THE PARADOX OF POLITICS

The debate staked out by deliberative democrats between decisionism and deliberation can be illuminating, but the two sides together crowd out an alternative understanding of paradoxes and an alternative approach to them that may be fruitful to democratic theory. These alternatives are accessible by way of a different paradox, one more fundamental than the two aforementioned and possessed of a different structure: the paradox of politics (Ricoeur 1984; Christodoulidis [2001], who calls it an aporia, not a paradox). First theorized by Rousseau, since commented on by many, and developed in more detail here, the paradox of politics confronts us neither with a conflict between two incommensurable principles or “logics” nor with a binarily structured combat between or within Rousseau and Kant. Instead, the paradox of politics catches us in a chicken-and-egg circle that presses us to begin the work of democratic politics in medias res, in a terrain grounded neither in the sort of universal principled justification embraced by deliberative democrats, nor in the groundlessness of pure decision that deliberativists imagine is the only alternative.³

Although the paradox of politics is sometimes referred to as a paradox of founding, it is more than that, for it is alive at every moment of political life and not just at the origins of a regime. Rousseau does at first see the problem as one of origins in the Social Contract (Book II, Ch. 7): “In order for a nascent people to appreciate sound political maxims and follow the fundamental rules of statecraft, the effect would have to become the cause; the social spirit, which should be the product of the way in which the country was founded would have to preside over the founding itself; and, before the creation of the laws, men would have to be what they should become by means of those same laws.” In order for there to be a people well-formed enough for good law-making, there must be good law for how else will the people be well-formed? The problem is: where would that good law come from absent an already well-formed, virtuous people?

Rousseau is said by most commentators to have solved the chicken-and-egg problem of founding by introducing a lawgiver, a good man prior to good law, an objective or virtuous figure who can found the polity. Unfortunately, the lawgiver also aggravates the problem; he cannot just solve it, because his entry onto the scene compromises the people's autonomy or equality, which Rousseau also seeks to secure as a condition of their ability freely to will the general will (Gildin 1983, 67–91; Johnston 1999, 55; Bertram 2004, 128–47). As we shall see, however, the limitations of the supposed solution of the lawgiver do not matter that much because the problem exceeds the proposed solution. The lawgiver only addresses the problem of origins. But the seeming quandary of chicken-and-egg (which comes first, good people or good law?) takes off and attaches to democratic politics more generally once we see that established regimes are hardly rendered immune by their longevity to the paradoxical difficulty that Rousseau names. Every day, after all, new citizens are born, and still others immigrate into established regimes. Every day, already socialized citizens mistake, depart from, or simply differ about the commitments of democratic citizenship. Every day, democracies resocialize, recapture, or reinterpellate citizens into their political institutions and culture in ways those citizens do not freely will, nor could they. The problem that Rousseau seems to cast as a problem of founding recurs daily.

The idea that the so-called paradox of founding, the vicious circle of chicken-and-egg, exceeds the period of founding and attaches itself to politics, more broadly, is broached by the democratic theorist, William Connolly, who appreciates Rousseau's elaboration of the paradox of founding but also criticizes him for solving it. Rousseau's solution, Connolly argues, is not simply the lawgiver over whom so much ink has been spilled but rather, more subtly, Rousseau's location of the problem in time, at the beginning of a new regime. By confining the chicken-and-egg problem to the founding period, Rousseau prevents it and the unwilled violence that resolves it (personified by the lawgiver) from spilling over into politics more generally. In sum, Rousseau casts the paradox of politics as a paradox of founding in order to reassure his readers, “to imagine another time when it could be resolved” (Connolly 1995, 137), a time when the lawgiver and all he represents would be unnecessary, and politics could be more truly free. In so doing, Rousseau leads his readers to infer that they must just somehow get through the founding, whether by way of a lawgiver's impositional guidance or if necessary by way of a more explicit violence that can produce by force that which will later come by way of education and culture. Hence his approval of the idea that people can be “forced to be free” (1968, I:1; though, as Johnston points out, education and culture can be coercive as well [Johnston 1999, Ch. 4]). If they can find that much-needed bridge over what Hannah Arendt calls the founding period's “gap” in time, the people might somehow limit to the founding period the violence that attends the paradox of politics. They might then avoid the violence that otherwise recurs daily in established regimes, in the name of law (which claims to be nonviolent by representing itself as purely self-grounding) or popular sovereignty (which claims to be nonviolent by representing itself as the true and total will of the people who are, however, not yet formed).⁴

As it turns out, the so-called paradox of founding— the vicious circle of chicken-and-egg—is not overcome nor is it just concealed, as Connolly argues, by way of unacknowledged, foundational violence in Rousseau. (One compelling example Connolly gives of this hidden violence are the naturalized relations of sexual inequality in the male-headed households of Rousseau's polity). It is also replayed ad infinitum in Rousseau's own text, as the paradox of politics. Close attention to Book II, Chapter 7, of the Social Contract indicates that each of Rousseau's several efforts to solve the paradox succeeds merely in moving it to another register where once again it defies resolution and inaugurates anew a contestatory politics that is neither deliberative nor decisionist, per se. The repeated reappearance of the paradox of politics in Rousseau's text supports the idea that the problem of how to generate or recognize a general will recurs daily in democratic regimes.

THE PARADOX OF DEMOCRATIC LEGITIMATION

The paradox of politics is not central to deliberative democratic theory. The problem of how to identify or generate the general will is framed, instead, as the paradox of democratic legitimation. Here is how the deliberative democratic theorist, Seyla Benhabib, understands the paradox of democratic legitimation:

Democracy's regulative fiction affirms the sovereignty of the people but also limits or shapes its actual manifestations by requiring that it aim toward a collective good. The regulative fiction motivates the quest for a “moral standpoint” to guide or assess popular willing. Benhabib begins with Rousseau because she credits to him the worthwhile articulation of the paradox of democratic legitimation, but in the end she prefers Kant because Rousseau does not answer to the need for a moral standpoint (1994a, 30).

Rousseau himself makes no mention of regulative fiction, but he does seem to acknowledge the insufficiency of mere majoritarianism to democracy when he considers, in Book II, Chapter 3, of the Social Contract, the possibility that the general will can err. His response? The general will cannot err because if it erred it would not be the general will, it would be the mere will of all.⁶

The fact that the general will might go one way and the will of all another could have led Rousseau to reject the idea of a general will as such, or to lose faith in the people whose willing legitimates the regime. But, Rousseau insists, “the general will is always right and always tends toward the public utility.” The people may not see it. Their deliberations may lack “rectitude” not because they are corrupt, the people themselves are “never corrupted,” rather they are “often tricked,” Rousseau says (1968, II:3). The goodness of the people may be beyond dispute, but it becomes increasingly clear to Rousseau that not even their goodness can guarantee their rightness and, with the general will now operating as an external standard by which popular willing can be judged, the people may be found to be on the side of the will of all, not the general will, even if through no fault of their own. The problem is so serious that Rousseau refers only three chapters later in the Social Contract no longer to “the people” but to the “blind multitude” (1968, II:6).

Benhabib rejects what she sees as Rousseau's two solutions to the paradox: the first because it fails and the second because, she says, it imports to the scene of deliberative rationality and its regulative fiction an idealized rationality that violates the legitimacy to which it ought to be committed. Rousseau's first solution Benhabib calls an “arithmetic solution,” in which the will of all, which “is merely a sum of particular wills” can be used to calculate the general will as follows: “take away from these same [private] wills the pluses and minuses which cancel one another, and the general will remains as the sum of the differences” (Rousseau 1968, II:3, cited in Benhabib 1994a, 28). Benhabib correctly argues that “Rousseau's ‘arithmetic’ solution does not satisfy, because it is not at all clear what the language of ‘taking away the pluses and minuses of individual wills'’ could mean concretely or institutionally” (Benhabib, 29). But, caught up in the search for an independent normative standard to which Rousseau was not himself committed, and seeking a solution to the paradox rather than a diagnosis of it, Benhabib may dismiss too quickly the import of this first, flawed effort.

What other than a right solution might this arithmetic puzzle have to offer us? We might perhaps read Rousseau's odd arithmetic symptomatically: Rousseau's pseudo-arithmetic solution might illustrate for readers the deep inextricability that marks the will of all and the general will. On such a reading, Rousseau's casual directive—“take away some of this and some of that and count what's left over”—is not meant to reassure but rather to illustrate for us the hopelessness of knowing for sure when we have our hands on only one and not the other. Might this be the lesson of Rousseau's fuzzy math—that the general will is inhabited by the will of all and that we cannot know for certain when we have disentangled them and cannot hope, therefore, to guide our politics by such knowledge?

This possibility is borne out by Rousseau's full discussion of the lawgiver. Benhabib sees the lawgiver as Rousseau's second attempt to solve the paradox of democratic legitimation. In fact, Rousseau's turn to a lawgiver represents not a second solution but a second slightly different problem. The lawgiver arrives in response to a move away from the paradox of democratic legitimation and toward the paradox of politics, construed by Rousseau initially (but only initially, as we shall see) in Book II, Chapter 7 as a paradox of founding. In the paradox of democratic legitimation, represented in Book II, Chapter 3, on whether the general will can err, the people loom as a problem for the general will that is supposed to be theirs. They may “err” by willing the wrong thing. They may be “tricked.” Or their deliberations may lack “rectitude.” In the paradox of politics, however, as in Book II, Chapter 7, the focus is on finding “the best rules” for an emerging nation on the brink of existence as such. Here the problem is not that the people might be misled or might miscalculate in their deliberations such that they mistake a particular will for a general will. Here the problem is that the people do not yet exist as a people and so neither does a general will. The solution cannot be the right procedure or standpoint, for the people are in the untenable position of seeking to generate, as an outcome of their actions, the very general will that is supposed to motivate them into action. They lack at this juncture all the necessary conditions of communal action. Without a public, there can be no public good. Or so it seems. The problem is clearest in the moment of founding but, as I suggested earlier, it attaches to democratic politics as such partly because the people are never so fully what they need to be (virtuous, democratic, complete) that a democracy can deny credibly that it resorts to violence, imposition, or coercion to maintain itself. In some sense that is, the “people” are always undecidably present and absent from the scene of democracy. That is why it is always part of the point of democratic political practice to call them into being, rhetorically and materially (Johnston 1999; Keenan 2003) while acknowledging that such calls never fully succeed and invariably also produce remnants.

In this paradoxical moment of founding, no member of the community can yet be said to possess the needed perspective, which can only come post hoc, to form the rules or advocate for a collective good by which the people need to have already been acculturated in order to be not a “blind multitude” but a “people” capable of the autonomous exercise of popular sovereignty (Rousseau 1968, II:6). Somehow the impasse is negotiated—(Hannah Arendt (1963) in On Revolution speaks in this regard of the several good fortunes that allowed the American people and their founders to found rather than founder)—but its trace remains. The general will can never be really equally in everyone's interest nor really equally willed by everyone. Even if it were so fully willed, its authors nonetheless experience it as alien when it becomes a source of rule, and they are no longer only its authors but also law's subjects (Honig 2001a). More to the point, given the vicissitudes of legislative processes, there is always some divergence between what people will, as authors, and what emerges as law over them, as subjects (Habermas 1996, 354–59). Indeed, it may even be that this uncanny law, always somewhat alien and perpetually reproduced as such even by democratic institutions, is marked by the paradox of politics itself: in that recurring paradox, again and again, the subject postulated by politics is seen as never quite the cause because also always the effect of political practice. As Peter Fitzpatrick colorfully puts it, this is a moment in which “time runs widdershins and the present precedes itself” (2001, 74).

On this reading, the lawgiver, seen by some of Rousseau's readers as a would-be solution to the problem (he is said to have the perspicacity to get the law really right for the people who cannot yet do so for themselves), actually works to mark the problem rather than to solve it: even if the lawgiver could get the law really right, he enables the people's self-governance by compromising their autonomy. Just as the earlier arithmetic solution showed the general will was always inhabited by the will of all, so here too the point seems to be that popular sovereignty is always haunted by heteronomy, that the people are always undecidably also a “multitude.” In Rousseau (this is one of his great strengths as a democratic theorist), notwithstanding his aspiration for autonomous popular sovereignty, there is a sense that the people are never just heroes of their own story but always also protagonists in someone else's (represented by the would-be lawgiver), the always undecidable bearers of forces larger than themselves.⁷

However, let us assume for a moment, with Benhabib, that the lawgiver is a response to the same problem as the earlier failed arithmetic solution. There is something attractive about this figure, a founder who addresses the paradox of democratic legitimation by setting the material and procedural conditions for successful general willing. At a minimum he facilitates or educates. At a maximum, he forcibly acculturates. To Benhabib, however, the lawgiver solution is no solution at all because the lawgiver is an “instance of idealized rationality,” not deliberative rationality. Rousseau, she argues, is willing to “trade off legitimacy (the will of the people) for rationality (the legislator, ‘an instance outside the united will of the people…whose rationality transcends the legitimacy deriving from the people'’)” (Benhabib 1994a, 29–30). So Benhabib looks elsewhere for the missing “moral standpoint” and moves on to Kant.

Benhabib is not alone in treating the lawgiver as a betrayal of Rousseau's democratic ideal. Whether we take the lawgiver to be a figure of idealized rationality who trades off legitimacy in the way Benhabib describes, or of some other force (whether rationality, foreignness, paternal authority, moral certitude, or just really good intuition about what a particular people needs to function as a unity), the lawgiver is taken by many of Rousseau's readers to represent the would-be sovereign people's dependence on an outside force to guide them to the general will that ought properly to be their own, free internal motivation and the expression of their political unity (Honig 2001a, 21; Fitzpatrick 2001, 83; Keenan 2003, 17). In short, it seems that whatever problems of founding or willing are solved by way of the lawgiver's agency amount to little by comparison with the problems caused thereby for a would-be democracy.

The failure of the solution is not what it seems, however. In fact, Rousseau's lawgiver is less of a solution, failed or successful, than most readers of the Social Contract assume. He simply does not provide the certainty or force that Benhabib and other commentators associate with him for good or ill. The lawgiver may offer to found a people, he may even attempt to shape them, but in the end it is up to the people themselves to accept or reject his advances. They may be dependent on his good offices, but he is no less dependent on their good opinion. This idea would come as no shock to Moses, one of Rousseau's favorite lawgivers, who led a people repeatedly referred to in the Hebrew Bible as “stiff-necked.” The ancient Israelites changed their minds more than once about whether to follow their would-be lawgiver or rededicate themselves to other leaders and gods (Exodus 34:9). Contrary to those who see Rousseau's civic religion as, per se, a limit on popular autonomy, the Israelite example suggests that even absolute divine power can be resisted and engaged. Here, indeed, is the ineliminable moment of popular sovereignty in a nation (whose myths describe it as having been) formed by a lawgiver: the true lawgiver (prophet, god) is no more clearly identifiable by the “people” than is the general will. For democratic theory, this is a good thing for Rousseau's lawgiver, therefore, cannot help but inaugurate or represent a contestatory politics. The lawgiver forms a people into a unity that may or may not stay true to his leadership and he also generates remnants with plural and contending visions of the public good. The lawgiver does not transcend the political fray; he (or really law itself) is at the center of it.

In short, Rousseau is not using law or a lawgiver or even ideal rationality to solve a political problem from the outside. Instead, his treatment of the lawgiver considered in full suggests Rousseau has a political understanding of law, and perhaps even of rationality. The point is made clear when Rousseau ends his chapter on the lawgiver with a warning against cheap impersonators. “The great soul of the legislator is the true miracle that should prove his mission. Any man can engrave stone tablets, buy an oracle, or feign secret intercourse with some divinity, or train a bird to talk in his ear, or find other crude methods of imposing his beliefs on the people. He who knows no more than this may perchance assemble a group of lunatics [a “blind multitude?” ], but he will never found an empire and his extravagant work will soon die with him. Pointless sleights-of-hand form a fleeting connection; only wisdom can make it lasting” (1968, II:7). Just as the will of all can masquerade as the general will, so too can the charlatan impersonate a true lawgiver. “Legislator and charlatan thus remain radically undecidable” (Bennington 1994, 222). The people, still and always also a multitude, never so fully formed that they are uninhabited by anarchic waywardness, must nonetheless discern or decide the difference between the legitimate lawgiver and the pretender. The decision, as it were, rests with them; it reflects (in true chicken-and-egg fashion) who they are, and/or it also forms them into the particular people they are and are to be.

Thus, with the introduction of the lawgiver, the problem of how to know and/or generate the general will has not been solved; it has been shifted to a new register. Now in place of the earlier, supposedly arithmetic problem, we have a political problem: the would-be people's, the multitude's, performative choice to reject the overtures of one lawgiver, who is, or is said to have been, a charlatan, and to commit themselves to another, proclaimed authentic. In so doing, they choose and are chosen by their destiny. The decision may also divide them, however. They may split and become antagonists, as occurred after the prophet Mohammed's death in 632 A.D., when the people who later became Sunni or Shia decided differently how to answer the fateful question of who was to be Islam's next rightly guided caliph: Ali, a direct descendant of the prophet, or Mu'awiya Ummayad who, after beating Ali in battle, declared himself caliph? The Shia went one way, the Sunni the other.⁸

Even without such extreme division, the irreducibly political condition of the community is not ameliorated at this point in Rousseau's text; it is unrelieved. For, sure enough, the problem occurs again on a third register and then on a fourth and fifth. Perhaps to break out of the difficulty of distinguishing in the moment between lawgiver and charlatan, Rousseau moves to another indicator—durability of the laws given. “The Judaic Law, which still exists, and that of the child of Ishmael, which has ruled half the world for ten centuries, still proclaim today the great men who enunciated them” (1968, II:7). Invoked by Rousseau as an authentic indicator of authentic lawgiving, durability does not function reliably, however. Not only can we not assess durability in advance and are thus not helped by this indicator if we are looking for guidance in the present moment; but those who assess durability post hoc may also disagree on the facts or on their significance. (“A durable state may always be a mere simulacrum of a good one,” [Bennington 1994, 222].) There is no way to distinguish deserved duration (based on wisdom) from contingent duration (based on chicanery or mere good fortune), so even judgments based on the fact of durability are contestable.

Indeed, as soon as Rousseau invokes durability as a true indicator of good lawgiving, he concedes the contrary. By arguing against those who disagree with his own assessment of Judaic and Islamic law as the products of good lawgiving, he shows that he himself knows better than to think that mere longevity precludes people from passing plural and conflicting judgments: “while pride-ridden philosophy or the blind spirit of factionalism sees in [the laws of Moses and Ishmael] nothing but lucky imposters, the true political theoretician admires in their institutions that great and powerful genius which presides over institutions that endure” (1968, II:7). If there is no sure way to distinguish the general will from the will of all, the people from the blind multitude, the true lawgiver from the charlatan, significantly durable institutions from those whose durability is a function of mere good fortune or successful violent imposition or, even further, the true political theoretician from the pride-ridden philosopher or blind factionalist, then the will or judgment of the people who are not yet (or no longer) a people remains crucial. There is no getting away from the need in a democracy for the people to decide—on which is the truly general will, whose perspective ought to count, who is a true prophet, what are the right conditions for their lives, which enduring institutions deserve to endure and which should be dismantled, which would-be leader to follow, whose judgments to take seriously, and so on. Appeals to god or divine power do not escape this predicament, they replay it; when the lawgiver resorts to signs and wonders, he seeks to end the cycle of contestation but now the people move from judging him to judging the authenticity of his signs. Disagreements may persist and are of lasting political import (as the historic split between Sunni and Shia Islam suggests). It is worth noting that this reading (by contrast with Bennington's) stresses the undecidability of the people rather than the lawgiver.⁹

Benhabib could concede all of this, however, and still insist that her fundamental argument survives intact. The paradox that interests her may occur on four or five registers in Rousseau and not just one or two, the terms of the paradox may even shift a bit as the argument unfolds but, fundamentally, she might say, we are still faced with the problem that defines democracy: when faced with distinguishing general from particular wills, true lawgivers from pretenders, or properly durable institutions from those that are falsely so, popular sovereignty must be exercised not in a decisionistic fashion that takes its bearings from mere, aggregate preferences but on the basis of a commitment to deliberative procedures that generate outcomes that pass the test of a moral standpoint of universalizability. The material conditions of successful general willing called for by Rousseau (defined and sheltered territory, small population, relative equality, civic religion) are, from Benhabib's perspective, simply not a substitute for such a standard, nor for that matter are they relevant to us in late modernity. Caught up in the search for an independent normative standard to which Rousseau was not himself committed, conceiving of the paradox as a binary conflict not a vicious circle, and seeking a solution to the paradox rather than a diagnosis of it, Benhabib concludes that the people need the missing moral standpoint that Rousseau never provided in order to mark out a procedural path and cut short the spiral of paradox into which Rousseau supposedly falls. In quest of this standard, Benhabib turns to Kant.

Why didn't Rousseau make this move? Perhaps Rousseau wanted to impart a different understanding of politics, one that takes its bearings from the real human world as we find it (“taking men as they are and laws as they might be,” he says at the outset of his Social Contract), with its plural, conflicting aspirations and vicious circles, and not from a regulative ideal, fictional or otherwise. Rousseau is a theorist of political culture, not of universality (pace Habermas 2001, 768). But he is not therefore irrelevant to us now. Quite the contrary. His material conditions of popular sovereignty may be impractical in our late modern world (as Benhabib points out) but they remain instructive. It is as true for us as it was in his day that under conditions of radical inequality it is difficult to generate, identify, string together, and fight for public goods and against narrow factionalisms. Instead of addressing the problems of politics by way of a principle or a regulative ideal or fiction, Rousseau illustrates for us, time and again, the mutual inhabitation of general and particular will, people and blind multitude, lawgiver and charlatan, properly durable institutions and those stabilized by force. These are not binary paradoxes and they are not soluble by philosophical inquiry; indeed, they are often generated by philosophical inquiries, which tend to harden tensions into hypostatized, polar alternatives. What is lost by way of such polarization? The fecundity of undecidability, a trait that suggests that our cherished ideals—law, the people, general will, deliberation—are implicated in that to which deliberative democratic theory opposes them: violence, multitude, the will of all, decision.

Rousseau does juxtapose the general will and the will of all analytically but he also repeatedly depicts them as inextricably intertwined. No law and no perspective, as such, could disentangle them. The best a democratic politics can do is set the material conditions of shared living in such a way as to relieve the propensity of these two to diverge and to harden into oppositional relation. Rousseau teaches that conflicts between general will and will of all will be more or less frequent and intense, and variously experienced, in different settings under different conditions. The frequency and intensity of their conflict are products of varying political circumstances and of the expectations we bring to those circumstances. Theories that lead us to expect such conflicts and speak of a quasi-logical paradox may themselves lead us into not out of the paradoxes they seek to avoid or resolve.

Thus, we return to the question with which we began: what problem might the paradox of democratic legitimation be solving for democratic theory? The paradox of democratic legitimation is a paradox much-beloved by democratic theorists who worry about mere majoritarianism (which is to say, it is beloved in some form by nearly all democratic theorists). Unlike the paradox of politics, the paradox of democratic legitimation seems soluble and its supposed solution—a moral or juridical standpoint of universalizability, or the rule of law—underlines the waywardness of the people (their multitudinous character) and their need for legal or procedural institutions that are cast by contrast as merely stabilizing or enabling, not themselves wild or impositional. In short, if the paradox of democratic legitimation (Rousseau 1968, Book II, Ch. 3) is preferred by many democratic theorists to the paradox of politics posed repeatedly in Book II, Chapter 7 (and only in its first iteration as a paradox of founding), that is because the paradox of democratic legitimation focuses our attention on law's regulative powers, specifically on the need to direct the energies of the people which are assumed to be independent of law rather than partly its products. The paradox of politics, by contrast, calls attention to law's formative powers, its never fully willed role in processes of subject-formation, and the need, therefore, in a democracy, periodically or regularly to subject law to democratization by way of amendment, augmentation, or nullification.

Amendment, augmentation, or nullification are all forms of re-founding which, as such, call the (not necessarily blind) multitude forth out of the people or manifest “the people themselves” (Kramer 2004) acting in resistance to or in popular support of the institutions that form them. These re-foundings respond to the paradox of politics. But they may also cast us once again into that paradox, leaving us with no firm criteria or ground from which to distinguish with confidence the will of all and general will, multitude and people, because the perspective from which to do so and the identities at stake are themselves in question or in (re)formation. Thus, if the paradox of democratic legitimation takes the place of the paradox of politics in Benhabib's (1994a) considerations, that is because the former paradox, not the latter, allows us to take for granted the distinction by which it is supposedly troubled, the distinction between general will and will of all, while also rescuing law and proceduralism from implication in the phenomena they are entrusted to constitute and regulate. That is part of the point and attraction of the paradox of democratic legitimation.

When Benhabib mis-takes the paradox of politics for that of democratic legitimation (substituting, as it were, the problem posed by Rousseau in Book 2, Ch. 3, of the Social Contract for that posed in Book 2, Ch. 7, or conflating the two), she sidesteps one of democratic theory's knottiest problems and renders the (revised) paradox soluble by way of that “missing moral standpoint.” She also paves the way to a kind of constitutionalism that serves as the moral standpoint's proxy in the human world of politics. And with that we arrive at a second paradox, also taken to be binary in structure: the paradox of constitutional democracy.

THE PARADOX OF CONSTITUTIONAL DEMOCRACY

The problem of how to secure good general will in the absence of antecedent good law by way of which good willing is shaped is seen by some as soluble by way of constitutionalism. A constitution limits the damage that can be done by a poorly directed, confused, or wanton people by taking some things, like human rights, off the agenda. But constitutionalism, on this account, seems to take the place of Rousseau's awkward lawgiver (as he is traditionally understood) and to betray democratic ideals. Why should the people whose will legitimates the regime be bound by something they have not themselves willed? The paradox of constitutional democracy seems to restage the paradox of democratic legitimation. (Indeed, Judith Shklar [1992] identified constitutions with the general will.) Have we simply traded in one paradox for another?

Yes and no. As we shall see, this new paradox does replay many of the issues at work in the paradox of democratic legitimation. Instead of will of all versus general will, we have popular sovereignty versus constitutionalism. However, there are important differences between the two paradoxes. In the translation of one into the other, something does change.

In the paradox of constitutional democracy literature, the problem that for Rousseau occurred synchronically at the founding and forever after—(that which the law presupposes as its cause can only be produced by it as its effect)—unfolds diachronically. In place of the present tense problem of (un)justifiable constraints on popular sovereignty that daily affect the people's relation to itself as both ruler and ruled (the paradox of politics), we get the rather different problem of constitutional democracy and its limits from the past on popular sovereignty in the present. Recasting the conflict in this way divides the ruled (the people) and ruler (law, the founders, or the constitution) and restages the paradox of politics as a generational divide, a problem articulated most memorably by Thomas Jefferson when he asked: “Should the dead have rights?” This new paradox of constitutional democracy is not a conflict that goes to the very heart of democratic politics, which impossibly promises both (self-)sovereignty and freedom or, as Emilios Christodoulidis puts it, both self-rule and law-rule (2001, 126). Instead, the tense elements of the paradox are split into two distinct objects: the constitution represents law-rule and the people represent self-rule and these are seen as at odds (Schmitt 1996, 2006; Mouffe 1999, 2000, 2005) or somehow self-fulfilling (Elster 1984; Holmes 1995; Rubenfeld 2001).

Thus, the paradox of constitutional democracy externalizes the conflict that the paradox of democratic legitimation, notwithstanding its flaws, subtly put at democracy's heart. The unwilled, constraining element of rule is now identified not with democracy, per se, but with the constitution which may be right or necessary, and the paradox is now not internal to democracy (which seeks impossibly to combine will of all and general will, rule and freedom); rather, it is a feature of one kind of democracy, constitutional democracy, which impossibly but necessarily combines written constraint with free popular sovereignty and then derives its legitimation from that impossible, tense combination. The result? We have come full circle, for we are left with the implication often, albeit erroneously, attributed to Rousseau: that a really, unmediated, unwritten, and unconstrained democratic regime could experience—simultaneously and without conflict or paradox—both freedom and rule, both general will and will of all, law-rule and self-rule. The implication here is that were it not for constitutionalism, we could have democracy. This attractive implication is surely one of the reasons the paradox of constitutional democracy commands more scholarly attention than the paradox of politics.

However, because democracy is, or threatens always to be, in effect, a self-consuming artifact (those wayward, multitudinous people, again, on whom everything democratic depends but by whom everything democratic is threatened), democracy is said nonetheless to require the order and constraint of constitutionalism.¹⁰

Taking this temporalized conflict between past and present as the problem to be solved, defenders of constitutionalism give many reasons for constitutional limits. Constitutions represent the general will or universal norms, or norms that represent the true political commitments of the people whether or not they would will them in a particular political moment (Peter sober rather than drunk, as Stephen Holmes puts it [1995, 135]), or they help shape the people into the stable agency of popular sovereignty that democracy presupposes and requires (Rubenfeld 2001), or they supply a people with rules that are enabling not coercive (Holmes 1988, 1995), or constitutions represent a multigenerational commitment that does not constrain freedom but rather actualizes it over time (Elster 1984; Holmes 1988; Rubenfeld; Habermas 2001, 774; but see Kramer 2004 on the erasure of “popular constitutionalism”).

Those who map the so-called paradox of constitutional democracy as a tension between past and present generations assume that governance across temporal distance is similar to governance across spatial distance in that both are alien to their subjects and impositional in nature. The analogy between time and space, temporal and geographic alienness, was deployed to good effect by Jefferson, who worried that an enduring constitution would (really, should) be experienced by subsequent generations as a foreign imposition. Thomas Paine, too, spoke of rule from the grave as the “most ridiculous and insolent of all tyrannies” (quoted in Arendt 1963, 233). Noah Webster shared the concern and memorably metaphorized the temporal imposition in the clearest, spatial terms: “‘the very attempt,’ he warned, ‘to make perpetual constitutions, is the assumption of the right to control the opinions of future generations: and to legislate for those over whom we have as little authority as we have over a nation in Asia’” (1787; quoted in Rubenfeld 2001, 22). For these thinkers, “the past is a foreign country” (Hartley 1953).

The analogy, particularly Webster's version, subtly identifies certain of the Framers with the powers against which they had just rebelled: If seeking to govern over the future is just like governing over those far away and is illegitimate for the same reasons then, Webster implies, framers like Madison and Hamilton who found a constitution for the ages are little different from the English King and Parliament. The point is underlined by Webster's reference to Asia, in which England had been involved since 1685 (Kennedy 1970, 68–77; Farrington 2002, 111–17), a fact known to have been elsewhere on Webster's mind (1843). But the analogy also has other effects. First, because the passage of time is held to be responsible for the alienness of the law, the analogy subtly implies that the people in the present might experience without conflict both freedom and rule. Indeed, it implies that the founders enjoyed that very experience, since the only expressed worry is about later generations, not the current one. Second, those who reject the space-time analogy deployed by Jefferson, Webster, and Paine do so by embedding temporal distance in national time. They point out that the founders may have lived a long time ago, but they are our founders. Any lingering sense of uncanniness of those who come before to those who come later is papered over by talk of beneficiaries and heirs, fathers and sons, or intergenerational community (Rubenfeld 2001). Habermas himself seems to do this when he notes the responsibility of each “generation” to fulfill the promise of constitutional democracy.

In his 2001 essay, “Constitutional Democracy: A Paradoxical Union of Contradictory Principles?”, Habermas intervenes in the constitutional democracy debates with the aim of showing that deliberative democratic theory is untroubled by their conundra. He counters the constitutional democracy literature's externalization and temporalization of democracy's conflict between freedom and rule. He rejects the idea that constitutional democracy represents a struggle between past and present. For him, constitutional democracy models the appropriate relationship of considerations of right (constitutionalism) to considerations of will (democracy). Deliberative democratic theory, he says, is committed to a view of these as mutually implicated. The problem of a possible conflict between them is unique to modern constitutional democracies because they have two distinct sources of legitimation, the rule of law and popular sovereignty: “This duality raises the question of how the democratic principle and constitutionalism are related,” he explains (2001, 766). But that question is answerable. The two principles are “co-original,” Habermas says, meaning that they are of equal conceptual import; neither is prior to the other. He differs from Mouffe. The two principles are not antagonists in need of articulation; they are always already “co-implicated.” The act of co-legislation itself postulates the rights and equalities entrenched by constitutionalism. Thus, each depends on the other, and the rights that issue from each—public and private rights, taken together as basic rights—“are constitutive for the process of self-legislation,” a kind of autonomy in which rule and freedom are experienced together (2001, 774). The co-originality and self-legislation arguments are helped along by Habermas's procedural treatment of constitutionalism, in accordance with his own discourse theoretic requirements, and not as the reflection of a people's substantive norms and culture, in accordance with republican requirements (1996, 463-490). With the assumption of a procedural constitution (i.e., one cast as less impositional than the one Jefferson feared), we are some way out of the conflict Habermas is trying to solve (though see Ferrara on the unreality of this assumption in actually existing constitutions [2001, 788] and Waldron on how even procedural constitutions impinge upon democratic autonomy [1998]).

Although he attends to the paradox of constitutional democracy, the term paradox does not come up as frequently in Habermas's writings as in Benhabib's. He is less inclined than she is to engage with those who criticize deliberative democrats for being mired in paradoxes they cannot solve. Nonetheless, in Between Facts and Norms, Habermas theorizes what Patchen Markell refers to as “the constitutive tension of law and politics,” one that needs “to be negotiated by citizens, not transcended by the theorist” (Markell 2000, 45–46; cf. Habermas 1996, 680). Similarly, in the essay examined here, Habermas turns to practice to aid theory in charting a way out of the paradox of constitutional democracy. Along the way, though, he runs right into the paradox of politics.

The problem arises in Habermas' “Constitutional Democracy” essay, in which he considers a question posed by Frank Michelman. Michelman appreciates the deliberative democratic view of constitutionalism as the nonconflictual partner to democracy but, Habermas notes, Michelman worries that the deliberative democratic solution is incomplete, and he raises a powerful objection: “‘A truly democratic process is itself inescapably a legally conditioned and constituted process… Thus, in order to confer legitimacy on a set of laws issuing from an actual set of discursive institutions and practices…those institutions and practices would themselves have to be legally constituted in the right way’” (Michelman 1999; in Habermas 2001, 773). In Habermas' parsing: The “chain of presuppositions of legitimation reaches back even beyond the constitution-making practice. For example, the constitutional assembly cannot itself vouch for the legitimacy of the rules according to which it was constituted. The chain never terminates, and the democratic process is caught in a circular self-constitution that leads to an infinite regress” (774).

In Between Facts and Norms, Habermas' response to this concern was “that we understand the normative bases of constitutional democracy as the result of a deliberative decision-making process that the founders—motivated by whatever historical contingencies—undertook with the intention of creating a voluntary, self-determining association of free and equal citizens” (2001, 772; cf 1996, Ch. 9). However, in his essay on the paradox of constitutional democracy, under pressure from Michelman, Habermas sees his earlier solution as insufficient: “the paradox seems to return [i.e., rather than disappear] when we trace matters back to the act of constitution-making and ask whether discourse theory allows us to conceive the opinion- and will-formation of the constitutional convention as an unconstrained democratic process” (772).

Habermas rejects the option of stopping the infinite regress with a “moral realism that would be hard to defend” (2001, 774).¹¹

Thus, Habermas' contribution to the constitutional democracy paradox literature is to endorse a thin constitutionalism, situated between a conceptual co-originality (the ground) and a practice of tapping (the horizon) that together work to secure the sense of freedom that constitutional rule might otherwise threaten. He braids together the two sources of liberal democratic legitimation (rule of law and popular sovereignty), casts them as mutually constitutive, not antagonistic, and insists that each is dependent on the other for eventual full realization: “The allegedly paradoxical relation between democracy and the rule of law resolves itself in the dimension of historical time, provided one conceives of the constitution as a project that makes a founding act into an ongoing process of constitution-making that continues across generations” (2001, 768). Launched into time, but anchored by a co-origin that is out of time, the paradox of constitutional democracy seems to Habermas to be resolved or dissolved. Yet Habermas provides one more argument. He goes on to supplement the conceptual origin, co-originality, with an empirical event, the constitutional assemblies of Paris and Philadelphia. He provides the supplement in the context of a passing remark about Kant.

Recall that in Benhabib's treatment of the paradox of democratic legitimation, Kant was said to serve the deliberative project better than Rousseau. She had in mind the Kant of the second Critique and so does Habermas, when he embraces (and modifies) Kant, who sought to render will and reason compatible by “subordinating law to morality.” But Habermas is also critical of Kant because elsewhere, in “Conflict of the Faculties,” Kant “went beyond the systematic boundaries of [his] philosophy and raised the French Revolution to the level of a ‘historical sign’ for the possibility of a moral progress of humanity” (2001, 768). Habermas seems genuinely puzzled by Kant here, but he misremembers Kant's position.

In “Conflict of the Faculties,” Kant sought some sign that mankind might improve over time. “In human affairs, there must be some experience or other which, as an event which has actually occurred, might suggest that man has the quality or power of being the cause and…the author of his own improvement.” What event might “serve to prove the existence of a tendency within the human race as a whole” (181; italics original)? Not the French revolution. Its meaning, contra Habermas, was too volatile and uncertain to serve as a sign of anything. The revolution, Kant said, “may succeed or it may fail. It may be so filled with misery and atrocities that no right-thinking man would ever decide to make the same experiment again” (182). Instead, Kant took hope from the spectatorial response to the revolution. The sign of possible human improvement was the universal sympathy for the revolution that “borders almost on enthusiasm,” and the fact that spectators outside France, themselves still subject to monarchical rule, risked expressing their sympathy publicly: “it proves,” Kant says, “(because of its universality) that man has a moral character, or at least the makings of one” (182).¹²

It may be that Kant's uncharacteristic enthusiasm for enthusiasm, or something that “borders on” it, departs too much for Habermas from deliberative democratic commitments to dispassionate deliberate proceduralism.¹³

Can rational discourse, the core element of Habermas' deliberative democratic theory, need the supplement of an event (or, more aptly, its “reasonable trace”)? Habermas here tries to take advantage of the exemplary power of the event without forsaking the transcontextual rationalism of discourse theory; (not the event, but its “reasonable trace,” not a constitution, but a “rational constitutional discourse”). His effort recalls that of Kant in the Groundwork to illustrate the moral law's power with phenomenal examples, while insisting that such manifestations of the law are always only partial. In both cases, the examples exceed and may even betray the reason or the law they are meant to serve. That is part of the attraction and risk of such examples, even to those who try to use them in a limited or constrained way. Habermas needs Philadelphia and Paris to motivate his “constitutional patriotism.” Without the events to conjure up a colorful human world of passion, loyalty, betrayal, idealism, and reason, the idea of affectively attaching to a constitution (which, after its characteristic nods to the people's virtue, is simply a list of offices, procedures, and rules) is about as attractive as kissing a typewriter.

With the place names, however, a Pandora's box opens. Philadelphia and Paris represent not simply “constitutionalism” but two distinct revolutions and foundings, each characterized by its own unique, contingent drama, intrigue, public spiritedness, and remnants. In the U.S. case, “Philadelphia” conjures not only the assembly that produced the new national constitution but also the many competing conceptions of the American experiment that were sidelined or minoritized by the assembly and its constitution. The revolution, the Articles of Confederation, the constitutional assembly in Philadelphia, the resulting constitution itself, the antifederalists who fought it, the diverse crowds considered too unruly to be part of the deliberations, the various practices of popular constitutionalism delegitimated over the years, and the confederal practices of some native peoples are all the origins of contemporary American constitutional democracy. Not all of these are compatible, but all are part of American popular and democratic constitutionalisms and all—including those defeated—played a role in the historical shaping of American democracy. If they are unrecollected in Habermas' invocation of “Philadelphia,” that is because they are not, for him, part of its “reasonable trace.” It is the trace, not the event, that he seeks to recollect. It is the trace not the event that he secures when he says that those who tap the system of rights must orient themselves toward a beginning from which they take their bearings and build a tradition. To do so, and by doing so, they must inhabit the perspective of the founders (they are “in the same boat”) and take up the unfinished project of founding: “a constitution that is democratic…is a tradition-building project with a clearly marked beginning in time” (2001, 774).

But which is to be the beginning? Philadelphia or Paris? Habermas refers to that “great dual historical event” as if it does not or must not matter. But if it really does not matter, why utter the names Paris and Philadelphia at all? Clearly Habermas is aware that the names inspire; he wants to avail himself of their inspiration without the (for him) problematic particularities that make them inspirational but not universal, formative of a people but also productive of remnants.¹⁴

It may seem to Habermas that the paradox of politics calls us back repeatedly to the political moment of origins wherein it pulls the rug out from under our feet, and it may seem to him that in such moments, upheaval rather than settlement necessarily dominates, but this need not be the case. The paradox of politics can be a generative force. Similarly, we need not participate in any “democratic mysticism” (Holmes 1993) to ask after the remnants of constitutionalism. We can be clear-eyed about the undemocratic aims of some popular and minority powers, but without falling for any “constitutional mysticism.” We can be clear-eyed about the ways in which constitutions may operate not just to “canalize” popular power, as Holmes says (1995), but also to cannibalize it. That is, it makes little sense to talk of constitutionalism versus democracy, as such: there are varieties of constitutionalism, including popular constitutionalism, many of which were casualties of Philadelphia and Paris (Kramer 2004).

When Habermas' tappers choose “Philadelphia” as the beginning of their tradition-building enterprise, the costs of alternatives foregone and still sidelined daily are not viewed. Does Philadelphia win out over other contenders now because it did so then, that is, because you go to politics with the constitution you have, not with the one you wish you had? If so, then this seems to be one of those moments of decision that critics claim are discernible in deliberative democratic theory, notwithstanding its protestations to the contrary.¹⁵

THE PARADOX OF POLITICS, REVISITED

Which came first, the chicken or the egg? Anaximander solved this, the oldest paradox, by postulating an infinite thing, an uncaused cause that functions not unlike Habermas's postulate of co-originality. Aristotle generated a different solution, the immortality of species (1986, 415b, II:4) which, Roy Sorenson says, posits “an infinite relationship between finite things,…an infinite sequence of parents and children, [in which] a parent could care for each child and there is no need to postulate an animal [or other] origin for human beings” (Lennox 1985; cf. Sorenson 2003, 11). Aristotle's solution did not survive the theories of Darwin, Mendel, and Lamarck (evolution, inheritance, and acquired traits), which suggest that chicken or egg can be ranked first. Might Aristotle's solution be instructive nonetheless for politics?

Arendt (1963) may have had in mind something like Aristotle's infinite relationship between finite things when she remarked that the American revolutionaries succeeded because they practiced self-governance for decades before they rebelled. Although theirs was a revolutionary beginning, it was preceded by decades of acculturation in democratic habits, mindsets, practices, law, and institutions (an infinite sequence, as it were, of parents and children) made possible by their contingent distance from sovereign power.¹⁶

Arendt (1963) is often referred to as a theorist of beginnings. She repeatedly emphasizes the inaugural powers of action but her resort to something like Aristotle's infinite sequence to praise the American Revolution suggests a different notion of beginning than the ab initio variety with which she is usually associated (for good reason [see, e.g., 205] by many of her readers, including this one [Honig 1993]) and with which, as we saw earlier, Habermas (2001) also affiliates: he celebrates Philadelphia and Paris as an “entirely new beginning” and endorses a constitutional democracy that “is a tradition-building project with a clearly marked beginning in time.” What are the two different ways of thinking about political origins? Michael Oakeshott is helpful when he distinguishes ab initio origin stories, which he disapprovingly associates with the opening line, “In the beginning,” from those that that begin with “Once upon a time.” The latter invite entry into another time and postulate many temporalities (that they imagine $a$ time, suggests there are others as well). They have “no unconditional conclusion; [their] end is the beginning of another story. [They have] no over-all meaning; [they tell of] occurrences understood in terms of the meanings they acquire from their evidential contingent relationships.” In the beginning stories, by contrast, posit one time, one beginning and, as in Habermas's tradition-building project with its clearly marked beginning in time, their energies are harnessed to an overall purpose. They are grand narratives, out “not to tell a story but to construct a myth” (1975, 105), Oakeshott says, though we need not accept his implication that the choice is between good stories and bad myths. The paradox of politics resists that binary as well. Most narratives have elements of myth and story, and partake of both sets of traits Oakeshott identifies here. They can be turned to either purpose (story and/or myth) or, perhaps better, we can be turned by them to either purpose, or both.

When deliberative democratic theorists and some of their critics substitute neo-Schmittian binary paradoxes for the vicious circle of Rousseau's paradox of politics, they evade the conditionality tracked here in the work of Aristotle (1986), Arendt (1963), and Oakeshott (1975). Deliberativists do so in the hope that constitutional democracy can be provided with a less contingent more legitimate ground, with what Oakeshott refers to (in this paper's opening quotation) as a justification rather than a vindication. As I have suggested, this lands them in the very paradoxes they then worry how to resolve, manage, or transcend. If deliberativists prefer binary paradoxes, however worrying they find them, to the paradox of politics, it is because the problems posed by the former (general will vs. will of all, constitutionalism vs. democracy) are themselves the deliberative democrats' solution (cf. Dietz 1998). In their preferred paradoxes, the people are a problem that rules might solve. In the paradox of politics by contrast, with its plural “once upon a time” and “in the beginning” temporalities, the problem is that the people are always also a multitude, the general will is inhabited by the will of all, the law(giver) is possibly a charlatan, and political theorists' objectivity is also partisan. Here, we get neither deliberation nor decision as such; we get a politics, in which plural and contending parties make claims in the name of public goods, seek support from various constituencies, and the legitimacy of outcomes is always contestable.

With their focus on conflicting principles (Habermas, Benhabib) or incommensurable “logics” (Agamben 1998; Mouffe 1999, 2000; Christodoulidis 2001), democratic theorists drive us into binary paradoxes that shuttle us back and forth between decision and deliberation. Even Mouffe's strategy of articulation cannot help but preserve the binary of law and democracy whose pragmatic settlements she seeks to rework.¹⁷

Rousseau's (1968) paradox of politics, like Aristotle's (1986) infinite sequence, explodes the familiar binaries that structure the debates between decision and deliberation. It points to alternative domains of political work by depriving us of postulated points of origin (landing us right in the conundrum of which comes first, good law or the wisdom of self-governance?) and inviting us to see how (admittedly to different extents, in different ways, in different regimes) law and its authors/subjects fundamentally fail to intersect in the present in ways that satisfy independent standards of legitimation. This is not, contra Jefferson, Webster, and Paine, simply because others authored the law by which we are later, distantly governed. Nor is it, contra deliberative democratic theory, because we lack a moral standpoint from which rightly to distinguish the general will from will of all. Nor is it because we have failed our responsibility to tap the systems of right by which we are governed (though we may well have). Rather it is because this infinite sequence is the condition in which we find ourselves when we think and act politically, when we demand that the lawgiving/charlatan institutions by which we are always already governed and shaped be responsive to the plural, conflicting agents who together are said to authorize or benefit from them: the ever-changing and infinitely sequential people, the multitude, and their remnants.