Cosmopolitan Regard: Political Membership and Global Justice (Contemporary Political Theory)

Cosmopolitan Regard Cosmopolitan theory suggests that we should shift our moral attention from the local to the global...

Author: Richard Vernon

63 downloads 1280 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Cosmopolitan Regard Cosmopolitan theory suggests that we should shift our moral attention from the local to the global. Richard Vernon argues, however, that if we adopt cosmopolitan beliefs about justice we must re-examine our beliefs about political obligation. Far from undermining the demands of citizenship, cosmopolitanism implies more demanding political obligations than theories of the state have traditionally recognized. Using examples including humanitarian intervention, international criminal law, and international political economy, Vernon suggests we have a responsibility not to enhance risks facing other societies and to assist them when their own risk-taking has failed. The central arguments in Cosmopolitan Regard are that what we owe to other societies rests on the same basis as what we owe to our own, and that a theory of cosmopolitanism must connect the responsibilities of citizens beyond their own borders with their obligations to one another.

richard vernon is Distinguished University Professor in the Department of Political Science at the University of Western Ontario.

C ON T E M P OR A RY P OL I T I C A L T H E ORY

SERIES EDITOR

Ian Shapiro EDITORIAL BOARD Russell Hardin

Stephen Holmes

Jeffrey Isaac

John Keane

Elizabeth Kiss

Susan Okin†

Phillipe Van Parijs

Philip Pettit

As the twenty-first century begins, major new political challenges have arisen at the same time as some of the most enduring dilemmas of political association remain unresolved. The collapse of communism and the end of the Cold War reflect a victory for democratic and liberal values, yet in many of the Western countries that nurtured those values there are severe problems of urban decay, class and racial conflict, and failing political legitimacy. Enduring global injustice and inequality seem compounded by environmental problems, disease, the oppression of women, racial, ethnic and religious minorities, and the relentless growth of the world’s population. In such circumstances, the need for creative thinking about the fundamentals of human political association is manifest. This new series in contemporary political theory is needed to foster such systematic normative reflection. The series proceeds in the belief that the time is ripe for a reassertion of the importance of problem-driven political theory. It is concerned, that is, with works that are motivated by the impulse to understand, think critically about, and address the problems in the world, rather than issues that are thrown up primarily in academic debate. Books in the series may be interdisciplinary in character, ranging over issues conventionally dealt with in philosophy, law, history, and the human sciences. The range of materials and the methods of proceeding should be dictated by the problem at hand, not the conventional debates or disciplinary divisions of academia.

Other books in the series Ian Shapiro and Casiano Hacker-Cordón (eds.) Democracy’s Value

Ian Shapiro and Casiano Hacker-Cordón (eds.) Democracy’s Edges Brooke A. Ackerly Political Theory and Feminist Social Criticism Clarissa Rile Hayward De-Facing Power John Kane The Politics of Moral Capital Ayelet Shachar Multicultural Jurisdictions John Keane Global Civil Society? Rogers M. Smith Stories of Peoplehood Gerry Mackie Democracy Defended John Keane Violence and Democracy Kok-Chor Tan Justice without Borders Peter J. Steinberger The Idea of the State Michael Taylor Rationality and the Ideology of Disconnection Sarah Song Justice, Gender, and the Politics of Multiculturalism Georgia Warnke After Identity Courtney Jung The Moral Force of Indigenous Politics Sonu Bedi Rejecting Rights

Cosmopolitan Regard Political Membership and Global Justice Richard Vernon

cambridge university press Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo, Delhi, Dubai, Tokyo Cambridge University Press The Edinburgh Building, Cambridge CB2 8RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521744379 © Richard Vernon 2010 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2010 Printed in the United Kingdom at the University Press, Cambridge A catalogue record for this publication is available from the British Library Library of Congress Cataloguing in Publication data Vernon, Richard, 1945– Cosmopolitan regard : political membership and global justice / Richard Vernon. p. cm. – (Contemporary political theory) ISBN 978-0-521-76187-1 (hardback) – ISBN 978-0-521-74437-9 (pbk.) 1. Cosmopolitanism. 2. Political obligation. I. Title. II. Series. JZ1308.V47 2010 306.2–dc22 2009053754 ISBN 978-0-521-76187-1 Hardback ISBN 978-0-521-74437-9 Paperback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party Internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Contents

Acknowledgments Introduction 1 Against associative obligations

page viii 1 11

2 Particularizing obligation: the normative role of risk

39

3 The social waiver

65

4 Compatriot preference and the Iteration Proviso

92

5 Humanitarian intervention and the case for natural duty

117

6 Associative risk and international crime

143

7 A global harm principle?

167

Conclusion: citizens in the world

193

Bibliography

210

Index

220

Acknowledgments

This book has two origins. One is in a course on “Global Justice” that I introduced and taught in the Department of Political Science, University of Western Ontario, and I am grateful to many students for the critical discussions that have led me to develop my approach, as well as to graduate students who have worked with me on related thesis projects. The second origin lies in the remarkable work of other scholars who, in barely three decades, have turned a persistently neglected area of inquiry into such a flourishing field. Many of them are mentioned in my footnotes, and I hope it goes without saying that my appreciation of their work is no less when I happen to disagree with their conclusions. I will single out my colleague and friend Charles Jones, whose clarity of thought and expression I cannot hope to match, and whose encouragement I have greatly valued. Earlier drafts of chapters in this book have benefited from criticism by Gillian Brock, Joseph Carens, Charles Jones, Steven Lecce, David Luban, Daniel Weinstock, and Christopher Wellman, as well as by members of the Moral, Political and Legal Philosophy Research Group at Western Ontario. John Horton, one of Cambridge University Press’s readers, offered not only welcome and generous encouragement but also many acute criticisms and helpful remarks, some (but not all) of which are acknowledged in footnotes below. If I were half as good a scholar as any of those named above, this would be a better book. Research for this book was supported by a Standard Research Grant from the Social Science and Humanities Research Council of Canada, 2006–9. Rhiana Chinapen provided able assistance in preparing the manuscript.

Acknowledgments ix

I have drawn upon previously published material. Chapter 1 is a revised version of “Obligation by Association?” Political Studies 55 (2007): 865–78; chapters 2 and 3 draw material selectively from two articles, “Compatriot Preference: Is There a Case?” Politics and Ethics Review 2 (2006): 1–18, and “States of Risk,” Ethics and International Affairs 21 (2007): 451–69; chapter 4 is a revised version of “Contractualism and Global Justice: The Iteration Proviso,” Canadian Journal of Law and Jurisprudence 19 (2006): 345–56; chapter 5 is a revised version of “Humanitarian Intervention and the Internal Legitimacy Problem,” Journal of Global Ethics 4 (2008), 37–49; chapter 6 borrows in part from “What is Crime Against Humanity?” Journal of Political Philosophy 10 (2002), 231–49; chapter 7 is a revised version of “A Global Harm Principle?” Critical Review of Social and Political Philosophy 12 (2009): 1–18. I am grateful to the editors and publishers of these journals for permission to draw from these articles here.

Introduction

The most remarkable development in political theory, over the past two or three decades, is its new orientation to issues of global justice. When Charles Beitz published Political Theory and International Relations in 1979, it stood virtually alone as a normative account of our duties beyond our political borders. By now, however, the literature is immense: dozens of books and hundreds of scholarly articles address issues of global distributive justice and related questions of what we owe to those who are not our co-citizens. In this literature the idea of “cosmopolitanism” – variously interpreted – emerges as a political concept of central importance. That term now denotes an idea of moral and political obligation that gives weight to the interests of all human beings, in ways that are taken to impose significant constraints on the pursuit of our own national self-preference. According to the editors of a particularly illuminating collection, “everyone has to be at least a weak cosmopolitan now if they are to maintain a defensible view, that is to say, it is hard to see how one can reject a view that all societies have some global responsibilities.”1 Likewise, from the nationalist side, it is agreed that cosmopolitanism’s “weak ethical version – formulated in terms of a principle of equal moral worth or equal moral concern – can be accepted by almost anybody barring a few racists and other bigots.”2 Disagreements are not, of course, any less significant for that reason, for the possibility of a common formula scarcely diminishes practical controversy. Stronger cosmopolitans believe that more follows from that weak premise than nationalists allow; they believe that it leads 1

2

Gillian Brock and Harry Brighouse (eds.), The Political Philosophy of Cosmopolitanism (Cambridge: Cambridge University Press, 2005), 3. David Miller, “Cosmopolitanism: A Critique,” Critical Review of International Social and Political Philosophy 5 (2002): 84.

2 Cosmopolitan Regard

directly to global responsibilities, while their critics maintain that, in the allocation of responsibilities, national or else civic obligations take up much or even most of the ground before we turn to what we owe to outsiders. Nevertheless, both cosmopolitans and their critics now occupy what we may term a “weak cosmopolitan plateau,” upon which struggles are waged not over whether we owe outsiders anything, but how much, and what it is. In this book I adopt the term “cosmopolitan regard” for the belief or assumption that what happens to everyone is of moral importance – that only “racists and other bigots” will defend the contrary view, that some lives are of less importance than “ours” (or certain people’s). I distinguish between this regard and anything that is determinate enough to be called a political responsibility, and in that respect I agree with critics of strong cosmopolitanism who believe that several intermediate steps must be taken before that basic moral belief issues in an obligation that falls to people in their capacity as citizens. On the other hand, I do not agree that global obligations are, as it were, residual, or that national or civic associations take up moral space independently of what can be said to be owed to outsiders. My view is that, given cosmopolitan regard, they take up their moral space only conditionally, and that their own moral sustainability implies strong if limited obligations to those who are outside it. As a citizen, one’s obligations to co-citizens and to outsiders rest on the same footing. In making this case, obviously the argument takes issue not only with strong cosmopolitans and cosmopolitan-skeptics of the nationalist kind, but also with theorists of “moral dualism” who maintain that cosmopolitan and local obligations are of different and incommensurable types, irreducible to one another. I shall shortly give a synopsis of the book’s argument and the sequence of steps that it takes, but before doing so let me explain its motivation. The significance of its attempt arises from what I believe to be an under-explored paradox. “Cosmopolitan” is of course an ancient term, coined and deployed by Cynics and Stoics in the late classical period in order to

Introduction 3

contrast local political membership with membership in the world – strictly speaking, the cosmos itself – as a whole. To describe the world as a polis, or “city,” was to transfer to it the then-standard term for the association of greatest importance to human life. There were, of course, different views about the implication of this for actual cities, ranging from Diogenes the Cynic’s rude dismissal of them to later Romanized Stoic views, such as Cicero’s, that effectively reasserted the value of the actual city and rendered the world-city somewhat marginal. Despite these differences, the cosmopolitan idea represented a normative re-weighting of life, conveyed by a term that executed a powerful metaphorical transfer; a transfer that undermines the moral autarchy of political associations, and broadens the context of moral justification. Concealed in this transfer, however, were some internally contradictory elements. For the world, after all, is not really a city, or even much like one. The “city” that served as the ground of the metaphor was a community of people connected to each other by strong ties of proximity, acquaintance, and defi nite legal relations – not to mention a sense of exclusiveness. To imagine the world as a city was obviously to abstract from all this for the sake of disclosing a bare, uninstitutionalized moral relation among humans; it is to say that they matter too. It is to foreshadow the plateau referred to above. And this in turn, as one scholar of Zeno’s Stoicism has pointed out, can rebound on the conception of the actual, political city. For if we adopt an idea of association from which all the contingency of proximity and acquaintance is removed, we have taken a large step towards ideas such as natural law, which require the just treatment of other humans simply as humans – an idea that “does not require for its intelligibility or acceptability any reference to citizenship at all.”3 In exploiting the normative power of the idea of the “city,” in 3

Malcolm Schofield, The Stoic Idea of the City (Cambridge: Cambridge University Press, 1991), 103. It should be noted that it is not uncommon to use the term “city” to refer to something like a horizon of awareness: as globalization proceeded, “there were fewer and fewer people whose city was their world, and more


other words, we may empty it of distinctiveness and just make it a part – so to speak – of the “world.” And if so, why is it so important that the world should be thought to resemble a “city”? If the city has distinctive features then it is potentially most significant that the world as a whole is said to resemble it, but if the city is just an instance of generic association the interest of the metaphorical transfer evaporates. Distanced though we now certainly are from the cosmopolitanism of Zeno, I believe there is a lesson in what Malcolm Schofield calls its “unstable” character, that is, its tendency to dissolve its own metaphorical ground. To say that we are citizens of the world is to place in question what our actual citizenship is, or what it means. It is to imply that what makes it different from bare human association cannot be normatively important – for if it were, then how could it possibly serve as a model for human association itself? The “instability” here can be resolved either by deciding that cosmopolitanism is really “only” a metaphorical term – which, of course, virtually deprives it of moral force – or else by accepting the implication that the features of citizenship are mere contingencies of no moral importance. The first solution simply leaves the moral autarchy of polities intact: it leads to a sort of gestural cosmopolitanism that has little hope of success against the established, solid and pre-emptive demands of nationality and citizenship – or else, one critic complains, to “a simple negation.”4 What (if anything) is wrong, though, with the second? One thing that is wrong with it is that, if we stage a political contest between a contingency-free view of association in which local identities count for nothing, and a view that finds a serious place for local identity, the latter will certainly win. That is, an

4

and more people for whom the world had become their city.” (Philippe Van Parijs, “International Distributive Justice,” in Companion to Contemporary Political Theory, ed. Robert Goodin et al. [Oxford: Blackwell, 2007], 638.) Without objecting to this usage, this book concerns citizenship as a political category. R. B. J. Walker, “Citizenship After the Modern Subject,” in Cosmopolitan Citizenship, ed. Kimberly Hutchings and Roland Dannreuther (Houndmills: Macmillan, 1999), 177.

Introduction 5

objector might say, only a political consideration: but while political theory cannot make itself hostage to political considerations, nor can it simply imagine away the most basic constraints in political life. Among these is the political division of the world, the consequences of which we can reasonably hope to modify in important ways, but the fact of which must be accepted unless attractive and viable alternatives become available. And while exclusive political divisions remain, it is a good thing that people should attach importance to their membership in them, because that can make them more attractive sites for living in. The second – and for the present purpose, more important – thing that is wrong is that, in the world as it is, the prospects for global justice can be achieved only if (actual) citizenship is valued even more than it currently is. For consider the main components of practical (as distinct from academic) cosmopolitanism. More or less during the years in which scholars have come to embrace cosmopolitanism, weak or strong, three endeavors (to be discussed in the last three chapters of this book) have emerged in the field of global politics. One is a movement, still very much in progress, to re-evaluate the idea of sovereignty so that it admits intervention when states commit atrocious acts. Another is an effort to impose constraints of international criminal law upon state leaders, and their followers, who commit atrocity. A third is the increasingly inescapable view that consumers in wealthy countries will have to accept economic changes in order to avoid exploitative relations with poor countries that are both morally wretched and politically inflammatory. Of course, none of those movements or tendencies are anywhere near complete, and they still face opposition; but they have won the support of many governments of the world, of influential segments of public opinion, and of important international organizations. Whatever one’s view of their prospects, those three endeavors comprise what cosmopolitanism, practically, is, and, I believe, it is by considering our obligations in those three contexts that we can best work out the implications of an abstract cosmopolitan premise.


But if we consider them, they all entail additional demands by states upon their citizens. Their prospects are nil unless states can provide compelling grounds for the sacrifices that their citizens will have to bear if these projects are to be achieved. Humanitarian interventions are very costly, in financial terms and, more importantly, in terms of lives inevitably lost. International criminal law, in the absence of an international police force, is effectively hostage to states’ willingness to bear the considerable political and financial costs of arrest, trial, and punishment. Changes in the international trading regime are very likely to increase the cost of living for citizens of wealthy countries and also to lead to employment dislocations to which they would be called upon to adapt. In short, the actions that would have to be taken to promote justice globally, whether by way of reforming the international political economy or of enforcing international criminal justice or of supporting effective aid programs or just interventions, can occur only if citizens accept such things as legitimate and necessary objectives of their states. They will impose economic and political costs that citizens have to bear. So there is an important sense in which global responsibility can be promoted only if local citizenship is taken to impose even more far-reaching demands than it does now – if, in other words, it becomes in some respects stronger, not weaker, as a source of moral attachment. This book tries to explain how it is that states can at once be more demanding of their citizens and more open to what they owe to outsiders. It rejects the weighing-and-balancing model according to which what one level gets entails a loss to what another might have legitimately expected. It attempts to outline a framework of entitlement that is common to insiders and outsiders. Insiders, it concludes, are indeed special for good reasons, but the reasons for their specialness, while strong, also carry with them strong reasons for concern for outsiders. Since the argument of this book draws on what it is that justifies local obligation, and seeks to draw broader implications from

Introduction 7

it, it must fi rst come to terms (in chapter 1) with the view that local obligation needs no external justification at all. The meaning of local obligations is falsified, some maintain, by the requirement that they be “derived from” or “reduced to” obligations of a more general kind. That view, developed in various sophisticated ways by theorists such as Oldenquist, Rorty, Scheffler, and Horton, is assessed in this chapter, which discusses the ideas of derivation and reduction and rejects the view in question on the grounds that it draws too tight a connection between the “phenomenology” of association and its moral supportability. It is acknowledged, however, that this critique falls short of resolving what is termed the “particularity problem,” that is, the problem of connecting a view of obligation supported by general morality with membership in an arbitrarily particular society. Chapter 2 addresses the particularity problem directly, acknowledging that it resists several proposed solutions, but arguing that the failed solutions make the mistake of relying, in various ways, upon the receipt of benefits as the source of obligation. Ultimately, it is argued, the time-honored receipt-of-benefits argument, although it corresponds to a good account of what justifies political society in general, simply cannot particularize obligation, because the bare fact of the current receipt of benefits cannot justify the exclusion of others from them. (It can provide justification but not legitimation, to borrow A. John Simmons’ distinction.) However, another feature of association, the fact of shared exposure to risk makes a solution possible. Members of a political association are complicit in a set of arrangements that exposes their fellow-citizens to distinctive kinds of risk, and, correspondingly, they have special obligations to their compatriots: obligations to ward off shared risk, and to devote resources to doing so. This argument is distinguished from a currently influential view that special compatriot obligations arise from common exposure to coercion, a view that relies on too sharp a line between state and interstate organizations (as other critics have rightly objected).


Chapter 3 addresses the background theory of justice implied by the preceding chapter. Chapter 2 distinguished between general justifications of political society, in terms of their conferral of benefits, and the legitimation of a particular political society to its members, in such a way as to support local obligation. Although the two levels of argument are importantly distinct, chapter 3 argues for a contractualist version of justice that sustains both of them. We may conceptualize political society in terms of a “social waiver” (of background rights), a waiver that responds to antecedent risks and (as the previous chapter argued) generates subsequent ones. It is in focusing on the notion of subsequent risk that the contractualism offered here differs from other versions. The legitimation of a society depends on an ongoing corrective process in which members are morally engaged by their complicity, a notion whose in-context meaning is explored. Since this approach is generalizable to other (substate) institutions, it provides a unified approach that offers an alternative to the moral dualism that Nagel and Scheffler deploy, replacing it with a model of “subsidiarity.” Chapter 4 situates the argument in terms of other views of particularized obligation or “compatriot preference,” such as those offered by Wellman, Mason, Goodin, and Nagel. Its main focus, however, is on the claim, common to Rawls and his more cosmopolitan followers, that any domestic version of the social contract has no direct implications, in terms of justice, for those who are not parties to it. But the social waiver version, it is argued, has strong iterative implications, so that it serves not only to justify but also to limit the moral basis of preference for insiders: the exclusiveness of one’s own political society, which imposes external costs, is justifiable only on the basis of respect for the parallel social projects of others, and implies a duty of aid when those projects fail, as well as a duty not to promote their failure. These external duties of justice rest on the same moral basis as one’s obligations to compatriots. The argument developed above is clearly cosmopolitan, in some sense, in taking as its starting-point a moral regard for other

Introduction 9

humans. But this regard is mediated, it is argued in chapter 5, by the consequences of political membership, duties arising from it by virtue of the argument from iteration. Other theorists have argued, however, that cosmopolitanism entails direct (unmediated) “natural duty.” Otherwise, they argue, there can be no obligations of justice beyond institutional borders, and hence no obligation to introduce justice in a non-ideal world. To explore this view, this chapter takes up the case of humanitarian intervention, for it is said to be an especially hard case for contractualist views in this regard; and it argues that the idea of iteration leads to a duty to intervene in cases of state failure or violent oppression. The serious practical obstacles to intervention suggest, however, that international law and economic justice, the topics of the next two chapters, often or even generally provide better ways of meeting the duty in question, in a longer-term view anyway. Chapter 6 continues the iterative argument in relation to the idea of crime against humanity. That idea expresses in its most radical form the category of associative risk introduced in chapter 2: it is the transformation of state power from benign form to extreme malignancy. It should be defined, it is argued, in terms of the perversion or travesty of three features of the state (administrative capacity, local authority, and territoriality) that are essential to its functioning. This provides a reasonable approximation to the legal elements of the crime, such as the “systematic” requirement. It is also defended against other proposed views (the stimulating views of Hannah Arendt, Larry May, James Bohman) as the best way of making sense of the moral core of crime against humanity. The fi nal chapter moves on to the context of international political economy, examining the idea (Pogge, Linklater) of a global harm principle as a basis for the policies and practices of wealthy countries and the international institutions that they support. The principle is endorsed. But just as the domestic harm principle, notoriously, needs an interpreting theory, so too does its global version, because a conception of harm requires a background idea of what


is owed. It is argued that the most coherent interpretation of global “harm” is supported by the framework developed in this book: drawing again on the model of iteration, the “harm” that counts is harm to others’ capacity to develop their own social projects. The distributive implications, although not directly egalitarian, are quite strong, but they are mediated through the political consequences of economic practices. The conclusion situates the book’s argument in relation to other theories of global citizenship. It offers a distinction between moral cosmopolitanism as “citizenship of the world” (variously understood) and cosmopolitanism as understood in this book – “citizenship in the world” – the point of the distinction being that the political requirements of the basic moral view are worked out through the implications of citizenship itself, by way of the idea of the social waiver and its iterative implications. The strength of that view, it is argued, is that it places duties of global justice on the same basis as the duties of citizenship itself. The two are not, as is often supposed, in competition, or else incommensurable with one another; on the contrary, to understand why one has duties to those within one’s borders is also to understand the need for cosmopolitan regard.

1

Against associative obligations

This book sets itself the task of showing that the reasons justifying local obligation imply wider duties. That project at once meets a very basic obstacle indeed: the view that local obligations simply need no justifying reasons of a more general kind. The strongest versions of that view hold that all morality whatsoever is essentially local, so that to require further justification of local morality is to make a vain appeal to an empty or imaginary category. More moderate versions accept that there is a general morality that is not tied to local context, but that local obligations do not depend on any more general moral considerations for their force; local and general moralities coexist. Whether we consider the claim’s more radical versions (which some would describe as relativistic) or its more moderate versions (best described as “dualist”), we confront, clearly, an issue that must be addressed. If local obligations are sustained by reasons that are themselves essentially local in their reach, the task set by this book is quixotic. If local obligations are self-justifying, we can hardly suppose that further obligations could be extracted from them. This chapter will maintain, however, that such claims about local obligation should be abandoned. Now it is not at all uncommon to explain our own or other people’s action simply by referring to some relevant relationship, or to one person’s special position in relation to another. More than that, referring to some relevant relationship serves not just to explain but to justify the action, or else to criticize it on moral grounds. So, for example, “He’s my friend” does not just explain why I am lending him some treasured object, but tells us why it is right for me to do so, and “But you’re married” would amount, in context, to an immediately intelligible critique of some proposed action or omission, which


it declares to be wrong. To an audience familiar with the expectations standardly connected with the relationship, elaboration is generally not needed. Knowing about the expectations is simply part of the set of practices to which the action in question belongs. But suppose we move to the reflective level, or the level of critical inquiry: is the reference (together, of course, with local background knowledge) enough? Here, interpreters differ quite sharply. Some maintain that the mere citation of relationships can never be enough for reflective purposes. A. John Simmons, for example, using the language of “position” rather than “relationship,” writes: “If a positional duty is binding on us, it is because there are grounds for a moral requirement to perform that positional duty which are independent of the position and the scheme which defi nes it.”1 It follows that, when the existence of the position is cited in support, it is always possible to examine those grounds, even when it is not necessary to do so. Likewise, Brian Barry has made an influential distinction between fi rst-order and second-order levels of moral practice: at the firstorder level, it is very often enough to carry out the special duties attached to a position or role, while at the second-order level we can be called upon for an account of why those special duties are justifiable in a general (“impartial”) context.2 Martha Nussbaum employs the example of childcare, which “will be poorly done if each thinks herself equally responsible for all, rather than giving the immediate surroundings special attention and care. To give one’s own sphere special care is justifiable in universalist terms.”3 (She believes, somewhat more controversially, that the example should govern our views of citizenship.) There are, then, on such views, no moral duties that cannot be framed, at a more general or abstract or reflective level, in terms of impartial morality. Particular role-obligations should be

1

2 3

A. John Simmons, Moral Principles and Political Obligations (Princeton: Princeton University Press, 1979), 21. Brian Barry, Justice as Impartiality (Oxford: Clarendon, 1995), 194–95. Martha Nussbaum, “Patriotism and Cosmopolitanism,” in For Love of Country, ed. Joshua Cohen (Boston: Beacon, 1996), 13.

Against associative obligations 13

understood as vehicles of larger purposes, and taken to be justifiable only to the extent that the larger purposes sustaining them are themselves justifiable. On the other hand, some interpreters insist there are cases in which citing the position is sufficient, that there are positional or – as they are now generally called – “associative” duties that do not need to be derived from some further moral validation. Steven Lukes, for example – contributing to a discussion provoked by the paper by Nussbaum, quoted above – remarks that moral universalism and local allegiance “are distinct aspects of the moral life and the second cannot be derived from the fi rst.”4 Susan Mendus, referring to the work of several other scholars, notes the “directly motivating or unmediated quality of acts” done from special relationships, a quality such that “no further argument or inference is needed” – in fact, as she points out, further argument or inference may entirely undermine the reception of the acts themselves.5 (Telling someone to whom one is doing a favor that “doing favors is good” certainly diminishes the value of the act.) David Miller, likewise, insists that alongside whatever universal duties we may have there are “independent, nonderivative principles with a more restricted scope” and that “our particular identities as members of communities, groups, and associations of various kinds have intrinsic ethical relevance.”6 In particular, Samuel Scheffler, in a book-length discussion that focuses largely on this very topic, draws attention repeatedly to “the unmediated moral significance of those special ties which bind members of a community to each other”: again, “to attach noninstrumental value to my relationship with a particular person just is, in part, to see that person as a source of special claims in virtue of

4 5

6

Steven Lukes, Liberals and Cannibals (London: Verso, 2003), 22. Susan Mendus, Impartiality in Moral and Political Philosophy (Oxford: Oxford University Press, 2002), 79–80. David Miller, “The Limits of Cosmopolitan Justice,” in International Society: Diverse Ethical Perspectives, ed. David R. Mapel and Terry Nardin (Princeton: Princeton University Press, 1998), 166–67.


the relationship between us.”7 So, on such views, to see special ties as vehicles of larger purposes is to falsify them. What, though, do terms such as reducibility and irreducibility, derivability and underivability, immediacy and mediation, mean in this context? I believe that the attractiveness of the second view – now generally termed “associativist” – depends on the conflation of a number of different claims, many of which have initial plausibility, and even have significant reach when considered, but none of which fi nally stand up to critical analysis. This chapter will distinguish between, and criticize, a series of different versions or interpretations of the associativist view, some concerning the origin of associative obligations, some their content, some their authority, and others (what is called) their “phenomenology.” It will, finally identify the strongest version – which, however, will need a chapter to itself.

I I shall begin with two related theses about where moral beliefs come from. The fi rst may be termed “the socialization thesis,” and it relates to the temporal sequence of our acquisition of duties. It amounts to the view that associative duties cannot be instances of universal duties for the simple reason that they “come fi rst.” Some of Martha Nussbaum’s critics appear to adopt this view in some form. Michael McConnell writes: “One who knows no heroes in his own land will feel nothing but contempt for the naiveté of those who honor heroes elsewhere. Before a child can learn to value others he needs to learn to value.”8 While the second sentence is analytically true it does not support the fi rst, because among those whom one fi rst values may be some who are not in one’s own land at all. Nussbaum replies, effectively, that “Good fairy tales are rarely about 7

8

Samuel Scheffler, Boundaries and Allegiances (Oxford: Oxford University Press, 2001), 35, 100 (emphasis added). Michael McConnell, “Don’t Neglect the Little Platoons,” in For Love of Country, ed. Cohen, 80.


Cambridge, Massachusetts,”9 and that the moral imaginations of children are formed in part not by the local but by the distant, or perhaps more precisely, by fables in which the difference between the local and the distant tends to evaporate altogether, through processes of imaginative merging. Moreover, long before fables of any kind come into play, children’s basic orientation to the world is fundamentally molded by the generic circumstances of infanthood, by hunger, dependency, bonding, fear, separation, identity-formation, all in relation to a parent or parents, but forming a general matrix for future relations with all other humans. But even if the critics’ view were true, it is hard to know what would follow for “reducibility,” for temporal priority does not confer fi nal authority on beliefs. No doubt it is normal to learn simple and specific things before we encounter more advanced and abstract considerations – we learn the arithmetical tables before we learn about number theory (if we ever do) – but that does not mean that the specific is irreducible to the abstract, for that issue is not about whether A precedes B, but about whether B can logically accommodate and explain A, and about whether A could be true if B were not. Closely related to the socialization thesis, often in fact fused with it, is a somewhat different “expansion thesis,” according to which relatively large identifications are necessarily arrived at by expanding relatively small ones. (The socialization thesis claims only that A comes before B, the expansion thesis claims that B actually is a larger form of A.) This idea, with deep roots in conservative thought, is clearly one that many people fi nd intuitively plausible. But the model of expansion seems questionable, particularly

9

Nussbaum, “Patriotism and Cosmopolitanism,” 143. Jeremy Waldron makes much the same point in commenting on a famous passage in Alasdair MacIntyre’s defense of the communitarian view: stressing the importance of roles in making sense of life, MacIntyre refers to the impact on children of stories about wicked stepmothers and wolf-children – stories, Waldron points out, from a variety of spatially and temporally distant cultures: Jeremy Waldron, “Minority Cultures and the Cosmopolitan Alternative,” in The Rights of Minority Cultures, ed. Will Kymlicka (Oxford: Oxford University Press, 1995), 107.


perhaps when we reach the level of the nation. Why should it stop expanding there, so that English people feel differently about their own national anthem than they do about the much more tuneful “Marseillaise,” so that admiration of national military heroes does not easily extend to their opponents, even if more heroic and skillful, so that those who celebrate their quite good national triumphs do not commiserate with the entirely dreadful national disasters of others? The most obvious anecdotal evidence suggests that patriotism is more likely to be morally absorbing, as it were, than morally expansive – that is, it may violently close off, rather than extrapolate, the idea of connectedness. Moreover, if there are consequential links between small and large identities, it would seem odd if they only ever ran one way, large ones never reacting upon small ones. Most interpretations of the rise of nation-states tell us that, far from arising seamlessly from family or village or tribe, they brutally extinguished local attachments. Tribes or perhaps small nations, to be sure, draw in part from the idea of the family: but the large modern nation established its claims only by representing itself as a proxy for humanity in general; and in doing so it tended to permeate itself, and its components, with ideas of equality that color the meaning of citizenship. The nation-state, then, may have to be understood not as the product of an expansive process, but as (in part) a top-down construct in which, at successive levels, smaller groups are interpreted and reshaped as proxies for larger ones. Within their own spaces, nations have done all the homogenizing things that globalization is accused of.10 So they cannot convincingly rally under the “partiality” banner in warding off the claims of cosmopolitan morality. The expansion thesis was, however, asserted with great confidence by Richard Rorty, who believed that the idea of “humanity” could arise only by way of extending or enlarging previous

10

See Kwame A. Appiah, The Ethics of Identity (Princeton: Princeton University Press, 2005), 239–40.


sentiments of partial loyalty.11 But in the case of European rescuers of Jews from Nazi genocide, and in other cases, evidence suggests that this was simply not so. Again and again, rescuers explained their actions on the basis of humanitarian, not neighborly, considerations – they “did not distinguish between in-group and out-group members when they risked their lives to save strangers. They neither asked nor appeared even to consider whether they had something in common with the people they saved.”12 And on the other side of the coin, neighborly considerations were plainly no barrier to collaboration in genocide. While strangers were sometimes rescued “because they were human,” neighbors were regularly abandoned despite familiarity.13 The relatively assimilated, some evidence suggests – that is, those who were more likely to have developed ties of intimacy and locality – were actually more likely to fall victim to genocide than the less assimilated. And in the most recent genocides (Rwanda, the former Yugoslavia) the slaughter of neighbor by neighbor was among the most horrific and morally baffling of their features.14 In the absence of supporting evidence, views such as Rorty’s must be seen as intuitively attractive but sadly optimistic guesses about the benign causality of closeness.

II The associativist view may be advanced, however, in terms of the content of particular obligations, rather than their origin. It may be proposed, for example, that one’s particular obligations do not consist wholly of elements that can be derived from general morality; or, in what I take to be another way of formulating the view, do not contain only elements that could be predicted from a knowledge of general 11

12

13

14

Richard Rorty, “Justice as a Larger Loyalty,” in Cosmopolitics, ed. Pheng Cheah and Bruce Robbins (Minneapolis: University of Minnesota Press, 1998), 45–58. Kristen Renwick Monroe, The Heart of Altruism: Perceptions of a Common Humanity, (Princeton: Princeton University Press, 1996), 199. Norman Geras, Solidarity in the Conversation of Mankind (London: Verso, 1995), especially chapter 1. See Victoria M. Esses and Richard Vernon (eds.), Explaining the Breakdown of Ethnic Relations: Why Neighbours Kill, Malden, MA: Blackwell, 2008.


morality. Their content would be “inexplicable” without reference to their local context, and they are in that sense “distinctive.”15 This version of the claim, however, sets very modest goals. One does not have to be an associative-obligation theorist to acknowledge that when principle meets context, local accommodations must be made. But that hardly diminishes the force or applicability of general morality. A familiar distinction between criterion and standard goes a long way here: a criterion tells us what kind of considerations it is appropriate to bring to bear, a standard tells us how, quantitatively, to measure them. So a common criterion may meet with different measures in different locales. Decency in dress is a simple example: different societies accept it as a common value, but define its practical requirements very differently (in for example San Diego and Tehran). What is required in any given context is indeed inexplicable without reference to local history, and in that sense is indeed distinctive. But the varied local requirements do not tell against the generally shared principle – if anything, they reinforce it, by showing that the shared principle has force in different contexts. What is at issue in local contexts, moreover, may be better understood in terms of a weighing of conflicting general principles, colliding in their requirements, than in terms of some unique and unmediated local imperative. In several parts of the world, for example, local jurisdictions have inherited a constitutionally entrenched policy of using public funds to support denominational schools. In one case (that of Ontario, Canada) that practice has been formally condemned by the United Nations Committee on Human Rights (UNCHR), on the grounds that equivalent provision is not made for the schools associated with other denominations. Suppose we support or reject the practice. We would reject it, as the UNCHR proposes, on the basis of an abstract idea of equality that dispenses with local knowledge. We would support it on the basis of a principle that agreements with constitutional status 15

John Horton adopts the term “distinctive” in his “Defending Associative Political Obligations: A Response to Richard Vernon,” Political Studies 55 (2007): 880–84.


should be kept – “pacta sunt servanda.” What we confront, in other words, is the collision between two general principles that the context compels attention to. It is not a matter of a collision between a general principle and an appeal to “what we do here,” which is not actually an appeal at all, but a morally vacuous description. The associativist thesis valuably reminds us of the complex ways in which moral goods interweave with other goods.16 Here, however, much depends on the examples that a proponent of the view has in mind. We may understand the “moral” in a narrow sense in which it does not embrace all normativity (all goods and values) but refers only to some smaller category, such as “duty,” perhaps, or at any rate a branch of normativity that concerns itself with general rules and principles. One kind of example of the distinction is provided by conventions, such as greeting rituals, that are certainly normative but which fall short of such things as promise-keeping or truth-keeping or the other standard examples of moral duty. Another would be those expectations, arising from local connection, that go, rather, above and beyond any duty specifiable as a universal rule: for example, the expectation that if a friend suffers a loss one will try to make it up, even if one has not caused it oneself.17 The first kind of example does not seriously strain the criterion/standard distinction. Greeting rituals are perfunctory in North America, more formal in France, quite elaborate in West Africa, but these different standards give operational form to a criterion of politeness, in turn sustainable by a general duty not to give gratuitous offense. Conceptualizing them in this moral way certainly misses local color, but not the point of the practices, (taking “the point” to mean what would have to be explained to a visiting anthropologist). The second kind of example is well chosen to defeat an ethical view that insists on deriving everything from rights: your friend has no right 16

17

John Horton, “In Defence of Associative Political Obligations: Part One,” Political Studies 54 (2006): 438. See Joseph Raz, “Rights-Based Moralities,” in Theories of Rights, ed. Jeremy Waldron (Oxford: Oxford University Press, 1984), 182–200.


to have a loss made up by you when you are innocent of causing it, but the expectation is woven into the meaning of friendship. But while a rights-based view is thus precluded, nothing except word-choice precludes expressing the expectation in terms of duties of friendship, in turn sustainable by (for example) a general principle about aiding those who particularly rely upon us. Those responses to both kinds of examples, however, can only be provisional, because they depend on accepting certain answers to other questions posed below.

III This leads us directly to another version of the claim: perhaps the irreducibility that matters is that of the authority of duties18 as opposed to that of their content. The content of local duties may resemble the content of general ones: but do they bind only because they are local? Associative duties call upon us authoritatively, and in a way that does not depend on their being somehow licensed to do so by cosmopolitan morality. Scheffler believes that the authority of associative duties creates a dilemma for theorists such as Martha Nussbaum. According to Nussbaum, as we have seen, it is better for all humans if we give more weight to the needs of our own children than to those of other people’s children, and to the needs of co-citizens than to those of foreigners. Things work better that way, so cosmopolitanism accommodates local duties. (Of course, this does not mean we should then give no weight to other people’s children, or to foreigners – it just means that equal worth does not require equal attention, because there is an obvious distinction between the level of basic justification on the one hand and the operational level on the other.) But all the same, according to Scheffler, Nussbaum faces a dilemma.19 If she is to support giving special weight to (e.g.) one’s own children, she must choose between (1) claiming that one’s own children are of greater value than other people’s, and (2) claiming that one’s preference for one’s own is grounded in overall human 18 19

Scheffler, Boundaries and Allegiances, 64. Ibid., 118.


benefit. For different reasons, neither fork is acceptable. (1) is clearly unacceptable because it denies the premise of human equality. (2) is unacceptable because it misrepresents the nature of our relationship to our children (etc.). This is because we do not first have a relationship with a child (spouse, sibling, friend, etc.) and then ground it in an obligation of some general kind: the relationship itself contains, is constituted by, obligations. It does not need to import obligations from elsewhere, because obligations are part of what it is. The cosmopolitan view that it depends on justification in universal terms is false. Can Scheffler’s objection be met? It is of course entirely implausible to think that I owe things to my children because my giving them those things is good for humanity. The universalist does not, however, confront exactly that gross absurdity, for we must distinguish, fi rst of all, between the justification of acts and the justification of general practices (e.g. between justifying what I do for my son, and justifying “what parents should do for children”). The effect of a practice-level justification is to relieve the acts which it covers of their immediate justificatory burden (e.g. if I have reason to think that parents should give preference to their children, that means that I no longer have to justify the preference I give to my own children). As far as it goes, this may seem only to move the argument one step back: for one could simply object that all the reasons for thinking that one’s obligation to one’s son needs justification arise again in connection with thinking that parents’ obligations to children need it. But insisting that the justification be a practice-level one opens up a new question. What is it, exactly, that practice-level justification accomplishes? I want to say that, in the case of extending benefits to “associates” (Scheffler’s useful generic term), what is relieved is the burden of justifying the exclusion of non-associates.20 It should be put in this way in order to make it clear that we do not need 20

See Jeff McMahan, “The Limits of National Partiality,” in The Morality of Nationalism, ed. Robert McKim and Jeff McMahan (New York: Oxford University Press, 1997), 118.


to justify the giving of benefits to people, because there is nothing wrong with it. It is the non-giving of benefits to other people that, potentially, needs to be justified. Is this just a verbal quibble? Could we not say that the issue is raised by “giving to some but not to others,” and that distinguishing between (a) giving to some and (b) not giving to others, is artificial? There is a twofold answer to this. First, there is nothing inherently wrong with giving to some but not to others, and if an instance of that is wrong, that is because of some additional feature. For example, suppose you give your last dollar to homeless person A, so that when homeless person B approaches you there is nothing left for him: that does not mean it was wrong of you to give money to homeless person A. The unequal treatment of A and B is due to sheer luck. It might be a different matter if, for example, the unequal treatment of A and B arose from some institutional arrangement, but the bare fact that one benefits while another does not is morally unobjectionable. So “giving to some but not to others” is as it stands a morally neutral phenomenon that acquires moral significance, if it does, only from some additional element. The bare fact that B does not benefit does not give grounds for depriving A of the benefit. Second, we can see the importance of the distinction at issue if we look at the disanalogy, in terms of justificatory requirements, between conferring benefits and imposing costs. Suppose I impose some sort of loss on person A, not on person B. Obviously I have to justify to A what I am doing to her. The question of justifying to B the non-imposition of costs on her does not arise. There is a parallel in the theory of punishment. James Rachels argues that if someone has done me a favor then he deserves to receive a favor from me in preference to some other claimant; and then he argues, by analogy, that if someone has offended then they deserve to receive punishment.21 But the analogy fails to be persuasive,

21

James Rachels, “Punishment and Desert,” in Ethics in Practice, ed. Hugh LaFollette (Oxford: Blackwell, 2000), 470–79. John Horton has suggested to me


because what the fi rst case establishes is only the justification of a negative act or omission (i.e. not giving someone something), whereas the second case requires the justification of a positive act (i.e. punishing). The fi rst case only shows, by analogy, that it is wrong to punish the innocent; it does not show that it is right to punish the guilty. The point is, then, that whatever it is that justifies the nongiving of benefits to non-associates leaves the reasons for the giving of benefits to associates entirely intact. It does not challenge them. It simply relieves those reasons (whatever they are) of a potential justificatory barrier. So the second fork of Nussbaum’s dilemma is removed, just because the relevant justifications are not justifications for giving (but for excluding). An example: we give birthday presents to our own (and friends’) children. That practice, I assume, survives appropriate tests of global justifiability. But the fact that one’s reasons for giving one’s children presents survive this test clearly does not mean that one’s reasons for giving them presents must be reducible to/derived from global principles: it means that those reasons (whatever they are) escape scrutiny because the nonrecipients have no rights in the matter. Gross nepotism, on the other hand, would face (and fail) scrutiny, because it causes harm to others who have legitimate expectations of fairness. At least, it would in societies such as ours, whereas similar favoritism in more strongly kin-based societies might be viewed more as present-giving: which leads us to acknowledge the interpenetration of associative duties and larger contexts of social principle, and thus the fact that they do not inhabit different moral worlds. Is it the case that associative obligations are independent of general moral principles in the sense that they are not required to that my argument here may press too hard on the act–omission distinction, a distinction that in some contexts is, of course, questionable. An omission is morally equivalent to an act when there is an unmet duty to perform it. But in this paragraph my argument assumes, in the case of punishment, that there is no prior duty to impose costs, and, in the case of gratitude, that there is no prior duty to confer benefits on the person who has conferred none on me.


be consistent with them? This view may take one of three forms: (a) associative obligations would be valid even if there were not any sustaining general principle; (b) they would be valid even if contradicted by a general moral principle – as we shall see, (b) has to be further subdivided; or (c) when there is a corresponding general principle of a supportive kind, the associative obligation does not depend on it. Kwame Anthony Appiah puts this last point very nicely: while there may be health reasons for eating chocolate, he points out, those are not actually the reasons why people eat it.22 Taking (a) first, we at once encounter the difficulty that the categories needed for comparison do not seem to be genuinely independent. Is something properly called an “obligation” if there is no general principle implied by it, or available for its justification? We distinguish, after all, between etiquette and ethics, and associations may be made up in part of shared practices that we would place in the former category, where “it is what is done” is sufficient justification, rather than the latter. That is not to say that something may not cross the line from a matter of etiquette to a matter of ethics – but that is because it crosses a line. It crosses the line because some failure of compliance can no longer be seen as mere ignorance of local practice, but has to be viewed as culpable ignorance (on the part of an outsider) or deliberate contempt (by outsiders or insiders). Culpable ignorance and deliberate contempt bring us into the realm of general morality: so when we rely on those elements to declare some local practice “obligatory,” we have already implicitly settled the question. I do not mean to say that this should lead us to one answer or another, just that (a) seems to lead nowhere, because the way words are generally used seems to rule out any possibility of finding an example. Let us turn to (b), the case of confl ict. Some of the requirements of social identity can be “obnoxious,” in Simon Caney’s phrase,23 and call out for rejection by those who assign moral weight 22 23

Appiah, The Ethics of Identity, 330. Simon Caney, “Individuals, Nations and Obligations,” in National Rights, International Obligations, ed. Simon Caney et al., (Boulder, CO: Westview, 1996), 127. In adopting Caney’s example I do not mean to imply that only local


to important general principles, such as, and especially, equality. If putative role or membership obligations can be cancelled when they conflict with important principles of general morality, that may seem to demonstrate their subordinate or conditional (if not actually derivative) character. And that is often taken to be a decisive objection to regarding them as obligations at all, except to the extent that they coincide with general morality at all by virtue of passing a filtering test of the sort sketched above. But Horton24 and Michael Hardimon 25 offer an important counter-objection that, if successful, would settle much controversy. There are also cases, they point out, in which obligations of general morality are cancelled by outweighing considerations, as, for example when someone makes a promise to perform a wrongful act, such as murder. We suppose that that voids the promise, without also being led to suppose, therefore, that promising has no moral weight. So, by parallel reasoning, we should not suppose that a role or membership obligation has no moral weight just because it is subject to filtering and, potentially, cancellation. Something can have moral weight, after all, without having absolute or final moral weight. Horton and Yael Tamir provide similar examples. Horton26 draws on the notion of honor among thieves – it is some merit in a thief, even if it’s overridden by our judgments about the wrong of theft, that he keeps faith with some arrangement that he has agreed to about the distribution of spoils. He would be worse if he both robbed and reneged on his co-robbers. Tamir27 also cites the case of someone who serves a criminal group, but who does so with honor

24

25 26

27

practices are obnoxious, for naturally many beliefs that are held to be general principles can be obnoxious too. John Horton, Political Obligation, (Atlantic Highlands. NJ: Humanities Press), 156–57, and “In Defence” (2006): 437. Michael Hardimon, “Role Obligations,” Journal of Philosophy 91 (1994): 344. Horton, “In Defence” (2006): 438. In citing these similar views from Horton and Tamir I do not mean to imply that their positions are identical. See John Horton, “In Defence of Associative Political Obligations: Part Two,” Political Studies 55 (2007): 8–9. Yael Tamir, Liberal Nationalism (Princeton: Princeton University Press, 1993), 101–02.


and integrity. He would be worse – she implies – if he served that group in a duplicitous way that (say) lined his own pockets beyond his share. I agree with their conclusions: the treacherous thief is even worse than the honorable thief, the corrupt Mafioso even worse than the honorable Mafioso. But I do not agree that this, as such, establishes that local obligation has weight. For we need to distinguish between several kinds of case in order to test the success of the parallel with the cancellation of promises. In one kind of case – which may present the clearest parallel to the “obnoxious identity” problem – we do not in fact regard the promise as having moral weight. If we did, then we would regard upholding the promise to murder as mitigating the wrong of murder, and thus find a promise-keeping murderer less blameworthy than, say, an impulsive one. In fact, of course, we are likely to do the exact opposite, finding impulsive murderers less repellent than reflective ones. The case of the dutiful murderer is thus distinct from three other kinds of case in which we reason differently. In one, an agent has to choose between keeping a promise and pursuing another good: this is the (trite) example of breaking an appointment in order to be a Good Samaritan to an accident victim. The promise to keep the appointment has weight, but not enough. In a second kind of case, what would otherwise be a wrong action would lose its wrongness because moral necessity imposes it: the storm trooper at the door demands to know who is in the basement, and you (lying) say that no one is. In a third kind of case, an on-balance evil person may have features that we regard as good – according to the movie Downfall, Hitler, in his bunker, was courteous and considerate to his office staff. To borrow a distinction from a classic paper by Charles Taylor,28 the three latter cases are cases of “overriding,” that is, they are cases in which some good act is recognized as good even though, in a balancing contextual judgment, something else – the pursuit of another good, the 28

Charles Taylor, “Neutrality in Political Science,” in Philosophy, Politics and Society, 3rd series, ed. Peter Laslett and W.G. Runciman (Oxford: Blackwell, 1967), 25–57.


condemnation of an evil – is more compelling. What we might term a “checklist” morality applies – we acknowledge the independent force of various considerations, and weigh up the pluses and minuses at the end. The case of the dutiful murderer, however, is one of “undermining”: that is, whatever force his promise might have had is voided, as opposed to outweighed, by the purpose to which he puts it. And if the weight of promises depends on the legitimacy of the purposes to which they are put, then the proposed parallel has negative consequences for the associative-obligation view: it means that the weight of putative role or membership obligations depends on their justifiability within general morality. The overriding (or checklist) model does not apply, because in such cases judgments are not discrete but mutually influential, and it is artificial to separate the evaluation of steps from the evaluation of goals. To this, however, there is a possible reply. The activity of promising, in general, retains its own moral force, even though in particular cases promises have no force because they are put to evil use. So, restoring the proposed parallel, could we not say that associative obligations have their own moral force even though in particular cases their force is undermined (not just overridden)? Although that would be a reasonable response, it would lead in a direction that takes us quite abruptly away from the associative obligation thesis. For it would require, first, that each of the associative obligations that we wished to rescue would have to be framed as a member of a general class (analogous to promises) and, second, that we would have to sanction each of these classes in terms of the view of general morality that we favored. We would have to frame particular acts as instances of, say, “loyalty,” and be prepared to give an account of that general value in terms of its contribution to some end of compelling general importance. But would we not then be on the terrain of general, not associative, morality? At most, we might have persuaded an unworldly general moralist of the importance of values and practices that pertain to local contexts – a worthwhile success, but one that stops short of establishing the associative obligation


thesis. We would be better described as usefully enlarging and/or refi ning the list of general values in the light of particular ends. These considerations cast doubt on a very common defense of the associativist view that stresses the element of meaning. Obligations are implicit, it may be held, in the very description of the associative relation, so that to ask for further support is simply to miss the point. This is the “one thought too many” objection, in a formulation made famous by Bernard Williams,29 in the course of criticizing the view that the normativeness of roles depends on some consideration of their general worthwhileness or utility: we will rescue our loved ones from a fire – despite William Godwin’s notoriously bad advice – because we love them, not because the world goes better if people favor those whom they love. If one gives preferential treatment to (say) a parent, that is adequately explained, and justified if justification is called for, by the consideration “s/he is my parent.” That the world goes better, given that consideration, may be true, but it is a redundant thought – redundant because it leads to the same conclusion by a longer route. But can we say both that general morality is redundant when it coincides with a role-obligation, and that it becomes material when it confl icts with one? That would seem odd. It would seem to imply that, for purely internal reasons, associative obligations had a sort of unconscious sensor that for the most part avoided collisions with general morality, occasionally breaking down. It would seem to imply that what happens inside a role, as it were, is autarchic, and that morality comes into play only sporadically at its outer boundaries. (Rather as states are sovereign, up to the point at which they violate people’s rights?) The idea that features connected with a valued role are morally validated without reference to anything outside the role is surely indefensible. Some familiar examples tell against it. One can endorse the role of “fatherhood” while rejecting the view that its features include (as was once believed, by, for example, Jean Bodin) the power of life and 29

Bernard Williams, Moral Luck (Cambridge: Cambridge University Press, 1981), 18.


death over children, one can endorse the role of “widowhood” while rejecting the view that its features include a duty of self-immolation on the funeral pyre, and so on. A supposed duty of self-immolation by widows, however, is (allegedly) justified exactly by its particularity, as a distinctive cultural feature – that is what widows, in our culture, do, some said. It is what widowhood “means.” Why, though, do we accept conceptions of family duty that involve rescuing people from fi res, while we recoil from conceptions of family duty that involve people having to throw themselves into flames? It seems quite implausible to say that failing to rescue a family member from a fire is a case of family duty not going far enough, while requiring a family member to burn herself alive is a case of family duty going too far. It is not a matter of the boundaries, small or large, of family duty; it is a matter of the very idea of the family itself. If it is an idea that involves utilitarian calculations of the kind that Godwin recommended, or relations of domination of the sort entailed by the practice of sati, it is not a good idea. The requirements of morality do not apply at the edges, they apply from the ground up. Thirdly, let us consider (c), the view that associative obligations do not depend upon the truth of general moral considerations that might be brought in their support. This differs from the cases discussed immediately above, where it was a case of there being (a) no general principle or (b) a conflicting general principle. Here we have parallel general principles and local practices, which are consistent in that the general principles could be called upon in support of them – but we are told that we need not do so, for the local practices are sufficiently valid already. The non-dependence thesis is puzzling when the general principle simply states the value of the local practice in a more abstract way. A particular act – in war, or sport, or childbirth, say – may be called “courageous,” and there is a corresponding general belief that courage is a virtue. But the corresponding general belief simply sums and generalizes the value of many disparate courageous acts. Can we


say that the local practice is “independent” of the general belief? As an imperfect parallel, we could know that the angles of an equilateral triangle add up to 180 degrees, and also that the same was true of all triangles. Does the former “depend” on the other? The question is odd, and can be answered either way. “No,” we might say, because the equilateral triangle’s angles would add up the way they do anyway (measure them), or “yes,” we might just as well say, because that fact would not be true of the equilateral triangle if it were not for facts about the properties of all triangles whatsoever. The parallel is imperfect, of course, because the contexts of life are multiple and varied, something that ideas of associative obligation rightly point out. But here it is useful to ask why, given the multiplicity and variety of life-contexts, there is any convergence between local practices and general principles at all, let alone the rather high degree of convergence that there is. Two reasons suggest themselves, one historical, the other conceptual. The historical consideration is that in all modern societies local practices have been constrained over a long period by the demands of civic equality. As is well known, James Mill, an Examiner (senior officer) at the East India Company, obtained a junior position at the Company for his seventeen-year-old son, John Stuart. The East India Company no longer exists, of course – another example of differentiation between public and private interest! – but we would certainly no longer accept such blatant nepotism in some equivalent setting. We can think of many similar examples of constraints on partiality, as a result of which what was once a virtue – looking after one’s own – has become a public offense. So in at least some cases, local practice and general principle converge because local practices have been through a process of slow but quite rigorous selection; elements that might otherwise have collided with general principle have already been discredited. On a much more grandiose level, we might associate this idea with nothing less than a philosophy of moral and religious history of the kind proposed by Henri Bergson, who sought to show that our current moral beliefs arise from a systematic compounding


together of “closed” or partial sentiments and “open” or egalitarian aspirations.30 But whether we adopt modest or grandiose versions of the idea, it confronts us with the possibility that local practices have an apparently free-standing appeal only to the extent that they have been through a process of relentless and minute screening over a period of centuries. Their free-standing appeal, in that case, would demonstrate, precisely, their non-independence – they have been subjected. So when we point to their apparently independent force, we are in fact pointing to their deeply conditional status. The conceptual reason, however, suggests a less antagonistic relation between the local and the general. It is that locally valued practices and generally approved principles do not respond to wholly different considerations. To anticipate a theme of chapter 3 below, suppose that we call things good, in a moral sense, if they tend to reduce the vulnerability of others to suffering. That desired tendency certainly guides judgments about local practice, in matters such as the appropriate treatment of family members or friends or associates: associations of all kinds both expose their members to possible pain and offer the means to avert it, and good associations are ones that do more to avert than to expose. Something much like that, however, could – more abstractly stated – capture much of what we see as important in general principles. Confl icts of level will arise in that reducing vulnerability to associates may increase vulnerability in a public context – protecting one’s sons by nepotistic means exposes other people’s sons to such things as undeserved unemployment: but as we have just seen, such collisions have been reduced in number by the slow triumph of public morality.

IV It is certainly true, however, that associative obligations do not require prior acknowledgment of general principles by which they

30

See Richard Vernon, “Pascalian Ethics: Bergson, Levinas, Derrida,” European Journal of Political Theory, forthcoming.


might be supposed to be supported. This is the undoubted jewel in the crown in the associative-obligation thesis – a “phenomenological” jewel, as it has been called. It is sustained by an easy demonstration of absurdity. I do not begin the day by wondering whose children to look after, my own or the next-door neighbors’, reflect that on the whole the world goes better if people look after their own, and thus decide to look after my own, after all. That would certainly be to have one thought too many, and so the associativist view would seem to be sustained by the most elementary phenomenology of moral experience. Theorists of associative obligation rightly make much of this. What follows from it, however, is where the real question is. Do justification and motivation have to coincide? Here quite a range of different cases may present themselves. In some, it is essential to a certain relationship that it be maintained in a certain kind of spirit, for example, in a spirit of affection. While children should be cared for, and families can be justified because they make caring possible, we should give no credit to a parent who looked after his children entirely with an eye to winning father-ofthe-year award, even if his attentiveness was exemplary. In other cases, there is some separability between the subjective side and the act in question, in the sense that the same subjective coloring may be associated with morally different acts. The spirit of “pride,” for example attaches to very different kinds of deed. There is the pride that one person takes in her volunteer work in children’s wards and retirement homes, and the pride that another person takes as a member of a gang that beats up gay people in public parks. In yet other cases it does not seem necessary that the acts be performed in any particular spirit: paying taxes, for example. But in the first type of case, the connection between spirit and deed is so intimate that the latter cannot be suitably identified without the former: “children should be cared for” seems an inadequately filled-out justifying principle for family life, for it falls well short of indicating the attitudes that must infuse caring acts. In the second type of case, our judgment of the agent’s project reacts upon our judgment of the


spirit involved, so that the good person’s pride in her achievement is admired while the homophobic thug’s is not. In the third type of case, we simply detach act and spirit. Clearly it is the second kind of case that is interesting here. In the fi rst, what justifies the action is the same as what must motivate it, so that the question of a relation between the two does not arise. In the third, the question of motivation is beside the point. But in the second, it is clear that the justifying purpose cannot merely override (to revert to an earlier distinction) but can undermine the value that motivates. In cases of that kind, then, what motivates an action may be distinct from what justifies it; but it is denied moral value unless the purpose is good. The upshot is, then, that outside a range of special (first-type) cases, “phenomenology” and morality need not coincide.

V Strong versions of associativism (such as Rorty’s or Oldenquist’s) claim to embrace all morality. More moderate versions accept that there is also a realm of general morality that appropriately operates in terms of non-associative principle. Now all kinds of views can accept that there will be conflicts between what we owe to close associates and what we owe as a matter of a general principle that we might endorse – utilitarian views can do so, for example:31 moderate associativism is distinctive, though, in claiming that, when such conflicts occur, it is not general principle that decides the matter. In fact, it seems, nothing – according to the associativist – can decide the matter, for what we face is the confrontation of two differently grounded moral views. (So local morality could not be the decider either.) Samuel Scheffler has set out and defended this view at length. There is, he says, “a deep and persistent tension between … two features of our moral thought,” and there are no “guarantees that we will be able simultaneously to accommodate both features to our own satisfaction.”32 31 32

See Peter Singer, One World (New Haven: Yale University Press, 2002), chapter 5. Scheffler, Boundaries and Allegiances, 109. I return to Scheffler’s view in chapter 3 below.


Some unrefi ned views of the confl ict between partial and general morality arise from a mistaken assumption about what it is that general morality requires. A confl ict is generated on the basis of a straw-person version of what general morality is. The paradigm case here is, again, that of “the storm trooper at the door.” You are sheltering a persecuted family in your basement, the storm trooper demands to know if you are, and general morality allegedly requires you not to lie. That entails a very formalistic and inflexible view of general morality which makes it the enemy of decency. Moreover, as already noted above, those who shelter the persecuted often do so as a matter of general compassion or a sense of shared humanity, and those who fail to do so are shamefully unaffected by the ties of neighborliness or friendship that we might think would restrain them: so the recruitment of partiality to the side of decency has a distorting effect on the question, and obscures the important way in which people are able to grasp the general in the particular, the human in the person. It may not be the particularity of the victim that matters, just their humanity-in-proximity.33 That point is also obscured by partialist views that, beginning with perfectly true facts about the specific and local character of moral learning, go on to expel from consideration the generalizing capacity that learning also requires. No doubt at least some people will learn more from Huckleberry Finn or To Kill a Mocking-Bird than from The Metaphysics of Morals: but how can they learn from the rich local context of those two fi ne novels without a power to generalize from them? Without any generalizing power, why should readers not regard their message as wholly contained by the immediate context of the story?34 Why should we take Twain’s novel to undermine racism, rather than, say, to celebrate friendship? The argument

33

34

For a defense of this view, see Richard Vernon, Friends, Citizens, Strangers: Essays on Where We Belong (Toronto: University of Toronto Press, 2006), chapter 5. A point made by Geras, Solidarity in the Conversation of Mankind, 98.


from the particularity of moral experience seems to demand, but not to explain, a powerfully intuitive generalizing capacity. This theme has been reinforced in a different way in discussions of one important form of local practice, that of professional ethics. That field certainly contains rich materials for the discussion of conflict, as professional codes often claim to empower practitioners to do things that would be morally objectionable if done without license: a lawyer’s duty to aid her client, for example, often imposing on third parties damage of a kind that would otherwise be outrageous. It is even claimed, sometimes, that these special professional duties spring “naturally” from the very character of the relationship between professional and client, just as special responsibilities inhere in the very nature of family relations. 35 In his discussion of this topic, Gerald Postema considers and rejects a dualist solution that he identifies with Montaigne (who wrote: “The mayor and Montaigne have always been two people, clearly separated”). According to the dualist solution, a professional’s conduct is properly regulated by a code which is quite different from, and does not interact with, her conduct in her non-professional life. Such a solution is clearly unable to make sense of a professional’s efforts to find the right answer – or the best available answer – when the stipulations of a professional code cause harms that are condemned by ordinary morality. These efforts call for all the moral resources available to the person in question.36 Adapting Postema’s initial distinction, we might ask if it is “the mayor” or “Montaigne” who makes these efforts? It cannot be “the mayor,” for from the mayor’s point of view the situation at hand is adequately captured by the terms of professional responsibility. “Montaigne,” on the other hand, is a whole person capable of reflecting critically on the varied facets of his being, including his 35

36

See David Luban, “Professional Ethics,” in A Companion to Applied Ethics, ed. R. G. Frey and Christopher Wellman (Oxford: Blackwell, 2003), 594–95. Gerald Postema, “Moral Responsibility in Professional Ethics,” New York University Law Review 55 (1980): 63–89.


mayoralty. If adjudication can take place, then, it must take place within the space of general moral reflection in which professional obligations are relativized as particular ones. Of course, that does not mean that professional obligations are therefore automatically trumped by general principles. Although they are relativized, they may turn out to provide the best answer to what to do, in context; but if so, that is not because of their own unmediated authority but because their claims are mediated by a process of reflection in which the whole person engages. This analysis transfers from the professional cases to other cases of special relationship easily enough. One reservation that one might have about the transfer is again that in some cases, such as familial ones, the affective tone of the relationship is importantly different from professional commitment of even an enthusiastic kind. We want the responsibilities of familial relationships to be carried out not just with the conscientiousness of a good professional but with deep affection. On some accounts, in fact, that element makes so much difference that it nullifies even the best moral answers – we may want parents to protect their children from the police even in the face of compelling moral reasons to turn them in. 37 I see no reason to adopt this view, however, in the context of the argument developed above, for that argument offers no reason to think that deep affection is discounted within morality. “Integrated” (Postema’s term) reflection may lead to the conclusion that to act from affection is morally best. That is hard to imagine in cases in which serious third-party harm will result, for example, when sheltering your own guilty child will lead to someone else’s innocent child being punished; but short of that extreme case, it is surely conceivable that a parent would attach such importance to maintaining a relationship of deep trust that some requirements of justice would be set aside. But that conclusion would be the parent’s conclusion about the right

37

See Susan Wolf, “Morality and Partiality,” Philosophical Perspectives 6 (1992): 243–59.


thing to do, not a compromising of the right thing to do for the sake of affection.

VI But I have left to the end a consideration that poses a very serious difficulty for the critic of associativism. This is the view that, in theorizing particular obligations as instances of general ones, one in effect theorizes particularity away. The criticism arises in the context of political theory. We are said to have obligations to particular states, those that we were born in or move to. Suppose we justify that by some general moral reason, such as, that states are valuable because they provide order, or that just ones provide justice. That would seem to give us a reason to support all states, or all just ones; perhaps some state other than our own is in more need of support, or much more just than our own – the argument tells us, surely, to direct our support (e.g. our tax payments) to it. Here, the associativist may claim, the basic flaw of moral universalism is exposed. While it can tell me that x is good in all its instances, it cannot tell me that my instantiation of x has a stronger claim on me than others do. An argument about the moral value of political association does not show us “why we have a special relation to the particular polity of which we are members,” a problem that is rightly taken to be “central” to the associativists’ critique.38 Above I have tried to show that while associativists certainly bring out the ways in which we have special relations with polities and other groups, they do not succeed in showing that the relations, independently of wider moral principles, have moral force. But critics of the view seem to have exactly the converse problem. They can appeal to morally compelling principles: it is not clear, however, that they can successfully connect these with special relations. They are thus said to neglect what some, following Godwin, have called “the magic in the

38

Horton, “In Defence” (2007): 10.


pronoun ‘my’.” There is, I believe, a non-magical way to surmount this objection, but it calls for a discussion of its own. Having tried to show that associative ties are not morally (as opposed to phenomenologically) free-standing, the argument now needs an account of why obligations are not free-floating.

Particularizing obligation: the normative role of risk

2

The “particularity requirement” was fi rst given its name by A. John Simmons, though as Simmons rightly points out it has long been present, though unnamed, in the history of political philosophy.1 It is the requirement that a successful theory of political obligation must ground an obligation to comply with and support a citizen’s own state – not an unconditional obligation, of course, but an obligation that is distinct from what one may owe to states in general, or, if our criterion is more selective, all those states that have some important and desirable feature. We can easily see that providing that ground may be problematic for approaches that take off from some cosmopolitan idea such as natural duty: if, after all, it is one’s human nature itself that generates obligations, it will be an uphill battle to explain why there should be any sort of special moral bond arising from the contingencies of membership – and perhaps, according to “philosophical anarchists” (a label adopted by Simmons and others), a losing battle too, for even a successful argument that states were good things to have, from a cosmopolitan standpoint, would not amount to an argument that a person had a reason to support a particular state. (That it was “theirs” would of course be questionbegging, or, to employ Godwin’s term, “magical.”) Other approaches, however, seem on the face of things to have a better chance of meeting the requirement. These are approaches that may be termed “transactional,” in the sense that they draw upon some past event in which the citizen has (or is said to have) participated, thus connecting her with a specific political order. Approaches based on contract, on consent, on the receipt (or acceptance) of benefits, 1

See Simmons, Moral Principles, especially 31–35.


or on gratitude, have been elaborated upon since Plato’s Crito dialog, which, remarkably, employs them all. Plato has Socrates say, in defense of his belief that he has a duty to obey Athenian law, even at the cost of his life, that he has made an agreement with the city and its laws, albeit tacitly, that he has received many benefits (including life) from it, that he has had opportunities to persuade the city to act differently. But many strong criticisms have also been elaborated, and philosophical anarchists remain unimpressed. A third general approach, discussed in the previous chapter, is now generally termed “associative,” and it offers to cut through the whole issue by making the obligations in question inherently particular; some obligations are simply part of what constitutes membership, the view claims. But considering this approach leads us back to the particularity requirement by a slightly less direct route, for, as we saw, the most stubborn part of the associativist case is, in effect, that non-associativists cannot satisfy it (on the grounds that what is inherently particular cannot be derived from something else), so that the associativists win, as it were, by default. Among the most convincing candidates for the “particularizing” role is the argument from reciprocity, or giving a fair return for benefits received, whether from the state itself (or “The Laws” as in the Crito) or from the fellow-citizens with whom one shares state membership. This chapter (sections I–II) critically discusses that approach as a way of grounding obligation, and attempts to offer an alternative view, also of a transactional kind: one based, however, not on the receipt of benefits but on exposure to risk. Philosophical anarchists and associativists are both right, in my view, to claim that the receipt-of-benefits argument fails – we cannot use it to satisfy the particularity requirement, despite sophisticated attempts to shore it up. Setting out what is wrong with the receipt-of-benefits case, however, provides us with a valuable checklist of issues which aids us in evaluating any alternative view. I go on (sections III–V) to set out and defend such a view, and then indicate what kind of obligations it will lead to. They are, I believe, limited and conditional,

Particularizing obligation 41

but the question of the limits to those obligations – in the form of obligations beyond one’s own political society – will be taken up in the following chapters.

I A receipt-of-benefits approach claims that in consequence of deriving advantages from a political order a citizen owes something to other citizens in return, for those advantages depend on the acceptance of costs by others. What we owe is “submission” to the restraints imposed by the arrangements that are of mutual benefit. In his classic formulation of the claim, H. L. A. Hart represented those restraints in terms of a loss of liberty: when a number of persons conduct any joint enterprise according to rules and thus restrict their liberty, those who have submitted to these restrictions when required have a right to a similar submission from those who have benefited from their submission.2

We may understand “loss of liberty” broadly, so as to include any loss of resources (which amounts, after all, to a loss of the liberty to deploy them as one might otherwise wish). That I should agree to suffer a loss of liberty because you have done so in a way that benefits me would seem, initially, to be an unproblematic requirement of fairness. But a problem immediately arises, one that we may term (following Simmons again) the problem of generality. In one form, that problem involves asymmetries of obligation that cannot be captured by the idea of reciprocal fairness. For one thing, as Robert Goodin has noted in his critique of the mutual-benefit model of political society,3 there are active participants (such as resident aliens) who contribute without receiving full benefits; and even if we 2

3

H. L. A. Hart, “Are There Any Natural Rights?” Philosophical Review 64 (1955): 185. Robert Goodin, “What Is So Special about Our Fellow Countrymen? Ethics 98 (1988): 667–68.


remedy that, out of fairness, it is still the case that societies extend their protection to people (such as the handicapped and the elderly) who are unable to provide benefits in return. Surely we approve of their doing so, and surely we approve on the basis that vulnerability deserves protection, whether or not the vulnerable person has done anything for us, our approval thus implying rejection of the mutualbenefit model. That point may be placed, as Goodin does elsewhere, in the context of a broader critique of transactionalist approaches, such as Hart’s, on the grounds that they place too much weight on what one has done in determining what one must do, or what one deserves. In order to do so, Goodin contends, they must stretch ideas such as that of “assuming” obligations, instead of turning directly to ideas of what one owes to others by virtue, simply, of their vulnerability to loss and pain. Taking promising as the paradigm case of special responsibility, the argument, he complains, has to adopt increasingly attenuated simulacra of promises in order to account for the special responsibilities that ordinary morality recognizes.4 But the problem of generality takes another important form too, one that involves the very concept of “receiving benefits,” rather than the scope of who might be said to deserve them. On a undifferentiated notion of “receipt,” the receipt-of-benefits argument would obviously be open to the famous and telling line of objection developed by Robert Nozick, by way of his set-piece example of people “benefiting” from a system of broadcast news and music, backed up by other examples of uninvited lawn-mowing and the spontaneous thrusting of presents (books) on people.5 If someone confers a benefit on you unasked, it is not clear that simply by virtue of that you owe them anything, let alone that you can be rightfully compelled to give them anything. If, for example, some neighborly benefactor takes it upon herself to mow all the front lawns on the block, uninvited, it is very far from clear what (if anything) her neighbors would be obliged 4

5

Robert Goodin, Political Theory and Public Policy (Chicago: Chicago University Press,1982), 30–32. Robert Nozick, Anarchy, State and Utopia (Oxford: Blackwell, 1974), 90–95.


to give her: an expression of thanks might be in order; but a coercively backed demand for payment (which is what the political analogy would require) would be absurd. In order to meet this objection, one may look for ways to show that something additional has come into play, in consequence of which recipience is not merely passive. We may, for example, adopt a distinction between “receiving” and “accepting” benefits.6 The distinction is between (for example) passively benefiting from the background of general security (receiving) and personally calling the police to deal with an emergency (accepting). A voluntary act of that kind – drawing water from a public supply would be another example – creates an obligation in a way that mere enjoyment (receipt) does not. So “accepting” would create an obligation. But of course, not everyone “accepts” in that sense, so the obligation cannot be a general one. It would apply only to some, and so could not be the basis of a political society’s claim to obedience on the part of all its citizens. It has been objected, however, that these conditions are unreasonably stringent, so that the receipt-of-benefits argument looks supportable in at least one important case.7 This is the case of nonexcludable public goods, goods that are “public” ones in the fullest sense: goods such that, when they are supplied for a society, it is impossible to prevent members of the society from benefiting from them. In such a case, Richard Arneson rightly points out, it is not possible to make the proposed distinction between (merely) “receiving” and (positively) “accepting.” It is also reasonable, he argues, for those who pay the cost of providing the goods to require unwilling beneficiaries to contribute too. To make this point, Arneson constructs two hypothetical cases in which the refusal to contribute would clearly be objectionable. In the first case, a contribution-refuser maintains 6

7

Simmons, Moral Principles, 126–27, 130. See also Richard Dagger, Civic Virtues: Rights, Citizenship, and Republican Liberalism (Oxford: Oxford University Press, 1997), 72–74, though Dagger argues that “acceptance” is more common than Simmons allows. Richard Arneson, “The Principle of Fairness and Free-Rider Problems,” Ethics 92 (1982): 616–33.


that the public goods in question fall like manna from heaven, and do not in fact depend on contributions from fellow-citizens. In the second kind of case, a racist believes that the cost should fall disproportionately on a racial group other than his own, or someone maintains against all the evidence that she is called upon to pay much more than her share. Both examples clearly make the point that those who contribute their fair share can rightfully override either “bizarre” beliefs about causality or prejudiced or uninformed beliefs that are plainly not “reasonable.” But while contributors could be expected to reject irrational or immoral grounds for non-contribution, we cannot ignore the fact that there may be reasonable refusals to contribute. One can think of three possible grounds, at least. The fi rst is an objection to the good in question: pacifists, for example, may have reasoned objections to the support of a military establishment, on the grounds that military establishments are a cause of confl ict, and that their use is never a justifiable response to confl ict. The second is an objection to paternalism. Some citizens might believe that some thing, although a good, should not be publicly provided. They might think, for example, that self-reliance is eroded by the public provision of health care or pension benefits, and that the good of a self-reliant society would better be promoted by a scheme that placed responsibility for meeting the needs of ill health or old age upon each individual, or upon families. (According to some “family values” theorists, for example, public provision of benefits to children has the capacity to undermine the intimate bonds on which families rely.8) The third is an objection to priorities. Some citizens may believe that while national defense and education (say) are both goods, and appropriately to be provided by public means, spending on the former is so out of proportion, in relation to the needs of the latter, that contributing a tax dollar to such distorted use is morally unacceptable. Now if the decisions in question – to provide for military protection, 8

Ferdinand Schoeman, “Rights of Children, Rights of Parents, and the Moral Basis of the Family,” Ethics 91 (1980): 6–19.


to provide compulsorily supported social benefits, to spend more on one good or another – are democratically adopted, that would bring a further moral dimension into play: dissenting citizens who have taken part in a democratic process could reasonably be regarded as bound by its outcome, on the grounds that to take part in a process (normally) implies a good-faith commitment to accept the decision that it produces. That would of course take us back to a version of the criterion of “acceptance,” in this case acceptance by political participation. If they had not taken part, though – perhaps because the dissenters had no realistic chance of persuading the majority, say – how would things stand? Certainly it seems true that in the case of pure (non-excludable) public goods the distinction between receipt and acceptance cannot be made. But it does not follow that the moral effects of the distinction do not apply, so that simple receipt generates what voluntary acceptance would generate – that if criteria of acceptance are unavailable, then mere receipt is just as good. Contributors certainly need reasonable grounds for compelling support from non-contributors. But while unreasonable refusals to give support can be discounted, reasonable ones cannot be, even in cases in which the goods supplied are purely public; so the familiar doubts about deriving obligation from receipt still apply, and the reciprocity argument remains open to strong objections.

II A second line of objection may be termed the “question-begging” critique: critique to the effect that the fair-play argument covertly assumes what it claims to establish. Once again, the critique takes two forms. The first is offered in an especially clear way by KokChor Tan and by Ryan Pevnick. Tan points out that while promises are normally binding, a promise to dispose of something that I cannot rightfully dispose of is not – standardly, the case of a promise to commit murder is often cited here; likewise, we should invalidate a supposed commitment to return a benefit in cases in which the


benefactor had no title to confer it, and/or the beneficiary had no title to what was expected in return. So if the fair-return-of-benefit argument is to justify giving preference to compatriots in conferring benefits, it depends on a prior argument that justifies the compatriots’ title to the benefits that they are exchanging.9 Along somewhat similar lines, Pevnick maintains that it is a defect in traditional fair-play arguments (such as Hart’s) that they assume the justice of mutual-exchange schemes to be independent of one another – that the fairness of a scheme can be determined without reference to its impact on other schemes.10 As one may point out in support of both of the arguments, it is always possible to impugn the justice of a scheme, whatever its internal character, on the basis of its external implications. Both Tan and Pevnick are concerned to draw out what this means for questions of global justice. This line of criticism takes us interestingly, but – for the present purpose – unproductively, in a circle. We can (in a procedural sense) fairly distribute among ourselves some good that is not fairly ours – the fable of robbers in their lair dividing their plunder is an ancient one, deployed by Plato, St. Augustine, and Rousseau to make this point. We cannot, however, appeal to fairness in support of the resulting distribution unless we are willing to examine the fairness of the original possession of the good. And then we encounter the larger distributive question that Tan and Pevnick rightly pose. What, though, is the answer to that question? One approach is to ask, as they do, about the requirements of global justice: what nations can justifiably dispose of is contained within the parameters of justifiable global distributions. This line of argument has at least two important merits. First, if it succeeds, it deals neatly with the conflict between local and universal claims – once the just claims of the rest of the world have been met, local associations are free with respect to what is left to them. (The 9

10

Kok-Chor Tan, Justice Without Borders (Cambridge: Cambridge University Press, 2004), 171–72. Ryan Pevnick, “Political Coercion and the Scope of Distributive Justice,” Political Studies 56 (2008): 399–413.


parallel with individuals within a domestic Rawlsian basic structure is clear.) Secondly, the argument is notably tolerant, in relieving local associations of any need to give an account of their own internal distributive practices or of the basis that they posit for their own specialness: they are not called upon to “derive” it from universal premises (a troublesome matter, as we saw in chapter 1).11 Despite these advantages, I am not sure that the strategy is convincing. It depends on the claim that “special obligations are limited by general obligations, not the other way around.”12 “The other way around” would be a view that outsiders get only what (if anything) is left after insiders have divided up what they have, and is plainly unacceptable. But the claim in question is acceptable only if one has decided in advance that the principle of cosmopolitan regard is to operate without reference to the jurisdictional division of the world. Here the ancient fable of judicious thieves in their lair is likely to mislead: for they are dividing among themselves a good that manifestly belongs to someone else, whereas the issue of global distributive justice concerns goods produced by some complex mix of domestic and global inputs, so that what is rightfully whose is a question rather than a background premise of the model. That consideration, I believe, simply tends to neutralize the “fairness” principle for the purposes of either compatriot preference or global redistribution: it cannot stand alone. The second line of critique under the “question-begging” heading is developed by Peter Singer.13 This concerns the circularity of justifying benefits in terms of membership when it is in fact membership that settles the distribution of benefits in the fi rst place. Members of wealthy societies have constructed exchange-of-benefits schemes for themselves, and in general do quite well out of them. They may try to justify the fact that outsiders do not benefit on the grounds that they do not contribute either. But of course, outsiders may wish that they could contribute (thus qualifying to receive corresponding benefits), and it is manifestly unfair to justify their 11 12

Tan, Justice Without Borders, 166. Ibid., 168. 13 Singer, One World, 169–70.


non-receipt of benefits on grounds arising from the fact that we have forcibly prevented them from contributing. Singer casts this point in terms of the situation of refugees who desire admission, a case that makes the point with especial force. It might be met by “lifeboat”type objections, that is objections on the grounds that the very features that make wealthy societies desirable would be undermined, counter-productively, if their borders were to become open – the lifeboat sinks if we try to rescue too many. The case for open borders is not, as it were, open-and-shut, for even cosmopolitans may recognize that some limits to the amount or rate of immigration can be justified. But even so, surely there is a larger point to be made along the same lines. Over and above the numbers of potential refugees from poor countries to rich ones, there are all those inhabitants of poor countries who, whether they want to migrate or not, suffer the effects of exclusiveness that is justified by an appeal to the sovereignty of rich countries. There are, for example, trading arrangements among (relatively) rich countries that allow external protectionism, to the great disadvantage of outsiders, and security arrangements that extend mandatory protections to insiders, arrangements that outsiders would often benefit from more than insiders do. Why should those outside not be let into those arrangements? There would, of course, be significant asymmetries of contribution: poor countries would gain more and contribute less. But we accept that asymmetry in the domestic case, and in the domestic case we would, as noted above, regard the requirement of symmetrical contribution as barbarous. It would deny benefits to the very young, the incapacitated, the unproductive, and the elderly. Singer’s version of the “question-begging” objection, then, should be not only accepted, but broadened in order to take account of the fact that it is not only refugee or immigration policy that confers benefits whose justification is basically circular.14

III A basic distinction between justice and legitimation stands in the way of benefit-based arguments of the kind that have been discussed 14

For a similar view, see Van Parijs, “International Distributive Justice,” 650.


above.15 We assess the justice of a political order in terms of its systemic features, whether the features that interest us concern its net consequences, or a principle of organization that it embodies. We must think of its legitimation, however, in terms of some feature of it that makes its commands binding on certain agents. Judgments about the justice of an order clearly raise the particularity problem, for they render conclusions about all the orders that happen to have the features in question. Judgments about legitimacy are, of course, proposed solutions to the particularity problem, for they necessarily seek to show why it is one order rather than another that can issue binding commands, of an exclusive kind, to particular persons. Now to move directly from justice to legitimation is to commit something like a distributive fallacy, that is, to suppose that if the benefits provided to a whole society are sufficient to render it just, then the benefits rendered to each individual within it are sufficient to render it legitimate. They are not, however, for the context of individual choice is not (generally) the same as the context of overall assessment. If we say, for example, that a political society is just because it is an alternative to violent anarchy, it does not follow that any individual has a reason to obey it – even if they do indeed value order over anarchy – because their own decision is without significant consequence for the society’s survival. We may of course replace such a directly consequentialist approach for one based upon fairness – each should bear her fair share of the restraints required by a just political order, for not to do so is to claim an unjustifiable moral privilege. That is at the heart of Christopher Wellman’s case for political obligation – to refuse to do one’s part is to help oneself to more than one’s fair share of discretion, as he illuminatingly puts it.16 As it stands, of course that case is no less vulnerable than Hart’s was to the particularity requirement – why

15

16

For extended discussion, see A. John Simmons, Justification and Legitimacy: Essays on Rights and Obligations (Cambridge: Cambridge University Press, 2001), especially chapter 7. Christopher H. Wellman and A. John Simmons, Is There a Duty to Obey the Law? (Cambridge: Cambridge University Press, 2005), 42.


should I not choose to make my due contribution (or even more than my due contribution) somewhere else? To particularize the obligation, Wellman introduces the further idea that what is specific about states is that they have the task of coordinating people’s behavior within the specific geographical space of their jurisdiction. They must solve a coordination problem, and they cannot do so if those within their jurisdiction are free to make their contribution where (or in the form) that they wish. Does this resolve the particularity problem? It certainly comes close: but it still seems vulnerable to a renewal of the original line of objection. The argument shows that, among the duties of fairness, there is an inherently local duty arising from the spatial basis of coordination. But does it assume, rather than show, that the duties of fairness are the relevant ones to apply, that is, does it assume from the start that a political society is to be viewed as a fair system of cooperation? If we press the particularity requirement very hard, as Wellman’s critic does,17 we may continue to be unsatisfied, and also perhaps to conclude that a theory that rests so much on fairness may fall out of the set of transactional accounts into another set of views, those based on natural duty. No brief account can hope to settle a debate that is not only ancient but that continues to be conducted with skill on both sides. Theory choice is rarely a matter of knock-down arguments, but a matter, in part, of which questions are judged most important to answer.18 We may regard theories as “research strategies” of the kind defended (in philosophy of science) by Imre Lakatos, strategies that cannot be directly refuted by counter-examples because they prioritize some questions while (temporarily) bracketing (or “quarantining”) others.19 Perhaps, for example, the benefit theory could be sustained if there were enough cases of acceptance? Or if we 17 19

18 Ibid., 186. Ibid., 47–48. See Imre Lakatos, “Falsification and the Methodology of Scientific Research Programmes,” in Criticism and the Growth of Knowledge, ed. Imre Lakatos and Alan Musgrave (Cambridge: Cambridge University Press, 1970), 91–196. Lakatos goes so far as to quote d’Alembert: “Allez en avant et la foi vous viendra” (142).


weakened the criteria for “acceptance” in ways that did not seem implausible?20 It is also quite proper to give weight to a candidate theory’s normative implications, and to persevere in its defense because, among other reasons, it foregrounds some attractive political features – the more people think of themselves as participants in a cooperative enterprise, “the more likely we are to find civic virtue in the form of a willingness to cooperate for the common good.”21 But surely it is clear that the benefits-received approach is open to recurrent challenges; and if some of its merits can be retained by another approach that is less exposed to critique, an alternative response to the particularity problem should be explored. A better alternative, I shall argue, draws upon the (transactional) fact of complicity. To understand this, we must turn from the topic of benefits received to that of costs imposed. As a member of a political order that exercises coercive power over its members, one lends support, passively and/or actively, to a system of arrangements that imposes costs on others. It is important for the present argument to note that these costs extend far beyond what we may term the constraints of justice, important though they are: to rest the argument on those would be to revert to an argument about fair reciprocity, and thus to lose track of particularity once again. Rather, the costs in question are those that arise from involvement in a system of collective decision-making, both social and public, that unpredictably imposes burdens, of varying severity. States – in fact, institutions quite generally, as chapter 3 will argue – derive their justification from the reduction in vulnerability that they bring to their members. Very few dangers can be successfully confronted by the entirely unassociated person, and political association derives its rationale from its promise to ward off dangers of a particularly basic kind. While it will not ward off the dangers of loneliness, for example – a danger that other associations (such as 20 21

I owe these two suggestions to John Horton. Dagger, Civic Virtues, 79.


dating services) promise to take care of – it does undertake to ward off threats to physical security from outsiders, coercion and intimidation by insiders, and loss of the means of subsistence. It may do more – one may want it to do much more – but it at least promises that. But the idea to be developed here is that in diminishing vulnerability in some respects, states increase it in others. Thomas Nagel, in developing his own approach to the topic of local preference, remarks: “once the state exists, we are in a new moral situation.”22 He means that what we gain from states (“coordinated conduct”23) creates new ties. But it may be that what we lose, rather, helps more to advance the question at issue in this discussion. For “once the state exists” we become subjected to several kinds of serious risk that we would not face in an imaginary condition of statelessness. We trade the risks of isolation for the risks of association. Let us start with the most dramatic of these. Over the past half-century, since the Nuremberg Trials, international theory has evolved – with painful deliberateness – the idea of crime against humanity. At the core of that idea is the double recognition that states must be held to account for what they do to their subjects, and that what states do to their subjects can rise to a level of unparalleled hideousness. The importance of this is underscored by the familiar but still chilling twentieth-century statistic about the numbers of people killed by their own states rather than by other states. States can accomplish this massive slaughter because they command administrative resources, have the authority to mobilize mass action, and enjoy territorial control that is sanctioned by international norms of sovereignty. Those are, of course, features of the state that are valued because of their role in protecting populations, and there is therefore a special perversity in their use as instruments of oppression. David Luban has written eloquently of this: “The legal category of ‘crimes against humanity’ recognizes the special danger 22

23

Thomas Nagel, “The Problem of Global Justice,” Philosophy and Public Affairs 33 (2005): 133. Ibid., 115.


that governments, which are supposed to protect the people who live in their territory, will instead murder them, enslave them, and persecute them, transforming their homeland from a haven into a killing field.”24 This is, then, a particularly vivid example of the double face of institutions, the simultaneous reduction and magnification of vulnerability that they bring about. We have “no alternative to living in politically organized communities,” as Luban writes, but “we are all, in effect, hostage to politics.”25 But this extreme case (to be discussed in more detail in chapter 6) is far from being the only one, or even the most salient. Normal, non-genocidal, states profoundly affect their members’ capacity to protect themselves. After all that has been written about the global political economy, there is no need to emphasize that external economic events have at least as deep an impact on people as their own governments’ policies, or that – as stressed by cosmopolitans in recent debates – there are elaborate and powerful institutions beyond the state, institutions that are state-like in many ways. The view defended here is very far from claiming that states are uniquely a source of risk, as opposed to sources of unique risk. For governments direct, select from, and can suppress political reactions to the effects of the global economy, in a way that global institutions cannot. The deliberate application of power to individuals is one of the things that is said to make states different from global institutions. As we shall see below, some have doubts about drawing a bright line between state and non-state power in light of differences of this kind. But this feature of states is essentially related to the point at issue here. The capacity of states to legitimize or delegitimize, tolerate or prohibit people’s responses to change is something that importantly 24

25

David Luban, “A Theory of Crimes Against Humanity,” Yale Journal of International Law 29 (2004): 117. For a similar view, see Richard Vernon, “What Is Crime Against Humanity?” Journal of Political Philosophy 10 (2002): 231–49. Luban, “A Theory of Crimes Against Humanity,” 138. It should be noted that the argument from state abuse is mentioned, though not developed, by Kwame Appiah in “Cosmopolitan Patriots,” in For Love of Country, ed. Cohen, 28: “States matter because they are both necessary to so many human purposes and because they have so great a potential for abuse”.


mediates between people’s lives and the global political economy, and bears crucially on their capacity to protect their interests from damage. That my state may increase my level of protection in general, in relation to some baseline, does not mean that in any particular case it will protect me from severe loss, if in its judgment my loss is demanded by a larger interest that cancels the legitimacy of defending my own. At an even more basic level, there is the issue of people’s capacity to protect themselves against each other. The collectivization of security is fundamental to what states are. We can consult an extensive social contract tradition for the view that we should think of state formation as essentially an event by which we surrender selfprotection for common protection. Or, if we prefer political sociology as a source, we can turn to Weber for the view that a state is nothing if it does not monopolize the means of violence. Or, we can consult instead Rousseau’s Discourse on Inequality for a classic statement of the heightened vulnerability arising from political society, and the progressive loss of independent capacity as labor is divided, as specialized institutions form, and as technology colonizes social life. Or we could consult J. S. Mill’s perhaps less well known essay “On Civilization,” a powerful account of the loss of personal defensive capacity. The “savage,” Mill says, “shifts for himself,” each relying on “his own strength and cunning,” whereas in modern society “the bulk of the community. . . rely for their security mainly upon social arrangements”; he continues, “the safety of all, constantly depend[s] upon the vigilant performance by each, of the part peculiarly allotted to him in the common task.”26 Or indeed we need only consult our own personal experience to see that there is, on the one hand, a sort of juridical surrender of self-sufficiency, magnified, on the other hand, by processes which constantly increase our reliance on the 26

John Stuart Mill, “On Civilization,” in Collected Works of John Stuart Mill: Essays on Politics and Society, vol. XVIII, part I ed. J.M. Robson (Toronto: University of Toronto Press, 1977), 120, 123. More recently, the connection between risk and modernity has been explored by Ulrich Beck: Risk Society: Towards a New Modernity (London: Sage, 1992).


predictability and restraint of others. Given all this, modern societies rely to a previously unheard-of extent on states as the definers and guarantors of personal security, by virtue of their role as regulators of patterns of expected behavior. Consider, as an example, the simple case of automobile traffic. Millions of our compatriots have in their hands the means to obliterate us in an instant. They have these means at their disposal because the state licenses them to operate them and creates a huge infrastructure of roadways permitting their operation. We survive this only by virtue of some uneasy mix of personal self-restraint and state surveillance.27 Fourthly, there is the important consideration that Goodin notes, in qualifying the received idea that we treat our compatriots better than we treat outsiders.28 Of course we do, in many ways but also in many ways, we treat them worse, or are less respectful of their freedom. Referring to domestic public law and public international law, Goodin points out that (for example) fellow-citizens are liable, as outsiders are not, to taxation and expropriation, conscription, the consequences of river diversions, toxic environmental effects, the legal limitation of commercial damages, and so on. Being a citizen, then, exposes one to a much closer set of controls, and much more far-reaching demands, than any state can legally impose upon citizens of other states. Finally, there is another matter, connected inseparably with the natural consequences of majority–minority difference. As members of organized societies, we have only limited opportunities to escape from the preferences of majorities. In part, of course, this is because legislation is substantially responsive to majority views, and minorities may have to live with officially sanctioned provisions that conflict with their own conceptions of justice. That can be avoided, and 27

28

For a full account, see Tom Vanderbilt, Traffic: Why We Drive the Way We Do (and What It Says About Us) (New York: Knopf, 2008). Vanderbilt notes that “The road . . . is a place where many millions of us, with only loose parameters for how to behave, are thrown together daily in a kind of massive petri dish in which all kinds of uncharted, little-understood dynamics are at work” (6). Goodin, “What Is So Special about Our Fellow Countrymen?” 667–71.


even then only partially, only by winning an “insular minority” status on the basis of history and faith – such as some Old Order Mennonite communities in North America enjoy – and thus virtually detaching oneself from the larger national society. But there is also what J. S. Mill terms “opinion,” or what T. M. Scanlon terms “informal politics,” that is, the inexorable effect of others’ decisions on the social context of one’s own life.29 Scanlon offers two examples: standards of dress, and the public expression of religion. Without any necessary recourse to legislation – though that may happen too – others’ tastes may result in a society in which there is more bodily exposure than some find it comfortable to view, or, alternatively, in puritanical standards that some find stifling. Likewise, others’ religious tastes may result in a proliferation of public symbols and ceremonies that offends people who want a public space that is secular and religiously neutral, or, alternatively, disapproval of religious expression may create a climate in which some communities suffer loss in terms of expression that they deeply value. So, as a member of an organized political society, one is constantly vulnerable to the possibility of acute discomfort, and painful demands may be made on one’s sense of tolerance. Those with whom we live, by virtue of both majoritarian political weight and their preferred use of public space, constrain the kind of life that it is possible for us to lead. Those are, then, in outline, some of the major ways in which national states – like all institutions – enhance our vulnerability in some respects in the course of diminishing it in others. I have suggested that it is their enhancement of our vulnerability that explains our special duties to those with whom we share membership in them. But that thought, to become clearer, needs to be distinguished, fi rst of all, from two related but, in my view, less promising ones.

IV The idea of building on the negative features of political society, rather than the positive features, is not wholly new. In particular, Michael 29

Thomas Scanlon, “The Difficulty of Tolerance,” in Toleration: An Elusive Virtue, ed. David Heyd (Princeton: Princeton University Press, 1996), 226–39.


Blake has constructed a provocative argument that takes as its starting-point the fact of coercion, and seeks to show that, because compatriots are part of a coercive system of relations, they “share more than common humanity.”30 Common humanity directs us to do what we can to ensure that all human beings, whether we are related to them in some way or not, reach a certain threshold of security and well-being – at least freedom from famine, extreme poverty, and “crippling social norms,” perhaps more. But it does not provide grounds for justice in distribution. Those grounds are provided only within the borders of organized national societies. That is because, within those societies, citizens impose on one another laws and policies that are backed by the use or threat of coercive force. Coercive force is a deliberate reduction in the autonomy of those to whom it is applied; and so its use needs to be accompanied by concern for the justice of the distributive shares that result. So we are to have special concern for our fellow-citizens, not because we care for outsiders less, but because outsiders do not come within the range of the special justifications that must accompany coercion.31 Here, then, we have an example of a type of argument that does not distinguish between insiders and outsiders on the basis of the receipt of benefits, but on the basis of exposure to constraints, that is, constraints of a coercive kind. I think one line of objection to the central logic of this argument, i.e. that reducing people’s autonomy calls for special concern, overlooks an important and relevant matter. Critics urge, first, that it is odd to think that the application of coercion creates some kind of compensatory requirement (even odder to think that we can compensate people for losing autonomy in one respect by giving them more concern in another), and, secondly, that just laws do not reduce autonomy – they may in fact extend it, as Locke noted in his metaphor of the fences that prevent us from falling over cliffs.32 The fi rst 30

31 32

Michael Blake, “Distributive Justice, State Coercion, and Autonomy,” Philosophy and Public Affairs 30 (2001): 260. Ibid., 264–65. See Richard Arneson, “Do Patriotic Ties Limit Global Justice Duties?” Journal of Ethics 9 (2005): 127–50, and Pevnick, “Political Coercion and the Scope of Distributive Justice,”: 399–413.


objection is supported by an example in which I justifiably use coercion (against someone who threatens me), and so, like the second objection, it implies an uncontroversially available view of justice. At the heart of politics, however, there are disagreements about what is just, and by virtue of whatever is the decision procedure some will have to live with decisions whose justice they contest. This is not to embrace any form of skepticism, but simply to point out that political theory must take serious account of the fact of disagreement, and the consequent fact that, in view of the need for collective decisions, we run the risk of having to live with laws and policies that we regard as unjust. Of course, there is procedural justice, which, let us assume, is consensually supported: but that still means that some people (the losers) will be compelled to swallow their own views of what is substantively just, and that is surely enough to sustain Blake’s argument, for they suffer what is in their own view (which is what counts here) a coercive defeat, even if they grant it procedural legitimacy. But another objection suggests the need for an alternative to the coercion-based approach. It arises, once again, from poorness of fit. Suppose coercion does give rise to special concern for the coerced: there are, beyond the borders of political society, relations of world-wide coercion, embedded in practices and institutions, that go beyond state borders. The facts are familiar: there are schemes of arrangements that impose themselves upon states, and which they could escape only at the extravagant cost of autarchic isolation. Beyond the facts, there may be questions about whether the kind of coercion involved is distinct in some significant way from state coercion: but all the questions seem answerable in the negative. State coercion is “deliberate,” Blake says, in its imposition of pain, and in that respect differs from policies that as a matter of fact cause pain, thus needing a special justification.33 For some purposes that distinction is important, but it does not mean that 33

Blake, “Distributive Justice, State Coercion, and Autonomy,” 293.


outcomes imposed on other societies, although not intended to cause pain, do not diminish autonomy to a similar or greater degree, when viewed from the receiving end. The subjects of states are required to “engage” in state action by their obedience – but so are debtor states required to “engage” by accepting WTO terms for essential loans, for example.34 The participation of developing countries in global trade and fi nance arrangements might better be described as “nonvoluntary” than as coerced, since there may be no direct threat of sanction, but non-voluntariness, Andrea Sangiovanni argues, has all the moral consequences of coercion anyway, for what is objectionable about coercion is that it takes away voluntary agency.35 Finally, even if we could somehow conclude that state coercion is relevantly different from the global imposition of constraints, why should we be content with a simple binary choice? As Cohen and Sabel argue, if states bring new kinds of more-than-humanitarian relations into being, why should other forms of association – such as global trade and fi nance arrangements – not bring other more-than-humanitarian relations into being, even if they differ from, and perhaps fall short of, the relations that accompany full citizenship?36 The fact of coercion, in short, generates boundary problems that make it unsuited to the task of distinguishing insiders from outsiders, and thus to the task of limiting the scope of justice.37

V The meaning and point of this chapter’s proposal may become clearer by way of a second important contrast – a contrast with a more familiar “Kantian” idea about the politics of risk. According to Kant and 34

35

36 37

For this objection, see Joshua Cohen and Charles Sabel, “Extra Rempublicam Nulla Justitia?” Philosophy and Public Affairs 34 (2006): 167. See also Tan’s critique, Justice Without Borders, 173–77. Andrea Sangiovanni,“Global Justice, Reciprocity, and the State” Philosophy and Public Affairs 35 (2007): 10–14. Cohen and Sabel, “Extra Rempublicam Nulla Justitia?” 149–53. The contrary view is defended by Mathias Risse, “What to Say About the State?” Social Theory and Practice 32 (2006): 671–98. For a critique of Risse’s and Blake’s view, see Simon Caney, “Global Distributive Justice and the State,” Political Studies 56 (2008): 487–518.


those who follow him here – such as Jeremy Waldron – people in an unregulated state of nature would be especially likely to harm those to whom they were physically close and with whom, consequently, they interacted. The likelihood of damage, coupled with a negative duty not to harm, would give rise to an obligation to create an institution that would protect others’ rights, and an obligation to support it when it came into existence. On this argument, as Waldron adds, it would also follow that outsiders would have an obligation to respect the operation of the institution and the secondary rules that it would have to generate in order to fulfill its rights-protecting task.38 And so, if this line of argument succeeded, we would be led from a natural duty to respect others’ rights to a duty to contract to protect them, and flowing from that a general obligation to accept and respect the arrangements of a world in which the duty to protect had been parcelled out on a local basis, thanks to the causal connection between proximity and the risk of damage. Natural duty would have been successfully particularized. But several obstacles stand in the way of the argument’s success. Political societies were never in fact “sorted” out from each other on the basis of degrees of threat of damage. Might we not even expect, as Simmons also says, that people who lived in proximity would be likely to have formed social bonds, and so be less likely to damage each other than they would be to damage more distant others? And since states in reality do more than protect citizens from each other, and make claims on the basis of these further functions, the argument, even if it survived objection, would license only very limited degrees of obedience.39 These objections are, I believe, very telling. But clearly they do not tell against the proposal made here, a proposal that rests not on the antecedent risks that motivate the formation of states, but on the subsequent risks. Like the Kantian argument, the proposal takes as its point of departure the vulnerability of others to the acts that

38

39

Jeremy Waldron, “Special Ties and Natural Duties,” Philosophy and Public Affairs 22 (1993): 22–50. Wellman and Simmons, Is There a Duty? 172–75.


we undertake in exercising our freedom – a point of departure that enjoys enormous intuitive force. Unlike the Kantian argument, it is not open to objections such as those immediately above. Let us consider them one by one. (1) It is true that it is quite implausible to think that political societies were ever “sorted” on the basis of risk reduction among citizens, and also true that if our exclusive goal were to prevent one person from harming another we could come up with expedients other than states. Once they are in existence, though, political societies bring distinctively new risks into being, of the kind listed above. They arise from the bare factuality of states – coupled, again, with the moral importance of vulnerability – and do not depend on the rationality of states’ boundaries or indeed the rationality of the institution of the state itself. (2) The objection derives much force from the us–them phenomenon, dispiritingly well documented by historians and psychologists, a phenomenon that could well incline us to believe that the risk of violence against out-groups was more serious than the risk of violence against neighbors. If so, then we might come to believe, with Rousseau – or with Rousseau in one of his personae – that states are a bad bargain, since the foreign deaths that they cause outnumber the domestic deaths that they prevent. Whatever the truth of that, however, we must set against the us–them phenomenon the no less dispiriting phenomenon of intimate violence. Neighbors do kill each other.40 Short of that, they damage and impede each other in many ways, and their capacity to do so is enhanced by the close ties of interdependence that political societies foster. On this basis, the proposal claims, we owe a special allocation of attention and resources to domestic risk reduction. (3) To conceive of states as instruments of rights protection only is to adopt a narrow and unusual view of them, to be sure, and if one’s duty to support them went only as far as to support their police function it would go only a fraction of the way towards legitimating the degree of support that states actually enforce. But the proposal made here does not rest upon, though it embraces, the antecedent risks posed 40

See Esses and Vernon (eds.), Explaining the Breakdown of Ethnic Relations.


by one citizen to another. It also embraces the risks subsequently posed by the intense interaction that political society encourages. It also embraces the risks posed by social and political majorities to minorities, obviously a political and not a natural phenomenon. It also embraces the risks posed by the state apparatus itself. Starting, then, from an important premise concerning vulnerability, we may arrive at conclusions about our local obligations, if we focus on the subsequent risks that states intensify even as (in good cases) they diminish antecedent ones. These risks are systemic, general, and documentable. This feature of them enables us to ward off a further objection that the Kantian proposal invites, namely, an objection based on the subjectivity of fear. Being a threat, as Simmons notes, is an objective notion.41 Feeling threatened is another matter – some may feel threatened by the robin on the lawn. Like the robin on the lawn, some proximate humans – in the Kantian state-of-nature scenario – may actually be harmless; so the anticipation of threat, on which the whole construct rests, may be irrational. So, given that the project to construct a political society is a prospective one, the robin-fearers will have an irrational degree of weight; and this irrationality will pass over to the constituted political society, to the extent that its legitimation rests on the thought that every citizen is a threat to every other, so that any citizen’s disobedience to law puts others in jeopardy. That thought is not, of course, true or plausible. But the proposal made here is immune to considerations drawn from neurotic robin-fearing. It rests on an account of the ways in which one citizen is exposed to risks posed by all other citizens in their social and political behavior, and by the collective institutions established to regulate their common life. The idea of common risk exposure, I believe, particularizes obligation while escaping the compelling objections that have been made to attempts to ground obligation on the receipt of benefits; or on the moral effects of coercion; or on the avoidance of antecedent risk. 41

Simmons, Moral Principles, 175.


VI The basic moral consideration in this chapter is this: if one is part of a set of arrangements that enhances the vulnerability of others to damage, then one has a special reason to adopt measures to avert that damage. The overall benefits of a political society make it just; the risks that it imposes make its special demands legitimate. Some familiar examples that suggest initial support for that intuition are drawn from the standard rock-climbing and lifeboat-rescue scenarios. Suppose you share a rope with other rock-climbers: you would have a stronger reason to secure it than you would to secure the ropes of other groups. Suppose you have scrambled aboard a lifeboat that is leaking: you have a stronger reason to help bale it out than you have to bale out other boats. But in both of those examples one’s own selfinterest may also be at stake, thus inviting the sort of objections that mutual-benefit models attract, so to clear that away let us try two others. You have given a party at which alcohol has flowed freely, and some of your guests seem to be planning to drive home: you have a stronger reason to dissuade them than your (uninvited) neighbor does. Or, you teach a course in which you have set some requirement, such as a grade for participation, which, you later come to see, will tend to penalize some highly meritorious students, conscientious and intelligent introverts: you have a good reason to do what you can to moderate the effect of the grade, even though your educational institution has no good general reason to favor that particular character type over others. I do not think anyone will have trouble thinking of other examples of cases for the principle adopted here. We simply owe more for the consequences of something when we are part of it than when we are not. But the extension of this idea to the case of compatriot preference will be less intuitively certain, for two different kinds of reason. First, it may seem unsatisfactory to rest the case for legitimacy on the empirical and hence variable features of states. Simmons objects to a view of local obligation that would “make . . . bonds vary in strength and direction as international political and legal affairs


changed.”42 But I do not see how a risk-based view of legitimacy can avoid doing this, for the role of states in imposing risks, and their relative importance as risk-imposers, must inevitably vary from state to state and from time to time. This is a topic to be taken up in the fi nal chapter of this book. But on the other hand, I do not see that this must be a source of weakness in a political theory, unless one supposes that political theory must deal with timeless objects. At the limit, after all, there could be something called a “state” that had no role at all as either a distributor or a mitigator of risk, its members’ life-chances being entirely determined by external economic and political forces. It seems an acceptable conclusion that in that limiting case, no special ties would be generated by civic membership, on the argument adopted here. There would be special ties of kinship and friendship, of course, and also humanitarian ties intensified by proximity, but it even seems odd to think that specifically political ties remained in an association that had, after all, lost all the features that are specific to a polity. Secondly, though, there is a more difficult conceptual question. In what sense is one part of a political order, by virtue of citizenship within it? Obviously it would be question-begging to appeal to the fact of formal citizenship, a response that would merely resurrect the particularity problem – what is the moral content of formal citizenship? The argument of this chapter rests on the claim that one is liable, as a citizen, not simply because co-citizens experience risk – a fact that does not distinguish them from citizens of other states – but that one is complicit in the general arrangements that impose them. Each citizen’s compliance supports a set of arrangements that exposes others to risk, I have argued – a consideration that withstands the important objections to benefit-based arguments for legitimacy. But what does “complicity” mean? The following chapter attempts to answer that question, by framing it in the context of a contractualist line of argument. 42

Wellman and Simmons, Is There a Duty?, 167

3

The social waiver

While adopting the distinction between justification and legitimacy noted in the previous chapter, the approach developed in this book nevertheless offers a single model for explaining both. That model – termed here “the social waiver” – sets out the basic justifying conditions for a good political society, and also demonstrates why a political society that meets those conditions can legitimately bind its citizens (in ways that it cannot bind outsiders). As we have seen, the fact that it meets those conditions could not be the (sole) reason for its legitimacy – it is only necessary, not sufficient; but, I shall argue, the same considerations underlie both the justifying conditions and the legitimating reasons. That will be the task of the fi rst part of this chapter. The second half, extending the notion of the social waiver, returns to the issue of moral dualism discussed in chapter 1. There, it was argued only that moral dualism is not necessary to make sense of the two kinds of obligation that people have, partial and universal; this chapter sets out to show how the two kinds of obligation are related.

I In his Second Treatise Locke gives the impression, at several points, that he envisages the social contract as an essentially hypothetical device. He says, for example, that to suppose that absolute monarchy is a justifiable response to the security problem is to imply an absurd “as if” proposition: “that men are so foolish, that they take care to avoid what mischiefs may be done them by pole-cats, or foxes; but are content, nay, think it safety, to be devoured by lions” (s. 93). Similarly, he argues, in discussing the contractual basis of political society, that “no rational creature can be supposed to change his


condition with an intention to be worse” (s. 131, emphasis added). Absolute power is rejected because “it cannot be supposed that [people] should intend, had they a power to do so … to put a force into the magistrate’s hand to execute his unlimited will arbitrarily upon them” (s. 137, emphasis added). Some interpreters have argued that such passages give us the best general interpretation of Locke’s contractualism.1 Of course, if we take that view, we deprive Locke of any argument for political obligation (as distinct from an argument giving us a reason to obey and support political orders that meet the terms suggested by hypothetical consent), for only actual, not hypothetical, consent can actually bind people. As a recent commentator on Locke has said, “in arguments using hypothetical consent, the role of … the reasons for why it would be reasonable to consent are really doing all the justificatory work, leaving the agreement part virtually inert.”2 Moreover, Locke’s effort to demonstrate actual agreement is generally agreed to be an abject failure. Ever since David Hume’s devastating objection (in “Of the Original Contract”) to the idea of “tacit consent,” commentators have found nothing good to say about that proposal of Locke – it strains the idea of actual consent to the breaking point, for while there is certainly something that we could call “tacit consent” – “I’ll go ahead with that unless I hear from you, OK?” – equally certainly it cannot be said to be given in the circumstances specified by Locke. In this chapter and the next, I shall advance a view that, I believe, entitles a (broadly) “Lockean” view to a claim about political obligation. For the moment, though, setting aside the question of whether Locke himself has a theory of obligation, and also setting aside the question of whether the idea of a hypothetical contract 1

2

For a classic statement, see Hannah Pitkin, “Obligation and Consent,” in Philosophy, Politics and Society, 4th series, ed. Peter Laslett et al. (Oxford: Blackwell, 1972), 45–85; more recently, Lena Halldenius, “Locke and the Non-Arbitrary,” European Journal of Political Theory 2 (2003): 261–79. References to Locke’s Second Treatise of Government (Indianapolis: Hackett, 1980) above are given in parenthesis in the text, by paragraph number. Halldenius, “Locke,” 273.

The social waiver 67

gives us the best interpretation of the treatment of consent in the Second Treatise, I want to focus on what that idea can contribute to the project at hand. First, it leads us to a distinction that is sometimes marked by the different uses of the terms contractarian and contractualist. Contractarians believe that morality itself can be generated out of the agreements made by amoral (self-interested) contractors. Contractualists, however, offer a political view that takes some already given moral view as its background.3 We may illustrate this point in two ways, one historical, one conceptual. Historically, contractualists such as Locke drew explicitly upon some view of natural law, that is, the doctrine that there are moral requirements that pre-exist any “conventions” or agreements that people may make, and, further, that such conventions or agreements are valid only to the extent that they operate within, and operationalize, basic moral requirements. Conceptually, as Hart pointed out in his famous article, the very idea of a binding agreement presupposes a background right (whether this is cast in terms of the framework of natural law or not). To agree to something is to give up a freedom to do otherwise, and that would literally make no sense unless one had a right to the initial freedom in the first place. If that is so, then we are committed to the idea that a political society presupposes some more basic moral relations among people, such that each has some regard for the agency of others, the very idea of right or entitlement thus being conceivable. The argument here is minimal in its assumptions, which require us to accept only a need for justification. The contractual device is understood here simply as a way of setting out what it means for citizens to explain to one another the mutual claims that they make. This of course contradicts views that present the social 3

I do not mean to imply that only contractualists do this. The point is sometimes made about states in general: they are “always (potentially) informed by claims about humanity in general rather than just … citizens in particular” (Walker, “Citizenship After the Modern Subject,” 173, and see also 198).


contract itself as the source of all rights, but surely the very idea of justification requires us to reject such views. Minimal though the right to receive justification may be, as a condition of accepting a legal or political constraint, it could not be possessed by literally rightless beings, and so at least that minimal right is presupposed by the idea of contract itself. It is in this sense that a contractualist argument, while leading us to recognize the special nature of contractually justified civic ties, must lead us at the same time to a notion of precontractual moral regard. We may then think of a political society as constructed by a waiver, in a purely notional sense, of background entitlements.4 The second point is that notional waivers must be subject to a reasonableness test, their justifiability depending on the value of the arrangements that they generate. They would have to be arrangements that could plausibly be claimed to enhance people’s agency overall and to reduce their vulnerability to serious obstacles. That is, of course, a very broad provision, which, within the space created by a notional waiver, leaves much room for actual consent in the construction of institutions. The third point, however, is that properly various as the institutions may be, Locke’s polecat-and-lion problem applies: they cannot be just if they effectively lead to an increase in people’s vulnerability to coercion or the threat of coercion. The polecat-and-lion problem is simply the most direct and vivid example of a general point about institutions: they increase vulnerability in the course of reducing it, so that their justification is ultimately an on-balance matter. Constitutions, of course, clearly acknowledge the ambivalent potential of institutions, in that in part they contain provisions intended to guarantee personal security, while in part they 4

Referring to Goodin’s view (see chapter 2, note 28) that compatriots can override each other’s negative rights, Pogge (World Poverty and Human Rights [Cambridge: Polity, 2002], 131) maintains that this is legitimate only on the basis of a “waiving” of them via a democratic procedure. This seems too demanding: surely there are other background conditions that allow “waiving” to be inferred?


also contain provisions intended to restrain and manage the use of power.5 Some provisions, such as political representation, may be thought of as doing both. All this sounds very Lockean, of course – as indeed it is, up to a point. There is no need to conflate the constitutionalist spirit of the Second Treatise with the (often parochial) letter of its recommendations, for the constitutionalist spirit may be embodied in institutions that differ from the European model – a point to be taken up again in chapter 7 below. Also, there may be quite insufficient provisions in Locke’s model for protecting minority rights against majorities, to which Locke sometimes seems to ascribe something resembling sovereignty – albeit within the constraints of natural law, though majorities (under God) also seem to be the fi nal arbiter of what natural law requires. Finally, and perhaps most significantly, the protection offered by political society should not be narrowly defi ned in terms of physical damage or property loss, nor should the risk posed by political society be imagined wholly in terms of tyranny; everything that one wants to say about those topics can be repeated in the context of economic life, unless one arbitrarily restricts the idea of coercion. What modern-day liberals would want to add to Locke is, above all, a much clearer sense of the ways in which socioeconomic inequality generates imbalances of power that impede people’s agency no less than crime or tyranny do. They would want to enhance Locke’s hypothetical contract in light of the equally fictitious but more strongly egalitarian version offered by Rousseau’s Discourse on Inequality. If we look at the society around us, Rousseau complained, it could have been created only by a contract in which one party (the rich) hoodwinked another (the poor).6 Again, the assumption is that no one can be supposed to have adopted an arrangement 5

6

“An interpretation of government as an expression of trust, where representatives have discretion over the lives of many people, must be accompanied by a rigorous system of checks and balances.” Patti Lenard, “Trust Your Compatriots, But Count Your Change: The Roles of Trust, Mistrust and Distrust in Democracy,” Political Studies 56 (2008): 312–32. Jean-Jacques Rousseau, Basic Political Writings (Indianapolis: Hackett, 1987), 69.


that is systematically to their disadvantage: if we suppose that the poor consented to their present condition, we must suppose them to have been deceived. They would have had to have been duped by the proposition that political society would guarantee everyone’s holdings, to everyone’s benefit, forgetting that this guarantee would be much more to the benefit of some than to others. In Rousseau’s clearly implied view, then, a set of social and political arrangements would be morally sustainable only if it could be justified without deceit – which is another way of saying “reasonably” – to those who benefit least from them. In this regard, Rousseau figures as the conceptual link between the proto-liberalism of Locke and the egalitarian liberalism of Rawls’ difference principle: arrangements must be justifiable to those who profit least. In very general terms, then, political societies will be justifiable to the extent that they offer basic protection to individuals, make power revocable or otherwise anticipate its abuse, shield minorities from the full force of majority will, and adopt social and economic policies that prevent domination by those who hold economic power. We can derive that much, at least, from the idea that a political society embodies a (notional) reasonable waiver, for without those conditions it cannot be conceived of as reasonable. But constitutions, or quasiconstitutional settlements expressing basic ideas of social justice, go only some way towards inhibiting the damage that political societies potentially cause. To explore this point, let us revert to the short list of “risks” outlined in the previous chapter. These risks arise, in large part, from the distribution of costs that accompany social and economic policies. Beyond whatever protection a constitution may provide, a fair political system – one that could be the subject of a notional waiver of freedom – could not be a winner-take-all system, but would have to include in its estimates of justice the incidence of burdens. A policy of less-than-full employment is unjustifiable if it is not accompanied by provision for the unemployed. A policy of exemption from a military draft is unjustifiable if its burdens fall on the relatively deprived. A sharply inflationary policy is unjustifiable if pensions are not indexed. Examples could be multiplied indefinitely,


but the general point is that, in a system that passes the “social waiver” test, the stakes of all citizens must be taken account of. In an illuminating discussion of “risk,” Wolff and de-Shalit connect the concept with the plurality of human wants.7 People’s capabilities are best thought of in terms of possible “sets,” that is, the resources that they have may be put to various possible combinations of purposes, though not to realizing every purpose that they have. Allocating resources to certain purposes, and thus sacrificing others, is part of the ordinary exercise of freedom. But sometimes, circumstances are such that in order to sustain one important purpose people have no option but to place some other highly valued purpose in jeopardy. To give two graphic examples cited by Wolff and de-Shalit: those who subsist by gathering honey in Bengal run the risk of attack by the Royal Bengal tiger; Bedouins in the Negev district of Israel accept health risks and traffic fatalities for the sake of maintaining their homes.8 Facing jeopardy of this kind is an important source of disadvantage in relation to others, and hence an important topic for social justice. We must take account of people’s capacity not just to adopt some purpose but to sustain it without making other vital purposes insecure. A “central way of being disadvantaged” is “when, in order to secure certain functioning, one is forced to make other functionings insecure, in a way that other people do not have to do.”9 This way of conceiving of disadvantage, with its emphasis on sustaining a capacity into the future, gives a central place to the temporal notion of risk, the prospects of sustaining or losing a capacity over time. In a political context, it focuses our attention on the ways in which political societies distribute insecurities, often in the course of providing securities, or of providing securities for some at the expense of placing others in jeopardy. In fact, we may see the idea of a social waiver as the paradigm case. The waiving of background rights initiates what is at once an enhancement of risk and a risk reduction project, the justice of which must be the subject of ongoing evaluation. 7

8

Jonathan Wolff and Avner de-Shalit, Disadvantage (Oxford: Oxford University Press, 2007), chapter 3. Ibid., 65, 67. 9 Ibid., 72.


Some of the risks arising from the social waiver are what we may term “background risks,” that is, risks that are predictably part of social and political arrangements. Any system of power is open to abuse. Any system conferring power on majorities needs to protect basic individual rights, and take account of the way in which minorities that can never become majorities are systematically endangered. Any attempt to institutionalize political equality will be undermined unless provisions are made to constrain the degrees and effects of economic inequality. Considerations such as this are important to good constitutions (or, to avoid legal formalism, to good quasi-constitutional settlements that constrain legislation and policy with equivalent effectiveness). They are the considerations that govern a political society’s justification. But there are also what may be termed “dynamic risks,” that is, risks emerging from a society’s political and socioeconomic development that fall through the constitutional net sketched in the paragraph above. Being a citizen entails constant exposure to the effects of others’ judgment, over and above the effects that can be constitutionally constrained. It therefore entails being part of a system that, likewise, exposes others. It is this that legitimates special concern for one’s fellow-citizens. Within a set of arrangements that are justifiable to the extent that they minimize background risks, one has a reason for special concern as part of a system that imposes dynamic risks on others. That is due to the fact that those arrangements are best conceptualized in terms of an implied waiver of the others’ background freedom, which enhances their vulnerability and thus gives rise to an entitlement to concern on the part of all who participate.

II Two questions arise at once. First, why do these risks not count as, simply, the costs and constraints imposed by justice itself? Systems of punishment, after all, impose risks of severe loss on lawbreakers, but that does not give rise to special concern for their fate. But we have already noted an objection to that line of argument, in rejecting


the view that it is the application of coercion itself that particularizes obligation. On the model adopted here, the background system is assumed to be just, as a precondition of (though not the basis for) legitimacy. But the requirements of justice may not all be consistent with one another: there are, after all, such things as confl icts of rights, which involve weighing what may potentially be required by one right against what may be required by another. There are reasonably diverging interpretations of justice. Then there are contested views of social and economic process, such that even those who interpret justice similarly may diverge sharply on policy. In short, there is politics, formal and informal, and those who are subject to political decisions must bear the costs of legitimately made decisions that affront their principles or interests. That does not mean, however, that their principles or interests are negligible. That view could be sustained within a political world as imagined by Carl Schmitt, in which winners take all, but that is not an imaginable world within the framework of a notional contract, in which, as we have seen, no one’s stake can simply be negligible. The second immediate question concerns the sense in which a person can be said to be part of a political society and thus complicit in the way suggested above. The idea offered here needs to be distinguished from three others, which are incompatible with the model of political society developed in this book. One is the idea of national identification. Another is the idea of collective intention. A third is the idea of implied agreement or commitment. The model of national identification suggests that membership should be thought of not in terms of political citizenship, or not basically in those terms, but in terms of a person’s identity and corresponding cultural values. David Miller suggests that we may ascribe collective responsibility on this basis.10 Without at all discounting

10

David Miller, “National Responsibility and International Justice,” in The Ethics of Assistance, ed. Deen Chatterjee (Cambridge: Cambridge University Press, 2004), 123–43.


the colonial past – for which rich nations may owe reparation to former colonies – or current systems of global exploitation, which must be corrected in terms of a criterion of fair treatment, Miller contends that we cannot hold nations blameless for choices that have contributed to their present circumstances. (Here he holds out against critiques of “explanatory nationalism,” that is, critiques – such as Pogge’s – of the view that nations should be held responsible for their own fate.11) There are at least hypothetical cases in which they should be: if a deeply religious society were to invest its resources in building sacred monuments, and if its beliefs were to forbid the exploitation of natural resources, under its control, that would be valuable on the world market, then we should not absolve them of responsibility for what follows. Especially we should not do so if we hold rich countries accountable, on the basis of their values, for what they have done to poor countries. Obviously many empirical matters crop up here. In how many cases can we say that economic development is connected in a causal way with a society’s values? But before we get to that issue, and the large comparative and methodological questions that answering it would entail, we should acknowledge the space between a society’s values and the actions that may be attributable to its members. Sometimes that space is very clear. If the elected president of Philatelists for Peace commits some atrocious crime, we have no trouble understanding that whatever made him do it is (almost certainly) extrinsic to the role that he plays and hence to the values of the association’s members. But for the most part, that space is an interpretative one, and the interpretation of values may be more consequential than the holding of them. Societies may collectively value, say, equality, or the sacredness of jihad. But given the striking openness of both of those ideas to interpretation, there is no reason to suppose that the distance between interpreters of either will be greater than the distance between the interpreters of each. Conceptually speaking, basic concepts rank 11

See Pogge, World Poverty and Human Rights, 139–44.


higher than conceptions (interpretations) of them, but politically speaking a conception of x may be closer to a conception of y than to some rival conception of x. Liberal ideas of social democracy, for example, may be closer to reformist versions of Marxism than to libertarian ideas that appeal to the same intellectual sources as the liberal egalitarians do. To assign responsibility on the basis of values held, without regard to an agent’s consent to their practical interpretation, would seem to run the risk of a too-radical understanding of the ends–means distinction. It is of course true that ends are prior to means in the sense that without a reason to act, action cannot be rational. But that cannot mean that the evaluative responsibility falls on ends alone, for the means adopted have consequences that must themselves be prospectively evaluated, and agents’ responsibility depends on their relation to that process of evaluation. To hold people accountable for what is done in the name of their values, then, is, fi rst, to neglect the fact that decisions about what means to adopt are as morally consequential as the adoption of ends, and, secondly, to neglect the primary agency of (presumably power-holding) interpreters. On both counts, their endorsement of basic values is not enough to connect them strongly to what is done in their name, for their initial endorsement may be decisively outweighed by their strong dislike of either the conception of basic values that is officially adopted or of the means adopted for its realization. The model of collective intention proposes that there is an unmediated connection between membership in a social grouping and obligation, in that to be part of a social grouping is to be part of a “plural subject,” and to be part of a plural subject is to be subject to obligations. So the connection is analytic (unlike the proposal developed here). In her careful and illuminating discussions of this topic, Margaret Gilbert offers as a clear example the case of “walking together.”12 Two people taking a walk together is hardly a model

12

Margaret Gilbert, Living Together: Rationality, Sociality, and Obligation (Lanham, MD: Rowman and Littlefield, 1996), chapter 6.


of elaborate social organization, but Gilbert shows that even this simple case points to the existence of constraints implied by the very endeavor itself. One cannot, for example, abruptly veer off without explanation, or suddenly begin walking at a furious pace that one’s companion cannot match. None of this has to be agreed upon: it is all part of what it is to take a walk together. The strength of this approach is that it preserves what is so valuable in a contractual model – one’s obligation is triggered by one’s own act: in this case, taking a walk – while dispensing with the awkward requirements for identifying something that resembles a contract. It can accommodate, moreover, forms of engagement that cannot be traced to a single time or date (the moment of deciding to walk with someone) but that gradually emerge over time (such as falling into a routine with a friend). That is an achievement that the view advanced here emulates. One kind of doubt about the plural-subject approach arises, however, in the course of extending it from the social to the political case, for as one does so the force and clarity of the initial face-to-face example seem to diminish. To help us enlarge the initial example, Gilbert sketches two further scenarios.13 In the fi rst, a large crowd assembles outside a corn-dealer’s house, and a leader (flouting J. S. Mill’s specific prohibition in On Liberty!) asks (rhetorically) if “we” should attack the house: the crowd (or most of it: NB) roars “Yes!” That affirmation is addressed to no particular individual, but is addressed by a collective to itself. So the person-to-person element of taking a walk is dispensed with. The second scenario dispenses with even the element of co-presence. News comes that all the villages in a certain region are to be attacked by a power that seeks to enslave its inhabitants; each village separately resolves to resist, messengers bring the news of each village’s resolve to each other village, and a plural subject (constituted by the shared resolve to resist) comes into being among communities that do not meet.

13

Margaret Gilbert, A Theory of Political Obligation (Oxford: Clarendon, 2006), 178–79.


While such scenarios narrow the gap between the simple social case and the political case, they do not seem to me to close it. I will state three reasons. The fi rst is that the two expanding scenarios still depend, as the original case does, on a purposeful intention or project. To sign on to such a thing is evidently also to will the means necessary to it (i.e. complying with the implicit terms). Barring some exceptional cases, belonging to a political society is not like that. If it rested on agreement, it would have to be an agreement to be bound to undertake yet unknown future projects (as indeed Gilbert’s formulation of the implied terms suggests: “We jointly accept that that body [i.e. the government] may make and enforce edicts throughout this territory”14). In the spirit of Gilbert’s own use of micro-examples, the equivalent simple case would resemble this: Rashid goes to a party, and discovers that in order to mingle with others he has to wear a paper hat. He readily agrees, picking a hat that happens to be red, and mingles happily with the other guests. After a bit, other guests begin spontaneously (?) to chant someone’s name, let’s say “Darleen,” and Rashid, not wanting to be unsociable, joins in, though he does not know who Darleen is. To general applause (in which Rashid also politely joins) Darleen leaps onto a table, waving a cardboard box. In this box, she says, there are instructions about what anyone wearing a red hat has to do … and so on. The more general and indefi nite the consequences of association, the harder it is to share the intuition that association and obligation are analytically tied together. Reverting to a distinction between overriding and undermining made above, the intuitive response to this story is not, in my view, that Rashid has an obligation to do what the message in the box says, but that it is defeasible. He would not, for example, be obliged to give some mitigating excuse for non-compliance with the instructions in the box. He has no obligation at all, beyond, perhaps, an obligation to be polite in declining. 14

Gilbert, Living Together, 370.


The second reason is that, in the context of large groups most of whose members are entirely unknown to any one member, agreement is an effect rather than a cause of association. When I join you for a walk, or roar approval of the assault on the corn-dealer’s house, or vow to defend my village hoping that other villages will join the resistance, I am pursuing a project of my own, however indispensable the participation of others is to its accomplishment. In the political case, the agreement of others is not just something indispensable to my pursuit of self-chosen goals, it is a condition that I have no choice but to adapt to. There may be some who participate in the national project out of their own personal beliefs – David Miller’s national valuers – but in the limiting case there may be none such, each person simply deferring to the similarly conditioned agreement of every other. “Everyone else does” is in a few cases a good reason – for example, we should drive on the right-hand side of the road if everyone else does – but that consideration is compelling only in cases in which sheer coordination is a priority. In other cases it is an understandable motive, but where we have reason to think that that is the motive we have correspondingly less reason to place moral weight on the fact of agreement. This time, the hapless Rashid agrees to go to a party because the organizer assures him that Candace, whom he likes, will be there; the organizer has persuaded Candace to go, however, by telling her that Marius, whom she likes, will be there, and in turn he has persuaded Marius to go because Rashid, whom he really likes, will be there. Let us call this a self-referential collective agreement – each agrees in the belief that others agree. When its basis is brought to light, surely any force that might attach to agreement will evaporate.15 The third reason relates to the dissenters. In Gilbert’s corndealer’s house assault scenario, some members of the mob apparently yelled “No!” (but to no avail). Are they or are they not, then,

15

For a Sartrean example, Jean-Paul Sartre, Critique de la raison dialectique (Paris: Gallimard, 1960), 615.


“members”? They have not just not agreed, they have expressly disagreed, and if they are still to count as “members” of the group that attacks the house then either we have to give unreasonable weight to an initial act of innocent subscription – Rashid’s party nightmare – or else we have to adopt a “who wills the end wills the means” principle of the sort criticized above in connection with David Miller’s view. If, however, they are not members, then by analogy we will find it hard to account for membership in societies in which dissenters are not just a handful, but often constitute more than 50 percent of the electorate. Gilbert develops enormously interesting devices in enlarging the simple model to the political case, but at some point the disanalogy between a consensual process and a politics of contestation seems bound to come to light. As in Miller’s alternative, nationalist, proposal, a large question is raised by the issue of dissenting holdouts and their apparently unjustifiable inclusion in responsibility. It is the third idea, however, that needs most attention here, as it raises the issue of how to distinguish the waiver model from two related but different views. For Hart, as we have seen, the idea of promising or undertaking a commitment is at the heart of special duties, and others are to be understood as being more or less distantly related to it. We have special duties because we have in some sense undertaken them. As we have also seen, Goodin fi nds Hart guilty of illicit conceptual stretching here: in a classic piece of thorough demolition, he takes us through the standard cases of special duties and shows how little they resemble the paradigm of promise-based commitments.16 Even contracts, he suggests, are better understood as a special (and clear) case of giving rise to reasonable expectations. Employment conditions, though employment is entered contractually, are hedged and regulated by moral and legal constraints, as are seller–buyer relations. Professional ethics are governed by codes of conduct that are not contractual in origin. As for families, the terms

16

Robert Goodin, Protecting the Vulnerable (Chicago: University of Chicago Press, 1985) 44–108.


of marriage are governed by legislation, and the requirements of childcare are nothing to do with promises, but spring directly from the vulnerability of (minor) children. In fact, the model of childcare gives us a better starting-point than the idea of a promise, suggesting as it does that what does the work in special obligations is a concern for protecting those who need our protection. “Your varying responsibilities toward [particular others] seem roughly proportional to the degree to which they are, in fact, dependent on you (and you alone) to perform certain services.”17 We can, I think, accept Goodin’s critique of Hart, without however accepting quite the conclusion that he draws from it. First, we may adopt the weaker form of Hart’s requirement. While taking promises as the paradigm, he effectively replaces the idea of deliberate commitment with the much broader notion of “some voluntary action” that gives rise to an obligation. Secondly, we could absorb much of the force of Goodin’s critique by accepting that even though the content of special obligations is not contractual in origin, those obligations are nevertheless triggered, in many cases, by “some voluntary action.” Consumer protection legislation, for example, greatly modifies the principle of caveat emptor, but it is still the case that one voluntarily becomes a consumer – and so on. Thirdly, beyond that, we should also accept the point that there are some obligations that do not depend on any previous actions at all: the obligations of the Good Samaritan, for example, or obligations grounded simply on the helplessness of the other. Nevertheless, there is something important to be said for retaining the element of previous voluntary action. The importance lies in the need to distinguish between what is due to others because of their vulnerability, and what is due from you to particular others because some action of yours has increased their vulnerability. It is the increase of vulnerability, not the bare recognition of vulnerability, that makes obligations “special.” After all, it is not the 17

Ibid., 33–34.


recognized degree of vulnerability that gives rise to a special case for action – for if it did, parents in rich and secure countries would have a stronger obligation to children in poor and strife-ridden countries than to their own – but your own contribution to that vulnerability. By taking care of a particular patient you have brought about the assignment of medical personnel to other cases. By taking care of particular children (biological or adoptive) you have excluded other potential care-givers. By befriending particular people you have absorbed their investment of social time. I hasten to say – recalling the distinction between motivation and rightness in the previous chapters – that these are not (except, allowably, in the first case) the considerations that should inspire one’s relationships. The inspiration is affective. But to the extent that the issue of rightness happens to arise – if it ever does – it is not the generic vulnerability of sick people, children and friends that would figure, but a history of previous contact with the particular people in question. For two reasons, the model offered here adapts itself better to the political case than the model of sheer vulnerability does. First, political obligation is systemic, and even if we conceive of it as owed to other citizens rather than to an abstract entity (“the state”) it is still owed to all of them, not to some special ones. Here the idea that the strength of an obligation varies with the specific vulnerability of the other seems out of place. Goodin defends his view against the apparently absurd implication that if we are responsible for those who are particularly vulnerable to us, we should avoid driving our sole commercial rival out of business, since a sole rival indeed satisfies the criterion of particular vulnerability.18 He wards off the objection by arguing that the public at large has an interest in effective competition, and that while no single member of the public has as much to lose as your rival does, “lots of little vulnerabilities and responsibilities can combine to trump some larger ones.” If we are to make sense of special civic obligations, then, we have to think of 18

Ibid., 119n.


them as being mediated through a whole set of social, economic, and political arrangements. And if that is so, then one’s responsibility must be triggered by the fact of civic membership. That civic membership enhances vulnerability, in certain ways – as argued above – explains why that is so. The alternative explanation would have to be that other citizens, regardless of one’s personal connection to them, were supposed to be more vulnerable than anyone else – a special system-wide obligation would have to be triggered by special system-wide neediness. But that would not be plausible (except in the single case of the neediest country in the world). That brings us to the second reason. Goodin sets out to show that if we accept the grounds for special obligations we are necessarily led to accept grounds for broader ones. In this he succeeds not only brilliantly but too well. He demolishes the idea (or unthought prejudice) that our duties to our nearest and dearest excuse us from any broader responsibility; but his case for broader responsibility has so long a reach – a global reach – that it cannot satisfactorily ground the special obligations of citizenship. We can provide a ground for them, however, if we shift the focus from the extent of vulnerability to the source of vulnerability, and simply posit that as a member of a civic association one is part of a system of relations that imposes endemic risk. To conclude this section, then, both the nationalist and the plural-subject approaches, in their different ways, attach too much weight to the fact of belonging. Getting from belonging to being obliged, I have argued, is neither a matter of unmediated valuing nor an analytic necessity. The vulnerability thesis, on the other hand, dispenses with any transactional element at all, in order to expunge the excessive voluntarism of Hart’s view. The social waiver thesis proposes to affirm a minimal but significant transactional element, the fact of participation in a system that enhances the vulnerability of other citizens and imposes risks upon them. That participation is not a sort of tenuous simulacrum of a promise: it implies that a recognition of responsibility is due. But it is not a recognition


of responsibility-in-general, something that does not satisfactorily distinguish between civic and Good Samaritan obligations: it is a recognition of what the civic relationship does to those with whom one shares it.

III The idea of “complicity” here is of course central to the argument, and it is important to clarify what it is and (no less crucially) is not. Perhaps the fi rst suspicion that might arise is that it resurrects “tacit consent,” the very idea whose dubious character is responsible for the problem at hand. Locke maintained that simple presence within a jurisdiction amounted to consenting to its authority. Many critics have pointed out that while there is indeed a (binding) phenomenon that could be called “tacit consent,” Locke’s example does not correspond to it. One can be said to tacitly consent when there is a clearly expressed understanding that silence means “yes,” and also when saying “no” carries no penalties or strongly negative consequences. Locke’s example meets neither test. No one supposes that travelling upon a highway carries with it obligations further to obeying the Highway Code; and the refusal to take part in any of Locke’s specified activities would entail a heavy cost (that of emigration, as Hume classically pointed out). But “complicity” in the sense intended here is not meant to be the basis for inferring a strained form of consent. It is not that you have somehow undertaken to do or accept something; it is that you have participated in a process that has increased others’ exposure to certain risks. That is not a consent argument of any kind, but a strict-liability argument. For a micro-example, consider the case of a store selling fragile goods, which posts a sign: “You break it, you buy it.” We can say that anyone shopping at the store consents to the rule, if it’s well enough publicized, when they touch the goods. But suppose, taking a step back, we ask whether the store has a good reason for adopting the rule? If we agree that it has, then liability to it is supported by a reason other than consent. The reason is one that is part of the very idea of responsibility itself. If we cannot


be connected with the consequences of our acts except through our consent, it is hard to see what meaning “responsibility” could have, for it could always be simply evaded. The idea must include a concern not only for what is purposefully assumed, but also for what is reasonably foreseeable. The second suspicion that might arise is that the argument may merely have resurrected the “plural-subject” views that were rejected above. To say that agents are strictly liable for what their actions, together with others’, bring about – how is that different from incorporating them in a plural subject? The answer, I believe – following some critical remarks made above – is that the pluralsubject view places much weight on the normative implications of identity. Responsibilities are ascribed on the basis of a new entity (a “we”) that association brings into existence. The view developed here is different, however, in two respects. First, it maintains that the normative implications of association are sometimes undermined, not merely overridden, in light of both the mode of association and the content and context of the responsibilities that are said to arise. Secondly, the basis of ascribing responsibility is not the coming-into-being of a new entity but the implied waiving of background rights, an argument that preserves the idea of simple individual agency. The model of individual agency relates the proposal to the meticulous study of complicity by Christopher Kutz, who, rejecting the plural-subject thesis, opts for the alternative view of individual agency in which persons become complicit through their intention to “play a part” in a collective process.19 It is the content of their will, not a share in a collective identity, that incurs responsibility. Because the content of people’s wills – over and above the will to play a part – is various, we too will want to ascribe varied degrees of complicity, depending, for example, on the extent to which

19

Christopher Kutz, Complicity (Cambridge: Cambridge University Press, 2000), especially part 3.


participants identify with the overall project. That general idea relates well enough to the idea of civic complicity, which likewise is triggered by participation inspired by varied motives and informed by varied degrees of awareness. But civic complicity, as envisaged here, differs from Kutz’s standard cases in three main ways. First, his standard cases – the bombing of Dresden, for example – involve discrete (if major) enterprises which have beginnings and ends, and which demand enquiry into the location of responsibility and its degrees. Being complicit by virtue of one’s citizenship is not like that. The enterprises undertaken by polities are too various, and the web of participation is too extensive and complex, for one’s citizendebt, as we may call it, to be computed. For particular enterprises – a war, a policy change – or omissions – culpable failures to act – it is necessary (however hard) to figure out degrees of responsibility; but for the risks of citizenship, which have no beginning or end but which form the permanent background to civic life, it can only make sense to impose a flat moral tax, as it were, since relations are too diffusely mediated for individual contributions to be satisfactorily assessed. Secondly, whereas the standard examples of complicity are two-party cases in which one group damages another – Bomber Command devastates Dresden, Lieutenant Calley’s platoon kills Vietnamese villagers – what is meant here by civic complicity concerns the risks that members of one party impose on fellow-members as conditions of their participation. So it relates not to a determinate act – the destruction of a city or village, say – but to a set of ongoing constraints that are imposed on others by an arrangement in which one takes part. Thirdly, whereas the standard cases concern damage for which responsiblity is to be retrospectively assessed, civic complicity concerns risk, that is, prospective and potential damage. The retrospective assessment of damage leads to judgments about liability to punishment or for redress. Complicity in risk-causing arrangements, however, is forward-looking: one’s liability is not for punishment or redress, but for taking part in or supporting steps to limit risk to others, or the impact of risks upon them.


IV At several points above, the argument moved forward by distinguishing between the political and the personal levels – claiming, for example, that the “walking-together” model could not be successfully extended to the political case, or acknowledging that there are cases (such as immediate rescue cases) in which vulnerability alone is enough. This may give rise to a suspicion that the model adopted is too political to cover much ground. And this suspicion would be reinforced by the fact that a social contract model, however attenuated, remains a fundamentally political idea, possibly appropriate to regulating public and coercive relations, but out of place elsewhere. The basic contrast between contractual and intimate relations has often been made, after all:20 contracts imply initial separation followed by (some degree of) connection, while in other contexts connection comes fi rst, separation only later. Unless this suspicion can be met, the argument will tend to resurrect the moral dualism that was questioned in chapter 1: that is to say, the view that the moral universe is not one but two, comprising two incommensurable languages, one of closeness and particularity, the other of cosmopolitan equality. But that conclusion should be resisted. The model of the social waiver, or the attenuated contractualism proposed here, rests on the idea of an implied surrender of a background right to freedom, the surrender of which by others gives rise to a case for special concern on your part. Now the whole idea of rights – surrendered or otherwise – is sometimes seen as basically incompatible with the idea of intimacy. One particularly clear example of this critique is provided by a much-cited paper by John Hardwig,21 which puts forward the view that the very idea of rights has no place in (good or healthy) relations between men and women,

20

21

Goodin, Protecting the Vulnerable, 70–92; Virginia Held, “Mothering and Contract,” in Beyond Self-Interest, ed. Jane Mansbridge (Chicago: University of Chicago Press, 1990), 287–304. John Hardwig, “Should Women Think in Terms of Rights?” Ethics 94 (1984): 441–55.


and that the contribution of women to moral discourse lies exactly in their capacity to dislodge and replace an essentially male rightsoriented way of thinking. What can be said in response to such critiques? Since Hardwig’s concern is with “healthy” relationships, his position can accommodate the objection that in the case of relationships that are not healthy, or have broken down, then rights have a crucial place – “justice replaces affection,” as one defender of the rights view has put it.22 Hardwig’s case is that it is better, indeed essential, that rights not be “present in” close relationships, in the sense that to appeal to them is to invoke separateness at the expense of partnership, respect at the expense of affection. Two substantive issues arise. One is whether all ordinary rights do in fact fall away, or whether there may not still be some (privacy, for example) that could still properly be respected and even invoked. An alternative view is that while it would be deadly for the rights typical of strangers to play a role, rights of a new kind are generated. Employing the language of “friends” in a generic sense, Michael Meyer contends that relations among friends bring into play expectations that can properly be expressed in terms of new rights, particularly demanding ones, in fact – rights to special concern, and so on.23 These are interesting issues, but for the present purpose a more relevant line of thought involves questioning what it means for rights to be present in a situation. Hardwig’s view is that it means that people are aware of them and invoke them when necessary. But there is a more nuanced sense in which they may be said to be “present” because it is their waiving that makes possible, and even constitutes, a relationship. Friendship would be insignificant, or perhaps even inconceivable, if there were no things that non-friends could not do. If anyone could enter your house, it would be of no significance that particular people could 22

23

Jeremy Waldron, “When Justice Replaces Affection: The Need for Rights,” in Liberal Rights: collected papers, 1981–1991 (Cambridge: Cambridge University Press, 1993), 370–91. Michael Meyer, “Rights between Friends,” Journal of Philosophy 89 (1992): 467–83.


reasonably expect to. If anyone could phone you late at night with a problem, it would be of no significance that particular people could. (Running a hotline service, after all, is not like having hundreds of “friends.”) The terms of exclusion, then, change the meaning of what is included. On this argument, polygamy would not be a matter of a man having many “wives,” in the sense that the term “wife” bears in monogamy: the meanings would overlap, but would not be the same, because the scope of exclusion would be different. If this is true, then rights may be said to be “present in” an arrangement even when they are not invoked or even thought of – their effect is present in what the arrangement is. Its shape would not be possible unless the background were ordered in a certain way. As Harry Brighouse has written, “A great deal of rights-thinking does not involve the assertion of rights. It involves waiving one’s rights … Waiving what one regards oneself as having a right to willingly, or enthusiastically, is a meaningful and intimate gesture which would be unavailable absent rights-holding within the relationship.”24 Correspondingly, one can, out of love for a partner, avoid putting them in a situation in which he or she would waive a right of their own, even if, reciprocally, they would willingly do so were the situation to arise. If all that is so, then the distance between a contractual model of the state and an adequate picture of partial relationships may not be as great as is often supposed. We may see both in terms of the setting aside of prior entitlements, and as conceivable only on the supposition that prior entitlements exist. Both are valuable because that setting-aside is essential to warding off risks (albeit of different kinds); and both (in different ways) achieve that good at the cost of exposing participants to new risks. That commonly created exposure generates a strong moral reason to sustain the engagement. It also gives rise to a set of internal requirements, much more easily specifiable, of course, in the political case – we may call it a constitution; 24

Harry Brighouse, “What Rights (if Any) do Children Have?” in The Moral and Political Status of Children, ed. David Archard and Colin M. Macleod (Oxford: Oxford University Press, 2002), 31–52.


but in the personal case its very unspecifiability may be essential to its meaning, calling as it does on continually renewed and highly contextual interpretation, rather than on specifiable duties of a sort that can be handed off to special agencies (such as courts). So that important difference between the two cases does not stand in the way of bringing both within the model suggested by the double potential of institutions, as simultaneously limiting and extending vulnerability, with moral consequences for what should happen within them. The contrary – “dualist” – view that the political and the partial levels are fundamentally different depends in part, as we saw in chapter 1, on an emphasis on “phenomenology,” that is, on the irreducibility of the experience of partial ties to some impartial moral theory. Some reasons were offered for not giving this as much weight as some believe we should do. There it was suggested that the importance of “phenomenology” is diminished once we accept that the internal ideas of appropriateness are constrained and modified by wider moral considerations. What motivates us to act and what justifies the action are indeed often different, but if our motivations are constrained, as they often are, by what can be justified – the case of nepotism was cited – the language of incommensurability seems to go too far. But the considerations above give a stronger reason to reject dualism. If the internal and subjective meanings of arrangements are not just limited and bounded at the margin, but are actually shaped, by wider moral beliefs in the way that the “waiving” model suggests, then surely we do not have two moral worlds that stand in competition with one another. We have a single moral world that is colored by a fundamental sense of the vulnerability of others to risks, and complicated by a sense that the solutions to vulnerability inevitably expose us to it in new forms. That is as true of families and friendships as it is of the state. But we may take this point further. Let us suppose that each person’s life is located within a series of progressively larger arrangements, at each of which the “double potential” referred to above


recurs. What might justify the inclusion of relatively small arrangements within relatively large ones? Well, the candidate reason that at once comes to mind is that each level would have a strong basis for jurisdiction if its role were to restrain or redress the risk-inducing features of the level immediately below. States have a strong basis for jurisdiction over family practices, to take the most obvious example, only because and to the extent that they can diminish the risks arising from familial intimacy. To generalize the point, requirements imposed at each level of organization demand justification in terms of a net reduction in vulnerability at levels below. Those requirements may be, for example, for a secured, substantive right of exit, something that may tend to restrain the abuse of local power, or else provide a remedy when it is abused; or – in cases more amenable to direct regulation – they may be for compulsory observance of the terms of trust, when voluntary observance fails. So, to take a classic example of this logic, from J. S. Mill’s On the Subjection of Women: the Married Women’s Property Act secures (for some women) a right of exit, thus (Mill hoped) tending to change the power relations within the marital micro-polity, while legalizing the voluntarily adopted non-standard marriage contract tends to substitute consent for domination. Whatever merits this schema has are, of course, as much of the problem-defi ning as of the problem-solving kind. Its abstractness should not be taken to imply neglect of the difficulties of applying it. It is simply meant to show that we can take account of the multilevelled nature of human life without having to resort to dualist models that construe choices as incommensurable (or tragic, or merely pragmatic). To the extent that the schema may govern the decisions made within it, two of its general features may be especially important. First, its implied stress on conditionality: there is a constant process of reservation that denies that humans are entirely defi ned or encompassed by their associations – a view classically expressed by social contract theory, though I hope to have escaped some of the critiques that the more literal-minded versions of that theory invite,


by reinterpreting it as an implied “social waiver.” Secondly, there is a strong element of subsidiarity, that is, of the view that levels of organization derive their authority only from the shortcomings of levels below. So there is a certain built-in bias towards the local: higher levels of organization face an uphill slope of justification that compels demonstration of the (counter-productive) injustice of relatively local control. Now those two features – the “reservation” feature and the “subsidiarity” feature – may seem to be at odds with one another in contemporary political thought, the one undermining, the other supporting, local forms of association. The argument of this book, that we have to take them together as complementary requirements, may therefore look like an attempt to square the circle. But the two features, I believe, differ only in their respective distributions of the onus of proof. The “reservation” feature tells us that local associations have no moral validity unless and until they meet a stated criterion of justice. The “subsidiarity” feature tells us that larger associations have no moral standing unless and until more local associations fail. But I see no contradiction in maintaining both that all associations are open to evaluation from a perspective larger than their own, and also that no association is automatically absorbed by those which surround it; some reasons for the consistency of this view were set out in chapter 1. The difference with regard to the onus of proof arises simply from the different contexts of justification and legitimacy, the starting-point for this chapter’s discussion. The more abstract context of justification naturally leads us to adopt criteria of justice that associations will be required to meet. The context of legitimation, based as it is in the concrete activities of civic membership, must take account of the fact that associations are never only instances or derivatives of general justice, but also embody local understandings.

Compatriot preference and the Iteration Proviso

4

Although broader ethnic or religious loyalties sometimes supervene, people all over the world attach special importance to the fate of their compatriots. How to measure the preferential factor is an intriguing question: perhaps we could measure the front-page space devoted to domestic and international matters respectively, or the extent of a foreign as opposed to a domestic disaster needed to cross the threshold of attention, or compare foreign-aid budgets with domestic welfare and social service expenditures. Or perhaps we could measure generally accepted “them–us” kill ratios in warfare. Some measures might yield a very high factor: possibly compatriots are given a thousand times more weight, perhaps more, in some respects. But of course, even if compatriots were (implicitly) judged to be worth only (!) twice as much as others, we should still want to know why. Sometimes psychological reasons are given: it is argued, for example, that Rousseau was right to claim that human attachments weaken as they extend, that they must stop somewhere if they are to retain any motivating force and remain reliable.1 We should, however, still want to know if we are justified in doing what we feel inclined to do. Moreover, Rousseau’s spatial model does not actually fit the facts at all well. Quite often people give more weight to relatively large attachments than to relatively small ones; they send their children off to war, for example – a reminder of the important fact that compatriot preference needs to be justified in relation to smaller local attachments, as well as to whatever it is that global justice demands.2 That consideration 1

2

See Clifford Orwin, “Distant Compassion: CNN and Borrioboola-Gha,” The National Interest, Spring 1996, 42–49. See Charles Jones, Global Justice: Defending Cosmopolitanism (Oxford: Oxford University Press, 1999), 131–33; Margaret Moore, The Ethics of Nationalism (Oxford: Oxford University Press, 2001), 47–50.

Compatriot preference and the Iteration Proviso 93

has particular weight in light of views that partiality at the subnational level is more readily justified than compatriot preference.3 There is a sense in which the very fact of valuing group membership will work to the disadvantage of non-members, to the extent that members’ contributions of time, attention, and resources to their group will necessarily diminish what can be given to those outside. The point was noted by Cicero, and in fact played a key role, in his hands, in transforming Stoicism from a cosmopolitan doctrine into a kind of self-righteous Roman patriotism: it would be wrong, Cicero believed, to give to outsiders anything that diminishes one’s ability to care for one’s own – any scarce good, in fact, so that we owe outsiders no more than the gift of renewable goods such as fire and running water.4 The same point is elaborated by Samuel Scheffler, in a thorough discussion that seeks to demonstrate that “associations” entail commitments that cannot be derived from, and may conflict with, obligations of general morality.5 Scheffler does not undertake to show that one should give preference to associative duties, only that those duties coexist with duties of general justice, without possibility of theoretical resolution. That view depends, of course, on interpretations of partial and universal morality that have been questioned above. If we reject those interpretations, how can partial and universal obligations be reconciled? We have already seen some difficulties that arise in the effort to do so. In particular, we have seen some difficulties that obstruct attempts to justify partial obligations directly in terms of the realization of some general value. One such attempt built upon the idea of fairness, but was open to the objection that there is no

3

4

5

See Henry Shue, “Mediating Duties,” Ethics 98 (1988): 687–704; Peter Singer, “Outsiders: Our Obligations to Those Outside Our Borders,” in The Ethics of Assistance: Morality and the Distant Needy, ed. Deen Chatterjee (Cambridge: Cambridge University Press, 2004), 11–32. Thomas Pangle’s view of Cicero’s cosmopolitanism as “moderate” is thus open to debate: see Thomas Pangle, “Socratic Cosmopolitanism: Cicero’s Critique and the Transformation of the Stoic Ideal,” Canadian Journal of Political Science 36 (1998): 235–62. Scheffler, Boundaries and Allegiances, 97–110.


reason to confine the site of fairness to one’s society. Another represented political societies as simply local sites for the efficient execution of general human duties, but met the difficulty of showing that other arrangements would not be more efficient, especially in tackling inequalities. A third put forward the view that respect for autonomy required special concern for those who are coerced; and although that view could be defended against some important objections, it was argued, it was unable to show that states are coercive to a decisively higher degree than other forms of organization. Some of the same difficulties arise for more selective accounts, those that offer an account of the mutual obligations that arise within a society on the basis of some general political value, of a kind that we are assumed to admire: a liberal, liberal-democratic, republican, or egalitarian kind. Two refi ned recent treatments of compatriot preference have advanced excellent reasons for giving special weight to the mutual obligations of citizens in societies with good institutions. Andrew Mason has compared the good of shared citizenship, in a polity that he terms “republican,” to the good of friendship: it is valuable not just for the benefits that it produces but also for the kind of relationship through which it produces them, a relationship that assigns equal status to all in the process of shaping the public environment.6 Along different lines, Christopher Wellman draws attention to several ways in which it is particularly important to advance equality among co-citizens: deprivation tends to be measured in relation to local rather than distant others, for example, and relative material equality is essential to the effective exercise of rights.7 Such arguments, for all their appeal, obviously leave unexplained the case of societies that do not take equality and rights to be normatively basic. More important, they are subject to the objections developed 6 7

Andrew Mason, “Special Obligations to Compatriots,” Ethics 107 (1997): 427–47. Christopher H. Wellman, “Relational Facts in Liberal Political Theory: Is There Magic in the Pronoun ‘My’?” Ethics 110 (2000): 537–62, and “Friends, Compatriots, and Special Political Obligations,” Political Theory 29 (2001): 217–36. In his contribution to Wellman and Simmons, Is There a Duty?, Wellman develops a view of civic obligation that is not regime-dependent.


by Simmons and Singer, that they give us good reason to support any society in which the values we admire are embedded, but no special reason to support our own. We have not, however, attended to arguments that operate in the reverse direction, that is, arguments that begin with a commitment to partial morality and seek to show that it can be extended or modified to account for duties beyond national boundaries. So this chapter begins with an examination of two influential versions of that approach.

I Someone who takes partiality as morally basic believes that “relations between persons are part of the basic subject-matter of ethics” and that agents are “encumbered with a variety of ties and commitments to particular other agents, or to groups and collectivities, and [that] they begin their ethical reasoning from those commitments.”8 Beginning our ethical reasoning from those commitments, it would seem that we could accommodate further – non-particular – commitments in one of two possible ways. We could admit another level of morality, universal in scope, of a subordinate or derivative kind; or we could stretch and enlarge the idea of particularity so that it became globally comprehensive. Michael Walzer and Richard Rorty, respectively, provide examples of these two strategies. According to Walzer, the “thick” nature of morality – its rootedness in locality and history – does not preclude a “thin” morality that extends beyond our national communities. This thin morality enables us to identify injustice and oppression in other, different, societies, and to empathize with movements of protest against them. Our understanding of those movements will necessarily be less than total, for they will have emerged from the thick morality of their own local cultures. Nevertheless, there is enough overlap for significant partial understanding. Thin (or minimal) morality “simply designates 8

David Miller, On Nationality (Oxford: Clarendon, 1995), 50.


some reiterated features of particular thick or maximal moralities”:9 “It consists in principles and rules that are reiterated in different times and places, and that are seen to be similar even though they are expressed in different idioms and reflect different histories and different versions of the world.”

10

It is not at all emotionally thin –

on the contrary, when it comes into play it is likely to be intense. Nevertheless it is derivative in the sense that one would only ever arrive at it through the prior experience of a thick morality. Contrary to theories that begin with some universal basic proposition, Walzer claims that “maximalism in fact precedes minimalism,” and if we did not have our own thick or maximal morality to begin with, we would never arrive at those moments of intense though necessarily sporadic empathy with the struggles of other societies.11 Now this proposal, if true, would give us a very compelling general reason to put compatriots fi rst. Nothing less than the future of morality would depend on it. It is in the specific local understandings among compatriots that morality itself has both its origin and permanent home: and if we tried to be cosmopolitans fi rst, we would undermine its very existence. No doubt we can make some room, within this view, for transnational sympathies, and even accept losses in responding to dire emergencies in other parts of the world.12 But the very structure of morality rests on compatriot preference and the complex and rooted understandings and arrangements arising from it. While this claims to tell us where morality comes from, it is less than perfectly informative about what it leads to, for the thin or minimal morality, once in play, has a critical force that may undermine thick or maximal expectations, in relation both to what we owe to compatriots and what we owe to others; and it is in general impossible to say what should happen when a thin critical demand 9

10 12

Michael Walzer, Thick and Thin (Notre Dame: University of Notre Dame Press, 1994), 10. 11 Ibid., 17. Ibid., 6, 13, 19. Ibid., 15–16, and Michael Walzer, Just and Unjust Wars, 3rd edn. (New York: Basic, 2000), xi–xvi.


and a thick settled expectation come into conflict. Thin morality may lead people to judge, for example, that an anti-polygamy protest in a foreign country is justified, even though we may know little about the culture in question; but it may also lead us to judge that gender roles in our own society are unjust – with more legitimate confidence, in fact, since we are acquainted with the culture in question. Or, someone may arrive at the view that resource-intensive consumption in rich countries like her own violated the requirements of thin morality in relation to people in poor countries; and she might then conclude that certain practices treasured by her compatriots – such as culinary traditions – were indefensible. It would obviously be arbitrary in such cases to say that the importance of maintaining tradition and sustaining cultural ties with compatriots must come fi rst, however highly we may value traditions and cultures. Implausibly, we might imagine situations in which compatriot duties should “come fi rst” in that sense – that is, take normative priority – because the very survival of a culture was at stake. The implausibility here is familiar from the well-trodden ground surrounding Lord Devlin’s claims about the effects of permissive legislation on national survival. As critiques by Hart and Dworkin made clear, the claim that abandoning some treasured practice will destroy one’s community is either circular – the community in question is implicitly defi ned as one containing that very practice – or else embodies improbably dire causal predictions that equate the effects of deviance with the effects of treason.13 To be sure, we may speculate about limiting cases in which some feature is so deeply and uniquely emblematic of a society that its loss would dissolve its identity. But the real-world issues that raise the matter of compatriot preference come at a point on the spectrum that falls very well short of posing survival problems of any kind. They are issues about economic protectionism, foreign aid, humanitarian intervention, and so 13

H. L. A. Hart, Law, Liberty and Morality (Stanford: Stanford University Press, 1963); Ronald Dworkin, Taking Rights Seriously (Cambridge, MA: Harvard University Press, 1978), 240–58.


on, issues that do not even invite either possible version of a Devlintype thesis. It could very much more plausibly be argued that, without at least a strong presumption in favor of compatriot preference, local moral community could be eroded in ways that damage its effectiveness and value. Here the familiar parallels are not in legal theory but in discussions of the family. Along lines very similar to those pursued by critics of cosmopolitanism, “family values” theorists have argued that to take general moral obligations as primary, or to see the family as essentially a vehicle for protecting general rights, is to deny its own moral nature. According to Ferdinand Schoeman, for example, to ground parental care on children’s rights is effectively to turn parents into public servants and to deny the intimacy and privacy of the parent-child relation.14 Because of the special quality of this relation, every transfer of functions from parents to public agencies has the potential to undermine it. Likewise, one might argue, compatriot preference is an essential part of maintaining ties of solidarity, and if the specialness of the tie between compatriots gives way to a principle that weights all humans equally, it too will be diminished. But first, considerations of this kind give us only defeasible conclusions. No defender of family privacy, no defender of communitarian solidarity, argues for absolute protection. “Clear and present danger” to the child (Schoeman) or state acts that “shock the conscience of mankind” (Walzer) open the family or the state to external judgments and interventions; and as before, it does not seem that the arguments from either privacy or solidarity can themselves tell us when they do or do not retain priority, because they leave untouched considerations that a broader context might supply. At best, they can give us a subsidiarity principle (of the kind defended in the previous chapter), or a presumption in favor of local decision-making. Secondly, arguments of this kind quite rapidly evolve into arguments of the kind that was briefly mentioned above, that is, arguments in defense of institutions 14

Schoeman, “Rights of Children.”


deemed to be good ones by some independent criterion, and then the weight comes to fall on the values that the criterion reflects. No one wants the state to intrude into well-functioning families characterized by love and mutual respect; no one doubts that citizens are strongly bound by decisions made by well-functioning and rights-respecting democracies; but the reasons do not arise from consideration for the simple integrity of the institutions in question, but, rather, from the belief that love and liberal democracy are good things. So, to sum up: starting from the idea that moral community is basic, one can add further layers of duty of a “thinner” kind, without necessary inconsistency. But the argument does not have the internal resources to establish the priority, in the relevant sense, of compatriot preference – not, at least, without making implausible claims or else collapsing into an argument based on the value of some version of democracy. One can see the attraction, then, of the alternative strategy, one that does not introduce two kinds of morality – thus raising questions of priority that cannot be settled within the argument itself – but which sticks firmly with one kind. This is the “justice as larger loyalty” thesis, which holds that all morality is a matter of loyalty to some group and its way of doing things. Now it might seem at once that such an approach would simply rule out at ground level any idea of obligations beyond groups, such as, for example, obligations based on ideas of “human rights.” But, some proponents maintain, such beliefs can be read into the approach, by way of extending the idea of a “group.” Human morality is just the loyalty internal to the human group as a whole, so cosmopolitan moralists have just been advocating a group loyalty all along – it just happened to be a very large group that interested them, in fact, the largest possible human group. Andrew Oldenquist, for example, hypothesized that if Immanuel Kant had met rational beings from another planet, he would have treated them as alien, because what he was teaching was really human species-loyalty.15 15

Andrew Oldenquist, “Loyalties,” Journal of Philosophy 89 (1982): 173–93.


(Evidence for that hypothesis is of course hard to come by.) Essentially the same view is advanced by Richard Rorty, in several well-known texts. He offers the (“non-Kantian”) view that “one’s moral identity is determined by the group or groups to which one identifies,” and that conflicts often described as conflicts between group loyalty and (universal) justice should be redescribed as “conflicts between loyalties to smaller groups and loyalties to larger groups.”16 Any idea of a human obligation should be grounded not in any Enlightenmentdescended rationalism but in a “community of trust” that the nations of the world might build.17 The fi rst and most obvious response to this is that it inverts the usual relationship between community and loyalty. Usually when we speak of loyalty we imply that an object of loyalty is already in being, an object to which people with a certain identity feel attached; in Rorty’s proposal, however, it seems that we are to create an object in order to have loyalty to it. He accepts Walzer’s view, in Thick and Thin, that morality begins with the thick commitments generated within partial groups,18 but maintains that a stretched kind of thickness could reach the scale of humanity as a whole. Whereas Walzer denies that humanity is a community – “it has no history and no culture, no customary practices, no shared understandings of social goods”19 – Rorty believes that it has community potential. It is here that the inversion of the usual relationship between community and loyalty becomes more than a matter of words; for we have to ask, if the community is only a potential one, what is it that

16 17 18

19

Rorty, “Justice as a Larger Loyalty,” 47–48. Ibid., 56. Ibid., 47–48, 54. Without reference to Rorty, Toni Erskine, assessing the “embedded” or identity-based approach to global justice, adduces as one example the identity that one acquires in subscribing to a cosmopolitan tradition or practice: Toni Erskine, Embedded Cosmopolitanism: Duties to Strangers and Enemies in a World of “Dislocated Communities” (Oxford University Press, for the British Academy, 2008), 225. We can accept that as an important basis for identity, while wondering about what takes priority here. Perhaps sometimes it is subscription that comes fi rst. Walzer, Thick and Thin, 8.


would motivate us to create it? Not all potentials compel realization, after all. Two possible answers suggest themselves. Suppose we take the view that communities are particular in the sense of having distinctive characteristics that set them off from other communities;20 we would then naturally enquire into the shared characteristics of human beings. Now that approach is explicitly ruled out by Rorty himself, on the grounds that “socialization goes all the way down,”21 the idea of a common human nature being one of the most emblematic relics of Enlightenment rationalism. So, it seems, the approach should be abandoned at once, as a way of making sense of Rorty’s view. But in his critique of Rorty’s work, Norman Geras has compiled quite a list of the ways in which the anti-Enlightenment claim is apparently qualified.22 While sharing with other animals the capacity to suffer pain, humans, Rorty says, are unique in suffering one kind of pain, humiliation; that capacity is linked to further unique qualities such as the use of language, and the “distinctively human” symbolic understanding of one’s life-context; and it is also made to yield normative conclusions about the kinds of cruelty that it is worse to employ. Now this is not the place to explore whether Rorty’s explicit denial of human nature is consistent or inconsistent with his apparent affirmation of it. The point for the moment is simply that, to the extent that the case for creating a global community draws at all on the kinds of human properties that are affirmed or implied, it is clearly not a partialist or community-based case. The second answer is more consistent with Rorty’s actual argumentation in Contingency, Irony and Solidarity and elsewhere. This does not rest at all on an appeal to shared properties. It is a matter of expanding the scope of a morality entertained by particular kinds of society, that is, liberal ones. It is a matter of “expanding the 20 21

22

Miller, On Nationality, 32–3. Richard Rorty, Contingency, Irony and Solidarity (Cambridge: Cambridge University Press, 1989), xiii. Geras, Solidarity in the Conversation of Mankind, 47–70.


range of the present ‘we’,”

23

as opposed to the universalist pretence

that there is a neutral standpoint (or else the Foucauldian project of creating an entirely new “we”). This is to be done by addressing people of other societies not as arrogant bearers of universal truth but as “frankly ethnocentric”24 persuaders who will paint attractive pictures of what we have achieved on the basis of our own language, metaphors, and practices. Now it is clear that all the work here is being done by the substantive content of our own moral community, not by the bare fact of community itself. First, whether or not one approves of the project depends entirely on one’s finding that content attractive, and secondly, whether or not it succeeds depends entirely on others’ fi nding it attractive – it is only from an external point of view, not from the point of view of the parties involved, that it can be described as a process of community expansion. As before, the “loyalty” is a dependent variable, not something with explanatory or justificatory force. What we have, again as before, is a preference for a particular kind of society, though the way in which this preference is filtered through the language of community removes from it any need to justify itself. So it does not seem that the “expanded loyalty” version of the partialist case is any more successful than the “add human rights” version. Certainly it allows for a community of global scope, but it does not explain why anyone would want one, unless it either reverts to a covert humanism, or else draws on and extends the moral and political content that just happens to comprise the ideology of societies such as ours. As in the case of Walzer’s version, the effort to include transnational duties either fails or else compromises the basic position.

II If the contractualist solution advanced here is distinct from all the approaches found wanting above, it is nevertheless clearly closer 23

Rorty, Contingency, Irony and Solidarity, 64n.

24

Ibid., 56.


to the moral universalist solution than the partialist one. For one thing, it requires local political arrangements to seek justification in terms of a model that begins by abstracting from them. For another, the model clearly requires a background morality, for without it the very notion of a contract is inconceivable. This is so at the levels of both justification and legitimacy. At the level of justification, it has been argued that a justifiable political society must be one that makes sense of the waiving of background rights by offering a more secure context for the exercise of agency. At the level of legitimacy, it was argued that participation in arrangements made possible by the waiver of others’ rights generates a special reason for concern for one’s co-citizens. Now the very idea of a contractualist solution may at once seem paradoxical, for contractualism is often seen, rather, as the problem. A full-length study by Martha Nussbaum, for example, develops the view that outsiders (along with the handicapped, and with non-human animals) are necessarily shut out of the moral picture by the social contract model, since only those who can participate will get their interests considered. So we should turn, she says, to “Grotius, not Locke,” that is, to a directly cosmopolitan view such as was offered by Natural Law.25 But even more interestingly, the idea that a contractual model of political society carries with it no concern for outsiders is shared by the most influential contractualists themselves. Rawls’ A Theory of Justice – following a classical example set by Hobbes, Locke, Rousseau, and Kant – adopted the model of a society, a bounded and exclusive entity, as the subject of justice. Many, including some notable Rawlsians, found this troubling in its neglect of the concerns about international justice that became increasingly salient in the following years.26 But Rawls’ later attempt to accommodate the international domain in The Law of Peoples was 25

26

Martha C. Nussbaum, Frontiers of Justice: Disability, Nationality, Species Membership (Cambridge, MA: Harvard University Press, 2006), 36–38. John Rawls, “The Law of Peoples,” Critical Inquiry 20 (1993): 36–68.


not to satisfy his more cosmopolitan admirers at all, for it merely confi rmed his original conclusion that the full requirements of justice do not apply beyond state boundaries, and that much weaker obligations are in play there. For those more cosmopolitan admirers, the attractive solution is, of course, to go global: to think in terms of a single world-wide contract that will make relations among all humans the subject of justice. Only a globalized version of the contract, they believe, has the critical purchase needed to address global inequalities among political societies.27 This chapter does not seek to add to the critical literature on either side of this question. It does not enquire into whether or not Rawls himself can escape the long-standing charge of arbitrariness in confi ning the scope of justice as he does,28 or whether, on the other hand, his more cosmopolitan admirers can manage to give a good account of the density of civic ties within states, which the idea of a single global contract may deny. It argues, rather, that both Rawls and the cosmopolitans neglect an important resource from within the contractual tradition. They share the assumption that the obligations of justice arising from the domestic version of the contract do not apply to anyone beyond the contracting parties – hence the hyper-cautious internationalism of The Law of Peoples on the one hand, and the demanding notion of a globalized contract on the other. But I will suggest that the contractual argument contains a necessary feature, termed here the Iteration Proviso, that immediately connects political obligations with global ones. The Proviso is a very simple one. It is the claim that a group of people can legitimately set out to confer special advantages upon each other if others, outside that group, are free to do the same in their own case. Given a good-faith requirement, an exclusive political group of people can claim legitimacy for their arrangements only 27

28

See especially Charles H. Beitz, Political Theory and International Relations, 2nd ed. (Princeton: Princeton University Press, 1999), and Thomas Pogge, Realizing Rawls (Ithaca: Cornell University Press, 1989). A charge made as long ago as 1973 by Brian Barry, in The Liberal Theory of Justice (Oxford: Clarendon, 1973), 128–33.


if they acknowledge two duties: a duty to aid other groups when the Proviso manifestly does not hold, and a duty not to harm the societybuilding work of other groups. These two duties are more demanding, on the one hand, than the international duties that Rawls sets out. They are less demandingly egalitarian, on the other hand, than the duties arising from the cosmopolitan one-contract argument. But they have the merit – if the argument developed below is successful – of resting very significant international duties upon exactly the same grounds as our duties to co-nationals. If we are entitled to give special weight to the claims of co-nationals, the argument will claim, then some demanding obligations to outsiders cannot consistently be refused, if obligations to those inside our own political society are not also to be undermined.

III The more-or-less uncontroversial starting-point for this case is that we should think of contractualism in terms of a view of reciprocity, in a particular sense of that term that distinguishes it from mutual advantage on the one hand and simple impartiality on the other.29 A contract of mutual advantage is possible among people all of whom are able both to contribute to, and to draw from, a pool of benefits, and is motivated by enlightened self-interest. A simple impartialist view makes any sort of contract redundant, for it expresses a commitment to deal equally with others without the mediation of any conventional arrangement at all. The reciprocity view, as distinct from either of those two alternatives, expresses a desire to act justly to others provided one has assurance that those others will also be bound by rules of justice. It rejects the criterion of advantage. It accepts the criterion of impartiality, but only conditionally. A social contract, then, marks out the boundaries of a group of people who give one another reciprocal assurance of just dealing.

29

See Alan Gibbard, “Constructing Justice,” Philosophy and Public Affairs 20 (1991): 264–79.


Now if an arrangement is conceived of in that way it is quite right that it should be accompanied by duties that do not apply outside the arrangement: springing as they do from conditions of reciprocity, they are not arbitrary even though they are exclusive. But those who are outside the arrangement cannot arbitrarily be excluded from moral concern altogether, by virtue of the background morality that is presupposed. The justice that we propose to extend to (a limited set of) others will be supported by some understanding of people’s morally relevant features. Let us say: my desire to act justly arises from my understanding that others, like me, are rational and vulnerable creatures who wish to pursue ends in the face of potentially damaging obstacles; so I support a social contract as the basis for a scheme of institutions that acknowledges and accommodates our shared rationality and vulnerability. But while I am entitled to confine the scope of my resulting duties to those who are likewise part of the scheme, nothing can entitle me to deny that those outside it also have exactly those morally relevant features that get the argument launched in the fi rst place. It is here that the Iteration Proviso plays its most salient role. Exclusive associations are, as Samuel Scheffler points out, doubly hard on the excluded: they sink lower in the affectionate priorities of those in the in-group, and in addition do not benefit from those special obligations that the in-group generates within itself. 30 But this unfairness evaporates if we simply suppose that those excluded from our in-group can form an in-group of their own, in turn excluding us from its special benefits. We can deal with the potential moral problem here, in other words, simply by supposing that the procedure that constitutes our group can be replicated an indefinite number of times around the world: for then our own exclusiveness does not imply an arbitrary denial of the morally relevant features of others, and thus it can be justifiable. 30

Scheffler, Boundaries and Allegiances, 84–85.


We need to ask, though, before moving on, why the contracting parties are supposed to have scruples that must be satisfied in this way. After all, are not most contracts more unworriedly self-regarding? If A and B agree to marry, for example, it would be odd to make their agreement depend on C, A’s disappointed suitor, finding a partner, and even odder to make the agreement depend on everyone else succeeding in fi nding one. But the practice of contracting within civil society is an institution, and depends on a background assumption that matters subject to contractual arrangement are suitably left to private initiative. (Contracts that would lead to rights-violating externalities are, after all, forbidden.) So A and B have the possibility of contracting just because C and D and E and F can make similar contracts, and that they are responsible for their own arrangements is a reasonable default assumption. No more than that, however, is required here in the case of the social contract itself: just as contracts within civil society take place against a given background scheme, so the idea of a social contract assumes the freedom of those outside it to create background schemes of their own. But it would be wrong to suppose that the Proviso merely leaves others to their own devices. What it requires is that in making special arrangements for ourselves we must be able to make a good-faith assumption that others can behave symmetrically. While belief in their agency would be an acceptable default position, there are situations in which that assumption cannot be held in good faith. Those are situations in which a state fails, or else turns criminal. A failed state is one which no longer has the resources to provide the basic conditions of security for its citizens. A criminal state is one which has resources, but which turns them atrociously against its own citizens. In such cases, the assumption that would make our own self-preference legitimate quite plainly fails. The victims of failed or criminal states are not in a position to help themselves. To continue on the basis of the default assumption would be like continuing with our family birthday party while the children next door


noisily attack one another with spears, or are publicly garrotted by their psychotic parent. The assumption sustaining the normal privacy of our arrangements would plainly fail. A duty to aid, then, would seem to follow in such circumstances, for no one could suppose in good faith that the assumption that underpins our own social project would still hold. There is of course a sternly doctrinaire liberal view, advanced by J. S. Mill, that tells us to refrain from aiding, on the grounds that the only real help is self-help by the agents themselves. 31 I do not think that view is entirely without merit, for global parentalism certainly has its limits. But surely there must be a tipping-point beyond which belief in self-help becomes merely doctrinaire: at a certain point, social collapse makes agency (individual or collective) impossible, or else tyranny makes it fatal to the agents. At that point, wherever it is, the doctrine of self-help runs out, and members of successful states must face the fact that the legitimacy of their own arrangements is undermined if they ignore what is happening to the victims of state failure or state crime. In the absence of aid, their own basic legitimating premise could be held only in bad faith.

IV But I believe much stronger requirements than that can be extracted from the Proviso, if we think of the social contract, more controversially, in the light proposed above – as a risk-taking project involving sustained hazard: a defense against the antecedent risks of a state of nature, that is also, necessarily, a source of subsequent risk, and which therefore makes sense only in terms of an ongoing quest for a positive net balance. Members of one such risk-taking project owe it to members of others not to hinder them in achieving a positive net balance in their own case. In this light, a social contract will be a reasonable arrangement only if it is understood to refer to a process of risk reduction over time: if it contains terms that ward off the 31

John Stuart Mill, “A Few Words on Non-Intervention,” Collected Works, vol. xxi, ed. John M. Robson (Toronto: University of Toronto Press, 1977).


threat of power abuse, guarantee a high level of personal security, and acknowledge the vulnerability of minorities to majority decisions. It is reasonable, in other words, if understood as a kind of project for collective risk management. Now taking part in such a risk-reducing project will be demanding. It will demand that its participants attach special importance to sustaining the arrangements that they share with others; to exercising restraint in social interactions with proximate others; to playing their assigned constitutional role; to warding off threats to the positive sum of risked results. Playing one’s part in sustaining the contract will carry with it local duties that do not extend beyond the contractors’ borders. And the fact that some, being outside the scheme, do not get the benefits of the scheme, cannot itself be a reason to extend the scheme to them. 32 Once again, though, the basic moral problem of exclusion here is resolved if we take the continuing requirements of the contract to be subject to the same Proviso as the original contract – “if others can do so too.” In the context of the original contract, the Proviso can be adopted in good faith only by accepting an obligation to aid the victims of failed or criminal states. In the context of the contract’s continuing requirements, the Proviso can be adopted in good faith only by undertaking not to impede the efforts of other states to attain the positive net sum that makes the contract reasonable. And there can be no doubt that in the world as it is rich states are in violation of this good-faith requirement, for in many ways they seriously impede the efforts of poor states to manage their own risks successfully. Some of their practices do so directly: they engage, for example, in a trade in small arms that contributes to both state failure and state criminality, and they support authoritarian regimes that halt any prospect of democratic development for societies that are unlucky enough to have valued natural resources or geopolitical importance. But more significant, in the longer run, are practices whose political effect is 32

See Brian Barry, Democracy, Power and Justice: Essays in Political Theory (Oxford: Clarendon, 1989), 445.


less direct. If we suppose that there is any connection between economic development and political development – a connection that the fi nal chapter below will defend – then the economic practices of wealthy states, and many of the policies of the international institutions that wealthy states support and largely control, have devastating effects not only in terms of poverty but also in terms of impeding political development elsewhere. None of those policies is justifiable on the version of the contractual argument advanced here, for on that version a project to prefer one’s own, while legitimate, cannot arbitrarily deprive others of what we shall term cosmopolitan moral regard. Cosmopolitan moral regard, on the model outlined here, does not require equal treatment but a global equal-opportunity principle, giving rise to a duty to aid when the opportunity for self-help has vanished, and a duty not to deprive others of the possibility of enjoying the political and economic conditions that rich countries standardly seek.

V Those considerations, however, necessitate a further distinction, for the approach adopted here does not take the facts of global interaction as morally basic. The approach adopted here takes as its basis the moral implication of political practice, not the consequences of it. The moral logic of contractualism, it is argued, lies in what is implied in the idea of a social contract, that is, respect for the contractually undertaken projects of others. Respect, of course, includes a duty not to undermine them. But it also includes a duty to aid, regardless of anything we may have done to generate the need for aid in the fi rst place. Other approaches, however, take as their startingpoint not some global version of the moral generalization principle favored here, but, rather, the facts of globalization. No one, these days, needs to be reminded of these facts. Calling for technical support on my Italian-designed, Chinese-manufactured pasta machine, I reach a call center somewhere in India. We are reminded at every moment of our interdependence with distant


others. According to some important theorists, that is why we have duties to distant others. We have duties to deal justly with those whose lives we affect. Onora O’Neill is a particularly forthright advocate of this position, which she regards as much superior to views – such as human-rights-based views – that try to derive duties from the bare status of the other, regardless of any relationship we may have with them.33 This makes a certain sense, perhaps, in light of the Kantian complexion of O’Neill’s argument: the injunction to treat others as ends rather than only as means may not come into play except in the case that we are treating others as something, and may simply have no bearing when we are not treating others as anything at all because we have nothing to do with them and perhaps do not even know that they exist. O’Neill seeks to make the point by appealing to our intuitions about the moral relations, or, rather, lack of such relations, between groups in pre-globalized society. In a globalized world, she rightly points out, we constantly engage in actions that imply belief in the agency and subjectivity of distant humans, this implied recognition linking us to others in ways that give rise to duties. But, she continues, “The inhabitants of Viking Dublin and their Peruvian contemporaries did not know of one another’s existence: they lived beyond the pale . . . of one another’s known world; they did not and could not premise action on assumptions about one another’s capacity to suffer.”34 The example may trade in part on our sense that Vikings occupied the biker-gang rather than the flower-child end of the sensitivity spectrum; and may also trade in part on the fact of ignorance rather than non-connection. Be that as it may, let us try two more scenarios that lead us in a different direction. First, suppose that a country falls under the control of a dictatorship that removes it from the global market and from participation in global institutions, imposing isolation upon it and practicing rigorous control of information. Let us 33

34

Onora O’Neill, Bounds of Justice (Cambridge: Cambridge University Press, 2000). Ibid., 195.


call this the “North Korea” problem: on the interactionist argument, we would seem to have no reason to care about the people of such a society, even if they suffered starvation and abuse. Secondly, imagine what we might call the “spoiled brat” problem: after engaging in profitable interactions with other societies, a rich country, discovering that it is expected to become subject to moral rules thereby, withdraws from engagement entirely and confi nes its exchanges to similarly wealthy partners; again, on the interactionist model, we could have no reason to complain. That one would have obligations as a party to an arrangement cannot be a reason either to enter into the arrangement itself, or to continue it once in. Another version of interactionism, if justified, would undermine the approach taken in this book. This is the view, advanced by John Charvet, that the social contract approach is itself parasitic on facts about interaction.35 It is entirely implausible, Charvet says, to think of the contract as generating a society out of previously atomic individuals. We should substitute for that fairy-tale notion a view of the social contract as a device put to use by already associated people in reflecting critically on their relations with one another and trying to regulate them by an agreed-upon impersonal principle. By extension, then, we should see association on a global scale as a necessary condition for extending the contractual model to the global level. “Moral life emerged in a multiplicity of separate associations and is likely to continue to require them for some time. Hence if there is to be a moral life encompassing all human beings, it must come about through relations between these associations.”36 The starting-point of this argument is convincing, and welcome for showing that the 35

36

John Charvet, “International Society from a Contractarian Perspective,” in International Society: Diverse Ethical Perspectives, ed. David R. Mapel and Terry Nardin (Princeton: Princeton University Press, 1998), 114–31. Ibid., 120. While I agree that “moral life” must in some sense “come about” in social relations, the idea of contractualism entertained in this chapter must obviously reject the critique of moral realism that Charvet associates with that idea. For some critical remarks on Charvet’s view, see Chris Brown, “Contractarian Thought and the Constitution of International Society,” in International Society, ed. Mapel and Nardin, 132–43.


social contract argument is compatible with a sociological realism that it is naively thought naively to exclude. But the conclusion does not follow. It appears to commit what we may term the fallacy of replication, i.e. the view that if B follows from A then it must have the same form as A. But I have argued above that B may follow, rather, from implication. What is implied in our critically-reflected-upon association with insiders may have direct consequences for what we owe to outsiders, whether or not we have an equivalent degree of association with them. At most, the claim that the conditions of the domestic contract would have to be globally replicated would tell against something resembling a global contract creating a global polity. The claim cannot, however, tell against an argument that the domestic contract – even if distinctively comprehensive – should lead us to acknowledge its other-regarding consequences. Those consequences, coupled only with a good-faith requirement, should lead us to accept some important transnational responsibilities.

VI In this fi nal section, I want to attempt two different but related things. One is to say something about the basic political idea that underlies this chapter. The other is to say something about what, politically speaking, follows from it. The general picture offered here is one of a world in which persons have a sense of their own rationality (their desire to look ahead) and vulnerability (the obstacles to that desire) and acknowledge the same features in others. For entirely natural reasons, that acknowledgment, combined with a desire for personal security, justifies the formation of exclusive societies, legitimately exclusive because the shared pursuit of security creates heightened risks, in view of which citizens can reasonably be expected to have heightened care for those who share the risks with them. That consideration validates special concern for compatriots. But the acknowledgment of facts about the rationality and vulnerability of others cannot stop short at national borders without moral arbitrariness: so that acknowledgment, and


what it implies, constitutes a limit to what national associations can require. The Iteration Proviso tells us that we can rightly give special concern to compatriots only if (i) outsiders are, truly, in a position to distribute special concern among themselves and (ii) we do not prevent outsiders from seeking the advantages that we seek in the course of pursuing advantages for ourselves. The picture, then, is of a world made up of parallel social projects, each legitimately entitled to a certain degree of self-preoccupation, but each of which also contains an implicit recognition of the respect due to other projects. There are situations in which this implicit recognition needs to be made explicit and connected with political action. The high moral cost of failing to do so is that the ties binding compatriots together could no longer be supported in good faith. They would come to depend on an arbitrary view that outsiders are of no moral account at all; in other words, they would become supportable only on the basis of a radically partialist view that rejects any need at all for impersonal justification.37 Now the contexts in which that consideration comes into play may not be ever-present: there are many contexts of action in which spatially limited recognition of others’ interests is beyond moral reproach. So the transnational duties that this argument attempts to sustain do not challenge compatriot preference as such – in fact, I have argued, they spring from a set of considerations that actually validate it: the considerations, outlined above, that explain why contractually allied risk-takers can properly give special weight to each other’s vulnerability. What, though, concretely speaking, would this argument call upon states to do? The following three chapters explore this, but let me give a brief synopsis here. First, the duty of aid. This part of the argument leads to more determinate results. As we have seen, when 37

I do not believe that one can evade the requirements by embracing the radical partialist option, at least if one wishes to retain a case for political obligation, for there simply is no reason to think that the bounds of obligation marked out by that option will resemble the boundaries of states.


another society’s risk-taking project has manifestly not paid off – John Locke’s polecat-and-lion problem – our own state’s legitimacy depends on its contributing to a solution, by means of military intervention or humanitarian aid or international criminal prosecution of state-sponsored atrocity, or all three. As a first step, the argument would call upon states to incorporate crimes against humanity within their own domestic criminal codes: because their own security-seeking projects imply that others can pursue the same security-related ends, they cannot ignore situations in which the security of outsiders is attacked by the outsiders’ own states, and so must criminalize attacks by other state leaders upon their own populations. Since that act alone has ineffectiveness written all over it, they must take two further steps to render the intention effective: they must support an international criminal court that will operate in a more reliable way than a regime of “universal jurisdiction” by individual states acting independently, and they must support an international body that has powers of arrest, backing it up with the necessary military force.38 Secondly, the duty not to harm. This part of the argument seems to lead to less determinate conclusions. The argument claims to have shown that the members of strong states, to avoid hypocrisy, must not act in such a way as to deny equal opportunities for political and economic development to less favored societies. But there is no consensus as to what “equal opportunity” means in practice, or the extent to which different capacities need to be rebalanced in order to achieve it. Moreover, as states of various sizes and capacities pursue their various paths, there will be collisions of interest which will repeatedly pose the question of how much one state must do for another. How much economic advantage would rich states be required to forgo in avoiding damage to weaker economies? This issue is difficult. But perhaps there is an institutional solution. 38

This is, of course, merely a brief abstract statement of a view that needs much concrete elaboration. Among many excellent recent discussions, see J. L. Holzgrefe and Robert O. Keohane (eds.), Humanitarian Intervention: Ethical, Legal and Political Dilemmas (Cambridge: Cambridge University Press, 2003).


Some duties can be formulated in a way that can be codified and legally enforced, such as a duty on the part of state leaders not to commit atrocities and a duty on the part of other state leaders to restrain or punish or remove them. But in the case of domestic political society, no one has ever supposed that all duties are of that kind, and there is even less reason to think that transnational duties are all of that kind either. The natural remedy, in the domestic case, is the representation of interests. We may have only incomplete principles about how to distribute goods fairly among competing claimants; but a fair outcome is much more likely if the interests of all claimants are made present in the process. That is the elementary consideration that lies behind Locke’s requirement of an elected legislature, and also Mill’s case for a working-class franchise, not to mention later demands for an identity-based “politics of presence.” That seems no less natural a solution in the transnational case too. The abstract duty not to cause basic harm would translate concretely into a requirement on the part of rich states to join regulatory institutions in which the interests of poor states are made present: institutions that would rule on appropriate terms of trade, investment, and loan conditions. This is, I think, not just an ad hoc solution, but one that is continuous with the whole argument of this book. The interests of others are already implicitly embedded in the arrangements that political associations make for themselves, and the two sets of institutional requirements that have just been sketched simply take what is implicit and give it explicit force. The requirements are far from novel: but I hope to have shown that our liability to them rests on reasons that are closer to home than is usually thought, and that the social contract tradition has untapped resources for explaining them. It explains them as political obligations, 39 and not as claims that rival what we owe to compatriots. 39

James Bohman also attempts to support such a conclusion, using a “republican” rather than a contractualist approach: but I believe his argument is exposed to the same difficulties as other versions of interactionism. See James Bohman, “Republican Cosmopolitanism,” Journal of Political Philosophy 12 (2004): 336–52 (discussed in chapter 6 below).

Humanitarian intervention and the case for natural duty

5

We are far from done, however, with the idea that the solution to the issues discussed in this book is to be found in a direct appeal to cosmopolitan “natural duty,” an appeal that would make any contractualist apparatus (however understood) redundant. That idea fi nds very strong support. Martha Nussbaum’s view – that we must give up on the very idea of a social contract if we are to make sense of our obligations other than to citizens – has already been mentioned above. Jeremy Waldron has proposed that the “special duties” of citizenship are best explained as derivatives of natural duty, as necessary (not contractual) implications of justice-promoting arrangements.1 Allen Buchanan – one of whose proposals will be examined at length below – argues that only an appeal to a strong “natural duty of justice” can justify one society’s coming to the aid of another, and that contractualism implies moral isolationism. But perhaps the most basic case for a direct appeal to natural duty is offered by Liam Murphy in his critique of Rawls.2 Rawls maintains that we have a natural duty of justice only in relation to associative requirements that “apply to” us. Given that we are part of an institution (a state) that issues binding requirements upon us, we have a duty to assess the justice of its requirements. But suppose there are injustices – inequalities, let us say – that fall outside the scope of any institution’s responsibility; what does Rawls’ theory tell us then? Apparently – according to Murphy – it tells us that there is no moral reason to create responsible institutions, and only an idea of natural duty can do so. Given that we live in an unjust world, in which no 1 2

Waldron, “Special Ties and Natural Duties.” Liam Murphy, “Institutions and the Demands of Justice,” Philosophy and Public Affairs 27 (1999): 251–91.


institution is charged with remedying injustice, how can we possibly arrive at the view that institutions so charged should be created, without appealing to a pre-institutional idea of what we owe regardless of the institutions that we happen to have right now? By now it should be clear that this book attempts, in effect, to answer Murphy’s question. Our contractual arrangements contain implications for responsibilities beyond our borders, both in the sense that they presuppose a cosmopolitan moral regard, and in the sense that their exclusiveness generates special concern for the excluded. Humanitarian intervention offers an important test to this view. It is true that, according to some, the international community already has a regime in place that authorizes intervention, so that duty of aid is not, as it were, in a pre-institutional state, or phase. 3 But that judgment is highly contested. There are those who believe, rather, that interventions are and should be unusual exceptions to a general regime of non-intervention and respect for national sovereignty – or even that these interventions should remain illegal even when they are legitimate. The issue between these views is a complex one that reaches far into the question of the very basis of international legality.4 But whatever our view on that issue, surely it is clear that we are very far from moral consensus on the matter, and that the rightness of intervention is the subject of continuous questioning. In that regard, it is proper to take it as a test case for the moral scope of contractualism. In a world that is not (yet) just, do we have to appeal to natural duty if we want to make it more so?

I The starting-point for this discussion is a very blunt question: why should members of one society accept great costs for the benefit of 3

4

See especially Fernando Tesón, “The Liberal Case for Humanitarian Intervention,” in Humanitarian Intervention, ed. J. L. Holzgrefe and Robert O. Keohane (Cambridge: Cambridge University Press, 2003), 93–129. For a more critical view, see Simon Chesterman, Just War or Just Peace? Humanitarian Intervention and International Law (Oxford: Oxford University Press, 2001).

Humanitarian intervention 119

outsiders? And even if some members of the society would accept them, why is it legitimate for them to impose those costs, imperatively, upon their compatriots? Humanitarian intervention poses that problem in its most severe form, for it raises the issue of insider versus outsider interests in so stark a way. Taking the hardest possible line on the matter, realists such as Samuel Huntington and Henry Kissinger have denied that citizens of one country should be required to accept any such costs at all. Huntington remarked, in the context of the US intervention in Somalia, that it was “morally unjustifiable” (as well as “politically indefensible”) that US citizens should be called upon to give their lives for strangers;5 and Kissinger once declared it entirely unusual (and by implication absurd) that “moral and humane concerns” for others should be thought to justify expending “not only treasure but lives.”6 Suppose we believe, though, that there are cases where such expenditure is called for – and let us leave aside here the very complex question of how such cases should be defi ned and by whose authority – how could we meet views such as Huntington’s and Kissinger’s? This issue is surprisingly under-explored. There has been much recent discussion (in Canada especially) of the “Responsibility to protect” (International Commission on Intervention and State Sovereignty), a doctrine now inching its way uncertainly onto the United Nations agenda. However, discussion of that responsibility has focused principally on the reasons for regarding sovereignty as conditional – the reasons, that is, why an abusive state loses the protection of sovereignty and becomes open to intervention by other states. That part of the case has been made well and in many ways: it has been argued, for example, that sovereignty is not a sort of natural feature of states but a status conferred by the international

5

6

Quoted in Michael Smith, “Humanitarian Intervention: An Overview of the Ethical Issues,” Ethics and International Affairs 1 (1998): 63. Quoted in J. Bryan Hehir, “Military Intervention and National Sovereignty,” in Hard Choices: Moral Dilemmas in Humanitarian Intervention, ed. Jonathan Moore (Lanham, MD: Rowman and Littlefield, 1998), 36.


community, a conferral that it would plainly be wrong to regard as unconditional and which would be lost if the (implicit) conditions of conferral were violated.7 But the second half of the argument – a necessary accompaniment to the first – has been very much less studied: that is, why members of the intervening states must carry the cost (in terms of “life and treasure”) of rescuing the victims of other states’ abuse. This has always been a problematic issue. Grotius himself, in a passage that is often seen as offering the fi rst (although very qualified) defense of humanitarian intervention, explained that while people must not overthrow their own abusive princes, other Good Samaritan princes can do so on their behalf, because they have never promised to obey anyone and so retain a sort of general power of guardianship.8 However, this argument from the special status of princes leaves us to wonder why the subjects of Good Samaritan princes must bear the costs of their Good Samaritanism, as taxpayers, soldiers in their armies and soldiers’ relatives. (“Because their princes tell them to” is one possible answer, of course, although hardly a usable one these days.) In a ground-breaking article – one of the very few recent discussions of the topic – Allen Buchanan called this the problem of “internal legitimacy” (as distinct from the much discussed external legitimacy issue). His discussion sought to show that the problem is fatal for the contractualist tradition.9 That, he said, is because a social contract is an arrangement for mutual advantage, in consequence of which the contractors acquire obligations to one another, but not to anyone else. It would, in fact, actually be wrong for their 7

8

9

See for example Henry Shue, “Limiting Sovereignty,” in Humanitarian Intervention and International Relations, ed. Jennifer Welsh (Oxford: Oxford University Press, 2004), 11–28. Hugo Grotius, The Law of War and Peace (Indianapolis: Bobbs-Merrill, 1975), 584. Allen Buchanan, “The Internal Legitimacy of Humanitarian Intervention,” Journal of Political Philosophy 7 (1999): 71–87. Buchanan’s view is further developed in chapter 2 of his Justice, Legitimacy, and Self-Determination (Oxford: Oxford University Press, 2004), but not, as far as I can see, in a way that modifies the claims discussed here.


government to require them to expend resources for some outsider’s benefit. So if we believe there are cases where humanitarian intervention is justified, or even permissible, we must abandon contractual models for an idea of natural duty. We must suppose that people have a natural duty to support just institutions and delegate this duty to their governments; otherwise we will be driven to adopt a pure “compatriotist” morality, as we may call it, accompanied by pure realism in foreign policy. Here it will be argued, however, that those who reject realism will fi nd that an argument of this type relies on unpromisingly strong premises and yields unsatisfactorily indeterminate conclusions, and that an approach to justice triggered by contractual agency is preferable in both of those respects.

II Many people would probably accept that there is something that could be called, optionally, a natural duty of justice. But what degree or level of natural duty are we to assign to people? Do we have a duty, for example, only to regulate our own dealings with just principles or do we have a duty to see that everyone’s dealings are justly regulated? Buchanan’s article helpfully proposes three possible versions of natural duty, ranging from NDJ1 (a duty to contribute to just arrangements with fellow citizens) through NDJ2 (a duty to contribute to just arrangements with all those with whom one interacts) to NDJ3: “Each person has a duty to contribute to the inclusion of all persons in just arrangements,” this being subject to the proviso “if one can do so without excessive cost.”10 Only NDJ3 is said to give us a solution to the internal legitimacy problem. Now, if true, it would certainly solve that problem. But is it true? It seems to face two kinds of difficulty. The fi rst is that there are minimal and maximal ways to understand it, with no apparent way to decide which is the one that we 10

Buchanan, “Internal Legitimacy,” 83.


should settle upon. A minimal reading is offered by another seminal article on natural duty, by Jeremy Waldron, to which Buchanan’s article refers in support.11 Waldron tackles head-on the very problem of reconciling the idea of natural duty and the requirements of a world of exclusive citizenships. He begins with the idea that there is a natural duty to support just institutions. He accepts that (as Kant and others have argued) there are good reasons to fulfi ll this duty on a relatively local scale. He points out that the duty to support local institutions will be refracted, as it were, into a number of secondary duties designed to make local institutions workable. Among these secondary duties there will have to be a duty not to obstruct the institutions’ working. And these, by virtue of the fi rst premise of the argument, will bind outsiders as well as insiders. So we can explain why, for example, it was wrong for the French government not just to sink a Greenpeace vessel in Auckland harbour but also to obstruct New Zealand’s attempt to prosecute the perpetrators. Now that is an excellent argument, but it only tells us not to harm (or obstruct), not to aid. One could of course claim that negative duties not to harm inevitably turn into duties to aid as well – the duty not to assault gives rise to a duty to aid the assaulted and punish the assaulters.12 That too is a very good argument, but it does not show, by itself, that all three duties must fall to the same agent. I certainly have a duty not to assault people anywhere in the world, but while someone should have the other two correlative duties, it (equally certainly) is not me. So the claim that states have a negative duty does not show, by itself, that they bear the other two correlative duties as well. On the other hand, we can take the natural duty requirement to be so strong that it virtually obliterates frontiers. This is because of the considerable weight that the very conditional argument of relative cost is made to bear. We have a spatially unlimited natural duty to support just institutions,

11 12

Waldron, “Special Ties and Natural Duties.” Buchanan, “Internal Legitimacy,” 80–81.


wherever. In practice, though, we support local ones for the most part, because the duty contains the proviso that it is limited by the costs that we have to bear in fulfilling it. That proviso is meant to direct us to relatively local duties, in the fi rst instance at least, on the grounds that the cost of fulfi lling duties on a global scale is very high: that is why we can rightly observe local needs fi rst.13 Now we can think of many reasons why the cost of aiding locally is smaller: proximity and existing institutional connections make local aid more effective, cross-cultural judgments are perilous, local knowledge makes it more likely both that we will understand problems and be effective in solving them, and so on; but while such reasons are by no means negligible, they are all too easily defeasible and give us only provisional results. (In a world of diasporas, for example, local knowledge is quite widely dispersed.) Moreover, the consideration that distant intervention is restrained only by its cost makes independent statehood hang by a slender thread indeed, and a world in which resources were matched with need only with regard to avoiding excessive cost would bear little resemblance, organizationally, to the world that we live in. I do not mean to criticize that alternative world, just to point out the maximal implications of the principle: it makes even prima facie compatriot preference hard to support, and it sanctions measures far more invasive than humanitarian intervention, opening boundaries to other kinds of penetration whenever cost permitted. From a conceptual point of view, the principle succeeds in justifying intervention at the (paradoxical) risk of undermining the moral legitimacy of the states on which the burden of intervention would actually fall. From a political point of view, the principle is so demanding that it actually risks discouraging, rather than sustaining, humanitarian interventions: that (some) demands for intervention are urgent is more generally compelling, surely, than the view that the state system can be justified in the way proposed. 13

Ibid., 86.


To sum up so far, then, I see no way to stabilize the natural duty argument between unsatisfactorily minimal and radically maximal claims. It gives us either less (as a negative duty) or more (as a positive duty) than we need, thus lending support to Rawls’ view that if we build universal altruism in at the base of the theory we cannot give others any reason to support the particular level of altruism that we happen to favor.14 And the internal legitimacy problem, we may note, exactly concerns the reasons that we can give to others in support of demands that are to be made upon them. The second objection is prompted by a thought-provoking article by Michael Green.15 We may place his discussion in the context of a dispute between theorists (such as Rawls) who see a difference between personal duties and our duties as citizens and others16 who deny that there can be any fundamental difference. Into this dispute, Green injects a useful articulation of the concrete differences between persons and institutions. Common-sense views of morality fi nd individual responsibility easier to handle than collective responsibility; they fi nd acts to be more weighty than omissions and they fi nd special responsibilities (among compatriots, for example) more compelling than obligations to unspecified others. For all three reasons common-sense views of morality falter in the face of largescale processes in which the consequences of acts are remote from the agent – a category which happens to embrace most current global problems. However, institutions such as states are not subject to the constraints that excuse individuals: they can collect information about the distant effects of actions; they can change incentives so as to modify mass behavior; they can distribute the costs of remedies so that no one pays too high a cost for forbearing from distantly damaging actions. It follows that institutions such as states can be held 14 15

16

John Rawls, A Theory of Justice (Cambridge, MA: Belknap Press, 1971) 148–49. Michael Green, “Institutional Responsibility for Moral Problems,” in Global Responsibilities: Who Must Deliver on Human Rights? ed. Andrew Kuper (New York: Routledge, 2005). Murphy, “Institutions”; G.A. Cohen, If You’re an Egalitarian, How Come You’re So Rich? (Cambridge, MA: Harvard University Press, 2000), especially chapter 10.


to have responsibilities that their individual members do not have and that the duties their individual members have cannot plausibly be called “natural.” Do these “institutionalist” considerations deal a fatal blow to the natural duty argument? The answer seems to me to be quite mixed – “yes,” in some respects, “maybe” in others, and “perhaps not” in others still. “Yes,” it undermines the natural duty argument by showing that, by virtue of their membership in political societies, people may (by implication) sign up to obligations that as individuals they could not possibly have. The topic of humanitarian intervention (not, I should mention, Green’s topic) makes this point particularly well. When it comes to distributive matters such as global poverty relief, individual and institutional initiatives are on a single spectrum, as it were: I can give a little to Oxfam, my state can give a lot more to overseas development. It is just a matter of large and small. However, when it comes to humanitarian intervention there is no corresponding personal contribution to place on the spectrum.17 (In the days when aristocrats raised private armies there would have been, for aristocrats.) Natural duty theorists could of course meet the objection by stating the duty concerned in sufficiently abstract terms – a duty to aid all humans, say. However, I see no way that such a duty could predictably survive competition against all the other duties that our engagement with the world gives rise to, in a way that would guarantee that it would be action-guiding or that, if it guided action, any particular level of action would be dictated. “Maybe,” because, as Green himself scrupulously points out,18 the argument may be incomplete. While institutional capacity can be said to bring new responsibilities into play, responsibilities that “natural” individuals would not have, it may seem to involve a debatable “can implies ought” premise: just because states and their 17

18

The difference is between “absolute” and “discrimination” thresholds, as they have been called: see Jonathan Glover, “It Makes No Difference Whether or Not I Do It,” Proceedings of the Aristotelian Society, supp. vol. 49 (1975): 171–90. Green, “Institutional Responsibility,” 129.


members have the capacity to aid abused populations, can they be said to have a duty? I think we can reinforce the worry here with two further considerations. First, many states, in fact most of them, lack some basic elements of the capacity for effective intervention (such as, at the time of writing, the type of expensive heavy-lift aircraft that effective rapid deployment demands). If we want them to acquire these elements it is not clear that we can do so on the basis of the capacity argument itself. (We could say, perhaps, that they have the capacity to acquire more capacity, but this seems to leave the argument looking a bit stretched – after all, anyone has the capacity to acquire more capacity.) Secondly, the argument of capacity does not pick out states uniquely. Many corporations have more resources than many states and the argument would thus seem to license a sort of corporate vigilantism which would be worrying, even if it turned out (remarkably) to be benign. For these reasons we need, in addition to capacity, an argument that brings the distinctive normative features of states into play. “Perhaps not” because the obligations of states require corresponding obligations on their members’ part. Meeting obligations to relieve poverty or attack oppression in other countries requires that states should be empowered to compel sacrifices on the part of their own members. We cannot rely on a “default” argument to ascribe a duty to states: not only because of the possibility that, although a need is great, we cannot assume that it is anyone’s responsibility to meet it, but also because it seems an important achievement of early modern political thought to have distinguished between what states could potentially do and what they have authority to require.19 So, do state obligation and individual obligation eventually come to the same thing? Here the difference between the two arguments may seem to be rather thin. Green’s article claims that since the emergence of institutions brings into play new responsibilities

19

For this argument, see Richard Vernon, The Career of Toleration: John Locke, Jonas Proast, and After (Montreal: McGill-Queen’s University Press, 1997).


that were not there before, it makes sense to speak of “institutional responsibilities,” even if one grants the obvious point that institutions are made up of and supported by people.20 Buchanan’s article, on the other hand, offers us a view of the powers of institutions as “enhancing” the natural duties that individuals previously had when they commanded only individual resources.21 The fi rst view is that states have distinctive responsibilities (which their members support because they are citizens); the other is that as humans we have duties that for practical reasons we delegate to states and which states, by virtue of their greater capacity, then magnify. This, as it stands, looks like a tie. Because the model of enhancement or magnification is available, I do not think we can say that the natural duty approach is disproved. How we settle this part of the issue will have to rest on our assessment of the claims of the natural duty argument overall in relation to alternative approaches.

III Buchanan’s case for the superior claims of natural duty is made in part disjunctively, that is, by showing that the contractualist case for conventional duty produces (what we are assuming here to be) unwelcome results: convention failing, nature is the only recourse left, by elimination. That is because, as noted above, his view of the contract – like many other views – depicts it as an arrangement for mutual advantage. The contractors agree to confer benefits on each other, acquiring obligations in the process, obligations which obviously do not extend to those who are outside the arrangement. The state is understood as the creation of a hypothetical contract among those who are to be its citizens, and the terms of the contract they agree on are justified by showing how observance of those terms serves their interests.22

20 22

Ibid., 127–28. Ibid., 74–75.

21

Buchanan, “Internal Legitimacy,” 85.


Principles of justice “are thought of as specifying the terms of cooperation among those who are bound together in one political society, rather than as specifying how persons generally must be treated.” Government is understood only as an agency dedicated to enforcing those terms of cooperation. Such an approach, Buchanan said, owes its popularity to many real advantages, which he listed: in particular it presents government as a servant of citizens’ interests, not as a body with distinct interests of its own. However, its shortcomings show up sharply in global ethics. In the contractual model “a government that engages in … pure humanitarian intervention [i.e. intervention in which its own citizens’ interests are not at stake] violates its fiduciary intervention: it fails to act in the best interest of its citizens.” To sustain the validity of (pure) humanitarian intervention we must adopt a more inclusive idea of the scope of justice: the only candidate principle that will do the work, as we have seen, is said to be that “Each person has a duty to contribute to the inclusion of all persons in just arrangements” (if able to do so, it should again be noted, without excessive cost). In the light of this principle the state is reconceptualized as an instrument of justice, not as an association for mutual advantage.23 Locke’s contractualism is taken as a reference point for this case, so let us also take that as a fi rst example. We do not need to delve deeply into textual analysis to recognize what this interpretation leaves in the margin, i.e. the background of pre-existing moral entitlement that the contractual model depends upon. First – unlike “contractarian” views from which it was distinguished above – it amounts, conceptually, to the waiving of a pre-existing moral right, i.e. a right to interpret and enforce the law of nature unilaterally, an idea that is of course literally unintelligible without the precontractual moral apparatus on which Locke relied. Secondly, it is presented explicitly in terms of the pursuit of justice. Locke’s contractors are driven by an imperative of general preservation limited only by the right to preserve oneself. 23

Ibid., 83.


They recognize a violation of natural law as “a trespass against the whole species” or an act of “war against all mankind” which anyone (in the state of nature) may punish “by the right he hath to preserve mankind in general,” a right which is “distinct” from the right to preserve oneself (Second Treatise, s.8). Of course, the idea of advantage-seeking does not come from nowhere: but the advantages sought are advantages sought (it is implied) by all humans and to represent the process as self-centered is to distort the motivation that Locke attributes to contractors. They seek common advantages, even though the right of self-preservation dictates that they need assurance that others will be obliged to cooperate in the process of seeking them. Calculations of self-interest come into play, it is true, in those passages in which Locke considered the terms of association that would be agreed to (ss. 50, 72–73, 93, 137). However, his discussion is concerned to exclude what “no man” would agree to, i.e. the interest that any person would have, and the appeal is not to personal or exclusive self-interest but to the interest of typical persons. So self-interest comes into play, fi rst, because we need not act in accordance with justice when the personal cost is, literally, lethal and, secondly, as a way of inducing reflection on shared or typical interests. Nor is all this, as some might suspect, a peculiarity of a model based in natural law . Much the same could be said of Rawls, for example, who employed his own version of the contract in interpreting the background entitlement that he assumed, i.e. an entitlement to equal moral consideration. First, with regard to security, Rawls accepted an account of what he was doing as something midway between mutual advantage and pure impartiality. Mutual advantage, on the one hand, defines justice in terms of reciprocal exchange. Pure impartiality, on the other hand, dispenses with any kind of reciprocity. Rawls, however, wanted us to think of his model as one in which we agree to treat others on a fair basis provided that we have an assurance that others will behave fairly too. Secondly, with regard to the practical definition of duty, Rawls offered a reason for employing self-interest, rather


than universal altruism, as a device in the original position.24 What it means to do good to others depends on knowledge that we cannot hope to command – we do not know “the relative strength” of benevolent desires (relative to each other, to non-benevolent ones or both?) and the assumption of (indefinite) benevolence is too strong to be useful in political argument. So, in both respects the appeal to self-interest, exactly as in Locke, is contained within a project for giving background morality (1) effective and (2) more definite political expression. This is both an historically more accurate and a more plausible account of what contractualists offer than the mutual advantage account gives. And it avoids the awkward conclusion arising from the mutual advantage view that no actual contractualist theorists are real contractualists, because they rely upon independent moral principles. What happens, though, to the background principles after that – to the initial universalism that the argument not only assumes but requires? One, seemingly implausible, view is that it simply vanishes. A Lockean contract, it has been argued, leads directly to realism in international ethics, sealing outsiders off from moral concern.25 That view is supported in part by reference to the self-preservation principle, which is said to transfer from the personal to the societal case, and in part by reference to Locke’s silence about issues of global morality. However, fi rst, the self-preservation principle is very weak. It tells us only that our duty to others does not entail our own destruction, a proviso that any moral view would surely acknowledge. (It is in fact weaker than Buchanan’s own “excessive cost” constraint on natural duty.) And if it offers only a weak restraint on justice in the personal case, it offers an even weaker restraint on justice in the national case, since nations are far less likely to face extinction than are individual persons, as Charles Beitz has pointed out in his critique of person–society analogies.26 So the fact that social contracts may confer a group right of self-preservation on 24 25 26

Rawls, Theory of Justice, 148–49. See Richard H. Cox, Locke on War and Peace (Oxford: Clarendon, 1960). Beitz, Political Theory and International Relations, 52–53.


political societies is a very long way from the realist view of global conduct. Secondly, whatever the merit in general of arguments from silence, they are unconvincing in cases in which silence has a ready alternative explanation. This is such a case. Locke obviously assumed that the process of contractual separation will be iterated. The argument is a typical one. Others will create their own political societies. So the simplest of reasons for not mentioning them is that the argument assumes that others will take parallel steps to secure their own collective preservation. In this way the duty to prevent or punish “trespasses against mankind” will be parceled out among political societies, each of which functions, as it were, as a proxy for “mankind.”27 Confronted with what we might term Scheffler’s question – what gives you the right to an exclusive association, which both diverts your care from outsiders and increases the level of benefit for those inside?28 – they can reply, simply, that others can do so too. It must of course be possible to make that reply in good faith, and here the argument may take a further step. There is an unsurprising self-preservation proviso built into the argument; but if a political society based its survival on contempt for (let alone assaults on) the survival of others, then the initial moral premises needed by the argument would be violated, since they require preservation of “mankind” as a whole. The real-world conditions of state failure and state abuse correspond exactly to those hypothetical conditions. They effectively prevent populations from seeking the conditions for their own preservation. This means that the assumption of iteration cannot be made in such cases and that compatriots lose the good-faith defense for their exclusiveness. They cannot justify the selective conferring of benefits on the grounds that those excluded can make parallel arrangements of their own in cases

27

28

This may go some way to resolving the tension between “man” and “citizen” that Kimberly Hutchings detects in Locke: “Political Theory and Cosmopolitan Citizenship,” in Cosmopolitan Citizenship, ed. Kimberly Hutchings and Roland Dannreuther (Houndmills: Macmillan, 1999), 13. Scheffler, Boundaries and Allegiances, 56–64.


in which some of those excluded are being forcibly prevented from doing so. A duty to aid in those cases if one can is therefore built into this justification of compatriot preference. Now this approach faces an obvious objection. If we have to begin with precontractual, natural, or generic humanist premises why not then build directly on those, instead of adopting an approach mediated through contractualism? Why not “Grotius” rather than “Locke,” as Nussbaum, as we have seen, recommends? There are two reasons for not doing so. The first is that contractualism has an important feature additional to those that Buchanan’s article lists. This is its appeal to immanence, or what was termed “complicity” in chapter 3 above. It appeals to the fact that citizens are actually engaged in ongoing relations of interdependence and makes only the assumption that they are interested in knowing whether or not their terms of association can be justified. It tries to draw out non-obvious conclusions from what they would obviously regard as fair and in this way to confront them with what could reasonably be described as implications of their own conduct. They should accept them, if the argument works, because they already depend upon them, if only implicitly, to justify the costs that are imposed by common arrangements upon their fellow citizens. The argument requires no more demanding a starting-point than domestic political theory requires – an openness to questions about what we owe to those around us. It requires no more remote a standpoint than that of critical reflection on one’s own situation. The second reason is that natural duty views can make no such argumentative appeal to immanence. Their appeal, in contrast, is somewhat discretionary. Is it even possible to give an argument for them to anyone who does not already find them intuitively compelling, as Buchanan acknowledges?29 Many, even most, people would surely find a principle of human respect compelling in some form. They would, for example, be likely to accept – at least – that all human beings, and perhaps other living beings too, fall within the scope of 29

Buchanan, “Internal Legitimacy,” 83–84.


our moral regard, in the sense that their interests cannot merely be discounted if they happen to conflict in any way with our own. Others are vulnerable to pain; they are subjects of a life. They are not mere things to be used at will. While that understanding is basic, indispensable, and powerful, it stands at some distance from conclusions about actions that must be taken, or specifiable duties. Here I think Rawls is right, if I correctly understand the passage quoted from him above. Such attitudes have an insecure place given that they compete not only with one another but also with imperative duties, and we need a mediating process (such as a contractual approach sets out to provide) to connect them with what we ought to do. Could it be objected that Rawls’ concern in that passage is with something properly called “benevolence” – and hence optional – rather than with “respect” – a matter of justice?30 The distinction is important, but even if we build in that stronger notion at the base I do not see that the problem of indeterminacy is resolved. Respect for someone amounts to a demand that their interests be taken seriously, not that their claims be automatically met – I disrespect you not by refusing you what you demand, but by not listening to and weighing the reasons for your demanding it, by treating you as inconsequential – it is a process term, not outcome-dependent. 31 In the context of competing duties – such as contexts in which duties to compatriots compete with duties to outsiders, as in the case at hand – the principle of respect can operate only with procedural rather than conclusive force. If so, then the need for a mediating process remains, as argued above.

IV The commitment of “life” (as distinct from “treasure”) is of course a particularly serious matter and one that has given rise to a sub-theme 30

31

See Terry Nardin, “International Political Theory and the Question of Justice,” International Affairs 82 (2006): 449–65. See Ronald Dworkin, A Matter of Principle (Cambridge, MA: Harvard University Press, 1985).


within the small literature on the internal legitimacy problem. Huntington’s view, quoted in the opening paragraph above, is that no loss of life for the sake of outsiders is acceptable. Others, however, accept the loss of compatriots’ lives, but with a major restriction. Fernando Tesón, in an otherwise forthright defense of humanitarian intervention, proscribes the use of conscripted troops in carrying it out.32 The argument rests on a distinction between wars of national defense and wars in pursuit of justice. A war of national defense, it is claimed, is a war to defend all citizens’ interests, and like all pursuits of collective interest it poses the problem of dealing with the free-rider. Hence the legitimacy of conscription, to compel people to do what it is in their own interest to do as long as they can be assured that others will be compelled too. A proposal for humanitarian intervention, however, rests on claims about what justice requires and on that matter citizens may have genuine disagreements: it would be unjust to compel them to serve an end whose justice they denied. Professional troops, however, have a contractual obligation that might reasonably be thought to outweigh personal judgments about the aims of a war (important issues of international legality aside). This view would seem to lead to quite controversial results. It would apparently show that, to the extent that US participation in the Second World War was based on justice or even sustained by an appeal to it, it was wrong for the US government to commit conscripted troops to the European (as opposed to the Japanese) theater. Nevertheless, it fi nds support elsewhere. Michael Walzer proposed a similar view.33 David Miller suggested that “a duty to respect the rights of compatriots – their right to liberty which conscription denies – decisively outweighs the duty to protect the rights of foreigners under threat from third parties,” whereas if a professional army is committed for that purpose that only raises issues about

32 33

Teson, “The Liberal Case,” 123–28. Michael Walzer, “The Argument about Humanitarian Intervention,” Dissent 49 (2002): 29–37.


the relative allocation of resources for foreign and domestic purposes (i.e. not an issue about compatriots’ rights).34 This discussion has not even touched on the question of whether it is right for governments to impose conscription for any purpose, and so cannot give anything like a rounded treatment to this issue. However, the argument that it has advanced must clearly work against the distinctions made in the previous paragraph, in presenting the duty to support humanitarian intervention as a membership duty as opposed to an optional and generous one. If conscription is justified for any purpose – and the argument is agnostic on that – it is justified as much by humanitarian purposes as by defensive ones – to the extent that (as argued) the former purposes are implicit in a social contract that makes the latter purposes explicit. The following considerations would seem to support that conclusion. First, the duty to intervene is said to be “imperfect” in that it does not fall to any particular state, and in that, for any particular state, there are likely to be more occasions than it can respond to. So there cannot be a duty to intervene in any particular case. Walzer’s view is that the imperfectness of the obligation passes down from state to citizen, so that each citizen – like each state – can exercise discretion. We may suppose professional soldiers to have waived this discretion by making a personal contract with the state, as citizens subject to a military draft have not. This does not seem plausible. Each individual, lacking the personal capacity to intervene, can never make good on his imperfect obligation except as a member of a state, and if the exercise of state discretion can be cancelled by individual discretion in this way then the obligation can never be acted upon. Secondly, it makes a difference, of course, that in the case of humanitarian intervention the support that citizens are called on to give entails risk of life, which, we may say, can reasonably be refused

34

David Miller, “Reasonable Partiality Towards Compatriots,” Ethical Theory and Moral Practice 8 (2005): 63–81.


in a way that (say) taxation demands cannot. That could be developed into an argument against conscription for any purpose. But can it be used to discriminate between a draft for the purpose of national defense and a draft for the purpose of humanitarian intervention? The case for believing that it can rests on the issue of free-riding. A war of national defense is in everyone’s personal self-interest, but to prevent some from free-riding on the willingness of others to contribute at the risk of their lives, a draft is a just solution to a collective action problem. The same argument is not available in the case of humanitarian intervention because we cannot in the same way invoke personal self-interest. But the consideration works only if one assumes to begin with that the sole purpose of the state is to solve collective action problems. Even on that basis, it implausibly assumes that in every case the defense of a national interest is valued by all individuals to the point of being willing to risk their lives, provided only that others do too. Thirdly, it cannot be supposed that an appeal to consensus is more available in the case of a war justified on defensive grounds than in cases of proposed intervention. It is hardly true that issues about national defense, unlike issues about justice, are uncontentious: the point needs little emphasis at the time of writing, when citizens of the USA are deeply divided over whether or not actions taken in the name of their security interests actually advance those interests or, on the contrary, tend to defeat them. Difficulties about the government’s goodness of judgment apply across the board, not just to humanitarian issues, and conscripted soldiers may be made to go against their own best judgment in either case. Being made to take part in a war spuriously justified as defensive would amount to being made to violate one’s sense of justice, for the damage done by such a war could not satisfy either the “right intention” or the “proportionality” criterion of justice in war. (All unjustified wars cause disproportionate damage.) Moreover, on the version of the contractual view adopted here, unlike the mutual-advantage version of it, advancing the national interest will be constrained by judgments


of general justice, and so what is or is not to be taken to be a valid interest is just as open to contention as any other policy. The risking of any soldier’s life, professional or conscripted, is a serious matter; the risking of a conscripted soldier’s life – if conscription is permissible – may be an even more serious matter, given the absence of consent, but there is no distinction of the kind proposed for the use of troops in humanitarian and defensive wars.

V This chapter has argued that people acquire external obligations by virtue of, not despite, their membership in political associations and that their “responsibility to protect” outsiders is the counterpart to their entitlement to exclude. Four claims have been made. (1) The social contract model depicts a hypothetical process by which spatially limited zones of obligation come into being, over and above whatever regard human beings owe to one another as human beings. These zones make it possible for those within them to develop complex kinds of interdependent relationships from which they benefit. Those outside do not benefit from them. (2) However, since those outside are not beyond the pale of moral regard – for anyone who accepts even the weakest of humanisms, of whatever kind – what was termed “Scheffler’s question” is posed: why do outsiders get less than insiders? The answer – iterative contractualism – is that they too can become insiders: inside their own political society, that is. To the extent that compatriots can honestly give that answer, their preference for each other’s interests is morally validated. (3) But that answer can be honestly given only to the extent that outsiders are also in a position to enjoy contexts in which they too can create flourishing civil societies. It cannot be honestly given in relation to those cases in which outsiders suffer the violent effects of political collapse or the violent oppression of abusive states, where the capacity to resist state terror or to organize for collective self-preservation has been lost. (4) So


in such cases citizens of successful societies can justify their own enjoyment of benefits only if they are willing to go to the aid of the victims of failed or abusive states. To be sure, the whole argument depends on an initial assumption that human beings deserve moral regard, a notion that can perhaps be worked up into a theory of “natural duty” that would circumvent the whole contractual logic (although this discussion has resisted the equation of “moral regard” with “duty”). However, while such a theory may be attractive to some, it is indeterminate in what it requires; it competes, with very uncertain outcome, against strong demands stemming from local duty; perhaps most important, it cannot claim to be grounded in an immanent way in the arrangements that compatriots make among themselves, since it stands, rather, in rivalry to them. That certain global obligations may rest on the same grounds as domestic ones does not, of course, show that they are of equal strength or equal in priority. Even if, as claimed above, contractualist arguments can be shown to entail obligations to non-contractors, it must surely be the case that in any sort of contractualist framework obligations to non-contractors come fi rst. That need not be because, as in more literal versions of the doctrine, we regard obligations as arising from an express or tacit promise that occurs among citizens only; it is also the case in the hypothetical version of the doctrine adopted here, in which a contract is imagined only as a test of the justifiability of shared institutions. Unless those bound by its requirements have some special relationship distinguishing them from others the contract would be vacuous, even as a hypothetical event. However, the argument developed here can, I believe, accommodate that basic fact and make sense of it. For, first, its appeal is to those who derive benefits from successful political associations, with a large margin of resources beyond security provision, and what it calls for is assistance in establishing the minimum conditions for political association by others. It is entirely consistent, therefore, with a policy that preserves the benefits of success for members of the intervening society, since it is exactly the


enjoyment of those benefits that constitutes the grounds on which the duty to assist is based. Nothing in the argument calls for equality at the distributive level and members of successful societies can justly continue to expend resources more generously on their own behalf than on that of others. It may also be the case that exploring the implications of contractualism will lead to further demands of a redistributive kind that constrain economic self-preference, but that topic is left for chapter 7 below. As far as humanitarian intervention goes, the argument here construes it as an obligation on the part of the successful to those whose political arrangements have failed, and nothing in it requires the bearers of the obligation to undermine the basis of their own success. As for the question of how much they would be required to sacrifice, I see no way to answer that at the level of analysis adopted here. Suppose interventionary assistance is, as argued here, a civic obligation: within civic obligation there are disputes about the demands of justice that cannot be settled simply by theories of what civic obligation is, and there is no more reason to think that the argument can determine the appropriate costs of intervention than to think that it can determine the appropriate costs of, say, domestic policing. There will be disputes. The point of this discussion has only been to show that the appropriate allocation for interventionary assistance must be on the list of priorities that citizens dispute. It cannot be zero. A society that arrived at zero allocation would be, morally speaking, like a society that failed to make any provision at all for its own members’ basic security needs. Secondly, there may be cases – for example, and especially, post-colonial ones – in which one country bears special causal responsibility for disasters in another, and so has a special moral responsibility to respond. In the standard case, however, no one country is picked out (by this argument) as the appropriate intervenor, and any moral responsibility to intervene will be diffuse. In that respect the obligation could be said to be an imperfect one. To whom then does it fall? The answer is that, special cases aside, the obligation to intervene falls to no one. Now this is not an unfamiliar


problem. In the domestic case we are accustomed to the idea that imperfect obligations, when important, should be made perfect by institutional means. When no one in particular bears responsibility for meeting the essential needs of other citizens, then special bodies are charged with the responsibility of meeting them. Again, then, if some responsibilities beyond borders are on the same footing as responsibilities within, the duty to respond translates into a duty to create institutions that can effectively carry them out. Since most states have virtually no independent capacity to intervene and no states (even the most powerful) have the capacity to intervene in all justifiable cases, the only solution is a multilateral one, i.e. the duty translates into a duty to support collective institutions with the appropriate capacity, and subject to appropriate safeguards, to respond to state failure and state abuse. It may at fi rst seem that we have been here before, in connection with “natural duty” arguments of the kind that were questioned above. Have we not simply resurrected Buchanan’s NDJ3, i.e. the view that “Each person has a duty to contribute to the inclusion of all persons in just arrangements”? The difference is the difference that is made by the contractual mediation of the argument. Certainly, contractual arguments must begin with an idea of what is due to “all persons” in their capacity as persons, for without that beginning, as noted above, no contract (express, tacit, or hypothetical) is conceivable. However, we cannot go straight from there to a duty to intervene in other societies’ arrangements, for reasons outlined above. We might go straight from there to a right to punish “offenders,” following Locke’s indications, and thus to the establishment of an effective system of international law that would hold individual “trespassers” accountable. We can get to the much more demanding idea of a duty to support intervention, however, only by taking a further step. That step is the exclusionary one, also noted above, by which we form associations that confer benefits selectively upon their members while implicitly relying upon moral conceptions, such as equality of respect, that are universal in their scope.


Its exclusionary nature is justifiable, given the security-based need to rely on others’ observance of the rules that one adopts for oneself: but those who are excluded are not thereby stripped, absurdly, of basic respect, and basic respect for them has the implications described above. They too must have the right to form exclusionary associations for security-based reasons, and when they cannot do so we can be called to account for our own security-based exclusions.

VI This discussion must end with two caveats. At the time of writing any defense of humanitarian intervention is likely to be confused with something else, particularly an ideology that calls for the dissemination of democracy. So the fi rst caveat is that nothing in the argument above licenses the attempt to impose political regimes, even good ones, on other countries: it is an argument enabling other countries to be placed in a position to make a choice, not for persuading them to make one choice rather than another. (Surely the response to the objection that this principle obliges us to intervene to assist other countries in establishing dictatorships is the same as Mill’s response to the issue of voluntary slavery: if we value choice, we have no reason to assist in its abandonment.) Of course, there are questions about what amounts to evidence for a choice having been made, particularly when circumstances discourage expression. Adaptive preferences are also a stumbling-block here, for given a history of oppression we cannot always give full value to the choices made. Democracy is, in principle, a partial solution to the evidencegathering problem, although not to the adaptive preferences problem nor to other problems that arise when background security is lacking. In any event, nothing in the argument above licenses any attempt to impose regimes of a favored kind. It favors local “choice,” whatever the difficulties in interpreting what that term requires. For the guiding (“iterative”) thought is that we should think of political societies as parallel exercises in self-government – and before we get to difficult questions about what “self-government” means we surely know


that there are circumstances so dire that it is farcical, even offensive, to suppose that a society must save itself from them. The second caveat is that the argument above has left entirely untouched all the crucial and currently much discussed issues of practicality that determine whether or not an intervention has a chance of success and whether or not the intervenors have the will or capacity to undertake the further work that may follow. The argument does not, however, underestimate the importance of those issues, taking “ought implies can” as read. It thus leads to the strong possibility that there will be cases where intervention is prima facie justifiable but ought not be undertaken, and I hope that, in addressing the internal legitimacy problem, I will not be taken to have minimized intervention’s many externalities. The gap between ethical theory and ethical practice is always an uncomfortable one, and it may well be that the gap is especially wide in relation to this topic, given all the well-known constraints of capacity and knowledge, and the “hard choices,”35 that come into play in interventionist efforts. The right response to that, however, would not be moral isolationism, but to explore ways of carrying out the duties of assistance by means that are more proactive and less fraught with risk. When duties confront real-world obstacles the moral effect is not to weaken them but to strengthen the case for alternative routes to their fulfi llment. The following chapters explore two such routes. One is the movement to bring about long-term “norm change” through the use of international criminal sanctions; the other is the move to “pre-empt”36 the need for interventions by modifying the global political economy. I shall try to show, however, that those issues too are best understood in terms of the model of political morality adopted in this book.

35

36

See Jonathan Moore (ed.), Hard Choices: Moral Dilemmas in Humanitarian Intervention (Lanham, MD: Rowmen and Littlefield, 1998). Thomas Pogge, “Preempting Humanitarian Interventions,” in Humanitarian Intervention: Moral and Philosophical Issues, ed. Aleksandar Jokic (Peterborough, ON: Broadview, 2003).

Associative risk and international crime

6

A new readiness to question the traditional protections of “sovereignty” has also emerged in the context of international criminal law. Just as proponents of humanitarian intervention have been urging that political boundaries should become permeable to military force in cases of state failure or state-induced atrocity, so too a movement in international criminal law, in the past six decades, has sought to hold individuals who commit atrocities accountable to the world. “In the past six decades” is a misleading phrase if it suggests slow though steady progress, for that is not what happened at all. The (effective) origins of the movement may be traced, clearly, to the Nuremberg and Tokyo trials at the end of the Second World War, when the victorious allies tried and punished individual political and military leaders. But after that no perceptible progress was made – during the whole period of the Cold War – until, in the 1990s, the United Nations Security Council established international criminal tribunals for the former Yugoslavia and for Rwanda, and the drive to establish a standing International Criminal Court made unprecedentedly rapid headway. While that Court’s future remains clouded, given the refusal of major powers to sign on to it, and given too, unresolved issues about the interface between law and politics, surely it is beyond doubt that it has a future, of some kind. If some of the idealism sustaining it has been over-optimistic, so too has some of the resistance to it been exaggerated.1 Naturally, there is already a large legal literature on these developments, which, in assigning criminal responsibility to individuals, 1

For a balanced view, see David Wippman, “Exaggerating the ICC,” in Bringing Power to Justice: The Prospects of the International Criminal Court, ed. Joanna Harrington et al. (Montreal: McGill-Queen’s University Press, 2006).


and thus challenging centuries of state-centered international jurisprudence, open up fundamental questions about the authority of law and its basis. Political theorists, however, have so far had less to say, even though the developments are at least as challenging to the boundaries of their own domain as to those of the lawyers’. The idea that crime is an offense against a state is no less essential to the history and definition of the state than it is to the history and definition of crime; the two histories, and the two definitions, are intertwined. Crimes are offenses against the state (“crown,” or “people”), as opposed to the victim, because states came into being, in part, by asserting territorial control over issues of public order, displacing local or communitybased justice-dispensing in the process of doing so, thus eventually becoming the monopolists of coercive force – a criterion of success that they effectively constructed in the course of striving to meet it. Given that long history, the very idea that something can be a crime “against humanity” – thus, not against a state – is quite simply stunning in its implications. It implies that, for certain purposes, individuals can be tried and punished as though they were citizens of the world, their boundaryless human status trumping their bounded protections. For its own important purposes, legal theory has evolved a juridically workable definition of what a “crime against humanity” is. It is set out in the Rome Statute, Article 7, which states that it is one or more of eleven categories of crime (murder, enslavement, torture etc.) “when committed as part of a widespread or systematic attack directed against any civilian population, with knowledge of the attack.” But what should political theorists make of it? How, if at all, should it affect their conception of the state? I believe that the general argument of this book is well positioned to answer those two questions. But before turning to that, let us explore the available range of views.

I Hannah Arendt’s discussion, in Eichmann in Jerusalem,2 was, remarkably, the fi rst approach by a political philosopher to the idea of 2

Hannah Arendt, Eichmann in Jerusalem: A Report on the Banality of Evil (London: Penguin, 1977).

Associative risk and international crime 145

crime against humanity, a discussion focused not on the Nuremberg trials (though it makes some very shrewd comments about them in passing) but on an event that took place a decade and a half later. Needless to say, that event had features that invited controversies of a legal kind, involving as it did the kidnapping of the defendant, the standard worries about the retroactive application of law and the partiality of the court, and other procedural issues. Arendt’s book is illuminating on all of these, as it is, of course, on the central theme for which it became famous: the difficulty of applying standard ideas of criminal accountability in the context of a “criminal state.” Eichmann, though morally deprived, was not morally depraved; he was a bureaucrat, a specialist in transport operations, sustained, apparently, by the powerful thought that he was doing his “work.” We shall return to this below, as it is obviously relevant to the theme of complicity. But for the moment, I shall take up Arendt’s very interesting and original contribution to the matter of defi nition. “Crime against humanity,” as a legal (rather than a rhetorical) charge fi rst appeared in the charter that established the International Military Tribunal (Nuremberg tribunal). There is general agreement that, for all its later importance, it was an extraordinarily illthought-out charge, driven politically (perhaps) by the desire to fi nd a charge that could not be returned against the prosecuting nations, and conceptually by the need to extend the established category of “war crimes” to something that was not really a war crime at all. The German translation of the document, Arendt points out, added confusion; the term chosen implied that the Nazi policies of extermination had been merely lacking in “humanity” (in the sense of human kindness) – not just “the understatement of the century,” but also something that fails to distinguish genocide from many ordinary crimes, and indeed from many things that do not even rise to the level of crime at all (such as unhelpfulness or ingratitude).3 In formulating her approach, Arendt refers approvingly to a gloss on the charge provided by the French prosecuting counsel: crime against 3

Ibid., 275.


humanity, he maintained, is best understood as an offense against “the human status.”4 As it stands, of course, that formulation does not advance far beyond the “inhumanity” suggestion, for it is still too broad. Kant, in one of his formulations of the duty of morality, defi ned it in terms of respect for the humanity of the other (as well as of oneself), and in terms of such a view all moral wrongdoing whatsoever could be described as an offense against the human status. But Arendt gives the idea a context of interpretation that rescues it from overgenerality. A crime against humanity is an atrocious act or policy that aims to “eliminate forever certain ‘races’ from the face of the earth,” in this case a policy of “refus[ing] to share the earth with the Jewish people.” Although the crime was committed “on the body of the Jewish people,” it was a crime against humanity – rather as a domestic crime, although committed on a victim, is a crime against the sovereign? – because diversity of identity is an essential feature of the human race itself. Without the acceptance of diversity, “the very words ‘humanity’ or ‘mankind’ would be devoid of meaning.”5 Arendt returns to this theme in the closing words of the book, where she states what in her view should have been the Eichmann court’s judgment: “we fi nd that no one, that is, no member of the human race, can be expected to want to share the earth with you. This is the reason, and the only reason, you must hang.”6 In his discussion of “hospitality” in Perpetual Peace, Kant also referred to sharing the earth’s surface as a basic human right. There is a right to “communal possession of the earth’s surface.” “Since the earth is a globe,” Kant continues, “[humans] cannot disperse over an infi nite area, but must necessarily tolerate one another’s company.”7 Pirates and colonizers, however, are “inhospitably” intolerant, forcibly denying the rights of other societies instead of seeking that unforced agreement that is the basis of an eventual 4 7

5 6 Ibid., 257. Ibid., 268–69. Ibid., 279. Immanuel Kant, Political Writings (Cambridge: Cambridge University Press, 1970), 106.


cosmopolitan society. Kant’s view here seems to underlie some passages in Arendt’s Human Condition that help to amplify the closing remarks in Eichmann in Jerusalem. In inhabiting the earth’s surface, she writes, human beings create spaces which serve as public contexts for performing and recording the activities distinctive to our species. The human species thus consists of a plurality of public worlds, spatially organized.8 And so, taking this back to the theme of Eichmann, an atrocity is a crime against humanity if it is driven by the belief that one or more of those public worlds – whose existence defi nes what it means to be human – can be destroyed. At the time of the Holocaust, the Jews of Europe were not spatially organized in public worlds. This was the very fact that made them vulnerable to genocide, and which also gave rise to the response that many of them consequently favored: to create such a space of their own. The best justification for the Eichmann trial, in Arendt’s view, lay in the fact that while other peoples had alreadyexisting states that enabled them to prosecute their Second World War oppressors in the years after 1945, the Jewish people acquired such a state only later. The retroactivity of the law under which they prosecuted Eichmann merely compensated, very partially though justly, for the fact that it was precisely the lack of a state that had enabled the crimes of Eichmann and others in the first place. That argument relies on a literal conception of political space in physical terms, a conception that accords well with Kant’s bluntly physicalist observation that the earth is a sphere, so that direction is necessarily fi nite, and the need for toleration thus inescapable. But Arendt’s line of argument risks circularity. Having a political space enables a people to take action against their oppressors in seeking justice for what happened to them before they had such a space: that is a compelling idea, and suggests that complaints about retroactivity are (to put it mildly) context-insensitive. But what is problematic is the implied view that what happened to them before they had a space can be 8

Hannah Arendt, The Human Condition (Chicago: University of Chicago Press, 1958), 1–3, 50–51.


defi ned in terms of violation of their space – as though they already had one. Once the state of Israel existed, (Israeli) Jews could then do what several European nation-states had already done – the trial of Eichmann should be viewed as a sequel to innumerable post-war trials in Western and Central Europe: but can what had been done to them be defi ned in terms of already-existing statehood? Arendt’s argument takes a more metaphorical turn, however, introducing an idea of “space” that is no longer physical, or connected with the fi nitude of a sphere’s surface. “Space” becomes a way of talking about group identity. The “space” that matters may sometimes coincide with physical space, but need not always do so if we think of it in terms of “the space between individuals in a group whose members are bound to . . . each other by all kinds of relationships, based on a common language, religion, a common history, customs and laws.”9 This turns the physicalist argument into an argument about culture. We can certainly agree that the attempt brutally to extinguish cultures is vile. But why is it a crime against humanity, if humanity’s interest in the matter is defi ned in terms of the physical, non-metaphorical, possession of space? Here a distinction between plurality and diversity seems important. The idea of plurality refers to the fact that the human race has become organized into multiple jurisdictions, each of which has a collective right that parallels every other jurisdiction’s collective right. An attack on one jurisdiction’s right could be taken to undermine other jurisdictions’ secure enjoyment of their rights, and in that sense could be taken to elicit a general concern on the part of spatially organized “humanity” (Arendt’s “comity of nations”). The idea of diversity, however, seems to point in a different moral direction altogether. If what is important is the possession of a distinct language, religion, and culture, we need another line of argument to show why such things bring a (universalist) human concern into play. It points in the direction of a nationalist argument, not in 9

Arendt, Eichmann, 262–63.


the direction of the kind of politics that Arendt celebrated, a view in which polities matter not because they embody differences but because they are self-constructed and serve important universal and distinctively human purposes. To be sure, plurality often leads to diversity, as David Luban has pointed out, in objection to this line of criticism: the existence of plural independent jurisdictions leads naturally to the formation of diverse ways of life.10 The link between plurality and difference is provided by the feature of “originality.” But all the same, it is important to separate the two lines of justification, for they do not map onto each other. In the fi rst place, if diversity figures as a consequence of plurality, it is not normatively basic, and not immediately connected to the basic facts about the human condition that provide so graphic a starting-point. In the second place, the value of diversity is excessively optional, is held by some much more strongly than by others, and even for those who hold it is rather too easily overridden by other values – such as security. In the third place, a justification based on the value of diversity may lead to unwelcome results: if a crime against humanity is an assault on diversity, it would not be a crime against humanity to assault assimilated minorities who had taken on the social and public culture of the majority society, people defi ned by a genealogy that had no bearing at all on their way of life. A diversity-based view leads, then, to only conditional results. One’s protection depends, on that view, on one’s difference, and the unconditional wrongness of an assault on one’s “human status” may be lost from sight.

II The idea of the “human status” is also important to what is to date the only full-length discussion, in political philosophy, of crime against humanity. The idea enters into Larry May’s interpretation of that crime in a very subtle way.11 May’s concern is to give a reasoned 10 11

Luban, “A Theory of Crimes Against Humanity,” 115–16nn. Larry May, Crimes Against Humanity:A Normative Account (Cambridge: Cambridge University Press, 2005).


account of the domain of international criminal law. Unless it is to displace domestic law altogether, of course, international criminal law’s scope must be demarcated by means of ends or principles that are distinctively its own. While international criminal law must have the protection of persons’ security as its object, a “security principle” is only a necessary, not a sufficient, condition for its role, for while states exist some presumptive reasons protect them from intrusions by other (foreign or international) bodies. In addition, May argues, we need an “international harm principle,” that will permit intrusions into sovereignty only when a further condition is met – harm to “humanity.”12 Just as domestic crimes are offenses against a community, so too international offenses must be offenses against a community – a larger one. What that means, however, is in need of careful interpretation. May argues for a view that invokes the “group-based” nature of certain crimes, by which he means, in part, that they are crimes which have as their victims persons qua members of groups. Why, though, does a crime that is group-based in that sense amount to an offense against international society, “humanity,” as a whole? May’s answer is twofold, involving both expressive and consequential elements. First, a crime committed against a group treats its victims as members of a group, thus denying their humanity as persons. Secondly, crimes committed against groups are especially likely to spill across borders, thus creating threats to common peace and security. The fi rst, expressive, element is said to involve a denial of basic human autonomy. “If an individual person is treated according to group-characteristics that are out of that person’s control,” May writes, “there is a straightforward assault on that person’s humanity. It is as if the individuality of that person were being ignored, and the person were being treated as a mere representative of a group that the person had not chosen to join.” That idea is then clarified or expanded by way of an appeal to the value of autonomy: humanity is 12

Ibid., 80–95.


said to be harmed when individuals suffer deliberate harm “according to things that are beyond their autonomous agency.” In suffering harm by virtue, simply, of group affiliation, it is as though their group membership superseded and nullified their membership in the human race. It is in this sense that “humanity” (as a whole) is “victimized.”13 There is also, as noted, a consequentialist element to the argument, as well as a second idea of “group-based” crime deriving from the idea that groups are the perpetrators of it, but before turning to either of these let us consider the thesis that harm to humanity is to be understood as a crime against human autonomy. The fi rst critical thought concerns the primary or even exclusive focus on ascriptive groups – groups that people can be said to belong to regardless of choice. That focus is understandable in light of certain paradigmatic examples, such as the Armenian genocide and the Jewish Holocaust. But there is an air of paradox in a view which, because of the importance of autonomy, allots special protection to precisely those groups that are the least likely to arise from autonomous choices. Moreover, the view seems to exclude, or move to the margin, other real or potential victim groups that less readily fit the ascriptive model. Political and social groups come to mind: it would seem, on this model, that systematic persecutions of, say, Communists, or rich peasants, would fail to qualify, if anyone’s being a Communist or a rich peasant could be traced to any significant extent to the exercise of autonomous agency, as, it seems likely, it could be. In other cases, the groups are constituted as such by the persecutor’s decision, and in that sense are non-dependent on their members’ agency, though their members’ decisions could not reasonably have been supposed to invite a group identity – the victims of the Cambodian atrocity being a leading example: the choice to move to the city, to buy a radio, or to wear eyeglasses, could no doubt be described as autonomous, but the fact that such acts amounted to membership in a group was constituted by another agent. So was 13

Ibid., 85.


membership in that group autonomous or not? In yet other cases, the focus on autonomy seems to miss the moral point. Homosexuals were among the victims of the Nazi Holocaust. It is currently and reasonably believed that sexual orientation is a matter of disposition rather than of choice. But is that why the extermination of homosexuals was a crime against humanity – so that it would altogether cease to be one if sexual psychology were one day to discover a component of choice in people’s orientation? The second critical thought concerns the close equations between autonomy and humanity and between their denial and victimhood. If the recognition of autonomy were connected with the recognition of humanity, and the denial of one’s humanity led to one’s being a victim, then the chain of connection would be strong. But it really is not clear that the perpetrators of crimes against humanity do deny the humanity of their victims. Tyrannies maintain their power by means of practices or threats to which they know human beings are vulnerable – they know that human beings fear pain and shame, need sleep and food, love their close relatives, and revere sacred symbols. If they did not know that their victims were human in those ways, they would not have the means to oppress them. Moreover, conquistadors, slave-owners and concentration camp guards had sexual relations with the people they dominated: did they really think they were committing acts of bestiality? The convenient idea that oppressors deny the humanity of their victims does not adequately fit the facts. What we face, rather, is, surely, the powerful human capacity to make exceptions in relation to one’s own desires and interests, not a failure to understand general abstract truths about what humans have in common. In part, then, May’s approach is to appeal to “humanity” in the sense of a distributed feature of humans, a feature that is denied by group-based atrocity. Crime against humanity is a crime that is not only directed against groups, but is directed by one, that is, a state or state-like entity. That second element allows his argument to account for some more of the central features that figure in mass


atrocity. As May himself notes, though, these two elements may strike readers as rather distinct.14 He insists that the group-based character of both relates them importantly together, and there is no reason to deny that it is an important common feature. But in terms of the present argument, the second element, while certainly groupbased in a significant sense, is not group-based in the sense of denying autonomy. So while the value of protecting autonomy from harm may get us to the fi rst group-based element – if it overcomes the questions and qualifications outlined above – it does not seem to get us to the second, even though the second may be described as groupbased for as good or even better reasons. In light of these considerations, then, one might question the expressive thesis that seeks to tie international crime to the violation of autonomy. As noted above, however, the expressive thesis is accompanied by a consequentialist one: crimes that have groups as their victims are especially likely to spill over borders with dangerous consequences for the international community as a whole. This version has both legal and political advantages. It has very clear legal advantages, since the promotion of international peace and security, authorized by Chapter 7 of the UN Charter, is among the strongest foundations for justifying intrusions into state sovereignty. It has political advantages of a motivational sort, for it appeals to one of the clearest global public goods, of benefit to all societies. Needless to say, it also has the defect of consequentialist arguments, namely, that it depends on causal stories that may not always obtain. Whether or not the story obtains seems quite conditional. It is a threat to international peace and security to persecute an ethnic group with a territory that straddles boundaries, or which has a large and influential diaspora in powerful countries, for example; it is much less of a threat when conditions are such that the persecution is likely to be ignored because it is geographically remote, for example, or has no influence on the supply of globally valued resources. But apart from 14

Ibid., 91.


issues of that kind, the connection between security threats and the persecution of ascriptive groups, as opposed to groups that in some sense are voluntary, seems weak or even counter-intuitive. For aside from the normative question of whether we want to give ascriptive groups a higher level of protection than voluntary ones, might we not argue quite the reverse of the empirical thesis? A threat to a group with feature x is a threat only to those who have feature x, and in that sense is contained – those who do not have the feature are exempt from threat. But suppose the threat is made to people who may be said to share some very much more elastic feature, such as an attitude or orientation? One may think of being an “intellectual” in Pol Pot’s Cambodia or a “bourgeois” in Mao’s China. Those descriptors are infinitely more flexible than, say, “gypsy.” Moreover, the more flexible descriptors are, the more easily they can be manipulated in ways that link maligned co-nationals with powerful and feared outsiders, with the cross-border effects that May’s consequentialist thesis invokes. I think there is, however, an important sense – to which May’s discussion points – in which crime against humanity can be regarded as group-based not only in relation to its perpetrators, but to its victims too. Communities are, as it were, individuals’ first line of defense against attack or deprivation. Even in modern societies, in which states have largely absorbed the protective function to themselves, to be without any sort of familial or communal protection is to be vulnerable; and totalitarian versions of the modern state make it their business either to crush or to coopt social groups so that individuals fall under their complete control. Atrocities become possible, then, both because states have exceptional means at their disposal for overcoming collective resistance, and because their absorption of the protective function necessarily increases their members’ exposure to potentially atrocious force. The group-based elements of crime against humanity, in other words, belong within an account of endemic political risk.


III A third proposed approach is an imaginative appropriation of the classical language of “republicanism” to the global scene – imaginative because it is conventional to think of republican values as being tied to the politics of sovereign domestic polities, to the exclusion of wider obligations. Nevertheless, “republican cosmopolitanism,” as James Bohman has shown, is thinkable, and in thinking it Bohman draws upon the idea of crime against humanity as a “conceptual clue” to understanding the international system in republican terms.15 For that idea contains the further idea that the international community as a whole is the addressee of human rights claims. It expresses the fact that, in an increasingly interdependent world, “a violation of rights in one part of the world is felt everywhere,” as Kant put it.16 But the facts of interdependence alone do not create community, nor is the nature of global interdependence satisfactorily captured by what may be termed a “cosmorawlsian” basic structure. Rather, its normative meaning is best conveyed by its role in intensifying domination, a term central to the republican vocabulary. Since the decisions of powerful institutions affect the life-chances of indefinite others, some people are necessarily made part of other agents’ plans without their agreement, so that the core republican value of non-domination is violated. It follows that, although the existence of a court prosecuting crime against humanity is, as it were, emblematic of a world republican order implicitly committed to the value of non-domination, that value is not satisfied by the prosecution of tyrannical acts; for tyranny is only one form of domination. We need political accountability, not just criminal accountability. If violently tyrannical acts are objectionable because they make some people unprotected potential instruments of the will of others, then many decisions made within the global political economy are objectionable on exactly the same grounds.

15 16

Bohman, “Republican Cosmopolitanism.” Kant, Political Writings, 107–08.


The fi nal chapter of this book will turn to some of the issues of accountability raised by Bohman’s article. For the moment, the relevant point is that its argument presents crime against humanity in the context of a globalized world (understood in a sophisticated way). Here we should distinguish between two ways of taking Kant’s remark, quoted above. Perhaps violations of rights are now globally “felt” simply because they are now globally known – informational globalization. That reading is of course fully consistent with a background belief in natural rights: rights aren’t brought into being by globalization – it’s just that their violation is brought to light. Perhaps, however, violations are “felt” because an interdependent world has come to constitute a single community in some sense, so that the reverberations of abuse don’t stop at national boundaries. This appears to be the reading adopted in Bohman’s article, which urges us to replace abstract moral fictions (natural rights) with the idea of rights of membership.17 One may not be so sure, however, either that membership comes fi rst or that it is essential. With regard to the former, the emblematic institution (the International Criminal Court) and, more broadly, the international human rights regime, owed their origins to what, from the republican point of view, may be fictions, since they preceded anything that we might call a political community dedicated to them, even if in some sense they may be said to emblematize one. With regard to the latter, there is, notably, what we termed above the “North Korea” problem. Suppose hideous abuses (torture, deliberate starvation) occur in societies whose leaders have excluded them from the globalized world? If we want the international community to take note of and respond to them, it cannot be that we suppose that being part of the globalized world is a necessary condition for enjoying moral standing. To the “North Korea” problem we added a more hypothetical – but still, I think, morally revealing – “spoiled brat” problem, as it was termed, relating to the case of a wealthy and powerful country discovering that its 17

Bohman, “Republican Cosmopolitanism,” 344.


interaction with less favored countries generates obligations on its part: it would be entirely unobjectionable, on an interaction-based view, if it were to respond by withdrawing from connections with countries whose features might be thought to trigger unwelcome and costly obligations to aid. So can interaction be morally basic?

IV The idea of crime against humanity is a rich one in moral, legal, and political terms, and all the more rich in combining all three, as it often does. Accounts such as Arendt’s, May’s, and Bohman’s succeed in capturing important elements of it, although, it was suggested above, they also leave out some things of importance. In the remainder of this chapter, I want to set out a third line of interpretation and argue for its greater adequacy. This view may be termed the associative risk thesis, and its basic point will be familiar from chapter 3 above. According to some anthropological evidence, pre-state societies were capable of (proportionate) levels of violence that even modern states, with all the means at their disposal, have been unable to match.18 While casualties in battle were typically smaller, for weaponry had limited killing power, and the slenderness of the economic surplus made sustained warfare impossible, hostilities were frequent. An annual casualty rate of only a few percent, repeated over a five-year period, leads to a larger proportional population loss, within a generation, than any modern society has experienced in its most destructive wars. In a graphic piece of reporting, Jared Diamond brings out the destructiveness and inefficiency of intergroup retributive processes in contemporary New Guinea – an account that certainly gives life to Locke’s bland remark that life before states was “inconvenient.”19 So on a long-term balance sheet of lethality and security, states seem to have been a good idea. 18

19

Lawrence Keeley, War Before Civilization (New York: Oxford University Press, 1996). Jared Diamond, “Vengeance is Ours,” New Yorker, April 21, 2008, 74–87.


But life in a pre-state condition, however lethal, is not actually the alternative prospect that faces us today. The prospect that faces us is the lethality of the state itself; and even though, taking the longterm view, states may be justified by the security that they bring to their members’ agency, their capacity to kill their own members, or those (if not members) who are under their protection, is astonishing. It has been documented in a series of careful studies by R. J. Rummel, who gives us the chilling statistic that in the course of the twentieth century many more people were killed by actions of their own governments – about four times as many – than by the armed force of other states.20 Rousseau once remarked that states had put an end to civil wars only to unleash foreign wars “a thousand times more terrible,”21 but that judgment was (or has become) wrong. Civil wars, if understood as violence between parties (including governments) within the borders of states, are even more terrible, if we measure death rates, than the appalling foreign wars made possible by industrialization and concentrated political power. The associative risk thesis is that crime against humanity is the atrocious use of a state’s protective resources to strip protection from those under its control and thus to render them helpless against its own violence. As David Luban puts it, “The leitmotif binding together all the legal features [of crime against humanity] is that of politics gone horribly wrong.” Further, “To criminalize acts of a government toward groups in its own jurisdiction . . . is tantamount to recognizing that the cancerous, autopolemic character of crimes against humanity represents a perversion of politics, and thus a perversion of the political animal.”22 This view helps us to understand how crime against humanity, thus understood, involves a distinctive kind of evil. It may not be the worst kind of evil – definitions should avoid competitive victimhood – but it involves elements that are distinctive to it, just 20 21 22

R. J. Rummel, Death by Government (New Brunswick, NJ: Transaction, 1994). Jean-Jacques Rousseau, Political Writings (Oxford: Blackwell, 1962), vol. I, 365. Luban, “A Theory of Crimes Against Humanity,” 108, 117.


because the state itself is a distinctive form of organization, with a distinctive claim to moral importance arising from its coercive scope. For there is a particular complex of events, reflecting three central features of state power, that attracts a particularly radical form of condemnation. The fi rst of these features are large-scale administrative capacity, and local authority. Both are features of the clearest cases of crime against humanity, which, at a minimum, require powerful mechanisms of collective action and the habits of obedience that sustain them, and an institution regarded as entitled to make binding declarations of rightness and wrongness. Writing of the Rwandan genocide, Gerard Prunier has remarked that “only states have had the physical means of industrial slaughter as well as the massive legitimizing capacity needed to convince their citizens that some of their number are subhuman and deserve to die.”23 With a somewhat different emphasis, the same two state capacities are implicitly invoked by Avishai Margalit and Gabriel Motzkin in an article on the Holocaust; its uniqueness, they write, stemmed from the fact that it sought simultaneously to destroy a group physically, through large-scale organized extermination, and to use its authority to humiliate and degrade it in the course of doing 23

Gerard Prunier, “A Well-Planned Genocide,” Times Literary Supplement, October 22, 1999, 30; see also Philip Gourevitch, We Wish to Inform You that Tomorrow We Will Be Killed with All Our Families (New York: Picador, 1998), 95; Helen M. Hyntjens, “Explaining the 1994 Genocide in Rwanda,” Journal of Modern African Studies 37 (1999): 245; David Newbury, “Understanding Genocide,” African Studies Review 41 (1998): 76; Aryeh Neier, War Crimes: Brutality, Genocide, Terror, and the Struggle for Justice (New York: Times Books, 1998), 219. It is often recognized that state criminality is an essential component – see for example Geoffrey Robertson, Crimes against Humanity: the struggle for Global Justice (London: Allen Lane, The Penguin Press, 1999), 224, and David Luban’s construct, “the criminal state,” in Legal Modernism (Ann Arbor: University of Michigan Press, 1994), chapter 5. Some delegates to the Genocide Convention in 1948 objected to making state action or complicity essential to the crime, since terrorist organizations can commit genocide: see Leo Kuper, Genocide: Its Political Use in the 20th Century (New Haven: Yale University Press, 1981), 37–38. On the argument developed here, “unofficial genocide” would not be a crime against humanity, since (some elements of) two components other than sheer capacity are required.


so.24 Only states have this double capacity of physical removal and moral exclusion. With regard to administrative capacity, here Arendt’s central argument comes into its own. The Rwandan case was atypical in that the process of slaughter was highly participatory, a hundred thousand or more perpetrators committing directly criminal acts within the “widespread or systematic plan.” Those who formulated that plan certainly had administrative capacity – commentators noted the tragic paradox that they succeeded only because Rwanda had one of the more effective states on the African continent – but generally speaking the administrative capacity of states does its work by coordinating non-criminal individual acts into a collective criminal act. Public administration continues and perfects that process of “civilization” that Rousseau, Mill, and others remarked upon, that is, the division of labor and parceling out of tasks that enhances collective power, at the cost of diffusing responsibility to the point where it seems to vanish. This is the “criminal state” that Arendt wrote about, and its genocidal potential is simply the most extreme and visible effect of the general complicity discussed in chapter 3 above. To be part of a modern society is to play a role whose nature is hidden by its minuteness. To an extent, as the case of Eichmann (or Eichmann’s selfinterpretation, anyway) suggests, an appeal to the role itself may act to excuse consideration of wider responsibilities. In other cases, mere shortsightedness may be too charitable an explanation. Doing one’s job may be reinforced by less presentable motives: Schadenfreude; discomfort with the Other that falls short of overt racism; neighborly frictions; factually dubious notions of what others deserve; misdirected notions of threat.25 But all such things, however culpable, fall well short of actual depravity. Few people never experience any 24

25

Avishai Margalit and Gabriel Motzkin, “The Uniqueness of the Holocaust,” Philosophy and Public Affairs 25 (1996): 65–83. See the contributions to Esses and Vernon (eds.), Explaining the Breakdown of Ethnic Relations, for discussions of all these topics.


of them. But in the context of Rousseauan/Millian “civilization,” in which fates are so compacted together and oppression is for the most part vicariously conducted, the sense of duty, and ordinary human moral weakness – or most likely the two in combination – are a sufficient basis for atrocity. To these two features we must add another, as their necessary precondition. State power is territorial. By territorializing identity, and seeking or tending to substitute territorially based services and functions for communal ones, states erode what would otherwise be the primary basis for security, that is, community based on familial or local attachment. In that way, it increases potential vulnerability, by taking the protective role to itself. And that potential becomes actual when the territorial character of the state is put to oppressive use, when it is exploited in order to immobilize a group and also to deny it external help – to attack it while denying it hope of the desperate last recourse of either flight or protection, so that its boundary-enforcing resources can be brought to bear against it in an unimpeded way, the means of “protection” morphing into the means of destruction. Administrative capacity, local authority, and territoriality come into being with the creation of states; when, therefore, they play an essential role in an attack on a group of a state’s subjects, that group suffers a distinctive kind of evil that only statehood can bring about. Does this account match the legal defi nition? Perhaps most clearly, the model makes sense of the “systematic” criterion (a criterion that makes no sense, of course, from the individuated victim’s perspective – persecution is persecution, systematic or not). In fact, the “systematic” requirement is virtually identical to the requirement of state initiative. (If only state complicity is involved – letting death squads do their work – the model here can easily accommodate the distinction, just as the drafters of the Rome Statute did.26) If

26

See the Final Draft Text of the Elements of Crimes, adopted by the Preparatory Commission for the International Criminal Court ( June 30, 2000), Art. 7, note 6.


that requirement bore upon the element of premeditation – rather as fi rst-degree is distinguished from second-degree murder – it would surely lead us to single out the political leaders who plotted atrocity, rather than the sad (or banal) non-entities who carried it out: but the Rome Statute holds responsible all those involved. In doing so, it implies a background model which, like the model developed here, imposes on everyone a duty to recognize and to resist the moral risks entailed in membership in a state; it is because the state is the kind of institution that it is, with the features that it has, that crime against humanity is possible. In this regard, the model outlined here may help to deal with the troubling question of the scale of atrocity. Here intuitions tend, awkwardly, in different directions. On the one hand, we can hardly close our minds to questions of scale altogether. Legally, a crime against humanity (unlike a war crime, which may be a single act) is a concerted persecuting effort, or a component of one; but even outside a legal context the greatness of the evil owes something to its extent. So from this point of view numbers seem to count. It is very unclear whether numbers are absolute (a body-count) or relative (a proportion of the target population) or time-sensitive (numbers killed per day), but they would certainly seem to come in somehow. On the other hand, of course, moral sense rebels at the thought that numbers count in this way: that the Holocaust would have been less of a crime if only (say) three million had been killed, or that we might tell whether the Holocaust or Stalin’s reign of terror was worse by simply counting bodies, or whether enough Kosovars were killed to make NATO’s action against Serbia justifiable – inviting the cruel and absurd question, how many would have been enough?27 The view developed above suggests why numbers matter and also why they do not, as such, count. A crime against humanity is one that requires the application of a state’s resources – the capacity to apply violence

27

See David Luban’s remark about “charnel house casuistry” in Legal Modernism, 344.


on a large scale, and the control of territory; and it also requires a target group substantial enough to have developed expectations of common survival. Numbers matter, then, as an indicator of something else, i.e. of whether state power is essential, in these two ways, to the events in question. But above this (variable) threshold, questions of number (proportion, speed, etc.) have no weight at all, as far as the idea is concerned, given its non-additive nature. The kind of evil in question is also distinguished, by this model, from violations of human rights. Human rights violations are individuated. What are termed “gross” human rights violations are additive; and they do pose the quantitative question of how many are acceptable before a response is called for. But crime against humanity is not an additive concept. It relates on the one hand to the necessary institutional prerequisites of an organized project, and on the other to damage suffered by individuals only by virtue of belonging to a group. Of course, from the victim’s perspective neither of these considerations can be crucial: damage is damage, whatever its organizational prerequisites and whether or not it is group-based. But the whole argument developed above, especially in rejecting the idea of crime against humanity as an offense against an individual status, requires a perspective which is not victim-based. Finally, what kinds of victim count? The Armenian genocide and the Jewish Holocaust were directed against ethnically (or ethnically and religiously) defi ned victims; and cases (such as Rwanda) where the bases of ethnicity may involve fictional elements, do not seem materially different – the core cases involve ascriptive groups, whatever exactly is the basis of the ascription. But does anything morally important distinguish the persecution of ascriptive groups from the persecution of non-ascriptive ones? One provocative thought was suggested by the Iranian representative to the United Nations committee on the Genocide Convention in 1948: the destruction of ascriptively defi ned groups, he said, “appeared particularly heinous in the light of the conscience of humanity, since it was directed against human beings whom chance alone had grouped


together.”28 For if a group defi ned by the chance of birth is persecuted, it is as though its members had not been born into the human race at all – their human status is denied from the start. It is not some subsequent choice that attracts the punishment of the state; it is a sort of atrocious practical category mistake that travesties the logic of punishment, imposing penalties for being rather than doing. This view fi nds an echo in May’s criterion of violated autonomy. Although this is initially a compelling idea, it is hard to see, on reflection, what distinguishes this case from the case of innocent groups in general. Everything that is wrong about it is also wrong about Pol Pot’s extermination of “intellectuals,” meaning schoolteachers or even people who had learned to read, even though those affected could be said to have made choices – choices that could not have been foreseen to be politically contentious. So to “chance” we need to add “innocent choice.” But once we have made this extension, what excludes politically controversial groups? Political groups are included among the groups protected in the preamble to the Rome Statute, although there has sometimes been resistance to their protection on the same terms as other groups.29 On the model developed here, there is no defensible basis for this resistance. On other models, such as that suggested by the chance/choice distinction, there may be a difference between, on the one hand, chances and choices that (as it were) chance to be dangerous, and, on the other, political choices known at the time, perhaps, to be risky. But since the defi nitive element in this discussion is the travestying of state powers, to any group’s detriment, the suggested distinctions do not seem decisively consequential.

V Why, though, should crime against humanity, defi ned in this way, amount to a demand that other states create and support a system of international criminal law? At the basis, the argument here is no 28

Quoted in Kuper, Genocide, 19–20.

29

Ibid., 24–30.


different from the argument for humanitarian intervention: states are justifiable only on the basis of elementary cosmopolitan regard, they can bracket this regard only on a good-faith assumption that others can bracket it too, and when that assumption becomes manifestly unsustainable, cosmopolitan regard reasserts itself. The idea of crime against humanity implies the same conclusion in a still more pointed way: it comes into play not just when state power fails, or is abused, but when it is travestied (or as Luban says “perverted”), that is, when the features of the state – the very features that justify its sovereignty – are put to use in ways that inspire a specific kind of repugnance, the repugnance that one feels at seeing an institution essential to the ordinary conduct of human life being systematically betrayed. There is a systematic inversion: powers that justify the state are, perversely, instrumentalized by it, territoriality is transformed from a refuge to a trap, and the modalities of “punishment” are brought to bear upon the guiltless. But we can, I think, further refi ne the connections here. The idea of crime, as we have seen, involves a three-party relation: there is a victim, a perpetrator, and a third party, which we may term a “sovereign” or “party in interest,” against which the offense is said to have been committed – the “Crown” in Commonwealth jurisdictions, the “People” in the United States, the “State” elsewhere. Looking behind the legal forms, the normative meaning of that third party is obviously bound up with general justifications of the state, of an (at least approximately) Lockean kind. States can make a claim to deal justice because of their relative impartiality – not (necessarily) in relation to general views of justice, but (necessarily) in relation to the particular interests of the first and second parties. Their role is triggered, in other words, by breakdowns in the associative capacities of their members, who are supposed to have strong interests in mutuality, interests that may, however, founder when personal interest and open-textured (thus diversely interpretable) principles are in play. There is a deep model of subsidiarity here. Without cosmopolitan (basic human) regard, we would have to see states as


the creators of order among atomic individuals who, lacking moral capacity, would be incapable of unforced association. Rejecting that unusual and surely implausible view, we will necessarily see states in terms of a tiered model of association, each tier responding to contested claims of justice on the part of the associated persons or groups that it includes. Suppose we adopt that general view as the context for our idea of the state: states figure within a general scheme in which more inclusive associations are justified by their role in relieving injustice arising within less inclusive ones. That general scheme, it was argued above, gives us the best available account of why states have a particular claim upon their members. If we accept the claims of states on that basis, then the implied principle leads us directly to the case of international criminal law. Suppose a government not only fails to respect, but violently tramples upon, that principle? Then the principle of subsidiarity unwinds in an upward direction. If each associative level must be subject to control by the next, when it is unjust, then when we reach the highest available level, another must be created. For it is exactly that principle that justifies states, and they can benefit from it only by following through on its global implications.

7

A global harm principle?

The approach taken in this book has been “iterative” rather than “causal” in its view of transborder duties; that is to say, it has adopted a version of a moral generalization principle in getting from what we owe to co-citizens to what we owe to those outside. It thus dissents from several important views that take as their starting-point the facts of global interaction, deriving ideas about what we owe to outsiders from the indisputable reality of globalization. In places, in fact, this dissent has been explicit, the discussion offering alternatives to interactionist proposals offered by Onora O’Neill and James Bohman, for example. This approach may seem to invite the charge of ignoring what should not be ignored. The charge is particularly likely to be pressed in connection with the idea of risk that played an important role in justifying compatriot preference, for part of the indisputable reality of an interdependent world is that what we do imposes severe risks not only on co-citizens but on distant strangers. An attempt was made to identify kinds of risk that apply to co-citizens distinctively: but all the same, a risk is a risk, and even if it is of another kind its imposition can hardly be ignored. But this approach does not do so. The iterative model tells us that participants in one social project have duties to aid, and not to impede, the social projects of others. If duties of aid are especially relevant to humanitarian intervention and support for an international legal system, the duty not to impede has particular relevance for the economic policies and practices of rich countries and of the international institutions that they support. Recent discussion has formulated this duty in terms of a global “harm principle” that forbids causing or sustaining poverty or crippling development in vulnerable societies. This chapter endorses


such a principle. It argues, however, that the idea of “harm” requires interpretation within a theory of what we owe. That theory is the one developed in this book. Once again, the appeal is to the contractualist view that the exclusiveness of one contract is justifiable only on the basis of respect for others’ contracts – or, to put it differently, for others’ self-determining capacity as a parallel to our own. On this approach, the evaluation of economic practices is mediated through their political consequences; such mediation, it is argued, is essential if the idea of “harm” is to be a usable one.

I The scope of positive duties will be controversial as long as there is disagreement over the basic questions of morality itself. Their scope becomes doubly controversial in the context of global justice, moreover, for in that context whatever it is that we owe to others in general faces competition with the claims of duty to those with whom we are particularly associated, claims that are supported by a familiar array of considerations, thus posing priority dilemmas. So the attraction of turning to negative duties is powerful, given the practical objective of arriving at consensus: for while we may differ about our positive duties to aid or to equalize or to make restitution, who, if morally sentient at all – regardless of their general moral view – could be content to harm others, or to be part of a system that systematically does so? The prohibition of harm is “something that can be easily agreed upon by people with a wide diversity of overall conceptions of the good”1 and international treaties and declarations already give some evidence of this.2 A further important consideration, stressed particularly by Thomas Pogge,3 is that, while relations among members of one society legitimately give rise to special

1 2

3

Barry, Justice as Impartiality, 88. Andrew Linklater, “The Harm Principle and Global Ethics,” Global Society 20 (2006): 329–30. Thomas Pogge, “Cosmopolitanism: A Defence,” Critical Review of International Social and Political Philosophy 5 (2002): 86–91.

A global harm principle? 169

obligations of a positive kind, from which outsiders do not benefit, those obligations cannot cancel any negative duties that are owed, from the outset, to outsiders: so the negative-duties approach is very well positioned to fend off objections to global justice based on the special claims of compatriots, co-citizens, or co-nationals upon one another’s concern and resources. This turn to negative duties – to a global harm principle – hardly makes global justice undemanding in a way that would rightly be seen as suspect. Rich countries damage poor countries in many ways, so that a prohibition on doing harm would constrain them significantly, requiring changes both in their policies and in their own members’ expectations. Notably, as we learn from a substantial recent body of work, it would constrain such things as the use of agricultural and textile-industry support policies that deny fair market opportunities to poor countries; the imposition of socially and politically destabilizing loan conditions; toleration (or active support) of a trade in small arms that has devastating consequences; the imposition of environmental costs; the systematic recruitment of locally trained professionals to service the health-delivery systems of rich countries; and complicity in international borrowing regimes that motivate oppression and corruption on the part of local tyrants. Finally, a powerful consideration of a motivational kind is offered by Andrew Dobson.4 The “cerebral” understanding of our shared humanity has only too evidently failed to change human behavior in ways that one might hope for, but a sense of our causal responsibility for damage to others is more likely to bring us to a sense of complicity that will lead to action. “Relationships of causal responsibility … trigger stronger senses of obligation than higherlevel ethical appeals can do,”5 and Dobson’s discussion – focusing for exemplary purposes on the causes of environmental damage – demonstrates the empirical basis for a theory of global harm.

4 5

Andrew Dobson, “Thick Cosmopolitanism,” Political Studies 54 (2006): 165–84. Ibid., 182.


Understandable and even promising though it is, however, this global harm principle invites two related questions at the level of political theory. The first is that the harm principle was developed, and has typically functioned, in the presence of other, complementary principles that can be taken for granted in the context of the state, but not in the interstate context; how to transfer the principle to the interstate context is, then, bound to be problematic, for a principle that may be useful for restricting action within a certain boundary may not be a useful choice of instrument for extending concern beyond it. The second is that its emergence has coincided, paradoxically, with increasingly skeptical thoughts about the usefulness of the idea of “harm” in the domestic context, and, in particular, about whether it can serve effectively and as a free-standing criterion for establishing the scope of duty. Many commentators now doubt that it can. This chapter reflects on these questions, and suggests grounds for some skepticism about the global harm principle. Nevertheless, it also claims, there is a distinctive if quite narrow interpretation of it that can usefully help to specify an important limit to one political society’s acts in relation to another. The discussion begins by briefly noting some revisionist thoughts about “harm” in its original (Millian) context, which should alert us to problems that lie ahead; it goes on to consider, and reject, two alternative ways of transferring the harm principle to the global context (especially in relation to international political economy), and then to propose a third (and, it is argued, better, though narrower) way – one that identifies harm in terms of damage to another society’s self-organizing capacity; some practical implications are also noted. Finally it considers a background view of justice that would support the version of the harm principle that is proposed.

II In its appealing minimalism the proposal of a global harm principle is continuous, in one very important and basic respect, with the archetypal use of “harm” by Mill in On Liberty. For Mill proposed


the appeal to “harm” as a way to accommodate diverse views of life: given divergent views, we need to fi nd ways to defi ne the limits of our action on grounds that can be generally accepted. We should not appeal directly to our “likings and dislikings” in regulating action in society;6 we should, rather, appeal to a principle that allows different sets of likings and dislikings to coexist without oppression. That basic ambition relates easily enough to its proposed global counterpart, for here, too, the project is guided by a sense of diversity, in two ways in fact: the diversity among “overall conceptions of the good” that Barry refers to, and the diversity among cultural measures of the good that makes it necessary to identify some common standard. In the global context as in On Liberty, then, the harm principle is intended as a strong but minimally divisive restraint, potentially supported by consensus, that recognizes two equally indispensable considerations: the other-regarding nature of our actions, and the otherness of those whom our actions affect. Now there is no reason why the global use of “harm” should track, in every respect, its domestic use, and still less of a reason why it should precisely track its use in On Liberty. Nevertheless, given the basic structural similarity between the two projects, it would be surprising if the long-sustained and careful enquiry into its tradition of use could predict nothing at all about its global usefulness. And in fact, its tradition of use seems to yield two relevant questions, or warnings, both of which concern the capacity of “harm” to operate as a free-standing principle. First, of course, there are very familiar complaints about the open-textured meaning of “harm” and its malleability in the hands of its interpreters. In recent years that objection has been made with particular force by critics who want the principle to be reliably libertarian, but fear that the record shows that it is not. Richard Epstein7 6

7

John Stuart Mill, On Liberty and Other Writings (Oxford: Oxford University Press, 1991), 11. Richard Epstein, “The Harm Principle – And How it Grew,” University of Toronto Law Journal 45 (1995): 369–417.


and Bernard Harcourt,8 for example, have such serious doubts about the expansive use of the harm principle by courts (in the United States) that they recommend its entire abandonment. Whether we focus (as Epstein does) on the extension of the state’s regulatory scope to take account of “aggregate” harms rather than Mill’s own notion of individualized harm, or (as Harcourt does) on moralistic policies that can too easily drape themselves in the language of harm prevention, we arrive at serious doubts about the capacity of the harm principle to survive its own inflation, as they view it, and, thus, to enjoy any independent force in determining policy. But other interpreters go significantly further than this, by questioning whether the harm principle, understood, as by the critics quoted above (and by many others), as a free-standing and libertarian doctrine, actually captures what Mill himself proposed to accomplish in On Liberty. Commentators have noted that the term itself is not Mill’s own, but Joel Feinberg’s, implying in doing so that the term may artificially stabilize the complex and fluid argumentation of On Liberty.9 The business of defining “harm,” and of determining what kinds of harm count in justifying prohibition, may have to be carried out, they argue, not by trying to give practical defi nition to a single principle, but in the context of the larger strategy of the text as a whole. The line that Mill tries to draw, in their view, simply does not correspond to the line between harm and non-harm, for there are harms that are not to be prohibited, and non-harmful acts that are to be. Despite setting out, as he says, to assert “one very simple principle,” it is far from clear that Mill proposes to rely on only one principle in developing his view. For example, he relies on a background set of institutional duties in order to define the harm caused by negligence.10 He relies on a background account of “manners” to 8

9

10

Bernard Harcourt, “The Collapse of the Harm Principle,” Journal of Criminal Law and Criminology 90 (1999): 109–94. See Richard Vernon, “J. S. Mill and Pornography: Beyond the Harm Principle,” Ethics 106 (1996): 624; Daniel Jacobson, “Mill on Liberty, Speech and the Free Society,” Philosophy and Public Affairs 29 (2000): 276. Mill, On Liberty, 190.


explain certain kinds of victimless offenses.11 And most importantly, after neglecting the definition of “harm” in his first three chapters, he shifts the definitional burden at once to “rights” – it is harm “to interests that ought to be considered as rights” that it is the state’s purpose to prevent.12 So “harm,” we are warned, needs to be supplied with context in a way that cannot be forwarded simply by reflection on the meaning of the word – we need other principles, express or implied, that set the boundaries for its operation. But the second warning is perhaps even more pressing. In Mill’s own apparent view, the operation of the harm principle itself is embedded within a more basic normative principle, that of fairness. That emerges in a context of direct relevance to much of the discussion of the global context, that of trade. In trade, Mill says, we harm each other all the time. To elaborate on his point: entrepreneurs and managers cut their competitors’ profit margins and drive them into bankruptcy, landlords harm their tenants by raising rents, employers downsize their workforces with devastating consequences for those laid off – and, to anticipate later discussion, investment and employment decisions made in one country have harmful consequences for people elsewhere. Such market effects are justifiable, according to Mill, on the basis of an argument that he regards as “solid” but “different” from the harm principle;13 and while he does not elaborate it is easy enough to fill in the missing argument in terms of his acceptance of the classical economists’ case for market freedom. That some suffer immediate losses is a condition of a larger efficiency that in the longer run enhances opportunities for all. Now just how to fit this into On Liberty’s overall argument may be controversial,14 but I want to suggest a very simple way of doing 11 12

13 14

Ibid., 109. See John C. Rees, John Stuart Mill’s On Liberty (Oxford: Clarendon, 1985), and John Gray, Mill on Liberty: A Defence (London: Routledge, 1996), 48–57, for discussion. Mill, On Liberty, 105. Arthur Ripstein proposes, rather, that we bring competitive loss under the umbrella of an “autonomy” principle – a competitor who defeats me is not attempting to take control of my own powers, but, rather, is exercising her own powers more effectively than I am exercising my own. But while that is indeed


so: we must, Mill says – in an intriguingly contractualist-sounding passage – accept equality, and an equal sharing of burdens, as a precondition of political society;15 and given that, plus his background belief in classical free-market principles, the requirement to accept market damage – even if “harmful” by some valid and important measure – would follow. What is harmful, in short, depends on a background view of what citizens owe to one another in terms of an overall conception of common benefit. Of course, any short summary of what to make of such a closely interpreted text as Mill’s can only be suggestive: but what it suggests is so basically important to any harm-based approach that it deserves consideration. To sum up: it suggests two issues. First, if we abstract the harm principle from the context of an organized society we will have to reassess the boundaries of its application, for we can no longer rely on a background that so readily picks out its appropriate application and scope (the range problem). Secondly, even if we can establish a useful place for it, it may still be subject to control, and require justification, by a more comprehensive background theory (the normative incompleteness problem).

III Suppose, to fill in the notion of “harm,” we give weight, as Mill did, to the central value of autonomy; that would match an important part of the appeal of the harm principle in both domestic and global contexts – its respect for otherness. But what would it mean to say that one society should avoid harming the autonomy-interests of another? One possible answer can be seen off unceremoniously. That is the idea that states understood as governments-in-controlof-territories should be invested with autonomy-interests identical

15

preferable to accepting that Mill was willing to accept reductions in liberty for the sake of a greater good, it seems not to accommodate Mill’s express view that the defence of markets rests on “different” grounds than his defence of individual liberty, the view taken here. See his “Beyond the Harm Principle,” Philosophy and Public Affairs 34 (2006): 238–39. Mill, On Liberty, 83.


or analogous to those of individuals, and should enjoy self-determining power as a consequence. That simply cannot be, for at least two familiar reasons. First, while any legitimate interest the government might have could only be derived from the interests of those under its control, its own interest in maintaining control may diverge from the interests of those whom it controls. To give governments the protective cloak of self-determination may be to confirm them in their purpose of destroying the self-determining capacity of (some of) their subjects, so that the normative ground of the justifying analogy is violently removed.16 Secondly, a further disanalogy between the individual and collective cases resembles Rawls’ famous critique of utilitarianism:17 when individuals choose between the packages of costs and benefits connected with alternative decisions, they are compensated for the costs borne by the benefits that they themselves enjoy; but in the case of public decisions made by governments, benefits may be borne by some, the costs by others, so that they cannot be set in the same account. Sometimes, moreover, the costs selectively imposed by governments go far beyond what is acceptable; and these are, typically, just the cases in which the self-determining authority of states is properly challenged, that is, cases in which state power is turned against internal enemies with an atrocious level of violence. That such cases effectively cancel the self-determining power accorded to states is now very widely accepted, as we have seen, in discussions of international law and global ethics. Mill himself did not make the mistake in question, even though “A Few Words on Non-Intervention”18 may give the impression that he drew an analogy between the rejection of international intervention and the rejection of paternalism in On Liberty. In both cases, after all, the appeal is to what we must do for ourselves – others cannot make us free: “The individual/community analogy 16 17 18

Beitz, Political Theory and International Relations, 71–92. Rawls, Theory of Justice, 22–27. In Collected Works, vol. XXI, ed. John M. Robson (Toronto: University of Toronto Press, 1977), 109–24.


was very much in Mill’s mind as he wrote,” Michael Walzer says.19 But whatever the merits of Mill’s rejection of intervention in the international case, it does not depend on any parallel between state and individual in the sense of attributing to the former any feature analogous to autonomy. What he says is quite consistent with his individualism: people within a tyrannized country must learn that political freedom rests on their own efforts – as Walzer rightly goes on at once to say, he claims that “the members of a political community must seek their own freedom” (emphasis added), so that we are dealing, in effect, with an identity rather than an analogy. Mill makes that plain in relating the justifiability of intervention to the developmental level (as he views it) of the people comprising the society in question. So it would be mistaken to enlist Mill’s authority in favor of the parallel disputed above, for he proposes an aggregate case rather than a parallel one.

IV We must, then, translate the global harm principle in terms of harm to persons – not to their states. In doing so, we confront an initial obstacle. Suppose that one state’s duty not to harm cashes out as a duty not to harm the autonomy-conditions of persons in other states’ jurisdictions. That makes sense in that harm (to humans) is intelligible only as harm to human persons, since it must be something capable of being experienced, and if collectives are to be said to be harmed, then collective harm must at some point be translatable into harm to their members. But harm to individuals needs specifying in relation to the capacity in which harm occurs: in their capacity as persons simply, or in their capacity as members. Given the organization of the world into distinct political memberships, people suffer market effects as members of one society or another; and it is also plain that we implicitly subscribe to the view that in the fi rst resort the protection of members of a society is the responsibility of the 19

Walzer, Just and Unjust Wars, 87.


corresponding state. So why should some states, acting unilaterally or in coalition or via international institutions, assume this responsibility for themselves? If local states consent to (or acquiesce in) the local effects of the global market, can other states substitute their own judgment for theirs, as though their relation to other populations was unmediated by a regime of sovereign political institutions? If we give much weight to local state responsibility, then of course we may let other states off the hook rather too easily, by making their role only a secondary one. But on the other hand, if we make responsibility essentially universal and spatially neutral, then we may erase two important aspects of locality: the contributory effects of local institutions, and the cultural specificity of other nations. The first issue arises in connection with “complex responsibility”20 where one state’s contribution is entangled with the contributions of other agents, including the local state’s. Harm is often jointly caused, and when local authorities themselves are deeply complicit in causing it, the question of who has harmed whom may not be straightforward. If rich countries create conditions that local elites then exploit to the detriment of populations, can the harm that results be attributed to those who create the conditions? That question especially relates to an important proposal that Thomas Pogge advances,21 a proposal concerning international borrowing rights that is a key element in his global harm-reduction project. At present, any government enjoying sovereignty can borrow from international financial institutions: dictatorial governments can exploit this opportunity to their own personal benefit, while burdening their societies with the costs of repayment. A provision that would cancel the obligation to repay debts assumed by dictators, after their fall, would certainly curtail the willingness to make loans to them. Taking this to be true, can we say, though, that the present arrangements are themselves the cause of the harm in question? 20 21

Linklater, “Harm Principle,” 338. Pogge, World Poverty and Human Rights, 112–26, and “Preempting Humanitarian Interventions.”


It has been objected that “there has been oppression, often motivated by the sheer desire to rule, much longer than a global order has created incentives for it” and that “authoritarian predators are also thieves, but they are mostly oppressors.”22 So if wealthy countries were to take up Pogge’s proposal, should we describe that in terms of their ceasing to do harm, or, rather, in terms of their preventing harm by others? The question is significant, if we view the obligation not to cause harm directly as higher in priority than the obligation to prevent others from harming. Now on Mill’s own argument, there is a case for equating the tempting of others to harm with the act of harming itself, for he notes that special considerations arise when we allow there to be “a class of persons” – such as, in his example, pimps and casino-owners – with a direct interest in promoting (what Mill takes to be) harm; and it would not take much to connect this point with Pogge’s argument about international borrowing rights and incentives for dictators. But all the same, we should also note Mill’s own thought that here we are somewhere close to the disputable edge of the harm principle, “an exact boundary line,” as he puts it, where joint agency distributes causation and brings an inconclusive balance of considerations into play.23 What we need here, to establish responsibility in the moral sense, is an argument that is independent of the causal argument. We need some argument that will ground and limit the actions of states towards each other. This issue is strongly analogous to that of joint causation in the ethics of war, when damage, or much of it, is caused through the mediation of the opponent’s response to one’s own action:24 it requires us to develop a view of the legitimate extent of national selfpreference that is partly independent of the destructive responses of other societies to what we propose to do. That view might turn out to be very restrictive of national self-preference, as in my view it 22

23 24

Mathias Risse, “Do We Owe the Global Poor Assistance or Rectification?” Ethics and International Affairs 19 (2005): 16. Mill, On Liberty, 110. Thomas Hurka, “Proportionality in the Morality of War,” Philosophy and Public Affairs 33 (2005): 46–50.


would, but the present point is that it cannot be based on explaining the idea of “harm” alone. The second potential objection is that the idea of (personal) harm may be too strongly connected with the moral and political cultures of Western societies which attach stronger value to the protection of individuals than more “communitarian” societies may do; there is, after all, something of a stereotypical view that tells us that it is Western-centric to put individuals first. To this, however, Andrew Linklater convincingly replies that the evidence is to the contrary: moral and political beliefs other than Western ones attach importance to the protection of individual persons from dangers to which they are particularly vulnerable.25 To suggest otherwise could in fact be an alternative form of Western-centricity that defines other cultures as lacking in humanity. The Western-centricity that sees difference where there is none is at least as bad, surely, as the Westerncentricity that denies it where it exists – it “others” in a different way, but no less culpably or destructively, and perhaps more so. Such arguments present us with the need for a threshold. Given that local states have primary responsibility for harm prevention, there has to be a significantly high threshold for transnational responsibility. Given that different cultures may interpret what is “harmful” differently, there has to be an intercultural threshold that can form the basis for in-principle agreement among states, if we are to arrive at a workable rule for international behavior. Neither of those requirements seems at all excessive, if Linklater’s intercultural evidence is as sound as it seems to be; but the point at which they seem likely to set the threshold is also likely to make the harm principle redundant. There is evidence of widespread agreement about the evils of genocide, apartheid, violence against women, and torture26 – we scarcely need to rely, after all, on a culturally specific defi nition of what personal “autonomy” requires in order to condemn such things. (The practice of arranged marriage, for example, 25 26

Linklater, “Harm Principle,” 331–36. Ibid., 329–30.


or of certain forms of child labor, might fall into a different category, for here we would have to take culturally different ideas of “family” into consideration.) But to raise the threshold of harm in order to secure universality is to move, effectively, to a rescue principle: for when people suffer deprivations severe enough to cancel any value that can plausibly be attached to their state’s autonomy, and which, in addition, amount to severe deprivations by any cultural standard, then we are brought within the reach of a duty to respond that is independent of the matter of how the harm in question was caused in the fi rst place. That such evils exist is a sufficient reason to respond, whether we caused them or not. But may we reach such a rescue principle by way of the harm principle, by extending the idea of harm to include harm by inaction as well as by action?27 That is compelling when – as in Mill’s argument28 – it is the case that inaction violates a definitely assigned duty; without that qualification, however, the harm principle seems to lose any distinctive capacity to pick out a duty-bearer, for it becomes, effectively, a suffering-reduction principle. We would need, then, some independent line of argument that would assign a duty in the first place, and the harm principle itself cannot do that for us. It is true, certainly, that the obligation not to harm extends to an obligation to repair harm that has been done;29 but the duty of repair does not necessarily fall to the same agent – when, for example, a third party is in a better position to repair30 – and is not conditional on the reparative agents’ having done harm themselves. At most, the question of who caused the harm might lead us to assign special prima-facie responsibility for remedying it, rather as former colonial powers sometimes seem to feel a special responsibility for responding to disasters in former colonies that they had pillaged. But that has what value it has

27 29

30

28 Ibid., 342–32. Mill, On Liberty, 190. Henry Shue, Basic Rights, 2nd edn. (Princeton: Princeton University Press, 1996), 52. David Miller, “Distributing Responsibilities,” Journal of Political Philosophy 9 (2001): 453–71.


only as a way of solving a coordination problem – given that many could respond, but all should not, then who must? – not as a groundlevel principle. Its subordinate nature is made clear by the fact that sometimes it may not always even be morally or politically acceptable that it should be the former aggressor that undertakes the task of peacebuilding – better that someone with cleaner hands should. In both respects, then, the harm principle falls short of what the argument needs. Given that not all policies or transactions are positive-sum in their consequences, we need to develop a position from which, at least in principle, we can assess the legitimacy of state actions, even when they have effects that fall short of the extreme and massive harms that can be consensually condemned.

V We can move towards such a position – surprisingly, perhaps – by giving weight, fi nally, to what is perhaps the most forthright objection of all to the global harm principle. Challenging a growing body of literature about the causes of poverty, Mathias Risse provocatively advances what he terms the “Institutional Thesis,” that is, the claim that the most important determinant of a society’s welfare or illfare is not to be found in external factors – such as the actions of other states – but in its own institutional character. As he defi nes the thesis, “Growth and prosperity depend on the quality of institutions, such as stable property rights, rule of law, bureaucratic capacity, appropriate regulatory structures to curtail at least the worst forms of fraud, anti-competitive behavior, and graft, quality and independence of courts, existence of trust and social cooperation, and thus on the overall quality of civil society.”31 That principle is to be distinguished from rival theses that place causal primacy, rather, in facts of geography, or else in the consequences of global market effects. In that last respect, of

31

Mathias Risse, “How Does the Global Order Harm the Poor?” Philosophy and Public Affairs 33 (2005), 355.


course, the Institutional Thesis is meant to offer a basic challenge to the claim that poor countries are poor mainly because external factors impinge upon them, the claim that, of course, initially brings a harm principle into play. Now no single discussion, let alone a discursive one such as this, can hope to adjudicate between the empirical claims at issue here. But if – or to the extent that – the Institutional Thesis is plausible, it points the way to a global version of the harm principle that can better survive objections. Among the serious consequences of one country’s actions for another is damage to, precisely, the latter’s institutional capacity. One country harms another, we may now propose, if it damages the other’s capacity to enjoy a path of development that flows from its own values and choices, by undermining or overwhelming it: and its capacity is above all a matter of the effectiveness of its public institutions. One society harms another, in other words, when it damages the legal, organizational, and political arrangements essential to its common life, or exposes them to strains that they cannot bear, or deprives them of the conditions for their improvement. There can be no doubt that societies’ institutional capacity is harmed in this way (see below); so to accept this line of thinking is not to commit the common and morally evasive fallacy of what Pogge terms “explanatory nationalism,” that is, the fallacy of supposing that all explanations begin at home.32 While some may begin at home, many certainly begin elsewhere – but a major effect of those that do begin elsewhere is their destruction of the local society’s collective capacity to respond. This argument can not only absorb the Institutional Thesis, then, but actually welcome it as an important possible clarification of the global meaning of “harm.” For it points towards an approach that does not propose to measure “harm” on the basis of the balance of pluses and minuses imposed on weaker countries by the global political economy. What it says about harm is entirely consistent 32

Pogge, World Poverty and Human Rights, 139–44.


with the view that, on some defensible measure or other, the South or some part of it may in some ways be better off than it was n years ago, as a result of its participation in world-wide trade, and its claims would be unaffected by the demonstration, or the disproof, or even the essential contestability, of that view. For a society’s capacity for autonomy may be negatively affected despite – or even perhaps, in some cases, because of – processes that bring about its relative enrichment. The same is true, of course, of individuals: there are certainly ways in which you can harm my autonomy-interests even if, in the process, you make me better off (by some measure, even a reasonable measure) in the process of doing so. If that were not so, in fact, then On Liberty would have no case against paternalism, for that case could make no sense at all unless it were possible for you to harm me even if, by some (even reasonable) measure or other, you improved my well-being or prospects by doing so. I could also dispute your measures, of course, and am very likely to do so, since they would seem ex hypothesi (in that case) to differ from my own, but that would not be what my case depended on. It would depend on respecting my power to choose. We may think of autonomy, then, as the effective capacity of a whole society to enjoy conditions in which its development reflects broadly supported values and choices of its own. That is also, arguably, the most natural way to extend the harm principle from the individual to the collective case. For its point, in On Liberty, is above all to support an approach to law and policy through which people can live lives that are “their own,” not lives that are imposed on them by the likings or dislikings of others. On Liberty, it is true, accommodates both weak and strong conceptions of what it is for something to be “one’s own.” For, sometimes, Mill stresses persons’ intrinsic (unchosen) features, as when, for example, he refers to the “physical and spiritual conformation” of a person as a given attribute, or implies that people, like trees or rivers, have a natural shape that demands respect; sometimes, however, he stresses the importance of choice, with the implication that a character or way of life is one’s


own, and to be respected as such, because it reflects a discriminated choice.33 Later theories of autonomy have found it important to distinguish between these conceptions,34 for they are not equally stringent in what they require. Nevertheless, Mill believes that to harm people is to diminish their ability to live lives that are “their own” in either of those senses. Likewise, to harm another society is to diminish its corresponding capacity – whatever that is taken to be.

VI But what is it? First, what is meant by a society’s autonomy, and, secondly, what amounts to harming it? It is of course essential to the view developed here that there should be a way to distinguish the independence or self-organizing power of societies from the autonomy of states, rejected above as an illicit group–person analogy. States, in the argument adopted here, are to be understood as usable elements in a society’s self-organizing project, not as impervious international personae, or unopenable black boxes. And so, in the fi rst instance – qualifications will soon follow – we should turn to the issue of democracy. To the extent that a society’s rulers are accountable to mass approval or disapproval, and to the extent that the choice of policies adopted for it is affected by public opinion, we may see an analogy with personal autonomy. Of course the analogy is imperfect, for reasons touched on briefly above: the costs and benefits of policy choice distribute among, not within, persons. But although imperfect, it is still meaningful: in the end, there is a difference, surely, between societies in which public opinion matters and societies in which it does not, and the basic institutions of democracy provide one obviously important means of establishing that difference, and of establishing some identity between state decisions and the interests of civil society. “Democracy” is taken here in no very strict sense: it simply refers to a range (or some mix within 33 34

Mill, On Liberty, 75, 69, 65. See for example Harry Frankfurt, The Importance of What We Care About (Cambridge: Cambridge University Press, 1988).


that range) of practices that make the exercise of power constitutionally conditional – unrigged elections, significant media freedom, the acceptance of legitimate opposition. In chapter 5 above it was argued that intervention to impose democracy is not justified: the obligation not to destabilize democracy where it exists, or is nascent, rests on the same principle of self-determination. So what are the conditions for democracy? Here the argument can advance only by way of some speculation about causes. A famous proposal, made long ago by S. M. Lipset, was that democratic development followed economic growth, particularly because economic growth tended to change a society’s shape from ‘pyramid’ to ‘diamond’ – that is, from a shape with small apex and large base, to a shape characterized by a larger middle class between apex and base.35 (This, Lipset proposed, favored democracy by relieving polities from the pressures of either the hierarchical right or the populist left; middle-class occupations, moreover, require educational qualifications, so that members of the middle class will tend to have the democratic attitudes that education is said to foster.) Now few if any analysts today would be willing to adopt the Lipsetian package – we need a much more nuanced case-by-case analysis, recent scholarship suggests;36 but in modified form, the claim that democratization depends on an economic context is a persistent one. In a particularly thorough and authoritative account, Przeworski et al. argue that the most significant single variable in explaining stable democracy is a secure economic margin: democracy is unlikely to emerge, or, where it exists, is fragile, when subsistence is insecure.37 Other accounts complicate the picture, certainly, in stressing the different impact 35

36

37

Seymour M. Lipset, “Some Social Requisites of Democracy: Economic Development and Political Legitimacy,” American Political Science Review 53 (1959): 69–105. E. F. Kisangani, “Economic Growth and Democracy in Africa: Revisiting the Feldstein–Horioka Puzzle,” Canadian Journal of Political Science 39 (2006): 855–81. Adam Przeworski et al. (eds.) Democracy and Political Development: Political Institutions and Well-Being in the World (Cambridge: Cambridge University Press, 2000).


of concomitant variables, such as economic inequality, which has a destabilizing (and de-democratizing) effect even when it is accompanied by economic growth, as at certain stages it may tend to be;38 or in warning of the socially polarizing effects of precipitous change that lead to the politically destructive scapegoating of minorities. 39 But even with such important qualifications, the general if imperfect connection between economic security and democracy may be among the best that we have. It is at the very least impossible that we would promote democracy by causing economic dislocation. If that connection holds, what follows for the global harm principle? It would follow that wealthy countries would have a duty to abandon policies that either suppressed economic growth in poor countries or directly exacerbated inequality within them. The former category would include, notably, policies of agricultural subsidy that deny agricultural sectors in poor countries a chance to compete fairly. It would also include denying poor countries the resort to protectionist policies that, in the longer run, might allow them to approach levels of productivity already achieved by rich countries, instead of requiring them to face direct competition with economies enjoying greater sophistication and leverage. The latter category would include structural adjustment policies that stripped the local poor of traditional margins while opening up lucrative entrepreneurial opportunities for the local rich. The duty to avoid policies in both of those categories is, I have argued, the closest possible approximation to what a global harm principle would require; and a harm principle focused on autonomy, in the sense proposed, is the most persuasive way to reach it. It does not make the principle hostage to empirical (or metrical) disputes about the history of global wealth and poverty, it does not depend on any fallacious analogy between

38

39

Edward N. Muller, “Economic Determinants of Democracy,” in Inequality, Democracy, and Economic Development, ed. M. I. Midlarsky (Cambridge: Cambridge University Press, 1997). Amy Chua, World on Fire: How Exporting Free Market Democracy Breeds Ethnic Hatred and Global Instability (New York: Doubleday, 2003).


persons and societies, nor does it ignore the fact that harm to persons is mediated through their national memberships. Rather, building upon that fact, it posits national societies as groupings of persons with legitimate desires for making their interests known and winning respect for them, and it defi nes the duties of the more privileged national groupings of persons in that light. They have a duty not to impede that legitimate desire, and to constrain their own legitimate profit-seeking desires so as not to undermine political development elsewhere. But the duty not to impede the emergence and stability of democracy requires correction in several related ways. First, “democracy” here is not a normative preference but only a plausible means to identifying collective autonomy: in the context of this argument, it only helps to satisfy an informational requirement – with democracy we may have more reasonable assurance that state decisions are broadly reflective of social values. We should, however, be open to the possibility that democracy in something like the Western sense is not the only way of achieving that, and that stable traditional societies may have evolved effective informal mechanisms of consent that would be undermined, objectionably, by the same factors that undermine democracy. While it may be hard to find (non-fictional) examples of whole systems corresponding to Rawls’ “decent consultation hierarchies,”40 at least the role of traditional mechanisms within polities must be explored.41 Secondly, even to the extent that the eventual emergence of formal democracy is a value, it is unlikely that assessments of local factors, positive or negative, can be accurately made at a distance. And thirdly, it is a condition of autonomy in the individual case, and by somewhat weaker extension in the collective case, that there should be consultation – the latter case, however, being complicated, of course, by the issue of whom to consult. 40

41

John Rawls, The Law of Peoples with “The Idea of Public Reason Revisited” (Cambridge, MA: Harvard University Press, 1999), 71–8. See the work of the Traditional Authority Applied Research Network, University of Calgary, www.ucalgary.ca/~taarn.


For information-dependent reasons of these kinds, the duty not to harm can be effective only where less-developed countries are adequately represented in the making of global policies.42 Their inadequate representation is a common theme in the analysis of international institutions, one taken up in, for example, Peter Singer’s well-known recent work.43 More detailed studies reinforce it. Constantine Michalopoulos, for example, notes that many countries lack any representation,44 while S. Javed Maswood complains of their deficit of negotiating capacity (in the World Trade Organization) even when they are represented;45 Aaditya Mattoo and Arvind Subramian stress, rather, poor countries’ relative lack of bargaining power.46 Behind all such complaints is, of course, a sound Millian point, from On Representative Government rather than On Liberty, but surely consistent with the harm-reduction project, that no one can have “much room in their thoughts for anything which they can with impunity disregard.”47 The general duty not to harm, then, flows into a further duty to take institutional steps to discover what it is that one is doing, from the standpoint of those to whom one is doing it.

VII None of these prescriptions is novel, of course; and the route sketched above is by no means the only way to reach a broadly similar set of

42

43 44

45

46

47

Darrel Moellendorf, “The World Trade Organization and Egalitarian Justice,” in Global Institutions and Responsibilities, ed. Christian Barry and Thomas Pogge (Oxford: Blackwell, 2005), 150. Singer, One World, 75–77. Constantine Michalopoulos, Developing Countries in the WTO (New York: Palgrave, 2001). S. Javed Maswood, The South in International Economic Regimes: Whose Globalization? (New York: Palgrave Macmillan, 2006). Aaditya Mattoo and Arvind Subramian, “The WTO and the Poorest Countries: The Stark Reality,” IMF Working Paper 04/81, 2004. See also Ngaire Woods, “Held to Account: Governance in the World Economy,” in Global Responsibilities: Who Must Deliver on Human Rights?, ed. Andrew Kuper (New York: Routledge, 2005), 251–66. Mill, On Liberty, 330.


views about global justice. Similar or overlapping pictures could be based directly on democracy as a value, for example; or on a strong conception of cosmopolitan citizenship; or on a liberal version of nationalism; or on a global version of tolerance; or on theories of sovereignty as conditional – to mention only the best-known current alternatives. But this chapter has certainly not tried to argue that the route from the harm principle is the only one, or even that it is the best. The object has only been to explore where we end up if, impressed by the harm principle’s minimalist appeal, we attempt to globalize it. The conclusion is that if the principle is to take a defensible form then we do not end up where some proposals suggest, while we do end up at or near a place that is usually reached from different starting-points – starting-points that are less minimalist, and for that reason less potentially attractive to those who advocate the harm principle. But we noted above, in briefly discussing Mill, that the harm principle, for all its minimalism, cannot, when pressed, claim to be a free-standing one. This account of its global counterpart confirms both of the aspects of that warning that were distinguished. First, the range problem: the narrow range assigned to it above is tolerable only if it is supplemented by other principles. In particular, it needs to be supplemented by a principle of rescue that is based on considerations of need alone, not on the negative duty not to harm. Other humans cannot simply be put beyond the reach of help by considerations about who is (causally) responsible for their desperate circumstances. It is particularly important to stress this, as the argument developed above has the effect – as far as the harm principle goes – of giving weight to the dominant general value-orientations of other societies, as a factor of obvious importance to assessing their selfdetermining capacity. The oppressive potential of that weight need hardly be stated, surely. Some value-orientations oppress and kill people. But that is not captured by considerations about how one society harms another. It falls under the potential reach of a rescue principle that requires stronger premises, the conclusions from


which would of course need to be moderated by the very practical factors that would determine the likely hopes – and thus the (political) legitimacy – of aid or intervention. As for other principles, at least two have been untouched in this discussion. Given a history of colonialism there are questions about potential duties of restitution, arising neither from present harm-causing behavior nor from absolute distress, though either of those things would of course reinforce any claims that it had; would we have responsibilities arising from acts in the (perhaps quite distant) past to which we or our institutions can plausibly be connected? And further, we may need to make room for a principle of non-exploitation to take account of imbalances that fall short of the defi nition of harm that was argued for above.48 The second consideration (about normative incompleteness) concerns the harm principle itself rather than its necessary supplements. It is not enough to show, as was attempted above, that one application of the principle is more defensible than others. For that application depends on other elements that require defense, and if they cannot be defended then the application – better than others or not – cannot be justified. It takes for granted, on the one hand, an international system composed of distinct political systems whose citizens have stronger duties to one another than to outsiders; and it assumes, on the other hand, that members of those societies will acknowledge the moral claims of others and so accept constraints on their advantage-seeking behavior. Neither of those conditions could be derived from the harm principle itself: they are among the range of background assumptions that have to be relied upon to arrive at any application of that principle, for to apply it we have to discriminate among harms. This does not, in my view, undermine the principle; Mill himself, as we saw, while “asserting” only one principle, had to rely on others, especially a principle of reciprocity that the 48

David Miller, “Justice and Inequality,” in Inequality, Globalization, and World Politics, ed. Andrew Hurrell and Ngaire Woods (Oxford: Oxford University Press, 1999), 187–210.


bare idea of harm itself does not supply. But it does remind us that a consensus about harm avoidance still permits serious disagreement about background conditions, even if the prohibition against harming others is intuitively powerful. Both co-citizen preference and cosmopolitan concern can be justified in a number of ways, including some that were mentioned above. But I want, finally – drawing upon the previous chapters – to sketch a view that, while it cannot be derived from the harm principle itself, seems naturally allied with it, and makes the fewest possible additional assumptions. Suppose we say that one person’s freedom should be limited only by another person’s vulnerability to harm; and also that everyone’s freedom and vulnerability count. We may then see organized political societies as local efforts to minimize vulnerability under conditions in which personal freedom is also secured. They impose constraints on the freedom of outsiders, however, and also make them vulnerable to external effects, without extending to them the political and legal protections that co-citizens get. If everyone’s freedom and vulnerability count, then clearly a justification for this is needed; and a natural one to give is that outsiders have the capacity to create and sustain political societies of their own, thus securing the same mix of benefits. If that is the justification, then, it morally precludes actions that prevent other societies from creating or sustaining the kinds of political arrangements that, on balance, improve their own mix. So the recommendation arrived at above, in discussing how the harm principle could be globally applied, could be supported by a larger moral argument that is likely to appeal to those who fi nd the harm principle attractive in the fi rst place, since it draws in part upon the same conception of political society. To repeat, though, that larger argument cannot be a conclusion of the harm principle itself, as opposed to a general view that is congenial to it. But we should not take that to mean that the global harm principle, because it needs the support of a more general view, is parasitic or vacuous. A principle can be important, after


all, without being independent, exhaustive, or foundational. As was argued above, it is neither independent nor exhaustive nor foundational in Mill himself. In fact, that connection can be drawn more tightly still. The view that would support the global harm principle is in some broad sense a contractualist one: it rests on a background assumption of equality, it implies that institutions’ purposes are defi ned in terms of benefits to their hypothetical founders, and it expresses the belief that arrangements must be publicly justified to those whom they affect. If, as some recent scholarship has come to suggest,49 Mill’s archetypal harm principle itself implies contractlike elements, then the global version has an authentic place in a Millian tradition – sharing in both its great intuitive force and its normative incompleteness.

49

Vernon, The Career of Toleration, 90; Ripstein, “Beyond the Harm Principle,” 224; Steven Lecce, Against Perfectionism (Toronto: University of Toronto Press, 2008), 71–74.

Conclusion: citizens in the world

This book has argued that our associative obligations are not selfjustifying, but invite a larger context of moral reflection, a context that will lead us sometimes to endorse and sometimes to qualify them; and sometimes even to strengthen their demands, given the role that the larger context may call upon states, and their citizens, to accept. Defenders of associative ties have, of course, resisted this initial contextualizing move, and they have been right to do so, it was acknowledged, when associations have been theorized as nothing but vehicles for the exchange of benefits – clearly, they are more than that. While the conferring of benefits may be a good way of justifying the existence of associations in general, the legitimacy of particular associations rests, rather – it was claimed – on the fact that, in rendering their members vulnerable in several important respects, they can rightly call on their members to exercise special concern for each other. Groups of citizens are particularized by shared exposure to risk. Such a view is best captured by a revised version of social contract theory, one that pictures political society in terms of an implied waiver of background rights, a waiver by virtue of which cocitizens acquire special obligations in return for subscription to an inherently risky project. That picture of political society, however, introduces a global dimension of concern – “cosmopolitan regard.” For it implies a precontractual and general right, a right not to be disregarded in the making of arrangements that are potentially to one’s disadvantage. It implies a general right to be given justifications. This challenges the view that a social contract has moral implications only for the contracting parties, and also the view that only a global contract can capture the moral entitlements of everyone. The background


cosmopolitan regard, it was argued, brings into play a logic of iteration: we can justify the exclusiveness and self-preference of our own society only if other societies can, likewise, seek their own best solutions to the balance of risks and benefits in political association, and give special weight to their own shared and exclusive interests in doing so. The implications of this line of argument were then explored in three practical contexts. First, humanitarian intervention: rejecting the view that the justice of intervention discredits contractualist political theory, it was argued that, on the contrary, it is contractualism’s “iterative” element – its necessary concern for the success of others’ parallel projects – that can justify intervention, without incurring some of the problematic burdens that the (rival) “naturalduty” approach may incur. Acknowledging that theoretical adequacy entirely fails to guarantee practical worth, the argument went on to discuss two further areas in which pre-emptive options to intervention are available. The idea of “crime against humanity” provides what may be the clearest of all examples of the “risk” thesis worked out above: state power is among the most powerful magnifiers of risk, and deterring or stigmatizing its atrocious use is a central project of international criminal law that – on the contractualist argument adopted here – should secure the support of states. Finally, the book’s argument turned from the questions of international aid to the question of international harm. On this book’s argument, it should be easier to sustain a no-harm thesis than a requirement to aid, but what “harm” means in a global context is not immediately clear. A solution was offered, however, in terms of the basic logic of contractualism – what one should not harm is the collective capacity of another society to form and execute its common project. Despite its focus on the state and on the requirements of citizenship, this book claims to be a cosmopolitan one. Other versions of cosmopolitanism, of course, do not require (and often exclude) the idea that a person’s connection to global society is mediated by the state, or by (literal) citizenship: people are, it is claimed, “citizens of

Conclusion 195

the world.” Why should we not take that (generous) view? The idea of “citizenship of the world” is a powerful one that – just because it is powerful – comes to be refracted through many disparate concerns and thus to take diverse moral and political forms in contemporary thought. In the conclusion to this book I attempt to situate the argument above in relation to some influential ideas of “world citizenship,” offering a view – indicated by the conclusion’s title – that on balance may be a more accurate way of thinking of what has been argued. Some initial doubts about “citizenship of the world” were briefly noted in the introduction, in particular, an element of paradox that it seemed to contain from its inception. To the extent that one’s relation to the world is citizen-like, the idea of citizenship has to be stripped of the distinctive features that make it normatively attractive in the first place. The idea of citizenship in the world, I shall suggest, escapes that paradox, while still drawing attention to the moral necessity of political duties that are unconfined by states’ borders. Several typologies of cosmopolitanism are available in the literature, and without adopting any of them, but drawing gratefully from several of them, I would like to distinguish between five variants, each of which is internally quite diverse. These are: the idea of moral equality, including its expression in law; the idea of a common standard of moral appreciation, understood at both personal and political levels; the idea of decenteredness, again both personal and political; the idea of global interdependence; and the idea, or fact, of common global institutions. The argument of this book is differently situated in relation to each of these variants, and in some cases is differently related to subvariants too. First, the fundamental idea of moral equality, which is surely a necessary (though not sufficient) criterion for cosmopolitanism. Many commentators express it in similar ways: according to Sen, it is matter of “not excluding any person from ethical concern”;1

1

Amartya Sen, “Humanity and Citizenship,” in For Love of Country, ed. Cohen, 115.


according to Couture and Nielsen, it is “the idea that the life of everyone matters, and matters equally”;2 according to David Miller, as we have already seen, cosmopolitans claim “that every human being has equal moral worth, or in a different formulation that each person is equally an object of moral concern.”3 But that, according to Miller, is the “platitudinous” formulation of the view, one that has no distinctive consequences for political theory. Now it can certainly function as a platitude, and can be inconsequentially held, as it often is. But surely it is consequential if it is taken to be the only basic moral premise, and that in working out its political implications no other basic moral starting-points can be introduced. Suppose its basic status is taken to preclude any appeal to the moral significance of entities other than human individuals – nations or cultures for example – other than as ways of expressing human equality: then the idea has an edge (indeed, a controversial one). The argument of this book took that path; it proposed that national societies were justified, not as sources of moral experience, nor as embodiments of intrinsic associative value and meaning, but as ways of protecting human persons from dangers to which they are commonly vulnerable. It further argued that, while considerations such as that could justify political societies in general, what legitimated any particular political society was its members’ complicity in arrangements posing subsequent or postcontractual threats, a complicity calling for particular care for those affected, or potentially affected. If all of this demands any departure from the simple defi nitions above, it is only in relation to the idea of equal concern, for Richard Miller has proposed that the basic idea is better termed “respect,” since equal “concern” could well be taken to entail that all others had an equally direct claim on our attention, and would deserve equal 2

3

Jocelyne Couture and Kai Nielsen, “Cosmopolitanism and the Compatriot Priority Principle,” in Political Philosophy of Cosmopolitanism, ed. Brock and Brighouse, 183. Miller, “Cosmopolitanism,” 81. See also Toni Erskine’s succinct formulation in Embedded Cosmopolitanism (15) – it is a view about “who counts.”

Conclusion 197

weight in our practices and policies.4 Likewise, it was conceded above that “respect” is best thought of as a notion with procedural consequences, and that the outcomes flowing from it would depend on the constraints adopted for the procedures that we put to use in interpreting its meaning. The Stoics, and later Kant, adopted the term “law” to express the respect owed to others in virtue of their humanity. For the Stoics themselves “law” seems to have been as metaphorical a notion as the “city” to whose members it was said to apply; Kant had a more literal idea of it, and while it is too much to claim Kantian inspiration for the various legal restraints on states that have progressively accumulated – especially on their use of violence – it is quite common to apply a Kantian validation to them. Now if special obligations arise among co-citizens, it is clear that universal moral constraints cannot, in the normal case, be directly applied to them, for otherwise the very idea of a special obligation would make no sense. Everything that was done on special grounds could at once be undone on general ones (as Niko Kolodny has argued, in the case of resource distributions5). But there could still be limiting conditions, most simply understood as conditions under which the argument for special grounds loses its force. That is surely the case with “crime against humanity,” defi ned above as an idea that is forced into play when violent state actions run out of any possible justification. Cosmopolitan regard for other humans does not give us a possible institutional basis for conduct – hence the justifiability of states; but it cannot, logically, be cancelled by institutional arrangements that claim to be justified by it, for it remains latent within them. How and when it should emerge from latency is, of course, no simple matter. For the states whose leaders and agents fall under the jurisdiction of international criminal law are the ongoing containers for human societies with local needs that 4

5

Richard Miller, “Cosmopolitan Respect and Patriotic Concern,” in Political Philosophy of Cosmopolitanism, ed. Brock and Brighouse, 128. Niko Kolodny, “Do Associative Obligations Matter?” Journal of Political Philosophy 10 (2002): 250–66.


escape the snap-photography of “crime.” A crime takes place in a moment, whereas a “criminal” regime may last for years. A crime is committed by (a determinate number of) criminals, while “state crime” implicates an indefi nite number of persons whose complicity is colored in many shades of gray. A crime takes place against a background of general law-abidingness, and punishment for it is well conceptualized as restoring the background equality that the criminal act has disrupted; the “criminal state,” however, is one in which the apparatus of law itself has been generally perverted, so that criminal acts stand out not as violations of norms, but as normative.6 For all of these reasons, the immediate application of international criminal law to state agents’ acts would be not just politically unwise but conceptually questionable too. Historically, we are at a stage of necessarily dubious prosecutorial judgments, but they are necessarily dubious because we have not arrived at the settled character that domestic legal systems have come to enjoy. Of course the contrast between international and domestic criminal systems invites criticism of the former. But to meet these criticisms we have only to ask if the critic is unmoved to see the Milosevics, Pol Pots and Charles Taylors of the world go free, or if, on the other hand, would agree that it would be better if we could contrive a world in which norms were so clearly understood that hideous atrocity would be off the table as a way of getting or keeping power. Given that (obviously) long-term goal, there are bound to be very hard choices between the in-context best decision, and what the goal itself demands – choices that may resist the help of theory. The second cosmopolitan idea is that of universal moral appreciation. Into this category we may inject two views that are rather different, but are nevertheless, I believe, basically related. The first (and I think) less contentious view is that individuals are to be admired if they can extend their moral understanding to embrace practices that, because unfamiliar, once seemed objectionable, by 6

See Rhiana Chinapen and Richard Vernon, “Justice in Transition,” Canadian Journal of Political Science 39 (2006): 117–34, for a fuller statement of this view.

Conclusion 199

coming to understand that it was only their unfamiliarity that made them seem so. That view has an obvious bearing on the kind of education that the cosmopolitan point of view would suggest: to be educated as a world-citizen is to be exposed to a variety of ways of living other than one’s own, and to be shown what is attractive in them. Students of this mode of cosmopolitanism should take as their rolemodels those notable individuals whose lives have bridged diversity. To the extent that this open-minded ideal invites controversy, it is apparently because of worries about what it excludes. Those who advocate it believe that it is fully consistent with appreciation of one’s own culture or way of life. Martha Nussbaum, for example, writes: “None of the major thinkers in the cosmopolitan tradition denied that we can and should give special attention to our own families and to our own ties of religious and national belonging,”7 probably an acceptable claim as long as we deny “major” status to the Cynics. But others fear that cosmopolitan breadth is bought at the cost of traditional identity. On this question, which opens up the large topic of multicultural policy, and the equally large area of educational development, this book has nothing to say. If, though, as chapter 1 argued, the border between associative morality and the sense of humanity is a porous one, that would seem to cast doubt of a general kind on the claims of the advocates of purity. More closely connected with the book’s argument, however, is the political status of crossnational judgments as these affect international policy and law. Here more needs to be said, for the topic invites familiar charges of imperialism, and some positions taken here seem liable to tarring with that brush. Although defenders of cosmopolitanism stress its openness to cultural differences, presenting curiosity about otherness as one of its principal virtues, its critics maintain that its vice is, on the contrary, the suppression of otherness. (Interestingly, then, cosmopolitanism is accused of suppressing identity at the personal level, while at the political level, 7

Nussbaum, “Patriotism and Cosmopolitanism,” 135.


the charge is, rather, that identity is imposed.) Scott L. Malcolmson, for example, attacks the “imperial pedigree” of moral universalism, noting Kant’s frank remark (in “Idea for a Universal History”) that Europe “will probably legislate eventually for other continents.”8 Robert Pinsky describes moral cosmopolitanism as the local belief-system of the Western intellectual community.9 Gertrude Himmelfarb, likewise, sees so-called human rights and “humanity” (in its affective sense) as “predominantly, perhaps even uniquely, Western values.”10 In the long debate over these matters one may pick out several short answers on the cosmopolitans’ behalf. One is to insist on the distinction between the provenance and the validity of ideas: where they come from does not necessarily tell us where they apply. Another, stressed especially by Amartya Sen, draws attention to (what he is too polite to call) the implicit racism of the view that Asians lack humane values, as well as to the historical absurdity of supposing that the West is the home of tolerance.11 But “imperialism” is best viewed, surely, as a political rather than an epistemic category, and the question that this book needs to face is whether or not its argument supports the forcible imposition of one society’s beliefs on another society. The denial that it does can, I think, be supported by two main reasons. First, any sort of cosmopolitan view, however weak, is an individualist one in the sense that it considers human persons to have a value that is independent of their social or national or cultural milieu. It follows that its evaluative stance will demand that institutions be of value to their members. We may call this, following Anthony Appiah, “ethical” as opposed to “substantive” 8

9 10

11

Scott L. Malcolmson, “The Varieties of Cosmopolitan Experience,” in Cosmopolitics, ed. Pheng Cheah and Bruce Robbins Minneapolis (University of Minnesota Press, 1998), 237. Robert Pinsky, “Eros Against Esperanto,” in Love of Country, ed. Cohen, 85–90. Gertrude Himmelfarb, “The Illusions of Cosmopolitanism,” in Love of Country, ed. Cohen, 75. Sen, “Humanity and Citizenship,” and “Human Rights and Asian Values,” in Ethics and International Affairs, 2nd edn., ed. Joel H. Rosenthal (Washington, DC: Georgetown University Press, 1999), 170–93.

Conclusion 201

individualism,12 that is to say, it necessarily demands that individual welfare be the ground of rights, but not necessarily that individual interests be the subject of rights. It is entirely consistent with ethical individualism to value collective identities and projects, ruling out only the view that they be valued for their own sake. Even then, it will not demand that every institution must justify itself at the bar of individual welfare, only that institutions that impose avoidable hardships on their members be subject to moral questioning. That the hardships are valued by their members is a defense within ethical individualism, though it is obviously a defeasible one, as noted above. How that issue is to be negotiated is no doubt almost always difficult, but the difficulties are ones that are also faced by ethical nationalists who accept “thin” conceptions of human rights. They cannot even be avoided by ethical nationalists who reject human rights, except by making unrealistic assumptions of homogeneity, given that the necessity of traditional hardships is questioned within cultural communities too. There were Indians who passionately protested widow-burning. There were Chinese parents who viewed footbinding as barbarous (and helped to put an end to it by spontaneous collective action). So we can accept that the cosmopolitan will face profound questions about the justifiable claims of cultures on their members. The point, though, is that everyone does. The second consideration is that the argument of this book is not only ethically (as opposed to substantively) individualist, but that it takes collective self-determination as a central political value. According to the model that it has advanced, co-citizens are engaged in a project that, by virtue of its riskiness, requires them to devote special resources to it, and also to be guided by moral iteration: that is, they are required to come to the aid of other societies whose risk-taking has failed, and they are required not to add, avoidably, to the burden of risk that other societies carry. The measure, in both cases, is the other society’s capacity to engage in anything 12

Appiah, The Ethics of Identity, 72.


than can plausibly be called self-determination. That criterion is plainly not met in cases of state collapse or brutal repression. That criterion is violated by rich states whose economic policies impede political development in poorer countries. “Democracy,” as chapter 7 pointed out, figures here in an informational role only, and the model is quite consistent with acknowledging that there are ways of being collectively self-determining that are different from those of Western democratic polities. In short, whatever may be said about the “imperial pedigree” of moral universalism, I believe that as far as this version goes at any rate the charge of imperialism fails. The next version of the cosmopolitan picture is that of the decentered subject, again a core notion that takes both personal and political forms. In its personal form, it depicts the cosmopolitan self as hybrid and eclectic, no longer identifying (as is said once to have been the case) with a single place or role or history, but enjoying multiple connections while being fully defi ned by none of them. (Obviously, despite what is often seen as its post-modern and antiEnlightenment stance, this picture overlaps somewhat with the typical Enlightenment ideal of the urbane and open-minded person.) In what must be one of the most cited passages in recent political thought, Jeremy Waldron describes the life of a person living in San Francisco, of Irish ancestry, who “learns Spanish, eats Chinese, wears clothes made in Korea, listens to arias by Verdi sung by a Maori princess on Japanese equipment, follows Ukrainian politics, and practices Buddhist meditation techniques.”13 Lest this be thought of as no more than an account of consumption decisions (in San Francisco), it must be added that Waldron intends the eclecticism to reflect a conception of the self, one that he fi nds in Salman Rushdie’s celebration of “hybridity, impurity, intermingling. . . mongrelization. . . changeby-fusion, change-by-conjoining,”14 that is, a conception not just of taste but of the idea of the person’s multiple composition. Moving to 13 14

Waldron, “Minority Cultures and the Cosmopolitan Alternative,” 95. Ibid., 93.

Conclusion 203

the institutional level, we fi nd something of a complement to this in the idea of multiple transnational constituencies as developed, notably, by David Held. In this idea, the multiple sources of influence on people’s lives call for a new politics that is no longer state-centric, but which captures under democratic control the many and varied sites at which decisions are made for us.15 Once again, this book’s theme leads only to very weak conclusions about how selves are or should be, since the only idea of specific identity that it has insisted upon is that of civic identity. It is obviously essential to its argument that citizens continue to grasp what they owe to co-citizens, as a precondition for defining what they owe to outsiders. To what extent shared civic identity requires other shared identity seems to me to be an open empirical question. Certainly it lends itself more naturally to ideas of “constitutional patriotism,” in which shared citizenship is defi ned in terms of political values, than to nationalist models in which political society has to be underpinned by thick cultural commonalities. It does not however actually reject thick commonalities either (except, on this book’s argument, to the extent that they may prevent the global obligations of citizenship from being inferred – a hypothetical rather than an empirically grounded worry). On the question of multiple constituencies as distinct from multiple identities, however, the argument must take sides with the nationalist critics of Held’s transnational democracy. On very standard democratic premises, it certainly make sense that those affected by issues should have a say about them, and since issues are in fact decided at many sites it would seem to follow that multiple sites of participation should be available too. Some, it would seem, should be subnational, others national, yet others supranational, depending on the interests affected and the externalities involved. But in terms of the actual conduct of politics, David Miller is on strong ground in raising the 15

See, for example, David Held, Democracy and the Global Order: From the Modern State to Cosmopolitan Governance (Cambridge: Polity, 1995).


question of what we may term inter-issue reciprocity.16 This is not the same as the reciprocity-of-benefit view that was discussed and rejected in chapter 2 above. The question concerns the matter of winning and losing in a sequence of votes. If we vote repeatedly in the same constituency, we may sometimes win and sometimes lose, sometimes be allies and sometimes opponents. Wins compensate for losses, and are recordable, as it were, in the same account. But a loss in one constituency is not similarly recordable in the same account as a win in another, for while I have a reason to accept defeat at the hands of those whom I aimed to defeat by the same procedures, I have much less of a reason to accept defeat at others’ hands in a different forum altogether.17 Here we arrive at a special but politically important application of the idea of complicity that was stressed in chapter 3. When someone participates in arrangements that impose the risk of loss on others, an obligation to respond to what follows is, as it were, immanent in their own action. We have special responsibilities in relation to arrangements of which we are a part. That principle gives rise to responsibilities in formal democratic politics as it does in the broader informal “politics” of social coexistence. The principle that those affected by an issue should vote on that issue, on the other hand, is no more than a proposition of democratic theory, one that, defensible though it may be, cannot be said to impose immanent obligations on anyone. You are complicit because of what you have done, not by virtue of a theoretical requirement. It has to be acknowledged, however, that civic complicity is meaningful only to the extent that civic processes determine meaningful issues. To be complicit in the arrangement of deck-chairs on the Titanic, to use the time-worn simile, is not after all to be complicit in the larger decisions or non-decisions whose impact seriously 16

17

David Miller, “Bounded Citizenship,” in Cosmopolitan Citizenship, ed. Kimberly Hutchings and Roland Dannreuther (Houndmills: Macmillan, 1999), 76–8. See also Will Kymlicka, Politics in the Vernacular: Nationalism, Multiculturalism, and Citizenship (Oxford: Oxford University Press, 2001), chapter 11. For a fuller account see Vernon, Friends, Citizens, Strangers, chapter 6.

Conclusion 205

loomed. As noted above, the risk-management-based approach to political legitimacy clearly depends on the assumption that politics has a significant role in the mitigation and distribution of risk. Here we move to the fourth idea of cosmopolitanism distinguished above, that of global interdependence. The facts about that are too well known to need an account here. The vulnerability of this book’s thesis to such facts was admitted above. I think, to repeat, that it is a strength rather than a weakness in a political theory to be vulnerable to facts, and I cannot quite imagine what an invulnerable political theory would be like.18 Still, to fly in the face of facts would not exactly be a strength, so in chapter 2 some reasons (elaborated in later chapters) were offered for believing that states distribute risk in an interdependent world. Most dramatically, they can oppress their own people even more effectively, it seems, than they can oppress outsiders. They can legitimize or delegitimize resistance to events or threats, whether their source is internal or external. They control (or fail to control) the social mechanisms through which their citizens impact each other’s lives, sometimes in life-threatening ways. They exercise a deeper and more minute and specific control over their citizens’ freedom than outside agencies can (at least typically) do. And they serve as the guarantors of processes, formal and informal, by which majorities exercise control over the conduct of minorities’ lives. As always, when several criteria are invoked, we face a boundary problem when some are met and some are not, or when some or all of them are met only in a compromised way. But to the extent 18

I do not think this view denies G. A. Cohen’s argument in “Facts and Principles” (Philosophy and Public Affairs 31 [2003]: 211–45). Cohen’s argument challenges the view that “our beliefs about matters of normative principle . . . include matters of fact among the grounds of affirming them” (213), for the reason that facts can ground affirmation only because the “principle-grounding power” (214) of (some) facts arises from some further principle. As Cohen himself notes (230) his case is consistent (via “ought implies can”) with the view that facts can make a candidate principle impossible to follow, for his case concerns the grounding, not the excluding, of principles. In any event, “beliefs about matters of normative principle” can include practical beliefs about whether or not or how or when to act on them – the point made in the text – as well as beliefs about their truth.


that most of the states in the world meet most of these criteria, the risk-management view of their legitimacy has weight. But the facts of global interdependence have a still deeper bearing on this project’s claims, in relation to global duty. According to some views, some discussed above, it is the fact of interdependence that triggers global concern. Onora O’Neill’s remark – that we should not require the Vikings to have cared about the Incas – was cited, and it was suggested that it derives its comical effectiveness from the fact that the Vikings (despite being notable world-travellers) did not know that the Incas existed, and that, had they known, they would, given their disposition, likely not have cared. But while non-knowledge necessarily excludes concern, and colonist culture excludes it, what about the knowledge of interdependent relations with other societies, on the part of non-predatory peoples? Its motivating force, among segments of the population of Western societies, is evident. But the argument above denied that the facts of interdependence were essential, at the basic level, to global duties. An iterative argument, that draws attention to the resemblances between political societies, is different from an interdependence argument that draws attention to the influences between societies. Two examples were produced to illustrate the difference: the “North Korea” problem and the “spoiled brat” problem, the former intended to remind us of what might be owed to societies that have been removed from the web of global interdependence, the latter intended to raise the (more hypothetical) issue of a rich society divesting itself of obligations by retreating into isolation. The point of these two examples was not merely to draw attention to (real or hypothetical) exceptional cases; rather, the point was to bring into question what was at issue in the core cases, where the logics of both iteration and interdependence apply at once. If the iteration view applies to both core and exceptional cases, the argument ran, it is preferable to its rival. If so, then the moral role of interdependence, while obviously not negligible, needs to be reassessed. That we now know about the conditions under which our coffee beans are gathered, or our T-shirts sewn – whereas the Vikings knew

Conclusion 207

nothing of the Incas – reminds us, vividly, of a human connection. But how we should respond to that connection, it was argued above, is a topic that needs discussion in a way that is mediated through the facts of political association. I do not see how it can be wrong for members of one society to benefit from the bare fact that labor costs elsewhere are lower. To object to that is to imply support for an as yet unexplained alternative economic model. But to object to international economic arrangements on the argument outlined in this book seems to me to be based on grounds no less solid than the justification of co-citizen preference itself. If the economic terms that we impose on workers in other societies are such as to increase their vulnerability to control, by either domestic or foreign agents, thus diminishing their capacity to participate in collective self-determination, they are unjustifiable. Only in this way, I believe, can we fi nd an effective response to claims that international trade produces relative enrichment, claims on the basis of which globalization is placed, implausibly, beyond critique. If enrichment produces servitude, it is no more justifiable than a contract to exchange freedom for wealth in the domestic case. There remains, fi nally, the case of international institutions. As is very well known, important institutions above the state level have a profound impact on people’s lives, especially through their control of or influence on the terms of international trade and investment. As we have seen, there is controversy about whether or not that structure of institutions is sufficiently like a state that the full force of political justice applies to it. Although empirical evidence obviously comes into play in this debate, the question is in part a conceptual one: and I am not sure how to decide that an institutional structure is or is not sufficiently like a state. That is not to deny the significance of the debate, but only to say that a different route to global justice is adopted here, one that is drawn from the iterated features of states rather than from the state-like (or not) character of the international order. But the argument here is strongly on the side of global-justice theorists who call for the democratization of


supranational institutions; it is on their side, not because it takes a stand on the conceptual question, but because representation is essential to the detection, prevention, and remedy of international harms. Supporting effective representation is therefore, on this argument, an obligation arising from civic obligation itself, for that obligation carries with it an implicit duty not to harm, and thus to take steps that are necessary to that duty. It is a duty, then, of citizens of states, not of citizens of the world. That distinction, as noted above, is one which this book is to end with. It has argued that some important duties – not to harm, and to aid – are implicit in the obligations of citizens to one another: literal, not metaphorical, citizens, engaged in a project of collective self-determination, and thus needing to know how their project situates them, morally, in relation to the parallel projects of others. The argument has relied on a contractual model, which usefully points both to the moral universalism that its justification presupposes, and to the special ties involved in its legitimation. The approach taken to legitimacy, by way of the idea of the riskiness of association, leads to a further argument that specifies what cosmopolitan regard demands in the case of outsiders: their exclusion from the process of risk reduction is justifiable if they can engage in risk-reducing processes of their own. A simple consideration of good faith leads to the two conclusions about aid and non-harm, and to a defi nition of what harm would entail. If this line of argument carries conviction, the outcome that I most wish to emphasize is that the duties that it leads us to are political obligations. They are as binding as our obligations to other citizens, for they are sustained by the same considerations of political morality. The reasons for supporting them are not additional to, in conflict or else incommensurable with, the reasons for favoring co-citizens. Citizens should arrive at the reasons for global justice, as defi ned here, on the basis of reflection on what considerations are immanent in their own social and political practice as members of a political society.

Conclusion 209

Not all of our obligations are political. If the global obligations outlined here, while demanding, still seem too limited to reflect the response that humane people feel is called for by global deprivation, that is because they are. There are also obligations of humanity. Those are obligations that we should be open to as human beings who have no excuse for being unaware of the vulnerability of other humans whose desperately poor life-chances have been arbitrarily assigned. That poor life-chances impede one’s participation in collective self-development – the criterion of “harm” adopted above – is hardly the most salient of many reasons for condemning them, after all. But it is the reason that gives rise to a binding political obligation. As human beings, we also have duties of aid that we should strive to carry out through both personal contribution and political agitation. If this short book has said little about those duties, that is part because its concern has been to connect a somewhat more restricted range of global duties with our political memberships; and also because, acknowledging a perhaps rather unsympathetic treatment of Richard Rorty above, it is not political theory that is best positioned to convey, as distinct from reflecting on, our deepest moral perceptions. This book is agnostic on the matter of moral motivation. Perhaps Rorty was right, and the range and sensitivity of identification would best be extended by “imagination,” understood (controversially) to embrace poetry and the novel but not philosophy. Perhaps, rather, the motivational solution lies in the formation and valuing of transnational communities that will tend progressively to blur the categorical boundaries between “us” and “them”. Or perhaps, as others claim, the simple perception of a common humanity – even under conditions of extreme stress – can (sometimes) be powerfully compelling. From the perspective of this book’s argument, there is no reason to hope for the failure of any of these rival proposals. What I have argued for is a limited, but stringent, conception of political morality.

Bibliography

Appiah, Kwame Anthony. “Cosmopolitan Patriots.” In For Love of Country, ed. Joshua Cohen, 21–29. Boston: Beacon, 1996. The Ethics of Identity. Princeton: Princeton University Press, 2005. Arneson, Richard J. “Do Patriotic Ties Limit Global Justice Duties?” Journal of Ethics 9 (2005): 127–50. “The Principle of Fairness and Free-Rider Problems.” Ethics 92 (1982): 616 –33. Arendt, Hannah. Eichmann in Jerusalem: A Report on the Banality of Evil. London: Penguin, 1977. The Human Condition. Chicago: University of Chicago Press, 1958. Barry, Brian. Democracy, Power and Justice: Essays in Political Theory. Oxford: Clarendon, 1989. Justice as Impartiality. Oxford: Clarendon, 1995. The Liberal Theory of Justice: A Critical Examination of the Principal Doctrines in A Theory of Justice by John Rawls. Oxford: Clarendon, 1973. Beck, Ulrich. Risk Society: Towards a New Modernity. London: Sage, 1992. Beitz, Charles H. Political Theory and International Relations, 2nd edn. Princeton: Princeton University Press, 1999. Blake, Michael. “Distributive Justice, State Coercion, and Autonomy.” Philosophy and Public Affairs 30 (2001): 257–96. Bohman, James. “Republican Cosmopolitanism.” Journal of Political Philosophy 12 (2004): 336 –52. Brighouse, Harry. “What Rights (if Any) do Children Have?” In The Moral and Political Status of Children, ed. David Archard and Colin M. Macleod, 31–52. Oxford: Oxford University Press, 2002. Brock, Gillian and Harry Brighouse, eds. The Political Philosophy of Cosmopolitanism. Cambridge: Cambridge University Press, 2005. Brown, Chris. “Contractarian Thought and the Constitution of International Society.” In International Society: Diverse Ethical Perspectives, ed. David R. Mapel and Terry Nardin, 132–43. Princeton: Princeton University Press, 1998. Buchanan, Allen, Justice, Legitimacy, and Self-Determination (Oxford: Oxford University Press, 2004).

Bibliography 211

“The International Legitimacy of Humanitarian Intervention,” Journal of Political Philosophy 7 (1999): 71–87. Caney, Simon. “Global Distributive Justice and the State.” Political Studies 56 (2008): 487–518. “Individuals, Nations and Obligations.” In National Rights, International Obligations, ed. Simon Caney et al., 119–38. Boulder, CO: Westview, 1996. Charvet, John. “International Society from a Contractarian Perspective.” In International Society: Diverse Ethical Perspectives, ed. David R. Mapel and Terry Nardin, 114–31. Princeton: Princeton University Press, 1998. Chesterman, Simon. Just War or Just Peace? Humanitarian Intervention and International Law. Oxford: Oxford University Press, 2001. Chinapen, Rhiana and Richard Vernon. “Justice in Transition.” Canadian Journal of Political Science 39 (2006): 117–34. Chua, Amy. World on Fire: How Exporting Free Market Democracy Breeds Ethnic Hatred and Global Instability. New York: Doubleday, 2003. Cohen, G. A. “Facts and Principles.” Philosophy and Public Affairs 31 (2003): 211– 45. If You’re an Egalitarian, How Come You’re So Rich? Cambridge, MA: Harvard University Press, 2000. Cohen, Joshua, ed. For Love of Country. Boston: Beacon, 1996. Cohen, Joshua and Charles Sabel. “Extra Rempublicam Nulla Justitia?” Philosophy and Public Affairs 34 (2006): 147–75. Couture, Jocelyne and Kai Nielsen. “Cosmopolitanism and the Compatriot Priority Principle.” In Political Philosophy of Cosmopolitanism, ed. Gillian Brock and Harry Brighouse, 180–95. Cambridge: Cambridge University Press, 2005. Cox, Richard H. Locke on War and Peace. Oxford: Clarendon, 1960. Dagger, Richard. Civic Virtues: Rights, Citizenship, and Republican Liberalism. Oxford: Oxford University Press, 1997. Diamond, Jared. “Vengeance is Ours.” New Yorker, April 21, 2008, 74 – 87. Dobson, Andrew. “Thick Cosmopolitanism.” Political Studies 54 (2006): 165 – 84. Dworkin, Ronald. A Matter of Principle. Cambridge, MA: Harvard University Press, 1985. Taking Rights Seriously. Cambridge, MA: Harvard University Press, 1978. Epstein, Richard. “The Harm Principle – And How it Grew.” University of Toronto Law Journal 45 (1995): 369 – 417. Erskine, Toni. Embedded Cosmopolitanism: Duties to Strangers and Enemies in a World of “Dislocated Communities”. Oxford: Oxford University Press (published for the British Academy), 2008. Esses, Victoria M. and Richard Vernon, eds. Explaining the Breakdown of Ethnic Relations: Why Neighbours Kill. Malden, MA: Blackwell, 2008.

212 Bibliography

Frankfurt, Harry. The Importance of What We Care About. Cambridge: Cambridge University Press, 1988. Geras, Norman. Solidarity in the Conversation of Mankind. London: Verso, 1995. Gibbard, Alan. “Constructing Justice.” Philosophy and Public Affairs 20 (1991): 264 –79. Gilbert, Margaret. A Theory of Political Obligation. Oxford: Clarendon, 2006. Living Together: Rationality, Sociality, and Obligation. Lanham, MD: Rowman and Littlefield, 1996. Glover, Jonathan. “It Makes No Difference Whether or Not I Do It.” Proceedings of the Aristotelian Society, supp. vol. 49 (1975): 171–90. Goodin, Robert. Political Theory and Public Policy. Chicago: University of Chicago Press, 1982. Protecting the Vulnerable. Chicago: University of Chicago Press, 1985. “What Is So Special about Our Fellow Countrymen?” Ethics 98 (1988): 663 – 86. Gourevitch, Philip. We Wish to Inform You that Tomorrow We Will Be Killed with All Our Families. New York: Picador, 1998. Gray, John. Mill on Liberty: A Defence. London: Routledge, 1996. Green, Michael. “Institutional Responsibility for Moral Problems.” In Global Responsibilities: Who Must Deliver on Human Rights?, ed. Andrew Kuper, 117–34. New York: Routledge, 2005. Grotius, Hugo. The Law of War and Peace. Indianapolis: Bobbs-Merrill, 1975. Halldenius, Lena. “Locke and the Non-Arbitrary.” European Journal of Political Theory 2 (2003): 261–79. Harcourt, Bernard. “The Collapse of the Harm Principle.” Journal of Criminal Law and Criminology 90 (1999): 109 –94. Hardimon, Michael. “Role Obligations.” Journal of Philosophy 91 (1994): 333 – 63. Hardwig, John. “Should Women Think in Terms of Rights?” Ethics 94 (1984): 441–55. Hart, H. L. A. “Are There Any Natural Rights?” Philosophical Review 64 (1955): 175 –91. Law, Liberty and Morality. Stanford: Stanford University Press, 1963. Hehir, J. Bryan. “Military Intervention and National Sovereignty.” In Hard Choices: Moral Dilemmas in Humanitarian Intervention, ed. Jonathan Moore, 29–54. Lanham, MD: Rowman and Littlefield, 1998. Held, David. Democracy and the Global Order: From the Modern State to Cosmopolitan Governance. Cambridge: Polity, 1995. Held, Virginia. “Mothering and Contract.” In Beyond Self-Interest, ed. Jane Mansbridge, 287–304. Chicago: University of Chicago Press, 1990. Himmelfarb, Gertrude. “The Illusions of Cosmopolitanism.” In For Love of Country, ed. Joshua Cohen, 72–77. Boston: Beacon, 1996. Holzgrefe, J. L. and Robert O. Keohane, eds. Humanitarian Intervention: Ethical, Legal and Political Dilemmas. Cambridge: Cambridge University Press, 2003.

Bibliography 213

Horton, John. “Defending Associative Political Obligations: A Response to Richard Vernon.” Political Studies 55 (2007): 880 – 84. “In Defence of Associative Political Obligations: Part One.” Political Studies 54 (2006): 427– 43. Political Obligation. Atlantic Highlands, NJ: Humanities Press, 1997. “In Defence of Associative Political Obligations: Part Two.” Political Studies 55 (2007): 1–19. Hurka, Thomas. “Proportionality in the Morality of War.” Philosophy and Public Affairs 33 (2005): 34 – 66. Hutchings, Kimberly. “Political Theory and Cosmopolitan Citizenship.” In Cosmopolitan Citizenship,

ed.

Kimberly

Hutchings

and

Roland

Dannreuther,

120–42.

Houndmills: Macmillan, 1999. Hyntjens, Helen M. “Explaining the 1994 Genocide in Rwanda.” Journal of Modern African Studies 37 (1999): 241– 86. Jacobson, Daniel. “Mill on Liberty, Speech and the Free Society.” Philosophy and Public Affairs 29 (2000): 276 –309. Jones, Charles. Global Justice: Defending Cosmopolitanism. Oxford: Oxford University Press, 1999. Kant, Immanuel. Political Writings. Cambridge: Cambridge University Press, 1970. Keeley, Lawrence. War Before Civilization. New York: Oxford University Press, 1996. Kisangani, E. F. “Economic Growth and Democracy in Africa: Revisiting the Feldstein– Horioka Puzzle.” Canadian Journal of Political Science 39 (2006): 855 – 81. Kolodny, Niko. “Do Associative Obligations Matter?” Journal of Political Philosophy 10 (2002): 250 – 66. Kuper, Leo. Genocide: Its Political Use in the 20th Century. New Haven: Yale University Press, 1981. Kutz, Christopher. Complicity. Cambridge: Cambridge University Press, 2000. Kymlicka, Will. Politics in the Vernacular: Nationalism, Multiculturalism, and Citizenship. Oxford: Oxford University Press, 2001. Lakatos, Imre. “Falsification and the Methodology of Scientific Research Programmes.” In Criticism and the Growth of Knowledge, ed. Imre Lakatos and Alan Musgrave, 91–196. Cambridge, Cambridge University Press, 1970. Lecce, Steven. Against Perfectionism. Toronto: University of Toronto Press, 2008. Lenard, Patti. “Trust Your Compatriots, But Count Your Change: The Roles of Trust, Mistrust and Distrust in Democracy.” Political Studies 56 (2008): 312 –32. Linklater, Andrew. “The Harm Principle and Global Ethics.” Global Society 20 (2006): 329 – 43. Lipset, Seymour M. “Some Social Requisites of Democracy: Economic Development and Political Legitimacy,” American Political Science Review 53 (1959): 69 –105.

214 Bibliography

Locke, John. Second Treatise of Government. Indianapolis: Hackett, 1980. Luban, David. “A Theory of Crimes Against Humanity.” Yale Journal of International Law 29 (2004): 85 –167. Legal Modernism. Ann Arbor: University of Michigan Press, 1994. “Professional Ethics.” In A Companion to Applied Ethics, ed. R.G. Frey and Christopher Wellman, 583–96. Oxford: Blackwell, 2003. Lukes, Steven. Liberals and Cannibals. London: Verso, 2003. Malcolmson, Scott L. “The Varieties of Cosmopolitan Experience.” In Cosmopolitics, ed. Pheng Cheah and Bruce Robbins, 233–45. Minneapolis: University of Minnesota Press, 1998. Margalit, Avishai and Gabriel Motzkin. “The Uniqueness of the Holocaust.” Philosophy and Public Affairs 25 (1996): 65 – 83. Mason, Andrew. “Special Obligations to Compatriots.” Ethics 107 (1997): 427– 47. Maswood, S. Javed. The South in International Economic Regimes: Whose Globalization? New York: Palgrave Macmillan, 2006. Mattoo, Aaditya and Arvind Subramian. “The WTO and the Poorest Countries: The Stark Reality,” IMF Working Paper 04/81, 2004. May, Larry. Crimes Against Humanity: A Normative Account. Cambridge: Cambridge University Press, 2005. McConnell, Michael. “Don’t Neglect the Little Platoons.” In For Love of Country, ed. Joshua Cohen, 78–84. Boston: Beacon, 1996. McMahan, Jeff. “The Limits of National Partiality.” In The Morality of Nationalism, ed. Robert McKim and Jeff McMahan, 107–38. New York: Oxford University Press, 1997. Mendus, Susan. Impartiality in Moral and Political Philosophy. Oxford: Oxford University Press, 2002. Meyer, Michael. “Rights between Friends.” Journal of Philosophy 89 (1992): 467– 83. Michalopoulos, Constantine. Developing Countries in the WTO. New York: Palgrave, 2001. Mill, John Stuart. “A Few Words on Non-Intervention.” In Collected Works of John Stuart Mill: Essays on Equality, Law, and Education, vol. XXI, ed. John M. Robson, 109–24. Toronto: University of Toronto Press, 1977. “On Civilization.” In Collected Works of John Stuart Mill: Essays on Politics and Society, vol. XVIII, part I, ed. John M. Robson, 117–48. Toronto: University of Toronto Press, 1977. On Liberty and Other Writings. Oxford: Oxford University Press, 1991. Miller, David. “Bounded Citizenship.” In Cosmopolitan Citizenship, ed. Kimberly Hutchings and Roland Dannreuther, 60–80. Houndmills: Macmillan, 1999. “Cosmopolitanism: A Critique.” Critical Review of International Social and Political Philosophy 5 (2002): 80 – 85.

Bibliography 215

“Distributing Responsibilities.” Journal of Political Philosophy 9 (2001): 453 –71. “Justice and Inequality.” In Inequality, Globalization, and World Politics, ed. Andrew Hurrell and Ngaire Woods, 187–210. Oxford: Oxford University Press, 1999. “National Responsibility and International Justice.” In The Ethics of Assistance, ed. Deen Chatterjee. Cambridge: Cambridge University Press, 2004. On Nationality. Oxford: Clarendon, 1995. “Reasonable Partiality Towards Compatriots.” Ethical Theory and Moral Practice 8 (2005): 63 – 81. “The Limits of Cosmopolitan Justice.” In International Society: Diverse Ethical Perspectives, ed. David R. Mapel and Terry Nardin, 164–82. Princeton: Princeton University Press, 1998. Miller, Richard. “Cosmopolitan Respect and Patriotic Concern.” In Political Philosophy of Cosmopolitanism, ed. Gillian Brock and Harry Brighouse, 127–47. Cambridge: Cambridge University Press, 2005. Moellendorf, Darrel. “The World Trade Organization and Egalitarian Justice.” In Global Institutions and Responsibilities, ed. Christian Barry and Thomas Pogge, 141–58. Oxford: Blackwell, 2005. Moore, Jonathan, ed. Hard Choices: Moral Dilemmas in Humanitarian Intervention. Lanham, MD: Rowman and Littlefield, 1998. Moore, Margaret. The Ethics of Nationalism. Oxford: Oxford University Press, 2001. Muller, Edward N. “Economic Determinants of Democracy.” In Inequality, Democracy, and Economic Development, ed. M.I. Midlarsky, 133–55. Cambridge: Cambridge University Press, 1997. Murphy, Liam. “Institutions and the Demands of Justice.” Philosophy and Public Affairs 27 (1999): 251–91. Nagel, Thomas. “The Problem of Global Justice.” Philosophy and Public Affairs 33 (2005): 113 – 47. Nardin, Terry. “International Political Theory and the Question of Justice.” International Affairs 82 (2006): 449 – 65. Neier, Aryeh. War Crimes: Brutality, Genocide, Terror, and the Struggle for Justice. New York: Times Books, 1998. Newbury, David. “Understanding Genocide.” African Studies Review 41 (1998): 73 –97. Nozick, Robert. Anarchy, State and Utopia. Oxford: Blackwell, 1974. Nussbaum, Martha C. Frontiers of Justice: Disability, Nationality, Species Membership. Cambridge, MA: Harvard University Press, 2006. “Patriotism and Cosmopolitanism.” In For Love of Country, ed. Joshua Cohen, 3–20. Boston: Beacon, 1996. O’Neill, Onora. Bounds of Justice. Cambridge: Cambridge University Press, 2000.

216 Bibliography

Oldenquist, Andrew. “Loyalties.” Journal of Philosophy 89 (1982): 173 –93. Orwin, Clifford. “Distant Compassion: CNN and Borrioboola-Gha.” The National Interest, Spring 1996, 42 –49. Pangle, Thomas. “Socratic Cosmopolitanism: Cicero’s Critique and the Transformation of the Stoic Ideal.” Canadian Journal of Political Science 36 (1998): 235 – 62. Pevnick, Ryan. “Political Coercion and the Scope of Distributive Justice.” Political Studies 56 (2008): 399 – 413. Pinsky, Robert. “Eros Against Esperanto.” In For Love of Country, ed. Joshua Cohen, 85–90. Boston: Beacon, 1996. Pitkin, Hannah. “Obligation and Consent.” In Philosophy, Politics and Society, 4th series, ed. Peter Laslett, W. G. Runciman and Quentin Skinner, 45–85. Oxford: Blackwell, 1972. Pogge, Thomas. “Cosmopolitanism: A Defence.” Critical Review of International Social and Political Philosophy 5 (2002): 86 –91. “Preempting Humanitarian Interventions.” In Humanitarian Intervention: Moral and Philosophical Issues, ed. Aleksandar Jokic, 93–108. Peterborough, ON: Broadview, 2003. Realizing Rawls. Ithaca: Cornell University Press, 1989. World Poverty and Human Rights. Cambridge: Polity, 2002. Postema, Gerald. “Moral Responsibility in Professional Ethics.” New York University Law Review 55 (1980): 63 – 89. Preparatory Commission for the International Criminal Court . Final Draft Text of the Elements of Crimes, Article 7. Adopted June 30, 2000. Prunier, Gerard. “A Well-Planned Genocide.” Times Literary Supplement, October 22, 1999, 30. Przeworski, Adam, Michael E. Alvarez, Jose A. Cheibub and Fernando Limongi, eds. Democracy and Political Development: Political Institutions and Well-Being in the World. Cambridge: Cambridge University Press, 2000. Rachels, James. “Punishment and Desert.” In Ethics in Practice, ed. Hugh LaFollette, 470–79. Oxford: Blackwell, 2000. Rawls, John. A Theory of Justice. Cambridge, MA: Belknap Press, 1971. “The Law of Peoples.” Critical Inquiry 20 (1993): 36 – 68. The Law of Peoples with “The Idea of Public Reason Revisited”. Cambridge, MA: Harvard University Press, 1999. Raz, Joseph. “Rights-Based Moralities.” In Theories of Rights, ed. Jeremy Waldron, 182–200. Oxford: Oxford University Press, 1984. Rees, John C. John Stuart Mill’s On Liberty. Oxford: Clarendon, 1985. Renwick Monroe, Kristen. The Heart of Altruism: Perceptions of a Common Humanity. Princeton: Princeton University Press, 1996.

Bibliography 217

Ripstein, Arthur. “Beyond the Harm Principle.” Philosophy and Public Affairs 34 (2006): 215–45. Risse, Mathias. “Do We Owe the Global Poor Assistance or Rectification?” Ethics and International Affairs 19 (2005): 9 –18. “How Does the Global Order Harm the Poor?” Philosophy and Public Affairs 33 (2005): 349 –76. “What to Say About the State?” Social Theory and Practice 32 (2006): 671–98. Robertson, Geoffrey. Crimes against Humanity: the Struggle for Global Justice. London: Allen Lane, The Penguin Press, 1999. Rorty, Richard. Contingency, Irony and Solidarity. Cambridge: Cambridge University Press, 1989. “Justice as a Larger Loyalty.” In Cosmopolitics, ed. Pheng Cheah and Bruce Robbins, 45–58. Minneapolis: University of Minnesota Press, 1998. Rousseau, Jean-Jacques. Basic Political Writings. Indianapolis: Hackett, 1987. Political Writings, 2 vols. Oxford: Blackwell, 1962. Rummel, R. J. Death by Government. New Brunswick, NJ: Transaction, 1994. Sangiovanni, Andrea. “Global Justice, Reciprocity, and the State.” Philosophy and Public Affairs 35 (2007): 3 –39. Sartre, Jean-Paul. Critique de la raison dialectique. Paris: Gallimard, 1960. Scanlon, Thomas. “The Difficulty of Tolerance.” In Toleration: An Elusive Virtue, ed. David Heyd, 226–39. Princeton: Princeton University Press, 1996. Scheffler, Samuel. Boundaries and Allegiances: Problems of Justice and Responsibility in Liberal Thought. Oxford: Oxford University Press, 2001. Schoeman, Ferdinand. “Rights of Children, Rights of Parents, and the Moral Basis of the Family.” Ethics 91 (1980): 6 –19. Schofield, Malcolm. The Stoic Idea of the City. Cambridge: Cambridge University Press, 1991. Sen, Amartya. “Human Rights and Asian Values.” In Ethics and International Affairs, 2nd edn., ed. Joel H. Rosenthal, 170–93. Washington, DC: Georgetown University Press, 1999. “Humanity and Citizenship.” In For Love of Country, ed. Joshua Cohen, 111–18. Boston: Beacon, 1996. Shue, Henry. Basic Rights, 2nd edn. Princeton: Princeton University Press, 1996. “Limiting Sovereignty.” In Humanitarian Intervention and International Relations, ed. Jennifer Welsh, 11–28. Oxford: Oxford University Press, 2004. “Mediating Duties.” Ethics 98 (1988): 687–704. Simmons, A. John. Justification and Legitimacy: Essays on Rights and Obligations. Cambridge: Cambridge University Press, 2001. Moral Principles and Political Obligations. Princeton: Princeton University Press, 1979.

218 Bibliography

Singer, Peter. One World. New Haven: Yale University Press, 2002. “Outsiders: Our Obligations to Those Outside Our Borders.” In The Ethics of Assistance: Morality and the Distant Needy, ed. Deen Chatterjee, 11–32. Cambridge: Cambridge University Press, 2004. Smith, Michael. “Humanitarian Intervention: An Overview of the Ethical Issues.” Ethics and International Affairs 1 (1998): 63 –79. Tamir, Yael. Liberal Nationalism. Princeton: Princeton University Press, 1993. Tan, Kok-Chor. Justice Without Borders. Cambridge: Cambridge University Press, 2004. Taylor, Charles. “Neutrality in Political Science.” In Philosophy, Politics and Society, 3rd series, ed. Peter Laslett and W.G. Runciman, 25–57. Oxford: Blackwell, 1967. Tesón, Fernando. “The Liberal Case for Humanitarian Intervention.” In Humanitarian Intervention, ed. J. L. Holzgrefe and Robert O. Keohane, 93–129. Cambridge: Cambridge University Press, 2003. Van Parijs, Philippe. “International Distributive Justice.” In Companion to Contemporary Political Theory, ed. Robert Goodin, Philip Pettit and Thomas Pogge, 638–52. Oxford: Blackwell, 2007. Vanderbilt, Tom. Traffic: Why We Drive the Way We Do (and What It Says About Us). New York: Knopf, 2008. Vernon, Richard. Friends, Citizens, Strangers: Essays on Where We Belong. Toronto: University of Toronto Press, 2006. “J. S. Mill and Pornography: Beyond the Harm Principle,” Ethics 106 (1996): 621–32. “Pascalian Ethics: Bergson, Levinas, Derrida.” European Journal of Political Theory, forthcoming. The Career of Toleration: John Locke, Jonas Proast, and After. Montreal: McGill-Queen’s University Press, 1997. “What Is Crime Against Humanity?” Journal of Political Philosophy 10 (2002): 231– 49. Waldron, Jeremy. “Minority Cultures and the Cosmopolitan Alternative.” In The Rights of Minority Cultures, ed. Will Kymlicka, 93–122. Oxford: Oxford University Press, 1995. “Special Ties and Natural Duties.” Philosophy and Public Affairs 22 (1993): 22 –50. “When Justice Replaces Affection: The Need for Rights.” In Liberal Rights: Collected Papers, 1981–1991, 370–91. Cambridge: Cambridge University Press, 1993. Walker, R. B. J. “Citizenship After the Modern Subject.” In Cosmopolitan Citizenship, ed. Kimberly Hutchings and Roland Dannreuther, 171–200. Houndmills: Macmillan,1999. Walzer, Michael. Just and Unjust Wars, 3rd edn. New York: Basic, 2000. “The Argument about Humanitarian Intervention.” Dissent 49 (2002): 29 –37. Thick and Thin. Notre Dame: University of Notre Dame Press, 1994. Wellman, Christopher H. “Friends, Compatriots, and Special Political Obligations.” Political Theory 29 (2001): 217–36.

Bibliography 219

“Relational Facts in Liberal Political Theory: Is There Magic in the Pronoun ‘My’?” Ethics 110 (2000): 537– 62. Wellman, Christopher H. and A. John Simmons. Is There a Duty to Obey the Law? Cambridge: Cambridge University Press, 2005. Williams, Bernard. Moral Luck. Cambridge: Cambridge University Press, 1981. Wippman, David. “Exaggerating the ICC.” In Bringing Power to Justice: The Prospects of the International Criminal Court, ed. Joanna Harrington, Michael Milde, and Richard Vernon, 99–140. Montreal: McGill-Queen’s University Press, 2006. Wolf, Susan. “Morality and Partiality.” Philosophical Perspectives 6 (1992): 243 –59. Wolff, Jonathan and Avner de-Shalit. Disadvantage. Oxford: Oxford University Press, 2007. Woods, Ngaire. “Held to Account: Governance in the World Economy.” In Global Responsibilities: Who Must Deliver on Human Rights?, ed. Andrew Kuper, 251–66. New York: Routledge, 2005.

Index

anarchism, philosophical 39 Appiah, K. 24, 200 Arendt, H. 144–49, 160 Arneson, R. 43 associative duty authority of 20–23 content 17–20 independence 23–28, 29–31 phenomenology 14, 32–33, 38, 89 socialization and 14–17 Augustine, St. 46 autonomy 59, 176–79, 184 collective 183–84 Barry, B. 12, 168 Beitz, C. 1, 130 benefits and obligation 7, 21–22, 40, 105, 193 receiving and accepting 43–45 Bergson, H. 30 Blake, M. 56–57 Bodin, J. 28 Bohman, J. 155–57, 167 Buchanan, A. 120–21, 127–28, 140 Caney, S. 24 Charvet, J. 112 Cicero 3, 93 citizenship requirements 6, 55 and risk 7, 82 of/in the world 2–4, 10, 195, 208 coercion 7, 57–59, 69, 73, 94 Cohen, J. 59 complicity of citizens 51, 73, 132 and consent 83–84 degrees of 85 and reciprocity 204 of states 177 conscription 134–37 contractualism

and consent 66 and contractarianism 67 critique of 65–66, 86 global 104 hypothetical 65–66 Locke and 67–70 mutual advantage view 120, 127–28, 136 Rousseau and 69–70 “social waiver” view 8, 65–67, 71–72, 82–84 cosmopolitanism components 5 kinds of 195 local duty and 20 origins of 2–3 weak and strong 1–2 crime against humanity 9, 52, 115, 144, 194, 197 Arendt’s view 144–49 groups and 150–54 republican view 155–57 and risk 157–61 scale of 162–63 victims of 163–64 Couture, J. 196 De-Shalit, A. 71 democracy 45, 141–42, 184–86, 202 transnational 203–04 Devlin, P. 97 diversity, value of 148–49 Dobson, A. 169 dualism, moral 2, 11, 36, 65, 89–90 Dworkin, R. 97 ends and means 75, 79 Epstein, R. 171 equal opportunity 110, 115 fairness 45–46, 47, 49, 93, 132, 173 family ties 16, 28, 32, 36–37, 44, 98–99

Index 221

Feinberg, J. 172 friends 19, 87 Geras, N. 101 Gilbert, M. 75–79 global economy 53, 59, 110–11, 155, 205–07 Good Samaritans 26, 80, 83, 120 Godwin, W. 29, 37, 39 Goodin, R. 8, 41–42, 55, 79–83 Green, M. 124–27 Grotius, H. 103, 120 Harcourt, B. 172 Hardimon, M. 25 Hardwig, J. 86 harm principle 10, 150 autonomy and 174–76 and contractualism 192 economic policy and 186–88 and institutional capacity 182–84 negative duties 168–70 and responsibility 177–80 and rights 173 skepticism about 171–74 vulnerability and 191 Hart, H. 41, 49, 67, 76, 79–80, 97 Held, D. 202 Himmelfarb, G. 200 Hobbes, T. 103 Horton, J. 7, 25 humanitarian intervention and contractualism 120–21, 127–33 costs of 142 and democracy 141–42 and natural duty 121–27 responsibility for 139–41 and sovereignty 118 Hume, D. 66, 79–80 Huntington, S. 119, 133 immigration 48 iteration principle 9, 167, 194, 206 and Locke 131 Justice institutional view 118 interactionist view 110–13 Locke and 128–32 as loyalty 99–102 non-ideal 9

procedural 58 social 69–70, 71 states and 56–59, 103–04 transactional view 39–40, 82 and vulnerability 106 war and 133–37 see also fairness, harm principle, reciprocity, rights justification and legitimacy 21, 48–49, 72, 103 Kant, I. 59, 61, 62, 99, 103, 111, 122, 146–47, 155, 197, 200 Kissinger, H. 119 Kolodny, N. 197 Kutz, C. 79 Lakatos, I. 50 Linklater, A. 179 Lipset, S. 185 Locke, J. 57, 65, 78–83, 103, 115, 116, 128–32, 140, 155, 165 loyalty 27, 99–102 Luban, D. 52, 149, 158, 165 Lukes, S. 13 majorities and minorities 55–56, 62, 69, 72, 109 Malcolmson, S. 200 Margalit, A. 159 Mason, A. 8, 93 Maswood, S. 188 Mattoo, A. 188 May, L. 149–54, 164 McConnell, M. 14 Meyer, M. 87 Michalopoulos, C. 188 Mill, J. S. 54, 56, 76, 90, 108, 116, 141, 170–74, 175, 178, 183–84, 188 Miller, D. 13, 73, 78–83, 134, 196, 203 Miller, R. 196 Montaigne, M. de 35 Motzkin, G. 159 Murphy, L. 117 Nagel, T. 8, 52 nations and identity 73–75 and partialism 16 nationalism, explanatory 74

222 Index

natural duty 9, 50, 60, 117–18, 121–26, 127 types of 121, 140 negative duties 122–23, 168–69 natural law 69, 103, 129 Nielsen, K. 196 “North Korea” problem 112, 206 Nozick, R. 42 Nuremberg Trials 52, 143 Nussbaum, M. 12–13, 14, 20, 23, 103, 122, 132, 199 obligation imperfect 135, 139–40 and promises 25–28, 45 and public goods 43–44 and roles 27–29, 74 see also associative duty, contractualism, natural duty, reciprocity Oldenquist, A. 7, 33, 99 O’Neill, O. 111, 167, 206 paternalism 44 Pevnick, R. 45–46 Pinsky, R. 200 Plato 40, 46 Pogge, T. 74, 168, 177 Postema, G. 33 professional ethics 35–36 proximity, moral effects of 3, 17, 34, 60, 64, 109, 123 Prunier, G. 159 Przeworski, A. 185 Rachels, J. 22 Rawls, J. 47, 70, 103–04, 117, 124, 129, 133, 175, 187 realism 121, 130 “responsibility to protect” 119 reciprocity 40, 51, 105, 129, 190 between issues 204 republicanism 94, 155 rights 19, 61, 68, 72 antecedent and subsequent 61, 72, 88 and capabilities 71 confl icts of 73 and harm 173 human 99, 102, 155, 163 institutional 52, 113 and risks 7–8, 60, 64, 154, 167, 201–06 social contract and 86–88, 177, 193

Risse, M. 181 Rome Statute 161, 164 Rorty, R. 16, 33, 95, 100–02, 209 Rousseau, J.-J. 46, 54, 61, 69, 92, 103, 158, 160–61 Rummel, R. 158 Rushdie, S. 202 Sabel, C. 59 Sangiovanni, A. 59 Scanlon, T. 56 Scheffler, S. 7, 8, 13, 20–21, 42, 93, 131, 137 Schmitt, C. 73 Schoeman, F. 98 Schofield, M. 4 Simmons, A. J. 8, 39, 60, 62, 63, 95 Singer, P. 47, 48, 95, 188 Sen, A. 195, 200 sovereignty 5, 48, 150, 179 special ties 13, 36, 79, 138 and vulnerability 56, 81 “spoiled brat” problem 112, 156, 206 states criminal 107, 109, 145 demands of 109 failed 109, 145 features of 107–25, 154, 159–61, 205–06 self-determination of 56, 175 Stoics 2, 197 Subramanian, A. 188 subsidiarity 8, 89–91, 98, 165–66 Tamir, Y. 25 Tan, K.-C. 45–46 Taylor, C. 26 Tesón, F. 134–33 thick and thin morality 95–98 UNHCR 18 universalism, moral 13, 21, 93, 103, 130 vulnerability 52, 56, 113, 133, 152 and institutions 31, 53, 80, 89, 106 Waldron, J. 60, 117, 122, 202 Walzer, M. 95–99, 100, 134, 176 Wellman, C. 8, 49, 93 Williams, B. 28 Wolff, J. 71

Cosmopolitan Regard: Political Membership and Global Justice (Contemporary Political Theory)

Justice Beyond Borders - A Global Political Theory

Global Civil Society? (Contemporary Political Theory)

Contemporary Empirical Political Theory

Political Theory and Feminist Social Criticism (Contemporary Political Theory)

Global Justice: A Cosmopolitan Account

Justice, Gender, and the Politics of Multiculturalism (Contemporary Political Theory)

Democracy Defended (Contemporary Political Theory)

National Responsibility and Global Justice (Oxford Political Theory)

Contemporary Political Theory: A Reader

Rejecting Rights (Contemporary Political Theory)

Contemporary Political Theory (Philosophical Topics)

Rejecting Rights (Contemporary Political Theory)

Liberty and Justice: Essays in Political Theory

Political Theory of Global Justice: A Cosmopolitan Case for the World State

Political Theory of Global Citizenship (Routledge Innovations in Political Theory)

Violence and Democracy (Contemporary Political Theory)

Violence and Democracy (Contemporary Political Theory)

Violence and Democracy (Contemporary Political Theory)

Political Theory and Global Climate Change

Criminal Justice and Political Cultures

Enquiry Concerning Political Justice

Justice: Alternative Political Perspectives

Global Civil Society? (Contemporary Political Theory) (Only 23 Pages)

Multiculturalism and Political Theory

Multiculturalism and political theory

Republicanism and Political Theory

Justice As Impartiality (Oxford Political Theory)

Nationhood and Political Theory

State and Political Theory

Multiculturalism and political theory

Cosmopolitan Regard: Political Membership and Global Justice (Contemporary Political Theory)

Justice Beyond Borders - A Global Political Theory

Global Civil Society? (Contemporary Political Theory)

Contemporary Empirical Political Theory

Political Theory and Feminist Social Criticism (Contemporary Political Theory)

Global Justice: A Cosmopolitan Account

Justice, Gender, and the Politics of Multiculturalism (Contemporary Political Theory)

Democracy Defended (Contemporary Political Theory)

National Responsibility and Global Justice (Oxford Political Theory)

Contemporary Political Theory: A Reader

Rejecting Rights (Contemporary Political Theory)

Contemporary Political Theory (Philosophical Topics)

Rejecting Rights (Contemporary Political Theory)

Liberty and Justice: Essays in Political Theory

Political Theory of Global Justice: A Cosmopolitan Case for the World State

Political Theory of Global Citizenship (Routledge Innovations in Political Theory)

Violence and Democracy (Contemporary Political Theory)

Violence and Democracy (Contemporary Political Theory)

Violence and Democracy (Contemporary Political Theory)

Political Theory and Global Climate Change

Criminal Justice and Political Cultures

Enquiry Concerning Political Justice

Justice: Alternative Political Perspectives

Global Civil Society? (Contemporary Political Theory) (Only 23 Pages)

Multiculturalism and Political Theory

Multiculturalism and political theory

Republicanism and Political Theory

Justice As Impartiality (Oxford Political Theory)

Nationhood and Political Theory

State and Political Theory

Multiculturalism and political theory

Recommend Documents