2. Theory, Practice, and Judgement

Some of Kant’s contemporaries and near contemporaries, most notably Hegel, questioned whether his metaphysics of morals could provide substantive guidance for action. Hegel claimed that the categorical imperative yields only empty tautologies, because any maxim could be universalized, depending on our particular social assumptions and expectations. We define a maxim’s key terms within a set of background institutions, any number of which are feasible, and that introduces contingency and variability into Kant’s demand for a priori rational consistency (Hegel Reference Hegel[1807] 1977, §§430–1).

In his response in ‘Theory and Practice’, Kant begins by explaining the distinction between theory and practice and why theory must be brought to bear on real-life practical concerns. According to Kant, applied moral philosophy, like all philosophy, must be a priori; our practical obligations must be established through reasoning rather than by generalizing based on empirical cases. The resulting principles are necessary in order to justify and give guidance to our actions, which, without theory, would amount to ‘groping about in experiments and experiences’ (TP, 8: 276).Footnote ¹ Thus he cautions against drawing on common practice in moral or legal issues:

When examining probable outcomes of adopting a principle, we risk making its adoption dependent on empirical generalizations rather than rational necessity. A guiding principle ensures that our actions are consistent with what is rationally required of us.

Although theory is crucial for justifying our decisions, Kant realizes that theory alone is insufficient to determine what we ought to do. Just as we apply the categories to sensible intuitions, we also relate rational ideals to proposed maxims and existing political practices through ‘an act of judgment by which a practitioner distinguishes whether or not something is a case of the rule’ (TP, 8: 275). Judgement is necessary to identify whether an action falls under a principle and which circumstances are relevant for properly characterizing the action – for example, whether a killing is a case of murder or self-defence. Further levels of theory could not settle these questions, or else we would face a regress: a rule would determine which actions fall under a rule, but we would also need a rule to determine whether the initial rule is properly applied in these circumstances, and so on ad infinitum (TP, 8: 275). Therefore, although theory is needed to justify and make sense of practice, judgement bridges the gap between them by characterizing our actions – picking out their morally salient features, and so on – and determining whether they are covered by the rule.

3. The Liberal Ideal

In a legal sense, judgement determines whether and how an action is covered by the law, which is similar but not equivalent to the process of determining whether a maxim is morally permissible. Since we cannot be coerced into acting for the right motives, the most that the law can do is make sure that our actions do not wrongfully impinge on others’ acting; it protects the exercise of people’s outer freedom, not their inner freedom (MM, 6: 396). Furthermore, although all of us pursue happiness, we are after different things to make us happy. Thus the law fulfils the negative function of restricting us when pursuing our happiness harms or limits others, but it does not serve the positive goal of making us rational or happy. It merely gives us the opportunity (to try) to act rationally and to become happy.

We consent to the law in accordance with the rational demand to respect others as free pursuers of ends, and we act rightly by constraining our actions through the government’s coercive measures. Only by agreeing to the enforcement of external laws do we fulfil our moral duty to further the ends of others – specifically, by thwarting those actions that are inconsistent with others’ outer freedom (MM, 6: 231). The threat of fines or a prison sentence discourages people from violating the freedom of others and promotes freedom in general. Thus Kant defends a liberal system of right, which he thinks is best supported by a republican state apparatus – that is, a constitution under which the executive and legislative powers are separate. Such states can avoid despotism because the executive power enforces laws that are (or at least could have been) freely consented to by the people, rather than uniting the two powers in a way that would make law the instrument of someone’s private will (TPP, 8: 352–3; WIE, 8: 39; TP, 8: 297).

Kant also takes a broader view in his theory of right, saying that we should establish international laws to govern relations among states. Kant realized that, by his time, the world had developed to the point where nations inevitably affect one another, both politically and commercially: ‘the (narrower or wider) community of the nations of the earth has now gone so far that a violation of right on one place of the earth is felt in all’ (TPP, 8: 360). One problem is that our freedom remains at risk as long as the only people who are told not to harm us, on pain of punishment, are our fellow citizens. We safeguard our rights more effectively if we also protect ourselves against aggression from other countries, and we maintain a more stable system of international relations if we work to prevent other countries from collapsing into despotism or chaos. The latter, preventative measures are preferable to armed conflict, of course, and contemporary cosmopolitans such as Garrett Wallace Brown and Alexandra Bohm have drawn on Kant to claim that we are obligated to promote global economic and social justice, which would help us avoid war and prevent crimes against humanity, and to engage in military intervention as a last resort and only to establish a more peaceful condition (Brown and Bohm Reference Brown and Bohm2015; Bohm Reference Bohm2013).

The maintenance and enforcement of international law cannot depend only on individual states holding themselves and each other to account, however. Kant believes that an international order based on a balance of power among autonomous states – a Hobbesian system in which peace is maintained through deterrence alone, such as the one established by the Treaty of Westphalia – will never succeed in bringing about perpetual peace (TP, 8: 312–13). He says that states should form an association to further the principles of liberal freedom and republican government to which the member countries are committed (TPP, 8: 354–7; MM, 6: 344, 350–1). Although states will not relinquish their independence by submitting themselves to ‘a world republic’ that has its own coercive power – Kant says that states cling to ‘their idea of the right of nations’ and thus reject this option, which is right in principle – they may form a voluntary federation, ‘a league of nations’, in which independent states hold themselves to international agreements out of a sense of moral obligation to avoid war (TPP, 8: 354, 357). This cosmopolitan association not only provides a forum in which to peacefully adjudicate disputes between nations but also a body to support protections for individuals considered as world-citizens.

4. The United Nations and the Non-Cosmopolitan Present

With his liberal conception of legality, his republican politics, and his cosmopolitanism, Kant has answered the question of what nations ought to become and how they ought to relate to one another. Presumably judgement should allow us not only to evaluate how existing legal systems fall short of the ideal, but also to devise strategies for improving things so that we can get closer to the ideal in reality. If judgement were merely evaluative, then Kant would still be susceptible to the charge that his theory has little relevance for actual practice. As we will see, however, it remains unclear how (or whether) nations whose governments are informed by rational principles should help to transform so-called rogue states that do not participate in or respect international agreements, or more generally despotic states that do not accept the liberal principles that Kant defends.

Kant envisions a form of international engagement in which republican nations diffuse possible conflicts through reasoned discussion and collective action. Members of a ‘pacific league’ would not subject each other to coercion, since they are no longer in a state of nature with regard to one another (TPP, 8: 356; MM, 6: 344, 351). This raises the question of how to characterize our current state of affairs. If the United Nations is a deliberative body in the Kantian mould, as some people believe it is, then military force would seem to be ruled out as counterproductive and contrary to the terms of international cooperation. Instead, states would engineer the reform of other states by means of UN resolutions and other kinds of diplomatic and economic pressure.

Kant conceives of the league as ‘a federalism of free states’, not despotic states (TPP, 8: 354). Among other things, republics are less likely to go to war since they depend on the consent of the people, and the people are more likely to consider the human and economic costs of war (TPP, 8: 350). It is precisely because the participants in the United Nations are not all republics that Roger Scruton (Reference Scruton2005) and others think it is incapable of achieving a Kantian peace. Indeed, some have argued that, because the majority of UN member countries are not free democracies, it often protects dictatorships and does little to stop genocide and terrorism (see e.g. Roth Reference Roth2001; Gold Reference Gold2004). According to some UN officials, a pervasive moral relativism underlies its ‘philosophy of neutrality and nonviolence’, which they admit is unsuited to addressing modern crises (Crossette Reference Crossette1999). Rather than working through the organization to promote peace and protect human rights, countries often use their position on the UN to advance their own national agendas. For example, Russia and China have vetoed six separate proposed UN Security Council resolutions on Syria, seemingly to protect their military and economic interests, which has frustrated the United Nations’ already limited efforts to address the problem (Fisher Reference Fisher2013; Yan Reference Yan2013). And since 2015 Russia has stepped up its military ties with Syria, launching air attacks against ISIS and other, US-backed opposition groups, in support of the Assad government, despite the fact that the US and EU both consider Assad’s rule to be ‘illegitimate’ (see Sengupta Reference Sengupta2015). Kant believes that coercive power should not be used against nations within the pacific league, yet the membership of all of these countries in the United Nations – the US, the UK, countries of the EU, Russia, China, and Syria itself – has not replaced war with diplomacy; war between member states remains a live option and each country remains the judge in its own case, just as it would in a state of nature (MM, 6: 346). Because of the power of despotic regimes to shape UN activities, the UN’s frequent deference to oppressive national policies, and its lack of enforcement mechanisms, Fernando Tesón concludes, ‘Not only does [the UN Security Council] lack moral legitimacy; it does not even secure order, let alone justice’ (2005: 18).

We are left in an international situation that is neither a state of nature nor a league of nations, but rather a complex system of treaties for mutual protection and unenforced or unenforceable UN resolutions. Jürgen Habermas, whose own cosmopolitan theory draws on Kant, gives a sombre evaluation of where we are: ‘The contemporary world situation can be understood at best as a transitional stage between international and cosmopolitan law. But many indications seem to point instead to a regression to nationalism’ ([1996] 1998: 183). In the absence of a fully developed league of nations, Kant’s practical philosophy cannot tell us how member states, even ones that are (for the most part) rightly organized, ought to respond to the most egregious forms of state-sponsored violence in other nations.

5. Scruton, Roff, and Habermas: Justifying Military Intervention

Kant generally supports a gradualist approach when it comes to reforming political systems – evolution instead of revolution. Specifically, he claims that societies should develop internally towards the ideal of right by means of public discussion, and externally by working through international organizations and commercial activity to nudge despotic states in the direction of republican self-rule. Although admittedly it is short of a cosmopolitan arrangement, the existence of the United Nations and numerous treaties and covenants seem also to entail that war would almost always be wrong, short of defending oneself against outright aggression. This has led several theorists, including Georg Cavallar (Reference Cavallar1999, Reference Cavallar2006), Thomas Mertens (Reference Mertens2007), and Howard Williams (Reference Williams2012), to defend versions of Kantian cosmopolitanism that verge on pacifism. For all of them, our efforts to defend human rights or intervene in despotic states must be constrained by existing external law, specifically the will of the international community that is represented by the UN. There is only a provisional entitlement (as opposed to a right) to go to war in very restricted circumstances, when the state itself is threatened, because the state has a legal duty to sustain itself. Thus they take a reformist outlook towards both the UN’s own imperfections and the misguided laws of individual nations.

Still, since the majority of the UN are non-republican states, and since it has repeatedly failed to halt wars and mass killings – in short, since it is not Kant’s league of nations in composition or accomplishments – some people have drawn on Kant’s philosophy to argue for military intervention, based entirely on what they see as cosmopolitan principles. Kant says that in a state of nature states have a right to go to war, for example, ‘to establish a condition more closely approaching a rightful condition’ (MM, 6: 344). In light of the United Nations’ inability to protect the rights of cosmopolitan citizens or spread republicanism, Roger Scruton (Reference Scruton2004, Reference Scruton2005) appeals to Kant’s political philosophy to justify the United States’ invasion of Iraq in 2003. According to Kant, a league of nations can intervene to protect states from other states and to maintain a relation of right among them. And, Scruton notes, such an international body can only be legitimate and can only function if each member nation is a republic, or a ‘representative government under a rule of law’ (2005: 97). Therefore, in the case of Iraq and other rogue states, military intervention can be justified because it seeks to replace the despotic government with a republic that can participate in a cosmopolitan system. Tesón summarizes the point succinctly in claiming that Kant applies ‘a rule of nonintervention … only among just states’, so Western powers have no duty to respect the sovereignty of Iraq under Saddam Hussein – or, presumably, Syria under Bashar al-Assad (1988: 70; see also Tesón Reference Tesón2005).

Heather Roff (Reference Roff2013) similarly argues for military intervention in Syria, although she says that such intervention is warranted by the responsibility to protect (R2P) – in this case, to protect the rights of the Syrian people against the Assad regime. Roff argues that, since there is no international body that is equipped to enforce relations of right among states, sovereign nations are in a state of nature with regard to one another. In anticipation of the establishment of a league of nations (or an effective league of nations), however, we have a ‘provisional duty’ to protect those who would be protected if such an international arrangement were to be achieved. This duty allows nations to override the presumption of sovereignty in cases of ‘supreme moral emergency’, such as the crisis in Syria, where hundreds of thousands of people have been killed and millions have been displaced since the conflict began in 2011 (Roff Reference Roff2013: 62).

The call for military intervention, with explicit reference to Kant’s cosmopolitanism, is not new. Habermas is committed to a theory of communicative action whereby people engage in social coordination in order to pursue goals that they think are reasonable. Democratic pluralism in individual nation-states values the autonomy of each individual to consent to the social order, which gives rise either to a global civil society where nations engage in common democratic activity (Habermas [Reference Habermas1996] Reference Habermas1998) or a negotiating system in which democracies manage relations with other democracies, thus forming a looser political community (Habermas Reference Habermas[1998] 2001). Both systems are consistent with Kant’s theory. Yet even Habermas, who seems to defend a gradualist approach (through communication and negotiation), supported both the US-led Gulf War in 1991 (Habermas Reference Habermas[1990] 1994) and the NATO invasion of Kosovo in 1999 (Habermas Reference Habermas1999) – the former in order to support the UN and protect Israel, and the latter in order to ensure Kosovar autonomy within Serbia and prevent the genocide of ethnic Albanians. Like Scruton and Roff, Habermas claims that military intervention in non-republican (specifically totalitarian) states by republican states is necessary to protect the rights of world-citizens and to safeguard international law, i.e. ‘to strengthen the authority of the UN sufficiently for the resolutions of the world community to be enforced’ (Habermas Reference Habermas[1990] 1994: 10). In short, on Habermas’s view, in the non-cosmopolitan present, military action can advance cosmopolitanism.

Theoretically, Scruton’s and Roff’s idea of ousting an authoritarian ruler and Habermas’s idea of supporting international agreements sound very appealing and seem like they would further Kant’s political ideals. However, there are many alternatives for how to respond to these situations, including (but not limited to) negotiations, sanctions, and full-scale invasions. Which option is appropriate will depend on political, economic, historical, and geographical factors whose consideration is distinctly un-Kantian. For example, what background institutions are we assuming in our deliberations – a future league of nations, a present approximation (in the form of the UN), or a state of nature? Can we promote regime change and freedom through targeted sanctions, or will they be ineffective and adversely affect average citizens more than the government, thus betraying the very responsibility to protect that they are intended to fulfil? The proper route to take depends in large part on what people perceive to be the case – not even what is the case – regarding motives and circumstances. Many people in the Middle East distrust the United States and Europe because they have a long history of advancing their own self-interest there, such as supporting whichever government, oppressive or not, would best protect their military and economic aims in the region. The Iraq invasion in 2003 spurred recruitment efforts by terrorist groups, provoked terrorist attacks, and installed a Shiite-dominated government that is plagued by corruption and has responded harshly to anti-government protesters, leading the Economist Intelligence Unit (2016) to label it a ‘hybrid regime’, somewhere between a ‘flawed democracy’ and an ‘authoritarian regime’ – hardly the lynchpin for a new democratic Middle East. Of course, Kant himself cannot be expected to devise specific strategies for intervention, and what we ought to do, even for Kant, always depends on the circumstances, which must be factored in when we formulate our maxim. The point is that we can use Kant’s philosophy to justify military or other interventions in unjust states, yet whether we are contradicting or furthering the ends to which we are committed in theory depends entirely on the strategies that we adopt based on judgement regarding background institutions, people’s conception of the situation (whether founded or unfounded), anticipation of consequences, and so on.

Because of these problems, Antje Vollmer (Reference Vollmer2004a, Reference Vollmer2004b) uses Kant to criticize the Iraq war, noting that it did not succeed in spreading democracy, protecting the Iraqi people, or fighting terrorism. The question is: Was it excusable for the United States not to have foreseen this, and, in light of this uncertainty, was the decision to invade reasonable or not? Scruton says yes and Vollmer says no, but both of them appeal to Kantian principles, and there is no way to sort out which conclusion is right based on Kant’s theory alone. To be sure, Kant believes that we need to have recourse to practical anthropology in dealing with the application of right, but doing so introduces uncertainty in this case rather than actually clarifying our duties.

Like Scruton, Roff is optimistic about the outcome of a more aggressive interventionist policy in Syria, and her argument depends almost entirely on the aims of the mission rather than its prospects. Yet we need to make evaluative judgements about which expectations are reasonable and strategic judgements about the consequences of military action. She dismisses the mere imposition of sanctions by the US and the EU, and compares the response to the Syrian crisis unfavourably to NATO’s swift use of force in Libya in 2011 (Roff Reference Roff2013: 131). At the same time, she realizes that the situation in Libya was significantly different: the geography there allowed for better intelligence and more effective bombing campaigns, Libya had less dense population centres and a weaker military, and there was a unified opposition that controlled parts of the country (Roff Reference Roff2013: 135–6). We are thus left with a vague responsibility to protect, with uncertainty about who is bound by it and lack of clarity about what they should do. We cannot know whether sanctions and other efforts short of military intervention are sufficient to discharge the duty, how to consider a potential invader’s capacity to intervene along with their prospects for success, and so on (Roff Reference Roff2013: 138). Of course, one may say that the duty to obey international law takes precedence over a responsibility to protect, but as we see with Habermas, sometimes fulfilling the responsibility to protect is thought to strengthen rather than undermine international law.

Roff is left to conclude that ‘we must continually work towards establishing the conditions of justice in the real world’ while admitting that ‘each R2P situation will leave judgment about who ought to do what open, and identifying the duty bearer to the private judgment of states and individuals’ (2013: 139, 138). Although the aims of a possible intervention are clearly established by Kant’s practical philosophy, theory has little role to play in judgements about how to intervene and who should do it, thus leaving us more or less without guidance when it comes to political practice. Decisions are left ‘to the private judgment of states and individuals’, which, given variable interpretations of the facts and what is reasonable, seems a lot like ‘groping about in experiments and experiences’ (TP, 8: 276).

6. The Duty of Humanitarian Intervention: Provisional, Imperfect, or Perfect?

Under Kant’s practical philosophy, republican states are obligated to protect the rights and dignity of their own citizens. Beyond their borders, however, it is difficult to establish the basis and extent of their obligations to other people. Those who appeal to Kant to justify humanitarian intervention talk about it as a provisional duty (Roff Reference Roff2013), an imperfect duty (Benhabib Reference Benhabib2004: 36-7; Ang Reference Ang2015; see also Williams Reference Williams2012: 137–8), or a perfect duty (Bagnoli Reference Bagnoli2006), but all three approaches depend essentially on judgement to decide whether we have such duties in any given case. Roff assumes that countries are currently in a state of nature with regard to one another, and that military intervention is demanded of us in the absence of an international body to protect oppressed people (2013: 88–90). But both claims are questionable. Williams argues that Kant never considered it an all-or-nothing affair – either a state of nature or a league of nations – and instead ‘gives the impression that within the international sphere there is an emergent system of law with which political leaders should work and seek to enhance’ (2015: 167). Indeed, collective action on the part of NATO is putting pressure on Assad through targeted sanctions, and the United Nations is leading negotiations between the warring parties – not an ideal league of nations, to be sure, but hardly a state of nature. Thus the responsibility to protect by means of military intervention may be neither provisional (because we are not in a state of nature) nor a duty (because there are other ways to address the problem effectively short of using force).

Seyla Benhabib (Reference Benhabib2004) and Jennifer Mei Sze Ang (Reference Ang2015) also do not settle any of the important questions. Although characterizing the so-called duty to protect as an imperfect duty gives us some latitude regarding how the duty is to be fulfilled, the latitude in international obligations introduces an oppressive level of vagueness. For one thing, an imperfect duty does not obligate any specific state to intervene, so it seems that the duty to protect cannot be ‘effectively claimable’ (Tan Reference Tan2006: 95, Bagnoli Reference Bagnoli2006: 122). Although countries such as Italy and Belgium (or, for that matter, Jamaica and Cambodia, which are also members of the UN) may have a general duty to help others, lack of resources or military options limit how they could fulfil a supposed duty to protect. And if the US and Germany adequately responded, those smaller countries would not have to do anything to fulfil the obligation. This uncertainty means that often a duty will exist, in some sense, without anyone being willing to take it up, because no one thinks, perhaps reasonably, that it applies to them. It would amount to a kind of standing obligation, which would only require that a country act under certain circumstances – even though it would remain unclear what those circumstances are.

Kok-Chor Tan (Reference Tan1995), Jennifer Welsh and Miranda Banda (Reference Welsh and Banda2010), and Ang (Reference Ang2015) all acknowledge that the imperfect duty to protect does not entail that any particular nation must intervene, but in their defence of that interpretation, they claim that political organizations must step in to establish clear and enforceable obligations. Tan argues that the international community should establish laws that specify when and who should intervene (1995: 41–2). Welsh and Banda add that there must be ‘specific remedies’ when states fail in their protective role, based on ‘legal liability of the wrongful party’ (2010: 228). Ang simply defers to the United Nations, which she claims definitively establishes the ‘law of nations’ (2015: 46–7). None of these approaches actually settles the question, however, since they do not justify the grounds on which these political decisions are made. That is, they do not explain the theoretical basis of practical decisions about who is responsible and when they have failed in their responsibility. Thus they either leave the international community ‘groping about in experiments and experiences’ (TP, 8: 276) or they beg the question by presupposing some unspecified higher-order theory that guides our collective political decisions (e.g. Tan’s appeal to ill-defined ‘special relationships’ and ‘special capabilities’ that would obligate some nations more than others: 2006: 96–102).

In her defence of humanitarian intervention as a perfect duty, Bagnoli (Reference Bagnoli2006) claims that the formula of humanity enjoins us morally to intervene in order to fulfil our duty of respect, and that countries are obligated to coerce other countries that act wrongly just as they punish or threaten punishment when their own citizens break the law. But these arguments are misguided, both because Bagnoli mischaracterizes the existing international situation and because she fails to appreciate the conditions that affect whether we should intervene. To begin with the second line of argument: states are not related to other states in the same way that states are related to their own citizens. Rather, Kant claims that states are related to other states as individuals are related to other individuals; each state exists as a ‘moral person’ (TPP, 8: 344). As long as its problems are internal to it – that is, they do not negatively impact other countries – there is a presumption that the state should be allowed to govern itself (TPP, 8: 346).

Bagnoli’s first argument, that our duty of respect justifies military intervention, is more intuitively grounded in the categorical imperative. Yet even this seemingly clear duty becomes complicated when we introduce the inevitable risks of war. On individual moral questions, Kant can give clearer guidance about what to do and why, because the circumstances that must be taken into account are more limited and manageable. It is obvious that I should not kill another person to get what I want, just as it is obvious that the state should not intentionally kill innocent civilians to achieve its political aims. But more complex situations pose more of a problem for state actions than for individual agents. A police officer should not shoot at a criminal who is running through a crowd of innocents because, in doing so, he or she knowingly puts them at an unreasonable amount of risk and uses them merely as a means. When it comes to actions by a state, Kant can say that the end of our action must be to protect others and promote their freedom, but that aim is vague enough that the proper method of achieving it is nearly impossible to determine. Whether innocent civilians are risked unreasonably depends on how precise the weapons are; how reliable the military is, including the particular soldiers who will mount the attack; whether the target is urban or rural, and how populous the site is; whether the intelligence about the civilians in the area is reliable; whether we can accurately distinguish civilians from soldiers who may not be in uniform; and so on. And different people in the government will have different opinions on these issues, so there are higher-order issues regarding whom to trust and whose expertise is most relevant, among other things. Therefore, while the police officer’s proposed maxim can be clearly formulated – ‘I will shoot into a crowd of innocents in order to apprehend the suspect’ – the state’s course of action is uncertain – ‘Based on the opinions of some government officials (which ones?), we will put civilians at some risk (how much?) in order to weaken the regime (enough to make the risk worth it?) so that we can make a republican government more likely (how likely?).’ In short, it is unclear when a military strike would be the equivalent, on a larger scale, of a police officer shooting into a crowd.

The conclusion to take from this is that we cannot assume, as Bagnoli does, that humanitarian intervention should properly be characterized as fulfilling a duty of respect. A government’s actions may be reckless under the circumstances, or it may have unreasonable expectations about the likelihood of success or the level of risk to innocent civilians. Furthermore, given the existence of the United Nations, there is no clearly defined role for individual countries. Thus Bagnoli concludes that she is unsure what implications her theorizing has for political practice:

Bagnoli is conceding that there is a gap between theory and practice, and that Kant cannot tell us who is obligated to act and how they ought to fulfil that obligation. It is thus a matter of judgement whether there is a perfect duty to intervene under these circumstances, even if there is such a perfect duty under some circumstances. In short, even if we were to grant that there is a provisional duty, an imperfect duty, or a perfect duty to engage in humanitarian intervention, we would still need to establish first whether a given case is really a justified humanitarian intervention or an unnecessary threat to innocent lives.

7. Syrian Refugees and the Duty of Hospitality

The current crisis in Syria provides a good test case to determine whether Kant’s cosmopolitan philosophy is sufficiently action-guiding to inform public policy. What we discover is that it yields conflicting conclusions about the key issues. First, there is the question of what duties, if any, Western powers have to Syrian refugees. The responsibility to protect has often been invoked recently, and it involves not only possible military intervention, but the provision of humanitarian assistance when necessary. Kant himself spells out one duty that states have towards foreigners: ‘Cosmopolitan right shall be limited to conditions of universal hospitality’, which is the obligation not to refuse people outright, but to consider their requests (TPP, 8: 357; see also MM, 6: 352). He says that it is ‘limited’ to hospitality, by which he means that we are under no obligation to engage in commercial (or other) relations with them or to allow them to settle in whatever country they choose. A state can decide on its own economic and immigration policies.

Kant claims that the duty of hospitality does not exhaust a country’s obligations, however, when someone is emigrating to avoid grave threats: ‘The other can turn him away, if this can be done without destroying [Untergang] him’ (TPP, 8: 358, emphasis added). Pauline Kleingeld notes that this should be interpreted broadly to include not only death, but also ‘mental destruction or incapacitating physical harm’ (1998: 77). In that case, the state presumably would have to provide asylum to the person, either temporarily or permanently. Kleingeld (Reference Kleingeld1998: 76–7), Benhabib (Reference Benhabib2006: 21–3), and Roff (Reference Roff2012) all draw on Kant’s philosophy to support asylum, and Roff appeals to the universal right of hospitality to make a claim on behalf of Syrian refugees in particular. Since the purpose of the state is to protect individuals – for Kant, the focus is always on the humanity of the person, which has incomparable value, rather than the state itself, which has only instrumental value – helping Syrians to escape a very dangerous situation seems to be justified by the formula of humanity.

Furthermore, Western nations are party to international agreements that enumerate clear duties towards asylum seekers: the Universal Declaration of Human Rights was adopted by the UN General Assembly in 1948 (Article 14 covers asylum), the Geneva Convention on Refugees and Asylum was ratified by the countries of Europe in 1951, and the US signed the Protocol Relating to the Status of Refugees in 1967. All three agreements exemplify the Kantian ideal of international cooperation and a recognition of something like world-citizenship and the corresponding duty to protect refugees. The principle of non-refoulement especially follows Kant, purposely or not, in prohibiting countries from returning emigres to places where they would face persecution.

Still, there are reasons why the duty to help Syrian refugees is not so simple, and indeed why it may be limited in ways that are obscured by appealing to the duty of hospitality. Although this may seem like merely a question of customary compliance to ethically informed international agreements, a nation also has other obligations, both to its own citizens and to the Syrian people, that complicate things. First, one must consider how taking refugees out of Syria will affect the country’s future prospects for self-governance. As Scruton notes, if other countries take in refugees, then oppressive regimes can simply get rid of their political opposition without having to kill them, and this helps to maintain their stranglehold on the state (2005: 99–100). Assad supporters are not fleeing the country, so those who could oppose his rule or could participate in a future republican government are the ones establishing roots elsewhere.

Second, any possible duty to give safe haven to refugees is qualified by a number of consequential factors, including possible security concerns in the host countries, and issues of justice, including how to distribute the burdens of caring for them. Even those who use Kant’s duty of hospitality to justify asylum for refugees acknowledge that nations have a sometimes competing duty to defend their own citizens’ welfare. The Syrian refugee crisis brings these confounding factors to the fore. A recent US government study found that Islamic terrorists have infiltrated and plan to continue infiltrating refugee flows into Europe and the United States, and it casts doubt on the idea that effective screening procedures could be developed (US House of Representatives 2015). Therefore, any government must weigh its duty to Syrian refugees against its duty to protect its own citizens, and it must make a series of judgements: how great the risk is, what level of risk is acceptable, what screening methods to use, whether refugees can be effectively monitored once they have arrived, and so on. Benhabib concedes that ‘the lexical ordering of the two claims – the moral needs of others versus legitimate self-interest – is vague, except in the most obvious cases’ (2004: 36). We are left with doubt, not a spur to action.

Furthermore, a refugee crisis of this magnitude has an enormous impact on host countries economically, politically, and socially. The Migration Policy Centre at the European University Institute estimates that 11 million Syrians have been displaced since March 2011, with 4.8 million having fled to Syria’s immediate neighbours and 1 million requesting asylum in the European Union. In addition to the costs of screening, housing, and feeding them, a large influx of Syrian migrants to a given country affects the job market there, since many refugees work illegally in the informal economy, and drives up the cost of housing rentals, among other things. For example, one out of every four people now in Lebanon has fled Syria (about 1.5 million total refugees), and as a result many of the nearly 1 million Lebanese living below the poverty line have been priced out of housing, the government is struggling to provide public services, and many workers are now unemployed because they were replaced with low-paid migrants (Marsi Reference Marsi2015). Even if we could establish that Syria’s neighbours and Western nations have a Kantian duty to take in Syrian refugees, it raises larger questions of how justly to distribute burdens. This is not settled simply by determining who can best afford it, since more than one nation could afford it. It depends on what states should have to afford.

Determining which nations are obligated to accept Syrian refugees thus hinges on a number of factors that are not addressed by a simple appeal to hospitality. Is Germany more obligated to accept asylum-seekers because they have more resources? It depends on our conception of justice and fair distribution. Should Belgium or France be exempted from their obligation in light of recent violence by Islamic militants? Screening for terrorists is a perplexing process for any nation. Should a country that is targeted by specific threats be excused from fulfilling the duty of hospitality? How serious do the threats have to be? And what about countries that have received no direct threats but are worried nonetheless? Although the Kantian duty towards refugees seems stronger as a general rule, all of these questions about specific accommodation are difficult to answer, and they involve judging what is reasonable without decisive recourse to Kantian principles.

8. ISIS and the Right of Prevention

The second issue concerns how the United States and its allies should respond to the Islamic State of Iraq and Syria (ISIS), which infiltrated Syria in 2011, recruited anti-Assad fighters and now has strongholds in northern and eastern parts of the country. Kant says that, although states must work with other members of the league of nations, there is no such legal obligation with regard to independent groups, especially ones that pose a threat to other countries. The fact that ISIS is fighting against the established government in Syria means that, because of Kant’s wholesale opposition to rebellion, they are acting contrary to the conditions of right (MM, 6: 320, 372; TP, 8: 299–300). Furthermore, nations may go to war in order to respond to threats, and Kant defines threats in broad terms to include both other parties’ preparing for aggression and ‘the menacing increase in another state’s power’ (MM, 6: 346). If ISIS acquires even more territory in Syria, it will increase their resources and provide them with areas in which to train fighters and from which to launch terrorist attacks beyond Syria’s borders. And, since ISIS has released videos threatening to attack the US and other Western countries, the United States seems justified on Kantian principles to strike ISIS in Syria, based on ‘the right of prevention’ (MM, 6: 346).

One potential problem with attacking ISIS fighters in Syria is that the US and its allies may accidentally harm and kill innocent civilians. There is a risk of killing innocents in any military attack, of course, but the risk is particularly acute with ISIS, since the group occupies heavily populated areas and uses civilians as shields, American intelligence has few sources in Syria that would indicate precisely where to target, and the US is coordinating with isolated rebel groups rather than a central government that would help to identify appropriate targets (Beauchamp Reference Beauchamp2014). Given how likely it is that military strikes will result in the death of innocent civilians, Nigel Dower (Reference Dower2014) makes the case that, in using the responsibility to protect to justify military intervention, our means are inconsistent with the end we hope to promote. If we are intervening in order to protect innocent human life, trying to achieve that end at the expense of innocent human life is irrational: using military force ‘to promote human rights or restore a just peace … involves violence and the violation of human rights in the process’ (Dower Reference Dower2014; see also Williams Reference Williams2012: 117). Thus armed intervention seems to grate against the hypothetical imperative: we are willing an end (peace and protection) without willing the appropriate (peaceful and protective) means to that end. More importantly, Kant is unequivocal in his condemnation of deliberately harming innocent people to achieve even laudable aims. That would amount to using them merely as means.

As noted earlier, the issue is not so simple. Whether the possible death of innocent civilians is practically relevant to our decision, such that it would override other obligations, depends on such factors as how likely we are to accidentally kill them. This is not a utilitarian calculation, since we are not weighing the deaths of civilians against the lives that are saved by averting the threat. Rather, we must make a judgement about how reasonable it is to think that civilians will die were we to engage in military strikes against ISIS, and whether any such civilian deaths would be inadvertent (or inadvertent enough?), such that it should not be considered as part of our motives – a case of the doctrine of double effect. Roff says that we cannot know whether accidentally killing civilians involves a failure to protect or not, and she leaves it at that (2013: 137–8). Again, Kant’s theory gives us little guidance.

9. Bashar al-Assad and the Duty of Non-Interference

Since the outbreak of the Syrian Civil War in 2011, the government of President Bashar al-Assad has engineered the murder and intimidation of members of the opposition and perceived members of the opposition, resulting in mass killings of civilians. Yet there are reasons both to support and to reject calls for Western powers to attempt to overthrow the Assad regime (either directly, through military action, or by funding rebel groups). On the one hand, Kant gives several arguments in favour of non-interference, even when a country is acting unjustly. The fifth preliminary article for perpetual peace states clearly that ‘no state shall forcibly interfere in the constitution and government of another state’. Republican states are not justified in taking sides in an internal dispute that ‘is not yet critical’ or in using force to coerce other states into an international relation of right (TPP, 8: 346; see also MM, 6: 344). Kant is particularly concerned about the threat of chaos and lawlessness. An unjust ruling power is preferable to no ruling power at all. If rebel groups or other countries could overthrow an existing government at will, it would be equivalent to a state of nature, since in that case the law would have no authority. They could simply reject the system and work to remove the people in charge, without there being any power that adjudicates between the two sides (TP, 8: 299–300).

States must allow other states to develop at their own pace and in their own way. Kant does, after all, enjoin citizens to obey even unjust laws; a system should be changed through rational reflection and public debate, not armed conflict (WIE, 8: 35–42). Thus Sabine Jaberg (Reference Jaberg2002) and Howard Williams (Reference Williams2011, Reference Williams2012) both argue that Kant’s cosmopolitan arrangement grows out of law-governed international relations among republican nations, rather than by having a self-appointed defender of cosmopolitanism forcing the system on despotic governments. We should respect the international law that exists, and seek to enhance it, rather than having one country assume unilateral authority over others. Intranationally, Jaberg and Williams recommend an incremental approach to political progress, with republican states serving as good examples to non-republican states: ‘encouragement’ rather than ‘active interference’ (Williams Reference Williams2011: 118) or ‘persuasion and not compulsion’ (Williams Reference Williams2012: 137). Kant says that the existence of a despotic state can even have a positive purpose: it should, like an immoral individual, serve as a ‘bad example’, showing other governments the harms of despotism (TPP, 8: 346).

On the other hand, when despotic states assert their right against non-interference, this seems to assume that our obligation to protect people’s basic rights and well-being extends only to our fellow citizens, when in fact the spirit of the categorical imperative is to consider the humanity in all people. On Kantian grounds, we could argue that our moral obligations override any supposed duty to respect a state’s political independence, especially if a government is murdering its own citizens. Furthermore, a sovereign whose publicly expressed will cannot be universalized is what Kant calls an ‘unjust enemy’. The ‘violation of public contracts’, for instance, shows that the government is not committed to peace, but rather is perpetuating a state of nature with regard to other states. To choose only one example, Syria has breached the International Covenant on Civil and Political Rights (ICCPR), which it ratified in 1976. Under the circumstances, one could argue that Western nations can intervene, militarily or otherwise, to get Syria to the point where it will ‘adopt a new constitution that by its nature will be unfavorable to the inclination of war’ – namely, one that establishes a republican order (MM, 6: 349).

Kant says that there is no right to rule, and that an unjust system ought to be replaced with a just system. A state’s authority is dependent on its popular sovereignty, or the fact that its people could consent to the state’s coercive enforcement of rights (WIE, 8: 39, TP, 8: 297). Merely formal authority – the fact that the state is enforcing some kind of law – is not enough to justify the sovereign’s position, even though the people have no right to revolution. Thus a state that violates its subjects’ basic rights is in some sense already in a state of nature; if sovereignty is established by a relation to right principles – if, as Kant says, it is ‘indispensably necessary to join the concept of right with politics, and even to raise it to the limiting condition of politics’ (TPP, 8: 372) – then there is no sovereign per se. Even Cavallar, despite his aversion to war on Kantian grounds, claims that a totalitarian state (as distinct from a despotic state) lacks moral personality and that we may intervene out of respect for its people, even if state sovereignty were ‘formally infringed’ (1999: 88–93). When such circumstances exist, it seems that other countries may intervene to protect the people, to establish a government with popular support, and to bring the state into a relation of right with regard to other nations – in short, ‘to establish a condition more closely approaching a rightful condition’ (MM, 6: 344). If the state is no longer a state – if, for example, warring factions have split it into two parts, each of which lays claim to the whole – then we can intervene without ‘interfering in the constitution of another state’, since, Kant claims, the state as such no longer exists (TPP, 8: 346).

Given the tension between a state’s right to be free from external interference and our obligation to defend the rights of the oppressed, Michael Doyle claims that Kantian liberalism is ‘congenitally confused’ in its analysis of situations that may call for intervention (1983: 330). What to do in the case of Syria hinges on whether it has entered a state of nature or it is merely an established system of law facing opposition from rebel groups. A recent editorial in The Economist (2013), one that explicitly references Kant, recognizes this dilemma and comes down on the side of intervention, claiming that there are two separate states both claiming rights over Syria, yet the editors give no reason why we should believe that to be the case rather than thinking of it as an internal dispute between an established government and a (rightfully) disgruntled subset of the population. The latter interpretation would make it an internal matter rather than a justification for humanitarian intervention. How we decide this question determines which part of Kant’s practical philosophy is relevant here – the responsibility to protect or the duty of non-interference. This too is a matter of judgement.

10. The Means to Progress

Kant puts no trust in the idea that progress will be achieved ‘by way of revolution, by a leap, that is, by violent overthrow of an already existing defective constitution’, which seems to speak against interventionism. Instead, Kant recommends ‘gradual reform in accordance with firm principles’ (MM, 6: 355; see also WIE, 8: 36). Thus Williams, citing the ‘Hobbesian moorings of Kant’s political philosophy’, calls for a ‘reformist outlook’ that respects national sovereignty (2011: 104, 108). But both Kant and Williams are too sanguine about the progress of history and the inevitable move to self-government. Kantians may assume that we can come to a conclusion collectively, through the public use of reason. However, the United Nations, the supposed venue for such negotiations, often acts not as a general will, but merely a place for states, including non-republican states, to advance their own national interests. Although Kant believes that states will reform internally through free public debate, in light of numerous counterexamples he claims that the nations of the world inevitably move towards greater freedom and perpetual peace because of ‘good fortune [Glück]’, ‘providence’, ‘fate’ or a ‘plan of nature’ (TPP, 8: 356, 360–2; IUH, 8: 18, 29–30; CB, 8: 120–3; Rel, 6: 122). Williams resigns himself to the fact that this is where Kant must end up: ‘it is more or less inevitable that Kant has to fall back on the idea of one true church and a benign nature or Providence to ensure the ultimate success of his project’ (1983: 259). Yet a faith – even a practically justified faith (see Lindstedt Reference Lindstedt1999) – in moral and political progress is no substitute for a plan to improve. The United Nations has taken a gradualist approach for the six years that the Syrian conflict has been taking place, so it seems that we have a ‘mere wish’ that Syria become a representative republic rather than willing it to be so (G, 4: 394). It is cold comfort to talk about providence while tens of thousands of innocent civilians are being killed by the Assad regime and millions more struggle to live as refugees.

One could defend Kant by claiming that he is less interested in the methodological question of how to deal with particular challenges than in providing a conceptual space for what he calls the ‘moral politician’. That conceptual space designates an area that is not occupied by philosophers, but by decision-makers who are philosophically informed and practically orientated. They are tasked with bringing the domestic and foreign policies ‘into conformity with natural right, which stands before us as a model of the idea of reason’ (TPP, 8: 372). Such a politician must guide the ship of state in a way that is informed by duty, and if he does so, Kant claims that the end will be accomplished of its own accord (TPP, 8: 377–8). The problem is that this merely pushes the question back. The moral politician must himself employ judgement to determine how best to achieve perpetual peace, and to move towards that end ‘as favorable circumstances arise’ (TPP, 8: 378). Kant gestures at the kinds of compromises such a person must make: allowing a despotic ruling power to continue until the people are ready for self-rule, refusing to give up a despotic constitution if doing so would put the nation at risk of being conquered, and so on (TPP, 8: 372–3). Whether it is necessary to allow despotic rule to continue is a matter of judgement, based on an assessment of human nature, historical circumstances, and political will, among other things. Even moral politicians who adhere to the same Kantian principles will arrive at different decisions depending on their analysis of, for example, whether people are in fact ready for self-rule.

11. Conclusion: That May Be Correct in Theory … Redux

With regard to the three most pressing issues arising out of the crisis in Syria, nations seem to have conflicting duties: a duty to protect and provide for their own citizens alongside a duty of hospitality to refugees, a duty to diffuse threats from ISIS while refraining from killing non-combatants, and a duty to promote self-governance while also respecting the rule of law (however flawed) and avoiding plunging Syria into a state of nature. Of course, Kant would phrase it differently: there are no conflicting duties, only a judgement regarding how to formulate the relevant maxim. The problem is that it is exceedingly difficult to characterize and evaluate our national policies given the complexity of international relations and our inevitable biases. Without an existing league of nations (as Kant understands it), Roff says that ‘we have nothing more than our private judgment in this matter’, but even such a private judgement would have to be justified by appealing to Kantian principles, and her implication is that those principles would lead to different judgements for different people, based on the same reasoning (2013: 138).

Other theories are not so problematically vague. For example, under utilitarianism, clarifying the facts can in principle solve the problem of what to do to produce the best consequences overall – e.g. whether a pre-emptive strike will cost more lives than it saves by preventing future attacks. And for Hobbes, our judgement is limited only by our powers to predict what will advance our national interests – e.g. whether respecting a non-aggression pact will weaken or strengthen our security. For Kant, by contrast, the theory can lead to different conclusions even if we agree on all of the facts; it depends on which facts are emphasized as most important or how they are perceived by the affected parties.

None of this means that Kant’s position is useless. But we must be clear about what Kant is doing and not doing. He gives us an ideal to pursue – self-government, protection of individual rights, care for those in danger, international association, the avoidance of war through rational discussion, and so on – but when it comes to the question of how to act now, in the non-cosmopolitan present, most of the heavy lifting is done by judgement. In short, Kant can convincingly explain that we need to promote freedom and individual rights throughout the world without guiding us on how to do so. This is where the theory–practice divide re-emerges, and where Kant’s political theory is susceptible to the empty formalism charge. As Hegel claims, what is reasonable varies based crucially on circumstances and on cultural and historical expectations. Such variability is out of place in Kant’s theory of absolute duties and problematizes his claim that what is true in theory must also be true in moral and political practice.Footnote ²