Who,the people? Rethinking constituent power as praxis

Abstract

Modern thinking about democracy is largely governed by the concept of constituent power. Some versions of the concept of constituent power, however, remain haunted by the spectre of totalitarianism. In this article, I outline an alternative view of the identity of the people whose constituent power generates democratic authority. Broadly speaking, constituent power signifies the idea that all political authority, including that of the constitution, must find its source in some idea of ‘the people’, whose authority is never exhausted by constituted power. The deficiency I seek to address is that of asking who the people is to whom any claim of authority refers, while avoiding the pitfalls of totalitarianism. I show the most famous totalitarian view of constituent power – advanced by Carl Schmitt – to be not only politically unsavoury but also ontologically unjustified. To outline my alternative view, I draw on Jacques Derrida’s concept of just decisions to argue that the undecidable inaugurates collective responsibility by demanding a response. This suggests a view of ‘the people’ as a doing rather than a being. I conclude by showing how this avoids totalitarian views of popular sovereignty by demonstrating its congruency with Claude Lefort’s democratic theory as opposed to totalitarianism.

I Introduction

Modern thinking about democracy is largely governed by the concept of constituent power. Broadly speaking, constituent power is the idea that all political authority must find its source in some idea of ‘the people’. Surprisingly, however, too little attention has been devoted to the question of who this people as the source of all authority is. To put the question another way, if the concept of constituent power is meant to advance a view of the source of democratic legitimacy and holds that such legitimacy must always involve some reference to the people or the demos, it is often overlooked that the demos itself also raises a question of legitimacy. ¹ In this article, I advance a view of the identity of the people behind constituent power as one based on praxis, such that the people does not pre-exist the emergence of its constituent power. I argue that this account is better able to avoid some of the main pitfalls plaguing modern conceptions of constituent power, particularly the purported threat of totalitarianism exemplified by the view offered by Carl Schmitt.

Two influential accounts of constituent power famously do consider this question in more detail. The first, widely credited with popularising the distinction between constituent power and constituted power, is that given by Emmanuel Sieyès in his 1789 What Is the Third Estate? ² It is doubtful, however, whether Sieyès’ account, pivoted as it is on the concrete situation of 18th-century France, is easily generalisable to contexts beyond that of the Third Estate in France. The second conceptualisation of constituent power which considers the question of who the people is, is that offered by Carl Schmitt in his 1928 Constitutional Theory. Schmitt’s view, however, remains haunted by the spectre of totalitarianism. Schmitt theorises what he calls the constitution-making power as the pre-political ontological existence of the people displaying substantial homogeneity. The people then politically affirm its homogeneity through the constitution by drawing a line between friend and enemy, to the possible detriment of all those falling on the wrong side of that division. ³

For this reason, some scholars consider the concept of constituent power to be tainted by its affiliation with totalitarianism and therefore suggest discarding it altogether. ⁴ In this article, I advance an alternative conception of the demos behind constituent power. I begin by outlining why Sieyès view does not seem very helpful in theorising the identity of ‘the people’ today. Subsequently, I argue why Schmitt’s position is not only politically unsavoury but also philosophically untenable. Schmitt’s legal philosophy partially rests on his famous thesis that all concepts of modern state law are secularised theological concepts. ⁵ This allows him to deploy an ontology derived from Thomas Aquinas, by which he equivocates ‘the people’ with God as necessary existence and creator and legal order with the world as contingent existence and created, such that legal order is radically dependent on the people. ⁶ While his argument works insofar as it pertains to the contingent nature of legal order, Schmitt cannot justify why ‘the people’ should be thought of as having ascended the throne of God. Schmitt cannot justify, that is, why we must think of the people as necessary and pre-political existence. As a result, we end up with a view that stresses the need to decide, but which lacks a convincing account of the agent doing the deciding.

I then proceed to outline an alternative view of constituent power. The operative concept in my account is collective responsibility, which I interpret as a response being called for. I draw on Jacques Derrida’s conception of the undecidability of just decisions to argue that the ultimate absence of determinate ground implies that the only way for a demos to exist is for it to exist in its activity. Simply put, in his seminal 1992 essay Force of Law, Derrida argues that any just decision must go through the undecidable, for if a decision were to be based on decidable (i.e. fully determinate) ground, it is not a decision but a calculation. Something is undecidable if the grounds for answering without residue, such that no further justificatory questions could arise, are absent. Decisions of this kind, for Derrida, do not rest on constative statements, and therefore have no truth conditions. Instead, they rely on performativity. ⁷

This suggests that the demos itself is governed by performativity. My account of the demos behind constituent power is thus praxis-based in that it argues that the people is ultimately a doing, not a being, in that it is irreducible to a set of theoretical principles but only appears in its activity (deciding the undecidable). This furthermore means that the demos only exists insofar as its activity exists. The demos therefore does not have a prepolitical existence. Having outlined this view, I argue that it is consistent with another theorist who famously (though less explicitly) advanced a view of constituent power: Jean-Jacques Rousseau. ⁸ My view of constituent power confirms Bonnie Honig’s thesis that the paradox of politics faced by Rousseau in chapter 7 of book II of The Social Contract ⁹ is constitutive of politics itself, ¹⁰ adding that it is in fact simultaneously constitutive of the people and its constituent power. Finally, I conclude by showing how my alternative account manages to avoid the pitfalls of totalitarianism by demonstrating its congruence with Claude Lefort’s opposition between democracy and totalitarianism. ¹¹

Before proceeding, two further remarks are in order: one concerning the use of either the term ‘nation’ or that of ‘people’ to denote all those considered to be, in one way or another, part of some political community, and one concerning wider philosophical debates about how the people should be viewed democratically. First, throughout this article, I prefer the use of ‘people’ or occasionally ‘demos’ over ‘nation’. While it is true that historically, authors such as Sieyès, Rousseau and Renan generally defined the nation as a voluntary association, the term nation has also become associated with more ‘naturalised’ or even ‘ethnicised’ views of political communities. And while such views may certainly be criticised, to enter into a further debate as to the purportedly natural or artificial origins of the concept of nationhood is to enter into an academic minefield. I therefore opt to follow, among others, Arash Abizadeh and Étienne Balibar in rejecting the concept of ‘nation’ altogether. ¹² I instead opt to discuss the ‘people’, except when discussing (mostly historical) authors who themselves employed the word ‘nation’ rather than ‘people’.

Secondly, while it is true that questions regarding the demarcation of the demos in democratic theory have previously been addressed in a debate that has become known as the boundary problem, ¹³ my argument in this article is distinct for two reasons. First, and most simply, questions regarding the identity of the people have not been sufficiently addressed within debates on constituent power specifically. As such, the philosophical debate on the boundary problem addresses the question of how to properly demarcate the people without considering constituent power, whereas the debate on constituent power by and large does not address the question of whose constituent power is at stake, by and large. This generates a substantial difference between these two debates. Most importantly, the debate on the boundary problem generally considers the question of who is to be included in established decision-making processes and within the contexts of generally accepted procedures and institutional settings. What is at stake in the boundary problem is thus not the legitimacy of these institutions – or constitutions – as such, but rather the extent to which only one specific action within that institutional setting is legitimate by virtue of its level of inclusion. Constituent power, by contrast, inquires into the origin of the legitimacy and authority of such institutions more fundamentally.

Additionally, much of the literature on the boundary problem pays only very scant attention to (collective) political agency. As Bert van Roermund has argued, a substantial difference exists between ‘we, both’ and ‘we, together’, where the latter is the mode of intentionality that is presupposed by collective agency. ¹⁴ Take, for example, the all-affected interests principle proposed by Robert Goodin. ¹⁵ Goodin argues for an expansive account of who is to be included in any potential debate concerning any possible decision to be made. This model is then to be superimposed on extant institutional frameworks, such that it enfranchises all those potentially affected by any possible decision. But such a top-down approach is almost by definition unable to give rise to any form of collective political agency. Indeed, it seems to treat democratic participation entirely in the mode of ‘we, both’, whereas, as Andreas Kalyvas has demonstrated, the very etymology of constituent power presupposes a level of togetherness that models such as Goodin’s seem unable – or at best awkwardly able – to accommodate. ¹⁶ Hence, the extant debate on the boundary problem has very limited bearing on the concept of constituent power.

II Constituent power and the Third Estate

The distinction between constituent and constituted power is widely attributed to Emmanuel Joseph Sieyès. Although the notion of constituent power had been used prior to Sieyès’s seminal What Is The Third Estate? by, among others, George Lawson, it is unlikely that Sieyès was acutely aware of earlier (British) uses of the concept. ¹⁷ Constituent power’s almost mythical association with Sieyès, moreover, lends further credence to the view that even if he did not invent the notion, his writings at least gave it the political and philosophical relevance it has today. The situation Sieyès’s essay responds to is the pervasive inequality – political and otherwise – in 18th-century France. The Third Estate, he writes, includes everything required to be a nation, yet it remains largely subservient to the two dominant estates (the nobility and the clergy). He famously sums up the situation as follows:

What is the Third Estate? Everything.

What, until now, has it been in the political order? Nothing.

What does it want to be? Something. ¹⁸

The Third Estate is everything, firstly, because it is numerically the majority of the French nation and because it contains within it all services and roles a nation needs. ¹⁹

In a move clearly inspired by Rousseau ²⁰ and later to be echoed in 1882 by Ernest Renan, ²¹ Sieyès defines the nation as ‘a body of associates living under a common law, represented by the same legislature’. ²² Consequently, Sieyès not only argues for it to be impossible for the nation to be free unless the Third Estate is free, but for the Third Estate to be the nation, and the privileged nobility to be wholly excluded from it. The privileges accorded to the nobility in particular are incompatible with the required commonality of the law that is essential to his concept of a nation. ²³ As a direct consequence of the privileged position of the nobility, the Third Estate is reduced – at least in terms of its political existence – to something less than it actually is. Hence, Sieyès suggests, without the privileged order the Third Estate would be ‘[e]verything, but an everything that would be free and flourishing’. ²⁴ In sum, the entirety of the Third Estate is sufficient to form a nation, and rather than supply an addition to that nation the nobility in fact subtracts from it, as its privileges undermine the commonality and equality that should be at the basis of any voluntary political association.

This disadvantaged position of the Third Estate furthermore explains why, until now, it has been nothing. If the nation is such a voluntary association, and the nation is contained in the Third Estate, it has hitherto been nothing because it has not been able or permitted to wield any substantial political power. Instead, the exercise of power has been concentrated in the hands of the other two estates. This effectively creates a rift in the population of a territory since whoever is accorded privileges cannot be part of the common order. This is so because ‘a nation is made one by virtue of a common system of law’. ²⁵ Their privileges remove the nobility from the commonality that the law ought to have, and therefore from the nation altogether. Seeing as the system of governance, as identified by Sieyès in his time, was dominated almost entirely by the first and second estates, the Third Estate was effectively denied any meaningful political existence.

Sieyès’s emphasis on equality as the basis of the nation is consistent with his discussion of the Third Estate seeking to become something. If it is currently nothing due to its being denied political influence, and if one deserts the nation as soon as one is accorded privileges, what is required for a restitution of the Third Estate’s political rights is for its influence to be (at least) equal to that of the privileged orders. Given his assertion that the Third Estate is everything, Sieyès admits that this is a modest aim, but here he is drawing on what the Third Estate itself wants. ²⁶ Accordingly, the concrete demands raised by the Third Estate are that it should be represented by deputies from its own order (rather than the nobility), that it should have representatives equal in number to that of the other two orders and that votes should be counted by heads rather than by separate orders. ²⁷ While Sieyès suggests that these demands will be insufficient for the Third Estate to attain equal influence, these are at the very least what is necessary to become something. One might doubt whether the Third Estate can indeed be fully actualised such that it becomes the everything it in another sense already is or whether this constitutes a ‘fantastical congruency’. ²⁸ What matters presently, in any case, is the equality at its root necessitating its power to become at least equal to that of the other estates.

The status of the Third Estate finally comes to a head when Sieyès discusses the constitution, and specifically constitutional creation. This is the point at which the distinction between pouvoir constitué and pouvoir constituant comes into clear view. Constitutional creation for Sieyès amounts to the emergence of positive law out of natural law. As such, the constitution constitutes government but cannot constitute the nation. If the nation itself were constituted by the constitution, it could not constitute government by means of the constitution, for that would require the nation to be prior to itself. In other words, if the collective agent that Sieyès thinks the nation is were constituted by the constitution, by extension it could not exist prior to that constitution. And if that were the case, no collective agent exists that could constitute anything at all. This is the same problem faced by Rousseau in the seventh chapter of book II of The Social Contract, which Rousseau solves by recourse to his famous but no less divisive figure of the legislator or lawgiver. ²⁹ Sieyès’ solution, however, is different. While Rousseau thinks that there is no collective agent prior to the social contract, for Sieyès, the nation can only be a product of natural law. ³⁰

Unsurprisingly, when considering the theoretical origins of the nation, the model Sieyès proposes shows distinct similarities to that offered by the contractualists. Political societies, Sieyès argues, are formed along three epochs. At its basis, in the first epoch, a substantial number of isolated individuals seek to unify, and it is this very fact that makes a nation out of an aggregation of individuals. ³¹ Here, there is no common will yet, but a multitude of wills whose activity consists in the process of association itself. It is crucial to read this section carefully. The process of association is not what transforms a multitude into a nation for Sieyès; rather, it is the very intention of unifying that gives rise to the nation. This is important to point out here because, as we shall see, this means that Sieyès, unlike Schmitt, does not consider the nation to have a pre-political or ‘natural’ existence. It does mean, however, that consideration of the identity of the nation behind constituent power will have to hinge on this idea of voluntary association.

For present purposes, the second and third epochs are only relevant insofar as they relate to the first. Sieyès calls the second ‘government by proxy’, signalling the point at which for the simple reason of size exercising its will has one has become impractical for the community. As a consequence, they give over part of the collective will to be represented. This process of representation is deepened in the third epoch, in which a true common will no longer acts, but instead the general assembly is carried out only by a representative common will. These points should be fairly familiar to students of social contract theory. Three further remarks about the representation of the common will are in order here: first, political society never divests itself of the right to will. What it transfers is exclusively the exercise of its will. ³² Second, the nation does not transfer its will in its entirety, but only as much as is necessary to keep public order. Finally, and accordingly, the will expressed by representatives is incomplete and limited. ³³

These points are important because they help firmly establish the distinction between constituent power and constituted power. Constituted power is that which is constituted by the constitution, consisting of institutions and political bodies and organisations, which could neither fulfil their functions nor even exist absent a constitution. That is, concretely, they must have been constituted, ergo created, by something. That something is the nation, creating the constitution by exercising its constituent power. Constituent power, however, is inalienable. This means that the authority of the nation and the legitimacy of its decisions remain wholly independent of the constitution it has given its society. At no point, thus, does the nation become dependent on its constitution. At every turn, by contrast, are the constitution and the governmental bodies established by it dependent on the consent of the nation to persist in their existence. These points serve to differentiate Sieyès account of constituent power firmly, for example, from Hobbes’s social contract theory, as Hobbes proposes to think of the contracting parties (the citizens-to-be) as fully dissolving into the sovereign on agreeing on the contract, such that the will of the sovereign is, in a metaphysical sense, their will.

Now, the above account does in some sense appear to be an answer to the question of who the people wielding constituent power is. Indeed, Sieyès answer is that constituent power belongs to the nation, and the nation is simply the product of a sufficiently large group of people with the intention to associate. It is important to bear in mind, however, that, as Sieyès (and other social contract theorists with him) is undoubtedly aware, the description of the origin of nations he is offering is not so much a historical account as it is a rhetorical device. Just as the different accounts of the state of nature provided by Hobbes, Locke and Rousseau serve to justify the particular way in which they envisage the social contract, so Sieyès account helps him justify the unequivocal priority he allots the nation, in the form of the Third Estate, over the other two orders. Hence, he asserts that, should disagreements over the (already existing) constitution arise, these must be resolved by asking the nation, rather than the elite or the experts. ³⁴ Otherwise, notables would have taken primacy over the nation, even though at best they could have represented parts of the common will of the nation. Clearly, having recourse to any other agent than the nation as a whole if the constitution is in dispute is inconsistent with Sieyès view as outlined above.

As we have seen, thus, Sieyès uses his account of the origin of nations and of political society to justify the primacy of pouvoir constituant over pouvoir constitué. While that distinction has proven hugely influential, his account of whose power constituent power is tells us little beyond the opposition of the Third Estate in 18th-century France other than that it consists of a voluntary association. At this point, it is appropriate to say a little more about why we need a further account that goes beyond the regulative ideal of genuine consent at the bottom of such a voluntary association. As we have seen, the situation Sieyès’s essay concretely responds to is the pervasive political inequality between what we might also call the common people and the elites in 18th-century France. It is doubtful whether that situation is easily generalisable to the context of contemporary democracies. Because Sieyès’s essay was written to confront a concrete situation, the question of who the nation was, and accordingly whom constituent power belonged to, was one whose answer was given in advance: the Third Estate. The question this article asks, instead, is that of who the nation – or, more appropriately, the people or the demos – is in a more general sense. Given, that is, that the distinction between constituent power and constituted power has attained relevance that clearly goes beyond the opposition between the Third Estate and the other estates, who is constituent power to be ascribed to in a more general sense?

III From nothing to something: Constituent power and the ontology of change

As we shall shortly see, Carl Schmitt has an answer to the question of who constituent power is to be ascribed to that is easier to generalise beyond one specific situation. Before moving on to discuss Schmitt, however, I want to take a brief detour through an ontology of change. This is relevant because it shows, first, how Sieyès and Schmitt, despite all their similarities, ultimately differ in the fundamental structure of their thinking. Concretely, while Sieyès’s account of the Third Estate shows some distinctively Aristotelian characteristics, the ontology underpinning Schmitt’s constitutional theory is Aquinian in its origins. The most fundamental difference between these two ontologies consists in how they conceive of the relation between potentiality and actuality. Secondly, this brief recourse to ontology is important because it lays the groundwork that allows me to subsequently show that Schmitt makes a theoretical move that, based on his own metaphysical commitments, he is unable to justify.

My discussion here will necessarily be brief and takes its cue from Hans Lindahl’s account of the ontology of change that underpins constituent power. ³⁵ Lindahl’s main objective is to point out how a Schmittian view of constituent power needs Aquinian rather than Aristotelian ontology. In his Metaphysics, Aristotle asserts that ‘things that come to be […] all come to be both as a result of something and of or from something, and they all come to be something’. ³⁶ What this means is that two types of origins underlie every being. First, everything comes to be as a result of something, which indicates its cause. Second, all comes to be of or from something, the most concrete instance of which is its material. The typical example used to explain this relation is that of a carpenter making a chair out of wood. The chair, concretely, has two beings underlying its existence: the carpenter making the chair and the wood from which it is made. Metaphysically, what this means is that the potentiality to be a chair was already encapsulated by the actuality of the wood, and hence that ‘the actual exhausts the possible’. ³⁷

This matters because it implies that on this view, all potentiality is already present in actuality, and hence what it requires is simply for it to be activated or formed. In Lindahl’s words, change for Aristotle amounts to ‘the determination of a determinable’. ³⁸ Heidegger famously criticised this ontological model, arguing that all potentiality amounts to in this model is a deficient mode of actuality and that possibility here is no more than the actualisable waiting to be actualised. ³⁹ Let’s briefly pause to go further into how this ontological model is operative in Sieyès account of the Third Estate. I believe this is most clearly apparent in the seeming ambivalence with which Sieyès treats the metaphysical status of the Third Estate or the people. While Sieyès appears to believe that the existence of the Third Estate is not a natural fact but dependent on voluntary association, he also seems to superimpose a concrete and rather unified view of the Third Estate onto the specific reality he was confronting. In other words, while Sieyès emphasises voluntary association, he simultaneously imposes a preconceived view of the people onto the Third Estate. This seeming ambivalence, moreover, is reflected in some later commentary on Sieyès’ work. Istvan Hont, for one, highlights the extent to which Sieyès superimposed a unified view of the Third Estate on the much more diversified and diffuse reality of 18th-century France. ⁴⁰ By contrast, the reading offered by Nicholas Xenos, stressing the explicit construction of the nation, at first glimpse seems notably different: while the former emphasises the imposition of a preconceived unity, the latter maintains that such unity was, at the time of Sieyès’s writing, in the process of being constructed. ⁴¹

While some of these differences are undoubtedly due to the fact that Sieyès’s What Is the Third Estate? is as much a political pamphlet – a call to action – as it is a philosophical essay – an investigation of theoretical principles – I believe the Aristotelian ontology just outlined goes some way towards explaining how these seemingly opposed views are in fact not mutually exclusive. The unified view Sieyès superimposes is then concretely the potentiality of the nation, whereas the construction of the nation which nonetheless must be engaged in is reflexive of the need for this potentiality to be actualised. This point is furthermore supported by Illan rua Wall remark concerning Sieyès’s predetermination of sovereignty. Given that Sieyès, in Wall’s reading, entertains the notion of a predetermined sphere of sovereignty, the Third Estate’s actualisation is concretely aimed at reaching that sphere. ⁴² While Xenos is thus correct in highlighting the element of construction in any type of peoplehood, Sieyès’s account remains tied to an Aristotelian ontology in which such construction is aimed at a determinate conclusion: the actualisation of what was already potential.

Thomas Aquinas’s thought about God brings about a substantive shift in the philosophy of change. The difference is clearest in chapter 12 of book II of Aquinas’s Summa contra Gentiles. Aquinas wonders whether something can be said of God in a relative sense, meaning in relation to that which proceeds from it. His answer is in the affirmative, and the relation is one of radical dependency. The distinction between God and the world is posited by Aquinas as one between necessity and contingency, such that God, as necessary being, created the world even though it lay perfectly well within his powers to not have created the world. Hence, God here is a necessary being and the creator of contingent being, that is, the world. This is, furthermore, the precise distinction that Schmitt’s famous thesis that all concepts of modern state law are secularised theological concepts is hinged on. ⁴³ As we shall see, Schmitt himself takes this distinction and transposes it onto the people (necessary being, creator) and the constitution (contingent being, created and therefore dependent), respectively.

First, consider how Aquinas here reconfigures the relation between potentiality and actuality. Having, as I pointed out, argued that something can be said of God relative to his creation, Aquinas now asks whether this relation itself exists within God metaphysically. This cannot be, Aquinas thinks, for it would undermine the necessity and independence of God. Recall that the relation between creator and created is one of dependence. If this sense of dependence existed within God, ‘it would follow that God’s substance would depend on something else extrinsic to it’. ⁴⁴ The difference between this ontology of change and that offered by Aristotle should not be underestimated. Whereas for Aristotle actuality encapsulates all potentiality, the continuous link between the two is severed by Aquinas, for which reason the possible is no longer already present in the actual. What this means is that the possible becomes independent of what already is, which opens up the means for thinking possibility much more radically, in the sense that what is possible does not yet exist in what is actual in any way. Therefore, in the words of Alison Stone, ‘genuine possibilities can come about only through my free action […], reaching beyond everything that is already given’. ⁴⁵ Stone thinks of this potentiality on an individual level, but this point inaugurates a way of thinking about agency that ultimately has much more radical and political implications.

Consider, before moving on to Schmitt, how Sieyès’s account of The Third Estate displays traces of an Aristotelian ontology. To see this, it suffices to look back at Sieyès famous opening lines quoted above. Here, he states that the Third Estate is simultaneously everything and nothing. This is not a contradiction, but corresponds to two modalities of being. The Third Estate, that is, is potentially everything. In the actuality, Sieyès confronts, however, he sees that it is nothing. What is deficient in the Third Estate is hence nothing but its actualisation in the political sphere. What this means, concretely, is that a substantive segment of the political is removed from the scene from the outset. If, in a more general sense, the people already is what it is, but only needs actualisation, what can in any case not come about through its political action is a concrete challenge to who the people is in the first place. To be sure, from Sieyès’s point of view, this need not matter. After all, his description of a concrete situation ascertained the availability of a determinate and circumscribed people. But for a democratic theory, more generally, it does matter because it means that the question of who the people is has been foreclosed before the people has even become politically active.

IV From necessity to contingency: Schmitt’s adoption of theology

As I have briefly alluded to, Schmitt adopts the aforementioned Aquinian ontology and transposes it onto his model of constitutionalism. Although Schmitt frequently alludes to and approves of Sieyès, a brief survey of the underlying ontology reveals that at its core, Schmitt’s constitutional thought is different from that offered by Sieyès. Like Sieyès, the people for Schmitt exists in a manner radically independent of the constitution. The constitution is merely the expression of the existence of the political community, ⁴⁶ and therefore the people must pre-exist the constitution. This means that the constitution is essentially a dependent concept, existing only as expression of the ‘unified whole’ of the people which is prior to it. ⁴⁷ Schmitt views this unified whole as ‘substantial homogeneity’. ⁴⁸ Note that Schmitt’s use of the word ‘substantial’ here is to be taken literally, in a metaphysical sense: it signifies that all those who are part of the people partake of some common substance. The presupposition of this common substance grounds Schmitt’s view of the pre-political existence of the people as an ontological given. The people then need to deny and ultimately eliminate internal differences. ⁴⁹ Indeed, this is Schmitt’s famous – or infamous – thesis that the essence of the political ultimately comes down to drawing a rigorous distinction between friend and enemy. Accordingly, the political existence of the people is pivoted on its ability to draw this distinction, ⁵⁰ for which reason Schmitt points out that ‘humanity’ is not a political concept because it does not allow for such distinctions. ⁵¹

In terms of the dynamic between constituted and constituent power, this means that the people ultimately cannot be bound by its own constitution. Indeed, Schmitt goes further than that: the people cannot be limited by a posteriori curbs embedded in the constitution because that would imply a people that is simultaneously sovereign and subject to constraints, which is incoherent; the people also could not be constrained by a priori limits because this would suppress the political component of law and curb democracy. ⁵² Now, for the purposes of this article, there are two essential elements in Schmitt’s constitutional theory: the existence of the people as its normative underpinning and the decision it gives itself, and which acquires normative force based on its own existence. It is important to note here that Schmitt’s decisionism thus hinges on what one might call his ‘existentialism’: absent the real existence of the people, Schmitt cannot justify his decisionism.

Concretely, the decision the people gives itself is that to express its own existence through a particular medium – the constitution, or more specifically the absolute concept of the constitution. ⁵³ This is congruent with Schmitt’s definition of the state, which cannot exist prior to the constitution, as ‘a specific entity of a people’. ⁵⁴ Now, the important point here is that Schmitt raises the ontological status of the people to one of necessity. This is perhaps surprising, given the fact that his emphasis on the need to actively draw the distinction between friend and enemy seems to suggest that he thinks of both terms as constructed rather than as pre-political. Note first, however, that what depends on this construction of an enemy is specifically the political existence of the people. Concretely, it is this distinction that inaugurates the possibility of a constitution, not the possibility of a people.

Note further that this point sheds further light on the trajectory from existence to positive law. When discussing the distinction between the constitution as such and constitutional laws, Schmitt notes that the possibility of constitutional laws presupposes the existence of a constitution. ⁵⁵ While this point may not be highly original, its importance lies in the fact that Schmitt takes this type of relation one step further: just like constitutional laws presuppose the constitution, the constitution presupposes the people purely as existence. This is belied by the fact that Schmitt repeatedly states that pure existence is the ultimate normative ground for the constitution, as ‘what exists as political power has value because it exists’. ⁵⁶

It is true that this particular point has been interpreted, most prominently by Chantal Mouffe, to argue that Schmitt’s model seems to imply the simultaneous possibility to deconstruct all instances of the people. ⁵⁷ At first glimpse, this point seems to be belied by the idea that Schmitt’s concept of collective political identity is relational in that it requires, as a criterion, an Other as a constitutive outside. ⁵⁸ But it is worth remembering, in this context, Schmitt’s repeated emphasis on the absolute sovereignty of the people. If Schmitt views the people as relationally constructed, he loses the ultimate reason to rule out any prior normativity, since the people is in that case no longer truly sovereign anyway. The important distinction, I believe, lies in the construction specifically of the people’s political existence. To draw the friend–enemy distinction is to exist politically. But if the people are to be absolutely sovereign, and therefore independent, its existence per se cannot depend on this distinction, and hence, it must exist in some other sense prior to the political.

I believe that Schmitt’s repeated emphasis on the absolute independence of the people from all normativity rests on his conceptualisation of the people as sovereign in a way that derives ultimately from a theological concept of sovereignty. ⁵⁹ For present purposes, this has two important implications. First, positing the people as absolutely sovereign enables Schmitt to theorise the constitution as absolutely dependent, which in turn enables his critique of liberalism and allows him to justify his view of the primacy of the people over everything. Ultimately, thus, this affirms the radical contingency of the constitution. Second, this immediately binds Schmitt himself in theorising the people, since this forces him to presuppose the existence of the people as necessary (in a metaphysical sense) and therefore pre-political. For to theorise the existence of the people as anything other than necessary and pre-political is effectively to have its existence depend on something outside itself. If that were the case, the people could no longer be sovereign.

This is fundamentally the ontological model that Schmitt adopts from Aquinas. For Aquinas, God as the sovereign creator of the world necessarily existed, and the world as the created was rendered radically contingent. For Schmitt, analogously, the people as the sovereign creator of legal order, of the constitution, necessarily exists, and the constitution by contrast becomes radically dependent and contingent. Mapping the concepts of ‘the people’ and ‘the constitution’ onto ‘God’ and ‘the world’, respectively, allows Schmitt to justify the superiority of the people over the constitution, and thereby assert the ultimate contingency of any legal order. This move thus underpins Schmitt’s normative priority of constituent power over constituted power. It asserts that a decision undertaken by the people, whose existence itself justifies that decision, is ultimately the mechanism with which constituent power constitutes legal order. But the very possibility of theorising constituent power along these lines is now hinged on Schmitt’s ability to prove that the people’s existence is indeed necessary and pre-political.

Let us set aside medieval debates as to the existence of God and presume that existence is a necessary predicate of God as a perfect being. The task for Schmitt, if he wants to appropriate the Aquinian ontological model, is to show why the same metaphysical status must be ascribed to the people. Schmitt fails to do so: he only assumes the necessary and pre-political existence of the people because doing otherwise would undermine his own argument as to the sovereignty of the people. Hence, while Schmitt convincingly shows that legal order, if it depends on a decision or on constituent power, might not have existed at all, he fails to convincingly demonstrate that the people’s existence is necessary. The people, that is, might not have existed in the first place. To put it in Bill Scheuerman’s terms, if Schmitt tends to collapse normative questions into existential ones, reducing every ‘ought’ to an ‘is’, ⁶⁰ his entire framework becomes endangered when the ‘is’ itself is put into question. Hence, while Schmitt’s view of the identity of the people behind constituent power, widely recognised to be at the basis of Schmitt’s worrisome political commitments, may be politically unsavoury, his account also turns out to be philosophically untenable.

This point throws Schmitt’s model into relief, for if there is no agent to decide a decision whose normative force is underpinned by the existence of that agent, it suddenly becomes hard to see how any such decision could retain any normative force at all. The image we are left with here is one in which constituent power consists in a decision but without anyone to do the deciding. One might at this stage be tempted to discard the element of decision altogether. Since the decision itself can no longer be given its normative force according to the Schmittian model, it seems pointless to maintain the concept to be at the heart of constituent power. Indeed, we seem to have ended up in a stalemate where legitimate constitutional order requires a decision, but that decision itself seems to no longer allow for justification. I want to suggest otherwise. In what follows, I argue that the clue to this riddle lies in the element of decision itself and the apparent impossibility, in the absence of such an ontological concept of the people as Schmitt proposes, of providing a definitive ground for that decision. This point addresses the issue of how to conceptualise the people on whose supposed authority the constitution is legitimised. Rather than being prior to the political decision that creates the constitution, I will argue that the people and its decision are mutually constitutive. The operative concepts here are responsibility and phronesis. I interpret these concepts to mean that responsibility indicates concretely a response being called for, or demanded, and that this inaugurates a kind of practice which is revelatory of the people such that the people is performative, and therefore irreducible to an abstract set of theoretical principles. ⁶¹

V The paradox of politics and just decisionism: Inaugurating the people

If the people do not pre-exist politics or the constitution, let me begin this section by noting that the theoretical problem we are facing here is that famously confronted by Jean-Jacques Rousseau in the seventh chapter of book II of The Social Contract. How is it possible, a perplexed Rousseau asks, that an aggregation of individuals without any further expertise on the matter should be able to perform a task so momentous as to give itself a constitution? After all, the surest way to ascertain the existence of good laws is to have them be created by a virtuous and well-educated citizenry. What, then, is necessary for the people behind constituent power to be able to create good laws? They seem to only become so by virtue of the previous existence of good laws. Hence, the problem at hand is what Bonnie Honig calls the constitutive paradox of politics. ⁶² Its paradoxical nature consists concretely in the fact that ‘before the existence of laws, men would have to be what the laws have made them’. ⁶³

At this stage, Rousseau famously introduces the legislator who offers or frames the laws to the people-to-be. But this does not solve the problem: for how is the people to know they are not being deceived and the legislator is not a charlatan? Rather than leading his readers into an infinite regress, however, Honig suggests that Rousseau is shifting the problem to a slightly different register. The offer made by the legislator may be appropriate or not, but at this level, this is no longer what matters. What matters is that a choice is forced upon the people, which inaugurates collective responsibility in that this forces a decision and hence a response is demanded. The very fact of being put before a decision, then, propels a collection of agents into collective willing and as such first reveals the contours of ‘the people’ which is exercising its constituent power. It is the concrete structure of this decisioning, then, in which I identify the locus of the people.

The concept of ‘decision’ is at the heart of a famous essay by Derrida which is as influential as it is, for some, ‘troubling’. ⁶⁴ The question he asks is that of when a decision can be said to be just. This is not, to be sure, to expound an entire theory of justice, but rather to ask what is required of something for the attribute of justice (or injustice!) to be applicable to it in the first place. Derrida’s answer seems simple yet confusing: it must be a decision. This point hinges on his distinction between decisions and calculations. If it is clear from the outset what the answer to some deliberation or hesitation is going to be, Derrida argues, what we have before us is not a decision but a computational calculation. ⁶⁵ There is a fundamental difference, that is, between a decision and a calculation. Suppose, to illustrate the distinction, I am taking part in a quiz, and the host asks me whether the highest mountain in Scotland is Ben Nevis or Ben Macdui. Suppose, furthermore, that I do not know that Ben Nevis is the highest mountain in Scotland. I am forced to decide which answer I think is the one that corresponds to the truth. Now, whatever answer I elect to give, the choice itself is not one to which we would normally attribute justice. The reason for this, Derrida thinks, is that the question before me ultimately allows for a definitively correct answer: Ben Nevis is the highest mountain in Scotland.

Now suppose that I am confronted with a profound moral dilemma. The specificities of the dilemma are not important here. What matters, rather, is that I am again put before a choice and must decide. Surely, I may have some moral principles I adhere to which guide me in my decision. But unlike knowing whether Ben Nevis is indeed the highest mountain in Scotland, my moral principles ultimately do not permit any definitive truth. The dilemma, that is, is undecidable, and ‘the undecidable […] must be gone through by any decision worthy of the name’. ⁶⁶ The undecidable goes further than simply not knowing what the right decision is: for something to be undecidable is for it to disavow of any definitive answers. Decidability, by contrast, enables conformity. It offers guarantees as to the outcome of the question at hand. The upshot of this is that decidability does not permit decisions in the true sense of the word. The problem, in short, is that justice is a powerful political demand or even ‘an essential quality of a well-ordered society, but […] is open to many conflicting definitions’, such that it is always unclear what is meant when one calls for justice and why any specific conception of justice should be considered the ‘correct’ one. ⁶⁷ For Derrida, this feature is constitutive of justice inasmuch as anything about which such doubts could not arise may be many things but in any case cannot be called ‘just’. Only when a decision is undecidable can we assess whether a decision in the true sense of the word is just. Something is thus not undecidable because the right decision is unknown, but because there is no decisively and determinately right decision; because, at the end of every justificatory chain, we can still probe further into the justification of a decision or the justification of a justification.

Note that this does not commit Derrida to a kind of radical relativism. Rather, as Simon Glendinning points out, Derrida navigates between what Glendinning calls ‘platonism’ and ‘conventionalism’. ⁶⁸ A conventionalist would commit to the view that what is right depends on a deductive structure of legitimisation. Hence, to take Hans Kelsen’s legal positivism as a clear example, what is right (or, in Kelsen’s legal philosophy, what is lawful and legitimate) depends on whether the decision was authorised by a higher norm, which norm was authorised by a still higher norm, such that we ultimately end up with the necessary presupposition of the Grundnorm. ⁶⁹ This will not do for Derrida, for ‘when we hit the ‘final convention’ we can still ask where its force comes from’. ⁷⁰ A conventionalist approach that defines justness in accordance with some first principle, hence, cannot definitively shut down further justificatory questions. Nor does Derrida commit to the “Platonist’ alternative, for he disavows the possibility of justice permitting an ultimate or natural(ised) truth. In the final analysis, for Derrida, the question at hand goes beyond the simple opposition between convention and nature. ⁷¹ Law and justice, for Derrida, exist in an unceasing oscillation between undecidability and the need to decide. This does not mean that decisions occur in a vacuum. Much to the contrary, decisions must be ‘neither pure improvisation nor pure conformity’, ⁷² following principles or assumptions but in the process approve or disapprove of them, confirm or disavow them and reinstitute them. ⁷³ Judgement, then, is iterative. ⁷⁴

The need for judgement, or for a decision, is then what inaugurates the responsibility that lies at the heart of the concept of ‘the people’ proposed here. It does so because responsibility here is interpreted as a response being demanded, which forces those of whom a response is demanded to address the specificity of the question before them. Together with the undecidable, this disavows of the possibility to reductively anchor both ‘the people’, its responsibility, and that to which it responds to an abstract set of theoretical principles. Only such theoretical principles would enable one to decisively and determinately identify ‘who’ the people is in abstracto, that is, abstracted from any concrete situation in which the people becomes apparent. Instead, I propose here to view the people whose constituent power is under discussion as a performative activity, rather than a more or less static being. For the present, this has the following three elements and implications, each of which will be elaborated upon below. First, this means that the people are first of all identifiable by its activity or praxis, rather than by a determinate ontologised ground. Put differently, this suggests that the collective subject of the people should be seen not so much in terms of a pre – given identity, but in terms of more or less active identification. ⁷⁵ Following from that, second, this means that the people is in each instance particular and singular, corresponding to the situation which demands its response. Ultimately, third, this suggests that the people are irreducible to any entity or principle which definitively shuts down any further politicisation and probing into the people’s own grounds or, since ‘politicisation is interminable even if it cannot and should not ever be total’. ⁷⁶ The answer to ‘who’ the people of constituent power is, thus, cannot ever be given in advance but only reveals itself in the performance of the people, and even then it can only ever be a provisional answer.

VI Performing the people

In this final section, I briefly elaborate upon the three elements of my new account of the people as outlined above. As we have seen, Schmitt could not justify his ontological presupposition of the necessary pre-political existence of the people. His emphasis on the element of a decision by which the people decided to give itself a concrete political existence, mediated by a constitution, nonetheless, is useful. Following Derrida’s conceptualisation of the structure of decisions, I have argued that such decisions are first introduced by the confrontation with an undecidable question. The responsibility this confrontation gives rise to is what first inaugurates the people. Now, this suggests that the people cannot be reduced to an abstract principle, ontological or otherwise, but that it can only be identified in this confrontation and its response to it. The people itself, that is, must go through the undecidable.

Derrida puts this, as he is wont to do, in terms of the contrast between constative utterances and performativity. First introduced by John Austin, constative utterances are all those that profess to convey some fact about the world. As such they may be true or false, or their veracity may be unknown, but in principle, it is knowable. Performativity, by contrast, cannot be assessed in terms of the truthfulness of a statement, but rather brings about some change in the world. ⁷⁷ Thus judgements and decisioning in Derrida’s sense are performative. It is in these terms that the people behind constituent power must be viewed. ‘Being’ the people is not predetermined by abstract ontological principles, nor by concrete legal means such as formal citizenship. If democratic peoplehood is about political agency, such assessments of pre-political grounds can only ever fail to justify why some agents are supposedly capable of acting politically while others are not. Rather, ‘being’ the people is about ‘doing’ the people; about ‘staging’ the people; ⁷⁸ about performing the people. In this way, performativity is immediately revelatory of the people and in a sense self-fulfilling: to raise a claim of ‘we, the people’ is immediately to be the people, regardless of previous authorisations to be the people. ⁷⁹

Theorising the people as praxis is to centralise the element of its activity and agency and to disavow of the possibility of universalising the people based on some underlying principles. A convenient way of thinking about this is in terms of what Aristotle in the Nicomachean Ethics calls phronesis or practical wisdom. Aristotle contrasts phronesis to scientific knowledge, which concerns the discovery of things that are and, crucially, could not be otherwise. Scientific knowledge, in other words, is concerned with necessities and known or knowable truths. But since the first principles of phronesis ‘might be otherwise’, they cannot be demonstrably proven in the same way as those of scientific knowledge, and hence ‘practical wisdom cannot be scientific knowledge […] because what is done can be otherwise’. ⁸⁰ Instead, phronesis is a virtue ‘concerned with action in relation to human goods’, ⁸¹ which always appears in its singularity and particularity because what is good is always good for specific agents in a particular time and place. ⁸² It thus disavows the possibility of anchoring it determinately to any central principle from which it must be derived by necessity since its principle can always be otherwise. The appropriate principle is only revealed in acting towards some human good and based on practical wisdom.

While Aristotle speaks of phronesis in terms of the practical wisdom exhibited by persons individually, there is no reason why this principle could not be extended to collective action. Indeed, it is precisely this kind of deliberation and contestation that matters to undecidable decisions: what it asks is not what is true, for it cannot be definitively ascertained what is true, but what is good. And the question of what is good is irrevocably tied to what is good in the here and now, and what matters to those confronted with the question. It therefore specifies the demos always in its particularity and singularity, because the good in search of which it is revealed ‘is always situated’ because ‘it is good for us’. ⁸³ The principle of law-creation – which is what constituent power ultimately aims at constituting – should accordingly not be seen as transcendent, independent of the act of constitution, but as immanent, which is to say intrinsically connected to the practice of constituting. ⁸⁴ Hence, the fundamental question that confronts the demos and reveals it as demos is not that of what is true. It is rather closer to that other pertinent question, famously asked before by Chernyshevsky, Tolstoy and Lenin: what is to be done? And that question does not transcend the instance of its being asked but is immanent in it. But even this question must be slightly modified in light of the inevitable singularity and particularity of the situation in which it is asked, and therefore the concomitant particularity of the (collective) agent whom it confronts. The question at stake for the demos-as-demos is specifically ‘what must we do now?’, where the construction of the ‘we’ is involved in the posing of the question and the considering of answers.

Finally, if the people and its constituent power are co-constitutive, this seems to imply that an answer is given as soon as the question is posed. But as the above discussion of phronesis affirms, this can only ever be a provisional answer. Any answer to the question of who the people is, to use Aristotle’s words, ‘can be otherwise’; the veracity of any answer is not impossible to ascertain because the true answer is unknown, but rather because there is no definitively true answer. Every answer is susceptible to further questioning, politicisation and requires ongoing justification, which in each instance may call forward a different circumscription of ‘the people’. As such, it should be clear how this view disavows the possibility of absolutising any particular view of ‘the people’. If constituent power is accordingly viewed as a new beginning, its story begins as Bonnie Honig has noted not with the words ‘in the beginning’, but with ‘once upon a time’, suggesting that this beginning is only one particular beginning rather than the beginning. ⁸⁵ The beginning or the origin of the people in concomitance with legal validity is thus, as Carlo Accetti has phrased it, one of ‘always-already’, ⁸⁶ disavowing of the politico-theological idea of an ultimate sovereign beginning prior to which there was nothing. No story begins from a vacuum, and accordingly no constituent power constitutes ex nihilo, because there was always already some normativity which it could reject or critically affirm.

This does not necessarily invite a permanent instability that never crystallises into some reasonably stable, legalised reality. Rather, I would suggest that it is exactly this kind of view of the people of constituent power that is appropriate to a Lefortian view of democracy. Lefort’s democratic theory is relevant here insofar as he constructs his account in opposition to his concept of totalitarianism, which is not explicitly but nonetheless quite clearly construed in accordance with the Schmittian view discussed above. This becomes especially clear in Lefort’s characterisation of totalitarianism as signalled by the denial of division internal to society. ⁸⁷ In so doing, Lefort’s conception of democracy manages to capture one of the main worries animating critics of Schmitt, and by extension of constituent power. ⁸⁸ Showing how the view outlined above is consistent with Lefort’s political thought will therefore serve to pre-empt criticism on the basis of constituent power’s alleged tendency towards totalitarianism or, at the very least, significantly ease those concerns.

Famously, the crux of Lefort’s democratic theory is the empty place of power. What this means concretely is that there is no substantial and fixed entity to which all principles of legitimacy can refer and in which they can be grounded. Power, in other words, has become disembodied. ⁸⁹ What grounds power remains a reference to the people, which can be construed as a reference to constituent power, but the crucial thing to note is that the people ‘can never be grasped in their ultimate definition’ because no ultimate definition of the people that is democratic could exist. ⁹⁰ At issue for Lefort is the question of the form of society, concretely ‘the space of relations in which the people is meant to live and coexist’. ⁹¹ The form of society, or its mise en forme as Lefort is wont to put it, that is specific to democracy is one of indeterminacy. The demise of absolutist monarchies, in which society was given a concrete form by its embodiment in the monarch, calls for a new model. While totalitarianism fully determines the form of society, ⁹² democracy, by leaving the place of power empty, specifies a mode of coexistence in which the terms of that existence are not pre-assigned their particular positions, but rather interact in a way enabling them to persistently question the legitimacy of power. ⁹³

Thus, democratic legitimacy, whether the legitimacy of institutions – of constituted power – or that of constituent power itself, ‘is no longer referred to an unconditional pole’. ⁹⁴ In the absence of determinate ground, democratic legitimacy for Lefort is a permanent question, traced by ‘the legitimacy of a debate as to what is legitimate and what is not’. ⁹⁵ This very groundlessness constitutes the ground of democracy, as it maintains the essential openness that ensures that further justificatory questions can always be posed – and this is true for ‘the people’ as much as it is for any institution. Accordingly, democratic legitimacy must in some way refer to a concept of the people, and thus remains fundamentally tied to constituent power as the ‘truth’ of democracy, ⁹⁶ but this does not mean that its legitimising principle is not similarly subject to questions of its legitimacy. And absent a determinate ground in which to anchor any such legitimisation definitively and determinately, legitimacy must be viewed as an ongoing process of questioning and justification: a phronesis by which the contours of the people are revealed.

This then suggests that ‘the people’ itself is undecidable. Faced with the responsibility to address the issue before it, a multitude of agents act in certain ways that reveal the contours of a demos, but in an always unpredictable way. Because the people as such can never be known in advance, it is senseless to seek to assess who the people is with a determinate manner. That is, the question of ‘who’ the people are is one devoid of truth conditions. No determinate circumscription of the people could ever be made that is, first, invulnerable to further politicisation and justificatory questions (why is the line between inclusion and exclusion drawn such-and-such, and not so-and-so), and second, no account of the people exists that is without residue, or as Rancière would put it without supplement. ⁹⁷ Every construction of ‘the people’ is explicitly political, not only in the sense of never being definitive and therefore deconstructible, but also in terms of its being always incomplete without possibly reaching completeness. ⁹⁸ Its answer, however provisional, is contained in its performance. Accordingly, this point entails a rejection of the idea of constituent power as a unified whole, as Schmitt would have it. Instead, the only sense in which ‘the people’ to which democratic authority must refer can be seen as unified is in its representation to itself as such, in virtue of which we can still call something ‘one’ society, but this representation remains purely symbolic. ⁹⁹

Before concluding, I would like to highlight how the interpretation of the concept of ‘the people’ proper to constituent power I propose can be brought to bear on the split within the concept of ‘people’ indicated by Giorgio Agamben. Throughout European history, Agamben points out, the concept of ‘people’ – whether articulated as ‘the people’, ‘the nation’, ‘demos’ or otherwise – has been marked by a notable split. On the one hand, the people suggest an ascendent people that is not one among many, but the people par excellence, such as is connoted in the German Volk or indeed in Schmitt’s concept of the people. On the other, however, the people have always meant the supplement, the rabble or those falling short of what is required to be truly represented in the people. ¹⁰⁰ In the end, Agamben comes to advocate a politics that manages to reach beyond this ‘oscillation’ between a split-in-two concept of the people. ¹⁰¹

To begin with, Agamben gives very little detail on what such a politics would entail. More importantly, however, while that problem might be resolvable, my analysis suggests that moving beyond a rift of the type indicated by Agamben may not be feasible or even desirable. Taking a more positive attitude towards a similar structure of politics, Rancière highlights the political force that comes from the side of the people excluded from the ascendant one. Indeed, Agamben’s analysis seems in many ways analogous to Rancière’s differentiation between police and politics. Reading Agamben’s account of the people through Rancière suggests that moving beyond this split is not to overcome the fundamental inequality at the root of it, but rather to obscure the demos from view. ¹⁰² Derrida’s concept of justice reserves an important role for suspension. ¹⁰³ Interpreting the concept of the people in line with this suggests that the rift highlighted by Agamben is never overcome, but only suspended. Likewise, it may be temporarily fixed, but always in a provisional manner. To move beyond this opposition, by contrast, is to do away with the possibility of future instances in which a renewed suspension may be called for. A politics in which this split is no longer relevant in any sense may be a horizon, but remains, to use Derrida’s terminology, always to-come.

VII Conclusion: Claiming the people

In conclusion, let us return to the aporia of Rousseau’s lawgiver. My account of the people whose power constituent power is affirms Bonnie Honig’s decisionistic reading of Rousseau’s paradox, but goes beyond it in considering the moment of decision in itself as the moment of construction of the people. The figure of the lawgiver confronts the not-yet-people with a responsibility to choose devoid of any certainties that any of the options will be correct. Still the question is forced upon them: what is to be done? I have argued, moreover, that the answer is always an answer, in its particularity, which is congruent with Rousseau’s view. ¹⁰⁴ Based on my account of the people, I therefore want to end by suggesting a slight modification of the concept of constituent power. Having discussed who the constituents of constituent power are, that is, I want to point out a few considerations with regard to what is involved in the power of constituent power.

In accordance with the people as phronesis, constituent power is perhaps best seen as raising a claim. Such a claim may be one that asserts a ‘we, the people’ as much as it could be one that claims that specifically this is to be done. What matters in any case is that any such claim opens up, rather than closes down, a space for deliberation and contestation – for political agency – which calls forward phronesis in the absence of any determinate ground on which to base the answer to the questions raised by the claim. The structure is hence similar to that of the Rousseauean lawgiver: raising a claim that this must be done is not to foreclose inquiry into other possibilities but to foreground the necessity of a decision which invites debate as to the rightness of the claim being advanced; it propels the people into existence by stressing the need for a decision and therefore inviting collective willing – a general will – to emerge. But it does so in a manner that always remains provisional.

Before finally wrapping up, I want to very briefly pre-empt one possible line of criticism. One might worry that my emphasis on the people and on constituent power as a doing rather than a being invites a permanent instability into democracy. Such, indeed, was the worry that led Hannah Arendt to reject constituent power. ¹⁰⁵ But to say that the people is not a determinate pre-political being is not to rule out the possibility of relative stability. While the absence of determinate ground suggests the ever-present possibility of further questioning and further politicisation, there is no reason to assume that such politicisation must always be total and can never be stabilised. It is worth remembering, in that regard, what constituent power ultimately aims at: to constitute something – law, in most views. And if law ‘represents the principle of social reproduction’, ¹⁰⁶ the very aim of constituent power seems to drive towards a degree of relative stability. The point is that constituent power can only turn into constituent authority, ¹⁰⁷ and therefore represent a successful attempt at legitimisation, if it goes through the process of legitimisation and keeps that process alive. The alternative is to definitively fix democracy’s principles of legitimacy, and that is to dispense with politics from the start, and ultimately to undermine democracy altogether.