INTRODUCTION
The Decentering of Democracies
For us, the primary characteristic of a democratic regime is the anointment by the people of those who govern. The idea that the people are the sole legitimate source of power has come to be taken for granted. No one would dream of contesting or even questioning it. "Sovereignty cannot be divided," as a great French republican of the nineteenth century put it. "One must choose between the elective principle and the hereditary principle. Authority must be legitimated either by the freely expressed will of all or by the supposed will of God. The people or the Pope! Choose."1 To answer the question was to evade the need for any kind of argument. We have yet to move beyond this stage. Yet the assertion blurs an important distinction: as a practical matter, it is assumed that the general will coincides with the will of the majority. There has been little discussion of this point. The fact that a majority vote establishes the legitimacy of a government has indeed been universally accepted as a procedure marking the essence of democracy. This definition of legitimacy at first seemed natural, since it marked a definitive break with the previous way of doing things, in which minorities dictated their law. Expressions such as the "great majority" or "vast majority" established the law of numbers, in contrast to the minority rule characteristic of despotic and aristocratic regimes. At first, it was the difference in the origins of power and the foundation of political obligation that was crucial. Later, the majority principle came to be recognized in a more narrowly procedural sense. In a classic formulation, "majority rule is one of those simple ideas that gain immediate acceptance. It does not favor anyone in advance and places all voters on the same level."2
Founding Fictions
The transition from the celebration of the People or the Nation, always in the singular, to majority rule is anything but self-evident, however, since the two terms are situated on different planes. One is a general, or,
1 Louis Blanc, "Reforme electorale," Revue du Progres, vol. 2, October 15,1839, p. 308. 1 Adhemar Esmein, elements de droit constitutionnel francais et compare, 8th ed. (Paris, 1927), vol. l,p. 330.
2   ■ Introduction
if you wish, philosophical positing of a political subject, while the other is a pragmatic selection procedure. Democratic election thus conflates a principle of justification with a technique of decision. The routine identification of the one with the other ultimately masked the latent contradiction. Indeed, the two terms are not of the same nature. Majority rule is persuasive enough as a procedure but more problematic if understood in sociological terms. In the latter case it inevitably takes on an arithmetic aspect: it designates only a fraction of "the People," even if it is the dominant fraction. The justification of power by the ballot box has always implicitly rested on the idea of a general will and thus on a "people" symbolically standing in for the whole of society. This sociological notion was reinforced by a moral insistence on equality and a legal imperative of respect for rights, an insistence that consideration be given to the intrinsic value of each member of the community. Out of this developed a certain ideal of unanimity, which has been one of the underpinnings of the democratic idea from the beginning: "democratic," in the broadest possible sense, means "expressive of social generality." But we behave as though the majority were the same as the whole, as though majority rule were an acceptable way of imposing stronger demands on the governed. This first blurring of distinctions was connected to a second: the identification of the nature of a regime with the conditions under which it was established. The part stands for the whole, and the electoral moment stands for the entire term of government. The legitimacy of democratic governments rests on these two postulates.
The problem is that this basic double fiction has little by little come to be seen as an intolerable distortion of the truth. By the end of the nineteenth century, just as universal (male) suffrage was beginning to spread across Europe, signs of early disenchantment began to emerge everywhere. The specter of mass rule, initially so frightening to liberals, soon gave way to the reality of regimes hamstrung by the narrowness of their own preoccupations. The words people and nation, which had previously fed expectations and imaginations, were somehow diminished, drowned as they were in partisan squabbling and electioneering. The party system, which none of the early theorists of democracy had foreseen or analyzed, established itself everywhere as the actual center of political life, and government became enmeshed in the rivalries of personalities and clans. The legislature, which from the beginning had been taken to epitomize the spirit and form of representative government, lost its preeminence, and the nature of its operation changed. The initial idea—that of a temple of public reason in which representatives would debate the definition of the general.interest—in practice devolved into a system of bargaining in thrall to special interests. Meanwhile, energy continued to be invested in elections, and genuine issues were discussed. But the electoral ritual
The Decentering of Democracies   • 3
itself, once a celebration of the apotheosis of the citizen through universal suffrage, had lost its luster. Throughout the period 1890-1920, which saw the publication of countless books aimed at explaining "the crisis of democracy," the idea that a majoritarian electoral system could somehow express the interests of the whole of society lost all credibility. Many felt that elections and parliaments led to governments in which the logic of special interests prevailed over the requirement of generality. To be sure, the principle that governments should be elected by majority rule remained unchallenged, but no one believed any longer that majority rule was an automatic guarantee of governmental virtue.
Double Legitimacy: The Inception and Decline of a System
Responding to this loss of confidence in the period before and after the Great War, 1890-1920, people searched for ways to revive the democratic ideal. The most extreme solutions were explored, including totalitarian ones. Amid this turmoil, however, something else emerged that would quietly change the nature of democratic regimes: an authentic administrative power, or bureaucracy/Everywhere states grew stronger and organized themselves more^efficiently. Indeed, the growth of the state was closely related to efforts to reformulate the basic principles of democratic government. One sought to portray the "bureaucratic machine" itself as a force for the realization of the general interest. To conceptualize this, the public service model was developed in France and the rational administration model in the United States. The former proposed a sort of corporatism of the universal, in which bureaucrats were urged to identify with their mission, to become "interested in disinterestedness." The latter envisioned a search for generality through scientific management. Old ideas of rational government and positive politics, which from the Enlightenment to Auguste Comte had encouraged efforts to promote public welfare beyond the clash of partisan passions, were thus updated and brought into the democratic realm.
Since the unification of individual wills proved problematic, a search began for more realistic and objective ways of achieving social generality. Concrete steps were taken in this direction. Little by little, without really conceptualizing the shift, democratic regimes established themselves on a dual foundation: universal suffrage'and public administration. No longer was the bureaucracy a mere instrument for the exercise of political power: it acquired a measure of autonomy, based on competence. Equal access to the civil service complemented equality at the ballot box. Those who aspired to represent or interpret social generality faced tests of two kinds: elections on the one hand and competitive examinations on the other.
4   • Introduction
Elections were a "subjective" choice, governed by the system of interests and opinions, while competitive examinations were an "objective" device for selecting the most competent individuals. In the French case, universal suffrage and public service became the two denning features of republican ideology. The "Jacobin mandarins" of the high civil service embodied the Republic as fully as the people's chosen representatives. Alongside the legitimacy of election—consecration by the ballot box—emerged a second type of democratic legitimacy: legitimacy through identification with social generality. In practice, this second form of legitimacy would play a crucial role in compensating for the decline of electoral legitimacy. Two major conceptualizations of legitimacy thus came together: legitimacy based on social recognition of some form of power, and legitimacy based on conformity to some norm or system of values. These two overlapping forms of legitimacy—procedural and substantial-^-afforded the democratic regimes of the twentieth century a certain solidity. But in the 1980s this solution to the problem began to come undone.
First, legitimation by the ballot box suffered from the. diminished prestige of elections, what might be called their "desacralization." In the "golden age" of the representative system, election bestowed an incontestable mandate that allowed the winner subsequently to govern "freely." It was assumed that future policies were implicit in the terms of the electoral decision simply because that decision was framed by a predictable universe of choices structured by disciplined organizations with well-defined programs and clearly understood differences. This is no longer the case. The function of elections has been whittled down: elections are simply the process by which we designate those who govern. They no longer provide a priori legitimation for-policies to be enacted later. Furthermore, the meaning of the word' majority has changed. Although the legal, political, and parliamentary definition remains clear, the sociological implications of the term are far less precise. The interests of "the greater number" can no longer be identified as readily as in the past with the interests of the majority. The "people" can no longer be apprehended as a homogeneous mass. It is felt to be rather a series of separate histories, an accumulation of specific situations. Hence societies today increasingly understand themselves in terms of minorities. A minority is no longer merely the "smaller number" (and therefore obliged to bow before the "greater"). It has become one of a series of diffracted expressions of the social totality. Society nowadays manifests itself as a long litany of minority conditions. "People" has become the plural of "minority."
In addition, the "administrative power," or bureaucracy, has been largely delegitimized. Neoliberal rhetoric has played a part in this by damaging the credibility of the state and proposing the market as the new regulator of collective well-being. More concretely, the new public
The Decentering of Democracies   • 5
management movement has cast doubt on the classic figure of the civil servant as the authorized representative of the general interest. The upper echelons of the civil service have been most affected by this development. They are no longer capable, it seems, of representing the future in a more open, less predictable world. (Admittedly, they have also been undermined by massive defection of the elite owing to the growing disparity between private- and public-sector salaries.) As the average level of education rises, society becomes less willing to take for granted the notion that technocrats are uniquely endowed with the virtues of rationality and disinterestedness. The old style—"benevolent" bureaucrats administering a society treated as an underage ward of the state—has become both economically untenable and sociologically unacceptable. The bureaucracy has thus been stripped of the moral and professional qualities that were once its strength. Its legitimacy has therefore suffered, along with that of elected representatives.
The New Age of Legitimacy
The collapse of the old system, with its dual legitimacy, and the various changes that provoked and followed from that collapse in the 1980s, did not simply leave a vacuum in their wake. Despite a powerful sense of loss or even decay, a quiet reconstruction also began. Citizens voiced new demands. The hope of achieving a government that would serve the general interest found novel forms of expression and embraced new ideas. Values of impartiality,1 pluralism,, compassion, and proximity were strongly emphasized.Teflecting a new understanding of democratic generality and thus of the sources and forms of legitimacy. Independent agencies and constitutional courts were created or expanded and assigned new roles. Finally, new techniques of governance emerged, with increased emphasis on image and communication. The whole landscape of government changed in ways that need to be understood and appreciated. Description alone will not suffice, however. We need to identify the concepts that can make sense of this evolving new world and try to make out the new forms of democracy that may emerge in the future. So although description of what was said and done is important, as is lucid appreciation of the inadequacy, ambiguity, and even risks inherent in what was accomplished, our goal is to develop ideal types that can help us to think about and shape the new system that is beginning to emerge. Nothing is yet set in stone. New possibilities are mixed up with incipient pathologies.
What happened in the 1980s? One major feature of the change was a latent reformulation of the terms in which the democratic imperative of expressing generality was understood. To take the full measure of this
6   • Introduction
The Decentering of Democracies
7
development, we need to look back at earlier formulations of the general interest. Universal suffrage rested on an aggregate definition: the general will was represented as the voice of the masses of citizen voters. The civil service invoked a more objective notion of generality: the idea was that public reason and the general interest were in some sense identified with the structure of the republican state itself. In both cases, generality was taken to be something with a palpable physical incarnation. With the collapse of electoral and bureaucratic legitimacy, three less direct ways of constructing social generality emerged:
1. Achievement of generality by way of detachment from particularity, through systematic rational construction of a point of view at some distance from any particular aspect of a given issue. This defines power in terms of un lieu vide, an empty place or vacuum. The generality of an institution is then reflected in the fact that no one can appropriate it. This is a negative generality. It is characterized by a structural variable (the fact of independence)-and a behavioral variable (the maintenance of distance or equilibrium). This negative generality is what allows an institution to oversee or regulate the activities of others and what distinguishes the bearers of such authority from elective branches of government.
2. Achievement of generality through multiplication of the expressions of social sovereignty. Here the goal is to realize the objectives of democracy by making the democratic subject more complex or by adopting more complex democratic forms. In this respect, an important aim is to compensate for the failure of electoral majorities to embody the general will. I call this a generality of multiplication. For example, a constitutional court partakes of this form of generality when it subjects decisions of the majority party to constitutional scrutiny.
3. Achievement of generality through consideration of the variety of situations, or of society as comprising a myriad of special cases. This form of generality arises out of radical immersion in particularity, marked by concern for concrete individuals. It exhibits certain behavioral characteristics. It results from the actions of a government that forgets no one, that involves itself in everyone's problems. It is associated with an art of government that lies at the opposite extreme from the nomocratic vision. Instead of defining society in terms of a legal principle of equality, equidistant from all forms of particularity, this third type of generality takes all existing situations into account. Such a practice can be described asa "descent into generality."31 call this the generality of attention to] particularity;.,
3 In contrast to the usual sociological notion of "ascent to generality," which is achieved by raking one's distance from each specific case in order to arrive at an encompassing general concept.
These three ways of envisioning generality have one thing in common: the social totality is understood neither as an arithmetic aggregate (with unanimity as an underlying ideal) nor as a monist unity (with the social interest thought of as a stable property of a collective body or structure). They are the result of a much more "dynamic" approach, of generalization conceived of as an operation, a type of action. In a sense, they correspond to three possible investigative strategies: one can examine an object with a telescope; examine various cross-sections of the object under a microscope; or explore the object by tracing a series of paths through it. In this perspective, generality constitutes a regulatory horizon. It is no longer a palpable, substantial thing, as it is taken to be in the concept of the "general" will or "general" interest.
From this approach we discover three new types of legitimacy, each associated with one of three types of social generality described above: the legitimacy of impartiality (associated with negative generality); the legitimacy of reflexivity (associated with the generality of proliferation); and the legitimacy of proximity (associated with the generality of attention to particularity). This veritable revolution in the conception of legitimacy partakes of a broader decentering of democracy. The diminished prestige of the electoral process is only one aspect of this decentering. In Counter-Democracy I described the emergence of new forms of political investment: the people as watchdog, the people as veto players, and the people as judge. Each of these new forms helped to counter the declining importance of the ballot box. Democratic politics became something more than merely electing representatives. There are now many more ways in which a regime can be recognized as democratic, some of which complement the consecration of the polling booth while others compete with it.
The new forms of legitimacy are defined by qualities, jn contrast to the older legitimacies of election and selection, whicfrderived from the intrinsic properties of certain institutions (the ballot box and the competitive examination bestowed a certain status on those who successfully survived these trials). Hence the new forms of legitimacy are never definitively acquired. They remain precarious, always open to challenge, and dependent on social perceptions of institutional actions and behavior. This is a crucial point: it reflects the fact that these new forms do not fit within the usual typology, in which legitimacy as social recognition is contrasted with legitimacy as conformity to a norm. The legitimacies of impartiality, reflexivity, and proximity include both of these dimensions. They are hybrids. They share with institutions the ability to embody values and principles, but at the same time they remain inoperative unless socially recognized as such. With the rise of these new forms of legitimacy it is therefore conceivable that democracy is embarking on a new era. The emerging "figure of legitimacy" transcends the traditional opposition
8   • Introduction
The Decentering of Democracies
9
between the guardians of "republican generality," concerned mainly with substance, and the proponents of "strong democracy," who are interested primarily in the intensity of social mobilization.
The new forms of legitimacy also enlarge another, -classic typology based on the opposition between what has been called I'input legitimacy" and "output legitimacy."4 This distinction is not without its uses. It reminds us that citizens judge their rulers by their actions and suggests that nonelective institutions may be deemed legitimate as long as they contribute to outcomes recognized as socially useful.5 What interests me here, however, is the broader question of the legitimacy of institutions themselves. For that reason I am also not satisfied with proceduralist approaches such as Habermas's. Habermas, too, wants to go beyond substantialist approaches to democracy and urges us to look at the general will in terms of discursive dissemination.15 Nevertheless, he remains within the confines of a monist vision of popular sovereignty. He merely shifts the locus of that sovereignty from a concrete social body to a diffuse space of communication. In my view, the redefinition of legitimacy starts with a deconstruction and reconstitution of the idea of social generality, which leads to a radical pluralization of the forms of legitimacy. The idea is that there is more than one way to act or speak "on behalf of society" and to be representative. The three new legitimacies that I have proposed constitute a system in which each complements the other two to establish a more exigent democratic ideal.
This change is all the more decisive because of the importance that the question of legitimacy has assumed in today's world. As the Utopian ideologies that once gave solidity to the political order from "outside" have receded, that order has had to seek its justification from within. Like trust between individuals, legitimacy is an "invisible institution." It establishes a firm foundation for the relation between the governing and the
4 See Fritz Scharpf, Governing in Europe: Effective and Democratic? (New York: Oxford University Press, 1999). The opposition between input and output democracy was proposed by Robert E. Goodin, Reflective Democracy (New York: Oxford University Press, 2003).
5 It is worth noting that this distinction was formulated in response to the question of whether there exists a "democratic deficit" in European Union institutions.
6 Jürgen Habermas, "La souverainete populaire comme procedure: Un concept normatif d'espace public," Lignes, no. 7, September 1989. The same thing can be said about the approach of Bernard Manin, who proposes to replace the impossible demand for unanimity with an idea of universal deliberation as a way of redefining democratic legitimacy. See his article, "On Legitimacy and Political Deliberation," Political Theory, vol. 1J, no. 3, 1987. He shifts the focal point of the unanimity constraint, but at bottom he clings to the traditional perspective of a legitimacy of establishment, to which he merely imparts a more realistic formulation, even if the ideal of "a free and equal deliberation of all" is itself materially very difficult to achieve—it, too, can only be approximated, so that in the end one still has to pretend that everyone has deliberated.
governed. If legitimacy in the broadest sense simply implies absence of coercion, democratic legitimacy requires something more: a tissue of relationships between government and society. The essence of democracy— the social appropriation of political power—depends on this. Democratic legitimacy exists when citizens believe in their own government, which cannot happen unless they have a sense of empowerment. The efficacy of public action depends on legitimacy, and the sense of legitimacy affects the way in which citizens judge the quality of their country's democracy. In these respects, legitimacy is an "invisible institution" as well as a "sensitive indicator" of the society's political expectations and the response to those expectations. A broader, more searching definition of legitimacy is therefore an essential component of any effort to expand the meaning of democracy.
A Revolution Whose Outcome Remains Indeterminate
The examples of legitimacy discussed thus far are closely connected with institutions of two kinds: independent oversight and regulatory authorities on the one hand and constitutional courts on the other. The former are conceived and organized in such a way as to enjoy, potentially, what I have called the legitimacy of impartiality. Some were created by legislatures for the purpose of checking and balancing an executive deemed to be overly partisan; others were created by the executive itself, in order to restore credibility by shedding certain of its own powers or to shift responsibility for policy areas in which it felt it lacked the necessary competence. By contrast, the function of constitutional courts is to subject legislation to scrutiny according to criteria of generality different from those of majority rule. The legitimacy of such courts exemplifies what I am calling the legitimacy of reflexivity. The growing influence of these two types of institutions has considerably altered the nature of legislative and executive power as conceived by the leading figures of the American and French revolutions. Traditional democratic theory has had little to say about them. Having increased their power everywhere, independent oversight authorities and constitutional courts have begun to change the way in which the question of democracy is framed. The importance of this change cannot be overstated. Indeed, it is striking in retrospect to see how stable the conceptualization of democratic institutions had remained over two centuries.7
7 With the exception of political parties and their relation to the democratic process, which were the subject of intense debates and projects of reform around the turn of the twentieth century (consider, for example, the question of primary elections in the United
10   • Introduction
The Decentering of Democracies   •   11
From the end of the eighteenth century to the 1980s, the discussion was framed in terms of concepts that hardly varied, as any historian of the great revolutions of modern times can verify. Throughout this period, the questions of representative government, direct democracy, separation of powers, the role of public opinion, and guarantees of human rights were posed in terms that remained more or less unchanged. The political vocabulary itself barely evolved. The term autogestion (self-management), a product of the 1960s, was one of the few neologisms of any real importance. Yet even this novel idea vanished fairly soon after it appeared, an indication that it marked a turning point of which it then became the first victim. The new grammar of democratic institutions, which encompasses both independent authorities and constitutional courts, marks a rupture with the previous order of things. But for want of theoretical elaboration (never having found its Emmanuel-Joseph Sieyes or its James Madison), the magnitude of this change has not been properly appreciated. It is a product of circumstances, a response to the latent expectations of citizens and to a wide variety of perceived demands on public management.
Because these kinds of institutions were not conceptualized as novel political forms, they did not find their proper place in the democratic order. Hence no transcendent logic governs the way in which they may develop. They may yet deepen our sense of democracy, or ihc> may simply reinforce anxious liberal attitudes toward popular rule. For example, the traditional understanding of constitutional courts is that their purpose is to limit the expression of popular sovereignty by bolstering the authority of law. The underlying distinction between "government by will" and "government by constitution" is an old liberal topos.8 What are the proper limits of the power of the majority? Here, the question is implicitly framed by the old denunciation of the "tyranny of the majority" by nineteenth-century liberals afraid of being submerged by the tide of universal suffrage. But the development of constitutional courts can also be seen as an instrument for limiting the government's room for maneuver and therefore a way of increasing social control over representatives. As one important nineteenth-century political commentator explained, a constitution can be seen as a "safeguard demanded by the people against those who do the public's business, so that they do not abuse their mandate."5 Similarly, independent regulatory and oversight authorities can also be seen in two contrasting lights.
States). I will have more to say later about the precursors of this recent change, which differ from country to country.
8 The distinction appears to have been formulated first by Henry St. John Bolingbroke in his Dissertation upon Parties (1733), in Henry Bolingbroke, Political Writings, David Armitage, ed. (Cambridge: Cambridge University Press, 1997), p. 90.
9 Edouard Laboulaye, Questions constitutionnelles (Paris, 1872), p. 373.
Clearly, no stable picture has yet emerged in these two areas of democratic theory. It is therefore essential to be clear about what the issues are. Only then can the democratic potential of institutions of these kinds be exploited, and only then can these institutions be designed in such a way as to reinforce the insistence on generality in the public sphere. The institutions in question can then yield indirect benefits similar to those usually ascribed to the procedures of direct democracy. On this basis it may be possible to construct a theory of indirect democracy to compensate for the deficiencies of electoral-representative democracy.
Unlike the first two types of legitimacy, the legitimacy of proximity is not associated with any particular type of institution. It grows, rather, out of a range of social expectations as to the behavior of those who govern. Our attention thus shifts to the development of a democratic art of government—a second dimension of the new democratic realm. Historically, reflections on democracy were aimed at defining the rules and institutions that constitute a regime of popular sovereignty (allocation of powers, modes of representation, forms of citizen intervention, etc.). The political sphere was conceptualized in terms of two categories: the regime type and the manner in which decisions are made (in other words, how "policies" are shaped). To take account of social expectations and demands, these categories were broadened to include the art of government. Numerous studies have shown that citizens are at least as sensitive to the behavior of the people in government as to the precise nature of the decisions they make.10 The use of a novel vocabulary to describe the desired bonds between government and society attests to a certain evolution in this regard. In addition to the traditional terms for describing the representative bond, we find a new insistence on attentiveness, openness, fairness, compassion, recognition, respect, and presence. Words such as participation and proximity, which were rooted in the traditional vocabulary and therefore relatively available, have become increasingly common in public discourse. Behind these words we find not only heightened citizen demands (and thus new scope for the application of democratic ideals) but also political rhetoric as well as sophisticated techniques for the manipulation of public opinion.
The purpose of this work is to develop a conceptual framework for evaluating the democratic potential of these still embryonic and often ambivalent institutions and practices. The only way to achieve this goal is to construct ideal types corresponding to each of the new paradigms of generality andSegitimacy discussed above. This will have the added benefit of revealing the conditions under which these new paradigms may have perverse consequences, in the hope of reinforcing their positive contribution to a more democratic politics.
10 The results of which will be presented in the text.
12   • Introduction
The New Democratic Dualism
Describing the advent of democracy in his own time, Tocqueville observed: "The idea of government has been simplified: number alone determines what is law and what is right. All politics is reduced to a question of arithmetic."11 Today one would have to say exactly the opposite. The striking fact is that democracy is becoming more complex. We see this in a pair of dualities: between electoral-representative institutions and the institutions of indirect democracy, and also between the realm of procedures and behaviors, and decisions. Democracy as regime type rests on the first dualism, democracy as government on the second. These two dualisms are superimposed on the tension between electoral democracy and counterdemocracy, which defines the sphere of citizen activity. Taken together, these dualities define the new democratic order.
To begin with, the institutions of electoral-representative democracy form a system with the institutions of indirect democracy. Their articulation makes it possible to reconcile majority rule with the ideal of unanimity: k tension is set up between these two poles in such a way as to respect the requirements of each. This tension is central to the democratic idea, and from it flow two pairs of contradictory requirements:
First, a contradiction between the recognition of the legitimacy of conflict and the aspiration to consensus. Democracy is a pluralistic regime, which implies the acceptance of divergent interests and opinions. Electoral competition is organized around these differences. Elections institutionalize the conflict and its resolution. Democracy cannot exist without clear means of resolving such differences. Democratic politics implies choosing sides, taking a stand. In societies marked by social divisions and uncertainty as to the future, this dimension of democratic politics is essential. Yet at the same time there can be no democracy without a shared world and recognition of shared values, so that conflict need not escalate to the extreme of civil war.12 If both dimensions are to be respected, there is therefore a need to distinguish between institutions of
11 Alexis de Tocqueville, Considerations sur la Revolution (material for L'Ancien Regime et la Revolution), in Tocqueville, CEuvres (Paris: Gallimard, 2004), vol. 3, p. 492.
12 In this regard, Nicole Loraux has often called attention to the trouble that the word kratos caused in Athens. It suggested the idea of "having the upper hand," of victory of one group over another. The art of coming to a decision by way of a majority vote was thus associated from its inception with the image of a conflict resolved by force. Yet at the same time, there was compensatory celebration of the united demos and an incantatory appeal to unity of all citizens. See Nicole Loraux, La Citee divisee (Paris: Payot, 1997), and "La majorite, le tout et la moitie: Sur l'arithmetique athenienne du vote," Le Genre humain, no. 22,1990. The failure of Greek democracy can be understood in this perspective as the result of an inability to articulate and balance these two dimensions.
The Decentering of Democracies   ■ 13
conflict on the one hand and institutions of consensus on the other. On one side, the subjective partisan realm of electoral-representative competition; on the other, the objective world of institutions of indirect democracy. Recognizing the distinctive character of the latter makes it possible to give both poles of the democratic tension their due. It also counters the enduring temptation to deny the legitimacy of conflict, which has so often manifested itself in the past, and to hypostasize the idea of unanimity (a fantasy that has repeatedly fed illusions and led to perverse consequences that have undermined democratic regimes).
Second, a contradiction between a realistic principle of decision (majority rule) and a necessarily more demanding principle of justification (unanimity). No democracy can exist unless it is possible to reach a decision and act in a timely manner and unless the necessity of arbitration and choice is recognized. Neither can there be democracy without institutions whose mission it is never to lose sight of the general interest and to contribute autonomously to its realization. A democratic polity must therefore separate and sustain the tension between majoritarian institutions and institutions governed by consensus justification.
The organization of this duality-requires full recognition of the fact that democracy rests on a; necessary fiction, the assimilation of the majority to the unanimous whole. Organization makes the tension explicit and arranges for the coexistence of the two elements from which it stems. Indeed, the problem is that this fiction has never been recognized as such. This is not usually the case with legal fictions. Normally, reliance on such fictions does not deceive anyone. Legal forms that involve proceeding "as if" something were true are not intended to hide anything. They are merely a way of gaining control, reducing complexity, or taming contradictions in the interest of governability. As Yan Thomas rightly points out, legal fictions "establish the power to control reality by ostensibly denying it."13 Their meaning is clearly limited by their function and makes no claim to change the real nature of things. The fundamental fiction of democracy was not understood in these terms. It was never made explicit but rather dissimulated and left unacknowledged. This was necessary in order to establish the democratic idea on a firm footing, since it was impossible at the time to conceptualize a decisive and effective political order without unanimity of decision. Recognizing duality is a way of escaping from this impasse. It makes yisible the separation of the two poles of the democratic idea and encourages citizens to unravel the implicit fictions that can distort that idea or divert its practical consequences.
13 Yan Thomas, "Fictio legis: L'empire de la fiction romaine et ses limites medievale," Droits, no. 21,1995, p. 20. A legal fiction, Thomas continues, "takes the form of a decision to counter reality" (ibid., p. 22).
14   ■ Introduction
Majority rule should therefore be understood, prosaically, as a mere empirical convention, which remains subject to the need for higher levels of justification. Its legitimacy is imperfect and must be strengthened by other modes of democratic legitimation.
Alongside this duality of institutions a second duality has emerged, a duality that structures democracy as a form of government. In considering the question of government, executive power was for a long time of only marginal interest to political theorists. "Government" was an idea with no solidity of its own. In practice, government remained hidden behind its decisions. For a long time this neglect was justified by the central-ity ascribed to the legislative power. This was true of the revolutionary period in France, when the legitimate power of generality, identified with "the law," stood in stark contrast to the suspect power of managing particularity, which was seen as the essence of the executive. Theorists were slow to recognize the relative autonomy of governmental action owing to the substantial intellectual obstacles that had to be overcome.14
But executive power was envisioned at the time solely in terms of the content of governmental actions and decisions. The enduring field of public policy studies attests to the permanence of this approach in contemporary political science. Recently, however, a new dimension of executive power has emerged: it bears on the conducts those in power. This looms large in the minds of citizens but ha^ yet to be theorized as such. This has given rise to a tension between the democracy of decision making (embedded in the strictly political dynamic of universal suffrage) and a democracy of behaviors (with its implicit requirement that the needs of all citizens be taken into account).
The two emerging continents of the democratic universe also constitute a system. The expectation is that, by a variety of routes, they will contribute to the creation of a more democratic society and thus to the realization of the democratic project, which is as much to institute a society of equal individuals as it is to establish a regime of collective sovereignty. These developments reflect contemporary demands for greater individualization on the one hand (with increased emphasis on the distinctiveness of each individual) and, on the other hand, greater awareness of the general interest (and thus of the need to reduce the influence of special interests on governing institutions).
i4 See Joseph Barthelemy, he Role du pouvoir executif dans les repitbliques modemes {Paris, 1907); and Michel Verpeaux, La Naissance du pouvoir reglementaire, 1789-1799 (Paris: PUF, 1991).