CONTEXTUALISM AND GLOBAL JUSTICE 1. Introduction There are two sets of questions that have featured prominently in recent debates about distributive justice. One of these debates is that between universalism and contextualism. While universalists maintain that principles of justice apply universally and across the board, contextualists emphasize the crucial role context plays in settling questions of justice. The second debate concerns questions of international justice and here the main rift is between cosmopolitans, advocating strong principles of international justice, and more skeptically-minded thinkers, arguing that international justice is a rather problematic notion. Thinking about how these debates are linked, one might suspect that the universalist, when it comes to questions of international justice turns out to be a cosmopolitan, while the contextualist naturally takes a more sceptical stance. For the universalist end of the debate this might be true, however, contrary to her own claims, the contextualist has trouble arguing against strong principles of international justice. Taking the position of David Miller, who is a pronounced contextualist and fervent opponent of cosmopolitanism, as a background, I shall subsequently argue that some of the most powerful arguments against cosmopolitanism are unavailable, if one is to maintain an interesting notion of contextualism. To develop my thesis about the tension between contextualism and certain arguments against cosmopolitanism, I briefly elaborate Miller s notion of contextualism, secondly I explain two of his central arguments against cosmopolitanism, and finally I explicate why a tension arises between these arguments and contextualism. Contextualism neither does it provide good reasons to be skeptical about global justice, nor does it offer appropriate tools to undermine the normative force of cosmopolitanism. 2. David Miller s Contextualism There are two ways to understand David Miller as advancing a contextualist position, or put differently, Miller offers two reasons why one should be concerned about context when thinking about justice. On the one hand, certain contextual conditions have to be met for questions of justice to arise in the first place. On the other hand, context is crucial for determining
66 what kind of distribution qualifies as just and which one does not. I consider each of these points in turn. The Circumstances of Justice A first central claim of Miller s theory of justice asserts that for any principle of distributive justice to be applicable at all there are circumstances of justice that have to be met (Miller, 1999 1 ff). If these circumstances or conditions are not satisfied, then, according to Miller, it does not even make sense to raise questions about justice. To meaningfully address questions of social and distributive justice there has to be (i) the background of a society as a sphere of shared understandings and values against which the question can be asked, (ii) an institutional framework that impacts on the individual shares of burdens and benefits, and (iii) an agency that is capable of bringing about the change that is required by principles of justice. In responses to critics of his position, Miller explicitly states that causal interaction between two agents does not suffice for making principles of justice an applicable means in determining the validity of, say, different claims to resources: My position here relies on a distinction between people being associated with one another, whether by participating in a common practice, being subject to the same set of institutions, belonging to the same community, [ ], and people merely having a causal impact on one another s situation (Miller, 2003 368). The former situation satisfies the circumstances of justice while the latter does not. Principles of justice only apply in contexts that are very strictly defined. Context as the Normative Point of Reference In Two ways to think about justice (Miller 2002), David Miller makes another case for a contextualist approach to justice. For a contextualist, different principles of justice (P1, P2,, Pn ) apply in different contexts (C1, C2,, Cn ) without being derived from a single master-principle. It is the relevant context Cn, in Miller s case a form of social relationship, that determines which principle Pn is applicable. No reference to an overarching principle is necessary. To settle questions of justice, instead, the interpretation of the context indicates which principle of justice applies. Although the resulting distributive principles are freestanding and independent from one another, contextualism, according to Miller, is not to be confused with conventionalism. The variation of principles across contexts has a logic of its
Contextualism and Global Justice 67 own that can be used in making normative assertions. The contextualist can for example argue that in context C1 the principle P2 does not apply, because the features of C1 require the different principle P1. 3. Two Contextualist Arguments Against Cosmopolitanism In refuting ambitious principles of international justice, such as for example global egalitarianism, Miller marshals a number of different argumentative strategies. Central among these are two arguments, each building on Miller s notion of contextualism. Pluralism, National Self-Determination and Global Justice The concept of the nation is a cornerstone of David Miller s political theory. According to Miller, national self-determination is valuable in the same way that individual self-determination is valuable (Miller, 2000). But moral requirements, for example requirements springing from obligations of global justice understood as distributive justice on a global scale, could come into conflict with national self-determination. This is because nation states are themselves engaged in the pursuit of an important form of justice, namely social justice (Miller, 2000 167). If the requirements of national social justice were identical with the requirements of global justice, the problem of conflict between self-determination and justice would of course not arise, but according to Miller, it is highly implausible to assume that the distributive requirements of national social justice and those of global justice do indeed coincide. In fact, because various national conceptions of social justice differ markedly from each other (Miller, 2000 168), they cannot all be identical with the distributive requirements of global justice. If Miller is right about this diversity of nationally or culturally dependent conceptions of justice, any attempt to provide comparative and comprehensive principles of global distributive justice, let alone egalitarian principles, is facing a tremendous challenge. Giving reasons for these differences and thus a pluralism of conceptions of justice, Miller offers a three-fold account, partly relying on forms of contextualism already familiar from above: (i) There are two reasons why the understanding of the subject matter of justice varies from society to society. Firstly, the meaning of social goods and thus of the distributive principle going along with them can only be understood within one society. More fundamentally, different societies have different understandings of what is to count
68 (ii) (iii) as a burden and what is to count as a benefit. On the international level, a pluralism of conceptions of justice is the result. Even if different social or cultural communities agreed on what is a relevant criterion of just distribution, such as for example distribution according to need or desert, the conception of justice may still be effectively different. This is because the notions of need and desert are themselves dependent on shared beliefs and understandings. The question of which factors contribute to establishing desert claims for example, may receive different answers in different societies. According to Miller s contextualism, we make judgements about which distributive criterion is a just criterion by examining the context, that is, the social relationship in which it is to be applied. Because social relationships and contexts vary from one society to another, so do conceptions of justice. Miller s pluralism with respect to conceptions of justice is thus composed of three elements that allow for variation and context-dependence. This pluralism constitutes the challenge that every formulation of positive principles of international distributive justice, aspiring to create global obligations, has to come to terms with. It thus renders cosmopolitan positions unavailable. The circumstances of justice - Objection to Cosmopolitanism There is a second way in which Miller attempts to use his notion of contextualism to rebut global egalitarianism and related positions. Here Miller s argument against comparative principles of international justice takes a form that mirrors his argument from circumstances of justice as developed above. In Principles of Social Justice (Miller, 1999 1 ff), Miller develops the concept of circumstances of justice. These have to be met for questions of justice to arise at all. Only a group in which the circumstances of justice apply can be a candidate for the implementation of comparative principles of justice (Miller, 2003 368). Within an association of people that lacks an adequate institutional framework for example, principles of social justice have no place. In the context of international distributive justice, the concept of the circumstances of justice plays an important role. The specifically international version of the argument can be reconstructed along the lines of [Argument 1]:
Contextualism and Global Justice 69 (P1) (P2) Comparative Principles of justice can only apply if the domain in which they are to apply meets certain conditions or circumstances, such as being part of an adequate institutional framework and having a background of common world-views. In the international domain, because of pluralism in beliefs and valuesystems, as well as the absence of an institutional framework, these circumstances are not met. (C) Thus, in the international domain, comparative principles of distributive justice are inapplicable. Again, this conclusion puts considerable pressure on anyone wishing to maintain a cosmopolitan position. 4. Two Problems for Contextualist Anti-Cosmopolitanism Miller s arguments as presented so far possess initial plausibility and avoid many of the pitfalls of alternative positions. However, there are two critical remarks to be made. These will support my central thesis, namely, that the contextualist arguments against cosmopolitanism are far from convincing. The Dilemma of the Argument from Pluralism and Contextualism In Miller s reasoning about justice, the notions of contextualism and different forms of pluralism play an important role. A pluralism of social relationships for example is the rationale behind adopting a contextualist approach to justice, which in turn gives rise to a pluralism of principles of justice. However, when establishing an argument about international justice, slightly different accounts of pluralism and contextualism are employed. As will be argued below, these accounts undermine the first version of contextualism. This dilemma leaves Miller s position unstable. In Two Ways to Think About Justice, Miller explicitly demarcates his form of contextualism from conventionalism. It is the quintessence of his contextualism that different principles of justice apply in different contexts, while it is the form of social-relationship that determines the nature of the context: An alternative view, one that I have defended myself elsewhere, defines contexts by reference to the kind of social relationship that exists among the parties between whom justice is to be done. (Miller, 2002 10/11). Because certain contexts require certain principles, it is possible to judge the application of certain principles as unjust in certain contexts: Contextualism assumes that the relevant social contexts are
70 characteristically reiterated across time and space, and that whenever a particular context appears, we can determine what justice requires in that context. So the claims that contextualists make about the contextual validity of principles of justice are themselves objective and universal in character. In other words, they take the form: in context C it is always right to distribute the relevant goods and bads according to principle P. So if we find a society that contains context C, but distributes goods in that context in some other way, then the contextualist will rule that society unjust (Miller, 2002 12). The problem of conventionalism, allegedly finding itself unable to make such judgements and having to accept as just whatever is taken as just by those concerned, can thus be avoided. Taking Miller s understanding of contexts as being characterized by the form of social-relationship, and considering the objective relation he identifies between context and principle of justice, it follows that the form of social relationship can serve as a basis for justice-claims and social criticism across space and time. A further implication is that what justice requires in S1 is not necessarily what the inhabitants of S1 believe justice to require in their society. Contextualism leaves room for the claim that people s beliefs about justice may in certain respects be mistaken, for instance, if they hold that P2, rather than P1, is the appropriate principle to apply in C1 (Miller, 2002 8). In National Self-Determination and Global Justice Miller refers to contextualism and pluralism in an attempt to support his claim that conceptions of justice can be expected to differ between nations. In the wider context of his argument, this claim plays an essential role in establishing that there cannot be uniform principles of global justice. As seen in the previous section, there are now three elements that allow for a pluralism in conceptions of justice and also contribute to constituting different contexts: To conclude therefore, we have found three sources of likely divergence in these conceptions different understandings of the goods whose distribution is at stake; different understandings of the principles that should govern distribution; and different understandings of the social contexts in which these principles are to be applied (Miller, 2000 171). This explanation for why different principles of justice apply in different contexts is distinct to the one offered earlier and presents Miller with at least one severe problem.
Contextualism and Global Justice 71 While the international account of pluralism and contextualism is necessary to argue effectively against comparative principles of global justice, it undermines the distinction made between contextualism and conventionalism in the first place. Contrary to what Miller claims, the account of pluralism and contextualism presented in National Self-Determination, does not maintain this distinction. In the general account of contextualism, the just principle was intended to spring from the context that existed independently from people s beliefs. In the second account, the context is itself constituted by people s beliefs, shared meaning and agreements. This point is key to Miller s argument against more substantial principles of global justice, at the same time it is responsible for collapsing contextualism into conventionalism. Miller is committed to the premises of the following [Argument 2], but, although the argument is clearly valid, he seems unwilling to admit the implications of the conclusion: (P1) (P2) (P3) (P4) (C) The meaning of social goods is socially constructed and embedded in people s shared beliefs and understandings (Miller, 2000 168). The understanding of the criteria and principles of justice varies between cultures/societies and depends on the overall world-view of those living in that society or culture (Miller, 2000 169). Social relationships (i.e. contexts in the original sense) take different forms as world-views, understandings and beliefs vary (Miller, 2000 169). The meaning of social goods, the understanding of the respective criterion of justice and the form of social relationship determine what is just (Miller, 2000 169). What distribution of goods is to count as just is determined by the shared beliefs, understandings, and world-views of those concerned and markedly differs between cultures and societies. There can be no doubt that (C) describes an essentially conventionalist position not bearing much resemblance to the form of contextualism Miller proposed in Two ways to think about justice. Although Miller is keen on denying any conventionalist, or using his own terminology in that passage, subjectivist (Miller, 2000 171) interpretation of his position, this denial cannot be substantiated. To the contrary, the two accounts of contextualism that Miller has to offer are clearly incompatible and the international account effectively amounts to a conventionalist position. Consider the following example: In a society S1 families are an important form of social organisation, within which people spend much time. Within families, there are burdens and benefits, material resources as well as
72 time and the devotion of attention, to be distributed. Religion plays an important role in society S1, and within the religious world-view predominant in S1, the notion of merit is central. It is a religious requirement to be meritorious; according to the religious conception, the meritorious pleases God. Because religious beliefs are very powerful, they also shape concepts in secular domains and thus the central criterion in allocating burdens and benefits within families is that of merit. It is a further feature of the shared understandings and beliefs of S1, that economic success, measured in terms of wealth and income, is central to the notion of merit. Within families, burdens and benefits are consequently distributed according to merit, that is economic success measured in terms of the respective contribution to the family budget. The economically most successful is the chief recipient of resources and attention, while the least successful has to do the dishes. Now two political theorists, not themselves members of society S1 are invited to judge whether the distributive principle applied within families in S1 is just or unjust. The two theorists hold the same values, beliefs and belong to the same community. The only difference is that theorist (T1) is a contextualist in the sense of Two ways, whereas theorist (T2) supports the form of contextualism and pluralism elaborated in National Self- Determination. In compliance with his version of contextualism, (T1) can be expected to answer along these lines: The distribution of burdens and benefits within families in S1 is unjust because people misunderstand the meaning of the social relationship of the family. They also fail to recognize that the family context requires the distributive criterion of need. Additionally, they misinterpret the distributive criterion of merit. The context of the family requires principles of justice organized around the criterion of need, and thus the situation described is unjust. Theorist (T2) on the other hand can be expected to offer a very different answer: Because social-relationships take different forms in different societies and different cultural milieux, there is no ground on which to object to an understanding of the family that requires merit as a criterion of just distribution. Because the understanding of the criterion of merit employed is consistently embedded in the wider world-view of S1, it is perfectly reasonable for the people in S1 to found distributive principles on it. The distribution of benefits and burdens according to merit coheres with the shared understandings and beliefs of those concerned and is thus a just one. There are then two points highlighted and supported by this example. Firstly, Miller s international account of pluralism and conventionalism effectively amounts to a form of conventionalism or relativism. And secondly, the two accounts of contextualism are mutually incompatible. Miller s middle-
Contextualism and Global Justice 73 ground position, on the one hand trying to maintain a form of contextualism that is distinct from conventionalism, and on the other hand rejecting global comparative principles of justice by referring to pluralism and contextualism is thus unstable. Addressing the Challenge of Global Egalitarianism? Prima facie, the argument from the circumstances of justice seems to successfully assail the position of the global egalitarian. [Argument1] in itself is perfectly consistent. It is however less clear whether or not the argument does indeed support the point Miller wants to make. Admittedly, the argument shows that egalitarian principles of justice cannot be applied globally, yet it does not show that egalitarian principles of justice are not required. One can comprehend this subtle point by taking a closer look at the word inapplicable in the conclusion of [Argument 1]. On the one hand, inapplicable can stand for something like impossible to be implemented or unfeasible ; on the other hand it can stand for not morally required by the form of interaction under consideration. Depending on which of these readings one applies, the notion of the circumstances of justice is interpreted differently. Because there really are two different issues at stake, a distinction between these two forms or readings is necessary. Considering both, the weaker and the stronger reading of inapplicable, the following two arguments illustrate why [Argument 1] fails to address the challenge of global egalitarianism. On the reading of inapplicable as unfeasible the argument from the circumstances of justice does not succeed in showing that strong comparative distributive principles are not a requirement of justice. Justice may require a certain distributive principle although the domain in which it would apply does not allow for its implementation. Miller s [Argument 1] may then succeed in showing that comparative principles of justice cannot straightforwardly be implemented, but it does not show that such principles are not required. There is a real distinction between the impossibility to formulate and enforce principles of justice, and such principles not being a requirement of justice. Even an appeal to ought implies can would not save Miller here. The way his argument is set up does not suggest that it is necessarily impossible to implement global principles of justice, but rather that given the present situation, principles cannot be implemented. If justice requires a specific distributive principle, but in order for this principle to be implemented effectively a change of institutional settings would be necessary, then the change of institutional settings itself becomes a requirement of justice.
74 Consider this example: Person A and Person B live as hermits on opposite banks of a river. They do not communicate with each other, because there is no bridge that could be used to cross the river. Person A is a farmer growing beans on her bank and uses the river to water her fields. Person B lives off fish from the river. At some point, person A starts using fertilizers to increase the bean harvest. Unfortunately, using the fertilizer on the banks has the negative side-effect of polluting the river and killing all fish so that person B cannot make a living anymore. How should one react to this scenario? Clearly, some comparative distributive principle to allocate clean water is required here. But obviously, Miller s circumstances of justice are not met. Neither do A and B form a community, nor do they participate in a common institutional practice. Regardless of the principle of justice one would opt for in this situation, whether egalitarian or other, it could not be implemented in this very situation because there is not even a bridge. At the same time, this impossibility does not have any impact on the distributive principle required and there are no fundamental obstacles that make implementation impossible in the future, for example after having built a bridge. Building a bridge may in fact be the first requirement of distributive justice. On the stronger reading of inapplicable, the conclusion of [Argument 1] denies that the domain of the international is at all amenable to assessment by principles of distributive justice. It would then not only be the case that principles of justice cannot be implemented, but that these principles simply have no application. Consequently, a given distribution of goods on the international level could not be criticized as unjust and the demand to move to a situation, in which principles of distributive justice do apply, could not itself be a demand of justice. If we understand Miller s [Argument 1] in this way, cases like the one of A and B on different banks of the river appear in a new light: It is not that there are practical obstacles to the implementation of principles of justice, but that a question of justice simply does not arise. Does this argument against global comparative principles succeed? Consider different answers this question might receive. One can respond to the argument that situations like the one of A and B cannot be assessed by standards of justice because the circumstances of justice are not satisfied in either of two ways. The first response grants the success of Miller s arguments and rejects any attempt to formulate principles of justice for this situation. The second response recognizes the case of justice and discards the idea of circumstances of justice. Our intuition behind the example involving A and B seems to support this second response: There is something wrong with the idea that the principles of justice only apply under certain circumstances. I shall not attempt to produce a conclusive argument against the idea of circumstances of justice; however, there are a number of examples to establish
Contextualism and Global Justice 75 at least a presumption in favour of the second response. Consider an analogous case: On Rawls account, circumstances of justice comprise the condition of moderate scarcity (Rawls, 1999 110). But why should questions of justice not arise in cases of severe scarcity? If Scott and three friends are members of the same expedition to the South Pole, while it turns out that their food-supplies are such that two men survive if they exclusively share the food between them, while everyone starves if everyone only receives a quarter of the supplies, Rawls circumstances of justice are clearly not satisfied. But still, some ways to deal with this situation would be more just than others. Would we think of Scott, who happens to own the only rifle, as a just person if he makes a random decision to shoot two of his companions? Or would we consider Scott as a just person if he decides to run a lottery in which everyone has an equal chance for survival? These considerations raise substantial worries about the significance of circumstances of justice, and consequently about any argument building on them. Because of these worries, Miller s argument falls short of conclusiveness even on the strict reading of inapplicable. It seems that in either case [Argument 1] fails to address the challenge of cosmopolitanism. 5. Conclusion As we have seen, contextualism and scepticism about global justice do not necessarily coincide. The most powerful contextualist arguments against cosmopolitanism, such as the argument from pluralism and the argument from circumstances of justice, do not succeed. The first argument fails because it is inconsistent with the central tenets of contextualism. The second argument fails because it leaves the normative force of cosmopolitanism intact. Whether, in view of these results we ought to revise our understanding of contextualism, abandon it altogether, or in fact embrace cosmopolitanism, is a question that cannot be answered here. To recognize the need to address it, however, is a step towards understanding the requirements of global justice. University of Oxford
76 REFERENCES Miller, D. (1987) Exploitation in the Market, in Modern Theories of Exploitation, Reeve, A. Ed., Sage, London. Miller, D. (1995) On Nationality, Clarendon Press, Oxford. Miller, D. (1995b) Complex Equality, in Pluralism, Justice and Equality, Miller, D. and Walzer, M. Eds., Oxford University Press, Oxford. Miller, D. (1999) Principles of Social Justice, Harvard University Press, Cambridge. Miller, D. (1999b) Justice and Global Inequality in Inequality, Globalization, And World Politics, Hurrell, A. and Woods, N. Eds., Oxford University Press, Oxford. Miller, D. (2000) National Self-Determination and Global Justice in Citizenship and National Identity, Polity Press, Cambridge. Miller, D. (2001) Distributing Responsibilities, in The Journal of Political Philosophy, 9, 4, pp. 453-471. Miller, D. (2002) Two ways to think about justice, in Politics, Philosophy and Economics 1, 1, pp. 5-28. Miller, D. (2002b) Cosmopolitanism: A Critique, in Critical Review of International Social and Political Philosophy 5, pp. 80 85. Miller, D. (2003) Response, in Forms of Justice, Bell, D. and de-shalit, A. Eds., Rowman & Littlefield, Oxford. Miller, D. (2004) Holding Nations Responsible, in Ethics 114, pp. 240 268. Rawls, J. (1999) A Theory of Justice, Oxford University Press, Oxford.