The popular sovereignty of Indigenous peoples: a challenge in multi-people states

ABSTRACT The doctrine of popular sovereignty holds that the ‘supreme authority of the state’ belongs to the people, not to the political institutions exercising public power. What are the implications of this view when there is more than one people in the territory of that state? The case of Indigenous peoples highlights this question, as they are unequivocally peoples who are distinct from the majority population. This paper subjects to criticism of the received view according to which the inclusion of Indigenous peoples in democratic institutions is sufficient for the realization of their popular sovereignty. Instead, we argue that their constituent power must be recognized – the power to create and negotiate the constitutional order. The realization of popular sovereignty in settler states thus necessitates a process where the constitutional order is negotiated by Indigenous peoples and the majority population in conditions where the two parties are mutually recognized as sovereign.


I. Introduction
A central idea in democratic thinking is that public power is legitimate only when it 'derives from' the people. The powers exercised by the state do not belong to the state, but to the people subject to them; peoples are the true bearers of sovereignty and 'supreme authority in the state' (Ochoa Espejo 2015;Buchanan 2002). The political vigor of this idea is testified by contemporary state constitutions where the ideal of popular sovereignty is affirmed, either by explicit acknowledgement of this principle or by provisions to the effect either that public power derives from the people or that the constitution has been adopted by 'We, the People' (Galligan 2013). In this article, we consider the challenge of popular sovereignty for multipeople states, specifically, by asking how 'the people' should be understood in states established on the lands of Indigenous peoples. We argue that the idea of popular sovereignty used to justify state power needs to be reformulated in order to account for Indigenous peoples as distinct peoples. The problem for settler constitutional orders is that their legitimacy depends on the popular sovereignty of Indigenous peoples, qua peoples, not simply as citizens of an already formatted constitutional order. The challenge of popular sovereignty then is about how to recognize Indigenous peoples as authors of their own status as citizens in a state situated in a territory that used to belong to them.
We offer illustrative (but not comprehensively representative) real-world examples of arrangements that aim to include Indigenous peoples in the democratic processes of the state, but which fall short of recognizing their popular sovereignty. We aim to show that a renovated approach to popular sovereignty that introduces a procedural conception of 'constituent power' can be used to evaluate and guide relationships between states and Indigenous peoples. This theoretical underpinning could also assist to augment settler official understandings of the constitutional role of Indigenous peoples and their claims.
We take as a starting point the fact that Indigenous peoples are among the peoples upon whose will the legitimacy of settler constitutional orders necessarily depends. Indigenous peoples are now (although not historically) included within the formal jural categories of citizenship in settler states, and their political citizenship is protected through the ordinary channels of democratic participation available to all citizens. Yet, in some states, Indigenous peoples are not just citizens but also recognized as distinct peoples with legal personality and collective rights (sometimes including rights to selfgovernment) in national constitutions and legislation, judicial decisions and in treaties or other agreements (see, e.g. IWGIA 2016; Lightfoot 2016). In first order expressions of popular sovereignty, however, such as referenda, even these statuses do not generate a distinctive role for Indigenous peoples -as peoples -in legitimizing the constitutional order. The fact remains that Indigenous peoples were excluded from the design and establishment of settler state constitutional orders and continue to be so excluded today. The deficits in dominant theories of popular sovereignty enable this exclusion. How to make good on popular sovereignty and to ensure that the authority of the state derives from all peoples subjected to it -including Indigenous peoples -is therefore a theoretical challenge just as much as a political problem.
Most democratic theories struggle to accommodate the notion of multiple peoples with a democratic order, though usually confined to questions about inclusion in democratic citizenship. As noted by Margaret Moore, the standard approach is to 'assert that this wrong can be corrected by fully including individuals -though not necessarily in the political community to which these individuals feel that they belong -as equals in the political system' (Moore 2010, 156). Popular sovereignty of Indigenous peoples may, however, better be understood as constitutional power or, as it is often called, constituent power. In order for the powers of the state to derive from the people, the constitutional order should be the product of agreements in which the people have been decisive. In places where Indigenous peoples exist as distinct politics, the constituent power of the people cannot be reduced to the participatory activities of citizens through democratic procedures. Citizenship status and the rights associated with it is a product of the constitutional order and hence cannot be the source of its legitimacy. Instead, arrangements for the political inclusion of Indigenous peoples as citizens must be the object of a legitimation process that takes the separateness of Indigenous peoples as a starting point. This requires recognition of the constituent power of the multiple peoples that shares the same territory. The realization of popular sovereignty in multi-people settler states is therefore premised on Indigenous peoples being able to negotiate the constitutional order with other peoples on equal terms -thus having an equal share of the constituent power of the peoples subjected to that order. The outcome of exercises of constituent power cannot be determined beforehand, as they depend on the particular circumstances and aspirations of the people concerned. The view defended in this article, then, is a procedural account of popular sovereignty where the constitutional order derives from the peoples if and only if it is the outcome of negotiations by Indigenous peoples and majority settler peoples.
This article necessarily appears against the backdrop of a vastly complex and lively conversation underway about the nature of the juridical and political sovereignty of peoples in settler societies. This body of thought, spanning the disciplines of law, political theory and anthropology, owes much of its intellectual vigor to the contributions of Indigenous scholars, leaders and practitioners. They have their own very longstanding political and legal theories about the authority of peoples. Many reject the idea of 'sovereignty' as one too compromised and limited to be of use in Indigenous political thought and action (Alfred 2005, 67-71). Others have critically reworked the idea in line with Indigenous knowledges to propose a grounded counter-balance to the state's absolutist claims (Simpson 2014, 11-12). Indigenous concepts of authority, legitimacy and the nature and source of law may differ fundamentally from those that are used to justify state power, and are typically less rigid, less absolute, less anthropocentric and more relational than the western political traditions that are the focus of this article (Black 2011;Borrows 2016b;Coulthard 2014, 171;Nilsson 2021). It is a premise of the argument set out here that the theoretical traditions of Indigenous peoples have authority independently of settler recognition or endorsement. Dominant strands of western political thought have shut down possibilities that should remain open, intellectually and practically. Our aim is to redirect attention to these possibilities. In this article, we ask whether settler concepts of popular sovereignty can be rehabilitated sufficiently to do justice to the sovereignties of Indigenous peoples and so allow them to shape their rights and duties as citizens.
In the following part II, we discuss Indigenous peoples as a critical case for understandings of popular sovereignty. In part III, we define popular sovereignty as our theoretical starting-point. Thereafter, in part IV, we discuss and critique three common models of Indigenous popular sovereignty by means of democratic citizenship: (i) Indigenous participation in representative institutions, (ii) special representation in sovereign bodies through, for instance, quotas, and (iii) separate political bodies for Indigenous peoples. We conclude that all these manifestations of democratic citizenship are insufficient from the vantage point of Indigenous peoples as sovereign peoples. Instead, we argue in part V that popular sovereignty is better understood in terms of constituent power.

II. Indigenous peoples as a critical case for popular sovereignty
The language of popular sovereignty invites the belief that 'the people' only exists in the singular and that the legitimacy of the state and the rights of citizens depends on powers ultimately possessed by this singular entity that populates the territory of the state. But the existence of 'plurinational' or 'multinational' states reminds us that this assumption is not always true (see, e.g. Colón-Ríos et al. 2021;Keating 2002;Tierney 2007;Tully 1995).
Many constitutional orders are created on the assumption that two or more constituent 'nations' exist that share sovereignty within a single territory by various forms of powersharing, allocated vertically and horizontally (see, e.g. Choudry 2008;Lijphart 1977).
The notion that Indigenous peoples are 'peoples' with a distinct legal status has been firmly established by developments in international law over the last three decades. Indigenous peoples are international legal subjects vested with the right to selfdetermination, as manifested in the 2007 UN Declaration on the Rights of Indigenous Peoples (UNDRIP) (see, e.g. Allen and Xanthaki 2011). This right is, however, typically interpreted by settler and international institutions as a right to internal self-governance within the borders of already existing nation-states, even while avenues for unilateral secession in exceptional circumstances have not been completely closed off as a matter of law (see, e.g. Reference re Secession of Quebec 1998). As it stands, then, selfdetermination largely 'applies to indigenous peoples through domestic recognition and action' (Åhrén 2016, 118. See also, e.g. Human Rights Committee 2019a, [8.8], 2019b, [6.9]). Hence, in states where Indigenous peoples live, there are multiple peoples within the same jurisdictional territory, all with the right to self-determination and -we argueall as peoples entitled to exercise popular sovereignty, since 'self-determination is the international analogue to the idea of popular sovereignty' (Stilz 2017, 101). The idea that popular sovereignty vests in Indigenous peoples does not require that those peoples are or aspire to be states; rather that they be enabled to participate as peoples in negotiating the terms of their co-existence as citizens within the constitutional order by being equally empowered over that order.
Of importance to our understanding of popular sovereignty, the constitutional orders in territories inhabited by Indigenous peoples differ in significant ways from other 'plurinational' or 'multinational' states. First, Indigenous peoples are typically small in numbers relative to settler populations and are frequently territorially dispersed; often as a result of settler-colonial policies of targeted killing, dispossession and forced relocation. These features make unilateral secession or independent statehood -either as a live option or as a threat -much less feasible than it might be elsewhere (Dies 1993;Hannum 2006;Levy 2003). 1 For the most part, Indigenous and settler peoples are destined to share the same territory.
Second, and more important, the historical creation of contemporary states commonly reflected a deliberate attempt to disqualify pre-colonial sovereign peoples from participating in the constitutionalizing of the state, efforts that were and are perpetuated by legal doctrines of discovery, occupation or terra nullius (Cronin 2017;Pateman and Mills 2007). This is true also in states, such as the Nordic countries, that are not commonly referred to as settler states but where Indigenous peoples have been subject to 'internal colonization', via the encroachment of majorities onto Indigenous territory and the exclusion of these peoples from political processes. Indigenous peoples thus defy the constitutional ethos of the state as envisioned by settlers at the time of its inception. In spite, or because of this fact, settler law has assiduously directed itself towards the denial of fully fledged Indigenous sovereignty.
Historic treaties concluded between imperial or colonial states with Indigenous nations show that the status of Indigenous peoples as peoples and as property-holders was recognized by colonizing powers in some circumstances. The historic treaties of Canada, the United States and Aotearoa New Zealand, and contemporary treaties concluded in Canada, are cases in point. Subsequently, however, these treaties were (and are) deemed by settler institutions not to entail relations between sovereigns in international or state law, although many have been recognized as constitutional instruments (a point to which we return later) (McNeil 2018). For many Indigenous signatories, treaty-making is expression of their sovereignty that does not prospectively limit it, even where (as in Canada and Aotearoa New Zealand), settler authorities insist that treaties do not recognize, or else effect the cession of, pre-existing Indigenous sovereignty (Mutu 2018;Starblanket and Stark 2018, 179-80;Stark 2012, 125). Even in the United States, where tribal 'domestic dependent' sovereignty is acknowledged as part of US law, in that body of law the sovereign status of federally recognized tribes was diminished by their incorporation within the United States, and subsequently subordinated to the plenary power of Congress. In no case, from the perspective of settler law and theory, did treaties legitimize the constitutional order of the new settler state. Nor in any case does the constitutional order of the settler state depend for its legitimacy on these treaties, at least as viewed through the lens of settler institutions.
In settler states, then, prevailing ideas of popular sovereignty have been enlisted in colonial projects to deny the constitutive status of Indigenous peoples, by insisting on the existence of a unified sovereign people, within which Indigenous peoples are forcibly included, prior and on-going treaty commitments notwithstanding. 2 As noted by James Tully, the settler state took popular sovereignty to justify the suppression of cultural and political plurality, and so served to 'eliminate cultural diversity as a constitutive aspect of politics' (Tully 1995, 63). Experiences in states characterized by internal colonization have been similar, where the existence of Indigenous peoples as peoples has been denied or consciously neglected in the state-building process.
Indigenous peoples thus challenge prevalent understandings of popular sovereignty in two unique ways. First, by casting doubt on the unified understanding of the people behind contemporary democratic constitutions, especially given their recognition in international law as distinct peoples with a right to self-determination (Kasuhal 2017), and the recognition some receive in domestic law as historic communities with legal personality and pre-colonial property rights. Second, by virtue of their existence under conditions in which regular avenues of 'shared sovereignty' are unavailable because of the deliberate historic and ongoing denial of their sovereign status by incoming settler peoples. The case of Indigenous peoples consequently compels renewed attention to the conditions under which the legitimacy of a constitutional order can accord with popular sovereignty in a multi-people state.

III. Popular sovereignty revisited
There is little consensus on the institutional prescriptions that follow from principles of popular sovereignty. This is at least partly due to the ambiguity of the term itself (Morris 2000). In the remainder of this article, we explore two distinct conceptions of popular sovereignty; the first based on democratic participation, the second as constituent power. It is first necessary, however, to introduce what we take to be the core elements of various accounts of popular sovereignty.
The first is the claim that peoples should enjoy opportunities to participate in the making of political decisions. This core reflects the assumption that peoples have interests that are not fully exhausted by the quality of the institutions under which they live.
Peoples also have interests as 'decision makers' that are realized only if they are allowed the status as equal citizens with participatory rights in the making of decisions (Stilz 2017).
The second is the claim that peoples' interests as sovereign entities are respected only when they can participate in the making of decisions about the 'rules of the game' that define their rights as citizens. This points directly towards the significance of constitutional orders and is reflected in the assertion that 'the ability to engage in constitutional change is a fundamental act of popular sovereignty' (Schwartzberg 2007, 6). The 'constitution' can refer to documents named as such, to legal norms or conventions of particular significance, or to legal norms subject to rules of amendment (Elster 1995). In this context, by 'constitutional order' we mean the body of legal norms that define and regulate the legal powers exercised by political institutions. The importance of constitutional legal norms, in comparison to other legal norms, can be accounted for largely by reference to the distinction between primary and secondary rules (Hart 1962). The constitution is 'constitutive' of the political and legal system by including secondary rules that determine the normative powers of public bodies and the rules for making and revising primary rules (Onuf 1994).
The implication is that popular sovereignty is concerned with participating in the making of constitutional rules. This stipulation allows us to separate our working concept of popular sovereignty from other phenomena that are sometimes described by the same term. For instance, it is not uncommon to hold that more direct forms of popular participation are closer to the ideal of popular sovereignty. The idea is that direct participation -by referendum for example -allows collective decisions to be decided according to the will of the people to a greater extent than indirect forms of participation (Setälä 1999). Following the usage proposed here, direct participation is nevertheless distinct from popular sovereignty. What defines acts of popular sovereignty is the object of the decision: popular sovereignty is about popular participation in the creation and revision of the powers that the constitution confers to them as citizens. Popular sovereignty is not exercised by citizens through democratic institutions, it is manifested in the creation of the constitutional order that gives effect to citizenship and democratic forms of participation.
Thus, our usage of popular sovereignty is distinct from the conventional view that citizenship rights in a democratic state are sufficient and necessary for popular sovereignty to obtain. In the following section, we identify three versions of this democratic model of popular sovereignty in political practice, all falling short of recognizing Indigenous peoples' popular sovereignty: participation in parliamentary elections on equal terms; participation in parliamentary elections with special representation for Indigenous people; and, participation in elections for special institutions representing Indigenous peoples.

IV. Popular sovereignty as democratic participation: three models
As noted, the conventional view of popular sovereignty is that it is exercised by the citizens through procedures for democratic decision-making. In conditions when two or more peoples co-exist within the borders of the state, albeit with common citizenship status, this translates into three alternative models for the exercise of Indigenous popular sovereignty by means of democratic participation. The first -and also the most common approach to representation in multi-people states -is based on the idea that Indigenous peoples hold a share in the exercise of sovereign power as equal citizens in the democratic system. Equal rights and forceful anti-discrimination legislation may thus be regarded as the most important features of such a constitutional order.
After voting in favor of the UNDRIP in 2007, for instance, the Swedish government's representative declared that 'the realization of the right to self-determination could be ensured through [. . .] participation in democratic systems, such as the current Swedish system' (UN 2007). Likewise, when Australia cast its negative vote (Australia 2007), it confirmed that it 'supports and encourages the full and free engagement of indigenous peoples in the democratic decision-making processes in their country, but it does not support a concept that could be construed as encouraging action that would impair, even in part, the territorial and political integrity of a State with a system of democratic representative Government'. Similar understandings of how the right to selfdetermination ought to be realized have been expressed by other states, including Canada (Canada 1999), and the US (Lightfoot 2016).
The attribution of popular sovereignty to Indigenous peoples, following this model, is analogous to the reasoning that attributes popular sovereignty to peoples generally. Given that the parliament is empowered to revise or initiate proposed revisions to secondary rules, the parliament embodies sovereign power. The electorate embodies sovereign power whenever it is authorized to appoint the members of the parliament that is mandated to revise or initiate proposed revisions to secondary rules. Accordingly, when Indigenous peoples are included in the electorate, they are thus thought to be part of a collective exercising sovereign power.
In most parts of the world, however, Indigenous peoples constitute permanent minorities and are as such unlikely to influence electoral outcomes in directions that reflect their interests and aspirations. The impact of Indigenous peoples on the exercise of sovereign power by parliaments is therefore marginal, at best. When that is the case, democratic elections are extremely unlikely to ever produce influence for Indigenous representatives in national parliaments. And if no Indigenous representative is elected, there is literally no influence by Indigenous peoples in decisions by parliaments over secondary rules. Thus, the claim that Indigenous peoples exercise popular sovereignty just because they are citizens with access to the same rights to democratic participation as others appears consistent with the conclusion that Indigenous peoples have no effective influence over decisions about secondary rules at all.
The second model of popular sovereignty as democratic participation proposes a solution to this specific problem. Representative quotas in national parliaments for Indigenous people guarantee some degree of representation and presence in every decision taken by these bodies, including decisions on secondary rules. The dedicated seats reserved for Māori in the national parliament of Aotearoa New Zealand are a paradigmatic example. There are currently seven reserved seats for Māori representatives in the national parliament, a figure which is adjusted every five years to be proportionate to the number of voters registered on the Māori electoral roll (Electoral Commission 2018; Xanthaki and O'Sullivan 2009). Voters on the general role have elected a further 22 members of Māori descent across all major political parties, creating a cohort that in 2019 made up 24% of parliamentary seats, a number well above population parity for Māori, at 15% of the national population. By having dedicated seats, the Māori can participate in the decisions by the supreme law-maker concerning the making, revision or abolishment of secondary rules, including the terms of Māori representation. Guarantees of proportional representation, however, appear in a different light in the case of what Noel Pearson calls 'extreme' Indigenous minorities (Pearson 2014, 38). In Australia, for example, Indigenous peoples are outnumbered by settlers 35 to 1; in Canada the ratio of settlers to Indigenous peoples is 24 to 1, in the United States 67 to 1, and in a state like Sweden, the Indigenous Sámi population is outnumbered 450 to 1.
Even when Indigenous peoples achieve significant parliamentary representation they hold but a minority of the seats in parliament and run the risk of being outvoted in constitutional decisions, including decisions that impact directly on their special interests and rights (in the case of the Māori: the constitutional status of the 1840 Treaty of Waitangi). 3 As argued by Paul Patton (forthcoming), the nature of settler society virtually guarantees that Indigenous people remain 'consistently and systematically on the losing side in majority decision-making processes'.
The prospect of Indigenous peoples' exercise of popular sovereignty is worse still if we agree it requires political power -that is, the capacity to determine outcomes (Morriss 1987). The mark of sovereign peoples is the power to change the constitution they are governed by (Colón-Ríos 2010, 212. See also, e.g. Winterton 1998). Once the exercise of popular sovereignty is understood as the power to secure a people's preferred outcome, the notion that Indigenous peoples are entitled to popular sovereignty can only mean that Indigenous people should somehow have the capacity to change the constitutional order they are subjected to.
Following the tenet that capacity to effect constitutional change is a precondition for the attribution of popular sovereignty to a people, it is difficult to see how it could ever be attributed to more than one people represented in a single institution designed to aggregate the will of the people. This presents a puzzle that directly challenges the legitimacy of the states in which Indigenous peoples live, something Duncan Ivison describes as a 'genuine aporia -an irresolvable contradiction at the heart of liberal political orders' (Ivison 2017, 118).
The third model of popular sovereignty as democratic citizenship is designed to take the distinctiveness of Indigenous peoples seriously. The idea is that of creating separate Indigenous bodies to ensure political representation in the democratic system, and provide a way to fulfil Indigenous peoples' right to self-determination. Once again, the Swedish government's representative at the UN General Assembly presented this as a Swedish interpretation of the UNDRIP: besides participation in the Swedish democratic system, self-determination for Indigenous people could be realized through 'a consultative process between institutions representing Indigenous peoples and Governments' (UN 2007). The advantage of separate representative bodies is that Indigenous peoples can control the decisions made by those bodies. As members of the demos in the election of the Sámi Parliament in Sweden (in Swedish Sametinget), for example, the Sámi do not run the risk of being outvoted by members of the majority population. Moreover, the representatives elected to that body are fully able to determine the decisions made by it (within the limits established by the legislator, that is, the national Swedish parliament). Democratic forms of participation in the Sámi Parliament thus confer real power on the Sámi people, in the sense that they have the capacity to determine outcomes on matters within their jurisdiction.
Power is not necessarily sovereign power, of course. The obvious defect of the Sámi Parliament and other national level Indigenous bodies (such as the Australian Congress of First Peoples, the Canadian Assembly of First Nations and the New Zealand Māori Council) is that they typically lack the capacity to either influence or determine the decisions made about the substance of the constitutional order itself. In effect, bodies for Indigenous representation are powerless with respect to decisions that regulate their own constitutional status. For instance, to translate the Sámi representative body in Sweden into Sámi Parliament in English is gravely misleading, since it -like its counterparts in Finland and Norway -lack legislative power. The powers exercised by the Sámi Parliament (and the powers vested in it) are in other words regulated by laws over which the parliament itself has no power, as is the case for national-level Indigenous bodies in other states. Its actual political powers have been delegated from the state and can thus be unilaterally revoked by the state (Mörkenstam, Selle, and Valkonen Mörkenstam, et al., forthcoming). What has been institutionalized -to use the words of Fiona MacDonald (2014) in her analysis of Indigenous autonomy in Canada -is an order where the Sámi 'are burdened with responsibilities for many complex socio-political situations despite a lack of meaningful decision-making power', while at the same time the Swedish government is left 'dangerously unaccountable for [its] continued role in the well-being' of the Sámi.
The vulnerable status of Indigenous legal bodies is illustrated also by the decision of the Australian government to abolish the Aboriginal and Torres Strait Islander Commission (ATSIC) in 2005. The ATSIC was established as a national level advisory body to the government with budgetary responsibility for administering Indigenous programs, and its main objective was to enhance the participation of Indigenous peoples in the formulation of government policies. In its role as an advisory body ATSIC had no power to exert influence on the secondary rules in the legal system, and when declared a failure in improving the socio-economic situation for Indigenous peoples, the government decided unilaterally to terminate its mandate, deeming it a vehicle for racialized separatism threatening national unity (Robbins 2010). The government in power noted '[w]e believe very strongly that the experiment in separate representation, elected representation, for Indigenous people has been a failure. We will not replace ATSIC with an alternative body' (PM Transcripts 2004).
Current Australian proposals for a constitutionally guaranteed 'Voice to Parliament' (Uluru Statement 2017), through an Indigenous representative body, may empower the First Nations of Australia and guarantee them a fairer say, but are far from fulfilling an ideal of Indigenous popular sovereignty as constituent power exercised on terms equal to those of other peoples. Indigenous Australians would be vastly outnumbered in the referendum required to affect constitutional change. Moreover, Australian Prime Ministers to date have rejected the proposal to constitutionalize an Indigenous representative body on the grounds that it would violate equality principles attending citizenship status, sometimes also mischaracterizing the representative body as one with legislative authority: '[o]ur democracy is built on the foundation of all Australian citizens having equal civic rights [. . .] A constitutionally enshrined additional representative assembly for which only Indigenous Australians could vote for or serve in is inconsistent with this fundamental principle' (Turnbull 2017).
As the above illustrates, the theoretical gaps attending the exercise of popular sovereignty by multiple peoples have real-world consequences. The assumption that colonization and the assertion of sovereignty over Indigenous territory suffices to ensure that Indigenous peoples are part of the settler body politic, incapable for exercising their own popular sovereignty, is a remarkably persistent feature of settler state legal and political thought. In 2020, the High Court of Australia considered the constitutional status of Aboriginal peoples in the course of deciding that non-citizen Indigenous persons could not be classified as 'aliens' in the terms of the Constitution, and so could not be deported. Repeating precedential statements denying Indigenous legal sovereignty, one judge took the opportunity to exclude the possibility of Indigenous popular sovereignty. The dominance of the idea of popular sovereignty in Australian constitutionalism, she held, meant that the 'people of Australia necessarily includes Indigenous peoples' because [r]ecognition of Indigenous peoples as part of the 'people of Australia' denies that Indigenous peoples retained, or can now maintain, a sovereignty that is distinct or separate from any other part of the 'people'. (Love v Commonwealth 2020, per Gordon J, [356]) This form of 'recognition', premised on coercive unilateralism, offers to Indigenous peoples what Glen Coulthard might call 'the cheap gift of political and economic inclusion' (Coulthard 2014, 173).
If a people is to partake in the exercise of sovereign power, it must hold power over the rules that determine and define the powers conferred on public institutions. In states where Indigenous peoples are represented by their own local institutions and governments alongside national-level representative Indigenous bodies, agreements and treaties could specify the rules that determine the ways in which the institutions of the majority society and Indigenous institutions cooperate (Curry 2004, 157). These could bring us close to meeting the criteria for Indigenous popular sovereignty to be exercised as constituent power, provided that we can properly regard these agreements as constitutive of the state. Agreements on power-sharing should, however, be constitutionally protected, so that they can only be changed by a further exercise of popular sovereignty by the peoples party to those agreements, precluding settler majorities to override Indigenous rights and interests through the use of referenda.
Two examples illustrate the significance of this risk. In 2002, the Premier of British Columbia, who resisted the approach to treaty-making advanced by the federal government, (Campbell v British Columbia 2000), called a referendum seeking the agreement of the electorate to his preferred treaty negotiating criteria (Fetzer 2016). The 2002 Aboriginal Treaty Negotiations Referendum was vigorously opposed by British Columbian First Nations who saw it as a move to undercut their capacity to secure constitutionally protected treaty rights. This proposal would have ensured that Indigenous peoples would be outvoted in setting the terms of negotiations by a 95% settler majority. The 1992 'Charlottetown Accord' proposals to recognize First Nations governments as 'third order' of Canadian government were regarded by many Indigenous leaders as promising (perhaps even coming close to an exercise of constituent power, having been negotiated with several Indigenous national representative bodies at the table) but was defeated in the accompanying national referendum (Turpell 1993). Similar questions have arisen in Aotearoa New Zealand, where political parties have campaigned on the promise to hold a referendum on the proposed abolition of the Maori Parliamentary seats (Muir 2017). Both examples show the inability of direct democracy to decide matters that involve agreements and prior commitments between settler states and Indigenous peoples and should be determined by negotiation between the parties. Operationalizing the question of whether an Indigenous people is able to exercise their sovereign power is, consequently, to ask whether they are empowered to determine the terms of their own ongoing participation in the constitutional order.

V. Popular sovereignty as constituent power
Existing democratic models leave few real alternatives for the recognition of Indigenous peoples' popular sovereignty. In fact, inclusion of Indigenous peoples as citizens forces them into -in the words of Hans Lindahl (2016, 22) -a 'political dilemma': where political and legal recognition of their distinctiveness is obtained only at the price of being 'participants in a project with which they do not want to be associated'.
In this section, we contemplate ways to transcend this dilemma. We argue that the realization of Indigenous claims to popular sovereignty requires a model that focuses on the origins of institutions for democratic citizenship; rather than on the exercise of democratic citizenship as such. Indigenous sovereignty must find expression in the design of constitutional rules and institutions that determine the rights and statuses of citizens, not just in participatory activities taking place within a pre-defined constitutional framework.
Following a distinctive tradition in the history of popular sovereignty, the powers of the people must be distinguished from the powers vested in institutions and legal arrangements. Legal powers are constituted powers, whether these are legal powers exercised by elected representatives or legal powers exercised by citizens. By contrast, constituent power is the capacity by bodies that are not part of the state to create and revise the constitution and the powers vested in legal institutions. Understood in terms of constituent power, then, the sovereign powers of the people must be located in a body distinct from legally construed citizenship in order for the constitution to be 'an expression of the constituent power of the people to make and re-make the institutional arrangements through which they are governed' (Loughlin 2014, 231. See also e.g. Colón-Ríos 2014).
The constituent power of the people is also distinct from legislative and representative institutions. Popular sovereignty is the right of a collective 'We' to create and authorize the constitutional rules that regulate the institutions in which we participate as individuals. As explained by Frank Michelman (1999Michelman ( , 1626 the idea of popular sovereignty as constituent power envisages the act of superior lawmaking as emanating from 'a People' with a 'collective character, or political identity' -a 'collective political self'. The source of constituent power is, however, disputed. Carl Schmitt famously located it in that body with the capacity to 'determine the exception', thus transcending the constitutional order. Alternatively, constituent power is a process; manifested in the never-ending political struggle 'about who represents the people' (Loughlin 2014). The idea of constituent power explains why even the power to revise the constitution through procedures defined by the amendment clauses of the constitution is inadequate. The point is that the constitutive powers of the people must be unlimited; not subject to legal constraints. Since constituent power is the power to re-arrange the constitutional order, the sovereign people must be a non-legal entity with constitutional power unrestricted by the constitution (Lindahl 2016, 149). The powers attributed to 'the people' as a category defined by constitutional law, however, can be no more than powers constituted (Pasquino 2017;Colón-Ríos et al. 2021).
To accept that Indigenous peoples hold constituent power is to accept that Indigenous peoples always retain the power to renegotiate and rearrange institutionalized structures of public power, irrespective of the extent to which these powers are legally recognized by those institutions and the officials that populate them. One way of thinking about this prospect is to accept that post-colonial settler-state legitimacy requires that Indigenous peoples are able to exercise constituent power commensurate with their right to self-determination, i.e. that they are able to participate as sovereign peoples in the design of the constitutional arrangements that facilitate their political participation. In such a case, the constituent power to rearrange the constitutional order cannot legitimately be exercised by one people unilaterally, at the expense of the other(s). Inclusiveness is critical for constituent power to legitimize the constitutional order; 'the legitimacy of the fundamental norms and institutions depends on how inclusive the participation of the citizens is during the extraordinary and exceptional moment of constitution making' (Kalyvas 2005, 237). In a multi-people context, this requirement does not simply mean that all citizens should be included. Instead, it implies that all peoples must be included on equal terms. Just as the inclusion of some citizens is not allowed to override the inclusion of other citizens, the inclusion of one people is not allowed to override the inclusion of other peoples.
In fact, the attribution of constituent power to Indigenous peoples serves to highlight many sites of Indigenous mobilization as instances where sovereign power is at stake. In cases of civil disobedience, political demonstrations, or even (so-called) violent resistance, Indigenous people are not only asking for recognition as sovereign entities but are in fact manifesting themselves as such in opposition to the asserted sovereignty of the settler state (see, e.g. Borrows 2016a; Coulthard 2014; Muldoon and Schaap 2012). Indigenous scholars and leaders have made this point forcefully, in elaborating political theories of Indigenous resurgence, in which the state's offer of subsovereign forms of recognition should be refused, on the basis that it has no legitimate authority to issue such an invitation (see, e.g. Corntassel 2012;Coulthard 2014;Simpson 2014, 54. See also Ivison 2017, 122). As Coulthard argues, Indigenous peoples should reject forms of recognition that are premised on the delegation of state power, including those that are actualized in land claim settlements, economic development initiatives, and self-government agreements, and instead turn 'toward a resurgent politics of recognition premised on self-actualization, direct action, and the resurgence of cultural practices that are attentive to the subjective and structural composition of settler-colonial power' (Coulthard 2014, 23-24). According to models of Indigenous resurgence and refusal, the sovereignty of the people is extra-legal and therefore not confined within existing political institutions. In order to detect the sovereign powers of the people, as they are manifested in struggles for constituent power, we must look elsewhere.
The potential of constituent power in this context is that it elucidates the significance of the procedures for the creation of the institutions that enable Indigenous participation as citizens. In the absence of such procedures, any institutional arrangement for Indigenous participation necessarily falls short of realizing the ideal of popular sovereignty. To recognize Indigenous peoples as sovereign is, then, not to define any specific model of democratic participation as required but to insist that the choice of any such model is made in a process where Indigenous peoples are recognized as sovereigns and therefore treated as political equals (Mörkenstam 2015). It also requires that in order to be legitimate, changes to the powers conferred on institutions must be renegotiated between sovereign peoples. In a territory where two or more peoples co-exist, the constitutional framework that determines the rights of citizens cannot be determined by one people unilaterally.
The exercise of constituent power is not enough to secure democratic legitimacy, however (Harel 2016). In order to consider outcomes of a political process legitimate, the procedure must be conditioned by equality and inclusiveness (see e.g. `Beckman 2009;Goodin 2007;Viehoff 2014). In places where Indigenous peoples and the majority population are (or should be) both recognized as sovereign peoples, the best guess is that the democratic legitimacy of a constitutional order is determined procedurally. A procedural account of constituent power, consistent with the principles of popular sovereignty, envisages a process of negotiation between the peoples involved by which they are afforded equal status in the determination of the shared constitutional order. James Tully argues for a similar understanding, noting that negotiations between settler and Indigenous peoples as equals is a 'democratic way' to achieve reconciliation 'acceptable to both parties': On this view reconciliation is neither a form of recognition handed down to Indigenous peoples from the state or a final settlement of some kind. It is an on-going partnership negotiated by free peoples based on principles they can both endorse and open to modification en passant (Tully 2008, 223). This negotiation amounts to the recognition of a 'nation-to-nation relationship' within a state, in which more than one people has constituent power, and is a relationship that ought to be publicly acknowledged in the constitution. As Tully puts it elsewhere, 'the primary good of self-respect requires that popular sovereignty is conceived of as an intercultural dialogue' (Tully 1995, 190). This idea of perpetual negotiation suggests a specific procedure to enable the creation of institutions in a multi-people state that can properly facilitate Indigenous democratic participation as sovereign peoples.
The first step in this procedure would be, in line with Tully's argument, to recognize Indigenous peoples as having a political standing equal to that of nation-states (Mörkenstam 2015). This, of course, is easier said than done, given the extent of the powers the state has amassed to itself. Where no constitutional relationship has yet been established, this process of negotiation and renegotiation depends on the willingness of both parties, and can only be established by political means. What we are proposing here are the minimum conditions that could secure the legitimacy of a state established on Indigenous lands. It flows from their sovereignty as peoples that they should have had a determinative role in constituting the new state. The legitimacy of settler states remains in question to the extent that this condition is not met. Indeed, as Ivison (2006) has observed, state-Indigenous relationships 'are quasi-international relations since, as we have seen, although these groups lack external sovereignty, they claim jurisdictional rights or political authority independent of the state'. A status as political equals would grant Indigenous peoples a position from which they could negotiate the conditions of their relation to settler peoples on their own terms, i.e. it would enable them to exercise popular sovereignty by playing a constitutive role -as a distinct people -in shaping the body with 'final and absolute political authority in the political community' (Hinsley 1986, 26). When two or more peoples live within the same territory, this sovereignty is necessarily delimited by the popular sovereignty of other peoples. We may well understand such a delimitation of peoples' sovereign powers in terms of mutual delegation: some sovereign powers are delegated or shared with other people(s), while the constituent power remains intact, enabling revisions and renegotiations. This idea of mutual delegation would require each party to concede some limitations on the scope of their constituted powers, while retaining their constituent power as sovereigns.
The second step in the procedure is the actual negotiations. At this stage, the parties need to agree on the criteria through which they accept or reject the other's representative credentials, a process that can be fraught in this context, as it is in other types of political engagement Gover and Baird (2002). It is possible, although perhaps not likely, that negotiations conducted on this basis would result in institutions akin to those that emerge from one of the three models of democratic citizenship discussed above. In that case, Indigenous peoples' participation as citizens would amount to an expression of their sovereign status as a people, since the powers vested in them as citizens would be the result of their exercise of constitutive power. If procedural requirements are satisfied, so that Indigenous peoples participate as sovereign equals, any arrangement for the democratic participation of Indigenous peoples (within a range of permissible democratic alternatives) is legitimate. This means that it is not necessary, though possible, that democratic institutions provide separate avenues for the political participation of Indigenous peoples. What matters is how they came about; that the arrangements are the product of a constitutional agreement coauthored by Indigenous peoples negotiating on equal terms with their non-Indigenous counterparts (cf. Waldron 2011). A further issue that the parties need to address the openness or closure of the arrangement to future changes. As sovereign equals, amendments of the arrangements agreed on should at least be subject to approval by both parties.
Treaties concluded between settler states and Indigenous peoples demonstrate the potential for bargains of this kind to form the basis of a legitimate post-colonial constitutional order. Historic and contemporary treaties have acquired a constitutional character in those states in which they exist, as a matter of legal doctrine as well as political theory, even while breaches of their terms by states are not uncommon (Borrows 2016b, 798). New Zealand's Treaty of Waitangi has been described as the 'Māori Magna Carta' (McHugh 1991;Williams 2015), and 'the founding document of the state' (Cooke 1992). Canadian historic and contemporary treaties are acknowledged in settler law as representing 'an exchange of solemn promises between the Crown and the various Indian nations' that is 'sacred', 4 and protected by the Canadian constitution (Canadian Constitution, s 35).
In line with agreements as a critical part of the constitutional order of settler states, the multilateral system of state-Indigenous treaties in North America has been described by Indigenous scholars as a type of 'treaty federalism' or 'treaty order' in which the accretion of multiple treaty arrangements between Indigenous nations and national and regional settler governments forms the basis of constitutional powers (Henderson 1994(Henderson , 2002. Tully (1995) likewise writes of the promise of 'treaty constitutionalism' as an alternative to the strictures and inadequacies of popular sovereignty in settler societies. The Indigenous vision of treaties as constitutive of settler states qua states is an idea that provides a path towards legitimacy for those states.
While the interpretation of existing treaties by settler institutions falls short of recognizing Indigenous sovereignty, treaties are undeniably agreements concluded between peoples and between nations. They therefore offer a mechanism for augmenting settler state constitutionalism so that it enables the expression of Indigenous popular sovereignty, by providing a locus for Indigenous constituent power in the creation and revision of constitutional arrangements. The challenge is, as we have noted above, to ensure that the parties to treaty negotiations participate on terms that could plausibly be said to be 'equal'. The failure of settler states to acknowledge Indigenous constitutive power in treaty-making processes and in the content and subsequent interpretation of treaties indicates that these fall short of the ideal of popular sovereignty heralded by democratic constitutionalism. They are not premised on negotiations between sovereign equals about the constitutional arrangements that govern their co-existence. Indigenous sovereignty, however, persists despite the inadequacies of settler law and theory.

VI. Concluding remarks
The aim of this work is to show what the ideal of popular sovereignty requires when more than one people is recognized within the same jurisdictional territory. Instead of focusing on the institutions through which sovereignty is exercised (constituted power) we should turn to the most fundamental democratic power, i.e. constituent power. This is what popular sovereignty requires in multi-people states, given the constraints of political stability and territorial integrity.
The claim that popular sovereignty requires constituent power is vulnerable to the objection that it is 'voluntaristic'. The charge is that a collective choice of constitutional order is not sufficient to conclude that it is either just or legitimate (Duke 2017;Larmore 1999;Lee 2016;Stacey 2016). Interpreting popular sovereignty in terms of constituent power does not, therefore, help explain when and to what extent a people confers legitimacy on public institutions.
It should be noted, however, that our view is not merely that Indigenous peoples and settler peoples should be able to decide, together, the constitutionalized procedures for political participation but also that the decision should be made in a negotiated process where participants are recognized as equals. These are normative conditions that are not simple articulations of the ideas of constituent power, but that are more accurately understood as expressions of standards of democratic legitimacy. The ideal of popular sovereignty requires that the constitutional order derives from the people, which we understand to mean that it responds to the people as the constituent power. In order for Indigenous peoples to be recognized as sovereign peoples, they must also be afforded constituent power. But in order for the exercise of constituent power to play the legitimizing role typically associated with popular sovereignty in contemporary constitutionalism, the constituent powers of Indigenous peoples and the majority population must be conditioned by principles of equality. In cases where Indigenous peoples compete with the majority population for constituent power, we advocate a solution where all peoples participate as equals in the process of negotiating or renegotiating a common constitutional framework. Consistent with ideas of popular sovereignty, public power derives from the people when these requirements are met, even when there is more than one sovereign people in the same territory.