."
34
He goes beyond
that model by insisting that, in the discourse whose Iunction is to establish or reestablish discourse, "there can be only participants," because no one can have "a
privileged access to truth."
35
The implication oI Habermas's argument, in other words, is not to Iorcibly replace the conditions oI one type oI discourse with those oI
another but to establish new Iorms side by side with the old ones and perhaps to revitalize existing Iorms oI public liIe. Indeed, Habermas explicitly denies a privileged
discourse oI intellectuals or political organizations that would play a "leading role" with respect to empirical processes oI communication.
36
It is Albrecht Wellmer, however, who goes the Iurthest in an antiauthoritarian direction by Irankly announcing that actual consensus necessarily means factual
consensus.
37
How, then, can we tell when an empirical consensus is rational? To doubt the rationality oI an empirical consensus means either to propose speciIic
counterarguments or to doubt the rationality oI the participants. The latter, however, cannot be grasped with the help oI the structural conditions oI the ideal speech
situation. The doubt

Page 362
remains an hypothesis that can be sustained only by carrying out a new discourse and arriving at a new agreement: The participants must recognize their previous
unreason. As Agnes Heller aptly puts it, the norms oI argument, together with the insistence on a real dialogue open to all, imply a democratic process oI will Iormation
such that the general will can, aIter all, only be the will oI all.
38
Even iI a consensus is the product oI a "rationally organized society" that allows Ior both discourse and
dissent in its public spaces, we may not assume that the rationality oI the procedure guarantees the absolute truth or rightness oI the result. The truth oI norms cannot
be established once and Ior all. The content oI a rational consensus is not necessarily truewe consider it to be rational because oI procedural norms, true because oI
good grounds that we oIIer in the discussion and that are accepted as such.
39
But we could be mistaken or, to put it better, the kinds oI reasons we are willing to
accept can change over time. At best, we can arrive at a rational grounding oI the conviction oI truth that we must treat as true but that nonetheless we, as reIlective
moderns, must consider to be Iallible and open to new arguments. Thus, the idea oI a rational consensus does not mean the attainment oI absolute truth. The possibility
oI agreeing on norms involves the possibility oI rational disagreement! In short, a rational empirical consensus, the product oI discourse, is open to learning and, oI
course, to dissent.
2. II the distinction between rational and empirical consensus (linked to the procedural dimensions oI collective will Iormation) can thus be protected against
authoritarian implications, the distinction between particular and general interests (tied to the principle oI universalization) exposes Habermas once more to these
charges, this time with regard to issues oI content rather than Iorm. As already indicated, discourse ethics tests the validity oI norms according to whether they
articulate generalizable interests. In both early and recent Iormulations, Habermas maintains that discourse ethics brings need-interpretations into discussions oI norms,
so that a constraint-Iree consensus permits only what all can want.
40
Only iI norms express generalizable interests, in addition to being the product oI a general will or
agreement, are they based on a rational and true consensus. However, given the thesis that, in Iormally democratic, capitalist, class-based societies, the results oI

Page 363
empirical discursive processes suppress "generalizable interests," Habermas has repeatedly resorted to the conditional language oI ascription: "would agree" "were
|they| to enter into an unconstrained discourse," etc.
41
The status he gives to such ascription is only that oI a social-scientiIic hypothesis that requires testing and
conIirmation in actual processes oI practical discussion. Nevertheless, the theory is ambiguous at this level. While ''discursively redeemable norms" or universalizable
general interests must be "both Iormed and discovered in processes oI practical discourse,"
42
Habermas also seems to imply that, strictly speaking, only Irom the
"perspective oI the third person, say oI the social scientist" could the model oI generalizable interests be applied critically. In earlier texts, Habermas has spoken oI
"suppressed generalizable interests" in order to relate the theory critically to those social systems that prevent the conditions necessary Ior practical discourse Irom
emerging. The apparently objective vantage point he postulates Ior social science (that oI the nonparticipant "monologically" arriving at true general interests) seems to
correspond to the old Leninist or Lukacsian point oI view Ior distinguishing between "real" "universal" vs. "Ialse" "empirical" particular interests. The ambiguous status
oI the concept oI suppressed generalizable interests thus opens Habermas to the charge oI authoritarianism.
One way to avoid this charge would be to argue that the model oI generalizable interests is not as central to discourse ethics as some interpreters, including Habermas
himselI, have maintained. To be sure, Habermas insists that the satisIaction oI interests need not be a zero-sum game, and that some interests in all societies are, in Iact,
generalizable. Yet one could argue that discourse ethics could survive an empirical situation largely to the contrary. Assuming only particular interests, the discourse
that is needed Ior agreement on the rules oI their coordination could still be seen as an expression oI the general. Even a stable compromise needs normative
grounding and rests on a consensus as to its binding character, whether traditional or discursive. Habermas has had a tendency to interpret plurality in individualistic
terms, group Iorms oI plurality as particularistic, and compromise as strategic.
43
Nevertheless, he now insists on the need to discursively map out the boundaries
between generality and plurality, consensus and compromise, giving all oI these terms a communicative Ioundation.
44


Page 364
In his most recent text on the topic,
45
Habermas has corrected his earlier Iormulation in which compromise appeared to correspond to the Iailure oI communicative
action. He still distinguishes between the attempt by all concerned to clariIy what is a common interest and the eIIort oI those seeking compromise to strike a balance
between particular, conIlicting interests. But he has come to see that binding compromise also requires speciIic conditions. Participants in a binding compromise
assume that a Iair balance can be achieved only iI all concerned can participate equally. "But these principles oI compromise Iormation in turn require actual practical
discourses Ior justiIication."
46
A nonstrategic relation to the structure oI compromise that involves accepting its underlying normativity is the sine qua non Ior stable
compromises to occur. The rules oI the game must be taken seriously. II the structure oI compromise itselI has the capacity to obligate, it is the common concern oI all.
The communicative Ioundations oI compromise among a plurality oI particular interests could be made stronger iI the typical case oI a "rational coming to an
agreement" were interpreted as that oI a rational argument Ior a plurality oI points oI view, Iorms oI liIe, or interests that could lead to compromise. According to
Wellmer, the index oI particularity that attaches to all human situations should be thought oI not as a "possible limitation to rational selI-determination and
communication" but as "moments oI situatedness" to be brought into the concept oI reason. "Exactly where uniIication cannot be attained, there at least all must have
the same right to get a hearing Ior their arguments and to participate in decisions."
47
Thus, generality is attached not to the content oI interests but to the structure that
allows all to articulate their particular interests, and this is what leads to valid and binding compromise.
As seductive as this solution to the problems raised by the conception oI general interest is, it is not entirely compelling. Indeed, Habermas has explicitly addressed a
version oI this argument, that oI Ernst Tugendhat, and rejected it. Tugendhat attempted to equate argumentation with processes oI collective will Iormation and to
excise the cognitive dimension Irom the theory oI communicative ethics.
48
On the basis oI the position that every

Page 365
"rational agreement" is in Iact an empirical one, he argues that the issue is solely one oI elaborating principles Ior equal, symmetric participation in acts oI collective
choice. Questions oI justiIication are not relevant hereacts oI collective choice are acts oI will, not oI reason.
Against this position, Habermas points out that the price oI excising the cognitive dimension oI discourse ethics is that we are no longer able to distinguish the de Iacto
social acceptance oI a norm Irom its validity.
49
II we replace the "Unparteilichkeit" oI judgment with the "Unbeinflussbarkeit" oI will Iormation, we cannot say why
even the product oI a unanimous collective choice would be binding, iI no principle besides a momentary agreement under-lies it. This is the classic objection against
theories oI democratic will Iormation and majority rule. A mere empirical consensus does not in itselI produce legitimate obligation. Nor, Ior that matter, is it stable.
Moreover, it has no authoritative character iI it can be changed at will and iI it depends only on our momentary agreement. Habermas thus repeats his stress on the
centrality oI the idea oI general interest to discourse ethics.
The insistence on the cognitive component oI norms is also meant as the basis Ior a reply to the inevitable decisionism that generally accompanies the Weberian thesis
oI the war oI the gods, that is, oI the irreducible plurality and even irreconcilability oI values in modern societies. Habermas maintains that we can give rational grounds
Ior the intersubjective recognition oI validity claims without resorting to metaphysics or dogmatism. The validity claims oI norms are not, accordingly, located in the
irrational volitional acts oI the contracting parties but in a "rationally motivated recognition oI norms, which can be questioned at any time."
50
The Iact oI pluralism need
not mean that it is impossible to separate, by arguments, generalizable interests Irom those that are and remain particular. Yet he insists that "the cognitive component
oI norms is not limited to the propositional content oI the normed behavioral expectations. The normative validity claim is itselI cognitive in the sense oI the supposition,
however counterIactual, that it could be discursively redeemed through the giving oI reasons and the gaining oI insightthat is, grounded in consensus oI the
participants through argumentation."
51


Page 366
Habermas conIuses several distinct issues. To insist that the cognitive character oI a general interest yields the validity oI a norm conIuses several meanings oI the term
"cognitive." It is one thing to argue that the principles oI argumentation can provide a metanorm to which participants can appeal in testing the outcomes (norms) oI an
existing empirical consensus. It is quite another to locate the standard Ior the validity oI norms in a concept oI general interest that is, by its very nature, ascertainable
Irom the social-scientiIic or observer's point oI view. The latter strategy revives the naturalistic Iallacy that equates the objective generality oI interests with the
universality oI norms. Indeed, Habermas seems to be conIusing two meanings oI "rationality" that he has elsewhere painstakingly diIIerentiated. The rational process oI
coming to an agreement involves principles oI argumentation that are cognitive in the sense that we can test them in discourse. Nevertheless, the processes oI raising
and arguing validity claims with respect to the rightness oI a norm is distinct Irom the rationality or cognitive character oI truth claims involved in statements oI Iact. To
treat normative validity claims like cognitive truth claims would be to conIuse the object domains explored by practical and by theoretical discourse, respectively.
Practical discourse reIers to a world (the "social world") experienced and even reconstructed in the perIormative attitude, that is, the attitude oI participants. It is
implicated in a double hermeneutic and always depends on the validity claims made by the relevant social actors. Theoretical discourse, even about society, requires
objectiIying the social actors and their actions. The language oI general or generalizable interests is theoretical in this sense. It replaces the opinions oI participants
about what they need, want, and desire with an objective judgment (based on an analysis) about their interests. Thus, Habermas's stress on the criterion oI general
interests, in response to Tugendhat's renunciation oI the rationality claims embodied in the metanorms oI argumentation, relies on the wrong discursive test. The
generality oI the interests does not yield the validity oI the norm. Indeed, the idea that the legitimacy oI a norm rests on the Iact that it reIlects a general interest makes
consensus superIluous, Ior it implies that because the norm reIlects such an interest (however this is ascertained), consensus on its validity should

Page 367
Iollow. In short, consensus would Iollow validity rather than the other way around.
52
Tugendhat's denial that the metatheoretical Ioundations oI discourse have any relevance and Habermas's insistence on the concept oI general interest as a standard Ior
testing the validity oI norms represent two mistaken solutions to the problem oI obligation. The Iormer involves arbitrariness; the latter, objectivism. To his credit,
Habermas is aware oI the embarrassment Iaced by a reIlective mode oI justiIication in the context oI value pluralism, posttraditional law, and postconventional moral
reasoning. The viable part oI his response is his insistence that the objectivity (Unparteilichkeit) oI judgment is rooted in the structure oI argumentation itselIit is not
brought in as a value Irom the outside because we happen to choose it.
53
Although every consensus can be only empirical, this does not mean that we are leIt with an
arbitrary collective will. Rational grounds can be given not Ior the truth oI values per se, but Ior their incorporation into sociopolitical norms. The principles oI
argumentation can provide a metanorm (symmetric reciprocity) to which participants can appeal in testing the outcomes (norms) oI an empirical dialogue. The
rationality oI a consensus can be tested by reIerring validity claims back to those metaprinciples that alone can make it valid and obligatory. Thus, the principles oI
discourse that imply both the consideration oI every rational argument and the respect oI everyone capable oI arguing allow us to arrive at what is normatively right.
This is the convincing part oI Habermas's position.
But this still leaves unclear the role oI the concept oI "general interest" and what the "principle oI universalization" adds to the principles oI argumentative procedure oI
discourse ethics. II "generalizable interests" reIers to "raw" need-interpretations, then the objection Iirst articulated by Hume and repeated by Agnes Heller that a
discussion around interests and needs can only be inconclusive would be legitimate.
54
On the other side, we have already shown that iI the concept oI general interest
reIers to the objective interests oI a group, then this cannot be used as the criterion Ior the rightness oI norms without authoritarian implications.
Nevertheless, the concept oI interest is important to our interpretation oI discourse ethics. We suggest that the term "general

Page 368
interest" must give way, or rather priority, to the idea oI "common identity." In societies characterized by a plurality oI value systems, modes oI liIe, and individual
identities, discourse ethics provides a way oI discovering or reaIIirming what, iI anything, we who come into contact with one another and who are aIIected by the
same political decisions and laws have in common. As stated earlier, we aIIirm and in part constitute through discourse who we are, and under which rules we wish to
live together, apart Irom our personal or particular identities and diIIerencesthat is, what our collective identitv as members oI the same civil society is. Interpreted
in this way, the discovery oI generalizable interests in discussion implies something prior, namely that, despite our diIIerences, we have discovered, reaIIirmed, or
created something in common that corresponds to a general social identity (which is itselI open to change). A public discussion can show us that, aIter all, we do have
something in common, that we are a we, and that we agree on or presuppose certain principles that constitute our collective identity. These become dimensions oI the
content oI legitimate legal norms and the Ioundation oI social solidarity. The collective identity oI a community can then provide the minimum criterion, with respect to
content, oI the legitimacy oI norms in the negative sense as that which cannot be violated.
In his writings on legitimation problems, Habermas has explicitly stated that the claim to legitimacy is related to the social-integrative preservation oI a normatively-
determined social identity. "Legitimations serve to make good this claim, that is, to show how and why existing (or recommended) institutions are Iit to employ
political power in such a way that the values constitutive Ior the identity oI a society will be realized."
55
Social integration, social solidarity, and collective identity are
the "societal" (in Habermas's terms, the liIe world) reIerents oI normative claims to political legitimacy on the part oI polities. While the political-administrative system
cannot create identity (or meaning), its claim to legitimacy involves the nonviolation oI collective identity and the reinIorcement oI social solidarity and social
integration.
56
One could object that resorting to the concept oI collective identity merely transposes to another level the problems mentioned earlier involving the concept oI general
interests. What

Page 369
precludes a given collective identity Irom being authoritarian? Whose interpretation oI group identity is to prevail? How could it be anything other than particular, and
why make universalist claims Ior its deIense? The answer lies in the peculiarities oI a collective identity that has, as a core component, the principles oI democratic
legitimacy and rights. The principle oI democratic legitimacy implies that the conditions oI justiIicationthe procedures and presuppositions oI rational agreement
themselvesobtain legitimating Iorce and become the legitimating grounds (metanorms), replacing such material principles oI justiIication as nature or god.
57
The
principle oI democratic legitimacy involves a level oI justiIication that has become reIlexive and a procedural principle that is universalizable. This means that the
modern procedural principle oI democratic legitimacy presupposes a postconventional, posttraditional orientation to our own traditions, or at least to those aspects oI
our tradition and collective identity that have become problematic. Moreover, it implies that only those aspects oI our collective identity and common tradition that are
compatible with the principles oI democratic legitimacy and basic rights can provide the content oI valid political norms. The Iact that discussion and democratic
principles constitute a part oI our tradition militates against the authoritarian thrust oI the concept oI collective identity, because it means that we can accept as valid
inputs into sociopolitical norms only those dimensions oI our political culture that do not violate the metanorms oI discursive conIlict resolution.
Let us try to clariIy our argument Ior replacing "general interest" with "collective identity" as the substantive reIerent oI a procedurally deIined discourse ethics and then
return to this issue. We propose our interpretation as an alternative to three positions that are unacceptable Ior diIIerent reasons. First, there is Habermas's own
position that makes generalizable interests the centerpiece oI a new principle oI universalization. This necessarily makes an objective categoryone that is open to
analysis Irom the third-person point oI viewthe core oI discursive will Iormation, but it has unavoidable authoritarian consequences that Habermas himselI does not
want. Second, there is the opposite position that bypasses the issue oI generalizable interests by identiIying all consensus as

Page 370
merely empirical and by making empirical agreement itselI the goal oI discursive procedures. The results oI empirical consensus then become, by deIinition, justice in
the political sense. We believe that Habermas's objections to this position (instability and an extreme variability oI results that could lead to moral skepticism similar to
that oI legal positivism and, especially, legal realism) are correct. A third position (such as that oI Karl-Otto Apel) seeks to bypass the issue oI generalizable interests
by insisting on rational (rather than empirical) consensus as an end in itselI. Those who engage in argument, according to this interpretation, must seek to institutionalize
and spread the institutionalization oI rational discourse in order to avoid perIormative contradictions. But this approach tends to devalue all actual or empirical
discourse in the name oI an ever-vanishing counterIactual, and thereIore it cannot even begin to speciIy its own conditions oI institutionalization.
Our position involves two interrelated steps. First, we start with empirical norms, traditions, and consensuses that claim to be democratic, but we hold that they can be
evaluated (by participants) in terms oI their possible degree oI rationality and democratization, that is, in light oI the metanorms provided by discourse ethics. Second,
we remain, nonetheless, aware oI the instability oI the results oI even rationally debated empirical consensus, and we seek to remedy this by an argument based on
collective identity in the Iirst instance, and on general interests and social solidarity in the second. We Iocus on actual processes oI public discourse that can, iI
rationalized or democratized, constitute or reaIIirm a rational, democratic collective identity or political culture. In such contexts, discourse ethics provides the
standards with which to select those aspects oI our tradition, collective identity, and political culture that we wish to maintain and develop and that can provide the
content Ior legitimate norms. Processes oI public communication constitute the we oI collective action, certainly beIore it can be asked (Iormally speaking) what the
interests oI a society or group may be and beIore the conditions oI solidarity oI its members with one another can be explored.
OI course, no collective identity is simply or exclusively selI-reIlective, nor can any collective identity be universal in all oI its aspects. The universalizable principles oI
democratic legitimacy

Page 371
and basic rights can be only components, not the totality oI a common identity. Their reproduction on the symbolic level presupposes the discovery or reappropriation
oI the traditions, collective memories, preexisting patterns oI interaction, established values, and relevant practices (liIe world) that are the sources oI solidarities that
can sustain the rational core oI a political collective identity. From the point oI view oI a discursive, postconventional collective identity, this reappropriation must be
posttraditional, that is, critical in relation to tradition. It must select out traditions oI discourse and oI empirical solidarity that are not compatible with a postconventional
collective identity and must establish a highly critical relationship to them. The latter cannot serve as the content oI political norms. Thus, while every collective identity
is by deIinition particular, those that are able to have a critical relation to their own traditions can develop content that is not incompatible with the principles oI
discursive conIlict resolution. Whatever we have been, we as members oI modern civil societies now partake in a political culture predicated on the principle that we
ought to resolve conIlicts discursively. In other words, we have more than the mere temporally bound and limited discursive procedure itselI to validate our decisions
and to comprise our collective identity (which otherwise would be paper-thin), and we do not have to rely on objective interpreted interests Ior this "more." The Iree
public discourse that aIIirms our identity has itselI a tradition that gives this identity substance over time. Thus, the level oI generality sought by Habermas can, in the
Iirst instance, be derived Irom participation in discourse. But it can more solidly be based on discourses aimed at renewing the traditions oI discourse that underlie the
principle oI democratic legitimacy in modern civil societies.
In our deIinition, common identity is not equivalent to general interest. Yet once a common identity is established or reaIIirmed, it is then possible to arrive at an
understanding oI what constitutes the general interests oI a community. These would entail those institutions or arrangements that are needed to reproduce
"materially" (as distinct Irom normatively) the relevant collective identity oI the community. Here the social-scientiIic standpoint has a place. It is possible to argue, Ior
example, as Habermas does, that diIIer-

Page 372
entiating between system and liIe world and reproducing some kind oI modern economy and political state are in the general interest oI all those who participate in a
modern liIe world and have the corresponding postconventional moral and cultural collective identity that it presupposes. It is possible, in short, to speciIy the
necessary structural preconditions Ior reproducing a common identity whose principles have received discursive validation. We would still have to argue Ior
generalizing these interests by raising cognitive validity claims as to their truth. Moreover, we must be open to debate on whether speciIic institutional arrangements are
needed Ior our common identity or whether a variety oI institutional arrangements might serve this purpose, some oI them possibly better than the existing ones.
Indeed, there is an important distinction to be kept in mind between the institutional requirements (general interests) that are necessary Ior reproducing a
postconventional collective ide