Habermas, “Civil Society and the Public Sphere,” in Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy, pp. 329-387

Page 329

8 Civil Society and the Political Public Sphere

In the early postwar period, the sociological study of democracy led to a theory of pluralism that still linked normative models of democracy with so-called "realist" approaches, that is, with economic theory on the one hand and systems theory on the other. If one disregards for the moment the recent revival of institutionalist approaches,¹ then one can hardly avoid the impression that as democratic theory has developed, the idealistic content of normative theories has been evaporating under the sun of social science. It was, in any case, only the liberal model of democracy-—hence the normatively least demanding—that offered sociology a point of contact. The sociological enlightenment seems to recommend a disillusioning, if not downright cynical, view of the political process. It primarily focuses our attention on places where normatively "illegitimate" power forces its way into the constitutionally regulated circulation of power. If one takes the administrative systemor the "state apparatus"—as the point of reference, then the political public sphere and the parliamentary complex represent the input side, where the social power of organized interests enters the legislative process. On its output side, the administration again encounters the resistence of social subsystems and large organizations that bring their power into the implementation process. The autonomy of social power vis-à-vis the democratic process fosters in turn endogenous tendencies in the administrative complex to become increasingly autonomous. Thus an increasingly independent administrative powerjoins forces with a social power affecting both the input and the output side. Together they form a counter-

Page 330

circulation that cuts across the "official" circuit of democratic decision making steered by communicative power. To be sure, most of the descriptions of these countermovements operate with empiricist concepts of power that level out the normative distinctions we have introduced from a reconstructive point of view. In particular, the construct of "communicative power" must appear tendentious in light of empiricist folklore. Political sociology usually either takes an action-theoretic approach, conceiving "power" as the capacity of actors to have their way against the opposing will of others, or it follows systems theory and splits power up into the power code of a specific, namely political, system on the one hand and, on the other, a general organizational power or, more accurately, the autopoietic capacity of systems to organize and reproduce themselves. I would like to show that the normative defeatism to which both lines of political sociology lead is not simply a result of sobering evidence but of misguided conceptual strategies as well. These strategies lose sight of what political power owes specifically to its formal constitution in legal terms.

After a global overview of theoretical developments, I first pursue Jon Elster's revisions of the economic theory of democracy. These speak for the empirical relevance of the procedural concept of deliberative politics (sec. 8.1). I then discuss Helmut Willke's concept of supervision, which is meant to explain how a decentered society that has allegedly disintegrated into autopoietic functional systems can cope with challenges to society as a whole. From the criticism of this proposal, and stimulated once again by the work of Bernard Peters, I develop a sociological model that focuses on the empirical weight of the constitutionally prescribed, hence official, circulation of power (sec. 8.2). This weight depends primarily on whether civil society, through resonant and autonomous public spheres, develops impulses with enough vitality to bring conflicts from the periphery into the center of the political system (sec. 8.3).

8.1 Sociological Theories of Democracy

8.1.1

The early theory ofpluralism already relied on an empiricist concept of power. Pluralist theory employs an instrumentalist understand-

Page 331 ing of politics—according to which political and administrative power represent merely different manifestations of social powerto build a bridge between the liberal model of democracy introduced above and empirical science. Social power is measured in terms of the ability of organized interests to have their way. Moving up through party competition and general elections, social power is converted into a political power distributed between the incumbent Government and the opposition. This political power is in turn employed within the framework of constitutionally allocated powers and authorities so that, via the legislative process and administrative apparatus, the policies emerging from the channeled interplay of social forces can be converted into binding decisions and programs to be implemented. Moving back down in the opposite direction, administrative power is deployed to affect parliamentary will-formation and the interplay of organized interests. These interests, for their part, also get the chance to have a direct influence on policy formation and the use of administrative power. According to this model, a circular process is established that connects the clients' social power with the parties' acquisition of political power, the legitimation process with governmental services and administrative operations, and this implementation process again with the clients' initial claims. For the normative evaluation of these processes, the decisive assumption is that social power is more or less equally distributed among the relevant social interests. Only then can the balance of social forces and pressures keep political power circulating in such a way that the political system copes with the submitted claims as effectively as possible and satisfies social interests as equally as possible. The sociological account of pluralism managed to link up with the normative model of liberalism by means of a simple substitution, namely, by replacing the individual citizens and their individual interests with large organizations and organized interests.^2It assumed that all politically relevant collective actors enjoy roughly equal opportunites to influence the decision-making processes that concern them; that the members of the organizations determine the politics of pressure groups and parties; and that the latter in turn are pushed by multiple memberships into a readiness for compromise and the integration of interests. Under these conditions, pluralist democracies should project the social balance of

Page 332

power onto the distribution of political power in such a way that politics remains sensitive to a broad spectrum of values and interests.³

After these assumptions were falsified, pluralist theory was revised along lines already evident in the work ofJoseph Schumpeter. Because the composition of interest groups is in fact highly selective, because members are largely inactive and have little influence on organizational policy, it was now assumed that power struggles are essentially conducted by elites. A further assumption also proved untenable: that politicians and administrators depend on a variety of collective actors having approximately equal weight in the competition for political influence. What remained was a theory of political elites that essentially reduced the role of the democratic process to plebiscites between competing leadership teams, and thus to the recruitment of personnel and the selection of leaders.⁴ Naturally, from a normative standpoint, the theory still had to explain how "a politics basically initiated by elites [can] also satisfy the interests of nonelites."⁵ With this question, a remnant of normative expectations was shifted from the input side to the output side of the administrative system. Because the unspecific and highly aggregated trust of masses of passive voters no longer can determine the policies of the competing leadership groups, only the rationality of the elites themselves, capable of decision and ready for innovation, can guarantee that the administration functions more or less in the equal interest of all. This gives rise to the image of an administrative system that, operating relatively independently of society, procures the necessary mass loyalty and determines political goal functions more or less by itself. From a normative standpoint, this poses the problem of the conditions under which the state apparatus, if not actually steered by societal interests, at least develops sufficient sensitivity for these interests. A self-programming political system must assume, on its own, the task of articulating publicly relevant needs and wants, latent conflicts, repressed problems, nonorganizable interests, and so forth.

Since the late sixties, though, the evidence for a more cautious assessment has mounted. The administrative system apparently can operate self-consciously onlywithin extremely narrow limits; its operations seem to follow a more reactive style of politics dictated

Page 333

less by planning than by avoiding crisis. On the output side, the ''active state" quickly reaches the limits of its steering (or regulatory) capacity, because social systems and large organizations stubbornly resist direct interventions. On the input side, the room for initiative on the part of Government leaders and parties is also restricted by the unpredictability of independent voters—whether enlightened or mobilized by populist movements—whose party ties are becoming increasingly loose. As politics grows ever more bleak, established parties have to fear the withdrawal of legitimation through protest votes and nonvoting. Both legitimation deficits and steering deficits push politics into a kind ofincrementalism that can hardly be distinguished from quietism.

This has led to a fork in theoretical developments. Systems theory cuts the last remaining ties with normative models, essentially limits itself to the self-referential problems of an autopoietic political system, and once again takes up the organizational problems of the classical theory of the state, translating them into steering problems. By contrast, the economic theory of democracy presupposes a methodological individualism and focuses mainly on the process of legitimation. From the viewpoint of systems theory, the mode of operation of the political system is gauged by a rationality of selfreflexive steering that has lost all traces of the normative content of democracy (beyond an alternating allocation of power between the incumbent Government and the opposition). From the viewpoint of economic theory, this normative content is equivalent to the rational-choice behavior of those participating in the democratic process. Meanwhile, it is evident that both approaches, each in its own way, have pushed the normative weight reduction too far. Problems arising within the theories have led to revisions that are suggestive, though not always consistently carried through.

The economic theory of democracy attempted to capture empirically some of the normative intuitions of liberalism by demonstrating the rationality of the behavior of voters and politicians.⁶ According to this model, in casting their ballots, voters translate a more or less enlightened self-interest into claims on the political system, whereas politicians who want to gain or hold office exchange these votes against the offer of specific policies. The transactions between rationally choosing voters and political elites supposedly yield

Page 334

decisions that are rational insofar as they take into account the aggregation of individual interests of equal weight. More recently, the discussion over the "voter's paradox" has inaugurated a certain shift. On the premise of exclusively self-interested behavior, the fact that citizens bother to vote at all could only be explained by a hypothesis that soon proved false: the level of voter turnout does not vary with the voters' expectation that a single vote possibly could decide a head-to-head race. Consequently, proponents expanded the egocentric rational-choice model by bringing in the concept of a metapreference, so that the model could include ethical, albeit self-regarding, considerations.⁷ In the end, however, the empirical evidence spoke against all models that were premised on egocentric decision making (however expanded) and that disregarded how changes in interests and value orientations relate to social context.⁸ The most recent revisions consider, for example, the "laundering" effect of institutional arrangements that give priority to normative reasons. In this way, institutionalized procedures can foster "responsible" political action: "To act responsibly, then, is for the agent to evaluate his or her own actions by methodically taking the critical perspectives, simultaneously and in the futurum exactum, of the expert, the generalized other, and of himself or herself. By assuming this triple perspective, the actor validates the criteria of action substantively, socially, and temporally."⁹ With the perspective of Mead's "generalized other,'' Claus Offe approaches the concept of an opinion- and will-formation that, as I will show, explodes the conceptual framework of an empiricist theory of action.

Systems theory immediately abandons the notion of individual and collective agency. In the face of immense complexes of increasing organizational density, it resolutely concludes that society should be conceived as a network of autonomous subsystems, each of which is encapsulated in its own semantics and has all the other systems for its environment. Only the modes of operation internal to each system, and not the intentions or interests of participating actors, are decisive for interactions between such systems.¹⁰ This choice of conceptual strategy results, on the one hand, in the rejection of a hierarchical concept of society centered in the state. Even the political system, which is specialized for generating

Page 335

collectively binding decisions, must assert itself opportunistically, without possibilities of privileged access, against all the other functional systems (including the legal system). On the other hand, the state-centered understanding of politics that was already present in the liberal model emerges in full force. That is, systems theory ascribes the opinion- and will-formation dominated by party competition to a public of citizens and clients that, cut off from its lifeworld roots in civil society, political culture, and socialization, is incorporated in the political system. Here the administration does not just consist in the complex with the highest organizational density. It also sets in motion a circulation running counter to the "official" circuit of power: the administration predominantly programs itself by steering the legislative process through its bill proposals, by extracting mass loyalty from the citizenry through parties that have become arms of the state, and by making direct contact with its clients.11 With the increase in social complexity, the balance tips in favor of this informal circulation, so that the question of "how political responsibility under such conditions is even possible" becomes meaningless.¹² A systems theory that has banned everything normative from its basic concepts remains insensitive to the inhibiting normative constraints imposed on a constitutionally channeled circulation of power. Through its keen observations of how the democratic process is hollowed out under the pressure of functional imperatives, systems theory certainly makes a contribution to the theory of democracy. But it offers no framework for its own theory of democracy, because it divides politics and law into different, recursively closed systems and analyzes the political process essentially from the perspective of a self-programming administration.

The "realism" that systems theory gains with this selective approach comes at the cost of a disturbing problem. According to systems theory, all functional systems achieve their autonomy by developing their own codes and their own semantics, which no longer admit of mutual translation. They thereby forfeit the ability to communicate directly with one another and as a result can only "observe" each other. This autism especially affects the political system, which also self-referentially closes itself off from its environment. In the face of this autopoietic encapsulation, one can

Page 336

scarcely explain how the political system should be able to integrate society as a whole, even though it is specialized for regulatory activities that are meant not only to rectify disturbances in functional systems but also to achieve an "environment-friendly" coordination among systems drifting apart. It is not clear how one should reconcile the autonomy of the different functional systems and the political system's task of holding them together: "The heart of the problem lies in the improbability of successful communication among autonomous, self-referentially operating unities."¹³

The history of the "realistic" approaches leads, on the one hand, to an economic theory of democracy that would inform us about the instrumental features of democratic will-formation and, on the other hand, to a systems theory that would inform us about the impotence of this will-formation. Both approaches operate with concepts of power that are insensitive to the empirical relevance of the constitution of power under the rule of law, because they screen out the internal relation between law and political power. This deficit ultimately stands behind the questions that Elster and Willke so instructively pursue. Elster's revisions lead to an unexpected rehabilitation of the concept of deliberative politics.

8.1.2

Rational choice theory still revolves around the Hobbesian problem. It cannot explain how strategic actors are capable of stabilizing their social relations solely on the basis of rational decisions. We do not need to go into the details of the penetrating self-criticism here.¹⁴ What interests me is how Elster handles the difficulties this theory encounters when it is applied to political processes. First of all, in this context it is unrealistic to start with a model that assumes that opportunities and preferences can be treated as something given; both of these change in the political process itself. Moreover, the individual preferences expressed in opinion polls do not reliably reflect the actual preferences of the individuals polled, if by "preferences" one means the preferences they would express after weighing the relevant information and arguments. For the political transformation of values and attitudes is not a process of blind adaptation but the result of a constructive opinion- and will-

Page 337

formation. Elster designates this as the "autonomous" formation of preferences: "autonomy is for desires what judgement is for beliefs."¹⁵ Moreover, it is especially unrealistic to assume that all social behavior is strategic action and can thus be explained as though it were the result of egocentric utility calculations. The explanatory power of this sociological model is obviously limited: "While there is always a risk of self-serving behavior, the extent to which it is actually present varies widely. Much of the social choice and public choice literature, with its assumption of universally opportunistic behavior, simply seems out of touch with the real world, in which there is a great deal of honesty and sense of duty. If people always engaged in opportunistic behavior when they could get away with it, civilization as we know it would not exist."¹⁶

This and similar considerations led Elster years ago to broaden the rational-choice framework by including socioethical commitments and moral reasons among preferences. In addition, he described the democratic process as a mechanism that changes preferences through public discussions.¹⁷ What matters most to him are the procedural aspects of such a rational will-formation.¹⁸ To render this idea fruitful, he had to make two drastic revisions in the rational-choice model.

First, he expanded the theory by adding an additional action type. Besides strategic or purposive-rational action, which is oriented to consequences (under conditions of incomplete information) and steered by the individual's own preferences, there is also norm-regulated action. The latter is an elementary action type, because it is not reducible to strategic action.¹⁹ In reply to the argument that norms serve only tojustify opportunistic action after the fact, Elster maintains that no one could deal strategically with norms in a particular case if he could not count on the intersubjective recognition of norms in general. To this extent, the social validity of norms enjoys a logical priority over the advantages one might gain by pretending to orient oneself to such norms. Equally unconvincing is the further objection that norm-conformative behavior is actually a purposive-rational avoidance of internalized sanctions (such as feelings of shame and guilt). Dealing rationally with the consequences of a behavior that by presupposition is irrational does not explain how this behavior (as the result of the

Page 338

antecedent internalization of a behavioral norm) comes about in the first place: one cannot rationally resolve to act irrationally.

However, Elster still clings to empiricist premises when he introduces the new action type. As he describes it, norm-regulated action differs from strategic action only by the lack of an orientation to expected consequences. Purposive rationality is what homo oeconomicus has over homo sociologicus.²⁰ Norms and value orientations seem to elude rational assessment; they merely ground counterfactually maintained behavioral expectations that are immune to learning. As a result, Elster splits the moral or ethical sphere into two domains, one cognitive and rational, the other emotive and irrational. He either strips moral norms of their obligatory character, in line with utilitarianism, and classifies them as decision rules for purposive-rational action; or he admits them as binding norms, in the sense of a rigid ethics of conviction, but strips them of any rational character.

As long as normativity and rationality exclude each other like this, the rationally motivated coordination of action can only take the form of a negotiated agreement among strategically oriented actors. The process of rational agreement becomes equivalent to "bargaining"—the negotiation of compromises. Indeed, such bargaining, which requires a willingness to cooperate on the part of strategic actors, is connected with norms that take the form of empirical constraints or irrational self-bindings. To this end, Elster develops a parallelogram of forces that explains normatively regulated bargaining processes as a combination of rational calculations of success with social norms that contingently steer from behind ²¹.

However, this empiricist way of introducing value orientations proves deficient if it is supposed to explain how participants can rationally change their preferences in the course of political willformation. Nor does it explain how new options can open up in a "convincing" manner. The political process involves more than compromises based on credible threats. This is why Elster introduces "argumentation" as a further mechanism besides "bargaining" for solving problems of collective action: ''Rational argumentation on the one hand, threats and promises on the other, are the main vehicles by which the parties seek to reach agreement. The former is subject to criteria of validity, the latter to

Page 339

criteria of credibility."²² With "criteria of validity," a new kind of communication and action coordination comes into play. Whereas parties can agree to a negotiated compromise for different reasons, the consensus brought about through argument must rest on identical reasons able to convince the parties in the same way. The consensus-generating force of such reasons is displayed in the idea of impartiality that governs practical discourses.²³

This step necessitates a revision of the first revision. With the idea of the impartial assessment of interest positions and action conflicts, those norms previously considered irrational are partly drawn into the vortex of argumentation. If Elster wants an additional mechanism of action coordination—not only the reciprocal influence exerted by actors oriented to success but also the communication among persons engaged in argument for the purposes of reaching understanding—then he must acknowledge a rational core to norms and value orientations and correspondingly enlarge his concept of rationality. This he can do with the help of the deontological concept ofjustice or practical reason, in the light of which rights can be justified.²⁴ It follows that the task of politics is not merely to eliminate inefficient and uneconomical regulations but also to establish and guarantee living conditions in the equal interest of all citizens.²⁵

With these basic conceptual revisions in place, Elster undertakes an empirical analysis of the discussions conducted in the constitutional assemblies of Philadelphia (1776) and Paris (1789-1791). His analysis starts from the theoretically motivated distinction between "bargaining" and "arguing," where argumentation includes, according to our terminology, not onlyjustice arguments but also ethical-political arguments referring to the "general welfare" of the nation. Elster goes on, by way of a comparative analysis of the first two modern constitutional processes, to examine the assumption that a parliamentary opinion- and will-formation of this kind cannot be adequately explained on the empiricist premises of a balance of interests steered exclusively by power. Rather, discourses and bargaining processes intertwine in such processes, though compromise formation often takes place spontaneously and to this extent does not satisfy the fairness conditions of regulated bargaining.²⁶ Elster's Storr Lectures permit two interpretations,

Page 340

depending on whether one refers to their obvious message or to the analysis of the role played by certain types of arguments selected from the debates. Read from the first perspective, Elster reconstructs a piece of legal history. He reaches the conclusion that "the will of the constitutional lawgiver" was to enact a system of rights that under the perceived circumstances was intended to guarantee the citizens' political autonomy by institutionalizing an impartial opinion- and will-formation. In this respect, Elster's contribution (at least implicitly) tests the discourse-theoretic reading of the constitutional state against its historical background.

His explicit goal, however, is to rationally reconstruct patterns of argument in order to show that the decisions of the political lawgiver were, to a certain extent, rationally motivated, namely, by rational discourse and mutual understanding in combination with a success-oriented exercise of influence.²⁷ Elster investigates primarily the interaction of these two mechanisms. In the process it turns out, not surprisingly, that the actual course of the debates deviates from the ideal procedure of deliberative politics. At the same time, however, presuppositions of rational discourse have a steering effect on the course of debate. The communicative presuppositions for a deliberative contest of opinions are institutionalized in such parliamentary bodies at least effectively enough that the democratic procedure filters arguments and gives legitimacy-producing reasons a privileged chance to come into play.

For example, not all interests can be publicly advocated. Hence the publicity of political communications (emphasized by Kant), in connection with the expectation that proponents are consistent in their utterances and explain their proposals coherently, already exerts a salutary procedural force. Under this condition, concealing publicly indefensible interests behind pretended moral or ethical reasons necessitates self-bindings that either on the next occasion expose a proponent as inconsistent or, in the interest of maintaining his credibility, lead to the inclusion of others' interests^.28

These and similar considerations suggest that one should seek the conditions for a rational political will-formation not only at the individual level of the orientations and decisions of single actors but also at the social level of institutionalized processes of delibera-

Page 341

tion and decision making. These can be seen as arrangements that have an effect on the preferences of participants; they screen the topics and contributions, information and reasons in such a way that, ideally, only the "valid" inputs pass through the filter of fair bargaining and rational discourses. With this, the perspective shifts from rational-choice theory to discourse theory: "These institutions [i.e., of the United States Constitution] were designed to play the role of 'congealed' or 'sedimented' virtue, which thus made the actualpracticeof these virtues, such as truthfulness, wisdom, reason, justice and all kinds of exceptional moral qualities, to some extent dispensable—on the part of both the rulers and the ruled."²⁹ To the degree that practical reason is implanted in the very forms of communication and institutionalized procedures, it need not be embodied exclusively or even predominantly in the heads of collective or individual actors. Elster's investigation provides some support for the assumption that the discursive level of observable political communication is a standard for the effectiveness of a "proceduralized reason" of this sort. The results of deliberative politics can be understood as communicatively generated power that competes, on the one hand, with the social power of actors with credible threats and, on the other hand, with the administrative power of officeholders.

8.2 A Model of the Circulation of Political Power

Systems theory does not have to deny the phenomenon of a communicative power generated within the parliamentary complex—nor that of an influence acquired in the political public sphere—but it describes the phenomena in a way that a fortiori exposes communicative power as impotent. From this viewpoint, once the law is fully positivized, the political system should be able to forego independent sources of legitimate law. As with all functional systems, politics, too, has become an autonomous, recursively closed circuit of communication furnished with its own code. In connection with the legal system responsible for securing legality, such a contingent, self-referential politics draws everything it needs for legitimacy from itself. The need for legitimation can be met paternalistically through lines of connection that, running

Page 342

from the ritualized confrontation between the incumbent Government and the opposition, reach via party competition into the loose network of the voting public. However, Luhmann's depiction of the self-legitimation of a politics anchored in the state apparatus begins to fall apart if systems theory is confronted with the task of "conceiving the theory of the state from the perspective of an ethically responsible and responsive society."³⁰ In section 8.2.1, I present an immanent critique of Willke's innovative attempt to develop such a theory of the state; in section 8.2.2, I then develop from this a model suitable for a sociologically informed use of the concept of deliberative politics.

8.2.1

In his Philosophy of Right (secs. 250-56), Hegel assigned the corporations the important task of mediating between civil society and the powers of the state. In light of the current neocorporatism debate, Willke is led to a view that reads like a modern systemstheoretic version of the Hegelian Ständestaat—only without its monarchical head. Concerted actions, round-table arrangements, and the various kinds of coordinating committees that arise in the gray areas between state and society are described by Willke as symptomatic bargaining systems. These systems should allow politics in a decentered society to assume the role of a therapeutically trained supervisor that preserves the encompassing social unity that the state itself can no longer represent. On the one hand, Willke, like Luhmann, sees that the political system has become one subsystem among others. No longer able to claim social primacy, it is relieved of the function of integrating society as a whole. On the other hand, through the back door, Willke reintroduces the state as guarantor of a neocorporatist social integration.

As surprising as this answer might be, the question is the logical result of the "autopoietic turn" taken by systems theory. Indeed, the logic of the functional differentiation of society implies that the separate subsystems are reintegrated at a higher level of society. If the decentered society could no longer preserve its unity somewhere, it would not profit from the growth in the complexity of its parts and would be, as a whole, the victim of their gains in

Page 343

differentiation. The society converted to "autopoiesis" actually seems to lead to this dead end, for the functional systems take the last step to autonomy via their own specialized semantics that, with all their advantages, have the consequence of breaking off a direct exchange of information with their corresponding environments. Each functional system forms its own picture of society. Subsystems no longer command a shared language in which the unity of society could be represented for all of them in the same way. Non-codespecific mutual understanding is passé. As a result, each system becomes insensitive to the costs it generates for other systems. There is no longer any place where problems relevant for the reproduction of society as a whole could be perceived and dealt with. The special languages like money or administrative power wear down ordinary language—as the functional systems do the lifeworld—so much that neither the one nor the other presents a sounding board that would be sufficiently complex for thematizing and treating society-wide problems. Under these conditions, the political public sphere cannot provide such a sounding board, because, together with the public of citizens, it is hitched to the power code and placated with symbolic politics.

On the other hand, the trends toward social disintegration pose a challenge especially for politics and law. In a sense, these trends revive a shock that both have experienced once before in the rise of religious pluralism and loss of an overarching religious legitimacy. The question of how the unity of society can be secured now that the state can no longer represent it is, to be sure, no longer posed today directly as a question of legitimation. The standard of "legitimacy" pertains only to legal and political questions; it cannot be applied to problems affecting society as a whole. Nevertheless, the routine business of procuring legitimation is burdened by problems of macrosocial irrationality, because law and politics have assumed a kind of surety for the cohesion of the entire system. In any case, Willke diagnoses the return of a legitimation problem that is at least induced through the inadequate integration of the whole of society, even if this problem is measured not in the specific systemic terms of "legitimacy" but by a "total system rationality." This kind of societal rationality should now be managed precisely through politically mediated processes of attunement (Abstimmungs-

Page 344

prozesse) among different functional systems. Willke's neocorporatist vision "aims at shaping intersystem relationships among autonomous, active, and interdependent subsystems that no longer submit to the primacy of one part [namely, politics]. Hence they do not derive total system rationality from the validity of the universal but from the reflexive attunement of the particular."³¹ According to his diagnosis, the political systems in Western societies are already on the road to a "supervisory" state.³² The following three points summarize the description of a society that would be both integrated by, and under the guardianship of, an intersystemic balance of this sort:

(a) The supervisory state looks to nonhierarchical bargaining systems for an attunement among sociofunctional systems that either suffer disturbances in their operations and performance and need "developmental aid," or burden their environment with externalities and must be prodded to "show consideration." However, the manner in which self-referential systems can be induced to structural transformations is already fixed by these systems' own structures. For this reason, the supervisory state must pursue an "options policy" that goes beyond incentives. This kind of policy is a familiar idea in economic planning. To steer a given system, an options approach first considers that system's mode of operation and degrees of freedom and then exerts influence on the system's self-steering through suitable changes in the context. The extensively researched difficulties in such policy fields as health care, technology, and science provide plausible evidence for the limits of direct state interventions; instead of this, the approach taken in business consultation is considered exemplary as a strategy that implicates the various operationally closed systems in a "productive and self-binding interrelationship."³³

(b) Even the politics that "counsels" systems must avail itself of the language of law, though no longer in the form of conditional or goal programs but as "reflexive" law.³⁴ Politics provides the manipulated systems with new "forms" for their own setting of priorities, so that the set of preferences internal to the system receives a different weight. Each system should play its own melody but with a new rhythm. To this end, an individualistic civil law must be transposed to the level of collective actors and converted from

Page 345

personal references to system relations. Examples are found in the legal protection pertaining to the new collective goods of the "risk society": protection from environmental destruction; protection from radiation poisoning or lethal genetic damage; and, in general, protection from the uncontrolled side effects of large technological operations, pharmaceutical products, scientific experimentation, and so forth. Law must not implement goals in the form of targets; rather, through "relational programs," it should induce and enable the system that is causing the danger to steer itself in a new direction. In this way, the law functions as a catalyst for self-monitored modifications.

(c) Though social integration would, according to this view, shift from the level of democratic opinion- and will-formation to the level of intersystemic monitoring, the "essential content" of a constitutional democracy transposed from persons to systems is supposed to remain intact. Willke speaks of the "establishment of societal discourses" and even of the "attunement of autonomous actors through rational discourses."³⁵ For the attunement processes follow the guidelines set by (democratic?) procedures that regulate the communication among the decentered units: "Consensus is required as a framework consensus regarding the foundations and bounds of dissent, so that dissent, taken continually further, does not lead to the system's dissolution." At this point, Willke adds ''that consensus is needed only as the continually reconstituted imaginary line, which is no sooner fixed than differences and dissent can fasten on it to dissolve it"; thus, even the idealizing moment in the discourse concept of truth and validity is captured in the language of systems theory.³⁶ Of course, this formulation, which Willke's simulation of democratic consensus borrows from the intersubjectivistic vocabulary of a different theoretical tradition, can have only a metaphorical meaning under the changed premises.

On the one hand, unlike the communicative practice of citizens, the "conversation" of functional systems no longer deals with norms, values, and interests. Rather, it is restricted to the cognitive goal of enhancing systemic self-reflection. Exchanges among experts who instruct one another about the operation of their respective functional sectors is supposed to overcome the specific

Page 346

blindness of self-enclosed systems. As the examples drawn from management literature show, this exchange resembles more a continuing-education program in which the training leader or moderator prompts the managers coming from various branches to act as consultants for one another's business problems. On the other hand, the "conference rules" governing the conversation forfeit the universalistic content of democratic procedure; the set of functional systems "in need of therapy" at a given point cannot claim to be representative.

Willke's interesting proposal for solving theoretically acute problems of macrosocial integration runs into difficulties that deserve comment, because they reveal the true proportions of the legitimation problem suppressed by systems theory.

In reply to (a) above: In its classical form, the Hobbesian problem was posed as the question of how the confluence of the egocentric perspectives of self-interested actors could yield an order that encouraged individuals to consider the interests of others. This problem, which rational choice theory still chips away at, reappears for systems theory in a different form. A self-stabilizing order must now be explained by the cognitive attunement of system perspectives. The practical dimension of action coordination has disappeared in the approach taken by systems theory, so that a reasonable "altruism" for the intercourse of functional systems becomes meaningless. In the purely epistemic version, however, the Hobbesian problem is posed still more sharply, because the egocentrism of clashing perspectives is no longer determined simply by each person's own preferences and value orientations but by each system's own grammarfor interpreting the world. Unlike individuals in the state of nature, autopoietic systems no longer share a common world. To this extent, the problem of successful communication among independent and self-referentially operating units, each with its own perspective on the world, corresponds almost exactly to the familiar phenomenological problem of constructing an intersubjectively shared world from the egological achievements of transcendental monads. No more than Husserl (or later, Sartre) solved this problem of intersubjectivity,³⁷ has systems theory managed to explain how autopoietically closed systems could, inside the circuit of self-referential steering, be induced to go beyond pure self-reference and autopoiesis.³⁸

Page 347

The reflexive spiral of the reciprocal observation of the other's self-observations does not escape the circle in which both external observation and self-observation are always a system's own observation; it does not penetrate the darkness of mutual opacity.³⁹ To "understand" the mode of operation and the self-reference of another system and notjust "observe" it, to be able notjust to form a ''picture" of this according to their own respective codes, the systems involved would have to have available an at least partially shared language. But this is excluded ex hypothesi:

	Successful communication presupposes that the parts present one another with reciprocally relevant information in such a way that it can be "read," i.e., can be understood even in the context of basic criteria that are different and foreign. The challenge is to establish compatibility between different "language games," whereby different realities and projections of the world are linked with the "language." It is thus true even of complex societies that the deep structure of their order is bound to the grammar of the transfer of comprehensible information.^40

The "transference rules" making up such a grammar, however, cannot be given simply by the grammatical rules of an ordinary language circulating throughout society. Rather, following the model of international private law, they must first be constructed as collision norms that, from the perspective of each system, set up bridges of mutual understanding with other systems. But to the extent that the involved systems generate such rules for themselves, they have not yet overcome their semantic perspectivism. At best they have created the basis for a new stage of development. The required intermeshing of perspectives must, therefore, wait for a new system of rules to emerge.

In the end, Willke has to conjure the conditions for the intersubjectivity of possible mutual understanding out of the hat of social evolution: "A new kind of rule comes into play here. For the first time, these rules are no longer anchored in subsystems but arise at the level of [whole] systems from the active and intentional interplay of the parts, which want [!] to combine into an emergent total system. This type of rule is the material from which decentralized context-steering can develop as procedures of political supervision."⁴¹ From the reciprocal observation and mutual groping of semantically closed systems, therefore, a language is supposed to

Page 348

emerge that simulates exactly what ordinary language—namely, the natural language from which the special semantics were originally differentiated—always already achieves. Willke's auxiliary construct of an emergent language is even less convincing in that the law, with its required "relational programs," must already link up with the ordinary linguistic achievements by which "comprehensible information" is transferred throughout society. If one follows the discussions in the analytical philosophy of language, it is at any rate not surprising that ordinary language functions as the "ultimate metalanguage." It forms the open medium of a language that circulates throughout society and can be translated into and from every specialized code.

In reply to (b): The supervisory state is supposed to ward off the "danger of absolutizing subsystem rationality at the expense of the rationality of the whole."⁴² This goal exceeds the reach of any one system, including the political system; it makes internal political legitimation depend on a successful context steering that is no longer under political control. This steering must instead proceed through corporate bargaining, that is, through institutions of intersystemic attunement and coordination. Even if politics should manage, without cognitive dissonance, to extend its own codespecific criteria of legitimacy by adding the standards of a comprehensive societal rationality that govern "supervision," this would shift the basis of decision making in such a way that the political system would no longer be able to attribute all the decisions in need of legitimation solely toitself But this implies an interruption of the internal process of self-legitimation. Intersystemic supervision is still effected in the forms of law. But when lawmaking powers are delegated to bargaining networks infected by the dynamics specific to different social systems operating self-referentially beyond politics, the reproduction of law and politics falls under the shadow of a "dual power" ambiguously divided between state administration and social units. The more the public administration gets involved in "societal discourses'' of the new kind, the less it can satisfy the constitutional imperatives of the official circulation of power. The same neocorporatism that is supposed to cope with the dangers of macrosocial disintegration and keep the new kind of legitimation problem in check disturbs the process of self-legitimation inside

Page 349

the political system. This objection would be off target only if the envisioned change in the type of legal regulation could be kept within the boundaries of democracy.

The more collective actors, sociofunctional systems, and large organizations act in place of individuals, the more clearly the basis for assigning responsibility for the consequences of actions shiftsand the less it seems possible to rely on rights for safeguarding those collective goods of the "risk society" that are worth protecting. For this reason, Willke considers it obsolete to anchor law in an individualistically designed system of rights. The "relational programs" he calls for are instead designed for the autopoietic reproduction and self-steering of systems; they no longer refer to the private and public autonomy of socialized individuals. Nevertheless, such a restructuring of law is by no means intended to abandon the idea of the constitutional state but only to interpret it differently. On this premise, a formally correct legal regulation of the bargaining networks would suffice to secure their legitimacy as well. Willke even insists that

	democracy can be preserved in highly complex societies only if this idea [of the constitutional state] is generalized for society as a whole. Furthermore, society must be specifically constituted in a way that permits and promotes the guarantee of the autonomy and differences of citizens as much as the autonomy and differences of its functional systems. The latter is not just an end in itself for the preservation of the achieved degree of functional differentiation but in addition serves to generalize the constitutional protection of the citizen.^43

In fact, this formulation reveals more than a break with just one historical—indeed, historically obsolete—reading of constitutional democracy. With the "idea of a society-wide, consensually institutionalized and binding constitution" that stretches "from its citizens as natural persons to include their organizations, corporate actors, and functional systems,"⁴⁴ the systems-theoretic adaptation of the Hegelian Ständestaat takes the place of the democratic constitutional state and undermines its individualistic basis of legitimation. This is already evident in simple examples. Neocorporatist bargaining networks can attune the growing complexity of each functional system with the others only by simultaneously stimulating this growth; yet there is no preestablished

Page 350

harmony between, on the one hand, this growth in the complexity of systems that are "entitled to take part in the arrangement" and, on the other, the realization of citizens' constitutional rights. Certainly one can often justify a high level of systemic differentiation even from a normative standpoint. And as long as rising complexity in public administration and the capitalist economy was paralleled by the increasing inclusion of citizens, one could assume, all in all, that gains in functional differentiation coincided with normative progress in the realization of equal rights. However, these parallels were a matter of contingent, and hitherto in no way linear, correlations. Moreover, there are many indications of counterdevelopments. To namejust one of these, in the fragmented societies of today's economically interdependent OECD-world, the prosperity and security enjoyed by a majority of the population is increasingly accompanied by the segmentation of a neglected and powerless underclass that is disadvantaged in practically every respect. Conflicts arise between the neocorporatistically negotiated policies and the constitutional protection of underorganized parts of the population at the periphery of society. These conflicts arise not only as the result of an unequal distribution of social rewards but also because the loss of collective goods affects different social classes selectively.

Hence the idea of government by law is vitiated if sociofunctional systems are constitutionally released from their instrumental role and promoted to "ends in themselves." The "autonomy and differences" of citizens must then compete with that of systems for legal protection, even within the "official" circulation of power. The constitutional structure of the political system is preserved only if government officials hold out against corporate bargaining partners and maintain the asymmetrical position that results from their obligation to represent the whole of an absent citizenry, whose will is embodied in the wording of statutes. Even in attunement processes, the bonds of delegation must not tear away from actual decision making. Only in this way is the connection with the public of citizens preserved, the public that is both entitled and in the position to perceive, identify, and publicly thematize the social intolerability of deficient or disturbed functional systems. Indeed, only in the corporatist arrangements do these systems supposedly

Page 351

first learn to overcome their specific blindnesses and observe themselves as subsystems in a larger system. Hence they depend on affected clients as citizens to instruct them about their external costs and the negative effects of their internal failures. If the discourse of experts is not coupled with democratic opinion- and will-formation, then the experts' perception of problems will prevail at the citizens' expense. However, every difference of interpretation of this sort must, from the standpoint of the public of citizens, be seen as further confirmation of a systems paternalism that endangers legitimacy.

In reply to (c): This narrowly cognitivist, managerial account of neocorporatist steering discourses is explained by the fact that the attunement among systems gives rise exclusively to problems of functional coordination. Here the knowledge relevant for steering, provided by various groups of experts, is supposed to be worked into policies and translated into corresponding legal programs by legal scholars enlightened by systems theory. This conception is based on the unrealistic assumption that one can separate the professional knowledge of specialists from values and moral points of view. As soon as specialized knowledge is brought to politically relevant problems, its unavoidably normative character becomes apparent, setting off controversies that polarize the experts themselves. This by itself shows that problems of functional coordination, when handled politically, are intertwined with the moral and ethical dimensions of social integration. It is against the life-historical background of violated interests and threatened identities that the effects of deficient system integration are first experienced as pressing problems. Therefore, it is counterproductive, not only from the viewpoint of legitimacy but also from a cognitive viewpoint, for attunement processes between governmental and societal actors to become independent vis-à-vis the political public sphere and parliamentary will-formation. From both viewpoints, it is advisable that the enlarged knowledge base of a planning and supervising administration be shaped by deliberative politics, that is, shaped by the publicly organized contest of opinions between experts and counterexperts and monitored by public opinion.

Page 352

8.2.2

The concept of a society of autonomous subsystems constituted as "estates" faces the objections analyzed in (a) through (c) above. These objections all point in the same direction: the integration of a highly complex society cannot be carried out in a systemspaternalistic fashion, that is, in a manner that bypasses the communicative power of the public of citizens. Semantically closed systems cannot be induced to invent on their own the common language necessary for the perception and articulation of the relevant issues and standards of evaluation that apply to society as a whole. For such tasks, an ordinary language is available, circulating throughout society and lying beneath the threshold of the special codes. In the peripheral networks of the political public sphere and in the parliamentary complex, this ordinary language is, in any case, already in demand for dealing with macrosocial problems. This fact alone makes it impossible to conceive politics and law as autopoietically closed systems. The constitutionally structured political system is internally differentiated into spheres of administrative and communicative power and remains open to the lifeworld. For institutionalized opinion- and will-formation depends on supplies coming from the informal contexts of communication found in the public sphere, in civil society, and in spheres of private life. In other words, the political action system is embedded in lifeworld contexts.

Paragovernmental bargaining arrangements lacking effective ties to the parliamentary complex and the public sphere give rise to legitimation problems. Because they are specialized for questions of functional coordination, such corporatist networks are also no match cognitively for the pressure of accumulating problems. Moreover, the concentration on steering problems distorts the proportion between traditional and new administrative tasks. The tasks of social integration (in its specific sense of maintaining order, income redistribution and social welfare, the protection of collective identities and the transmission of a shared political culture) still have a place second to none on the political agenda. A fixation on highly organized social complexes gives rise to a onesided picture. Functionally differentiated societies certainly in-

Page 353

volve more than a multiplicity of self-referentially closed subsystems. The systems paradigm corresponds most closely to the capitalist economy and to the public administration specialized in planning and welfare. However, the inner logic with which many highly organized spheres, such as the educational system or science, oppose direct interventions is due not at all to a specific code or a steering medium analogous to money but to the logic of the special questions addressed in these spheres. Besides, the "constitutionalization" of action systems, which is the purpose of the "context-steering" strategies of a supervisory state, means something different in communicatively integrated spheres, such as the family or school, than it does in systemically integrated large organizations or networks (such as markets). In the former case, the legal constitution is only superimposed on an existing normative infrastructure; in the latter, it aids in the functional coordination of social relationships that are legally created. Finally, participatory forms of administration that bring implementing agencies into discourses with their clients (taken seriously as citizens) have a different sense than do neocorporatist bargaining arrangements. These differences must not disappear in the gray-on-gray descriptions of systems theory.

If we want to answer the questions posed at the end of the last chapter, then we must find a route other than those of systems theory and rational-choice theory. Elster's reconstructive analysis of the constitution-making process directs our attention to the procedural rationality of democratic opinion- and will-formation; however, whatwe see there does not extend beyond the production of communicative power. From a perspective broadened by systems theory, Willke concentrates on an administration overburdened with steering problems, problems that on his analysis can be solved only by circumventing communicative power. But this diagnosis underestimates what a multifunctional ordinary language can achieve precisely in virtue of its lack of specialization. Ordinary language is the medium of communicative action through which the lifeworld reproduces itself; in it, too, the components of the lifeworld interpenetrate one another. The action systems specialized for cultural reproduction (education) or socialization (family) or social integration (such as law) are not totally separated in

Page 354

their operation. Through the shared code of ordinary language, each of these systems also concomitantly satisfies the functions of the other two and thus maintains a relation to the totality of the lifeworld. The core private spheres of the lifeworld, which are characterized by intimacy and hence by protection from publicity, structure encounters between relatives, friends, acquaintances, and so on, and link together the members' life histories at the level of face-to-face interactions. The public sphere has a complementary relation to this private sphere, from which the public, as the bearers of the public sphere, is recruited.

The lifeworld forms, as a whole, a network composed of communicative actions. Under the aspect of action coordination, its society component consists of the totality of legitimately ordered interpersonal relationships. It also encompasses collectivities, associations, and organizations specialized for specific functions. Some of these functionally specialized action systems become independent vis-àvis socially integrated spheres of action, that is, spheres integrated through values, norms, and mutual understanding. Such systems develop their own codes, as the economy does with money and the administration does with power. Through the legal institutionalization of steering media, however, these systems remain anchored in the society component of the lifeworld. The language of law brings ordinary communication from the public and private spheres and puts it into a form in which these messages can also be received by the special codes of autopoietic systems—and vice versa. Without this transformer, ordinary language could not circulate throughout society.

In the following discussion, I make use of Bernard Peters's model to give a more precise form to, and seek a tentative answer to, the question of whether and how a constitutionally regulated circulation of power might be established.⁴⁵ According to Peters's proposal, processes of communication and decision making in constitutional systems display the following features: they lie along a center-periphery axis, they are structured by a system of "sluices," and they involve two modes of problem solving. The core area of the political system is formed by the familiar institutional complexes of administration (including the incumbent Government), judicial system, and democratic opinion- and will-formation (which

Page 355

includes parliamentary bodies, political elections, and party competition). Hence this center, distinguished from the periphery in virtue of formal decision-making powers and actual prerogatives, is internally organized as a "polyarchy." Within the core area, to be sure, the "capacity to act" varies with the "density" of organizational complexity. The parliamentary complex is the most open for perceiving and thematizing social problems, but it pays for this sensitivity with a lesser capacity to deal with problems in comparison to the administrative complex. At the edges of the administration, a kind of inner periphery develops out of various institutions equipped with rights of self-governance or with other kinds of oversight and lawmaking functions delegated by the state (universities, public insurance systems, professional agencies and associations, charitable organizations, foundations, etc.). The core area as a whole has an outer periphery that, roughly speaking, branches into "customers" and ''suppliers."

On the side of implementation, for different policy fields, complex networks have arisen among public agencies and private organizations, business associations, labor unions, interest groups, and so on; these networks fulfill certain coordination functions in more or less opaque social sectors. One should distinguish this type of clientele bargaining from the "supplier" groups, associations, and organizations that, before parliaments and through the courts, give voice to social problems, make broad demands, articulate public interests or needs, and thus attempt to influence the political process more from normative points of view than from the standpoint of particular interests. The spectrum extends from organizations representing clearly defined group interests; through associations (with goals recognizably defined by party politics) and cultural establishments (such as academies, writers' associations, and "radical professionals"); up to "public-interest groups" (with public concerns, such as the protection of the environment, the testing of products, and the protection of animals) and churches or charitable organizations.⁴⁶ These opinion-forming associations, which specialize in issues and contributions and are generally designed to generate public influence, belong to the civil-social infrastructure of a public sphere dominated by the mass media. With its informal, highly differentiated and cross-linked channels

Page 356

of communication, this public sphere forms the real periphery. Naturally, as the debate over corporatist bargaining shows, the distinction between output-oriented "customers" and input-oriented "suppliers" is not a sharp one. But the actually observable fusion between influencing the implementation of policies that have already been adopted, on the one hand, and influencing the formulation and adoption of policies, on the other hand, is not in agreement with constitutional principles.⁴⁷

After a descriptive overview, Peters introduces two explanatory elements: the "sluice model" and the two modes of problem solving that determine the direction in which communication flows. If binding decisions are to be carried out with authority, they must pass through the narrow channels of the core area:

	However, the legitimacy of decisions depends on processes of opinionand will-formation at the periphery. The center is a system of sluices through which many processes in the sphere of the political-legal system must pass, but the center controls the direction and the dynamics of these processes only to a limited degree. Changes can startjust as much at the periphery as at the center. . . .. The idea of democracy is ultimately based on the fact that political processes of will-formation, which in the schema here sketched have a peripheral or intermediary status, are supposed to be decisive for political development. This is not predecided by the present schema.^48

This sociological translation of the discourse theory of democracy implies that binding decisions, to be legitimate, must be steered by communication flows that start at the periphery and pass through the sluices of democratic and constitutional procedures situated at the entrance to the parliamentary complex or the courts (and, if necessary, at the exit of the implementing administration as well). That is the only way to exclude the possibility that the power of the administrative complex, on the one side, or the social power of intermediate structures affecting the core area, on the other side, become independent vis-à-vis a communicative power that develops in the parliamentary complex.

To be sure, the normal business of politics, at least as it is routinely conducted in Western democracies, cannot satisfy such strong conditions. The countercirculation cutting across the "official" circulation of power certainly involves more than just a contradic-

Page 357

tory, disdainful social facticity. Many of these communications flowing in the opposite direction serve to relieve the burden of unavoidable complexity from the official circulation by breaking problems down into smaller components. Peters takes this circumstance into account with the help of a second explanatory element. For the most part, operations in the core area of the political system proceed according to routines. Courts deliverjudgments, bureaucracies prepare laws and process applications, parliaments pass laws and budgets, party headquarters conduct election campaigns, clients exert influence on "their" administrations—and all these processes follow established patterns. From a normative standpoint, the only decisive question concerns which power constellations these patterns reflect and how the latter can be changed. This in turn depends on whether the settled routines remain open to renovative impulses from the periphery. In cases of conflict, that is, processing matters according to the usual conventions is eclipsed by another mode of operation.

This latter mode of operation is characterized by a consciousness of crisis, a heightened public attention, an intensified search for solutions, in short, by problematization. In cases in which perceptions of problems and problem situations have taken a conflictual turn, the attention span of the citizenry enlarges, indeed in such a way that controversies in the broader public sphere primarily ignite around the normative aspects of the problems most at issue. The pressure of public opinion then necessitates an extraordinary mode of problem solving, which favors the constitutional channels for the circulation of power and thus actuates sensibilities for the constitutional allocation of political responsibilities. True, even during the "normal" course of business, parliaments and courts attempt normatively to limit the discretion enjoyed by an administration operating largely in a goal-oriented manner. But in cases of conflict, the constitutional scheme for regulating access to normative reasons acquires a sharper profile. Only in this way can parliaments and courts—the two branches of government that alone are formally empowered to deal with normative reasons in a constructive or reconstructive manner—also actually determine the direction in which communication circulates. When conflicts become this intense, the political lawgiver has the last word. To be

Page 358

sure, considerable evidence shows that by itself the parliamentary complex usually lacks the strength "to turn cases into cases of conflict." Institutions that decide under time pressure have a weak capacity to detect latent problems, which are apprehended either insufficiently or not at all by settled routines. And they have little initiative to stage newly emergent problems in a successful and dramatic manner.

The distinction between normal and extraordinary modes of posing and solving problems, however, can be rendered fruitful for a sociological translation and realistic interpretation of the discourse concept of democracy only if we introduce two further assumptions. The illegitimate independence of social and administrative power vis-à-vis democratically generated communicative power is averted to the extent that the periphery has both (a) a specific set of capabilities and (b) sufficient occasion to exercise them. The first assumption, (a), refers to the capacities to ferret out, identify, and effectively thematize latent problems of social integration (which require political solutions); moreover, an activated periphery must then introduce them via parliamentary (or judicial) sluices into the political system in a way that disrupts the latter's routines. The second assumption, (b), is less problematic. As we have seen, the links between decentered, increasingly autonomous social sectors loosen in the course of progressive functional differentation. There is thus a growing need for integration that renders crises permanent, stimulates the public sphere, and makes accelerated learning processes necessary.

The problematic assumption is (a). It places a good part of the normative expectations connected with deliberative politics on the peripheral networks of opinion-formation. The expectations are directed at the capacity to perceive, interpret, and present societywide problems in a way that is both attention-catching and innovative. The periphery can satisfy these strong expectations only insofar as the networks of noninstitutionalized public communication make possible more or less spontaneous processes of opinionformation. Resonant and autonomous public spheres of this sort must in turn be anchored in the voluntary associations of civil society and embedded in liberal patterns of political culture and socialization; in a word, they depend on a rationalized lifeworld that meets them halfway. The development of such lifeworld

Page 359

structures can certainly be stimulated, but for the most part they elude legal regulation, administrative control, or political steering. Meaning is a scarce resource that cannot be regenerated or propagated as one likes. Here I take "meaning" as something like a limiting value for social spontaneity. Like all empirical variables, this one, too, is conditioned. But the conditions lie in lifeworld contexts that limitfrom within the capacity of associated citizens to organize their common life for themselves. What ultimately enables a legal community's discursive mode of sociation is not simply at the disposition of the members' will.

8.3 Civil Society, Public Opinion, and Communicative Power

Up to now, I have generally dealt with the public sphere as a communication structure rooted in the lifeworld through the associational network of civil society. I have described the political public sphere as a sounding board for problems that must be processed by the political system because they cannot be solved elsewhere. To this extent, the public sphere is a warning system with sensors that, though unspecialized, are sensitive throughout society. From the perspective of democratic theory, the public sphere must, in addition, amplify the pressure of problems, that is, not only detect and identify problems but also convincingly and influentially thematize them, furnish them with possible solutions, and dramatize them in such a way that they are taken up and dealt with by parliamentary complexes. Besides the "signal" function, there must be an effective problematization. The capacity of the public sphere to solve problems on its own is limited. But this capacity must be utilized to oversee the further treatment of problems that takes place inside the political system. I can provide only a broad estimate of the extent to which this is possible. I start by clarifying the contested concepts of the public sphere (sec. 8.3.1 ) and civil society (sec. 8.3.2). This allows me in sec. 8.3.3 to sketch some barriers and power structures inside the public sphere. These barriers, however, can be overcome in critical situations by escalating movements (sec. 8.3.4). I then summarize those elements the legal system must take into consideration when it forms its picture of a complex society like ours (sec. 8.3.5).

Page 360

8.3.1

The public sphere is a social phenomenon just as elementary as action, actor, association, or collectivity, but it eludes the conventional sociological concepts of "social order." The public sphere cannot be conceived as an institution and certainly not as an organization. It is not even a framework of norms with differentiated competences and roles, membership regulations, and so on. Just as little does it represent a system; although it permits one to draw internal boundaries, outwardly it is characterized by open, permeable, and shifting horizons. The public sphere can best be described as a network for communicating information and points of view (i.e., opinions expressing affirmative or negative attitudes); the streams of communication are, in the process, filtered and synthesized in such a way that they coalesce into bundles of topically specified public opinions. Like the lifeworld as a whole, so, too, the public sphere is reproduced through communicative action, forwhich mastery of a natural language suffices; it is tailored to the general comprehensibility of everyday communicative practice. We have become acquainted with the "lifeworld" as a reservoir for simple interactions; specialized systems of action and knowledge that are differentiated within the lifeworld remain tied to these interactions. These systems fall into one of two categories. Systems like religion, education, and the family become associated with general reproductive functions of the lifeworld (that is, with cultural reproduction, social integration, or socialization). Systems like science, morality, and art take up different validity aspects of everyday communicative action (truth, rightness, or veracity). The public sphere, however, is specialized in neither of these two ways; to the extent that it extends to politically relevant questions, it leaves their specialized treatment to the political system. Rather, the public sphere distinguishes itself through a communication structure that is related to a third feature of communicative action: it refers neither to the functions nor to the contents of everyday communication but to the social spacegenerated in communicative action.

Unlike success-oriented actors who mutually observe each other as one observes something in the objective world, persons acting

Page 361

communicatively encounter each other in a situation they at the same time constitute with their cooperatively negotiated interpretations. The intersubjectively shared space of a speech situation is disclosed when the participants enter into interpersonal relationships by taking positions on mutual speech-act offers and assuming illocutionary obligations. Every encounter in which actors do not just observe each other but take a second-person attitude, reciprocally attributing communicative freedom to each other, unfolds in a linguistically constituted public space. This space stands open, in principle, for potential dialogue partners who are present as bystanders or could come on the scene andjoin those present. That is, special measures would be required to prevent a third party from entering such a linguistically constituted space. Founded in communicative action, this spatial structure of simple and episodic encounters can be expanded and rendered more permanent in an abstract form for a larger public of present persons. For the public infrastructure of such assemblies, performances, presentations, and so on, architectural metaphors of structured spaces recommend themselves: we speak of forums, stages, arenas, and the like. These public spheres still cling to the concrete locales where an audience is physically gathered. The more they detach themselves from the public's physical presence and extend to the virtual presence of scattered readers, listeners, or viewers linked by public media, the clearer becomes the abstraction that enters when the spatial structure of simple interactions is expanded into a public sphere.

When generalized in this way, communication structures contract to informational content and points of view that are uncoupled from the thick contexts of simple interactions, from specific persons, and from practical obligations. At the same time, context generalization, inclusion, and growing anonymity demand a higher degree of explication that must dispense with technical vocabularies and special codes. Whereas the orientation to laypersons implies a certain loss in differentiation, uncoupling communicated opinions from concrete practical obligations tends to have an intellectualizing effect. Processes of opinion-formation, especially when they have to do with political questions, certainly cannot be separated from the transformation of the participants' preferences and attitudes, but they can be separated from putting these dispo-

Page 362

sitions into action. To this extent, the communication structures of the public sphere relievethe public of the burden of decision making, the postponed decisions are reserved for the institutionalized political process. In the public sphere, utterances are sorted according to issue and contribution, whereas the contributions are weighted by the affirmative versus negative responses they receive. Information and arguments are thus worked into focused opinions. What makes such "bundled" opinions into public opinion is both the controversial way it comes about and the amount of approval that "carries" it. Public opinion is not representative in the statistical sense. It is not an aggregate of individually gathered, privately expressed opinions held by isolated persons. Hence it must not be confused with survey results. Political opinion polls provide a certain reflection of "public opinion" only if they have been preceded by a focused public debate and a corresponding opinion-formation in a mobilized public sphere.

The diffusion of information and points of view via effective broadcasting media is not the only thing that matters in public processes of communication, nor is it the most important. True, only the broad circulation of comprehensible, attention-grabbing messages arouses a sufficiently inclusive participation. But the rules of a shared practice of communication are of greater significance for structuring public opinion. Agreement on issues and contributions develops only as the result of more or less exhaustive controversy in which proposals, information, and reasons can be more or less rationally dealt with. In general terms, the discursive level of opinion-formation and the "quality" of the outcome vary with this "more or less" in the "rational" processing of ''exhaustive" proposals, information, and reasons. Thus the success of public communication is not intrinsically measured by the requirement of inclusion either⁴⁹ but by the formal criteria governing how a qualified public opinion comes about. The structures of a power-ridden, oppressed public sphere exclude fruitful and clarifying discussions. The "quality" of public opinion, insofar as it is measured by the procedural properties of its process of generation, is an empirical variable. From a normative perspective, this provides a basis for measuring the legitimacy of the influence that public opinion has on the political system. Of course, actual influence coincides with

Page 363

legitimate influence just as little as the belief in legitimacy coincides with legitimacy. But conceiving things this way at least opens a perspective from which the relation between actual influence and the procedurally grounded quality of public opinion can be empirically investigated.

Parsons introduced "influence" as a symbolically generalized form of communication that facilitates interactions in virtue of conviction or persuasion.⁵⁰ For example, persons or institutions can enjoy a reputation that allows their utterences to have an influence on others' beliefs without having to demonstrate authority or to give explanations in the situation. "Influence" feeds on the resource of mutual understanding, but it is based on advancing trust in beliefs that are not currently tested. In this sense, public opinion represents political potentials that can be used for influencing the voting behavior of citizens or the will-formation in parliamentary bodies, administrative agencies, and courts. Naturally, political influence supported by public opinion is converted into political power—into a potential for rendering binding decisions—only when it affects the beliefs and decisions of authorized members of the political system and determines the behavior of voters, legislators, officials, and so forth. Just like social power, political influence based on public opinion can be transformed into political power only through institutionalized procedures.

Influence develops in the public sphere and becomes the object of struggle there. This struggle involves not only the political influence that has already been acquired (such as that enjoyed by experienced political leaders and officeholders, established parties, and well-known groups like Greenpeace and Amnesty International). The reputation of groups of persons and experts who have acquired their influence in special public spheres also comes into play (for example, the authority of religious leaders, the public visibility of literary figures and artists, the reputation of scientists, and the popularity of sports figures and movie stars). For as soon as the public space has expanded beyond the context of simple interactions, a differentiation sets in among organizers, speakers, and hearers; arenas and galleries; stage and viewing space. The actors' roles that increasingly professionalize and multiply with organizational complexity and range of media are, of course,

Page 364

furnished with unequal opportunities for exerting influence. But the political influence that the actors gain through public communication must ultimately rest on the resonance and indeed the approval of a lay public whose composition is egalitarian. The public of citizens must be convincedby comprehensible and broadly interesting contributions to issues it finds relevant. The public audience possesses final authority, because it is constitutive for the internal structure and reproduction of the public sphere, the only place where actors can appear. There can be no public sphere without a public.

To be sure, we must distinguish the actors who, so to speak, emerge from the public and take part in the reproduction of the public sphere itself from actors who occupy an already constituted public domain in order to use it. This is true, for example, of the large and well-organized interest groups that are anchored in various social subsystems and affect the political system through the public sphere. They cannot make any manifest use in the public sphere of the sanctions and rewards they rely on in bargaining or in nonpublic attempts at pressure. They can capitalize on their social power and convert it into political power only insofar as they can advertise their interests in a language that can mobilize convincing reasons and shared value orientations—as, for example, when parties to wage negotiations inform the public about demands, strategies, or outcomes. The contributions of interest groups are, in any case, vulnerable to a kind of criticism to which contributions from other sources are not exposed. Public opinions that can acquire visibility only because of an undeclared infusion of money or organizational power lose their credibility as soon as these sources of social power are made public. Public opinion can be manipulated but neither publicly bought nor publicly blackmailed. This is due to the fact that a public sphere cannot be "manufactured" as one pleases. Before it can be captured by actors with strategic intent, the public sphere togetherwith its public must have developed as a structure that stands on its own and reproduces itself out of itself This lawlike regularity governing the formation of a public sphere remains latent in the constituted public sphereand takes effect again only in moments when the public sphere is mobilized.

Page 365

The political public sphere can fulfill its function of perceiving and thematizing encompassing social problems only insofar as it develops out of the communication taking place among those who are potentially affected. It is carried by a public recruited from the entire citizenry. But in the diverse voices of this public, one hears the echo of private experiences that are caused throughout society by the externalities (and internal disturbances) of various functional systems—and even by the very state apparatus on whose regulatory activities the complex and poorly coordinated subsystems depend. Systemic deficiencies are experienced in the context of individual life histories; such burdens accumulate in the lifeworld. The latter has the appropriate antennae, for in its horizon are intermeshed the private life histories of the "clients" of functional systems that might be failing in their delivery of services. It is only for those who are immediately affected that such services are paid in the currency of "use values." Besides religion, art, and literature, only the spheres of "private" life have an existential language at their disposal, in which such socially generated problems can be assessed in terms of one's own life history. Problems voiced in the public sphere first become visible when they are mirrored in personal life experiences. To the extent that these experiences find their concise expression in the languages of religion, art, and literature, the "literary'' public sphere in the broader sense, which is specialized for the articulation of values and world disclosure, is intertwined with the political public sphere.⁵¹

As both bearers of the political public sphere and as members of society, citizens occupy two positions at once. As members of society, they occupy the roles of employees and consumers, insured persons and patients, taxpayers and clients of bureaucracies, as well as the roles of students, tourists, commuters, and the like; in such complementary roles, they are especially exposed to the specific requirements and failures of the corresponding service systems. Such experiences are first assimilated "privately," that is, are interpreted within the horizon of a life history intermeshed with other life histories in the contexts of shared lifeworlds. The communication channels of the public sphere are linked to private spheresto the thick networks of interaction found in families and circles of friends as well as to the looser contacts with neighbors, work

Page 366

colleagues, acquaintances, and so on—and indeed they are linked in such a way that the spatial structures of simple interactions are expanded and abstracted but not destroyed. Thus the orientation to reaching understanding that is predominant in everyday practice is also preserved for a communication among strangers that is conducted over great distances in public spheres whose branches are quite complex. The threshold separating the private sphere from the public is not marked by a fixed set of issues or relationships but by different conditions of communication. Certainly these conditions lead to differences in the accessibility of the two spheres, safeguarding the intimacy of the one sphere and the publicity of the other. However, they do not seal off the private from the public but only channel the flow of topics from the one sphere into the other. For the public sphere draws its impulses from the private handling of social problems that resonate in life histories. It is symptomatic of this close connection, incidentally, that a modern bourgeois public sphere developed in the European societies of the seventeenth and eighteenth centuries as the "sphere of private persons come together as a public." Viewed historically, the connection between the public and the private spheres is manifested in the clubs and organizational forms of a reading public composed of bourgeois private persons and crystallizing around newspapers and journals.⁵²

8.3.2

This sphere of civil society has been rediscovered today in wholly new historical constellations. The expression "civil society" has in the meantime taken on a meaning different from that of the "bourgeois society" of the liberal tradition, which Hegel conceptualized as a "system of needs," that is, as a market system involving social labor and commodity exchange. What is meant by ''civil society" today, in contrast to its usage in the Marxist tradition, no longer includes the economy as constituted by private law and steered through markets in labor, capital, and commodities. Rather, its institutional core comprises those nongovernmental and noneconomic connections and voluntary associations that anchor the communication structures of the public sphere in the society

Page 367

component of the lifeworld. Civil society is composed of those more or less spontaneously emergent associations, organizations, and movements that, attuned to how societal problems resonate in the private life spheres, distill and transmit such reactions in amplified form to the public sphere. The core of civil society comprises a network of associations that institutionalizes problemsolving discourses on questions of general interest inside the framework of organized public spheres.⁵³ These "discursive designs" have an egalitarian, open form of organization that mirrors essential features of the kind of communication around which they crystallize and to which they lend continuity and permanence.⁵⁴

Such associations certainly do not represent the most conspicuous element of a public sphere dominated by mass media and large agencies, observed by market and opinion research, and inundated by the public relations work, propaganda, and advertising of political parties and groups. All the same, they do form the organizational substratum of the general public of citizens. More or less emerging from the private sphere, this public is made of citizens who seek acceptable interpretations for their social interests and experiences and who want to have an influence on institutionalized opinion- and will-formation.

One searches the literature in vain for clear definitions of civil society that would go beyond such descriptive characterizations.⁵⁵ S. N. Eisenstadt's usage reveals a certain continuity with the older theory of pluralism when he describes civil society as follows:

	Civil society embraces a multiplicity of ostensibly "private" yet potentially autonomous public arenas distinct from the state. The activities of such actors are regulated by various associations existing within them, preventing the society from degenerating into a shapeless mass. In a civil society, these sectors are not embedded in closed, ascriptive or corporate settings; they are open-ended and overlapping. Each has autonomous access to the central political arena, and a certain degree of commitment to that setting.^56

Jean Cohen and Andrew Arato, who have presented the most comprehensive study on this topic, provide a catalog of features characterizing the civil society that is demarcated from the state, the economy, and other functional systems but coupled with the core private spheres of the lifeworld:

Page 368

	(1) Plurality: families, informal groups, and voluntary associations whose plurality and autonomy allow for a variety of forms of life; (2) Publicity: institutions of culture and communication; (3) Privacy: a domain of individual self-development and moral choice; (4) Legality: structures of general laws and basic rights needed to demarcate plurality, privacy, and publicity from at least the state and, tendentially, the economy. Together, these structures secure the institutional existence of a modern differentiated civil society.^57

The constitution of this sphere through basic rights provides some indicators for its social structure. Freedom of assembly and freedom of association, when linked with freedom of speech, define the scope for various types of associations and societies: for voluntary associations that intervene in the formation of public opinion, push topics of general interest, and act as advocates for neglected issues and underrepresented groups; for groups that are difficult to organize or that pursue cultural, religious, or humanitarian aims; and for ethical communities, religious denominations, and so on. Freedom of the press, radio, and television, as well as the right to engage in these areas, safeguards the media infrastructure of public communication; such liberties are thereby supposed to preserve an openness for competing opinions and a representative diversity of voices. The political system, which must remain sensitive to the influence of public opinion, is intertwined with the public sphere and civil society through the activity of political parties and general elections. This intermeshing is guaranteed by the right of parties to "collaborate" in the political will-formation of the people, as well as by the citizens' active and passive voting rights and other participatory rights. Finally, the network of associations can assert its autonomy and preserve its spontaneity only insofar as it can draw support from a mature pluralism of forms of life, subcultures, and worldviews. The constitutional protection of"privacy" promotes the integrity of private life spheres: rights of personality, freedom of belief and of conscience, freedom of movement, the privacy of letters, mail, and telecommunications, the inviolability of one's residence, and the protection of families circumscribe an untouchable zone of personal integrity and independent judgment.

The tight connection between an autonomous civil society and an integral private sphere stands out even more clearly when

Page 369

contrasted with totalitarian societies of bureaucratic socialism. Here a panoptic state not only directly controls the bureaucratically desiccated public sphere, it also undermines the private basis of this public sphere. Administrative intrusions and constant supervision corrode the communicative structure of everyday contacts in families and schools, neighborhoods and local municipalities. The destruction of solidary living conditions and the paralysis of initiative and independent engagement in overregulated yet legally uncertain sectors go hand in hand with the crushing of social groups, associations, and networks; with indoctrination and the dissolution of cultural identities; with the suffocation of spontaneous public communication. Communicative rationality is thus destroyed simultaneously in both public and private contexts of communication.⁵⁸ The more the bonding force of communicative action wanes in private life spheres and the embers of communicative freedom die out, the easier it is for someone who monopolizes the public sphere to align the mutually estranged and isolated actors into a mass that can be directed and mobilized in a plebiscitarian manner.⁵⁹

Basic constitutional guarantees alone, of course, cannot preserve the public sphere and civil society from deformations. The communication structures of the public sphere must rather be kept intact by an energetic civil society. That the political public sphere must in a certain sense reproduce and stabilize itself from its own resources is shown by the odd self-referential character of the practice of communication in civilsociety. Those actors who are the carriers of the public sphere put forward "texts" that always reveal the same subtext, which refers to the critical function of the public sphere in general. Whatever the manifest content of their public utterances, the performative meaning of such public discourse at the same time actualizes the function of an undistorted political public sphere as such. Thus, the institutions and legal guarantees of free and open opinion-formation rest on the unsteady ground of the political communication of actors who, in making use of them, at the same time interpret, defend, and radicalize their normative content. Actors who know they are involved in the common enterprise of reconstituting and maintaining structures of the public sphere as they contest opinions and strive for influence differ from

Page 370

actors who merely use forums that already exist. More specifically, actors who support the public sphere are distinguished by the dual orientation of their political engagement: with their programs, they directly influence the political system, but at the same time they are also reflexively concerned with revitalizing and enlarging civil society and the public sphere as well as with confirming their own identities and capacities to act.

Cohen and Arato see this kind of "dual politics" especially in the "new" social movements that simultaneously pursue offensive and defensive goals. "Offensively," these movements attempt to bring up issues relevant to the entire society, to define ways of approaching problems, to propose possible solutions, to supply new information, to interpret values differently, to mobilize good reasons and criticize bad ones. Such initiatives are intended to produce a broad shift in public opinion, to alter the parameters of organized political will-formation, and to exert pressure on parliaments, courts, and administrations in favor of specific policies. "Defensively," they attempt to maintain existing structures of association and public influence, to generate subcultural counterpublics and counterinstitutions, to consolidate new collective identities, and to win new terrain in the form of expanded rights and reformed institutions:

	On this account, the "defensive" aspect of the movements involves preserving and developingthe communicative infrastructure of the lifeworld. This formulation captures the dual aspect of movements discussed by Touraine as well as Habermas's insight that movements can be the carriers of the potentials of cultural modernity. This is the sine qua non for successful efforts to redefine identities, to reinterpret norms, and to develop egalitarian, democratic associational forms. The expressive, normative and communicative modes of collective action . . . [also involve] efforts to secure institutional changes within civil society that correspond to the new meanings, identities, and norms that are created.^60

In the self-referential mode of reproducing the public sphere, as well as in theJanus-faced politics aimed at the political system and the self-stabilization of public sphere and civil society, the space is provided for the extension and radicalization of existing rights: "The combination of associations, publics, and rights, when supported by a political culture in which independent initiatives and

Page 371

movements represent an ever-renewable, legitimate, political option, represents, in our opinion, an effective set of bulwarks around civil society within whose limits much of the program of radical democracy can be reformulated."⁶¹

In fact, the interplay of a public sphere based in civil society with the opinion- and will-formation institutionalized in parliamentary bodies and courts offers a good starting point for translating the concept of deliberative politics into sociological terms. However, we must not look on civil society as a focal point where the lines of societal self-organization as a whole would converge. Cohen and Arato rightly emphasize the limited scopefor action that civil society and the public sphere afford to noninstitutionalized political movements and forms of political expression. They speak of a structurally necessary "self-limitation" of radical-democratic practice:

First, a robust civil society can develop only in the context of a liberal political culture and the corresponding patterns of socialization, and on the basis of an integral private sphere; it can blossom only in an already rationalized lifeworld. Otherwise, populist movements arise that blindly defend the frozen traditions of a lifeworld endangered by capitalist modernization. In their forms of mobilization, these fundamentalist movements are as modern as they are antidemocratic ⁶².

Second, within the boundaries of the public sphere, or at least of a liberal public sphere, actors can acquire only influence, not political power. The influence of a public opinion generated more or less discursively in open controversies is certainly an empirical variable that can make a difference. But public influence is transformed into communicative power only after it passes through the filters of the institutionalized procedures of democratic opinion- and will-formation and enters through parliamentary debates into legitimate lawmaking. The informal flow of public opinion issues in beliefs that have been tested from the standpoint of the generalizability of interests. Not influence per se, but influence transformed into communicative power legitimates political decisions. The popular sovereignty set communicatively aflow cannot make itself felt solely in the influence of informal public discourses—not even when these discourses arise from autonomous

Page 372

public spheres. To generate political power, their influence must have an effect on the democratically regulated deliberations of democratically elected assemblies and assume an authorized form in formal decisions. This also holds, mutatis mutandis, for courts that decide politically relevant cases.

Third, and finally, the instruments that politics has available in law and administrative power have a limited effectiveness in functionally differentiated societies. Politics indeed continues to be the addressee for all unmanaged integration problems. But political steering can often take only an indirect approach and must, as we have seen, leave intact the modes of operation internal to functional systems and other highly organized spheres of action. As a result, democratic movements emerging from civil society must give up holistic aspirations to a self-organizing society, aspirations that also undergirded Marxist ideas of social revolution. Civil society can directly transform only itself, and it can have at most an indirect effect on the self-transformation of the political system; generally, it has an influence only on the personnel and programming of this system. But in no way does it occupy the position of a macrosubject supposed to bring society as a whole under control and simultaneously act for it. Besides these limitations, one must bear in mind that the administrative power deployed for purposes of social planning and supervision is not a suitable medium for fostering emancipated forms of life. These can develop in the wake of democratization processes but they cannot be brought about through intervention.

The self-limitation of civil society should not be understood as incapacitation. The knowledge required for political supervision or steering, a knowledge that in complex societies represents a resource as scarce as it is desirable, can certainly become the source of a new systems paternalism. But because the administration does not, for the most part, itself produce the relevant knowledge but draws it from the knowledge system or other intermediaries, it does not enjoy a natural monopoly on such knowledge. In spite of asymmetrical access to expertise and limited problem-solving capacities, civil society also has the opportunity of mobilizing counterknowledge and drawing on the pertinent forms of expertise to make its own translations. Even though the public consists of

Page 373

laypersons and communicates with ordinary language, this does not necessarily imply an inability to differentiate the essential questions and reasons for decisions. This can serve as a pretext for a technocratic incapacitation of the public sphere only as long as the political initiatives of civil society fail to provide sufficient expert knowledge along with appropriate and, if necessary, multilevel translations in regard to the managerial aspects of public issues.

8.3.3

The concepts of the political public sphere and civil society introduced above are not mere normative postulates but have empirical relevance. However, additional assumptions must be introduced if we are to use these concepts to translate the discourse-theoretic reading of radical democracy into sociological terms and reformulate it in an empirically falsifiable manner. I would like to defend the claim that under certain circumstances civil society can acquire influence in the public sphere, have an effect on the parliamentary complex (and the courts) through its own public opinions, and compel the political system to switch over to the official circulation of power. Naturally, the sociology of mass communication conveys a skeptical impression of the power-ridden, mass-media-dominated public spheres of Western democracies. Social movements, citizen initiatives and forums, political and other associations, in short, the groupings of civil society, are indeed sensitive to problems, but the signals they send out and the impulses they give are generally too weak to initiate learning processes or redirect decision making in the political system in the short run.

In complex societies, the public sphere consists of an intermediary structure between the political system, on the one hand, and the private sectors of the lifeworld and functional systems, on the other. It represents a highly complex network that branches out into a multitude of overlapping international, national, regional, local, and subcultural arenas. Functional specifications, thematic foci, policy fields, and so forth, provide the points of reference for a substantive differentiation of public spheres that are, however, still accessible to laypersons (for example, popular science and

Page 374

literary publics, religious and artistic publics, feminist and "alternative" publics, publics concerned with health-care issues, social welfare, or environmental policy). Moreover, the public sphere is differentiated into levels according to the density of communication, organizational complexity, and range—from the episodic publics found in taverns, coffee houses, or on the streets; through the occasional or "arranged" publics of particular presentations and events, such as theater performances, rock concerts, party assemblies, or church congresses; up to the abstract public sphere of isolated readers, listeners, and viewers scattered across large geographic areas, or even around the globe, and brought together only through the mass media. Despite these manifold differentiations, however, all the partial publics constituted by ordinary language remain porous to one another. The one text of "the" public sphere, a text continually extrapolated and extending radially in all directions, is divided by internal boundaries into arbitrarily small texts for which everything else is context; yet one can always build hermeneutical bridges from one text to the next. Segmented public spheres are constituted with the help of exclusion mechanisms; however, because publics cannot harden into organizations or systems, there is no exclusion rule without a proviso for its abolishment.

In other words, boundaries inside the universal public sphere as defined by its reference to the political system remain permeable in principle. The rights to unrestricted inclusion and equality built into liberal public spheres prevent exclusion mechanisms of the Foucauldian type and ground a potentialforself-transformation. In the course of the nineteenth and twentieth centuries, the universalist discourses of the bourgeois public sphere could no longer immunize themselves against a critique from within. The labor movement and feminism, for example, were able tojoin these discourses in order to shatter the structures that had initially constituted them as "the other" of a bourgeois public sphere.⁶³

The more the audience is widened through mass communications, the more inclusive and the more abstract in form it becomes. Correspondingly, the roles of the actors appearing in the arenas are, to an increasing degree, sharply separated from the roles of the spectators in the galleries. Although the "success of the actors in the

Page 375

arena is ultimately decided in the galleries,"⁶⁴ the question arises of how autonomous the public is when it takes a position on an issue, whether its affirmative or negative stand reflects a process of becoming informed or in fact only a more or less concealed game of power. Despite the wealth of empirical investigations, we still do not have a well-established answer to this cardinal question. But one can at least pose the question more precisely by assuming that public processes of communication can take place with less distortion the more they are left to the internal dynamic of a civil society that emerges from the lifeworld.

One can distinguish, at least tentatively, the more loosely organized actors who "emerge from" the public, as it were, from other actors merely "appearing before" the public. The latter have organizational power, resources, and sanctions available from the start. Naturally, the actors who are more firmly anchored in civil society and participate in the reproduction of the public sphere also depend on the support of "sponsors" who supply the necessary resources of money, organization, knowledge, and social capital. But patrons or "like-minded" sponsors do not necessarily reduce the authenticity of the public actors they support. By contrast, the collective actors who merely enter the public sphere from, and utilize it for, a specific organization or functional system have their own basis of support. Among these political and social actors who do not have to obtain their resources from other spheres, I primarily include the large interest groups that enjoy social power, as well as the established parties that have largely become arms of the political system. They draw on market studies and opinion surveys and conduct their own professional public-relations campaigns.

In and of themselves, organizational complexity, resources, professionalization, and so on, are admittedly insufficient indicators for the difference between "indigenous" actors and mere users. Nor can an actor's pedigree be read directly from the interests actually represented. Other indicators are more reliable. Thus actors differ in how they can be identified. Some actors one can easily identify from their functional background; that is, they represent political parties or pressure groups; unions or professional associations; consumer-protection groups or rent-control organizations, and so on. Other actors, by contrast, must first

Page 376

produce identifying features. This is especially evident with social movements that initially go through a phase of self-identification and self-legitimation; even after that, they still pursue a self-referential ''identity politics" parallel to their goal-directed politics—they must continually reassure themselves of their identity. Whether actors merely use an already constituted public sphere or whether they are involved in reproducing its structures is, moreover, evident in the above-mentioned sensitivity to threats to communication rights. It is also shown in the actors' willingness to go beyond an interest in self-defense and take a universalist stand against the open or concealed exclusion of minorities or marginal groups. The very existence of social movements, one might add, depends on whether they find organizational forms that produce solidarities and publics, forms that allow them to fully utilize and radicalize existing communication rights and structures as they pursue special goals.⁶⁵

A third group of actors are the journalists, publicity agents, and members of the press (i.e., in the broad sense of Publizisten) who collect information, make decisions about the selection and presentation of "programs," and to a certain extent control the entry of topics, contributions, and authors into the mass-media-dominated public sphere. As the mass media become more complex and more expensive, the effective channels of communication become more centralized. To the degree this occurs, the mass media face an increasing pressure of selection, on both the supply side and the demand side. These selection processes become the source of a new sort of power. This power of the media is not sufficiently reined in by professional standards, but today, by fits and starts, the "fourth branch of government" is being subjected to constitutional regulation. In the Federal Republic, for example, it is both the legal form and the institutional structure of television networks that determine whether they depend more on the influence of political parties and public interest groups or more on private firms with large advertising outlays. In general, one can say that the image of politics presented on television is predominantly made up of issues and contributions that are professionally produced as media input and then fed in via press conferences, news agencies, publicrelations campaigns, and the like. These official producers of

Page 377

information are all the more successful the more they can rely on trained personnel, on financial and technical resources, and in general on a professional infrastructure. Collective actors operating outside the political system or outside large organizations normally have fewer opportunities to influence the content and views presented by the media. This is especially true for messages that do not fall inside the "balanced," that is, the centrist and rather narrowly defined, spectrum of "established opinions" dominating the programs of the electronic media.⁶⁶

Moreover, before messages selected in this way are broadcast, they are subject to informationprocessing strategies within the media. These are oriented by reception conditions as perceived by media experts, program directors, and the press. Because the public's receptiveness, cognitive capacity, and attention represent unusually scarce resources for which the programs of numerous "stations" compete, the presentation of news and commentaries for the most part follows market strategies. Reporting facts as humaninterest stories, mixing information with entertainment, arranging material episodically, and breaking down complex relationships into smaller fragments—all of this comes together to form a syndrome that works to depoliticize public communication.⁶⁷ This is the kernel of truth in the theory of the culture industry. The research literature provides fairly reliable information on the institutional framework and structure of the media, as well as on the way they work, organize programs, and are utilized. But, even a generation after Paul Lazarsfeld, propositions concerning the effects of the media remain controversial. The research on effect and reception has at least done away with the image of passive consumers as "cultural dopes" who are manipulated by the programs offered to them. It directs our attention to the strategies of interpretation employed by viewers, who communicate with one another, and who in fact can be provoked to criticize or reject what programs offer or to synthesize it with judgments of their own.⁶⁸

Even if we know something about the internal operation and impact of the mass media, as well as about the distribution of roles among the public and various actors, and even if we can make some reasonable conjectures about who has privileged access to the media and who has a share in media power, it is by no means clear

Page 378

how the mass media intervene in the diffuse circuits of communication in the political public sphere. The normative reactions to the relatively new phenomenon of the mass media's powerful position in the competition for public influence are clearer. Michael Gurevitch and Jay G. Blumler have summarized the tasks that the media ought to fulfill in democratic political systems:

1. surveillance of the sociopolitical environment, reporting developments likely to impinge, positively or negatively, on the welfare of citizens;

2. meaningful agenda-setting, identifying the key issues of the day, including the forces that have formed and may resolve them;

3. platforms for an intelligible and illuminating advocacy by politicians and spokespersons of other causes and interest groups;

4. dialogue across a diverse range of views, as well as between powerholders (actual and prospective) and mass publics;

5. mechanisms for holding officials to account for how they have exercised power;

6. incentives for citizens to learn, choose, and become involved, rather than merely to follow and kibitz over the political process;

7. a principled resistance to the efforts of forces outside the media to subvert their independence, integrity and ability to serve the audience;

8. a sense of respect for the audience member, as potentially concerned and able to make sense of his or her political environment.⁶⁹

Such principles orient the professional code of journalism and the profession's ethical self-understanding, on the one hand, and the formal organization of a free press by laws governing mass communication, on the other.⁷⁰ In agreement with the concept of deliberative politics, these principles express a simple idea: the mass media ought to understand themselves as the mandatary of an enlightened public whose willingness to learn and capacity for criticism they at once presuppose, demand, and reinforce; like the judiciary, they ought to preserve their independence from political and social pressure; they ought to be receptive to the public's concerns and proposals, take up these issues and contributions impartially, augment criticisms, and confront the political process with articulate demands for legitimation. The power of the media

Page 379

should thus be neutralized and the tacit conversion of administrative or social power into political influence blocked. According to this idea, political and social actors would be allowed to "use" the public sphere only insofar as they make convincing contributions to the solution of problems that have been perceived by the public or have been put on the public agenda with the public's consent. In a similar vein, political parties would have to participate in the opinion- and will-formation from the public's own perspective, rather than patronizing the public and extracting mass loyalty from the public sphere for the purposes of maintaining their own power.⁷¹

The sociology of mass communication depicts the public sphere as infiltrated by administrative and social power and dominated by the mass media. If one places this image, diffuse though it might be, alongside the above normative expectations, then one will be rather cautious in estimating the chances of civil society having an influence on the political system. To be sure, this estimate pertains only to a public sphere at rest. In periods of mobilization, the structures that actually support the authority of a critically engaged public begin to vibrate. The balance of power between civil society and the political system then shifts.

8.3.4

With this I return to the central question of who can place issues on the agenda and determine what direction the lines of communication take. Roger Cobb,Jennie-Keith Ross, and Marc Howard Ross have constructed models that depict how new and compelling issues develop, from the first initiative up to formal proceedings in bodies that have the power to decide.⁷² If one suitably modifies the proposed models—inside access model, mobilization model, outside initiative model—from the viewpoint of democratic theory, they present basic alternatives in how the public sphere and the political system influence each other. In the first case, the initiative comes from officeholders or political leaders, and the issue continues to circulate inside the political system all the way to its formal treatment, while the broader public is either excluded from the process or does not have any influence on it. In the second case, the

Page 380

initiative again starts inside the political system, but the proponents of the issue must mobilize the public sphere, because they need the support of certain groups, either to obtain formal consideration or to implement an adopted program successfully. Only in the third case does the initiative lie with forces at the periphery, outside the purview of the political system. With the help of the mobilized public sphere, that is, the pressure of public opinion, such forces compel formal consideration of the issue:

	The outside initiative model applies to the situation in which a group outside the government structure 1) articulates a grievance, 2) tries to expand interest in the issue to enough other groups in the population to gain a place on the public agenda, in order to 3) create sufficient pressure on decision makers to force the issue onto the formal agenda for their serious consideration. This model of agenda building is likely to predominate in more egalitarian societies. Formal agenda status, . . . however, does not necessarily mean that the final decisions of the authorities or the actual policy implementation will be what the grievance group originally sought.^73

In the normal case, issues and proposals have a history whose course corresponds more to the first or second model than to the third. As long as the informal circulation of power dominates the political system, the initiative and power to put problems on the agenda and bring them to a decision lies more with the Government leaders and administration than with the parliamentary complex. As long as in the public sphere the mass media prefer, contrary to their normative self-understanding, to draw their material from powerful, well-organized information producers and as long as they prefer media strategies that lower rather than raise the discursive level of public communication, issues will tend to start in, and be managed from, the center, rather than follow a spontaneous course originating in the periphery. At least, the skeptical findings on problem articulation in public arenas accord with this view.⁷⁴ In the present context, of course, there can be no question of a conclusive empirical evaluation of the mutual influence that politics and public have on each other. For our purposes, it suffices to make it plausible that in a perceived crisis situation, the actors in civil society thus far neglected in our scenario can assume a surprisingly active and momentous role.⁷⁵ In spite of a lesser organiza-

Page 381

tional complexity and a weaker capacity for action, and despite the structural disadvantages mentioned earlier, at the critical moments of an accelerated history, these actors get the chance to reverse the normal circuits of communication in the political system and the public sphere. In this way they can shift the entire system's mode of problem solving.

The communication structures of the public sphere are linked with the private life spheres in a way that gives the civil-social periphery, in contrast to the political center, the advantage of greater sensitivity in detecting and identifying new problem situations. The great issues of the last decades give evidence for this. Consider, for example, the spiraling nuclear-arms race; consider the risks involved in the peaceful use of atomic energy or in other large-scale technological projects and scientific experimentation, such as genetic engineering; consider the ecological threats involved in an overstrained natural environment (acid rain, water pollution, species extinction, etc.); consider the dramatically progressing impoverishment of the Third World and problems of the world economic order; or consider such issues as feminism, increasing immigration, and the associated problems of multiculturalism. Hardly any of these topics were initially brought up by exponents of the state apparatus, large organizations, or functional systems. Instead, they were broached by intellectuals, concerned citizens, radical professionals, self-proclaimed "advocates," and the like. Moving in from this outermost periphery, such issues force their way into newspapers and interested associations, clubs, professional organizations, academies, and universities. They find forums, citizen initiatives, and other platforms before they catalyze the growth of social movements and new subcultures.⁷⁶ The latter can in turn dramatize contributions, presenting them so effectively that the mass media take up the matter. Only through their controversial presentation in the media do such topics reach the larger public and subsequently gain a place on the "public agenda." Sometimes the support of sensational actions, mass protests, and incessant campaigning is required before an issue can make its way via the surprising election of marginal candidates or radical parties, expanded platforms of "established" parties, important court decisions, and so on, into the core of the political system and there receive formal consideration.

Page 382

Naturally, there are other ways in which issues develop, other paths from the periphery to the center, and other patterns involving complex branchings and feedback loops. But, in general, one can say that even in more or less power-ridden public spheres, the power relations shift as soon as the perception of relevant social problems evokes a crisis consciousness at the periphery. If actors from civil society then join together, formulate the relevant issue, and promote it in the public sphere, their efforts can be successful, because the endogenous mobilization of the public sphere activates an otherwise latent dependency built into the internal structure of every public sphere, a dependency also present in the normative self-understanding of the mass media: the players in the arena owe their influence to the approval of those in the gallery. At the very least, one can say that insofar as a rationalized lifeworld supports the development of a liberal public sphere by furnishing it with a solid foundation in civil society, the authority of a positiontaking public is strengthened in the course of escalating public controversies. Under the conditions of a liberal public sphere, informal public communication accomplishes two things in cases in which mobilization depends on crisis. On the one hand, it prevents the accumulation of indoctrinated masses that are seduced by populist leaders. On the other hand, it pulls together the scattered critical potentials of a public that was only abstractly held together through the public media, and it helps this public have a political influence on institutionalized opinion- and will-formation. Only in liberal public spheres, of course, do subinstitutional political movements—which abandon the conventional paths of interest politics in order to boost the constitutionally regulated circulation of power in the political system—take this direction. By contrast, an authoritarian, distorted public sphere that is brought into alignment merely provides a forum for plebiscitary legitimation.⁷⁷

This sense of a reinforced demand for legitimation becomes especially clear when subinstitutional protest movements reach a high point by escalating their protests. The last means for obtaining more of a hearing and greater media influence for oppositional arguments are acts of civil disobedience. These acts of nonviolent, symbolic rule violation are meant as expressions of protest against

Page 383

binding decisions that, their legality notwithstanding, the actors consider illegitimate in the light of valid constitutional principles. Acts of civil disobedience are directed simultaneously to two addressees. On the one hand, they appeal to officeholders and parliamentary representatives to reopen formally concluded political deliberations so that their decisions may possibly be revised in view of the continuing public criticism. On the other hand, they appeal "to the sense of justice of the majority of the community," as Rawls puts it,⁷⁸ and thus to the critical judgment of a public of citizens that is to be mobilized with exceptional means. Independently of the current object of controversy, civil disobedience is also always an implicit appeal to connect organized political willformation with the communicative processes of the public sphere. The message of this subtext is aimed at a political system that, as constitutionally organized, may not detach itself from civil society and make itself independent vis-à-vis the periphery. Civil disobedience thereby refers to its own origins in a civil society that in crisis situations actualizes the normative contents of constitutional democracy in the medium of public opinion and summons it against the systemic inertia of institutional politics.

This self-referential character is emphasized in the definition that Cohen and Arato have proposed, drawing on considerations raised by Rawls, Dworkin, and me:

	Civil disobedience involves illegal acts, usually on the part of collective actors, that are public, principled, and symbolic in character, involve primarily nonviolent means of protest, and appeal to the capacity for reason and the sense of justice of the populace. The aim of civil disobedience is to persuade public opinion in civil and political society  . . . that a particular law or policy is illegitimate and a change is warranted. Collective actors involved in civil disobedience invoke the utopian principles of constitutional democracies, appealing to the ideas of fundamental rights or democratic legitimacy. Civil disobedience is thus a means for reasserting the link between civil and political society  . . ., when legal attempts at exerting the influence of the former on the latter have failed and other avenues have been exhausted.^79

This interpretation of civil disobedience manifests the self-consciousness of a civil society confident that at least in a crisis it can increase the pressure of a mobilized public on the political system

Page 384

to the point where the latter switches into the conflict mode and neutralizes the unofficial countercirculation of power.

Beyond this, the justification of civil disobedience^80 relies on a dynamic understandingof the constitution as an unfinished project. From this long-term perspective, the constitutional state does not represent a finished structure but a delicate and sensitive—above all fallible and revisable—enterprise, whose purpose is to realize the system of rights anew in changing circumstances, that is, to interpret the system of rights better, to institutionalize it more appropriately, and to draw out its contents more radically. This is the perspective of citizens who are actively engaged in realizing the system of rights. Aware of, and referring to, changed contexts, such citizens want to overcome in practice the tension between social facticity and validity. Although legal theory cannot adopt this participant perspective as its own, it can reconstruct the paradigmatic understanding of law and democracy that guides citizens whenever they form an idea of the structural constraints on the selforganization of the legal community in their society.

8.3.5

From a reconstructive standpoint, we have seen that constitutional rights and principles merely explicate the performative character of the self-constitution of a society of free and equal citizens. The organizational forms of the constitutional state make this practice permanent. Every historical example of a democratic constitution has a double temporal reference: as a historic document, it recalls the foundational act that it interprets—it marks a beginning in time. At the same time, its normative character means that the task of interpreting and elaborating the system of rights poses itself anew for each generation; as the project of ajust society, a constitution articulates the horizon of expectation opening on an everpresent future. From this perspective, as an ongoing process of constitution making set up for the long haul, the democratic procedure of legitimate lawmaking acquires a privileged status. This leads to the pressing question of whether such a demanding procedure can be implemented in complex societies like our own and, if it can, how this can be done effectively, so that a constitution-

Page 385

ally regulated circulation of power actually prevails in the political system. The answers to this question in turn inform our own paradigmatic understanding of law. I note the following four points for elucidating such a historically situated understanding of the constitution.

(a) The constitutionally organized political system is, on the one hand, specialized for generating collectively binding decisions. To this extent, it represents only one of several subsystems. On the other hand, in virtue of its internal relation to law, politics is responsible for problems that concern society as a whole. It must be possible to interpret collectively binding decisions as a realization of rights such that the structures of recognition built into communicative action are transferred, via the medium of law, from the level of simple interactions to the abstract and anonymous relationships among strangers. In pursuing what in each case are particular collective goals and in regulating specific conflicts, politics simultaneously deals with general problems of integration. Because it is constituted in a legal form, a politics whose mode of operation is functionally specified still refers to society-wide problems: it carries on the tasks of social integration at a reflexive level when other action systems are no longer up to the job.

(b) This asymmetrical position explains the fact that the political system is subject to constraints on two sides and that corresponding standards govern its achievements and decisions. As a functionally specified action system, it is limited by other functional systems that obey their own logic and, to this extent, bar direct political interventions. On this side, the political system encounters limits on the effectiveness of administrative power (including legal and fiscal instruments). On the other side, as a constitutionally regulated action system, politics is connected with the public sphere and depends on lifeworld sources of communicative power. Here the political system is not subject to the external constraints of a social environment but rather experiences its internal dependence on enabling conditions. This is because the conditions that make the production of legitimate law possible are ultimately not at the disposition of politics.

(c) The political system is vulnerable on both sides to disturbances that can reduce the effectiveness of its achievements and the

Page 386

legitimacy of its decisions, respectively. The regulatory competence of the political system fails if the implemented legal programs remain ineffective or if regulatory activity gives rise to disintegrating effects in the action systems that require regulation. Failure also occurs if the instruments deployed overtax the legal medium itself and strain the normative composition of the political system. As steering problems become more complex, irrelevance, misguided regulations, and self-destruction can accumulate to the point where a ''regulatory trilemma" results.⁸¹ On the other side, the political system fails as a guardian of social integration if its decisions, even though effective, can no longer be traced back to legitimate law. The constitutionally regulated circulation of power is nullified if the administrative system becomes independent of communicatively generated power, if the social power of functional systems and large organizations (including the mass media) is converted into illegitimate power, or if the lifeworld resources for spontaneous public communication no longer suffice to guarantee an uncoerced articulation of social interests. The independence of illegitimate power, together with the weakness of civil society and the public sphere, can deteriorate into a "legitimation dilemma," which in certain circumstances can combine with the steering trilemma and develop into a vicious circle. Then the political system is pulled into the whirlpool of legitimation deficits and steering deficits that reinforce one another.

(d) Such crises can at most be explained historically. They are not built into the structures of functionally differentiated societies in such a way that they would intrinsically compromise the project of self-empowerment undertaken by a society of free and equal subjects who bind themselves by law. However, they are symptomatic of the peculiar position of political systems as asymmetrically embedded in highly complex circulation processes. Actors must form an idea of this context whenever, adopting the performative attitude, they want to engage successfully as citizens, representatives,judges, or officials, in realizing the system of rights. Because these rights must be interpreted in various ways under changing social circumstances, the light they throw on this context is refracted into a spectrum of changing legal paradigms. Historical constitutions can be seen as so many ways of construing one and the

Page 387

same practice—the practice of self-determination on the part of free and equal citizens—but like every practice this, too, is situated in history. Those involved must start with their own current practice if they want to achieve clarity about what such a practice means in general.