Throughout Gary Jacobsohn’s writings a core concept has been his idea of “constitutional identity,” most explicitly in Constitutional Identity and Constitutional Revolutions (Jacobsohn Reference Jacobsohn2010; Jacobsohn and Roznai Reference Jacobsohn and Roznai2020). As Jacobsohn observes, the reference to constitutional identity necessarily points to the working of constituent power; the same power to create a constitution is the power required to replace, amend, or reconceive its underpinnings. The assignment of constituent power to “the people” means both that the people create a constitution with the capacity for an identity – or multiple claims to identity existing in disharmonic competition – and that the people have the authority to change a constitution’s identity without thereby affecting their own.

These seemingly simple observations raise a whole series of questions: Are there real and “sham” exercises of constituent power?Footnote ¹ Democratic theorists talk about assertable claims of representation and constituency; should we think of “constitutional identity” and “constituent power” as the names of other assertable political claims rather than descriptions of political phenomena and thus turn our attention to the relative strength of different versions of that assertion? (Saward Reference Saward2006).Footnote ² Are there normative standards for good or genuine expressions of constituent power – as we may speak of constitutional orders that produce evil results – or would the normative evaluation of a claim of constituent power reduce to a judgment of superiority among “peoples”? (Jacobsohn Reference Jacobsohn2010).Footnote ³

In this chapter, I propose to get at some of these questions by considering constitutional identity in relation to a particular theory of constituent power, that of Jean-Jacques Rousseau. Rousseau is recognized as one of the three major modern or early modern theorists of constituent power along with Sieyès and Schmitt, but of the three Rousseau’s stands out for his attempts to deal with what Bonnie Honig calls “the paradoxes of democracy”: the problem of legitimating a founding moment, the relationship between constituent power and time, and the question of the identity of a people in relation to its constitutional commitments.Footnote ⁴

Rousseau’s model of constituent power is nicely captured in his description of three moments of the people in action following the creation of the state.

The same three moments appear in three dimensions of citizens’ self-aware actions. “[T]hey collectively take the name people; individually they are called citizens, insofar as participants in the sovereign authority, and subjects insofar as they are subjected to the laws of the state” (Rousseau Reference Rousseau and Dunn2002). The active sovereign is the people engaged in constitution making, “the action of the entire body acting upon itself – that is, the relationship of the whole to the whole, or of the sovereign to the State.” The distinctions among the State, the Sovereign, and Power capture three moments (or dimensions) of constituent power. An exploration of the points of consistency and difference between this model and Jacobsohn’s theory of constitutional identity demonstrates both the power and possibly some problems for the latter.

1.1 Rousseau’s Theory: Sovereignty, Representation, and the Paradoxes of Founding

Rousseau asserted the classical republican position that the constitution (in the Aristotelian sense of the term) shapes the people.

Maintenance of such national identities required homogeneity and separateness, as in the case of Moses and the Israelites. “In order to keep his people from dissolving among foreign peoples, he gave it morals and practices incompatible with those of other nations; he overburdened it with distinctive rites, ceremonies” (Rousseau Reference Rousseau and Kelly2014, 172). In turn, that meant that republics had to be small to succeed. “Almost all small states, republics and monarchies alike, prosper because they are small…. All great peoples crushed by their own mass groan” (Rousseau Reference Rousseau and Kelly2014, 183). In order to achieve the virtues of a republic in a larger, modern state the only solution was confederation, but confederation in a form that maintained a shared and singular sense of national peoplehood capable of sustaining an assertion of constitutive power. “In the present state of things, I see only a single means of giving it that stability it lacks; that is to infuse, so to speak, the soul of the confederates into the whole nation” (Rousseau Reference Rousseau and Kelly2014, 174).

Rousseau’s description raises the paradox of founding: By what democratic processes can the demos itself be constituted? Rousseau’s answer, of course, is his particular model of the social contract. “The people, being subject to the laws, ought to be their author: the conditions of the society ought to be regulated solely by those who come together to form it” (Rousseau Reference Rousseau and Dunn2002, 193). What about nonsigners? “They are foreigners among citizens. When the state is instituted, residence constitutes consent; to dwell within its territory is to consent” (Rousseau Reference Rousseau and Dunn2002, 250). The theme is echoed in his description of the liberator General Paoli’s address to the people of Corsica: “Corsicans, be silent. I am going to speak in the name of all. Let those who do not agree leave, and let those who do agree raise their hands” (O’Brien Reference O’Brien and Dunn2002, 302).

An implication is that “the people” of a given state are not coextensional with “all persons residing in the state.” And indeed, Rousseau frequently refers to the idea of citizens as an elite and empowered class. “The real meaning of this word [citizen] has been wholly lost in modern times; most people mistake a town for a city, and a townsman for a citizen. They do not know that houses make a town, but citizens a city” (Rousseau Reference Rousseau and Dunn2002, 75). As a theoretical matter, these are distinctions that emerge subsequent to the social contract; it is the sovereign – the people exercising constituent power – that has the authority to “set up several classes of citizens and even lay down the qualifications for membership of these classes” (Rousseau Reference Rousseau and Dunn2002, chap. 6).

Rousseau famously declared that the sovereign cannot be represented, but Rousseau’s sovereign is the people exercising its power in assembly wherein each participant expresses their view of the general will.

It is estimated that Geneva’s General Council numbered fewer than 1,600 in a “town” of 20,000 (Rosenfeld Reference Rosenfeld1987, 100). This was the legislature that exercised constituted authority, an elite and representative body charged with making laws as opposed to constituting loi. Yet in Rousseau’s telling the General Counsel also spoke for the general will. “In a State such as yours [Geneva], where the sovereignty is in the hands of the People, the Legislator always exists…. It is assembled and speaks authentically only in the General Council” (Rousseau Reference Rousseau, Kelly and Grace2001, 236). In its operations, the General Council appealed to this role in the two questions with which it opened each session. The first is, “Does it please the Sovereign to preserve the present form of government?” The second is, “Does it please the people to leave its administration in the hands of those who are actually in charge of it?” The formulation “does it please the people” indicates that the actions of the General Council only became actual exercises of constituent power once they were ratified by the people. “The laws, although received, only have a lasting authority so long as the people, being free to revoke them, nevertheless does not do so” (Rousseau Reference Rousseau and Dunn2002, 324), leading Colon-Rios to describe the initial creation of a constitutional order as “provisional” (Colon-Rios Reference Colon-Rios2020).

Yet this seems to raise problems for Rousseau’s theory. How is the General Council not a representative of the sovereign, something Rousseau says is impossible? One answer turns on the meaning of the term “represent.” The General Council cannot act for or stand for the people, but it can assert a claim to speak by expressing the general will. Hence the significance of subsequent ratification. The expression of constituent power, in other words, would be episodic. In between such moments – apparently marked by the calling together of a constituent assembly containing all the citizens – the state would operate through a government that represented the popular will by the authorization of the general will. In these periods of the operation of the General Council and the other councils of Geneva’s government, is its constituent power present or does it lie dormant?

Rousseau confronted an even more difficult second paradox. How is “the spirit of the Constitution” to be “infused” among the Polish people before the creation and adoption of that same constitution? A people would learn to be virtuous citizens, he said, through the experience of living in a republic – that is, under an appropriate constitution. “[T]he voice of duty replaces physical impulse and right replaces appetite.” Yet until they became virtuous a people would lack the will to act collectively to create such a constitution: “the effect would have to become the cause” (Rousseau Reference Rousseau and Dunn2002, 272).

Rousseau’s famous solution was the founding Legislator, a figure capable of using religious authority to compel a people to live virtuously, as if they were collectively possessed of virtue. But, as we have already seen, Rousseau also identifies “the Legislator” with the general will, articulated through the actions of the General Council. Prior to the time when the people have been made virtuous by the experience of their constitution, do they possess constituent power? Or is the legitimate authority of the people to create or alter a constitution something that arises by virtue of the constitution’s effects over time?

The easy reading of Rousseau is that constitution making is an episodic exercise of constituent power/sovereignty, after or in between which moments liberty consists in self-government through representative institutions. But this reading becomes problematic because the question of exclusion in representation – a principle that follows analytically if not practically from the imagined moment of foundation – makes authorship of constitutional identity the authority of a closed and elite group who then rule over others as well as each other. This is consistent with some modern descriptions of constitution making in practice (Hirschl Reference Hirschl2007; Ginsburg Reference Ginsburg2014), but by freezing the identity of the people it makes constituent power little more than the name attached to an outcome, essentially definitional and hence tautological. As Rosenfeld puts it, “The contract is unanimous only in an axiomatic way, as a logical deduction from the meaning of the terms themselves: it is unanimous among the contracting parties, which, of course, it would have to be, if not they would not be contracting parties” (Rosenfeld Reference Rosenfeld1987, 93).

1.2 Constitutional Identity, Constitutive Power, and the Question of Legitimacy

Jacobsohn wrestles with some of the same potential problems that troubled Rousseau. The idea of constitutional identity immediately points to a consideration of constituent power as the authority of the people to engage in constitution making in the first place. But while Rousseau conceives of constitution making as episodic and formal, Jacobsohn sees constitutional identity as something that emerges over time. “[A] constitution acquires an identity through experience … this identity exists neither as a discrete object of invention nor as a heavily encrusted essence embedded in a society’s culture, requiring only to be discovered” (Jacobsohn Reference Jacobsohn2010, 7). The process of acquisition of identity is not only dialogical; it is “transactional” (Jacobsohn Reference Jacobsohn2010, 326). The term “acquisition” is particularly interesting here; in Jacobsohn’s view the creation of a constitutional identity is not simultaneous with the creation of the constitution. But just as constituent power is required to legitimate the creation of a constitution, it is equally an essential element of legitimate actions creating or altering a constitution’s identity. What, then, is the role of the people – Rousseau’s sovereign – in imbuing a constitution with an identity or altering that identity through constitutional revolution?

All of which raises the question of whether a constitution has an identity at all prior to the initiation of that dialogue. That is, like Rousseau’s people learning virtue, does a people living under a constitution learn to invent its identity not present at the founding?

Of course a constitution may also be founded or changed another way, through a violent moment of disruption; is there a constitutional identity in that moment? If so, that identity cannot derive from the exercise of constituent power because constituent power cannot be authorized until after there is a constitutional expression of its recognition, creating a problem of infinite regress. “The people who participate in the referendum … were already constituted. Rules regarding who would be eligible to vote and participated in such an expression of the people’s will were already established: thus the famous paradox of constitutional democracy – ‘the constitution constitutes the people who in turn constitute the constitution’” (Jacobsohn and Roznai Reference Jacobsohn and Roznai2020, 248). Consider Srinivassan’s description of the first example of a constituent assembly. Srinivassan identified the source of the concept in seventeenth-century English Levellers’ 1648 call for authorized representatives acting on behalf of “the well-affected of every county” to meet with representatives “chosen by the Army” to create a new local political order (Srinivassan Reference Srinivassan1940). This makes the claim of constituent power nothing more than the assertion of “brute historicity” (Schweber Reference Schweber2021); it was the case that in this place at this time a sufficient number of people accepted the process of selection as legitimate as to make the conclusions reached by the assembly enforceable over objectors, essentially Rousseau’s solution of treating nonparticipants as “foreigners.” Can this satisfy any meaningful definition of constituent power in a way that provides a legitimating theory for the exercise of political authority?

As Jacobsohn notes, several solutions to this potential problem have been proposed. One solution is to simply abandon the idea of constituent power altogether. Treat the creation of constitutions as a purely positivistic act, and thereafter the assertion of authority in terms of competing legal and political forces seeking control. Another is to treat constitutive power as a purely narrative device, a purely symbolic expression of a claim of representation. “By telling ourselves a fictional story about ‘the people,’ we satisfy a ‘sort of psycho-legal need’…. According to this conception, the people’s constituent power should not be regarded as an ‘actual aggregate entity in the real world,’ but rather as ‘a concept that helps explain the normative basis for a constitution’s claim to authority’” (Jacobsohn and Roznai Reference Jacobsohn and Roznai2020, 252; Tushnet Reference Tushnet1983). And yet a third solution is to propose that constituent power is a kind of political dark matter, visible only retrospectively in its effects. Interestingly, the approach that favors purely retrospective assertions of constitutive power is consistent with theories that claim that power is always potentially present, as in Albert Venn Dicey’s theory of parliamentary supremacy limited only by the ever-present possibility of resistance from the populace (Venn Dicey Reference Venn Dicey1885).Footnote ⁵ Nonetheless, that approach cannot gain explanatory purchase on the idea of constitutive power as an element of founding, only as an element of the operation of the fully formed Rousseauian triumvirate of people/state/government.

Jacobsohn rightly rejects the solution of abandoning constituent power as a legitimating concept. But he recognizes the paradox of trying to find an expression of constituent power before such power has been constituted in an institutionalized form.

Colon-Rios makes the same point with specific reference to the idea of constituent assemblies as occurring somehow outside of law. “Both extraordinary and periodic assemblies could only be convened in accordance with the law: only a public meeting that complies with the forms established by the entire citizenry in their constitutional framework could be taken as authorised to pronounce the people’s voice” (Colon-Rios Reference Colon-Rios2020, 104). Andrew Arato describes “post-sovereign” constitution making in which procedural legality defines the exercise of constituent power (Jacobsohn and Roznai Reference Jacobsohn and Roznai2020, 234); Jacobsohn’s and Colon-Rios’ arguments point to the possibility that procedural legality is a necessary element of an exercise of sovereignty as anything more than a Schmittian exercise of will.

Thus we are returned to the problem of legitimacy. For an assertion of constituent power to be considered legitimate, it must be considered true; that is, even as a purely positivistic description of the nature of the governing narrative, one must recognize that the internal norm is the assertion that “the people” have truly acted. The possible tests for such validity seem quite different depending on whether one is talking about a moment of violent rupture occurring outside an existing constitutional system or a revolutionary change in meaning occurring within the operation of existing institutions.

Jacobsohn seems to try to elide that possible difference by making the distinction between real and sham assertions of constituent power turn on procedures.

For example, Jacobsohn points to Ireland’s experiment with deliberative jury of randomly selected ninety-nine citizens as part of the process of constitution making along with referenda and representative democratic politics; taken together these elements present a “legal approximation of constituent power” (Jacobsohn and Roznai Reference Jacobsohn and Roznai2020, 255). Subject to these limitations, constituent power may be exercised in either a moment of founding or a moment of revision. “Constituent power does not collapse into constituted power. The former can be exercised outside the constitutional order or through constituted organs” (Jacobsohn and Roznai Reference Jacobsohn and Roznai2020, 250–51).

But the two paradoxes – the problem of infinite regress and the problem of majoritarianism – remain troubling, particularly insofar as they seem to pose different problems for exercises of constituent power outside or through the existing constitutional order. Is “a legal approximation” of constituent power sufficient to resolve the tension between a legitimating norm of constituent power and democratic norms of ongoing deliberation, contestation, and pluralism?

1.3 Revisiting the Rousseauian Model of Constitutional Identity

Return to Rousseau’s resolution of the paradox of virtue by the invocation of a Legislator able to impose constitutionalism before there exists a popular sovereignty capable of creating it. For Rousseau, the idea was that, over time, people become virtuous, then citizens, then imbued with spirit of the constitution. This is a classically republican ideal in which the constitutional order of the polity shapes, not reflects, society and psychology. In Jacobsohn’s terms, at least until such time as public virtue is universally realized, such a constitution is “militant” rather than “acquiescent” (Jacobsohn Reference Jacobsohn2010, chap. 5), operating as a “sail” that “serve[s] as a foundation for the construction of a social order embodying the transformative hopes of its revolutionary promise” (Jacobsohn Reference Jacobsohn2010, 215).

One potentially superior alternative is to reconsider Jacobsohn’s distinction between a dialogic emergence of constitutional identity and a violent, disruptive moment of constitutional establishment. Considered fully, Jacobsohn’s argument presents a powerfully persuasive case that only constitutional change occurring within an existing constitutional system can express constituent power, because only in that situation is there the possibility of a developed constitutional identity. That is, constituent power is subsumed into constitutionality not as a matter of distribution of decision-making power but as an expression of identity. One cannot have constituent power without a people, nor a people without an identity in a system based on an ideal of constitutionalism that required a fully formed constitutional identity.

This is very close to Habermas’ solution – later developed further by both Jacques Derrida and Bonnie Honig – of the infinite regress problem. It is, indeed, impossible to legitimize a moment of founding rupture based on appeals to any subsequently emergent norms and values. Instead, according to these arguments, one treats the founding moment as the establishment of a set of discursive resources that are drawn upon in the ongoing, dialogic, contested process of creating a constitutional identity (Honig Reference Honig1991).

Jacobsohn acknowledges that a constitutional identity may be unknowable at the moment of founding. “Can we then know with certainty what is and is not irrevocable…? I would suggest that an affirmative response ought to be received skeptically. The reason for such uncertainty … lies in the dynamic quality of identity and the dialogical process by which it is formed and develops” (Jacobsohn Reference Jacobsohn2010, 332). An “unfinished symphony…” (Jacobsohn Reference Jacobsohn2010, 333)

But is this enough? Are these discursive resources not necessarily poisoned at the well by the brute historical fact of majority? If the mere assertion of the claim to constituent power is enough to bind future generations, then the answer is “yes” and the project is self-defeating: The only authoritative constituent power is one which lacks all the essential characteristics of constituent power that has any capacity for legitimation. The US founders were concerned with economic rights and interests, so they created a constitutionalist reservoir of meaning oriented around a conception of rights as a species of property. It required a violent disruption (such as the American Civil War) and new founding text (the Reconstruction amendments) to create a new reservoir of meaning on which we now draw in formulating our assertions of constitutional identity. An assertion of identity requires stability across time. “Imagining a polity in which the live hand of the present was the animating and sole directive source for its constitutive choices is to imagine a polity without a constitutional identity” (Jacobsohn Reference Jacobsohn2010, 324). But, at the same time, the sail model implies an aspirationist and transformative understanding of constitutional identity, which raises the question in the US case: Should we conceive of those aspirations as already present at the beginning – as in Lincoln’s invocations of the Declaration of Independence – or as the product of the dialogic process of creating constitutional identity?

Presumably any good answer will take the form of “both”, but moments of disruption – including Rousseau’s imagined signing of a social contract – cannot be conceived as exercises of constituent power. At most they are moments in which a bunch of potentially interesting ideas are recorded. Those ideas are not in any way determinative; in the dialogic process of contesting and creating a constitutional identity one or another of them may prove useful, but the choice of whether and how to use those ideas is the true exercise of constituent power that results in the emergence of a constitutional identity.

Two final thoughts emerge from this discussion. First, the demands of democratic citizenship are quite high. There would appear to be an affirmative obligation to engage in dialogue about constitutional identity and thus to participate in its creation and development. There is the expectation here of citizens being changed by the experience of that dialogue and of the dialogue itself changing its contours over time. This is the perfectionist element of the story – constitutional identity working itself pure would be the ideal Dworkinian case – that is the parallel to Rousseau’s unsatisfying reliance on the inculcation of virtue through the experience of citizenship.

The second closing observation is that while each individual assertion of constitutional identity must be assessed in terms of its specific elements, history, and conditions, the model in which constitutive power inheres only in the operation of a constitutional system appears as a universal analytical necessity. As Jacobsohn says, “What this suggests is that the dynamics of constitutional identity are less the result of any specific set of background cultural or historical factors than the expression of a developmental process endemic to the phenomenon of constitutionalism” (Jacobsohn Reference Jacobsohn2010, 348).