Toleration,decency and self-determination in The Law of Peoples

Introduction

According to Stephen Macedo, Rawls’ The Law of Peoples (Rawls, 1999b [hereafter cited as LP]) tries to find a balance between the Scylla of imposing one’s way of life on others and the Charybdis of letting respect for diversity, alone, determine the bounds of legitimate difference (Macedo, 2004: 1733). I believe that LP achieves a reasonable compromise between these extremes. Even more importantly, I believe that such compromise is conceptually less weak than many have alleged. While positive appraisals of Rawls’ account of international toleration do appear in the literature (see Porter, 2012; Dogan, 2010; Jenkins, 2010; Reidy, 2010; Avila, 2007; Freeman, 2007a; Mertens, 2005; Hayfa, 2004), this article concentrates on two objections that have drawn comparatively less attention and that, if successful, would prove fatal to the Rawlsian project.

While there has been considerable critical attention concerning whether liberal peoples should tolerate decent ones, the latter debate has not convincingly addressed the possibility that a reason for not tolerating decent peoples is that they represent, by conception, an internally incoherent idea. This is precisely the objection I address in the first part of the article. The latter claims, in short, that the very idea of a decent people does not cohere with Rawls’ understanding of reasonable pluralism and sanctions the oppressive use of state power by the government of a decent hierarchical people (see Neufeld, 2005; Caney, 2002). ¹

In the same way, while many have criticized Rawls’ account of human rights (see, for example, Buchanan, 2000), much less attention has been given, in the Rawls-friendly literature, to the actual choice of principles in the first international original position of LP and whether the latter can be understood and justified without giving in to a desire for moral compromise with decent peoples. The second part of the article addresses the core of this worry. According to many liberal cosmopolitans, liberal peoples should agree to a different and more expansive set of principles in the first international original position of LP: representatives of liberal peoples agreeing on a conception of international justice would include a more substantive set of human rights and some form of egalitarian distributive principle (see Pogge, 2006, 2004, 2001, 1994; Beitz, 2000). This would automatically exclude decent peoples from the scope of toleration as they would clearly be unable to respect, for the right reasons, a law of peoples which would require all societies to be internally liberal.

I find both objections unconvincing. The first objection does not properly distinguish between the use of state power aimed at curtailing difference and the oppressive use of state power. To define any use of state power as oppressive we need a benchmark, and such benchmark is provided by the essential elements of a conception of justice. In turn, Rawls’ critics seem to use a liberal conception of justice as the relevant benchmark against which to judge the use of political power within a decent hierarchical people. But to do so means that we fundamentally premise our judgement of the internal political life of a decent hierarchical people on whether it meets the standards provided by the core elements of a liberal conception of justice. This is, simply put, a non-starter as it premises our understanding of a non-liberal society on the background assumption that it should respect a liberal conception of justice.

The second objection fails to see that transposing a more liberal egalitarian set of principles in LP would entail the replication of the same entitlements within different levels of governance. In the first international original position, representatives of the parties know that they represent liberal peoples, so there is no clear reason for them to duplicate within LP’s principles what they already know is part of the conceptions of domestic justice adopted by the societies they represent (see Freeman, 2007a, 2007b). Furthermore, expanding the set of domestic issues overseen by principles of international justice affects one of the basic interests of a liberal people: its self-determination (see also Buchanan, 2013).

The article begins by providing a brief overview of the account of international toleration in LP. It then addresses in the next three sections the apparent tension between decent hierarchical peoples and Rawls’ understanding of reasonable pluralism. Dealing with challenges to the outcomes of the first international original position occupies the three final sections. The contribution of this article to the existing literature is twofold. First, the article reshapes some of the standard objections levelled against Rawls’ account of international toleration and provides a new set of arguments to address them. Second, the article highlights how the alleged incoherence between Rawls’ domestic and international theories is often premised on the assumption that all societies should be internally liberal and that the correct view of global justice has to follow a specific interpretation of cosmopolitanism; namely, one in which the value of collective self-determination is taken to be marginal compared with the realization of standard liberal individual rights. While I do not argue against the content of these views, the article highlights how they should feature as conclusions rather than assumptions when we address the coherence of the Rawlsian project and the question of international toleration.

Toleration in The Law of Peoples

Rawls’ account of toleration concerns the status of decent peoples. ² According to Rawls, ‘[a] main task in extending the Law of Peoples to nonliberal peoples is to specify how far liberal peoples are to tolerate nonliberal peoples’ (LP: 59). For Rawls toleration does not simply entail refraining from coercing certain subjects (ibid.); it means the extension of respect to those who are tolerated, and in the specific case of LP, to respect decent peoples as members in good standing (bona fide) of the Society of Peoples. Seen from the standpoint of a liberal democratic order, the way in which a decent people treats its members is clearly unacceptable (ibid.: 83). Yet a decent people respects human rights, ³ consults its members when important decisions have to be taken, and follows a just law of peoples in its foreign policy. These three factors, when considered together, mean that the institutions of a decent people have some features that deserve respect, and meet the standards required to ‘override the political reasons we might have for imposing sanctions’ (ibid.).

There are many possible ways of organizing decent political institutions. In LP Rawls proposes to survey the main criteria of decency for a decent hierarchical people (LP: 64). According to Rawls, in order for a decent hierarchical people to be a member in good standing of the Society of Peoples it has to meet two criteria. The first is external: it states the permissible foreign policy aims for any member of the Society of Peoples – namely, a non-aggressive foreign policy. Decent hierarchical peoples respect ‘the political and social order of other societies’ and if they seek influence internationally, they do so ‘in ways compatible with the independence of other societies’ (ibid.). The second criterion of decency is internal, and is divided into three parts. First, the system of laws of a decent hierarchical people must secure basic human rights (ibid.: 65). Second, ‘a decent people’s system of law must be such as to impose bona fide moral duties and obligations (distinct from human rights) on all persons within the people’s territory’ (ibid.: 65–6). Third, ‘there must be a sincere and not unreasonable belief on the part of judges and other officials who administer the legal system that the law is indeed guided by a common good idea of justice’ (ibid.: 66). According to Rawls, the first two parts of the internal criterion of decency require that the system of law of a decent hierarchical people be guided by a common good idea of justice (ibid.: 71). Part of the definition of a common good idea of justice requires the specification of a decent consultation hierarchy. A decent consultation hierarchy ‘must include a family of representative bodies whose role in the hierarchy is to take part in an established procedure of consultation and to look after what the people’s common good idea of justice regards as the important interests of all members of the people’ (ibid.).

A recurring objection to Rawls’ account of toleration (see especially Tan, 2005, 2000, 1998) is that the latter is based on the false analogy between comprehensive doctrines within liberal democracies and decent non-liberal peoples in international society. Rawls’ critics believe that the latter analogy is clearly flawed: while domestic toleration is based on the acceptance of the same (or a family of) liberal political conception(s) of justice, international toleration is arguably not based on the endorsement of liberalism as a constraint to the permissible differences in political organization (see, for example, Tan, 2000: 29). A commitment to ethical neutrality necessarily implies a commitment to a particular political system – one that protects ethical neutrality in the first place. Yet, at the international level, Rawls advocates the toleration of peoples that do not have liberal political institutions.

In what follows I will not address this objection to Rawls’ view. This is for two reasons. First, because the alternative objections I will discuss have, so far, received comparatively less attention in the critical literature more sympathetic to the Rawlsian view. Second, I believe that the disanalogy objection I have just presented has recently been given a strong answer by Thomas Porter (2012). Porter highlights how at the basis of the analogy between comprehensive doctrines in Political Liberalism [hereafter cited as PL] and decent peoples in LP lies a fundamental principle of political justification. Such principle states that liberals are required to provide a justification to non-liberal agents who are willing to comply, for the right reasons, with a liberal conception of justice. Nonetheless, even accepting Porter’s argument, the Rawlsian view of international toleration still requires to be defended against the objections I have mentioned in the introduction. Porter’s argument, if successful, will show that we should aim to provide a justification to decent hierarchical peoples that they can reasonably accept. But it still leaves open two basic questions. Can the very idea of a decent hierarchical people really be considered as consistent with the Rawlsian concern for reasonable pluralism? And what are the principles that decent hierarchical peoples should be asked to comply with if they are to be members in good standing of the Society of Peoples? Without a clear answer to these questions the case for Rawlsian international toleration cannot be complete.

Reasonable pluralism at home and abroad

Rawls’ PL responds to a crucial problem concerning the very coherence of the liberal democratic project. As Rawls tells us, PL seeks to answer the following question: ‘How is it possible for there to exist over time a just and stable society of free and equal citizens, who remain profoundly divided by reasonable religious, philosophical and moral doctrines?’ (PL: 4). The fact of reasonable pluralism sets limits to the depth and breadth of the type of agreement on political principles that we can realistically hope to achieve. In addition, the fact of reasonable pluralism implies that comprehensive doctrines cannot form the basis of a justification for the use of political power in a liberal democratic polity (see Scanlon, 2003).

Rawls supports his analysis of the legitimate use of coercive power in a liberal democracy, and its associated requirements of public justification, by setting out what he believes to be three enduring facts that characterize the political culture of a democratic society. First, the diversity of moral, religious and philosophical doctrines peculiar to democratic societies is not a mere historical accident but a permanent condition. Second, the continued allegiance of such societies to any comprehensive doctrine would require the oppressive use of state power. Third, the enduring survival of a democratic regime not rife with divisions and conflict necessitates the free and willing support of its citizens (PL: 36–8).

These three facts are central to our discussion of international toleration in LP as they seem to cast a shadow over the idea of decent hierarchical peoples. They are the basis of a significant alleged inconsistency between Rawls’ domestic and international theories of justice (see Neufeld, 2005). First, decent hierarchical peoples are, at least to some extent, pluralistic in nature. Since religious minorities and political dissent are part of what characterizes a decent hierarchical society and a decent consultation hierarchy, we should not imagine decent societies as monolithic. Diversity is clearly one feature of their public cultures, and there is no reason to believe that such diversity will not be permanent. Second, decent hierarchical peoples have a common good conception of justice which is based, at least according to Rawls’ own view, on a comprehensive (or at least partially comprehensive) doctrine (PL: 65–6; Neufeld, 2005: 289). Thus, as suggested by Rawls’ analysis of a liberal democratic polity, their social unity seems to require the oppressive use of state power. Third, decent hierarchical peoples are, according to Rawls, legitimate in the eyes of their citizens and are freely supported by at least a majority of them. Yet, given the fact of diversity and the fact of oppression (as stated above), this conclusion seems unwarranted.

Much of the appeal of the account of a decent hierarchical people rests on the fact that its institutions respect human rights and are freely supported by their members. However, if we follow Rawls’ analysis of the consequences of pluralism and the need for the oppressive use of state power to counter pluralism, then, the ideal picture of decent hierarchical peoples seems to be inconsistent with Rawls’ discussion in PL. In short, in PL and LP Rawls seems to draw different conclusions for what look like similar circumstances. If this were the case, the very idea of a decent hierarchical people as imagined by Rawls – its being well-ordered according to a common good conception of justice – would seem impossible to obtain. In this picture, the idea of a decent hierarchical people would lose much of its appeal too: it would be based on what Rawls himself sees as oppression.

A liberal benchmark for non-liberal peoples?

The critical force of the argument I have just presented is based on how we understand the term oppression and, relatedly, what kind of interpretation we make about the genesis of the idea of reasonable pluralism. In what follows I show that the alleged inconsistency between Rawls’ PL and LP presented above is premised on eliding the distinction between the oppressive use of state power and the use of state power aimed at curtailing difference. As we have seen above, one of Rawls’ main concerns in PL is the oppressive use of state power. What we should immediately recognize regarding the oppressive use of state power is that simply curtailing difference or pluralism in social life is not, per se, normatively impermissible and cannot, ipso facto, be equated to a form of oppression (see Macedo, 2000). This observation should alert us to the fact that by designating any exercise of state power as oppressive we are not simply signalling that it is a coercive exercise of power for the curtailment of difference; if that were the benchmark, most exercises of state power would be oppressive. This is a point worth stressing as it signals that defining something as an oppressive use of state power seems to presuppose a prior conceptual exercise, namely that of defining the limits of legitimate diversity. The very act of defining something as an oppressive use of state power seems to presuppose an account of how state power should be used. And, in turn, such account is provided by the essential elements of a conception of justice. ⁴ This should give some indication as to why the use of state power within decent peoples is not necessarily oppressive, even when it is based on some form of comprehensive reasoning. It is only when basic constitutional freedoms of the liberal democratic tradition have already been taken for granted that the use of state power to require citizens’ allegiance to any comprehensive doctrine can be defined as oppressive. In a liberal democracy, the use of state power to maintain citizens’ stable allegiance to a comprehensive doctrine is oppressive because it is impossible to do so (to maintain such allegiance) without violating the very rights and liberties that a liberal society recommends (see Cohen, 1994). Extending the same type of analysis to decent hierarchical peoples, we would simply define as oppressive any use of state power that does not guarantee the expressive and deliberative freedoms that are part of the liberal democratic tradition (and that define the basic content of any liberal conception of right). Yet, this would not be a sound way to proceed, as part of our task when discussing the idea of a decent hierarchical people, is precisely to describe a non-liberal society and to establish whether such society is to be tolerated.

Put differently, to define something as an oppressive use of state power necessitates a benchmark: oppression is a derivative concept. The benchmark that Rawls adopts for liberal societies is a liberal one: the use of state power that is required to unite citizens under the same comprehensive doctrine is oppressive when it violates the basic constitutional guarantees that are part of the liberal tradition. If we were to use the same benchmark to define as oppressive the use of state power in a decent hierarchical society we would in fact presuppose that all types of regimes need to protect exactly the same basic liberties that are part of the liberal democratic tradition. However, to do so would premise our analysis of decent hierarchical peoples on the assumption that all peoples should be internally liberal. Leaving aside the counter-intuitive nature of this conclusion, it is important to note that we cannot use such conclusion as the starting point of our analysis of decent hierarchical peoples and of the coherence of the Rawlsian account of international toleration.

Here, Rawls’ critics could provide two distinct rebuttals. First, they may wish to maintain that there is nothing inherently unsound in seeing liberal political theory as universal. Why are we not to assume that ‘all societies should be liberal’? Is not Rawls himself explicitly committed to this view given that he sees decent hierarchical peoples as not fully reasonable? Second, they may add that if toleration means refraining from using coercion (broadly understood) for practices that we find morally problematic, then, why should we not simply see decent hierarchical peoples as inherently oppressive and internally unstable and, at the same time, claim that they are to be tolerated for other, independent, reasons?

Let me answer these concerns in reverse order. It is not accurate to state that the Rawlsian view of toleration just concerns our reasons from refraining to use force or coercion. Many accounts of toleration are based on the latter idea, but the Rawlsian one sees toleration as an attitude based on mutual respect (McKinnon, 2006). If this is correct, then, we cannot simply allow decent societies to be internally oppressive. If we did, the attitude of respect for these societies that lies at the heart of the Rawlsian view of toleration would prove to be at the very least counter-intuitive. One way charitably to reconstruct the Rawlsian view, then, is to say that Rawls is suggesting that we see decent hierarchical peoples as not fully unreasonable, and that not being fully unreasonable is compatible with liberal peoples respecting them. Decent hierarchical peoples are not fully unreasonable for two reasons. The first has to do with the way in which their internal political institutions are organized, that is, the fact that they respect human rights, and allow their members to provide input in the political process. The second, and even more important, reason is because decent hierarchical peoples endorse for the right reasons the same conception of international justice that liberal peoples give to one another. Both reasons allow LP to create a special conceptual space for toleration. Decent hierarchical peoples are not fully just but we can still respect their institutions and foreign policy.

The first objection, on the other hand, is harder to respond to in the space available here. Nonetheless, I would like to suggest that it fails to hit its mark. In the first section of the article I have argued, following Thomas Porter (2012), that a strong element of commonality between Rawls’ domestic and international conceptions of liberal political philosophy concerns his view of political justification. According to Rawls, we start with a view of political justification in mind and then approach the question of toleration from this vantage point by asking what kind of societies we should tolerate internationally if we were consistently to respect our commitment to this specific account of political justification. Needless to say, this is a controversial view. Yet, it is a view that provides a distinctively liberal approach to toleration (and political philosophy more broadly), if, as many have argued, we see it as grounded in respect for persons (Larmore, 1999, 1996, 1994). In this picture, then, the universality of liberalism is provided by a more complex picture than the objection above seems to suggest. In this complex picture, our commitment to an account of political justification and our commitment to a liberal conception of domestic social justice come into conflict in the international arena: the first tells us that some non-liberal political societies are to be respected while the second maintains that they are not fully reasonable because they fall short of what a liberal conception of justice requires (see Reidy, 2013: 191–3). How this conflict is to be solved, and how we are to judge the trade-offs that it implies, is not something I will settle here. Rawls seems to believe that the best way forward is to see these non-liberal political societies as members in good standing of the Society of Peoples, and hope, over time, that their own ability for internal political change coupled with the historical success of liberal democratic modes of governance will do the rest. However, it is worth noting that recognizing this tension allows us to defuse the worry expressed by the first objection above: that my argument seems to require that we abandon the universal scope of our liberal commitments. If the tension I have highlighted is genuine, it is precisely our liberal commitments that will push us to be cautious about the way in which we interpret the meaning of liberal universalism and how it should inform our views about toleration. ⁵

Political liberalism and the genesis of reasonable pluralism

Rawls’ critics may concede that we cannot simply judge decent peoples using the main tenets of a liberal conception of right as the relevant benchmark. Nonetheless they could maintain that the pluralism that is likely to develop within a decent hierarchical people needs to be addressed in some way. Within the bounds of a liberal democratic polity Rawls offers a strategy to address the fact of reasonable pluralism. Why should we proceed differently within a decent society? Replying to this objection essentially requires an account of the genesis of the idea of reasonable pluralism. Paying attention to this genesis allows us to see that the very idea of reasonable pluralism cannot easily be transposed into what is, by definition, a non-liberal society. Stated differently, reasonable pluralism is not a fact of political life; rather, it is a fact about life under a very specific type of institutional system, namely a liberal democratic one. In what follows I will largely assume Rawls’ explicit concern with the stability of justice as fairness as the guiding theoretical reason for Rawls’ political turn (see Weithman, 2011; Mulhall and Swift, 2003, 1996). ⁶ In this picture, we start from the third part of A Theory of Justice [hereafter cited as TJ]. The latter’s central aim is to show that the ideal of a well-ordered society stable for the right reasons is achievable. Rawls’ original argument for stability has two main threads: the first concerns the moral psychology of citizens growing under just institutions (TJ: 429–40); the second pertains to the congruence (to use Rawls’ term) between citizens’ good and their affirming justice as fairness (ibid.: 496–505). Taken together, Rawls’ arguments aim to show that there is nothing in human nature that prevents citizens who are born under just institutions developing a sense of justice and experiencing their acting on this sense of justice as a good (see Freeman, 2007b). Both threads, according to the Rawls of PL, face an important internal problem (Cohen, 1994). The congruence argument is perhaps more obviously problematic as it directly relies upon the assumption that all citizens will come to share one and the same way of making their conceptions of the good congruent with justice as fairness. Yet, given the burdens of judgment, this is an unreasonable and unrealistic picture. One of the crucial aspects of Rawls’ political turn, and the main message that the idea of the burdens of judgment conveys, is that when citizens’ basic deliberative and expressive liberties are protected, even conscientious uses of human reason, unimpeded by simple bias and self-interest, will not deliver agreement on a single vision of the many non-political values that constitute the core of what Rawls calls a comprehensive doctrine (see Quong, 2011; Dreben, 2003). Under liberal democratic institutions, congruence between citizens’ conception of the good and their allegiance to principles of justice cannot rest on one and the same basis.

A different way of looking at the ground just covered is to say that given PL’s analysis of the use of practical reason under free institutions, we can see that a liberal conception of right plays a dual role for Rawls. On the one hand it determines the boundaries of permissible diversity and sets limits to the types of comprehensive doctrines that citizens can hold. On the other hand it also creates the very conditions for reasonable pluralism to develop in the first place. And it does so by protecting the basic deliberative and expressive liberties that are familiar from the liberal constitutional tradition. It is precisely by attending to the dual role of liberalism for Rawls, as both origin and limiting condition of diversity, that we can better appreciate the internal tension within the third part of TJ. If that is correct, then the oppressive use of state power, at least as it is presented by Rawls in PL, is a problem internal to the liberal democratic tradition (Maffettone, 2011; Quong, 2011). It is thus very difficult simply to transpose the problem to the context of a decent hierarchical people.

Here, Rawls’ critics might object that my defence of Rawls’ view depends on the unreasonable assumption that a decent people would not feature any significant form of moral, religious, or philosophical pluralism (see also Caney, 2002). I wish to deny that. I am not claiming that a decent political society is monolithic, or that no form of pluralism would develop within it. Nor am I claiming that such pluralism has no moral significance and could simply be coercively eliminated. Although pluralism would certainly occur in a decent society, we would still have to adjust our understanding of ‘the idea of reasonable pluralism’ to the moral diversity that is likely to develop within a system of decent (rather than liberal) institutions. In a decent people – that is, in a system of decent rather than free institutions – the division of persons along moral, philosophical and religious lines might be less deep and less irreconcilable. In other words, given the institutions under which persons develop in a decent society, its citizens might be less divided on fundamental moral, philosophical and religious issues.

To repeat, this does not entail that no doctrinal division is likely to emerge; or that such divisions can simply be met via the use of state power. Decent peoples are capable of moral learning (LP: 61) and the dissent of their members is an important aspect of the internal dialectic of change and reform that is part of decent institutions (ibid.: 66 ff.). Not only that, but public officials are required to take this dissent seriously and not to dismiss it as a sign of incompetence. In other words, as Stephen Macedo correctly argues, a decent society respects dissent and is transparent and capable of self-reform (2004: 1734–5). Instead, the point I have been making is that the dissent that would likely emerge within, and be permitted by the government of, a decent political society cannot simply be evaluated against the benchmark of a liberal conception of right. Nor can the pluralism that grounds this dissent be understood to have the same origin or extension as reasonable pluralism within a liberal democratic polity.

I have just argued that the way in which Rawls develops the idea of a decent hierarchical people, the object of Rawls’ theory of international toleration, does not reflect an inconsistent approach to the idea of reasonable pluralism. The next problem I will address concerns, this time, the terms of toleration. According to Rawls liberal peoples should offer decent peoples the same terms of cooperation that they are prepared to offer one another, namely the 8 principles of LP. As we have seen in the first section of the article, this is an important aspect of the argument for toleration in LP. Liberal peoples accept decent ones in the Society of Peoples precisely because decent peoples are able to agree upon the same conception of international justice and because they are willing to comply with such conception for the right reasons. Is it plausible, then, to argue that liberal peoples themselves would adopt the 8 principles of LP as the principles of justice to regulate their mutual undertakings? Are the 8 principles of LP what liberal peoples would agree to in the first international original position of LP?

Minimalism and accommodation

Some critics (see, for example, Pogge, 2006; Beitz, 2000) have denied that we can answer these questions in the affirmative. They have argued that LP is skewed towards the needs of decent peoples. In their view Rawls has artificially expanded the bounds of toleration in order to respond to global diversity. He has, allegedly, done so by limiting the extent to which liberal societies agree with each other. Rawls’ international theory starts by imagining only liberal peoples meeting in a first international original position (LP: 32). Rawls maintains that when representatives of liberal peoples meet to structure their mutual undertakings, they would choose the 8 principles of LP. The 8 principles of LP are, according to Rawls’ critics, minimalist in character and do not incorporate the main aspects of a liberal and democratic understanding of individual rights. Why, the critics ask, do liberal people not agree with each other on a much broader set of liberal principles in the first original position of LP? Since they all share the main aspects of liberal democratic regimes, why do they not agree on establishing these features as principles of a law of peoples? If they did, decent societies would be excluded from LP, since it would be impossible for them to accept international principles of political organization that require domestic political institutions to match liberal standards. In short, the critics claim, the first original position of LP is inclined towards minimalism to accommodate the needs of decent peoples (see also Pogge, 1994: 215–16).

It would be tempting to picture this disagreement between Rawls and his critics as one concerning human rights and their philosophical justification. For if we agreed, for example, that a sound conception of human rights is more demanding than the one that Rawls allows for in LP (for example, if we saw democracy as a human right), then, decent peoples would not be able to endorse the principles of LP. Nonetheless, note two important reasons not to reduce the argument discussed here to a form of disagreement about human rights. First, the fact that the first run of the international original position includes only liberal peoples entails that the latter could have a minimalist view of human rights but nonetheless add further liberal principles to the conception of international justice they are prepared to endorse. Thus, the potential discrepancy between what liberal peoples could agree to and what decent peoples could agree to from the standpoint of the original position, is something we have reason to investigate independently from a conception of human rights. Second, because it is difficult, from a Rawlsian perspective anyway, to justify a given conception of human rights without using the main justificatory tool within the Rawlsian tool-kit, namely, the original position. As Mitchell Avila correctly observes, it would be unclear, conceptually, where to justify a conception of human rights ‘from’ without using the original position (2011).

Let us first consider that there seems to be no real need to include the major features of a liberal democratic regime as the outcome of the first international original position in LP. The basic reason for this is that the rights protected by a liberal constitutional regime are already guaranteed within the different domestic jurisdictions of liberal peoples (Freeman, 2007a: 276). In the first original position in LP, representatives of peoples know that they are representatives of liberal peoples. Thus they know that the main aspects of a liberal democratic regime are already part of their domestic basic structures. What would be the point of replicating that protection at the international level? If we assume, as Rawls does, that we are within ideal theory – that is, we assume that all liberal societies are well-ordered according to their liberal conceptions of justice and face favourable conditions – what is the point of replicating internationally the protection of something that is already secured domestically?

One possible reason for duplication ⁷ is that making a given set of rights part of an international law of peoples is to afford that set of rights a more robust protection. In other words, if more peoples are collectively responsible for the protection of basic liberal rights in the jurisdictions of all liberal peoples, then those rights will be, ceteribus paribus, more secure. If the violation of all relevant liberal rights (however this set is defined) is legally felt in all liberal peoples wherever it might happen, there will be a greater chance that those rights will be secured if conditions for their protection become less favourable within any given liberal polity. Yet, this argument drives together ideal and non-ideal theory scenarios in an improper way. We cannot premise the main reason for duplicating a set of liberal rights within LP on the fact that those rights can be violated in a given liberal polity; the latter idea would entail that the content of a conception of international justice should depend on assumptions concerning non-ideal theory circumstances. But this is not how Rawls’s model of theory construction proceeds.

Duplication and self-determination

Note also that, if accepted, the ‘duplication strategy’ would not only be redundant. It would also have costs in terms of each liberal people’s self-determination. ⁸ Rawls’ critics seem to rely in their argument on the following, apparently inconsequential, assumption (A):

(A): If liberal peoples protect X at the domestic level, then, their representatives in the first original position of LP will agree to incorporate X within the international principles of LP

(where X is the preferred set of fundamental liberal rights).

Would that be problematic from the standpoint of the first international original position in LP? In order to understand the deliberations in the first original position in LP we need to have an account of the basic interests of the agents who are represented in that original position (see Wenar, 2006). For Rawls, liberal peoples are well-ordered according to a liberal political conception of justice and ‘seek to protect their territory, to ensure the security and safety of their citizens, and to preserve their free political institutions and the liberties and free cultures of their civil society’ (LP: 29; emphases added). The interests of liberal peoples thus include what we can call a concern for the free exercise of collective political self-determination.

Such interest is not the only one they have and, of course, the curtailment of the self-determination is not necessarily impermissible. For example, Rawls claims that the function of human rights is precisely to limit the internal political autonomy of peoples (LP: 25–7). But the observation fails to capture the basic issue at stake. The point is not that the self-determination of peoples is untouchable (it is emphatically not). Rather, the issue is whether assumption (A) above can be the basis on which we decide whether or not to include a given set of liberal rights within an international law of peoples. Proceeding through (A) in order to understand what should be part of the principles of a liberal conception of international justice would simply bypass the relevance of self-determination. In doing so, we would fail to consider the types of reasons that representatives of liberal peoples have in the first international original position of LP.

Why would a more expansive and liberal egalitarian set of principles affect the self-determination of liberal peoples? An example may be useful here. Imagine that all liberal peoples (L1 … Ln), at time T = t1, agree that the difference principle is a central feature of a liberal political conception of justice. If we accept the validity of (A), then it would seem natural to assume that the difference principle would be transposed into an international law of peoples and would be protected by international legal agreements. Now, further imagine that at time T = t2, a given liberal people (Ln; n = 2) decides, through democratic procedures, that it does not wish to adhere to the difference principle any more. Can L2 withdraw its allegiance from the difference principle? The answer is: no. Agreement on a public conception of international justice cannot be changed retrospectively by one of the parties.

The idea of a change of heart from times t1 to t2 is of course fictitious given that agreement on a public conception of justice is obtained by using a constructivist tool – the original position – which is a device of representation lacking a temporal dimension. Note though that the idea can be restated without taking into consideration temporality. The basic point is that the domestic public reason of a liberal democratic polity is not a static entity. Nor is the public reason of an international society of liberal peoples. But how are we to know that the two will evolve in the same direction? And what happens when the two come into conflict? ¹⁰ While Rawls does not seem to offer an explicit answer in LP, ¹¹ the implicit idea is that we should try to avoid conflict between these two public reasons by putting in place some form of division of labour: domestic conceptions of justice cover principles of social and political justice for the basic structure of a given liberal polity, while a law of peoples mostly concerns itself with foreign policy. And the (main) reason to proceed this way is precisely to protect the self-determination of liberal peoples, not to accommodate non-liberal societies. ¹² Note also that a strict separation between the two standpoints, that is, a strict division of labour between domestic and international public reasons, is not to be expected. As I have noted above, for Rawls, basic human rights are an important point of contact between the two. The point, once again, is that the integration of the two has costs and implies trade-offs (Buchanan, 2013: 245 ff.). ¹³ The trade-offs cannot simply be ignored. Whether or not we agree with the Rawlsian solution is a different issue. The crux of the matter is that in order to criticize it we need an argument that recognizes this trade-off, not one that ignores it and that is precisely what accepting (A) implies.

Rawls’ critics might retort that the example of the difference principle stacks the deck against them. At least from PL onward, a feature of Rawls’ theory is that liberalism can be interpreted in different ways and that what qualifies a conception of justice as liberal need not include the difference principle (PL). What if, instead, we selected the basic constitutional liberties that are covered by the first principle of justice as fairness? Surely a given liberal people could not claim that its internal democratic procedures have brought it to abandon basic constitutional freedoms. Surely, then, these rights should be part of an international (liberal) law of peoples?

I believe that the answer is not as clear-cut. Even if we were to agree that the rights and liberties that are at the core of a liberal democratic constitution are definitional of a liberal polity (those covered by the first principle of justice as fairness), this does not entail per se that these rights and liberties should feature in an international law of peoples. If they did, the way in which these rights are implemented, the interpretation of their concrete meaning and how they are entrenched in the domestic constitution of a liberal people, would all be turned into matters of international concern and partially removed from the democratic process of each liberal people. To illustrate, in TJ Rawls is clear that the first principle of justice can only provide broad guidance and how the basic liberties that it refers to should be adjusted to one another to cohere into a single scheme of basic liberties. Such adjustment should, to some extent, track the specific conditions of a given society and cannot be simply specified ex-ante (TJ: 52 ff.; Pogge, 2007: 85 ff.). For example, one can conjecture that different basic liberties may have comparatively different values according to prevailing social, economic and political conditions in different societies. In the same way, when Rawls discusses the curtailment of basic liberties he states that the latter may happen only for the sake of the basic liberties themselves (TJ: 54). Yet, decisions concerning the curtailment of basic liberties require some sensitivity to local circumstances. For example, it would not be implausible to think that a liberal people that faces recurrent political terrorism could decide to have more restrictive provisions on political speech indirectly supporting the terrorists, and that such provisions should (in this specific case) ultimately track the collective deliberations of the public in question, not the deliberations of the Society of Peoples.

To repeat, the point here is not necessarily that one needs to endorse the Rawlsian solution, that is, his way of balancing the value of collective self-determination for liberal peoples with other liberal individual rights. Rather, the point is that even considering the liberties covered by the first principle of justice there is a trade-off between their entrenchment in an international conception of justice and the self-determination of a given liberal polity. Those who wish to criticize Rawls have to acknowledge the tension and need to provide an argument to the effect that the balance of values suggested by Rawls is unsustainable. As Allen Buchanan correctly states (2013: 246–8) this ultimately depends on the type of account of global justice that one adopts and on the relative weight that one assigns to self-determination in the latter. One may legitimately disagree with Rawls, but the disagreement cannot be based on the accusation that Rawls is simply compromising the way in which he articulates the results of the first international original position in order to accommodate non-liberal peoples. Finally, note that the latter claim is also consistent with what I have argued above when I have stressed the continuity of LP with Rawls’ general approach to political justification. The Rawlsian toleration strategy is liberal at its heart: it is concerned with the self-determination of liberal peoples and only concludes that decent peoples are to be tolerated derivatively mainly because they are prepared to abide by exactly the same principles liberal peoples would give one another.

Self-determination and liberal citizens

At this stage Rawls’ critics may retort that I have based my discussion on the underlying premise that we should proceed by weighing the interests of liberal peoples in self-determination against the inclusion of a more extensive set of liberal individual rights within an international conception of justice. In other words, the structure of the discussion seems to forget that what ultimately matter in a liberal framework are the interests of persons, not those of the collectives of which they happen to be members. At the very least, the critics may add, we should start with a conception of liberal peoples whose basic interests reflect the basic interests of liberal citizens as understood by Rawls within the bounds of his domestic theory of justice. ¹⁴ Yet, within the Rawlsian domestic framework, citizens do not seem to have a specific interest in the self-determination of their political community.

I want to suggest two replies to this objection. The first reply takes the form of what we can call a reasonable interpretive conjecture. The first step of this conjecture is to note the following structural element of the Rawlsian domestic theory of justice, namely, the assumptions under which a conception of the basic interests of liberal citizens is framed. More specifically, when developing an account of a domestic justice, the fundamental interests of liberal citizens are framed under the assumption that society is closed and self-contained (see also R. Miller, 2011). In other words, in the Rawlsian domestic framework, liberal citizens cannot show any interest in the self-determination of their political community and of their ways of life as a whole simply because they may take it for granted. So, the question we need to ask if we want to assert that there is congruence between the interests of liberal peoples and those of liberal citizens is: What would happen (conceptually) if we relaxed this assumption? Should we revise our understanding of the interests of liberal citizens to take into consideration the fact that when discussing a conception of international justice we cannot assume that society is closed and self-contained? We can develop a reasonable (interpretive) conjecture that the answer would be affirmative, within the Rawlsian framework, by looking at Rawls’ remarks concerning the value of political society within his domestic theory of justice itself. For example, in PL Rawls discusses the relationship between justice as fairness and ideas of the good (PL: lecture V). According to Rawls, liberal citizens experience political society as a great good for reasons that are both individual and collective, and range from the very possibility of exercising their fundamental moral powers (ibid.: 203), to the good of justice and the social bases of self-respect (ibid.), and the good of social cooperation (ibid.: 204). ¹⁵ It would thus not be unreasonable to therefore conjecture that were we to relax the assumption that a liberal society is closed and self-contained, then, liberal citizens could be characterized as having an interest in the self-determination of their political community.

The interpretive soundness of this conjecture, however, will not settle the normative problem concerning collective self-determination since a conception of citizens’ interests is a normative idea and thus requires a normative justification. Reasons of space preclude a full discussion of this issue (but see Beitz, 2009b; Jones, 2009). Instead, I want to end by highlighting the following point: there is no reason, in principle, to be concerned that a liberal individualist conception of citizenship cannot provide for ample space to the value of collective self-determination. The usual reason for believing that self-determination should be taken cautiously from a liberal perspective is that liberals see (moral) group rights as problematic given their commitment to the idea that individuals, not groups, are the basic unit of moral concern. But conceptions of moral group rights can be thoroughly based on accepting the premise that individuals are the basic unit of moral concern. There is no need to believe that a group is entitled to be self-determining in virtue of some property of the group itself other than the contribution that the group makes, for example, to the well-being of its members (Raz, 1986; Raz and Margalit, 1990). ¹⁶ Collective (rather than corporate) conceptions of group rights claim precisely that: moral group rights such as self-determination can be understood as group rights and yet be grounded solely on the contribution that such rights make to the lives of the individuals making up the group in question (see Altman and Heath Wellman, 2011; Jones, forthcoming, 2009, 1998). ¹⁷ So, in principle, there is no inconsistency between adopting a liberal individualist standpoint and claiming that individuals forming certain groups are entitled to be collectively self-determining. In other words, to see self-determination as a moral group right, there is no need to understand groups as natural persons, or to ascribe moral standing to groups as such. The latter point would also allow us to clarify that liberal peoples’ interest in their self-determination should simply be understood as a collective moral group right grounded and ultimately justified by reference to the interests of liberal citizens. As I have mentioned above, this still leaves open the task of actually justifying the presence and relative weight of such interest for liberal citizens. ¹⁸ Yet accepting the collective conception of group rights allows us to see that the congruence between liberal peoples’ interests and the interests of liberal citizens requires no fundamental departure from the liberal axiom of moral individualism.

Conclusion

In this article I have defended the Rawlsian account of international toleration against two important objections. Ultimately, the arguments I have sketched give us a more complex picture of Rawlsian liberalism: one in which liberal ideas and principles travel a more complex journey across domains of justice and where the self-determination of liberal peoples (indeed, of liberal citizens) is more important than we would otherwise have predicted. These are, from a liberal cosmopolitan standpoint, often seen as weaknesses in the Rawlsian view. I disagree, but this is not the main message that I want to conclude the article with. The message is instead that the criticisms to Rawls’ view I have discussed seem to presuppose that they are and thus claim that Rawls’ account is inconsistent. I think I have shown that this is not a sustainable way to proceed.