To get our reflections going, here are some concrete examples – what Americans would call “hard cases”. Suppose that after a referendum or a majority vote of the elected representatives of the people, women’s right to abortion were to be repealed. Can the political will of the people do such a thing, simply because it is the will of the people and universal suffrage is the instrument of the legitimacy of governments and decisions in a democratic system? Or can that will be blocked, not by discussions or disputes using moral, philosophical and sociological arguments, but by courts using legal arguments? More precisely: in a democracy is it possible – and, more importantly, is it legitimate – to stop the people (or their representatives) from doing away with abortion, by depending on the constitutional right of women to do what they want with their bodies? Of course, this is a fictional case and no one here imagines that it could one day become a real one. Nevertheless, in some ways it does resemble some actual cases.
In August 1993 the Constitutional Council (the French judicial body that exercises constitutional review) overturned some provisions in a proposed law concerning immigration control, on the ground that they entailed “excessive infringements on fundamental and constitutional rights that are recognized to be held by everyone residing in the Republic’s territory”. The Interior Minister – Charles Pasqua, who had introduced the bill – immediately accused the judges of preventing the government from implementing an immigration policy that had been debated by the nation, approved in the March 1993 legislative elections, and adopted “unanimously by the partisan majority in both the National Assembly and the Senate”. In September 2005, the Justice Minister, Pascal Clément, defended in the Assembly a bill applying a security device – the electronic bracelet – retroactively, to people already in custody. Being aware of – in his own words – “the constitutional riskiness” of this retroactivity, Clément declared that “recent events” (two convicted rapists had been accused of reoffending earlier that week) “impel me to accept that risk, and every member of the national legislature can share the risk with me, simply by ignoring the Council’s decision, and those who recognize it will take on the political and human responsibility of preventing the new law from being applied to those currently in custody”.  Clément’s declaration received the following response from Pierre Mazeaud, the President of the Constitutional Council: “Respecting the Constitution is not a risk, it is a duty.” In February 2008, the Council decided it was contrary to the legal principle of non-retroactivity to apply a law’s provision for preventive detention at the end of a prison sentence to persons already convicted. The President of the Republic, Nicolas Sarkozy, immediately asserted that the principle of non-retroactivity should not be available to the most dangerous criminals, because “the immediate application of preventive detention to convicted criminals is a legitimate objective for the protection of victims”.  He wrote to the chief justice of the Court of Cassation (the highest court of appeal) asking that legal means be found to achieve that objective.
The arguments are the same in all of these cases: to the judge who reminds elected officeholders to respect constitutional rights and where necessary punishes their violations of them, the officeholders reply: universal suffrage! parliamentary majority! public opinion! Indeed, on what grounds can the people be forbidden to want what they want? Limiting the power of the people is justified in a libertarian system, in which equilibrium is the principle of legitimacy; in a dictatorial system, where the party is that principle; in a theocratic system where the legitimating principle is God; and so on. But where is the foundation for limiting the power of the people in a democratic system, in which the legitimating principle is universal suffrage? In 1981 a Socialist member of the National Assembly found fame by strongly and crudely expressing this dilemma, saying to the opposition: “You are legally in the wrong because you are politically in a minority”!
As a matter of fact, this issue is not new; one could go back to the “case” of Pontius Pilate, which Hans Kelsen comments on in The Essence and Value of Democracy: when the Sanhedrin convicts Jesus, and Pilate has to choose the punishment, “because he does not know what truth is and – as a Roman – is accustomed to think democratically, Pilate appeals to the People and conducts a vote”.  Everyone is familiar with the outcome. Would Jesus have been crucified if Pilate himself had made the decision? If not, then imagine how different history would have been!
Bearing in mind these cases as a point of departure – in both senses of that expression: starting with them but also moving on from them – it will be useful to focus on the assumptions that underlie them, the political and social philosophies that impel them, and the political and constitutional developments that can follow from them. The idea is to step back in our memories of these cases to understand better their contemporary meaning and to discover their possible futures. To do this, it is obvious that several different kinds of knowledge need to be used, and in the research project in my laboratory they are; but here we will draw only on knowledge about constitutionalism.
What is a constitution?
In the context of constitutional intelligibility, the foregoing cases are meaningful only if the constitution is thought of as a way of limiting power. However, this view assumes a particular political function of the constitution – limiting power – which is not part of the essence of the constitution, but has grown out of the history of politicians’ strategic requirements, and has acquired stability and rationality by means of legal doctrines. To positivists a constitution is simply the particular document that organizes the state legally, and its legal validity resides in its relation to the fundamental hypothetical-deductive norm that says we must obey the constitution – not the constitution that limits power, but the constitution regardless of its content. For positivists, the political function of a constitution is a matter of politics, not of law, and is therefore an issue that is indifferent to the character and the validity of the constitution; legally, the term “constitution” applies to any document legally organizing the state, whether with a separation or a confusion of powers, with or without a recognition of fundamental rights, and whether it limits or facilitates the arbitrariness of power. Obviously, in this understanding of the word “constitution”, a constitution does not imply democracy, and constitutionalism is merely a doctrine that declares the need for a formal constitution.
To continue highlighting the assumptions in the foregoing cases: these cases make sense only if the constitution is thought of as a norm, i.e. a norm of law, and of hard law too, issuing commandments and prohibitions – obligations to do or not to do. This view assumes a normativeness in the constitution that does not belong to its texture but is constructed by social actors, especially judicial actors. For legal realists, the constitution is simply a set of words – just “marks on paper” – at most, in Kelsen’s words, “subjective propositions about norms”, but is not itself a norm. Thus, Pierre Avril writes, a constitution “does not say anything”; it becomes a norm by the work of interpreting the words, work done by those who use it, mainly by judges; the norm is not in the textual statements in the constitution, it is in the meaning assigned to those statements. It follows that if the constitution does not say anything, it cannot tell us anything about the democratic character of a political regime, just as it cannot be a limit or constraint on the exercise of power.
The advantage of first of all going back to these legal assumptions is that it shows that, contrary to convention wisdom, the idea of a constitution is not consubstantial with the idea of democracy. For legal positivists and realists, these are separate and autonomous ideas with no relationship or reciprocal influence between them. That probably explains why – the positivist school having dominated the field of law for so long, and still remaining very strong – jurists have moved away from studies, research and reflections on democracy , abandoning this idea to philosophers, historians, sociologists, etc. In other words, the ideas of constitution and democracy are related to each other only in the epistemological context of a doctrine – constitutionalism – that considers a constitution as a means of democracy – “democracy through law”. This expected political function of a constitution is presented as a necessary result of the three properties of a constitution. First of all, a constitution is a written text; having written rules about exercising power enables the people to see whether or not actual exercises of power respect this text, and to punish any violations. That was the explicit intention of the authors of the Declaration of the Rights of Man and of the Citizen in August 1789, who said one of the purposes of publishing a declaration of rights was “so that the acts of the legislative as well as of the executive power can at any time be compared with the purpose of all political institutions, and therefore be more respected”. Next, a constitution is a text that organizes the separation of powers, in which internal mechanisms – checks and balances – prevents one institution from seizing all powers, thus producing an institutional balance supporting citizens’ political freedom. Finally, a constitution is a text that sets out the rights that citizens are entitled to claim against actions by governments. This too is stated in the preamble of the Declaration of Rights of 1789, to be published “so that citizens’ demands, henceforth based on simple and undeniable principles, will always redound to the maintenance of the constitution and to the happiness of everyone”. With these three properties – being written, organizing separate powers, and recognizing fundamental rights – a constitution is, in Benjamin Constant’s famous formulation, “the guarantee of the liberty of a people”.
A crisis in the meaning of constitutionalism
So that is the idea of a constitution conveyed by constitutionalism, which makes it a doctrine of democracy. Today, this constitutionalism faces a crisis; or, to avoid a word damaged through overuse, let’s say instead that the meaning of this constitutionalism has become an issue. The first reason for this is in the balance among the three basic properties of constitutions. Initially, i.e. in the 18th century, the dominant property that gave meaning to the idea of constitutionalism was the separation of powers: a constitution is what determines the separation of powers. Of course, the third property – acknowledging citizens’ rights and liberties – was not forgotten, but it did not have its own protective mechanism, because it was regarded as the necessary result of the limitation of power that came out of it being divided: as Montesquieu stated, political liberty would be a lost cause “if the same man or the same body of men – whether the nobles or the people – exercised the three powers of making the law, executing public resolutions, and judging crimes”. Gradually this property lost some of its strength and authority because of the political practices that showed that whatever the choice of constitutional organization, legislative and executive powers are, thanks to the logic of elections by majorities, united in the hands of the head of the executive, whether in those of the President of the Republic as in France or in those of the Prime Minister as in Spain, Germany, and the United Kingdom. So it appeared that the constitution-as-separation-of powers was no longer the way to guarantee democracy and citizens’ political liberty. This relative decline brought about an increase of the strength of the third property, respect for rights, and particularly encouraged the emergence of an institution to look after this: courts with the power of constitutional review. With the protection of rights not being guaranteed mechanically by the operation of the separation of powers, it has to be done by a special device, the legal review of the constitutionality of laws, which implies the possibility of a judge taking action against legislative infringements of constitutional rights. The constitution that supports democracy is therefore no longer the constitution that guarantees fundamental rights through the separation of powers, but the one that does this through constitutional review. It is no longer the constitution-as-separation-of powers, but the constitution-as-fundamental-rights.
Added to this first, internal shift, readjusting the properties of the constitutional idea, is a second, external shift, changing the target of constitutions. In the 18th century, constitutional restraints on the exercise of power were addressed to institutions not based on universal suffrage; their targets were institutions that were hereditary (kings) or based on limited suffrage (parliamentary bodies). In that understanding, a constitution can be a guarantee for the people, whom it protects against the potential tyranny of powers that they do not control. But today, with the extension of universal suffrage, these restraints target institutions based on popular voting. So if a constitution is still defined as an act of mistrust, the object of the mistrust has become universal suffrage and the institutions arising out of it. In this understanding, a constitution is a guarantee against the people: it protects constitutional rights against the will of the people or their elected representatives.
These two shifts sum up the crisis in the meaning of constitutionalism. While the original idea of a constitution was a political device to separate powers in order to oversee and to limit the exercise of powers with no electoral legitimacy, today it works as a legal device to protect fundamental rights against the exercise of powers that do have electoral legitimacy. This radical change of perspective requires reopening the question about the meaning of constitutionalism, for the meaning of the phrase “democracy by constitution” depends on whether the powers outlined in the constitution have an electoral origin. There is a great temptation to question the contemporary meaning of constitutionalism by contrasting the new concept of a constitution – guaranteeing fundamental rights and checking up on constitutionality – with the concept of electoral democracy, implicitly or explicitly presented as the legitimate and natural idea of democracy as “government by the people”. However, this temptation is to be resisted, for it generally rests on a definition of democracy formulated before the birth of modern constitutionalism, and this is an epistemological barrier to understanding constitutional and democratic modernity. If our purpose relates to the present – if (to paraphrase Michel Foucault commenting on Kant’s What Is Enlightenment?) our purpose is to understand what happens in democracy today – then we should turn the question around and wonder about the idea of democracy that is produced by the constitution-as-fundamental-rights and gives meaning to constitutionalism today.
The suggested hypothesis of this research is that the idea of the constitution-as-guarantee-of-fundamental-rights produces a kind of democracy with three distinct elements: an important political gap, deliberation, and individuality.
Constitutions and the Democratic Gap
In principle, the constitution-as-guarantee-of-rights constitutes a radical break in the image of the relations between governors and the governed: whereas the constitution-as-separation-of-powers produces a fusion between the representative body and the body of those represented, the constitution-as-guarantee-of-rights establishes a gap between the two. This differentiation is the logical result of checking up on constitutionality. The same scene reappears in every decision exercising constitutional review: enactments by the body of representatives – laws – are judged, for example by the French Constitutional Council, in their relation to the rights of the represented – the constitution. This means conceiving two distinct spaces – that of the representatives and that of the represented – with two potentially contradictory normative wills. In concrete terms, the constitution gradually defines a space that, through legally overturning a law, both symbolically and practically ensures the autonomy of the represented from the representatives. And this space grows and strengthens as new constitutional rights are “discovered” – the principle of respect for the dignity of the human person, the right to adequate housing, the freedom of marriage, the right to a normal family life, and so on. The logic of this gap was clearly applied in the Council’s decision of 16 January 1982,  in which it opposed the will of the elected representatives of the people to deny the constitutional importance of the right to property, a will contrary to the will of the people, who “in the referendum of 5 May 1946 rejected a bill of rights that included in particular a statement of principles different from those that had been proclaimed in 1789”, and on the other hand, “in the referendums of 13 October 1946 and 28 September 1958, approved of texts that recognized the constitutional authority of the principles and rights that had been proclaimed in 1789”.
Constitutional review thus produces a new image, an image of distance between the ruled and the rulers, by establishing the rights of the former in a separate body from the rights of the latter: the legal charter of constitutional rights and liberties symbolizes the space of the governed, while the laws symbolize the space of the governors. This image of a gap is profoundly different from the image of fusion that still prevails in our minds. In fact, in its most common understanding, the democratic ideal demands the people’s ever greater involvement in power – by the extension of universal suffrage, for example – and its full realization by the fusion of the people in the national representative political body. While the “people’s democracies” each with its single party may take this logic of fusion to its extreme, the “bourgeois democracies” also join in, with more moderation. Raymond Carré de Malberg gave us a perfect description of parliamentary regimes based on the identification of the governed with the governors, and on jumbling together the people and their representatives, thus the general will and the parliamentary will, putting parliament on a par with the sovereign, or rather, as Carré de Malberg expressed it, effectively erecting parliament as the sovereign. 
In spite of its democratic pretensions, in reality this kind of politics is only a transposed reproduction of the principle of monarchy, in which the body of the nation and that of the king are one and the same. As Louis XV declared in a speech to the Parlement de Paris on 3 March 1766, “the rights and interests of the nation, which some have dared to make into a body separate from the monarch, are necessarily united with my rights and interests, and rest in my hands alone; I will not accept the introduction of an imaginary body that could only disturb this harmony.”  The revolutionaries of 1789 certainly wanted and thought they had achieved the separation of these two bodies; indeed, they thought that the revolutionary act was precisely this bold assertion of the autonomy of the body of the nation and that of the King. However, in reality they reconstituted the unity of the bodies, simply giving the nation a new body to fuse with: the representative body. This constitutional continuity was probably dictated less by a doctrinal continuity than by the constraints of the political combat for legitimacy in 1789. Politically, the revolutionaries could not contest the unity of the body of the King by holding up the social diversity of the people. They had to assert the unity of the body of the people-nation; otherwise they risked weakening the already uncertain legitimacy of their claim to power. But by doing this, they committed the Revolution not to a path that broke with the Ancien Régime, but to one that simply “modernized” the political system of representing the nation. First, because the physical people disappeared, absorbed into and by the concept of the nation; next, and by logical necessity, because the nation, being an abstract entity, could express itself only through the intermediary of physical persons qualified to represent it. The upshot of this is the (con)fusion of the people and the nation, the (con)fusion of the representatives of the nation and the people-nation: since the people are the nation and the nation can express itself only through its representatives, there can be no expression of the will of the people other than that expressed by the representatives of the nation. This was clearly the claim of Siéyès when he bluntly asserted that “a people can speak and act only through its representatives”.
The constitution-as-guarantee-of-rights, with its call to check up on constitutionality, breaks up this fusion. Before the existence and development of constitutional law, representatives’ legislative activity was attributed directly to the people’s will, without the people being able to protest, because by constitutional definition there was no separate, independent way for them to have a will other than the one expressed by their representatives. With constitutional review, representatives are still qualified to express the will of the people, but the fusion of the two wills is no longer possible. Judges exercising constitutional review, by the charter of fundamental rights that they construct, and which defines the space for the autonomous representation of the people’s sovereignty, are always in a position to demonstrate (“in light of the Constitution” as is said in their decisions) and if necessary to take action against any gaps between constitutional requirements and interpretations of them in legislation passed by the representatives. By thus “demonstrating” that the two spaces may not coincide and that when they conflict the former takes precedence, judicial constitutional review forbids representatives to claim that they are the sovereign, and reveals their position as mere delegates who can always be reminded to respect the rights of the sovereign. In a way, judicial constitutional review reveals something that representation tries to hide: forgetting the people. In fact, even when asserting their existence, representation makes the people disappear, by laying down as a constitutional principle that the people can be present, as Siéyès said, only in the persons of the representatives. Judicial constitutional review brings the people back into existence as an autonomous and sovereign person, by representing representation. It demonstrates that representation is a scene in which two actors play two different roles: elected representatives are delegates of the sovereign, and the people are that sovereign.
In philosophy, this logic of the democratic gap can be seen in the debate between Heidegger and Cassirer in Davos in March 1929, about the idea of limits in Kant. In response to Heidegger’s argument that knowledge can only tumble poetically from being onto man, man being limited in his access to knowing, Cassirer reasoned that man’s finitude pointed to a gap between him and being, a gap that meant that knowledge of being is constructed through the mediation of language, culture and symbolic politics. Here too, where the general will is being (whether in the being of the king or that of the nation) and men are limited (“on their own, the people always will the good, but on their own, they do not always see it,” observes Rousseau ), it is left to the representatives of the nation to “reveal” the general will to men; or else the gap between the general will and men is such that they are limited to being able only to approach it through language. In the first case, we see democracy by acclamation; in the second, democracy by deliberation.
Constitutions and Deliberation
When it comes to portraying the mode of production of the general will, the constitution-as-guarantee-of-rights is again at the centre of a radical break: while the constitution-as-separation-of-powers promotes a monopoly in the production of the general will, the constitution-as-guarantee-of-rights gives rise to competition in formulating the general will. And this shift is clearly beholden to judicial constitutional review, since judges themselves set the rules by a new definition of law. For example, the French Constitutional Council, in its decision on 23 August 1985 , stated that “enacted law does not express the general will except insofar as it respects the Constitution”. This contrasts with the definition given in Article 6 of the Declaration of Rights of 1789: “Law is the expression of the general will.” This change in the definition of law – moving from the affirmative to the negative and introducing the constitutionality as a condition – also implies and legitimizes a change in the mode of manufacturing the general will.  The mode implied in Article 6 is fabrication of the law by Parliament alone; the representatives alone are qualified to express the general will, and the law expresses the general will by the mere fact of having been produced by them. The law’s status and normative validity do not depend on any other body; parliamentary sanction is enough to make the law the expression of the general will. In contrast, the new definition of law in the Council’s decision in 1985 implies the mode of fabricating the law by a plurality of actors – the governmental and parliamentary actor of course, but also the judicial actor. In fact, it follows from this new definition that the parliamentary “manufacturing” of the law is no longer sufficient to guarantee its normative validity; the law can claim to express the general will if and only if it respects the Constitution. In other words, only if the Constitutional Council judges that the text passed by Parliament does not infringe this or that constitutional right or principle; for if it does, i.e. if the text is judged to be contrary to the Constitution, it cannot express the general will, and therefore the status of being a law cannot be recognized in it.
This competitive mode of producing the general will involves deliberation as a necessary operative principle. Indeed, when the general will is not situated in a single institution, when the will of the representatives is not by itself the general will, when the general will is constructed by comparing the texts passed by the representatives to constitutional demands, the general will can only be the product of a deliberative process, an argumentative exchange between different actors in the manufacturing of laws. When this system is based on a single actor or assumes the identity of the body of the people and the body of the representatives, there is no need for deliberation or debate in order to construct the general will; it is in the body of the nation and since there is that identity, it is enough that the king or the representatives express themselves for their will to enjoy the status of being a law. However, when the bodies are separated and neither of them can claim to have sole possession of the general will, deliberation becomes a necessary part of the mode of producing laws.
This principle of deliberation compels us to return to an often forgotten property of constitutions: they are texts consisting of words. Jurists naturally tend to think that a constitution consists of norms and “hard” law, so that there is no need to deliberate to know what the constitution requires or forbids; all you have to do is read the constitution, and if necessary its preparatory work, in order to discover the norm, principle, or rule residing therein. In this view, an institution speaks authoritatively and the people acclaim. In contrast, by making deliberation an operative principle, a competitive system of formulating the general will takes on board the polysemy of legal words, and gives rise to a juridical politics of argumentation in order to determine – to construct – their normativeness. Because the norm is not in the legal words and because these words are polysemous, debate is necessary to select from all of the possible meanings that a given word can carry the one that will determine the norm.
By this deliberative activity, the competitive system for producing the general will has two characteristics that are (or at least can be called) “democratic”. The first of these is that the norm cannot be transformed into a fetish or an untouchable verity. Having come out of deliberation and a choice among several competitive meanings, it remains under discussion, and can change if further deliberations “recognize the rights” of other meanings that had not previously been favoured. This is what jurists coyly call a reversal of precedent! To this first characteristic, which makes the competitive system constantly open to the society, must be added the necessity of recognizing fundamental rights. When the general will is not issued by an authority that possesses this will within itself, and is constructed by deliberation among competing actors, one of the necessary conditions for the formation of the general will is that there be guarantees for the freedoms that allow that deliberation: the freedom of expression in order to defend the different meanings that a constitutional principle can have, the freedom to come and go in order to defend these differences everywhere, rights of defense in order to be protected, and freedom for demonstrations, association, pluralism, etc. Without these rights – some of them protecting persons in their private activities, others in their public activities, but each shaping and reinforcing the other – the principle of deliberation would be a dead letter. In the competitive system of formulating the general will, deliberation is an active principle only through the medium of the law, in particular through these fundamental rights with constitutional status, which define the rules for achieving deliberative activity.
Of course, there remains the constantly recurring issue of governing the judges. When judges exercising constitutional review intervene in this competitive system in order to say whether a bill passed by the legislature deserves to have the status of being a law, when they can refuse to accord this status to a bill passed by the representatives of the people, does that not transfer the real normative power from the elected representatives to the judges? Though it is often polemical, this criticism must be taken seriously, for it poses an alternative tht we must reflect upon: either the sovereign is regarded as being capable of producing the general will directly and with no mediation, or judges are meant to create, at their discretion and unconstrained, the meaning that will become a norm. Either the law is what the king – or the president, or the parliament – says it is, or it is what the judges say it is. We must get beyond this simplistic alternative. The sovereign’s word acquires depth and efficacy only in a complex relation between this word, written in words that appear in the constitution, and everyone who uses those words. It is only in this relation – and not in any unilateral, wilful and solitary action by one of the parties in this relation – that the meaning of constitutional statements gets constructed and the sovereign’s word becomes effective. In this complex interaction producing meaning, a judge exercising constitutional review is only one of the actors – the one who forces the others to argue for their reading of a given statement, to substantiate their claim that their interpretation is valid; and the one who subjects the relevance of the arguments to criticism, and in the decision recognizes the meaning of the constitutional statement that the deliberative exchange had reached at the moment of judicial intervention. At first sight, the meaning presents itself as an arbitrary creation of the judge – for example, the “creation” of the principle of the dignity of the human person by the Constitutional Council’s decision of 27 July 1994 . But well before that point, the formidable hermeneutical labour that prepares such constitutional “creations” has taken place: in legislative assemblies, in national and international courts, in learned committees and academies, in discussions of the legal doctrine, in political, trade union and societal associations, and also in the press. All this is not the expression of the will to power of judges delivering all alone the constitutional truth, absorbing and treating all the others as their oracles. Judges are only one link in the chain of argumentation. They intervene at a certain time to confirm by their decision the meaning of a constitutional statement, without however thereby halting the chain, which continues to grow, for the meaning that has been produced opens up, in assemblies, courts, and doctrinal discussions, new debates and reflections that can later produce a new interpretation. Judges exercising constitutional review do not govern; they are the regulating body in the competitive system of formulating the general will.
Constitutions and Societies of Individuals
Lastly, the constitution-as-guarantee-of-rights is at the centre of a radical break in the profile of the “res publica”: whereas the constitution-as-separation-of-powers has the state as its object, the project of the constitution-as-guarantee-of rights is a society of individuals. This shift is also a logical result of checking up on constitutionality: because judges exercising constitutional review handle cases related to the family (e.g. civil unions), workers (working hours), consumers (competition regulations), patients (healthcare provisions), students (universities), television viewers (broadcasting regulations), and all those on the receiving end of administrative decisions, these judges are in fact necessarily drawn into laying down what Georges Vedel called “the constitutional bases” of individuals’ social and private activities, and no longer just the constitutional bases of the activity of politicians. So the constitution is no longer the constitution of the state, but the constitution of the society, since all of the activities of individuals that are handled by the law can be related to the constitution; in legal parlance, this is reflected by various “constitutionalizations” – of civil law, labour law, corporate law, commercial law, administrative law, criminal law, and do on – i.e., by the idea that all branches of law, not just political law, find their principles in the constitution.
All things considered, the constitution being the act that (in the philosophical sense) informs society amounts to a break only with regard to the habit of thinking of the constitution as an act that organizes the public powers, leaving society to the civil code (which Jean Carbonnier said was the true constitution of France). Looking at the Declaration of Rights of 1789, this societal concept of the constitution is less a break than a continuity. Indeed, according to Article 16 of the Declaration, the object of the constitution is not the state, but the society: “any society in which the guarantee of rights is not assured and the separation of powers is not determined has no constitution”. Note: “any society”, not “any state”. When Montesquieu imagined the ideal constitution, he started with an analysis of society, of “social powers” – the nobility, the bourgeoisie, and so on – and he then looked for a structure of power that expressed the social structure; when Rousseau wrote his draft constitution for Corsica, he explicitly took as the basis and the object of his work the structure of Corsica’s social body. This concept of the constitution-as-expression-of-society faded away during the 19th century, replaced by the idea that a constitution was just the particular arrangements regarding the rulers; it logically reappears today with the emergence and development of a constitutional law that encourages applying the constitution to all social activities.
Indirectly but necessarily, this societal concept of constitutions has also influenced the application of the separation of powers. When the object of the constitution is reduced to the state, the separation required by Article 16 of the Declaration of 1789 applies only to the powers of the state: executive, legislative and judicial. But if the constitution has society as its domain, the requirement for separation applies to all of the powers at work in society: economic, media, religious, etc. So the constitution must address these powers, these third powers (to echo Siéyès on the third estate) which, though they are everything in the functioning of society, up to now do not figure in the constitutional order, but ought to. For examples, the constitution could take responsibility for the “fourth power” (the media) and set out principles designed to guarantee for citizens its independence in relation to political and economic powers; or it could give to a transformed Economic, Social and Environmental Council the means to encourage the participation of organized civil society in the formation of the general will.
By addressing itself to society, a constitution also addresses itself to the individuals who compose it and in this way it participates in the construction of their identity, in a particular moment of political history. In fact, today the political issue is not the issue of the individual, nor even of societies made up of fluid individuals (to borrow Pascal Michon’s expression ). No doubt, capitalism, with its mystique of individual interest, its devices for individualizing work contracts, and its right of property, has shaped this individualistic process, but socialism has not gone against it, since according to Marx himself the actual society of the future will be one “in which the free development of each is the condition for the free development of all”. Therefore while the social and historical process is that of a society of individuals, the political question is how to organize this social fluidity, how to make this fluidity coherent so it does not produce a chaotic society, how to find an instrument for constructing common things, for having some generality in this fluid society. However, today this question is unanswered. More accurately: the answers of the past no longer work. The god, the nation, the state, and the social classes that have given individuals a feeling of belonging together – in Christian, national or socio-professional communities – are no longer efficient producers of shared understanding among individuals. In this historical situation, constitutions, as they are renewed by constitutional review, can be the instrument common to individuals, in which they can recognize in their particularities and their own rhythms, but also in shared values, the common constitutional values that Habermas calls “constitutional patriotism”. Veritable magic mirrors, constitutions act as secular texts, as sets of shared principles, as places where modern disenchanted individuals can reconstruct a common identity. Worker, consumer, voter, parent, owner, believer, free thinker, and so on – all of these kinds and rhythms of life find themselves precipitated (in the chemical sense) into the heart of law, into the subject matter of law that individuals encounter in constitutions. Although individuals have become fluid, constitutions are these texts that stop them from drifting by giving them a fixed point where all their activities can be articulated. And also where they can be reflected upon, discussed, criticized, or judged. For each person’s image of the constitution contains more of desires and promises than it does of objectivity: equality between men and women, individual liberty, and fraternity (among other things) are the characteristics that are desired, dreamed of, and hoped for from law – characteristics that are denied in the everyday world of exclusion, inequalities, injustices and domination. And it is precisely from this gap between the constitutional promises of equality, liberty and solidarity and the misery of the world that there arise the possibility of criticism of this reality and the possibility of political action to change it. And that is why constitutional identity remains ahead of us, as something to come, something on the horizon of democracy’s demands.
Conclusion: Democracy as a Horizon
Gap instead of fusion, deliberation instead of revelation, society of individuals instead of state – these new characteristics distinguishing contemporary constitutionalism and transforming democracy’s configuration are obviously up for debate. Some see this “democracy by constitution” not as a more demanding democratic theory but as the return of the theological-political, with law as religion and judges as high priests . Others condemn a libertarian  or aristocratic concept of power that, through constitutional review, tries to keep the people out of politics . And some sociologists detect in the idea of constitutional democracy a will to power by jurists, in particular by professors of law serving a sophisticated mechanism for taking power away from citizens and giving it to judges, and for legitimizing this transfer of power .
These different views of the meaning of constitutionalism must be taken seriously. Firstly, because, to adapt Montaigne’s words, we can never be sure when we are in our own back room; next, because they demand more academic rigour in the presentation of this new constitutionalism; and finally, because they invite us to reflect on the reasons for these different ways of understanding this constitutionalism. Without claiming to identify all these reasons, there is one that appears clearly in all of these critiques: support for an essentialist definition of democracy. For it is when electoral democracy with universal suffrage is taken to be the sole principle of democratic legitimacy that “democracy by constitution” is seen as establishing a political regime that is libertarian, or aristocratic, or at any rate against the citizens.
This way of understanding the meaning of contemporary constitutionalism leads to accepting a historical form of democracy, and this prevents us from thinking about the present. The electoral form of democracy, a product of history, can be surpassed. It was in that spirit that in 1992 I suggested the idea of “continuous democracy”, by which I meant that democracy does not end with elections, that it is constructed by cooperating with the political rhythms, that this cooperation can take various forms including judicial ones, and that a constitution is the place where these different rhythms – electoral and non-electoral – can be given coherence and thus can make sense. I also meant, more modestly, that democracy remains a horizon!