No cosmopolitan morality without state sovereignty

Introduction

Cosmopolitan normative commitments are often considered incompatible with recognition of state sovereignty as a basic principle of international law. Although cosmopolitans do not necessarily reject the normative importance of sovereignty completely, proponents of an influential strand of contemporary cosmopolitanism that I call anti-statist cosmopolitanism tend to ascribe to it a mere derivative significance, dependent on its instrumental value for protecting human rights. In this article, I take issue with this kind of cosmopolitanism, arguing that there is a stronger connection between individual freedom and state sovereignty. Taking a conception of justice informed by Kant’s philosophy of right as a point of departure, I claim that state sovereignty is not only compatible with, but also essential to the recognition of individuals as units of ultimate concern. Justice among persons, understood as each person’s right to be independent from subjection to other people’s arbitrary choices, presupposes that their interaction is regulated by coercive public institutions (that is, state authorities). Accordingly, respect for the rights of persons requires respect for sovereignty, which entails norms such as non-interference and self-determination.

An important step in the argument is to challenge the specific way in which anti-statist cosmopolitans typically frame or conceptualize discourse about justice. Borrowing terminology from Iris Marion Young, I argue that most anti-statist cosmopolitan positions tacitly presuppose a problematic distributive conception of justice, ¹ which, briefly put, is a conception where justice is understood as the fair allocation of certain pre-politically defined outputs (rights). My suspicion is that the ascription of a mere derivative significance to sovereignty is a consequence of adherence to such a conception. I therefore argue that justice is better conceptualized with a direct focus on interpersonal relations, and that Kant’s idea of an innate right to freedom is a promising alternative in this regard. Conceptualizing justice with a primary view on interpersonal relations rather than outputs not only brings us closer to what justice is really about, but also makes it easier to see why there is an internal connection between individual freedom and state sovereignty.

The present defence of an internal connection between individual freedom and state sovereignty rests on a non-remedial view of the need for a state-sanctioned system of law and justice. Based on an understanding of Kant’s philosophy of right as a free-standing doctrine vis-à-vis his ethical theory, ² a central claim is that subjection to public authority with coercive power is the only way individuals can interact on terms of equal freedom, ‘however well disposed and law-abiding human beings might be’. ³ On this view, the most fundamental reason we need coercive public institutions is not Humean or Rawls-ian ‘circumstances of justice’ – that is, moderate scarcity and limited benevolence. ⁴ Rather than a remedial response to a troubling human condition, such institutions are partly constitutive of interacting persons’ equal independence. Absent coercive public institutions, structural features of the situation would cause those who cannot avoid contact to subject each other to arbitrary choice. For this reason, respect for each person’s right to independence requires subjection to public authority, even if resources are abundant and interacting persons are generally well disposed towards each other. ⁵

An implication of this view is that one can consider sovereignty as an international counterpart to freedom in the domestic realm. This view is sometimes dubbed the domestic analogy, since it implies that states have a right to territorial integrity much in the same way that individuals have a right to bodily integrity. Such an analogy is often criticized, either for overlooking important differences between states and individuals or for putting undue emphasis on communal integrity. The present defence of the analogy is not susceptible to this kind of criticism. I do not presuppose that states have moral faculties similar to those of human agents. Nor do I ascribe any independent moral value to community. States are instead understood as necessary institutional frameworks or arenas for realizing individual freedom, and being such arenas they have a right to be protected by the principle of non-intervention. That being said, I share the cosmopolitan commitment to the idea that all persons are equal units of normative concern generating obligations on every other person. The present defence of state sovereignty is in other words cosmopolitan in nature.

The argument proceeds as follows. In section I, I present the basic position of anti-statist cosmopolitans. In section II, I criticize the distributive conception of justice implicit in this position. In section III, I present an alternative relational conception, taking Kant’s innate right to freedom as a starting point. In section IV, I explain why this conception also implies that there is an internal connection between individual freedom and state sovereignty. In section V, I argue that the line of reasoning pursued in the two former sections provides a defence of the domestic analogy that circumvents the criticism typically raised against such an analogy by anti-statist cosmopolitans. Finally, in section VI, I discuss how defending non-intervention as a basic principle of international law leaves room for interventions in order to stop barbaric and inhumane use of force.

I Anti-statist cosmopolitanism

By cosmopolitan normative commitments, I understand commitments entailed in the core idea of moral cosmopolitanism – the idea that each person is to be recognized as an equal unit of concern generating obligations on every other person. Thomas Pogge has spelled out this idea by identifying three features uniting diverging strands of cosmopolitanism. Individualism: the ultimate units of concern are individual human beings or persons rather than human groups of various sorts; universality: the status of ultimate unit of concern attaches to every living human being equally; and generality: this special status has global force, which is to say that persons are ultimate units of concern for everyone. ⁶

Moral cosmopolitanism is commonly distinguished from institutional or legal cosmopolitanism, ⁷ which refers to positions advocating institutional schemes that bring states under the authority of some kind of supranational agency. Such advocacy is not implied in moral cosmopolitanism, but is not ruled out by it either. Commitment to moral cosmopolitanism can lead to support for cosmopolitan institutional schemes, but there is no necessary relation between the two. ⁸

In view of this openness with regard to institutional issues, there need not be any conflict between moral cosmopolitanism and an international legal order of sovereign states. The latter, sometimes dubbed a ‘statist’ order, is an order where all states qua sovereigns have legal standing and are recognized as equals. This means that they are formally subject to the same general rights and duties, most importantly the right to self-determination and the correlative duty of non-intervention. In a statist order, sovereignty implies that a collective actor has legal personality, and thereby can be a subject of international legal processes and a party entering into international treaties. It also implies the entitlement to organize legislative, executive and adjudicative institutions as it sees fit within a specific territory as well as the obligation to respect the territorial integrity of other sovereigns. Ideas congenial to a statist order are clearly present in the UN Charter, where paragraph 2(4) says that ‘members shall refrain in their international relations from the threat or use of force against the territorial integrity or political independence of any state’ and paragraph 2(7) says that nothing in the charter ‘shall authorize the United Nations to intervene in matters which are essentially within the domestic jurisdiction of any state’.

Despite the compatibility of moral cosmopolitanism and a statist international legal order in principle, the latter is today challenged by what I have termed anti-statist cosmopolitanism. Advocates of this strand of cosmopolitanism – they include philosophers such as Brian Barry, Charles Beitz, Allen Buchanan, Simon Caney, Darrel Moellendorf and Fernando Tesón – support global legal reforms in a decisively individualistic direction. The idea is that we should move away from an order based on the sovereign equality of states toward an order where respect for basic human rights serves as the exclusive criterion for judging the legitimacy of political and legal institutions. This view is based on the idea that political borders have no fundamental normative significance or the idea that there should be congruence between domestic principles of justice and international or global principles of justice. ⁹ Whatever principles of justice apply internally to states should also apply in the international realm. And since anti-statist cosmopolitans usually conceptualize justice in terms of human rights, so ‘the core of justice, protection of human rights, should be a primary goal of the international legal system’, ¹⁰ much in the same way as protection of human rights is the standard by which we ought to assess domestic political orders. State sovereignty is thereby reduced to an instrumental value whose importance is dependent on its effectiveness in promoting and protecting basic human rights. ¹¹ Individuals, not states, should be recognized as the ultimate subjects of international law, whereas the international standing of states should depend on the legitimacy of their domestic orders.

An important implication of this view is a rejection of non-intervention as a basic international norm. Given the normative primacy of individuals, protecting or promoting basic human rights is considered a just cause for intervention – including military intervention. Some, but not all, proponents of this view even argue that such interventions are not only permissible, but a duty on the part of states capable of helping those whose basic rights are violated. ¹² This is not to say that human rights violations taking place on the territory of a state provide sufficient justification for military interventions. All interventions are subject to standard jus ad bellum constraints: the use of military force must have a reasonable prospect of success; be a means of last resort; stand in proportion to the injustice it is meant to rectify; and so on. While this limits the scope of cases where interventions are justified, the norm of non-intervention is not, however, recognized as a basic norm governing international relations. As Charles Beitz puts it, ‘there is a right against intervention, but…it does not apply with equal force to all states’. ¹³ Sovereignty is a derivative of the more basic concern with justice to persons, and ‘only just states deserve to be fully protected by the shield of sovereignty’. ¹⁴

By the same token, the validity of claims to self-determination, as raised by former colonies in the 20th century, depends on whether or not liberation would be favourable with regard to reducing injustice in the relevant territory. People living under foreign rule can invoke no fundamental right to govern themselves against colonial powers. Self-determination, like non-intervention, is not a basic principle. It is just ‘a means to the end of social justice’. ¹⁵ Only if there is reason to believe that decolonization will lead to a less unjust society is there a right to self-determination.

It seems reasonable to say that anti-statist cosmopolitanism belongs to what Gerry Simpson has called a tradition of ‘liberal anti-pluralism’ characterized by ‘lack of tolerance for non-liberal regimes’. ¹⁶ Transforming sovereignty into a function of a state’s human rights record implies discrimination between states on the basis of their internal features. In addition to what I have already mentioned, such discrimination is reflected in proposals that would pull the international system in a less egalitarian direction were they to come into effect. These include proposals that representation in the UN should be restricted to democratic states that respect human rights, ¹⁷ or that there should be established a coalition of democratic states that under certain circumstances can trump the UN Security Council with regard to authorization of preventive use of force. ¹⁸ They also include proposals that regime change, or advancing justice in the basic structure of states, should be acknowledged as a just cause for military intervention. ¹⁹

Recently, Jean Cohen has criticized this anti-statist trend among contemporary cosmopolitans for being ‘normatively flawed and politically dangerous’. ²⁰ In her view, it is a form of cosmopolitanism that risks becoming an imperial ideology of powerful states in need of an excuse for going to war and, more generally, seeking an exceptional status for themselves. The crucial mistake of the anti-statist cosmopolitans, she further claims, is that they seek cosmopolitan reforms without acknowledging the fundamental significance of the sovereign state. ²¹ I agree with her on both points. The anti-statist cosmopolitan project seems to bring with it the risk of undermining one of the most important innovations of 20th-century international law: the prohibition against aggressive war. And contrary to the advocates of this project, I think a sound defence of human rights should not come at the price of degrading sovereignty to a function of a state’s human rights record. It is a mistake to construe the relation between individual freedom and state sovereignty in such a way that whoever ascribes fundamental normative significance to the one is compelled to ascribe a derivative or secondary significance to the other. We should avoid understanding the two as hierarchically related concepts. In the following, I will therefore suggest a way in which individual freedom and state sovereignty can be understood as co-original aspects of one and the same normative package, so to speak.

II A critique of distributive conceptions of justice

How can the normative tenets of moral cosmopolitanism outlined in section I be reconciled with the recognition of self-determination and non-intervention as fundamental principles of international law? An important first step is to call into question what can be termed a distributive conception of justice implicit in the anti-statist cosmopolitan view. I do so partly because I assume that it is precisely because they understand justice in distributive terms that anti-statist cosmopolitans cannot attribute more than an instrumental value to sovereignty. I also do so because I believe such conceptions are misleading and tend to produce erroneous reasoning about justice. ²²

Characteristic of a distributive conception is that justice is defined in terms of fair allocation of certain outputs. In the words of Iris Marion Young, distributive conceptions define ‘social justice as the morally proper distribution of social benefits and burdens’. ²³ Precisely what are regarded as morally relevant outputs or ‘benefits and burdens’ vary somewhat, but typical examples are civil and political rights, duties, material resources, and opportunities.

As far as anti-statist cosmopolitans are concerned, the output to be distributed is basic human rights grounded in certain fundamental human interests. In other words, the term ‘justice’ refers to the realization of a set of rights, usually including the right to life, the right to security of the person, the right to freedom of belief, the right to freedom of expression, the right to freedom of association, the right to freedom of movement, the right to protection against enslavement, the right to due process and equality before the law, the right to means of subsistence, and so on. These rights are justified as protections of interests of fundamental importance to individual human beings. Such fundamental interests can refer either to personal autonomy or to well-being and human flourishing. On this latter point there seems to be some disagreement among the anti-statist cosmopolitans that I have referred to. ²⁴ Yet whatever differences there are regarding the justificatory ground of rights, a common idea is that human rights are protections of whatever fundamental interests one takes as a starting point. Whoever is seriously committed to justice should seek to establish conditions that secure the non-violation of these rights.

This way of conceptualizing justice has an impact on what role one ascribes to legal and political institutions, not least the institutions that make up a state. Insofar as one thinks of justice in terms of allocating morally desirable outputs, institutions can serve only as more or less useful means with which we approximate these allocations. By that, I do not suggest that those who adhere to a distributive framework consider legal and political institutions to be of no or little significance. ²⁵ The point is rather that conceptualizing justice in terms of pre-politically defined outputs seems to imply that institutions and justice are externally related as means to an end. The reason for having institutions exercising the powers of making, applying and implementing laws is to make it more likely that the right results are realized, and the legitimacy of such institutions depends on their effectiveness in this regard – that is, their effectiveness in generating the morally desirable outputs.

Such a view on institutions is easily traceable in the writings of anti-statist cosmopolitans. It seems to be implied in the claim that state sovereignty is an instrumental value dependent on its effectiveness in promoting and protecting human rights. It is unambiguously expressed by Brian Barry: ‘the value of any political structure…is entirely derivative from whatever it contributes to the advancement of human rights, human well-being, and the like’. ²⁶ In a similar vein, Allan Buchanan emphasizes the ‘teleological’ nature of moral reasoning about institutions. Even if it need not be guided by the goal of maximizing welfare or happiness, and even if all efforts at achieving morally worthy goals should be subject to deontological constraints, such reasoning is nevertheless fundamentally goal-guided, in the sense that assessment of institutions takes the form of evaluating the institutions’ effectiveness in achieving the end they were made to achieve. ²⁷

The distributive framework for thinking about justice implicit in the anti-statist cosmopolitan view provides us with a problematic ‘image’ of justice. It should therefore be questioned. Importantly, by conceptualizing justice within such a framework one tends to lose sight of the fact that justice is a concept that applies only to interpersonal relations. Whatever the demands and entitlements of justice are, they can never apply to persons living isolated from other persons. I believe few people would deny this. Yet the relational nature of justice is played down to the extent that justice is conceptualized in terms of distribution of outputs. If justice is understood primarily as a question regarding proper allocation of goods or rights, persons are first and foremost seen as recipients of justice. What a person has a right to is in the first place specified independently of his or her relation to other persons. Only in a second step, after clarifying what output each person can rightfully lay claim to, do other people come into the picture as those against whom claims of justice can be made. In my view, this is to distort the phenomenon at hand. It is a misrepresentation that tends to cause erroneous reasoning about justice.

One kind of distortion caused by conceptualizing justice in distributive terms is the blurring of important distinctions in a way that severs the link between demands for justice and actual injustice. A primary focus on outputs does not seem to allow one to distinguish adequately between cases where people suffer as a consequence of natural events and cases where people suffer as a consequence of what other people do to them. Nor does it seem to allow one to distinguish adequately between cases of rights violations due to the exploitative acts or practices of other people and cases of rights violations due to our own acts and practices. This is not to say that someone adhering to a distributive point of view cannot recognize these distinctions or assess the various scenarios differently. Yet inasmuch as justice is identified with a specific output it seems to follow that all the cases raise justice-based demands on the ‘supply-side’, so to speak. Since what matters is the realization of a certain pattern of distribution, it is in each case required that we remedy the bad situation of those who suffer in order to fulfil our duties of justice. This is to conflate what we owe to others as a matter of solidarity with what we owe to others as a matter of justice. ²⁸ It is a confusion of aid to others out of sympathy for their suffering with righting a wrong. Such confusion is reflected in the view that we have a duty to assist militarily people who are denied basic human rights by their government, a view defended by many anti-statist cosmopolitans. ²⁹ It is also reflected in Allen Buchanan’s claim that we have a ‘Natural Duty of Justice’ to ensure that all persons have access to institutions protecting their basic rights even if we are not interacting directly or indirectly (via institutional schemes) with these persons. ³⁰

Another, and in this context more important, distortion caused by adherence to a distributive conception of justice, at least in the specific form of anti-statist cosmopolitanism, is insufficient attention to the issue of who can legitimately decide how abstract principles of justice should be specified, applied and implemented in particular cases. With Raymond Geuss, one could describe distributive conceptions as approaches that ‘complete the work of ethics first, attaining an ideal theory of how we should act, and then in a second step…apply that ideal theory to the action of political agents’. ³¹ Characteristic is a primary focus on what are appropriate principles of justice. What matters is that justice is done. The questions ‘Who is to determine what are justified claims?’ and ‘Who is entitled to ensure that justice is done?’ are either not addressed or else thought to rely on the extent to which the relevant agent meets objective criteria of justice. ³² This is particularly unsatisfactory insofar as the demand for justice is linked to the use of coercive means, as in the case of military intervention. For the anti-statist cosmopolitan it becomes hard to identify any normatively significant difference between coercion by domestic political authorities and coercion by foreign governments. ³³ Yet this is to ignore domestic context as the most important arena for specifying and concretizing what should count as each person’s legitimate rights. ³⁴ It implies a form of expert rule where political processes and decision-making involving the rights holders themselves is replaced by normative reflection carried out by the moral philosopher.

III An alternative Kantian conception of justice

In view of the considerations brought forward above, it is worthwhile to consider whether there are better ways of conceptualizing justice. To my mind, a promising alternative is to think of justice in terms of what Kant calls an innate ‘right to freedom’, defined as a right to independence from being subject to constraints arbitrarily imposed by other people. ³⁵ It is promising not only because it seems to steer clear of the problems connected to distributive conceptions of justice, but also, and particularly important in this context, because it better enables us to see the internal connection between individual freedom and state sovereignty.

The idea of a right to freedom is one that squares well with the basic features of moral cosmopolitanism, and it should therefore have some appeal to anti-statist cosmopolitans. It is individualistic in the sense that it acknowledges individual human beings as ultimate units of concern. It is universalistic in the sense that the status of ultimate unit of concern attaches to every human being equally. And it is general in the sense that all persons are ultimate units of concern for everyone.

The right to freedom is further linked to a concept of right which Kant says can be located ‘directly in the possibility of connecting universal reciprocal coercion with the freedom of everyone’. ³⁶ This formulation expresses the familiar idea of reciprocal and coercively protected spheres of freedom within which one is free to choose as one pleases. Inasmuch as the anti-statist cosmopolitans are political liberals committed to the ideal of freedom and equality, this should also appeal to them.

At the same time, conceptualizing justice in terms of Kant’s right to freedom differs remarkably from distributive conceptions that articulate justice in terms of human rights protecting fundamental human interests. First of all, the idea of equal freedom, as Kant understands it, does not track what are claimed to be fundamental human interests. Although the capacity for rational agency forms the ultimate ground for claims of justice, ³⁷ the right to freedom is not a protection of some underlying interest attributable to persons considered individually. Its focus is exclusively on interpersonal relations, or what Kant calls ‘the form in the relation of choice’ on the part of interacting persons. ³⁸ Rather than an output that can be specified without reference to one’s relations to other people, freedom as independence is a claim of each person against all other persons that they do not subject him or her to their arbitrary choice. As such, freedom ‘is not a good to be promoted’, but ‘a constraint on the conduct of others’. ³⁹ The normative base-line is that everyone should have the right to make choices of their own provided their exercise of this right does not encroach on anyone else’s right to make free choices. Every claim of justice must somehow be founded in this right to equal freedom, which is an unconditional constraint on any effort at promoting other desirable ends.

Second, due to its direct focus on interpersonal relations the idea of a right to freedom is not a distributive idea. It does not refer to the equal distribution of a pre-politically defined set of rights or of an equal range of equivalent opportunities. Nor does it refer to freedom as one good among others that have to be promoted, possibly in competition with other goods. The idea is strictly relational, in the sense that it concerns the standing of persons vis-à-vis other persons. This standing should be one of mutual independence. Everyone should be free to decide for herself or himself what ends to pursue, and no one should be in a position to impose her or his arbitrarily chosen ends on others. Justified restrictions on the right to pursue ends of one’s own choice must be reciprocal and non-contingent. They must restrict everyone equally, and they must not merely represent the particular view of one person or group concerning what should count as binding prescriptions for interaction. Enabling relations of mutual independence is the rationale for establishing legal and political institutions, and the idea of such relations is the standard by which these institutions are assessed.

On this basis, there is no straightforward way of justifying a duty to aid other people whose fundamental interests are at stake. Inasmuch as questions of justice are not primarily questions about what ends to pursue, but questions about who gets to decide what ends to pursue, it is neither sufficient nor necessary that someone is badly off in order to generate duties of justice in other people. Accordingly, in Kant’s right to freedom there seems to be no support for a duty to intervene militarily in order to help victims of tyrannical state coercion ⁴⁰ or for a ‘Natural Duty of Justice’ to help bring about just institutions for all persons.

As regards the internal connection between individual freedom and state sovereignty that I am aiming at in this article, it is crucial that the right to freedom is not only a principle for assessing the legitimacy of legal norms and institutions, but also an idea that requires a state authority. Freedom, understood as a system of reciprocal and non-arbitrary constraints, is not possible to uphold in the absence of a public authority that organizes legislative, executive and judicial public institutions. On this conception, we can interact in a fully rightful way only in a civil condition, of which the state is constitutive. If one accepts that justice should be thought of in terms of a right to freedom, one should therefore reject the view that legal and political institutions are mere tools for promoting desirable outputs. Such institutions should rather be seen as constitutive of justice. For the same reason, we should avoid thinking of state sovereignty as an instrumental value. If the state is a necessary condition for mutual independence, then recognizing the equal sovereignty of states is part and parcel of respecting each person’s right to freedom.

IV The complementarity of state sovereignty and individual freedom

The core of the problem that makes the state a necessary condition for interaction on just terms is that the right kind of independence is not possible in a hypothetical state of nature. Absent public institutions that make, apply and enforce laws, interaction on terms of equal freedom is not possible because of certain structural defects. With Arthur Ripstein, we might call these defects the problems of unilateral acquisition, indeterminacy and assurance. ⁴¹ Although the problems are distinct, they refer to defects that are parallel in their structure. In all cases, the point is that interaction and coordination of action plans in a state of nature must be based exclusively on the private judgements of interacting parties, which means that we unavoidably subject each other to arbitrary choice as long as there is not established a public institutional framework governing our interaction.

The problem of unilateral acquisition concerns appropriation of things external to oneself. Since a general prohibition against appropriation of things separate from us is an arbitrary restriction of freedom, any legitimate legal system must permit such appropriation. ⁴² External objects must further be acquirable without the consent of others. If consent were required, acquisition would depend on the arbitrary choice of others, which is not compatible with freedom of choice subject only to universal law. At the same time, unilateral acquisition restricts the free choice of all others by placing obligations on them that they would not otherwise have. This is equally incompatible with the requirement of equal freedom. ⁴³

The problem of indeterminacy concerns the demarcation of each person’s sphere of equal freedom when there is disagreement about rights. General rules and principles are indeterminate and in many cases leave room for reasonable disagreement about where the boundary between mine and yours should be drawn. The challenge is to resolve such disagreement consistently with each person’s right to freedom. Kant’s claim is that there is in fact no way in which we could solve such problems of indeterminacy in the absence of public legal institutions, because without these institutions any judgement about the distinction between what is mine and what is yours is a private judgement. Whoever decides where to draw the line subjects someone else to arbitrary choice.

The problem of assurance concerns enforcement of acquired rights. It must be possible to prevent, via coercive means, others from using whatever we have acquired a right to. Yet in a condition without a publicly authorized enforcement agency, there can be no rightful enforcement of property rights. Where there are no public authorities, any coercive act is performed by a private agent who cannot possibly suffice for generating reciprocal restraints. A private enforcer is what Kant calls a ‘unilateral will’, and the acts of enforcement by a unilateral will are arbitrary from the perspective of everyone else. ⁴⁴ At best, a private enforcement agency could restrain everyone but itself, which means that the problem of assurance remains unresolved with regard to the relation between the enforcement agency and all other agents. ⁴⁵

These defects of the state of nature are problems even under ideal conditions where people relate to each other in good faith. They do not depend on assumptions about moral baseness on the part of interacting persons. Since we are moral beings, we can always act according to what we recognize as being the right thing to do. Yet even if we assume human agents to be so fair-minded and well-disposed toward each other that they are not inclined to violate anyone’s right to freedom, the form of the relationship between interacting parties would still be wrong as long as all coordination of action relies exclusively on their arbitrary choices. Not human malevolence, but structural features of the state of nature alone explain why one cannot exercise the right to free choice in a way fully compatible with everyone else’s right to exercise their right to free choice. This, I take it, is the point when Kant writes that even if ‘the state of nature need not, just because it is natural, be a state of injustice…it would still be a state devoid of justice (status iustitia vacuus), in which when rights are in dispute (ius controversum), there would be no judge competent to render a verdict having rightful force’. ⁴⁶

The only way to overcome the systematic dependencies that would exist in a hypothetical state of nature is in Kant’s view to establish a state – that is, a public authority organizing legislative, executive and adjudicative bodies. Inasmuch as one agrees that any justified restriction on freedom must be for the sake of freedom itself, I think one should agree with him on this point as well. Absent public institutional bodies that coercively regulate interaction on a territory shared by several persons, there is no agent who can adjudicate or enforce rights in a way compatible with each person’s right to freedom. In a situation where such institutional bodies did not exist every conflict over rights would have to be resolved on the basis of arbitrary choices made by the conflicting parties. Rather than mutual independence there would be systematic dependence between persons. For this reason, creating a public authority that represents the will of all citizens united seems to be the only way there is to create a system of reciprocal and non-arbitrary constraints.

Yet if the state can reasonably be seen as a necessary condition for relations of mutual independence, it seems mistaken to contrast human rights with state sovereignty, or to reduce sovereignty to an instrumental value. Negatively put, this is so because the idea of mutual independence is not fully specifiable without reference to the legal and political institutions that make up the state. Rather than a pre-politically defined ideal, the right to freedom is a constraint on what can be recognized as a legitimate legal and political order. It is an abstract normative criterion that rules out laws and institutional arrangements that entitle some person or persons to subject others to their own choice or that create systematic relations of dependency between persons. What this concretely entails, however, is an open question. In what specific form the abstract right to freedom should be transformed into positive rights and how such rights should be interpreted in particular cases can be determined only within the institutional framework of a state. In other words, prior to political processes and procedures leading up to legally binding decisions issued by public and authoritative institutions there is no given or fixed moral output in relation to which sovereignty could be understood as an instrumental value.

Put more positively, state sovereignty is a necessary complement to each person’s right to freedom, because it secures the autonomy of the political processes and procedures necessary for giving the abstract idea of mutual independence concrete content and binding force. To recognize the principle of non-intervention as a basic principle of international law is simply to approve of the state’s role as an institutional framework enabling rightful exercise of individual freedom. By contrast, a right to military intervention for the protection of human rights is the same as a right to jeopardize the freedom-enabling institutional framework of the state. It is a right to wage war, which in turn is to put the state-sanctioned public order at risk. And this conflicts with each person’s right to freedom. If states are not protected from interventions in their internal affairs their function as public authorities is effectively undermined. Each of their decisions could then be contested and opposed by foreign powers, and individuals would thereby be deprived of a final authority that could determine the rightful boundaries of their freedom. For this reason, I believe Michael Walzer is completely right when he writes that ‘the recognition of sovereignty is the only way we have of establishing an arena within which freedom can be fought for and (sometimes) won’. ⁴⁷

It should also be noted that understanding states as freedom-enabling institutional frameworks is far from understanding them as guarantees that the equal freedom of citizens will in fact be realized. According to Kant, the ideal constitution for the state authority constitutive of civil society is the republican constitution that binds executive power to the legislative will of the people. Yet there is nothing in his argument or in the argument that I have put forward that makes a perfect republican constitution a criterion for recognizing the sovereignty of a state. Qua enabling frameworks states are structures where freedom can (but need not) take on concrete shape. Understood relationally, freedom is not a gift or something that can be imposed on a people from the outside, but a common practice, something which co-citizens must continuously strive for themselves. Such common practice needs an arena where reciprocal ascription of rights can take place. States are such arenas. And these arenas, even when they are less than perfect, should be protected by the principle of non-intervention. ⁴⁸ Whoever is concerned with individual freedom should therefore be equally concerned with state sovereignty.

V A novel defence of the domestic analogy

If the preceding argument is sound, it provides a defence of the so-called domestic analogy that circumvents criticisms commonly aimed at such an analogy by anti-statist cosmopolitans. Based on this analogy, sovereignty can be understood as an international counterpart to individual freedom in the domestic realm. In much the same way as individuals have a right to bodily integrity, states have a right to territorial integrity. In both cases, integrity is an external condition for autonomous conduct, although the right to integrity is not conditional on such conduct. The right to integrity is an essential aspect of the right to freedom, and it includes the right to make one’s own mistakes, on the part of individuals as well as states. As in the case of individuals, there are things we cannot do to states, even if it is for their own alleged good. In other words, there is no direct relation between the internal features of a state and its international standing. ⁴⁹ A state can be unjust domestically, but still be recognized as sovereign internationally. Accordingly, the domestic analogy implies equality of all states. In contrast to the asymmetrical view defended by anti-statist cosmopolitans, defending the analogy means defending non-intervention and self-determination as basic principles of international law covering both just and unjust states.

This parallel between interpersonal and international relations is frequently criticized by anti-statist cosmopolitans for being ill-founded. In their view, advocates of a statist international order have so far not succeeded in coming up with convincing arguments for recognizing non-intervention and self-determination as basic principles of international law. Such arguments are considered either (1) to overlook important differences between individuals and states or (2) to put an undue emphasis on the value of communal autonomy to the detriment of the rights of individuals.

(1) The first line of criticism is often aimed at an argument in favour of territorial integrity and non-intervention that can be traced back to political theorists such as Christian Wolff, Samuel Pufendorf and Emer de Vattel. The basic idea of this argument is quite simple: starting from the assumption that interstate and interpersonal relations are relevantly similar, one concludes that states should interact according to norms analogous to the norms governing interpersonal interaction. Like persons, states are morally equal actors, and just as persons generally are held to have a right to pursue ends of their own independently of what ends others think they should pursue, so states have a right to arrange their own affairs as they see fit without interference by foreign powers. Accordingly, states should be recognized as equal sovereigns possessing a right to self-determination and a duty of non-intervention in the internal affairs of other states. ⁵⁰

Anti-statist cosmopolitans generally reject this argument on the ground that it neglects relevant differences that count against recognizing states as autonomous agents worthy of respect comparable with the respect we owe persons. In contrast to persons, states are collective actors that cannot form beliefs or make choices of their own, at least not in any straightforward sense. States are not human beings writ large. As Beitz puts it, ‘states qua states do not think or will or act in pursuit of ends; only people…do these things’. ⁵¹ It therefore does not appear to be any compelling reason to invest states with rights analogous to the rights persons have to pursue ends of their own choice (as long as they do so in a way compatible with the rights of others to do the same). Since states lack relevant moral properties justifying territorial integrity on an analogy with integrity of the person the argument seems to fail.

(2) The second line of criticism usually addresses Michael Walzer’s communitarian defence of the domestic analogy. This is a defence that does not ascribe moral agency directly to states. The idea is not that states are moral agents comparable with persons and therefore also have rights analogous to persons. The idea is rather that states are political representations of underlying social entities that have a right to autonomy – that is, a right to stake out their own paths toward freedom without interference by foreign powers. ⁵² On this view, then, foreign intervention is less an offense against the state than an offense against the community represented by the state.

Although Walzer frequently speaks of communal autonomy as if it is a right possessed by political communities as such, he ultimately grounds it in the right of individuals to live in self-determining communities. In the final resort, the rights of communities rest on the tacit consent implicit in ongoing cooperative practices among community members. Thus, the deepest moral foundation of community is, as he puts it, ‘a contract, Burkean in character, among “the living, the dead, and those who are yet to be born”’, whereas ‘the idea of communal integrity derives its moral and political force from the rights of contemporary men and women to live as members of a historic community and to express their inherited culture through political forms worked out among themselves’. ⁵³ And to such historic communities foreigners owe respect, which Walzer juxtaposes with ‘a morally necessary presumption: that there exists a certain “fit” between the community and its government and that the state is “legitimate”’. ⁵⁴ Even if governments sometimes can turn against their citizenry, foreigners do not have enough historical knowledge or direct experience to form adequate judgements about a government’s actual legitimacy. Generally, they are not in position to judge whether a government uses the coercive apparatus of the state merely in its own self-interest, or whether it governs in a way that fits with a community’s traditional way of life. In combination with the right to live in a self-determining community, this lack of contextual understanding requires that we act as if governments or other states are legitimate – except in certain special circumstances. ⁵⁵ Failure to do so would be to disregard the rights of other states’ citizens. For this reason, Walzer concludes, ‘states can be presumptively legitimate in international society and actually illegitimate at home’. ⁵⁶

Anti-statist criticism of this communitarian defence of the domestic analogy typically puts into question the connection between respect for communal integrity and the presumption of fit between community and government. For one thing, there seems to be no obvious reason for assuming that we do not know enough in order to make sound judgements regarding such a fit. On the contrary, there is usually an abundance of relevant sources of information, such as social scientists, emigrants, experienced travellers, diplomats, scholarly works, and so on, that makes what Walzer calls a morally necessary presumption implausible. ⁵⁷ Moreover, it is rarely the case that there is a clear match between state and community. For the most part, if not always, there will be a plurality of communities within one and the same state territory. And even if it makes sense to speak of a state representing one and only one community it does not necessarily follow that we can expect internal unanimity regarding the interpretation and importance of different traditional values. In view of such pluralism, it is not clear that non-intervention is the best way of respecting communal integrity. Abstaining from interfering in an intra-state conflict could instead be interpreted as partisanship in favour of the dominant party. ⁵⁸ In other words, Walzer’s argument does not explain why non-intervention is a principle that should protect states, and therefore fails as a defence of the domestic analogy.

I do not want to challenge the anti-statist cosmopolitan criticism of these two defences of the domestic analogy. Instead, I claim that the present defence of the analogy is not susceptible to these lines of criticism. In no way have I assumed that states have moral faculties similar to those of human agents. Nor have I assumed that communal integrity has any independent moral value or that we are morally required to presume any fit between government and community. On the account given here, states are understood as necessary arenas for realizing individual freedom in a rightful way, and qua such arenas they ought to be protected by the principle of non-intervention. Sovereignty in the international realm is a necessary complement to such an understanding of the state, whereas any right to wage wars in the name of justice would endanger the state as a freedom-enabling institutional framework, which is therefore in conflict with each person’s right to freedom. Just as persons have a right to independence vis-à-vis other persons so states have a right to independence vis-à-vis other states, because only if states have such a right can they perform their role as public authorities properly. Against this defence of the domestic analogy, none of the considerations brought up above can count as relevant counter-arguments.

However, by taking the analogy a step further one could conceivably question a central claim made in this article, namely that we should regard states as equal sovereigns. Since states interact with other states, the idea that they have a right to independence might seem to support the view that there is need for a second-order global public authority analogous to first-order state authorities. Yet if states were to subject to a world state, then they would no longer be sovereigns. At most, they would be relatively autonomous sub-units of a global federal state vested with the power to coerce its members.

This apparent complication relates to Kant’s repeated comparison of international relations and the interpersonal state of nature. The comparison builds on the ideas that states, like individuals, can disagree about the scope of their respective rights and that no state has the authority to decide how to settle such disagreement. The merits of claims that one state makes against other states are similar to the merits of claims that private parties make against each other in the domestic context. For this reason, states unavoidably subject one another to arbitrary choice as long as they lack access to public procedures for resolving disputes over rights in an impartial way. Accordingly, an international system based on the balance of power is a ‘condition of war (of the right of the stronger)’, which is ‘wrong in the highest degree’, ⁵⁹ and states that cannot avoid interaction are obliged to end such a system by joining an association of states where they can decide disputes ‘as if by a lawsuit’. ⁶⁰

Many commentators claim that the analogy between an interpersonal and an international state of nature implies that a ‘state-like union…between existing states’ should be established. ⁶¹ The claim rests on the idea that if individuals must form a state with legislative, adjudicative and executive authority in order to resolve their conflicts in a rightful way, then states must form a state of states in order to do the same. In contrast to this view, I believe there are good reasons to reject the world state in favour of a voluntary league of states, which is usually held to be Kant’s preferred solution to the problem. ⁶² Unlike a world state, the league is vested only with an analogue to the adjudicative authority of first-order state authorities. Even if it is a precondition for rightful conflict resolution, one cannot force membership in the league upon a state. As an association of free and independent states, it is an organization that ‘each neighboring state is at liberty to join’ and that ‘can be dissolved at any time’. ⁶³

Two main considerations speak in favour of voluntary membership in the league. First, if membership were an enforceable duty, then states would have a right to wage war in order to exit the international state of nature. Such a right is in conflict not only with the main purpose of the league – to establish a condition of peace – but also with the idea that a state-sanctioned public order is a necessary background condition for interaction on terms of equal freedom. Since going to war is putting the freedom-enabling institutional frameworks that states are at risk this ‘is analogous not to founding a state but to a revolution which fails and leads to a state of nature’. ⁶⁴ Second, membership must be voluntary in order to allow states to step back and remain neutral whenever other people choose to resolve their conflicts by fighting a war. Unless permitted to avoid entanglement in violent conflicts between others, states would be dependent on the arbitrary choice of those who do not manage to solve their disagreements peacefully. A duty to get involved in international conflicts would violate the right to maintain one’s own public authority, and, by the same token, the basic right to mutual independence. ⁶⁵

The above considerations support the view that establishing a world state with coercive powers is the wrong way of overcoming the international state of nature, but they do not yet show why a voluntary league whose authority is limited to adjudication can do the job. If a coercive public authority is constitutive of just interpersonal relations, how can such an authority be dispensable at the international level? I can at this point only sketch an answer to this question, but I believe it is possible to square Kant’s defence of a voluntary league of states with the claim that international relations absent a shared authority is a state of nature if one acknowledges that the domestic analogy is only a partial, and not a perfect, parallel. The basic idea is that because of the essentially public nature of states only one of the three structural problems presented in the previous section arises in the international case. This is the indeterminacy problem, which can be solved by creating an international analogue to a domestic system of courts. ⁶⁶

Unlike individuals, states do not set their own purposes freely. As public entities, they can pursue only purposes that serve one basic end: to maintain a rightful condition where citizens can interact on terms of equal freedom. States’ right to independence in relation to other states concerns their entitlement to pursue this end freely. Although limited in the purposes they are entitled to pursue, states are free to organize as rightful conditions in the way they find most appropriate without answering to any higher authority.

Two implications of recognizing the public nature of states are of particular importance for the present discussion. For one thing, states are justified only in fighting defensive wars, since only defensive wars serve their own maintenance. Moreover, territory is not an external asset that a state owns, but its ‘spatial manifestation’. ⁶⁷ Unlike private property, which is the means with which a person has an exclusive right to pursue whatever end she or he chooses, territory primarily demarcates the limits of a state’s jurisdiction. We can for this reason conceive of a state’s territory as its embodiment, which sets it apart from other states. Rather than means with which it pursues private purposes, territory is constitutive of the international personhood of a state.

The distinction between property and embodiment is important because the problems of unilateral acquisition and assurance in the original state of nature between persons arise only with respect to external things. Since territory is not an external asset, but the international analogue to a person’s body, the demand that other states respect a state’s territorial integrity is not a unilateral imposition of a contingent obligation. Surely, the particular borders of states fall where they fall because of contingent historical processes, but in resisting aggressors, a state does not act as a unilateral will that enforces property rights. It simply acts defensively to maintain itself as a freedom-enabling institutional framework.

At the same time, it seems reasonable to say that there is a problem of indeterminacy inherent in an international system without a public authority. States can, for instance, reasonably disagree about border drawing or the use of natural resources that transcend borders (e.g. water or oil). Absent a shared authority to decide on such issues, any judgement is the particular judgement of one state. Accordingly, there is no way to settle disputes of this kind without subjecting someone to arbitrary choice. This explains Kant’s (partial) analogy and the corresponding need for a league of states. The league may not guarantee against aggression, but in providing public and impartial procedures for deciding disputes, it enables states to resolve their conflicts in a way compatible with their right to independence. Put otherwise, the league provides the necessary institutional precondition for rightful international relations. If this is correct, then defending the domestic analogy is compatible with recognizing states as equal sovereigns. Becoming members of a voluntary league allows states to remain sovereigns rather transform into sub-units of a global state.

VI Concluding remarks on permissible interventions

In the preceding pages, I have argued that having a just inner constitution is not a precondition for being recognized as a sovereign state in the international realm, which is to say that issuing unjust or oppressive laws does not provide other states with a just cause for intervention. A worry that probably will cause anti-statist cosmopolitans to resist this standpoint is that a general prohibition against intervention in a state’s internal affairs apparently can serve as protection for governments that commit grave violations of human rights, such as genocide or ethnic cleansing. Sometimes the coercive apparatus of a state is turned systematically against individuals and intra-state groups on a large scale. Are we really obliged to stand by and watch as atrocities are going on? I think not, but unlike anti-statist cosmopolitans, I also do not think that permitting interventions in such extreme cases should prevent us from acknowledging non-intervention as a basic principle of international law.

In order to see how defending the principle of non-intervention can be consistent with permitting intervention in certain extreme cases it can be useful to point out the obvious: that organizing a state is not the same as organizing a monopoly of force on a territory. Although the latter is an essential aspect of the former, a state must also establish an institutional order where conflicts over rights can be resolved through legal procedures and binding decisions made by public officials acting on mandate. Such an order can certainly be defective. It can be unfair, oppressive, or too invasive in private matters. Yet the legitimacy of a state is not dependent on its satisfying some ideal standard of justice. It is sufficient that it establishes conditions that make it possible for persons to interact on rightful terms. Regimes that commit genocide or ethnic cleansing do not satisfy this requirement. They are not simply defective public orders, but organizations that unilaterally employ force against other people. Permitting foreign intervention against them is fully compatible with defending the equal sovereignty of states as well as the principle of non-intervention. In such cases, military intervention is not to undermine a state-sanctioned public order that enables individuals to interact on terms compatible with the equal freedom of all, but to prevent mass murder and arbitrary expulsion of people from a territory.

Some brief remarks on Kant’s infamous rejection of a right to revolution can help clarify this point. There are several aspects to Kant’s argument against such a right. The one most relevant to the present discussion turns on the idea that public authority is constitutive of individuals’ mutual independence. This idea is the basis of the claim that sovereigns have only rights against and no enforceable duties to their subjects. ⁶⁸ There can be no right to overthrow a government because any attempt to do so threatens to throw everyone into a state of nature. Since subjection to public authority enables interaction on terms compatible with each person’s right to freedom, subjects are obliged to ‘put up with even what is held to be an unbearable abuse of supreme authority’. ⁶⁹

It is important to note that this account of the prohibition against revolution does not imply an obligation to comply with the strongest bully around. ⁷⁰ Legitimate authorities can fail to govern justly, either by adopting non-democratic legislative procedures or by enacting excessively restrictive laws. ⁷¹ While injustices of this kind are objectionable, one should nevertheless comply with an unjust government if it rules according to laws that treat all subjects as equals. Yet this is far from suggesting that any organization claiming authority over some territory is a legitimate authority. Authority rests on the provision of an institutional arrangement that secures each person from subjection to the arbitrary choice of other private persons. Regimes that turns their coercive apparatuses against specific groups or that deny certain parts of the population the rights necessary for enjoying independence vis-à-vis others do not satisfy this criterion of legitimate authority. Such regimes are not instances of unjust government, but of inhumane organizations that annihilate the freedom of some people in their relation to others. ⁷² Resisting such regimes does not conflict with the obligation to comply with unjust government. Instead, resistance is a permissible step towards a condition where each person’s freedom can be reconciled with the freedom of everyone else. For the same reason, other states are permitted to intervene to stop barbaric and inhumane use of force. An entitlement to intervene against regimes committing atrocities is consistent with a duty to respect the integrity of sovereign states because in such cases no freedom-enabling institutional framework is in place.

Having said that, I should add that it does not seem particularly useful to think of intervention as enforcement of individual human rights. More adequately, it is conceived as an emergency measure that can be justified in extreme cases so as to bring an exceptional situation to an end and to establish a normal situation where individual rights can be ascribed and enforced. In all normal cases, non-intervention is the basic principle we should adhere to. The concern with genocidal regimes should not mislead us to relativize sovereignty, turning it into an instrumental value in the service of morally desirable outcomes. Justifying general rules on the basis of concerns related to exceptional cases is a bad habit. Just as in the domestic domain, so in the international: hard cases make bad law.

Still, it is not necessarily an easy matter to determine exactly where one should draw the line between a defective public order and an illegitimate regime against which intervention is permissible. Unlike Walzer, I do not think we can presuppose a clear-cut chasm between brutal and oppressive regimes on the one hand and regimes guilty of systematic massacre on the other hand. ⁷³ While there can be cases that leave little room for doubt other cases are less obvious, and any general criterion regarding where to draw the line will open up for reasonable disagreement in particular cases. If nothing else, the cases of Kosovo in 1999 and Libya in 2011 seem to make it clear that moral stakes can be fundamentally unclear.

Given such indeterminacy, Walzer’s defence for a unilateral right to intervene for alleged moral purposes should also be rejected, because such a right reintroduces the kind of asymmetry and hierarchy between states that I have criticized in this article. If it is up to particular states to decide in accordance with their own arbitrary moral judgement whether the criteria justifying humanitarian intervention are met, the permission to intervene in exceptional cases seems to translate too easily into an exceptional status for powerful states. For this reason, legitimate interventions must be anchored in an international public body authorized to specify and apply general criteria for when the basic principle of non-intervention can be overruled. ⁷⁴ Authorization by such a body for any use of force short of self-defence is the only way intervention as an emergency measure can be reconciled with the equal sovereignty of states. Only if the permission to intervene goes along with commitment to an inclusive international organization representing all states is there a chance that it does not become a means of self-empowerment on the part of powerful states.