Toleration in Political Conflict 9781107468740, 9781107040328

Citation preview

Toleration in Political Conflict

Political disputes over toleration are endemic, while toleration as a political value seems opposed to those of civic equality, neutrality and sometimes democracy. Toleration in Political Conflict sets out to understand toleration as both politically awkward and indispensable. The book exposes the incoherence of Rawlsian reasonable pluralist justifications of toleration, and shows that toleration cannot be fully reconciled with liberal political values. While raison d’état concerns very often overshadow debates over toleration, these debates – for example about terrorism – need not be framed as a conflict between toleration and security. Framing them in this way tends to obscure objectionable behaviour by tolerators themselves, and their reliance on asymmetric power. Glen Newey concludes by sketching a picture of politics as dependent on free speech which, he argues, is entailed by the demands of free association. That in turn suggests that questions of toleration are inescapable within the conditions of politics itself. g l e n n e w e y is Professor of Political Theory at the Université Libre de Bruxelles. He has taught and written about toleration for over twenty-five years and is the author of Hobbes and ‘Leviathan’ (2008), After Politics (2001) and Virtue, Reason and Toleration: the place of toleration in ethical and political philosophy (1999). He has also written many articles on the theory and practice of toleration and is a frequent reviewer and commentator for publications including the London Review of Books, Independent, New Statesman and Times Literary Supplement.

Toleration in Political Conflict Glen Newey

University Printing House, Cambridge CB2 8BS, United Kingdom Published in the United States of America by Cambridge University Press, New York Cambridge University Press is part of the University of Cambridge. It furthers the University's mission by disseminating knowledge in the pursuit of education, learning and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781107040328 © Glen Newey 2013 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 2013 Printed in the United Kingdom by CPI Group Ltd, Croydon CR0 4YY A catalogue record for this publication is available from the British Library Library of Congress Cataloguing in Publication data Newey, Glen. Toleration in political conflict / Glen Newey. pages cm Includes bibliographical references. ISBN 978-1-107-04032-8 (Hardback) 1. Toleration. 2. Toleration–Political aspects. 3. Democracy–Philosophy. I. Title. HM1271.N52 2013 303.6–dc23 2013005733 ISBN 978-1-107-04032-8 Hardback 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.

For John Horton

Contents

Acknowledgements

page viii

1

Introduction: toleration in trouble

2

Tolerating politics

21

3

Democratic toleration?

40

4

Toleration as sedition

60

5

The trouble with respect

81

6

How not to tolerate religion

102

7

Liberty, toleration, security

123

8

Toleration and power

142

9

Tolerating ourselves, tolerating terror

163

Toleration, free speech and the right to lie

183

Epilogue

207

Bibliography Index

211 222

10

1

vii

Acknowledgements

I take this opportunity to thank the usual suspects for help and support in preparing this book for publication. I thank the hosts of forums in which versions of several chapters were presented, notably my former colleagues Monica Mookherjee and Sorin Baiasu, of Keele University; Shaun Young, University of York, Ontario; Tim Heysse, University of Leuven; Barbara Segaert, of UCSIA Antwerp; Dario Castiglione, University of Exeter, and Catriona McKinnon, University of Reading. I also thank Peter Jones, lately of the University of Newcastle, for debate and contestation which have spanned a long period. He always meets objections (which are reiterated below) with implacable courtesy and fair-mindedness, and his own writing offers a model of clarity and intellectual integrity. I also take the opportunity of thanking Sue Mendus, now at last retired from the University of York, for having first sparked my interest – that must be the word – in the concept of toleration. I went to York in the 1980s to study toleration mainly because of the money – the Morrell Trust offered what seemed princely sums to students to study on its MA course in toleration, a programme that happily continues thanks to the generosity of the trustees. As with Pascal’s view of religious worship, what was at first undertaken for purely egoistic reasons has over time become ossified into the gainless habit we know as virtue. As Sue and others who taught me at the time have often reminded me, my presence in seminar groups and in tutorials itself offered them a constant practical exercise in toleration. The oldest hath borne most. Thanks also to those who discussed the ideas in this book in these and other settings, including (in no special order) Peter Balint, Matt Matravers, Leif Wenar, Graham Long, Katrin Flikschuh, Paul Kelly, Melissa Williams, Stephen Macedo, George Kateb, Iain HampsherMonk, Rainer Forst, Ingrid Creppell, Linda Holt, Veit Bader, Andrea Baumeister, Derek Edyvane, Corey Brettschneider, Paul Bou-Habib, Dudley Knowles, Jonathan Riley, Anthony Duff, Victor Tadros, Glyn Morgan, Gerald Gaus, Ronnie Beiner, Dan Weinstock, Bluey Hick, Scott Davis, Anna Elisabetta Galeotti, Jon Parkin, David Owen, viii

Acknowledgements

ix

Tim Stanton, John Charvet, Petter Korkman, Véronique Muñoz-Dardé, Jo Wolff, Jim Gibson, Arto Laitinen, David Heyd, Charlie Husband, and the late Brian Barry. I have also benefited from questions and interventions in seminars and conferences from participants whose identities were unknown to me. My thanks to them, and to anyone else whom I may have omitted from the list of names above. Special thanks also to Carrie Parkinson, of Cambridge University Press, for help in preparing the typescript, and to John Haslam and Hilary Gaskin, also of Cambridge University Press. Many thanks also to Pat Harper, who discharges the often thankless duties of copy-editing with exemplary attention to detail. My major debt is recorded in the dedication. Versions of some chapters (chapters 3, 4, 6 and 10) have appeared in print already, though all have been modified for the present volume. Chapter 3: ‘Democratic toleration?’ was published as ‘Is Democratic Toleration a Rubber Duck?’ in Dario Castiglione and Catriona McKinnon (eds.), Toleration, Neutrality and Democracy (Dordrecht: Kluwer 2003), repr. from Res Publica 7:3 (2001), 315–36 and is used with kind permission from Springer Science + Business Media B.V.; Chapter 4: ‘Toleration as sedition’, was published in Critical Review of International Social and Political Philosophy 14:3 (2011), 363–84 and is used with kind permission from the Taylor & Francis Group, http://www. tandfonline.com; Chapter 6: ‘How not to tolerate religion’, was published in Monica Mookherjee (ed.), Toleration, Recognition and Religious Diversity (Dordrecht: Springer 2011) and is used with kind permission from Springer Science + Business Media B.V.; Chapter 10: ‘Toleration, free speech and the right to lie’, was published as ‘Free Speech and Bad Speech: Nike v. Kasky and the right to lie’, Bijdragen: International Journal in Philosophy and Theology 71:4 (2011), 407–25 and is used with kind permission from Bijdragen.

1

Introduction: toleration in trouble

When the King of Prussia entered Silesia in 1740, a small Protestant borough, jealous of a neighbouring Catholic village, came humbly to beg permission to put all the inhabitants of that village to the sword. The King replied: ‘If that village came to ask me for leave to cut your throats, would you think me right to grant it to them?’ To which they replied: ‘O gracious sovereign, the case is very different: we are the true Church.’1

Defenders and critics of the United States agreed that the 11 September 2001 attacks targeted not just the nation’s people and buildings, but its ideals and values – the ‘American way of life’. Some interpreted the attacks as a traditionalist backlash against US or western liberal pluralism, the assertion of the Way against the West’s pluralistic decadence, while others saw them as an essentially modern or postmodern phenomenon.2 At any rate, the 9/11 outrages offered the most graphic illustration of non-assimilation to western ideals, and particularly the rejection of liberalism, since the end of the Cold War. The 11 September attacks, and the subsequent bombings in Madrid, Bali, London, Istanbul and elsewhere, are only the most visible instances of the religious and cultural divisions that mark the post-Cold War period. Pope Benedict XVI’s Regensburg lecture about Islam, the dispute over the Anglican Primate’s remarks in 2008 predicting the absorption of Sharia into English law,3 religious fundamentalism in education and in relation to security policy, artistic censorship, such as the withdrawal of the play Behzti by Birmingham Repertory, the Jyllands-Posten cartoons and the BBC television screening of Jerry Springer: the Opera, and British Airways’ ban on the wearing of crucifixes by its employees also exemplify political disputes over toleration. Similarly, the murder of the Dutch film-maker Theo van Gogh in 2004 and the forcing into hiding of the 1 2 3

Voltaire, Treatise on Tolerance and Other Writings, ed. Simon Harvey; tr. Brian Masters (Cambridge University Press 2000), 132. John Gray, Al Qaeda and What It Means to Be Modern (London: Faber 2004). See Chapter 6.

1

2

Introduction: toleration in trouble

Somali Dutch activist Ayaan Hirsi Ali were held up as exemplars of religiously motivated intolerance. The stance of some religions towards homosexuality has also provoked conflict, both as disputes within the Anglican Communion over ordaining homosexual priests, and the proscription of Roman Catholic adoption agencies, which refused to make infants available for adoption by same-sex couples. Controversy attended the proposed Koran-burning by a US fundamentalist pastor, the proposal to build a ‘Ground Zero Mosque’ (in fact an Islamic community centre) near the site of the 11 September attacks in New York, and the French Senate’s proscription in 2010 of wearing the burqa in public. So toleration remains politically contentious. A notable feature of these disputes is that attributions of intolerance are not only made by one side rather than the other: the language of toleration readily lends itself to appropriation by either party to the disputes.4 For example, the controversy over Catholic adoption agencies invited accusations of intolerance on each side: by same-sex couples denied access to adoption by the agencies, and by the agencies who regarded this denial as a matter of religious conviction. The threatened Koran-burners could be seen as intolerant, but could also lay claim to toleration under the First Amendment. Similarly, those who disseminated an incendiary film about Islamic intolerance on the internet in September 2012 could be seen – while taking advantage of the tolerant culture of free expression – as themselves sponsoring intolerance against Muslims. Was the proposed Islamic centre near Ground Zero intolerant of the memory of victims of the 2001 killings, or were activists who sought to prevent its construction acting intolerantly, or perhaps both? Similarly, the French burqa ban might be regarded as intolerant of differences in religious dress, or as itself an act of toleration, in that it bore down on female repression within Islam. Appeals to toleration often seem not to resolve political conflicts, but to perpetuate them.

Toleration in political conflict This book does not try to set out another theory of toleration. Writers who do offer such a theory generally aim either to analyse toleration as a concept,5 or to give an account of justified toleration that can explain what should and should not be tolerated. Sometimes, again, the aim is 4 5

Glen Newey, Virtue, Reason and Toleration: the place of toleration in ethical and political philosophy (Edinburgh University Press 1999), ch. 5. Newey, Virtue, Reason and Toleration ch. 1 offered such an analysis.

Toleration in political conflict

3

to provide a rhetorical anatomy of toleration, for example in the form of an immanent critique. This is a little closer to the goal of this book, in that it takes seriously the idea that the language of toleration in liberal democracies is vulnerable to political manipulation and that theories that ignore this fact risk becoming, in the pejorative sense, ideological. But the ulterior aim is not to debunk toleration – to expose it as a bogus ideal. It is rather to reaffirm a feature of toleration that is in danger of becoming obscured both in the heat of political argument and in philosophical theory’s quest for conceptual clarity.6 That feature is the awkwardness of toleration or tolerance7 as a political ideal, evident in the examples already given. The awkwardness itself helps to explain why toleration comes under political and theoretical pressure. In the face of this pressure it becomes hard to reassert the value of toleration against the impulse in political theory and practice towards normative simplification. The attractions of simplicity are plain. But it can lead to distortion, and if what is distorted has normative content, a misrepresentation and indeed loss of value may result. I explore this process in the chapters that follow. Toleration grinds and jolts against other political values such as equality, democracy, security and justice. It then becomes tempting to discard it, downgrade it, or transform it into something thought to be more normatively manageable – such as one of these other political values. Its vulnerability comes out both in theories of toleration and in the uses to which toleration is put in political advocacy. To say that toleration has been distorted assumes, of course, that there is something that toleration is, which is being misrepresented. My aim is to keep this content largely implicit. That said, the underlying thought is that there is political space for the idea that, though a practice may be distasteful, regrettable, even deplorable, there is reason not to stop or censure that practice. This is not, it should be clear, a lexicographical point. At this high level of generality, what matters is whether a concept with this shape has a political point, or role. That in turn leads one to ask what it is for something to have a political point. The idea is that politics is a public arena which is incompletely deliberated – which poses a question about collective action. The fact that some practice is deplorable, distasteful, etc., raises a political question about the response appropriate

6

7

For a defence of this role for toleration, see John Horton, ‘Why the Traditional Conception of Toleration Still Matters’, Critical Review of International Social and Political Philosophy 14:3 (2011), 289–305. I shall not distinguish these terms.

4

Introduction: toleration in trouble

to it, and the very openness of that question leaves room for a concept with the broad shape just mentioned. This way of seeing things gives priority in political action to deliberative accessibility. It is often a real question – a political one – whether a certain course of action is, from here, accessible to political deliberation. Any political concept is liable to undergo deformation or, less tendentiously, ‘decontestation’8 as part of the deliberative process. Political actors adopt a specific interpretation of the concept in arguing their case, such as a libertarian conception that stresses freedom to appropriate, rather than freedom from others’ acquiring exclusive property rights. The very idea that political actors decontest key concepts counsels against naively reading off the concept from its political use. But that does not mean that the concepts should or even can be abstracted from their role in political argument. To do so risks offering merely another form of decontestation, not obviously more authoritative than others. Insofar as this book presents a single argument, it is that a concept with the general shape already outlined is likely to persist – to have an enduring political use – just because of ineradicable remainders in political thought and action. That does not mean that policies of toleration, or policies answering to a concept shaped like the one above, are ineradicable. It means that the concept is liable to have a deliberative use to the extent that there is politics. The instability lies in the tension between the openness of deliberation and the drive in politics to reach a decision; this is a constitutive tension. The impulse to closure in the face of deliberative remainders pulls against the open-endedness of political action. Faced with a state of affairs or practice that certain interests object to and want suppressed, the initially open question of what to do can be closed by complying with or opposing their demands. In this sense, as long as there is the option of using political means to deal with the practice, a political question exists. The resolution that comes with a decision is not final. That could only come from an end not only to the objected-to situation, but to the possibility of its arising again. The fantasy of a final determination is in this sense one of an end to politics – an end which lies in eliminating deliberative remainders. This suggests that attempts to formulate ideal conditions for toleration, or at least for failing to suppress or censure what is found objectionable, may commit a category mistake. Such theories offer a 8

Michael Freeden’s term. See his Ideology: a very short introduction (Oxford University Press 2003), 54–60; also his Ideologies and Political Theory (Oxford University Press 1998). For an acute analysis of the role of disagreement in political argument, see Andrew Mason, Explaining Political Disagreement (Cambridge University Press 1993).

Toleration in political conflict

5

static analysis of well-ordered societies in which toleration prevails, and so does not pose itself as a political problem. The theories explain why, if a problem does emerge, one side or other has made a mistake, and perhaps suggest what it or the political authority should do to rectify it. Theorists who highlight the political significance of normative controversy respond to problems of toleration, to which such controversy is central, by making further controversial normative claims. One can insist on a normative standard which political life fails to meet. But it is possible to take a different view, by noting the persistence of conflict in political life and the very general sources from which it springs – the openness of deliberation to which I have already referred, and the fact of normative complexity. So this book is less concerned with prescribing when to tolerate, than with understanding how this complexity works itself through in political practice. This is not only because concepts, including that of toleration, have a history. That history includes its deployment in political argument as well as in theoretical writing – not that those, such as Przypkowski, Milton, Walwyn, van Limborch, Bayle, Spinoza, Locke, Toland and others, whose writings on toleration have come down to us from the early modern period, saw these as distinct activities. It was not their aim to devise a conceptual analysis of toleration and then to demonstrate how this analysis had political implications to which those who were reasonable were committed. It was rather to argue on political and religious grounds that a certain regime was desirable or justifiable – a form of advocacy. That already makes the relation between argument and rhetoric less discontinuous than more recent styles of analysis often assume.9 It also means that arguments about toleration show less or more than their proponents aim for.10 Early modern advocates of toleration find themselves espousing arguments at odds with other commitments, as in the tension between Hobbes’s support of an autocratic basis for political power and his own preference, expressed in the English Leviathan’s ‘Review and Conclusion’, for religious pluralism.11 None of this means that philosophical analysis is of no use in understanding toleration. Indeed, one purpose analysis serves is precisely to 9 10 11

See Quentin Skinner, Visions of Politics, Vol. I: Regarding Method (Cambridge University Press 2002), esp. essays 4, 5 and 10. I take Locke as a case in point in Chapter 6. See Hobbes, Leviathan, ed. Richard Tuck (Cambridge University Press 1996), e.g. 486. Hobbes cut these remarks from the Latin edition of Leviathan published in 1668. I examine the tensions in Hobbes’s views further in Hobbes and ‘Leviathan’ (London: Routledge 2008), ch. 9.

6

Introduction: toleration in trouble

highlight tensions between political positions and the justifications offered for them. Conditions of public justification constrain what can be said, and the courses of action political language can be used to justify. Perhaps political actors can be brought within some entirely general scheme of justification. But it will not be adequately supported by norms that abstract from the actors’ circumstances and then pronounce what they should accept, or would accept in some extravagantly contrary-tofact situation. Often there is no reason to think that the counterfactuals have determinate truth-conditions, or that even when they do, their truth is dispositive for real-world actors.12

Toleration To tolerate something is to accept it, despite disagreeing with or disapproving of it. Someone who has the power to take action against it, whether by preventing it from happening at all, or by censuring or otherwise penalising those who have done it already, decides not to use this power. This decision need not involve doing nothing, though often it will. In deciding to accept it rather than preventing or penalising it, the tolerator need not – in fact, probably will not, at least in the short run – regard it as a good thing that the action goes ahead. In tolerating an action, one need not regard it, or the people who do it, as being as good as some other set of actions or people. These statements may well have the air of truism. If so, this is a good thing to the extent that truisms are at least true. I have set out the claims blankly, and without supporting argument, although each of them has been disputed by theorists in the extensive literature on the theory of toleration. Some of them are discussed in more detail in the following chapters, notably that about power and toleration,13 about whether toleration entails disagreement,14 and about whether the actions tolerated or the doers of them need be seen as good in themselves.15 I shall deal here with a further claim, that the tolerator need not feel disapproval or distaste towards whatever she tolerates. Certainly, the term ‘tolerant’ can be applied to people who are not disposed to feel disapproval of others.16 For instance, the Netherlands has often been described as a 12 13 16

See Ronald Dworkin, ‘The Original Position’, in Norman Daniels (ed.), Reading Rawls: critical studies of ‘A Theory of Justice’ (Oxford: Blackwell 1975), 20f. 14 15 See Chapter 8. See chapters 4 and 5. See Chapter 7. For an argument that a negative attitude is not necessary for toleration, see Peter Balint, ‘Toleration as a Liberal Political Practice’ (unpublished paper). Also Balint, ‘Not Yet Making Sense of Political Toleration’, Res Publica 18:3 (2012), 259–64.

Toleration

7

tolerant society,17 which is usually taken to mean not that the Dutch habitually feel strong disapproval which they strenuously suppress, but that they readily accept the differences of others. This might be called an intransitive understanding of toleration, since there is nothing and nobody at whom the toleration is directed: unlike emotions, indifference is not directed towards an object. By contrast, the disapproval that marks transitive toleration must have an object – persons, actions, or states of affairs. This book mainly deals with transitive toleration. There is nothing wrong with the intransitive usage. But it leaves a puzzle: why describe people as tolerant if they feel no disapproval, but are simply indifferent to or actively approve of others? Can you be said to tolerate my clavicle? An obvious response is that the intransitive sense is just that, so it has no bearing when a certain object is in view. In this sense it signifies an absence, namely the lack of disapproval towards the idiosyncrasies of others. As by definition people are tolerant in the intransitive sense when there is no object towards which attitudes are directed, ‘tolerant’ in this sense signifies non-disapproval of what others might disapprove of, elsewhere or at other times. But where even the possibility of disapproval does not arise, nor does toleration, even in the intransitive sense. In this the concept of toleration resembles that of politics:18 not everything is an object of toleration, just as not everything is political, but a concern becomes so when its being disapproved of, or an object of joint action, comes under consideration. However, nothing can be tolerated if the very concept of toleration is incoherent, as numerous writers have argued. The paradoxical nature of toleration has been argued for on a number of grounds.19 On some views, there is a contradiction in valuations that justify both the disapproval of some practice, and restraining oneself from acting to stop or censure it.20 Suppose one disapproves of the wearing of the burqa in public by Muslim women. The toleration question that then arises is whether or not to prevent or censure it. But if a person calls on some principle that justifies not doing so, it may seem that contradictory 17

18 19 20

See Arend Lijphart’s classic study, The Politics of Accommodation: pluralism and democracy in the Netherlands (Berkeley: University of California Press 1968). The association of the Netherlands with toleration goes back a long way. See e.g. William Baron, The Dutch Way of Toleration, Most Proper for Our English Dissenters Written at the Request of a Friend (1699) (Ann Arbor: University of Michigan Digital Library, Eebo editions 2010). See also chapters 2 and 10. See John Horton, ‘Three (Apparent) Paradoxes of Toleration’, Synthesis Philosophica 7:1 (1994), 7–20. David D. Raphael, ‘Toleration, Choice and Liberty’, Government and Opposition 6:2 (1971), 229–34. See also Preston King, Toleration, 2nd edn (London: Frank Cass 1998), 29.

8

Introduction: toleration in trouble

practical judgements are in play: one that says it is bad to permit the wearing of the burqa (say because it exemplifies patriarchal repression) and another that says it is bad to prohibit the wearing of the burqa (perhaps because this interferes with freedom of religious expression). It may be thought that these are not just conflicting but contradictory normative judgements. But, at the most general level, the tolerator’s position merely exemplifies the familiar possibility that reasons for action may point in opposite directions. Some argue that there is a ‘paradox of toleration’ on the following grounds: ‘a tolerant society is always at risk of tolerating those who are intolerant, and allowing movements to grow which foster intolerance’.21 For example, in responding to the resurgence of religious fundamentalism, liberal societies face what is sometimes thought to be a constitutive dilemma. Either these societies ‘stand firm’ against the ‘threat’ of terrorism by retrenching on basic liberties, or else they leave themselves ‘open’ to further attacks by creating an environment in which the planning and execution of terrorist atrocities can occur. The fact that such a situation could occur shows that there is no paradox – which I take to be a collection of statements each of which seems to be true, but which cannot be true together.22 At most it would show that there is something pragmatically self-defeating or self-undermining about toleration.23 Whether that is really true is a matter for investigation. On a more optimistic view, toleration is more like a homoeostatic mechanism such as a servo, which is self-limiting but not self-defeating. If so, toleration is not self-defeating even in a negative way. After all, nobody thinks that toleration should or can be limitless. Some things merit intolerance, including by those who in other respects act tolerantly. Liberals think that people’s differences in religion should be tolerated, but not human sacrifice. So religious practices like the Aztecs’ should not be tolerated. A perhaps superficially similar idea is that there is a paradox in ‘tolerating the intolerant’. On this view, the question whether to tolerate the 21

22 23

Anthony C. Grayling, Liberty in the Age of Terror: a defence of civil liberties and Enlightenment values (London: Bloomsbury 2009), 76. Cf. Karl Popper, The Open Society and Its Enemies, 2 vols. (London: Routledge 1946), and ‘Toleration and Intellectual Responsibility’, in David Edwards and Susan Mendus (eds.), On Toleration (Oxford: Clarendon Press 1987). On a stronger reading, a paradox consists in a pair of statements, the truth of each of which entails the falsity of the other. Worries on this score led John Rawls to move from the ‘comprehensive’ liberalism of A Theory of Justice (Oxford University Press 1971) to the ‘political’ liberalism of his later Political Liberalism (New York: Columbia University Press 1996). Rawls came to doubt that the institutions of liberty need engender allegiance among beneficiaries of liberty and thereby become self-reinforcing.

Toleration

9

intolerant poses a dilemma for liberals:24 either the question is answered ‘Yes’, and so tolerators promote or at least permit intolerance; or it is answered ‘No’, and toleration seems to fall prey to pragmatic contradiction. Others think that intolerance itself constitutes a due limit to what can be tolerated.25 Grayling resolves the problem in this way.26 But in fact there is no contradiction, even on the surface, in thinking that one should tolerate people who are themselves intolerant. This is a political problem rather than a paradox. There are reasons for disapproving of actions, and these reasons may include the fact that the actions are intolerant. If one took a purely consequentialist view, one might think that it would be odd to increase toleration by allowing a higher incidence of intolerance, though even then one could think that it was a matter of weighing the one against the other – again, a familiar and unparadoxical exercise. But if the question is how I should act, then consequentialist thoughts may not be uppermost: the question is what a tolerant person would do in my situation, and whether I should do that. Then again, it can be argued that the ethical or other dispositions constitutive of toleration as a virtue are in radical tension with one another.27 The ostensible paradox of the ‘tolerant racist’ illustrates the point. Suppose a white man, from previously having held no views about members of other races, acquires strong prejudices against black people. He continues to act in the same way towards them as before. Previously, he was neither tolerant nor intolerant towards black people; now, simply by dint of acquiring racial prejudices and behaving in the same way as before, he seems to have become tolerant towards them. This, if not a paradox, seems strongly counter-intuitive. Some think that the problem should be solved by identifying certain attitudes as intolerant in themselves, so that someone who has, say, racial prejudices is necessarily intolerant.28 However, the idea that there is a fixed list of attitudes that are necessarily intolerant raises the spectre of an illiberal liberalism: who decides on the list? If the response is that this is decided subjectively, the problem resurfaces, and leaves no room to say that someone whose attitudes are very different from one’s own 24 25

26 27 28

Grayling appears to conflate this problem with the idea discussed earlier, that toleration is empirically self-subverting. Cf. Jeffrey Murphy, Kant: The Philosophy of Right (Basingstoke: Macmillan 1970), 95: ‘one makes a mockery of the value of toleration if one includes active intolerance among those things we ought to tolerate’. See Grayling, Liberty in the Age of Terror, 77f. Popper, The Open Society and Its Enemies, Vol. I, 265. The problem was first identified by Horton, ‘Three (Apparent) Paradoxes of Toleration’. I discuss it in Virtue, Reason and Toleration, ch. 3. Horton argues for this position. For a contrasting view, see King, Toleration, xxi–xxii, 67f.

10

Introduction: toleration in trouble

may, nonetheless, act tolerantly. There is also the problem that if toleration is already moralised, and hence a pro tanto good, one can no longer say that toleration can be misplaced, either because the disapproval itself, or the attitudes that counsel against interference despite the disapproval, are inadequately grounded. It is more persuasive to say that the ‘tolerant racist’ need not instantiate a paradox. As often with philosophical examples, it makes a difference how the story gets filled in. Some attitudes seem intolerant, but much depends on how the disapproval is acquired and expressed, and what keeps it in check. The question is both how the man gained his beliefs and, more important, how he manages them. It is not just a question of what the man believes, but how he expresses the belief, and how far his holding it allows for readjustment in the light of further experience. Does he simply bite his lip out of fear, perhaps because he knows that expressing racial prejudice is legally prohibited or socially unacceptable? Or perhaps, despite himself, he subscribes to a principle of equal treatment or equal respect for persons. This may lead him to doubt the basis for the prejudices he feels, though he finds he cannot fully shake them off. In this case, there seems no compelling reason to deny that the man has gone from being non-tolerant to being tolerant, whatever is said about his belief that black people are inferior to others. One possibility is that the man tolerates, but does not display the virtue of toleration. Toleration involves a struggle between one’s own dispositions, and an ability to keep due distance from them. Like other virtues, it involves not just doing certain things but doing them, as Aristotle says, in a certain way.29 More specific worries can attend to specific justifications of toleration such as that based on autonomy:30 if toleration rests on autonomy, it faces the fact that the social context of toleration – that of Mill’s ‘punishment by opinion’31 – is liable to prejudice the autonomy of those at whom toleration is directed. Scepticism is often thought to ground toleration, on the grounds that one cannot enforce views that may be mistaken; but everyone who believes in toleration believes also that there can be justifiable intolerance, scepticism notwithstanding; so they

29 30

31

Aristotle, Nicomachean Ethics, tr. H. Rackham (Cambridge, MA: Harvard University Press 1934), 1105b8–9. I argue this at greater length in Virtue, Reason and Toleration, ch. 3. For a defence of the autonomy-based justification, see Joseph Raz, ‘Autonomy, Toleration and the Harm Principle’, in Susan Mendus (ed.), Justifying Toleration: conceptual and historical approaches (Oxford University Press 1988). John Stuart Mill, Utilitarianism and On Liberty, ed. Mary Warnock (Oxford: Blackwell 2003).

A political problem

11

also think that scepticism does not invariably defeat toleration. Neutrality and equality pose prima facie problems for toleration.32 It may also be said that the circumstances in which toleration is called for politically are the very ones in which it is no longer a political option, so that in this sense it becomes ‘impossible’.33 This again is not a paradox, but situations of this sort often mark modern political battles over toleration. Not infrequently, each side accuses the other of intolerance, while claiming toleration for itself. For instance, greater toleration of homosexuals is sometimes said to have led to greater intolerance towards those who regard same-sex families or lifestyles as inferior to heterosexual ones, while these attitudes may be said themselves to be intolerant.34 A similar structure marks debates in many western countries over religious symbolism and dress in the public sphere, or such matters as smoking, noise abatement, grammar, Leyland cypresses, public censorship and broadcasting standards, religious marches, public nudity, Sunday closure of shops, the use of the tax system to discourage certain activities and promote others, the imposition on parents of prescribed lists of forenames for their offspring, home schooling, the use of profanity such as putatively obscene or blasphemous language, the teaching of Darwinism on the state school curriculum, dress codes, and the policing of prostitution. In liberal societies, the political circumstances of toleration often arise from demands for intolerance. These rest in turn on the claim that some practices, such as those just listed, are intolerable. To support such a claim, its proponents need to show that the practice menaces the basis of fundamental political goods, such as the polity itself. They point at a ne plus ultra that the practice will allegedly breach. Whether justified or not, the discursive move aims to set a limit to what the polity can encompass.

A political problem As in the Roman Catholic adoption dispute, toleration is liable to be outflanked by protagonists who claim either that what is tolerated is intolerable, or that supposedly tolerant policies are intolerant or are 32 33 34

See Chapter 5. See Bernard Williams, ‘Toleration: an impossible virtue?’ in David Heyd (ed.), Toleration: an elusive virtue (Princeton University Press 1996), 18. ‘[T]he group Focus on the Family . . . has suggested that programs aimed at anti-gay bullying are themselves “intolerant” of Christians, and represent an effort not to protect children, but to promote a “gay agenda”.’ See the Opposing Views website (accessed 22 August 2011): www.opposingviews.com/i/group-complains-anti-bullying-programsintolerant-of-christians.

12

Introduction: toleration in trouble

themselves intolerable. Toleration may be a value in its own right, but it faces endemic problems of political justification. Once it is given specific content it can seem parochial, given that the very problem of toleration frames itself as one between different local justifications. This could be called the replication problem: when questions of justification come up, giving content to toleration risks replicating a normative conflict of the very sort that made toleration a problem to start with. Then it can seem that toleration can fend off challenges from advocates of ‘difference’ only insofar as it sheds content, for example by reframing itself as a relation between different evaluative beliefs. One way to understand the replication problem, and indeed the persistent air of paradox that surrounds toleration in general, is as a back-projection from the conflicts that attend toleration in political practice. The conflicts are inevitable, however, not because toleration is conceptually incoherent or ‘essentially contested’,35 but because this is what makes disputes involving toleration political ones. They arise in a public forum that is imperfectly deliberated, in the sense that public action, ex ante, is only imperfectly guided by reference to ambient values, including those upheld by liberalism. Disputes over toleration arise when some interest-groups seek to use their own power or that of the liberal state to control the action of others of whom they disapprove. So the allegedly paradoxical nature of toleration can be referred not to features inherent to its conceptual structure, but to its embattled role within the politics of liberal democracy. This is not to say simply that the aporias of toleration are simply ‘practical’ rather than ‘theoretical’. Rather they arise from the ‘strains of commitment’36 encountered by efforts to justify practices of toleration using the normative vocabulary available to democratic actors. On the one hand, toleration is usually upheld as a core liberal value. But democratic politics prove inimical to toleration: it gets squeezed between the robust egalitarian idiom of liberal democracies, which sees toleration as second-best to full civic equality, and the pressure on politicians to retrench on personal liberties. On one view, toleration is at best a stalking-horse for wider pluralisation in society, needed only to usher in a regime of equality enshrined in civic rights.37 For example, ‘toleration’ may be rejected as 35

36 37

W. G. Gallie, ‘Essentially Contested Concepts’, Proceedings of the Aristotelian Society 56 (1956), 167–98; I criticise the thesis in ‘Philosophy, Politics, and Contestability’, Journal of Political Ideologies, 6:3 (2001), 245–61. John Rawls, Collected Papers, ed. Samuel Freeman (Cambridge, MA: Harvard University Press 1999). See e.g. Chris Beneke, Beyond Toleration: the religious origins of American pluralism (Oxford University Press 2008).

A political problem

13

a description of liberalised attitudes towards, say, homosexuality precisely because it retains connotations of (unwarranted) disapproval and so falls short of full civic equality. Toleration is thought unsatisfactory because it relegates the tolerated to second-class civic status. At the level of basic political and civic entitlements, the language of democratic politics is, on the whole, robustly egalitarian. In this way, the normative environment of democratic politics proves inimical to toleration. ‘Merely’ to tolerate those who deserve full equality gives toleration a bad name. More generally, toleration either becomes identified with specific policies, and then the attitudes towards it simply mimic those for or against the policies themselves; or it acquires a practically contentless status and is colonised by attitudes now of endorsement, now of rejection. So the judgements underlying specific normative uses of ‘toleration’ lie open to subversion or inversion. For those whose disapproval of a practice extends to support for restricting it, it is a short step to affirming that the grounds for disapproval are so strong that they outweigh the reasons for not acting on this disapproval. Hence the recent rhetoric of ‘zero tolerance’ – that is, ostensibly justified intolerance. On the other hand, those whose approval of the practice extends to opposing its restriction often think that its acceptability is a matter of justice,38 so that any case for disapproving of it – or at least, acting politically on this disapproval – lapses. Its ambiguity explains the frequent appeal to ‘toleration’ as a rhetorical resource in political argument, and why toleration is embattled as a political ideal. The under- and over-tolerant can trump the claims of toleration, respectively for demanding too much, or settling for too little; in this toleration’s structure could be said to resemble that of Aristotelian virtues such as courage, seen as a mean between excess and deficiency.39 At the same time, those who argue for toleration regarding a given policy or attitude can be challenged by arguing that they are in fact intolerant, and their opposite, tolerant: an obvious example is the familiar wrangling over ‘political correctness’. The fact that toleration lends itself to this inversion, coupled with the availability of the trumping manoeuvre, makes it very hard to sustain as a robust and distinct political value. Its very availability to opposing parties is also toleration’s weakness. 38

39

Peter Jones, ‘Making Sense of Political Toleration’, British Journal of Political Science 37:3 (2007), 383–402; see also my ‘Political Toleration: a reply to Jones’, British Journal of Political Science 41:1 (2011), 223–7; Jones, ‘Political Toleration: a reply to Newey’, British Journal of Political Science 42:2 (2011), 444–7, and Balint, ‘Not Yet Making Sense of Political Toleration’. Newey, Virtue, Reason and Toleration, ch. 3.

14

Introduction: toleration in trouble

These competing democratic political demands echo, in turn, strains in modern liberalism, as exemplified by the replication problem. Either liberals insist on the particularity of their values, and then liberalism risks becoming merely one ideology among others – as John Rawls put it, ‘just another sectarian doctrine’;40 or liberalism retains its generality at the cost of vacuity, as in the well-known remark by Robert Frost that a liberal is a man too broadminded to take his own side in a quarrel. It is not surprising, accordingly, that many liberal theorists have tried to theorise its difficulty away. Some respond by rejecting toleration outright on egalitarian grounds, as do Dennis Thompson and Amy Gutmann.41 Other writers try to convert it into an unequivocally celebratory value, as in Anna Elisabetta Galeotti’s theory of toleration as recognition.42 What risks being lost, on either side, is an appreciation of toleration as both politically difficult, and indispensable.43 Toleration and political change Toleration as a political ideal arose in the early modern period, when it could be granted or withheld at the whim of political plenipotentiaries, as with the Edict of Nantes, issued by Henri IV in 1598 and revoked by Louis XIV in 1685. Indeed acts of toleration (such as that of 1689 in England)44 sometimes resulted from initiatives towards inclusiveness or ‘comprehension’ by monarchs that were strongly opposed within the legislature. In Hungary the Edict of Turda of 1568 promulgated by King John II Sigismund gave congregations the right to dismiss preachers whose views were uncongenial to them; the Edict was revoked, however, after John was deposed. Similarly, in personal life, one party may choose to permit an action of which he or she nonetheless disapproves. In each case, toleration combines contrasting attitudes within a single agent. On the other hand, in modern political conflicts, action is called for only when one party fails to tolerate the actions of another. If the disapproval is not acted upon, no political intervention is necessary; but if intervention is called for, it is because at least one side (if not both) fails to tolerate the 40 41 42 43

44

John Rawls, ‘Justice as Fairness: Political, not Metaphysical’, in Rawls, Collected Papers. Amy Gutmann and Dennis Thompson, ‘Reply’, in Stephen Macedo (ed.), Deliberative Politics: essays on democracy and disagreement (Oxford University Press 1999), 251. Anna Elisabetta Galeotti, Toleration as Recognition (Cambridge University Press 2002). For an appreciation of the difficulties faced by those tolerated, see Leslie Green, ‘On Being Tolerated’, in Matthew H. Kramer (ed.), The Legacy of H. L. A. Hart: legal, political, and moral philosophy (Oxford University Press 2008). Andrew Murphy, Conscience and Community: revisiting toleration and religious dissent in early modern England and America (University Park, PA: Penn State University Press 2001), 157ff.

Toleration and political change

15

other. The problem of ‘tolerating the intolerant’, often considered a special case, or paradoxical as already mentioned, may be rather seen as the political norm.45 This highlights a further, structural, aspect to the regulation of modern societies, which distinguishes them from their early modern predecessors. It is a feature that has been strongly echoed in recent liberal theory. This is that the state or political authority often comes to disputes primarily as an arbiter rather than as an actor in its own right. When civil-society groups or interests conflict on some issue, the state may be less concerned with who is right than with controlling the dispute so as to secure or preserve other public goods; for example, so as to keep the peace, or prevent the causing of other kinds of nuisance. In this setting, the political authority is in effect a third party holding the ring between vying civil-society groups rather than a protagonist in its own right. The state is nothing if not coercive, and its prime role in the circumstances of toleration is to impose a solution when the protagonists reach deadlock. This is not the same as acting tolerantly. The private judgements that often trigger problems of toleration cannot be translated into public policies of toleration, since they come to matter politically precisely because they start by rejecting toleration.46 Of course, any resolution that a political authority seeks to impose will be supported by norms, such as those that appeal to the public interest. But in general the state is less concerned with tolerating in the face of its own disapproval than to achieve a resolution which secures key public goods. In this situation, the state does not itself act tolerantly. It may not disapprove of the actions that are disputed. It may well be more concerned that the due forums, legal or other, are in place to adjudicate the dispute according to public norms. Even where norms of toleration play a role in formulating policy, these will generally be understood as an ideal of individual rights or of coexistence that the authority imposes in order to inhibit those who would otherwise act on their own intolerance. The political authority may, directly or through subsidiary institutions, enforce a resolution that requires in effect that one of the parties submit to a practice despite their own disapproval of it. In so acting, the state itself does not act tolerantly. It acts not to restrain its own disapproval, but to restrain that of the party concerned, whose behaviour threatens public goods. Thus the authority may intervene in toleration disputes to force a resolution not based on reasons supporting toleration itself. For example, 45 46

Newey, Virtue, Reason and Toleration, ch. 5. Williams, ‘Toleration: an impossible virtue?’, 18–19.

16

Introduction: toleration in trouble

modern liberal states devote resources to policing demonstrations and counter-demonstrations, between supporters and opponents of Israel, of fox hunting, abortion, homosexual rights, the Iraq War, censorship, nuclear weapons, sectarian marches, and other cases. In preventing pro-abortionists from expressing their hostility towards anti-abortionists, or conversely, political authority does not restrain itself from acting on its own disapproval: it acts so as to mitigate the effects of others who act on theirs. In so doing, the authority does not act tolerantly, but acts to contain the intolerant behaviour of others. Its prime concern is with maintaining public order, or with forestalling or mitigating certain kinds of nuisance, or with protecting individuals and groups from suffering greater or lesser evils. On the other hand, when the state does involve itself in regulation, the upshot is often inimical to toleration. For example, European Union countries, including the United Kingdom, have in recent years introduced legal restrictions on smoking in public places, so that it is now illegal to smoke in restaurants and bars. Reasons for disapproving of this practice, including the nuisance to non-smoking patrons and the health risks to workers in these establishments, figured prominently among the grounds cited for introducing a legal ban on smoking in them. It was not seriously debated that there might be a case for not imposing such a ban even if the reasons for disapproval were sound in themselves. Had this been considered, more flexible approaches to dealing with the matter of public smoking might have come to light, such as the possibility of granting entertainment premises different categories of licence, depending on whether or not they wished to permit smoking. This would still have permitted a ‘steer’ from the authorities by imposing more onerous conditions (such as a higher tariff) on those restaurateurs and publicans who sought a smoking licence. This illustrates one perverse effect of a political culture where claims about toleration are commonplace: they can foreclose political options that might serve to accommodate difference. It might be thought that political actors’ failure to be motivated by theoretical accounts of their best reasons for action merely posed a problem in ‘partial compliance’ theory,47 the branch of political philosophy devoted to determining what one should do when others, such as the public authorities, fail to comply with the demands of justice. Then problems of toleration can be explained as the result of a failure to entrench the right motives of justice in a ‘well-ordered’ society.

47

Cf. Rawls, Theory of Justice, 8; §25, §39.

Toleration and political change

17

Following an ideal-theory model, agents who act tolerantly when this is what the theory prescribes may be seen simply as failing to act on their own best reasons for action. But liberals’ failure to proselytise some of the population is not simply bad luck, or the result of people’s failure to act on appropriate reasons. It is also to some extent endemic in the liberal project. While in certain respects the political ideal of toleration poses specific problems of justification, it also, as I have suggested, exemplifies wider problems facing liberal attempts to deal with difference. Toleration is the product of reasons for action: first, a reason for disapproving of a practice; and second, a reason for not acting on this disapproval, e.g. to ban or otherwise interfere with the practice.48 In the previous section, I argued that what are often perceived as anomalies internal to the concept of toleration are better explained by the ambiguities of toleration in political practice. But this is not to deny that conceptual analysis can help to clarify the practical aporias of toleration. Contemporary political theory is well attuned to the political problems posed by disagreement in judgements of value. Disputes over toleration raise these problems in an acute form, since toleration disputes are usually about the justification and scope of private judgements, which are often evaluatively idiosyncratic. In democracies, toleration faces added problems, given the assumption that private judgements – ‘liberal’ or otherwise – should not be translated directly into public policy but subject to tests such as that of public reasonableness.49 The call for toleration is likely to be heard only when factions have failed to check their disapproval of each other and are at loggerheads. Such difficulties also exemplify the obstacles to making liberalism politically effective. For liberalism finds itself on treacherous ground when attacked by anti-liberals such as fundamentalists. If accommodating difference is liberalism’s raison d’être, it is liable to be embarrassed when anti-liberals invoke the claims of difference in attacking liberalism. Either liberalism amounts to a contentless defence of ‘difference’, and proves justifiable, or at least unexceptionable, at the cost of vacuity; or liberalism insists on the particularity of its values, and so no longer looks like a meta-ideology, but like ‘another sectarian doctrine’.

48

49

One can add that there will in addition have to be a set of reasons which give grounds for justifiable intolerance (or non-tolerance) – reasons which justify banning or stopping those practices deemed to be intolerable. Cf. ‘The Idea of Public Reason Revisited’, in Rawls, Collected Papers. Cf. chapters 4 and 5.

18

Introduction: toleration in trouble

Of course it may be said that liberalism is not reducible simply to a contentless defence of ‘difference’: it consists in a commitment to core values such as liberty and equality, and norms that call for accommodating difference rest on those core values. But the more specific the values become, the harder it becomes to pretend that any real rapprochement with ‘difference’ – particularly the difference which rejects those values – is on offer. This is the counterpart, in liberalism as a whole, of the dilemma for toleration which I have already set out. The problems arise in defending specific political values as being ‘justified’ regardless of whether anti-liberals accept them, if the nature and scope of justification are themselves what is at issue. Then the question of justification rapidly turns into a question of authority – with what warrant the values hold sway over everyone, anti-liberals and liberals alike. On the standard answer to this latter question, the warrant for the institutions’ authority comes from us, and we limit its scope by being joint authors of our own rule. But this answer comes under heavy pressure when we tackle issues of toleration. These arise from the subjective force of private judgements, just as they did when discussed by Spinoza and Locke. The question then might seem to be how heavily the institutions’ authority weighs against that of the inner voice. However, although toleration is born of dispute, it is not always politically unviable. The political problem of toleration arises when there are groups that fail to share deliberatively effective reasons in the face of conflict. An effective understanding of toleration in practice needs to accept that the political problem of toleration begins precisely when the theoretical reasons that purport to justify accommodation or suppression have failed to prove deliberatively effective. This does not mean accepting at face value the normative claims that either side may make in political conflicts over toleration. But it does mean acknowledging that strongly corrective theories of reasons for action will be unlikely to resolve the problem in its own terms. To refer the political problems of toleration back to the concept of toleration is likely to make them seem more intractable rather than less. Take the problem already referred to, of ‘tolerating the intolerant’: as I have suggested, this case, often thought marginal, could rather be seen as endemic to the politics of toleration. At the level of conceptual analysis, the idea that there is a paradox is mistaken, because neither alternative is impossible in itself. Tolerating the intolerant is something that liberal societies can and often have reason to do, for example when the intolerant benefit willy-nilly from liberties that cannot be reserved to the tolerant. On the other hand, failing to tolerate the intolerant does not involve even a pragmatic contradiction: there are some things

Conclusion

19

that liberal societies should not tolerate, and certain intolerant acts figure among them. Intolerance is sometimes, but not always, intolerable.50 The reason for not tolerating it is not merely that it is intolerant, but that it imposes certain burdens or public nuisances on others, or damages goods that the public authorities can reasonably think that they are entitled, even required, to protect. Toleration is a prima facie good, and the lack of toleration consists in the lack of this prima facie good. But it does not follow that toleration is always and everywhere good, or that intolerance is always and everywhere bad. So the problem is not insuperable. It can be overcome by rejecting the assumption that private judgements should be translated directly into public policy. Once we see the state as a corporate entity not directly beholden to its constituent parts, we can recover the idea of its acting tolerantly as a corporate individual.51 This still leaves room for the state, as the state, to make judgements of disapproval and restraint. The state would and should have little compunction, for example, in prohibiting the use of aircraft as guided missiles against its own citizens.

Conclusion I have taken and will take ‘toleration’ to refer to a class of situations in which someone who regards herself as having reasons for censuring or otherwise opposing some practice acts on countervailing reasons not to do so. As already indicated, I am less concerned to argue for this as the canonical notion of toleration – against, say, the intransitive interpretation of it distinguished earlier – than to understand the political pressure to which a concept with this shape finds itself subject, especially in modern democracies. One side effect of this pressure is to retool toleration as a more tractable concept, of which the element of disapproval, for example, has been purged. Two contentions will come to the fore in the following chapters. First, not just the political discourse of intolerance, but also those who argue for toleration, rest on political decontestations that are as open to dispute as any others. This includes, on the one side, security-based arguments for intolerance,52 such as those based on public order, sedition, and public health; but also well-rehearsed grounds for toleration, such as the idea that conscientious intuitions provide a trump card 50

51

As John Rawls says of the intolerant: ‘[their] freedom should be restricted only when the tolerant sincerely and with reason believe that their own security and that of the institutions of liberty are in danger’: Theory of Justice, 220. 52 See Chapter 3. See Chapter 8.

20

Introduction: toleration in trouble

which individuals can play against the law.53 In these matters the framing thought is that ‘we’ are the prospective tolerators, for whom the question of toleration arises in relation to a prospectively threatening other.54 Often these arguments cut both ways, or prove either more or less than is aimed at by the political advocacy in whose aid they are pleaded. In some cases, the justifications face incoherence, or indeterminacy at the very point where they need to be determinate.55 Second, toleration cannot ignore power. Even were there an intellectually compelling demonstration that a particular (say, liberal) theory of political design was uniquely justified, some people would still not believe it, and then there would be a question about how to use power in order to implement the uniquely justified theory.56 Those who try to accommodate difference are liable to sell the pass. Some of the core values of liberalism, including toleration itself, lend themselves to appropriation by liberalism’s radical opponents. Due modesty about the justificatory power of liberalism should not spill over into false modesty about how best, politically, toleration should be enforced. In this setting the use of power is not only inevitable, but admirable.

53 55

See especially Chapter 6. See chapters 4 and 5.

56

54 See chapters 2 and 9. See Chapter 7.

2

Tolerating politics

Like an aged forebear, toleration is at once central to liberalism’s genealogy and a source of some embarrassment to it. Few dispute that liberalism originated with the rise of toleration in the early modern period; but modern liberals’ commitment to equal status and neutrality makes their present-day relationship with toleration uneasy. Rawls’s political liberalism aims to forge a consensus on the basic design of society between the moral outlooks, or worldviews, of reasonable people, their disagreement on other normative matters notwithstanding. Political liberalism, in attempting to fashion a basis for agreement, avoids conceptions of political design that rest on the truth of some particular worldview. It also eschews decisions reached by powerplay – by the strength of force or bargaining power. The overriding aim is to provide a basis for stability, which Rawls thought necessary for a successful theory of justice. He was explicit that arrangements resulting from the exercise of force majeure are excluded from the scope of the political, on his understanding of politics: agreements based on force are, in a striking phrase that he used repeatedly, ‘political in the wrong way’.1 As this phrase suggests, Rawls’s definition of ‘political’ was stipulative. To be political in the right way is to ground basic institutions and procedures not on power, but on reason: power lies outside the scope of politics. It is not immediately apparent what other category is to be applied to powerplay, once the label ‘political’ has been denied it, but one possibility is that of war.2 It is however a basic feature of politics as commonly understood – an understanding which accepts the use of power – that the use of political power is something other than war. Political engagement may include, but is not exhausted by, the use of force. Even so, the implication is that when it achieves its outcomes by applying force, politics is war: and war not by other means, but by the same means. 1 2

Rawls, Political Liberalism, 142; cf. his Justice as Fairness: a restatement (ed. Erin Kelly Cambridge, MA: Harvard University Press 2001), §56. Bernard Williams, In the Beginning Was the Deed (Princeton University Press 2005), ch. 1.

21

22

Tolerating politics

It is an open question how far a Rawlsian well-ordered society leaves room for political conflict of the kind familiar from the non-theory world. But Rawls’s ideas seem to leave no scope for political conflict over toleration. That has already been ruled out by his political construction, which rests on agreement, despite initial appearances, about what is reasonable. Insofar as the arguments serve to define the ‘basic structure’ of society, they are designed to take it off the political agenda by insulating it from political contention. ‘Political’ toleration is then identified as a regime in which each liberal citizen recognises that he or she cannot justifiably impose sincerely felt disapproval via public policy unless this disapproval is itself reasonable – that is, unless it is reasonably rejectable.3

Politics anyway One risks misunderstanding to start with a stipulative definition of politics, and then classify acts or concerns as political according to whether they fit in with the definition. The content of politics is largely given in advance of theoretical delimitations of it. Politics confronts us as a public arena in which concerns or interests contend with one another and make different claims (including normative claims) about the direction of policy. That these concerns surface in this way is not some contingent or marginal feature of the political scene: to an important extent it is the political scene. It is more than usually obvious with the concept politics that there is no purely lexicographical fix. Even would-be encompassing definitions – ones that broaden the meaning of the term towards vacuity – can be confuted fairly easily. Take, for instance, the well-known claim by Lasswell that politics is about ‘who gets what, when, how’.4 This seems a pretty broad definition. But it fails if, at least, a necessary condition of success is capturing what people mean when they use the term ‘politics’. Lasswell’s definition underlines the fact that much political argument addresses questions of distribution, and it would indeed be an odd understanding of politics that made such questions peripheral to it. But definitions attempt to set the limits of intelligibility. If ‘politics’ really refers to activity concerning ‘who gets what, when, how’, then it is not merely false but unintelligible for someone to argue that a certain distributive question, such as the distribution of honours, or sporting or academic contacts, lies beyond the bounds of politics. 3 4

See chapters 4 and 5. Harold Lasswell, Politics: who gets what, when, how (London: Peter Smith 1990).

Politics anyway

23

Someone who says that such matters are ‘not political’ surely makes a claim that is intelligible, even if false. Nor does it help to claim that this person is implicitly making a normative claim – that, if they are political, at any rate sporting contacts ought not to be. Then the person is saying that only what ought to be political, is political. But then this normative claim will exclude from its scope certain questions concerning who gets what, when and how: an intelligible usage escapes the definition. The transitive object of talk about politics – the thing that people take themselves to refer to when they engage in such talk – is a presupposition of it, even while the scope for disagreement about politics makes its objective status uncertain. There is a certain pathos of objectivity here, an assertion in the face of circumstances that call its assertibility into constant question. For this reason I do not suggest there is some other and better definition that needs to be put in place of Lasswell’s. The mistaken inference to draw is that ‘politics’ means whatever anyone might take it to mean, or that, as is sometimes said, ‘politics is about everything’. A more cautious, and readily defensible, inference is that no prior limits can be placed upon the content of politics. What does ‘prior’ mean here? It means, in effect, ‘prior to politics itself’, where for example theorists engaging in political design stipulate what politics is, or what politics is when understood ‘in the right way’. It is not, again, that these stipulations are false, when set against some objective understanding of what politics is. The point is rather that any attempt to circumscribe politics by defining it, or by judging certain concerns to be external to it, can only raise once more the question it aims to settle. To put the point summarily: what ‘politics’ is taken to be is a political question. It follows immediately that attempts to delimit the political while remaining outside politics fail. These attempts are pragmatically self-defeating. ‘Politics’ falls within its own extension in that attempts to define its extent are themselves political acts.5 These remarks do not in fact apply without qualification to Rawls, since he allows that there is something that is political in a sense distinct from that of his ‘political’ liberalism, albeit ‘in the wrong way’. But insofar as this marks an attempt to demarcate politics, rightly understood, from misconstruals of it, the same applies. It does not, however, follow that ‘politics’ means whatever one takes it to mean: the fact that efforts to define ‘politics’ fail does not force the conclusion that uses of the word face no constraints on intelligibility. However, definitions of it are liable either to be wrong or to 5

I argue this in more detail in ‘Philosophy, Politics and Contestability’, and in After Politics: the rejection of politics in contemporary liberal philosophy (London: Routledge 2001), ch. 2.

24

Tolerating politics

be unhelpful. For example, it is unexceptionable, if uninformative, to say that ‘politics’ means what people usually mean when they talk about politics. The fact that this uninformative gloss on the word nonetheless imposes constraints on intelligibility becomes clear when one asks whether a really meaningless term could be intelligibly substituted for ‘politics’ in all contexts where the latter occurs. Can anything else be said at this level of generality? As often in philosophy, a fork looms between a defensible but vacuous account, and one that gains content at the cost of plausibility. However, we can ask what life without politics would lack, and by that means we can hope to give content to politics without risking a naturalistic definition. Suppose, for instance, that there were no means for concerting collective action, a fantasy made vivid in accounts of the state of nature. In such a world, as Hobbes graphically described, there would be no room for deliberation over joint action, or for procedures to put decisions into effect.6 This would be a world without politics. It might well, as Hobbes emphasises, contain power imposed by one individual on another; but that shows only that there are non-political uses of power, just as there can be non-political forms of collective agency, such as military command structures. The fact that there are forms of action that involve both collective agency and the use of power, but are not political, leaves the question what else might be necessary for politics. An obvious answer is that politics is at least partially deliberative. The basic political question is What do we do? which is, in this collective version, also the basic deliberative question.7 Then it seems that a world lacking the mechanisms for concerting collective action through deliberative channels would lack politics, as it is usually understood. It would clearly lack the formal institutional apparatus needed for deliberation and for implementing decisions. It would also lack an awareness that there is anything that is the object of collective deliberation. An interestingly hybrid possibility is that of systematically convergent collective deliberation, of the sort imagined by some anarchist thinkers; in News from Nowhere, William Morris envisaged a world in which formal politics has been left behind – the Palace of Westminster, for instance, is used for storing dung.8 But in Morris’s fantasy society retains deliberative forums, in which however 6 7

8

E.g. Leviathan, ch. 13, 88–9. Military command structures may include deliberative elements. To this extent, they can also be partially politicised. However, at the level of engagement with the enemy, warmaking excludes politics. That is, to go to war with somebody is to abandon deliberation with them. William Morris, News from Nowhere, ed. David Leopold (Oxford University Press 2003).

Politics anyway

25

an ultimate coincidence of interests is assumed to yield agreement in judgements. Hence these forums are innocent of factionalism, and dispense with the use of coercive as distinct from coordinative power, since people’s judgements converge. Here there is certainly a ‘we’ that decides, and on whose behalf decisions are made. But it does not contain what most would recognise as politics. Why? An obvious reason is that the society lacks conflict. Politics dramatises deliberative options in ways that generate or accentuate the opposition between them. It is important to understand how this comes about. Some writers, such as Mouffe,9 who emphasise the ‘agonistic’ side of politics tend to explain it by reference to conflict elsewhere, for example in ‘deep’ incompatibilities between competing values. In fact, the opposition between practical alternatives exists as a straightforward consequence of the fact that the activity is deliberative. As with the firstperson singular version of the question, there will only be something to deliberate over if the answer to the question What do we do? is not immediately obvious. Where it is not obvious, deliberation will typically consider a range of different actions that are severally possible but not compossible. It will involve, in part, a contest between the rival merits of the options. When the deliberation does not issue in an immediate conclusion, it is dramatised as public contention. This explanation does not say that there must always be some background conflict, of interests, say, or of classes or values, that is played out politically – though it also does not deny that political engagement may take this form. Instead the prevalence of conflict in politics is the consequence of the simple facts that politics is deliberative, that it is usually not obvious what to do, and that there are usually several options for action, each of which has something to be said for (and against) it. Then there are institutional features of ‘high’ politics, such as party competition and the formal existence of an opposition, that often serve to accentuate conflict. Looked at in this way, ‘politics’ is simply the label applied to public deliberation that is not immediately resolved but gives rise to deliberative remainders. So the explanation of politics is not traced to pre-existing sources of conflict, though conflicts of belief, interests and value are pervasive. It is rather that the mark of politics lies in the very fact that public deliberative remainders exist. Any number of potential sources of contention may fail to lead to real conflict. They may fail to do so because of deficient resources, including lack of knowledge, understanding or the social bases

9

Chantal Mouffe, The Return of the Political (London: Verso 2005).

26

Tolerating politics

of power.10 In such cases, the claim that a certain concern is ‘political’ may have a performative status, as in claims that ‘the personal is political’. In other words, where a pre-existing order of denial or repression has excluded certain concerns from politics, the assertion that these concerns are political should be seen not as describing an extant state of affairs, but as the willing into being of a wider or altered political programme. Such claims, in isolation, are not self-verifying. They need to reach a critical mass. How that happens is a complex precipitate of the interests and values that contribute, with external factors, to public opinion formation in democracies and other political systems, a process itself open to political critique.11 An obvious example of this movement in our time is the upsurge in political concern for the environment, which previous generations regarded as being at best marginal to core political concerns over production, ownership and distribution. There is nothing internal to the concerns that propel them into the political realm, beyond their perceived relevance to fundamental human interests. The politics of race is another good example. It is not clear that there is anything at all that corresponds systematically to notions of ‘race’ as they have played out politically. They proved however to be murderously potent during the Rwandan genocide of 1994, despite the fact that the distinction between Hutu and Tutsi groupings was largely an artefact of the colonial period and lacked any sound ethnographic basis.12 Sometimes, as with the Turkish genocide of Armenians during World War One, there is a systematic – in this case, linguistic – basis for discrimination, but a political choice made the difference salient. If and when such notions become obsolescent, there is nothing that need sustain them as objects of political concern. Politics emerges not as the outworking of preconceived norms, but as the persistent residue of the sub-ideal.13 It is precisely the failure of the ideal – in the form of value, the right, the reasonable or the rational – to determine a unique course for policy, which makes for unfinished political business. To this extent, ‘politics’ simply is the name given to the

10

11 12 13

One implication of this is that potential political issues may fail to surface, for example because of false consciousness, or the systematic denial to some parties of relevant information. In such cases, what is accounted political may be seen as epiphenomenal, as in some Marxist analyses of power under capitalism. See e.g. Andrew Willis, Silent Accomplice: the untold story of France’s role in the Rwandan genocide (London: I. B. Tauris 2007), ch. 1. I argue this further in ‘Gray’s Blues: pessimism as a political project’, in John Horton and Glen Newey (eds.), The Political Theory of John Gray (London: Routledge 2007).

Non-political toleration

27

clearing-house of proposed policies whose merits cannot be decisively judged in the abstract. Thus there is still a political question to answer, namely, What do we do about these people? – even if we have decided that they are acting unreasonably, or in other ways deviate from the demands of the ideal; often such judgements constitute the circumstances of toleration. To pre-define politics as ‘agonistic’, for example on the strength of value-pluralism, is to risk putting the cart before the horse. In other cases, plausibly, neither side need be acting unreasonably, and then there is simply a gap between what the ideal demands and what determines policy.14 The gap can open up when reasonable disagreement prevails over what disagreements are reasonable.15

Non-political toleration There are certainly non-political forms of toleration. Interpersonal toleration, whether based on moral or other forms of disapproval, occurs when an individual uses her discretion not to act on this disapproval. My table manners strike you as coarse, but you decide neither to pass comment, nor to impose other sanctions (such as not inviting me for dinner again). You may consider that my lack of table etiquette embodies a moral failing, or some non-moral defect. The presence of a moral objection is not, accordingly, sufficient to mark an act of toleration as political rather than interpersonal. Conversely, the basis for disapproval in political toleration need not be moral. It may be that certain inherently political goods, such as public order, are jeopardised by some action – but, even so, there may be grounds for permitting it. A public protest may be judged exceptionable not on intrinsic moral grounds (as might be thought to be the case, for instance, with the march which the National Socialist Party of America proposed to conduct in Skokie, Illinois, in 1977), but because it threatens civic peace. But there may still be reasons to let it go ahead, for example on grounds of free expression. When political authorities decide to tolerate something, they may do so on moral or other grounds. There is also a general oddity in the moralist approach to political toleration. Moralists try to lay down norms to decide what should be tolerated. These include constraints on public reason or argument, and 14

15

In some cases it may be demonstrable that one side or other is behaving unreasonably. But on most views – including those which countenance reasonable disagreement – it will often be the case that neither side can be shown to be unreasonable, and in such cases reason simply under-determines action. See Chapter 4.

28

Tolerating politics

principles such as reasonableness, autonomy, or equal respect for persons. These norms are double-edged.16 They may favour tolerating an action; on the other hand, they may give reasons for disapproving of it – and so prompt the question of whether to tolerate it – or show why the act is intolerable. Theorists usually try to impose some ranking on the reasons in play. But people will only raise the question of toleration if the reasons recognised by the theory, or the ranking it imposes, are not dispositive for them. For example, some may not see their actions as breaching others’ autonomy, as in paternalist curbs on ‘hard’ drug use; or pay autonomy little heed, as with Brahmins’ attitudes towards Dalits under the Indian caste system. Superficially the problem is that people have different opinions about the subject of the theory – for example, about whether legalising heroin harms autonomy, or whether autonomy is relevant to the status of Dalits. Opinions also differ about whether autonomy should be the basis at all for deciding what to tolerate. At the root of questions of toleration, however, lies a dispute simply about what to do. Politics starts when a collective agent is presented with different, and more or less exclusive, practical options, and has to choose between them. At this point the question is whether, say, possessing heroin, or trading it, should be a criminal act, or simply illegal, or legal. It is not usually whether or not these acts should be tolerated, though that does not mean that toleration as a concept cannot be applied: it can be said, for instance, that possessing or dealing in heroin is not tolerated under current English law. The trouble is that the slew of considerations that typically come to bear on a political decision about drugs laws bears little relation to the interplay of reasons, such as contrasting views of autonomy, envisaged by the theory. Some abolitionists appeal to the value of autonomy or freedom, of course, but neither they nor their opponents usually marshal the complex of reasons that weigh reasons for disapproval against reasons for permission. Abolitionist arguments tend to downplay reasons for disapproval, while retentionists see no good reasons beyond those. Two objections may be made. First, theories of toleration, as accounts of the reasons which make toleration appropriate, cannot be judged by the reasons that people cite in the hurly-burly of actual political argument. Second, as a specific value, autonomy or self-determination already allows for the possibility of disapproval – it marks out a protected area for individual action despite the disapproval of others.

16

A point noted by D. Raphael; see his ‘The Intolerable’, in Mendus (ed.), Justifying Toleration.

Non-political toleration

29

Each objection has some merit. In fact, the second seems to me sufficiently weighty to warrant a chapter to itself.17 To foreshadow the conclusion I reach there, it is possible to insist that autonomy is the only basis for toleration. But that relies on an extreme counterfactual view of the conditions for justifiably exercising power that is at odds with the circumstances of politics. In the end, politics confronts us in its facticity. How a problem is dealt with – whether, say, by the methods of politics or of war – is deliberable; that there is a practical decision to make, even if it is to do nothing, is not. The matter becomes deliberable at the moment of recognising that it is there. This broadly Aristotelian insight gives a basis, in my view, for a naturalistic account of politics rather different from the contractuarianism and constructivism currently popular.18 That suggests that political decisions are typically a bricolage created from found objects, and that goes for toleration as much as anything else. The first objection relies on the idea that the reasons framed by theories of toleration can be distinguished from those that people cite, sincerely or disingenuously, in political argument. But the political conditions in which toleration is at issue prove unreceptive to abstract structures of reasons yielded by theory, such as those imposed by notions of public reason. It is incongruous to impose on a conflict between parties, in the name of toleration, a structure of reasons that either party, or each, rejects. This being so, in what sense does either tolerate the other? The theoretical claim is that each should act on reasons that neither accepts in the prescribed form, otherwise the conflict would not exist. What is in view is a set of constitutional or other arrangements that would, if stuck to, have forestalled the original conflict. To set out the norms that the adversaries should have acted on to avoid it does not help much in dealing with the situation or in understanding it. From a political standpoint, modern discussions of toleration tend to miss the point.19 The modern democratic state seldom acts as a protagonist in moral conflicts in its own right. On liberal neutralist views, the state cannot commit itself between rivalrous ideas of the good, at least if these are severally reasonable. Internecine conflicts between, say, Hispanic and black gangs in Los Angeles, or between Pashtun tribal groups in Afghanistan, more intractable than disputes about religious symbols such as crosses or the burqa, surely involve intolerance. As often, such conflicts rage over control of resources, and so are soluble 17 18 19

See Chapter 8. Another work that develops from a similar starting point, though in a different direction, is Stuart Hampshire, Justice Is Conflict (Princeton University Press 1999). See Chapter 3.

30

Tolerating politics

only insofar as conflicts over resources are soluble. Perhaps for this reason, they attract little attention from theorists of toleration. But in some structural respects they resemble less murderous and more humdrum struggles. When toleration by either lobby is not to be had, the state has to step in and make its choice not for toleration, supported by some canonical norm, but for political values such as public order or security, or to protect resources. In this way a mismatch between political practice and theory marks modern problems of toleration. Part of the reason for this mismatch is historical. Toleration was first conceived of not as an abstract set of constitutional provisions, but as a practical response to a political problem. Early modern thinkers such as Milton, Bayle and Locke saw toleration as a call on behalf of the excluded for inclusion, or ‘comprehension’. It was a call to rulers such as Oliver Cromwell, Charles II or Louis XIV, who could exercise personal prerogative power for or against toleration of minorities such as dissenters.20 This bipartisan structure is superseded in modern democratic politics, since the political sovereign, the demos, wields power through an executive which holds the ring between different civil-society interest groups. To lay down a constitutional structure that guides the executive when it acts as a third-party referee between these groups – and on modern liberal neutrality, this is all it can justifiably do – fails to address the problem of toleration between the groups. When one set of people acts intolerantly towards another, this is not addressed as a question of toleration by formulating norms that a third party, the executive, should observe in dealing with them. On this score, modern discussions of toleration tend to miss the point. Other reasons for the mismatch can be found in the norms that theorists impose on political action, especially the requirement to comply with equality of respect. The main problem is that toleration seems to entail disapproval, and this sits uneasily with equality of respect and, not coincidentally, with neutrality.21 Some relax the tension by purging toleration of disapproval, so that ‘tolerant’ becomes predicable of people who precisely fail to disapprove of some practice.22 I have already noted this ‘intransitive’ use of ‘toleration’,23 whereby a set of people or the norms they live by accept difference; but, to repeat, this only qualifies as tolerant if there is some question of not accepting it. That, once framed, is a political question – it becomes political in the moment of tabling it 20 22 23

21 See Murphy, Conscience and Community. See Chapter 5. E.g. Balint, ‘Toleration as a Liberal Political Practice’. I am grateful to Balint for discussions on this point; cf. the Introduction above. See Chapter 1.

Political toleration

31

for joint consideration. It follows that efforts to limit toleration to the intransitive use, in effect depoliticise it. Concern about the political volatility of toleration may also help to explain other responses to the disapproval notion. Some writers take disapproval as a reason to reject toleration as a political value.24 Others, while wishing to salvage toleration, are also troubled by its apparent denial of equal respect. Some solve the problem by inverting toleration into a celebratory value, for instance in the form of ‘recognition’ for hitherto disadvantaged groups such as homosexuals: political authority should positively affirm the identity of those whom it tolerates, rather than stigmatising them as fit objects of disapproval.25 Yet other writers accept that toleration entails disapproval, but annex its expression from the public to the private sphere.26 All these responses, in my view, risk misunderstanding. They seek to domesticate or tame a value that is awkward in the same way that political life itself is awkward. Each objection, then, takes an insufficiently political view of toleration. It is not enough simply to frame reasons in the abstract and ignore their political implementation, or take them as read. Nor is this simply because ‘empirical’ facts somehow impede the ‘ideal’ principles laid down by theory. The problem lies in democracy, and its associated political values of equality and autonomy: the problematic political status of toleration is implicit in democratic political values.27 Above all, theorists who try to lay down (usually moral) norms to characterise toleration risk ignoring the fact that these reasons have failed to motivate the contending parties. They risk presenting a resolution in the name of toleration where in fact none of those involved, including the state, can be seen as acting tolerantly.

Political toleration I take a mark of political action to be that it is collective. Though perhaps not all collective action is political, all political action is collective, in being done by or on behalf of a collective agent. A political authority may act on 24 25 26

27

E.g. Amy Gutmann and Dennis Thompson, Deliberative Democracy and Disagreement (Cambridge, MA: Belknap Press 1996). E.g. Galeotti, Toleration as Recognition; I review Galeotti’s book in Utilitas 18:3 (2006), 310–12. E.g. Jones, ‘Making Sense of Political Toleration’; Rainer Forst, ‘Toleration, Justice and Reason’, in Dario Castiglione and Catriona McKinnon (eds.), The Culture of Toleration in Diverse Societies: reasonable tolerance (Manchester University Press 2003). I examine the problems posed specifically by democracy as a political value for this abstractionist tendency in Chapter 3.

32

Tolerating politics

its own account, or see itself as an umpire between individuals or groups. These are importantly different possibilities for political toleration.28 Reason often fails to decide political action, because it underdetermines decisions, or because people act unreasonably. The gap left by reason is often filled by the exercise of power. The circumstances of toleration exemplify this functional aspect of politics, because they arise precisely where the ideal has failed to forestall conflict between the actors and those who wish to impede them. As I suggest later,29 since any theoretical basis for reasonable disagreement will have to allow for reasonable disagreement about the reasonable itself, public reason will not be able to determine in all such cases who is acting reasonably. But even when it can, this does not in general decide the political question. In some such cases it may be possible to decide which course of action is the more tolerant, though as I already noted, that will still not tell us what we should do. The circumstances of political toleration exemplify this functional point about politics. In these circumstances, the one-sided or mutual antagonism has reached the point where one or both parties are prepared to act on this antagonism against the other. Here the two sides will offer sharply contrasting proposals about what should be done. One side wants a ban on wearing certain kinds of religious dress in public, say, while the other opposes such a ban; or one party wants smoking to be permitted in public places, while the other wants it to be prohibited; or similar disputes, mutatis mutandis, with sectarian marches; or allegedly blasphemous plays or art; or fox hunting; or gay sex; or the availability of drugs, or poisons; and so on. There is no prior reason why any specific issue must lead to conflict. But when it does, politics is the public forum in which conflict is played out. Then the standing political question – What do we do? – arises with regard to the conflict. Because the question poses itself in the first person plural, its answer implies not simply a course of action, in the form of public policy, but also a certain identity for those who take it. Further, because political action is usually carried on by one group of people on behalf of another, the political agents implicitly define those for whom they act. Political argument usually appeals to principles that purport to serve the interests of the public as a whole, rather than some interest-group. Toleration is no exception. Appeal is routinely made in arguments over toleration to notions of individual or group liberty, to social cohesion, to sanctity of conscience, to the security of the state, and

28

See especially Chapter 3.

29

See chapters 4 and 5.

Political toleration

33

so on. On some occasions, the selfsame principles are cited by each side in support of their opposing policies. When appeals to public interest on one side are met by countervailing public interest claims on the other, arbitration is needed. Here a conceptual fix proves tempting. To put it quickly and schematically: political action expresses political thought, and political thought has conceptual content; so political concepts are ‘prior’ to political action, and philosophy, insofar as its business is with clarifying concepts, can determine political thought and thereby political action. Political action certainly expresses political thought, and that thought has conceptual content, but the inference that political concepts are prior to action is faulty. One can take a fully realist – that is, verification-transcendent – view of the identity-conditions for concepts, while denying the priority claim. One could also accept that concepts are individuated by their possession-conditions,30 but deny that the content of a concept could be fully specified ahead of encounters between people using it. Whether someone possesses or grasps a concept may itself be a political matter. Has someone who believes that the concept of freedom allows, say, for people freely to enter into slavery contracts, simply failed to grasp the concept? Tyler Burge gave the example of a man who uses ‘arthritis’ to include rheumatic pains in one’s leg, as well as pains in the joints.31 In some other world, the extension of the term ‘arthritis’ might have included rheumatic pains; but that doesn’t show that that term would then have named our concept arthritis. So, on Burge’s view, in the counterfactual world the content of the man’s belief about his condition would be different from the content that is attributable to him in this world.32 On this view, the terms used to refer to concepts rigidly designate across possible worlds. Identical dispositions to verbal behaviour can stem from different concepts, given identical beliefs other than those about the concept itself. But in politics, disputants usually do not have identical verbal dispositions; realism about concepts gets little practical purchase. The appearance of agreement may mask underlying disagreement, and 30 31 32

See Christopher Peacocke, A Study of Concepts (Cambridge, MA: MIT Press 1992). Tyler Burge, ‘Individualism and the Mental’, Midwest Studies in Philosophy 4:1 (1979), 73–121. A claim characteristic of the position often referred to as content-externalism. One question that arises is why if concepts rigidly designate, one does not say that the man’s pseudo-concept arthritisþ already designates something different from the thisworldly concept arthritis. Concept-externalism is not defended here. All that is claimed is that arthritis, and the man’s arthritis-plus-rheumatism, are distinct concepts; rather than that because they are distinct, content can only be characterised by reference to conventional meaning.

34

Tolerating politics

vice versa, even if one holds that there is an objective truth about whose understanding is right. So concepts each of which is picked out by the English word ‘freedom’, for example, may be distinct irrespective of English-users’ dispositions to verbal behaviour. Fixing labels matters less than what political job a specific concept does – such as one that allows that people may freely enter slavery contracts. Parallel remarks apply to political toleration. As a term in English, ‘toleration’ can certainly be used to designate the free and easy acceptance of others or their conduct, as well as self-restraint despite disapproval. The argument, though, is not that those who uphold this view of toleration must have made a mistake about the concept. Debate about that question is, as I have already argued, political, and potentially interminable. It is that the conditions of politics predictably give rise to questions like this: How should we deal with those who, because of their own practices or others’ attitudes, excite disapproval? It is the basic political question, of the What should we do? kind. In the circumstances of toleration, that usually means deciding whether some action, or acts of a certain kind, can go ahead. Then there is the broader question, whether there is political use for a concept – whether labelled ‘toleration’ or not – that has a certain shape, one where reasons for disapproval are counterbalanced by reasons for non-censure. Take discussions of female genital mutilation or modification (FGM),33 often thought to raise questions of toleration.34 An autonomy theorist might argue that FGM in any of its forms should be tolerated when, but only when, it does not breach the autonomy of those on whom it is practised: autonomy, on this showing, divides warranted toleration from the intolerable. But this is hardly the end of the argument. Fairly obviously, autonomy as such is unlikely to be deliberatively effective for those women or girls on whom FGM is practised in traditional societies. Does that make it intolerable? The judgement does not seem compelling. Perhaps ‘autonomy’ can, after all, be applied to actions that conform to acculturated norms, insofar as such a notion even makes sense.35 But perhaps, by contrast, the tolerability of such practices should not turn on autonomy anyway. How can we, as those who pretend to tolerate the practice or not, impose our reasons on them, the practitioners of FGM, in the name of their autonomy? Rousseau’s paradox looms large. On the 33 34 35

Sometimes called ‘female circumcision’, though not all instances of FGM involve cutting. E.g. Xiaorong Li, ‘Tolerating the Intolerable: the case of female genital mutilation’, Philosophy and Public Policy Quarterly 21:1 (2001), 2–8. It is a substantial question whether or not ‘our’ concept autonomy answers to anything systematically recoverable from the deliberations of those in the traditional culture.

Political toleration

35

theory, genital modification is tolerable when but only when it does not breach autonomy. If it does breach autonomy, it is intolerable; if not, it should be permitted, but then there is no evident basis for disapproval at all. This point is not met by the possibility of distinguishing those, such as adult women, who can autonomously consent to FGM from those, such as infants, who cannot. Either way, political toleration seems to disappear. The danger is that theorised norms fail to do justice to normative complexity, and erode toleration’s distinctive political shape – or, not to beg the question just raised, a concept with the shape identified in Chapter 1. One might think that the above arguments applied peculiarly to autonomy-based justifications for toleration. But they spill over to others, such as those based on public reason. It is not clear whether banning or allowing FGM will pass the public reason test. Someone may say that a ban rests on grounds that all affected can reasonably accept, or none can reasonably reject, because genital modification involves a gross violation of physical integrity. On the other hand, it may be said that nobody should be prevented from pursuing an integral part of their idea of the good; though liberals will want to add the proviso that unbreachable curbs, such as not inflicting harm on others, must be respected.36 Again, passing the test of public reason – for example, that the reasons grounding policy should be reciprocally and generally acceptable to those subject to them37 – makes permission and prohibition all-ornothing matters. So, at the public level, the reasons in favour of allowing FGM pass the test, and it is not merely tolerated, but fully accepted; or they fail, and the practice is deemed intolerable. Consider, by contrast, male genital mutilation or modification (MGM). This gets less discussion as a question of toleration than does FGM. It is hard to see on the basis of the procedures themselves or their associated cultural practices why this is so. It is tempting to put this difference down to the fact that culturally prominent groups such as Jews and Muslims practise male circumcision as, for prophylactic reasons, do many infidels and goyim, whereas FGM is carried out by small minorities in western countries and by globally marginal groups in poor parts of sub-Saharan Africa. The practice is also strongly opposed by some western feminists. Since the bris (penile circumcision in orthodox Judaism) is performed on neonates, it cannot plausibly be depicted as an autonomous

36 37

A point which raises the question whether the harm proviso rather than public reason itself does the real justificatory work. For this formula, see Rainer Forst, Contexts of Justice: political philosophy beyond liberalism and communitarianism, tr. John M. M. Farrell (Berkeley: University of California Press 2002), e.g. 39.

36

Tolerating politics

choice made by the individual subject to it; nonetheless, autonomy is often cited as a ground for opposing FGM. The point is not that the justifications fail because they fail to distinguish morally between the male and female versions of the practice. It is that choosing or interpreting certain kinds of reason as relevant to deciding the issue – something that could help to answer the question What do we do? – is a condition of its being a political question in the first place. It is not that there is no case to answer. According to one report on the practice, [a] 1999 study of 48 boys who had complications from traditional male circumcision in Nigeria found that haemorrhage occurred in 52% of the boys, infection in 21% and one child had his penis amputated.38 Other complications can include concealed penis, urinary fistulas, chordee, cysts, lymphedema, ulceration of the glans, necrosis of all or part of the penis, hypospadias, epispadias and impotence.39 In 2010, Bollinger[40] estimated a death rate of 9.01 per 100,000, or 117 per year in the United States.41

Many would argue that such complications at least made the practice one that merited public investigation. Pressure groups in the United States and the United Kingdom aim to secure a legal prohibition on both female and male genital modification, at least as practised on minors. In 2011, activists in California tried to lay a proposition in favour of outlawing MGM before the San Francisco electorate. Their campaign was baulked by Judge Loretta M. Giorgi after opponents, including religious groups and doctors, sued to have the motion struck from the ballot.42 Those groups followed through a strategy of preventing male genital cutting from being considered politically. Admissible reasons were thus limited to juridical rather than more broadly political ones. Judge Giorgi ruled that the question was not eligible for political decision in view of the legal embargo in California on derogating from the practice of medical procedures approved elsewhere in the state. Hence, at the political level, no question of ‘tolerating’ MGM arose. To ask whether ‘we’ should tolerate FGM is already to frame the political subject in a 38

39 40 41 42

Figures cited from A. Ahmed, N. H. Mbibi, D. Dawam, and G. D. Kalayi, ‘Complications of Traditional Male Circumcision’, Annals of Tropical Paediatrics 19:1 (1999), 113–17. Quoted from George W. Kaplan, ‘Complications of Circumcision’, Urologic Clinics of North America 10:3 (1983), 543–9. Dan Bollinger, ‘Lost Boys: an estimate of US circumcision-related infant deaths’, THYMOS: journal of boyhood studies 4:1 (2010), 78–90. Citation from the original text. Cf. the Genital Autonomy webpage: www.genitalautonomy.eu/#/mgm/4543432632, which cites the 1999 American Medical Association report. LA Times, 29 July 2011, ‘Judge orders San Francisco circumcision ban off ballot’: http:// articles.latimes.com/2011/jul/29/local/la-me-circumcision-ban-20110729.

Political toleration

37

certain way, and is liable to create blind spots. The United States founding fathers asked what demographic value they – in their first-person role as ‘we, the people’ – should assign slaves for the purposes of apportioning congressional representation and taxes, and finally decided that one slave equated to three-fifths of a free man – even while upholding the doctrine that all men are created equal.43 Perhaps the reasons for banning FGM are stronger than those for banning MGM. It is however quite unlikely that theory can offer an authoritative account of the balance of these reasons, which shows why an asymmetrical ban is justified. The political situation is not well understood by judging, on the basis of such considerations that, say, FGM is intolerable while MGM is tolerable. Very often, the issue is precisely which actions count as ‘reasonable’.44 Justifications sway the political agenda not so much because they are reasonable, but because politically articulate and cohesive groups voice them in terms which they and others find persuasive. It becomes tempting to say that this politics of power-distorted reasons should cede to a politics responsive to the abstract balancing of theory. Political toleration comes up because actors fail to adopt reasons for toleration, or accepted reasons that show their actions are intolerable. One can appeal to notions such as autonomy, or justifiability to those affected, but part of the problem is precisely whether autonomy or directed justification are relevant here, and how they get calibrated against other kinds of reason. One might hope, when confronted with examples like the Three-Fifths Compromise, that duly enlightened political actors will light at last on correct reasons for action, rather than those that simply express local prejudices. As Alexander Hamilton said in the 1787 Compromise debates, ‘it is the unfortunate situation of the Southern States to have a great part of their population, as well as property, in blacks’.45 But the circumstances of political action, especially when it involves toleration, inhibit this process of epistemic ascent. They add what might be called a pathos of distance to the reasons on which embodied political subjects think and act. It is a reference-frame illusion to imagine that ‘we’ look down on an earlier ‘them’ from ground that is necessarily higher up. 43

44

45

See the materials in The Debates in the Several State Conventions On the Adoption of the Federal Constitution, as Recommended by the General Convention at Philadelphia, in 1787, ed. Jonathan Elliot (Washington, DC: J. B. Lippincott & Co. 1866). The idea of a common or shared world is sometimes used to denote a matrix of shared normative agreement. See e.g. Ingrid Creppell, ‘Toleration, Politics and the Role of Mutuality’, in Melissa S. Williams and Jeremy Waldron (eds.), Toleration and Its Limits: NOMOS XLVIII (New York University Press 2008), e.g. 322. The Debates in the Several State Conventions, 237.

38

Tolerating politics

Induction suggests that the same will go for our reasons, when viewed from the standpoint of our descendants. No doubt scaling down en bloc the civic value of a whole population group is not an option for us, as it was for the US founding fathers. But, to repeat, the belief that that vantage point will necessarily be ‘higher’ than ours may rest on a frame illusion: that they will think so follows simply from the fact that they will have different beliefs from ours – perhaps about genital modification – and that having a belief entails believing that it is true. In this respect one cannot but see changes in belief as progressive. Political action often proves resistant to reflectiveness, and this is as true of democracies as other political forms. In the circumstances of toleration, especially, reflectiveness is often in short supply. The obvious retort is that, at any rate, the actors ought to be more reflective in these circumstances, or more responsive to the reasons that theory says they should accept. But it seems futile to respond to a political conflict by merely insisting on the primacy of certain reasons which, had they been followed, would have forestalled it. Of course, it can be useful to learn where one went wrong. But the fact that directions can, more or less, be right or wrong may simply attest to disanalogy rather than analogy. The claim that some reasons are right and others not does not resolve the political conflict, but perpetuate it. A further implication of my approach is that no clear line can be drawn between ‘descriptive’ accounts of the politics of toleration, and theorising about toleration. On the one hand there will be an irreducible theoretical component in the political phenomena which we are looking for. At the same time, the phenomena themselves constrain what is theoretically avowable, not least because of the pressure of other normative ideals such as democracy, or equality. Practice helps to determine what is politically possible; but practice itself is of course not uncontaminated by theory. The upshot is a kind of reflective disequilibrium between pragmatism in political theory and idealism in political practice. The role for political theorists, as far as this goes, is to come up with ways of living with incoherence. Conclusion Political pressure to transform toleration into something else, such as equality or mutual respect, is almost as old as toleration itself. Distortion looms if theorists first try to come up with the ‘true’ account of toleration – based for example on reasons that political actors should accept – and then castigate or ignore these actors’ failure to act on them. Whatever is political in the ‘wrong’ way is still political; so for Rawls,

Conclusion

39

accommodations based on powerplay remain political. In Rawls’s sense, political life is ‘political in the wrong way’ most of the time. There would be a political issue to address even if politics could be defined uncontroversially as the domain of the reasonable. If the political construction at which Rawls aimed worked, we would then have a univocal account of tolerable and intolerable action. It must be common ground that, relative to this theory, some people would still act in an unjustifiably intolerant way. Along with other liberals, Rawls is prepared to say that unreasonable action can be a fit object of coercion: the concept of the reasonable can tell us when the use of power is or is not justifiable. But even where this concept gives a clear verdict on whether an action is intolerable, it is an open political question how to deal with its intolerability, given that the action may legitimately be prevented. Whether or not it is a good idea to do so will depend on irreducibly political considerations, such as the threat to public order, well-being – or what kind of polity political actors take theirs to be.46 That may suggest a robustly étatiste approach to political toleration. Public order and security certainly play an important role in the political background to problems of toleration, and constrain the set of solutions to them. But invoking security and public order as values does not determine a solution: for example, one that seeks to ‘securitise’ the political agenda.47 The idea that security is the political summum bonum or sine qua non is as open to challenge as any other such claim. It does not predetermine decisions about liberty or toleration – for example, that they must be sacrificed to security.48 The wider point is that politics appeals to a notion of freedom as undetermination,49 even while the circumstances of political decision-making strongly limit what political actors can do, and what they can justify to themselves and others.

46

47

48

A further difficulty: presumably this prevention counts as ‘curbing agency’ if anything does. If so, this form of response to the intolerable will always take the form (albeit, by hypothesis, justifiable) of intolerance. Made current by the ‘Copenhagen school’ in international political thought: see Barry Buzan, Ole Waever and Jaap de Wilde, Security: a new framework for analysis (Boulder, CO: Lynne Rienner 1998). 49 See Chapter 7. See Chapter 10.

3

Democratic toleration?

According to the old ‘Whig’ interpretation, British – or in its more ambitious versions European, even world – history was marked by an ascent from autocracy to liberalism. In one version of this story, the emergence of toleration in the early modern period heralded the advent of liberalism, which in turn called forth democratic ideals and institutions. The junction between liberal toleration and democracy lies, on this view, in the immanent commitment within toleration to the equality of persons, which extends, in its political aspect, to equal civic rights, including the franchise; and to autonomy, which toleration upholds in the form of other civic freedoms integral to democracy, such as freedom of association. This chapter’s less sanguine argument is that democracy and toleration tend to conflict with each other and do so foreseeably, for reasons that lie within the institutions and culture of democracy itself. This casts doubt on whether toleration remains a coherent political value in modern conditions.1 One problem is that the structure of political toleration today differs from that in early modern Europe. More generally, as I indicated in Chapter 2, the threshold of politics lies where agreement, including rational agreement, proves unavailable. That disagreement can and does extend, in the political circumstances of toleration, to understandings of freedom as a constraint on the exercise of power. The main prospects for democratic toleration are pre-emptive, in closing down opportunities for civic intolerance.

Structural transformation of toleration When the ideal of toleration was formulated by thinkers such as Bayle and Locke during the early modern period, the predominant form of 1

To this extent the situation may be more dire than that envisioned by those who think that toleration cannot be instituted politically as a moral value (e.g. because of moral disagreement) but that it can be given a purely political grounding, in roughly the sense of ‘political’ adopted by Rawls in Political Liberalism.

40

Structural transformation of toleration

41

government in Europe was personal and prerogative rule. The sovereign stood as individual person to the collectivity of the citizenry or subjects, a point made explicit in Leviathan’s account of authorisation.2 In these circumstances, toleration could be exercised on a basis strictly analogous to – in fact, as a special case of – the interpersonal situations already considered. A century later, Kant could still write in What Is Enlightenment? that ‘[a] prince who does not regard it as beneath him . . . in religious matters, not to prescribe anything to his people, but to allow them complete freedom, a prince who thus even declines to accept the presumptuous title of tolerant, is himself enlightened’.3 But Kant’s remarks here look forwards as well as back. His underlying point is that to tolerate is to wield power where power ideally should be absent or, at least, disseminated among the population at large. Paine makes a similar point in The Rights of Man: Toleration is not the opposite of Intolerance, but is the counterfeit of it. Both are despotisms. The one assumes to itself the right of withholding Liberty of Conscience, and the other of granting it. The one is the pope armed with fire and faggot, and the other is the pope selling or granting indulgences.4

The exercise of power in democratic polities is liable to make toleration impossible precisely because the personal-prerogative use of power that provides Kant’s and Paine’s tacit model of political action can no longer be justified within democracies. On populist views, the sovereignty of the demos lies at the core of democracy, and this is usually taken to require that the populace has control over the political process.5 This control embodies the core democratic values of political equality and autonomy:6 each citizen has an equal say in political decision-making, and their participation in it is the only way for citizens to subject themselves to a law which they have given to themselves. Given a suitable procedure for putting the popular will into effect, the test of democratic 2 3 4

5 6

Though he allows for representation by a corporate sovereign, Hobbes makes clear in Leviathan ch. 19 his preference for monarchy. Kant, ‘An Answer to the Question: “What Is Enlightenment?”’, in Kant, Political Writings rev. edn tr. and ed. Hans Reiss (Cambridge University Press 1991), 58. Cf. Paine, Rights of Man, ed. Henry Collins (Harmondsworth: Penguin 1969), 107. As with Kant’s remarks, this presupposes the discretionary and personal exercise of political power. For the most influential statement of this view, see Jean-Jacques Rousseau, Du Contrat Social, tr. G. D. H. Cole (London: Dent 1913). Thomas Christiano, The Constitution of Equality: democratic authority and its limits (Oxford University Press 2008) bases the justification of democracy on a prior commitment to the equality of persons. For an autonomy-based defence of democracy, see e.g. Henry Richardson, Democratic Autonomy: public reasoning about the ends of policy (Oxford University Press 2002).

42

Democratic toleration?

legitimacy, applied to a given measure, is whether it stems from the popular will as voiced by this procedure.7 The question is how well this account of democratic legitimacy can accommodate toleration. It is important for toleration that the type of reason for non-intervention be of the right kind.8 For instance, it may simply be a calculation of interest based on cowardice, or mercenary reasons for inaction. At the limit, someone pursues non-intervention because no alternative is possible. Beyond the limit, as it were, the possibility of intervening lies in the hands of somebody else, because the person, far from acting tolerantly, is the one whose action – whose failure to tolerate – makes toleration an issue. In personal life, toleration usually involves two individuals judging and acting on their own behalf. One person elects to tolerate her friend’s weakness for soya milk or Pre-Raphaelite paintings or musical kitsch, and the matter ends there. Forbearance marks the dispositions of the tolerant individual. Of course, matters may prove more complicated. For instance, gross failures of taste may provoke a gradual disengagement from the other person, or signal a wider loss of affective ties. More interesting, the internal conflicts to which the situation gives rise may prompt the tolerator to question her own unreflective responses to the other, and modify her patterns of reaction; part of what it means for toleration to be an executive virtue9 is that the executive takes account of events, so to speak, on the ground. In general, however, interpersonal toleration means that one person takes responsibility for her own attitude of disapproval, and acts as she sees fit. An important difference marks the modern political circumstances of toleration compared with interpersonal cases. Whereas in the standard interpersonal situation the tolerator stands to the tolerated as first party to second, the authorities, in the political circumstances of toleration, are typically related to the conflicting groups as a third party. This is particularly clear in neutralist theory, where political authority is seen as a normative vacuum. Since toleration requires that the tolerator have reasons for disapproving of the practice, and must nonetheless have reasons for regarding non-intervention as good, the normative vacuum is filled, and neutrality disappears.10 Nonetheless, some liberal neutralists,

7

8 10

I omit public-goods and other telic justifications of democracy. For a critique of populist conceptions of democracy, see William Riker, Liberalism versus Populism (San Francisco, CA: W. H. Freeman 1980). 9 See also Chapter 8. Newey, Virtue, Reason and Toleration, ch. 3. See also my ‘Metaphysics Postponed: liberalism, pluralism and neutrality’, Political Studies 45:2 (1997), 296–311.

Structural transformation of toleration

43

like Charles Larmore,11 believe that toleration is not merely consistent with, but entailed by neutrality.12 So the reasons in play are likely to differ from the interpersonal case, when the political authority acts as a third party with respect to two vying groups in civil society. A fundamental difference is that the introduction of a third party brings with it the exercise of power over others. The possibility of toleration always demands power in some sense,13 but individual forbearance demands only that one exercise power over one’s own actions, often pre-emptively. By contrast, if those in political power act to stop the second party acting against the first, coercion has entered the picture. Prima facie this demands justification, since it is one thing to restrain myself from intervening against a practice I disapprove of, and another to restrain somebody else from doing so. For one thing, the value of autonomy is engaged when I use my power to prevent another from acting as he or she would otherwise act.14 If coercion raises the bar for justification, the grounds for enforcing non-intervention on others will have to be stronger than those for restraining oneself. But the introduction of coercion is not the only novel feature of the three-party case. This comes out if we compare the authorities’ reasons for action with those of the original protagonists – those carrying on the practice, and those disapproving of it. Take the case of a march or demonstration, to which another group objects: they call on the authorities to ban the march. Clearly the reasons on which the authorities act, if they act tolerantly, cannot merely duplicate those of the protagonists, since the political problem is created by their failure to tolerate one another: it makes little sense to say of either party that it tolerates itself. But, whatever they decide to do, the authorities act on reasons that may bear little relation to those operative in two-party cases. Suppose that they decide to enforce acceptance of the demonstration. Then it will be a matter of coercing the objectors so that they do not intervene against it. The reasons which support this policy concern the consequential benefits of using force in this case, and the would-be coercive actions of the objectors to the practice will not be tolerated. The state enforces restraint on those who will not restrain themselves, as it does with laws that restrain people from taking their disapproval of abortion, say, or homosexuality to the point of attacking its practitioners. If, on the other hand, the authorities enforce a ban on the demonstration, 11 12 14

See Larmore, Patterns of Moral Complexity (Cambridge University Press 1987), e.g. 51. 13 See Chapter 5. See Chapter 8. The thought that underlies Robert Paul Wolff ’s Kantian anarchism: see his In Defense of Anarchism (New York: Harper & Row 1970).

44

Democratic toleration?

they also use coercion. Again the active reasons will concern the consequential benefits of force, and those who demonstrate are not tolerated. Land clearance policies and other attempts to purge public spaces of persons thought undesirable illustrate the point.15 Political questions arise when some body, such as local residents, objects to the presence of traveller encampments and calls on the state to intervene. So, whatever the authorities’ reasons justify, they justify withholding toleration. Both residents and travellers have at their disposal the resources for accusing the other of intolerance. This is true a fortiori when political power is mobilised to enforce either acceptance or prohibition of the disputed practice. Of course some liberals, and most non-liberals, reject the idea that the state must remain neutral between ideas of the good. What is then in view is a ‘perfectionist’, confessional, or ‘romantic-expressivist’ state instead.16 Such a state freely makes normative judgements in formulating public policy and deliberating executive actions. So it may, for instance, decide to tolerate cannabis use by failing to enforce laws against possession of the drug, or by decriminalising possession, but not the act of supplying it; the state may do this despite upholding the belief that it is better not to use cannabis. It may nonetheless strictly prohibit the sale and consumption of ‘hard’ drugs. In doing so, it acts in a way which is consonant with the structure of toleration. There is nothing incoherent in this picture, assuming the correctness of the basic analysis of toleration as reasons for disapproval held in check by principled reasons for self-restraint. The state – at least, the non-neutral state, or one in which non-neutral outcomes are delivered by neutral procedures17 – can adopt certain kinds of reason for nonintervention on its own account, and this qualifies the argument. In fact, a good deal of social policy seems to involve a structure of this sort: for example, aspects of the state’s regulation of pornography, speeding and some other road traffic offences, prostitution, cigarette smoking, reliance on state pensions, certain kinds of state-funded prophylaxis, gambling, and the availability of certain drugs, such as alcohol, and poisons. In such cases the state may be thought of as a first party in relation to a second who carries on an activity which the state perhaps disapproves of, or would not wish to become endemic in the population, such as gambling, and this seems a fair analogy with the interpersonal case.

15 16 17

See Jeremy Waldron, ‘Homelessness and the Issue of Freedom’, in Waldron, Liberal Rights: collected papers 1981–1991 (Cambridge University Press 1993), 309–38. ‘Perfectionist’ is Joseph Raz’s term. See his The Morality of Freedom (Oxford: Clarendon Press 1986). On the ‘romantic-expressivist’ state see Larmore, Patterns of Moral Complexity. See Chapter 5.

Structural transformation of toleration

45

However, the fact that certain practices occupy a legislative demi-monde between outright prohibition and straightforward acceptance does not show that the practices themselves are in fact being tolerated. As in the interpersonal case, we have to examine the antecedents of action. It is also necessary to be clear about the aspect under which politicians act when legislating or imposing policy. I will take these points in turn. Where the authorities disapprove of a practice, but decide despite this not to interfere with it, some reasons for the decision will not constitute it as an act of toleration. Often, given the clash between different lobbyor interest-groups over a practice, the concerns that guide policy may be more pragmatic. The most obvious type of situation is one in which those in power hold views qua private citizens about the practice, but are guided in their political actions by other considerations. This has certainly been true in the United Kingdom of past members of Conservative governments, for example, with regard to certain aspects of penal and social policy, such as the availability of abortion on the National Health Service or the abolition of capital punishment. Where the conflict is fierce, policy makers may be less concerned about their appraisal of the practice’s merits than the wider consequences of the courses of action open to them. What may then emerge is a familiar pattern of compromise in which both lobbies get some, but not all, of what they want. This is often true regardless of whether those who make the decision have conscientious views of their own about the practice. The abstract reasons for formulating such a compromise, or for coming down on one side or the other, may be less significant than the political pressures shaping the decision. These pressures include political considerations, more or less broadly construed, such as electoral advantage, party strategy, or the need to appease influential lobbies (such as newspaper editors). But they can also run in parallel with a bona fide concern with public goods. Prominent among these is public order, but those in power may also be concerned to protect the constitution, to nurture diplomatic or foreign trading links, to promote social engineering, and so on. Acts of toleration may be based on concern with public order, or other collective goods. But reasons to tolerate need both to be of a certain sort, and to play a certain role in deliberation. The relation between the agent, which in this case is corporate, and the policy differs here from the interpersonal case. As argued earlier, toleration is an executive virtue, and the tolerator’s deliberation reflects certain features of his agency, including other ethical dispositions, such as his dispositions to value. Thus the deliberation which issues in an act of toleration involves critical reflection on the deliberator’s own characteristics as an agent.

46

Democratic toleration?

In political toleration, deliberation addresses the actions of others, together with the valuations which accompany these actions. The very fact that the valuations mount politically pressing claims is what calls for a political decision in the first place, and the conflicts to which they give rise have to be negotiated politically. The valuations say on the one hand that the disputed practice should be permitted, and on the other that it should not. But these valuations need not duplicate the reasons for disapproving of the practice while nonetheless not intervening to stop it. The political conflict may be between two opposed but singleminded bodies of opinion. In such circumstances the reasons for tolerating are not those on which the defenders of the disputed practice characteristically mount their political claims. It is a significant fact about the political culture of toleration that there are also strong forces opposing the language of toleration itself. The political rhetoric of ‘zero tolerance’18 provides an expression of this way of thinking, which subverts toleration by appropriating its vocabulary of commendation.19 On the other side, toleration is rejected by those who, like Kant and Paine, demand more. To be tolerated requires disapproval, whereas the characteristic demand made by political lobbies is for equality, on the one side, and for suppression, on the other. Their claims are, for example, that this practice is demanded by the one true religion. In the circumstances of political toleration each side can brand the action the other demands as intolerable. The reasons which animate the protagonists in the conflict cannot, so to speak, simply be taken over by the political decision-makers. The reasons behind attempts at a political resolution will often differ from those which made the question politically pressing. The dual aspect of the legislator in representative democracies is telling: it provides a case where those directing law or policy may genuinely stand as one party to a second, as in the interpersonal case. But it is significant that this form of representative democracy marks the survival 18

19

‘Zero tolerance’ did not originate with the policing strategies implemented by Mayor Rudolph Giuliani in New York, but has been popularly associated with them. The term has become widespread in political rhetoric elsewhere. A random sampling of ‘zero tolerance’ websites turns up applications of the term to the following, among many others: violence against women, swearing at school, racist language, breaches of etiquette in bridge clubs, zero tolerance itself. For background to the debate over zero tolerance, see e.g. Roger H. Burke (ed.), Zero Tolerance Policing (Leicester: Perpetuity Press 1998); the influential ‘broken windows’ justification for zero tolerance appears in James Q. Wilson and George L. Kelling, ‘Broken Windows’, Atlantic Monthly 249 (1982), 29–38. In this regard the talk of ‘zero tolerance’ provides a case in point of the rhetorical transformations described by Quentin Skinner. See his Visions of Politics, Vol. I.

Toleration and democratic politics

47

of personal prerogative into the age of democratic politics: representatives take decisions as private individuals, e.g. on conscientious grounds, rather than being delegated by their electors. But in a pure direct or delegatory form of democracy, these interstices for private conscience no longer exist within the political process. Here, the possibility of toleration requires local derogations from democracy in political systems which are, on the whole, democratic.

Toleration and democratic politics As argued in Chapter 2, politics starts from the question What do we do? for a collectivity that may not be well defined before it is posed. The question clearly may concern toleration – usually, as in the early modern period, it is posed by one group, who have power, about another group, which does not. In such cases, as in the early modern period, the tolerators can exercise power over themselves to restrain their own disapproval of others. In an edict issued in 1568, the Transylvanian voivode János II Zsigmond Zápolya, a convert to Unitarianism, granted to congregations the freedom to choose a preacher of their own religious persuasion. This superseded a decree of 1557 that had given subjects religious liberty without restriction. The 1568 edict applied to Lutherans, Calvinists, Catholics and Unitarians; eastern Orthodox Christians, Jews and Muslims, although not given the guarantees granted to western Christian faiths, were also tolerated. János enacted the decree at the bidding of his court preacher, Ferenc Dávid, a Transylvanian Unitarian preacher.20 The Hungarian reformation was more eirenic than its western European counterparts, not least, perhaps, because of a long history of tolerating Greek Orthodoxy. It is not clear how far the edict enjoyed popular support, a notion of uncertain denotation at a time when most of the population played no role in making political decisions, and religious practice depended on the say-so of local magnates.21 Soon after his death the edict was rescinded by János’s successor Prince Báthory István.

20

21

István Lázár, Transylvania: a short history, tr. Thomas J. de Kornfeld (Safety Harbor, FL: Simon Publications 1996), 91–2. See also Péter Katalin, ‘Tolerance and Intolerance in Sixteenth Century Hungary’, in Ole Peter Grell and Bob Scribner (eds.), Tolerance and Intolerance in the European Reformation (Cambridge University Press 1996); and Alexander Sándor Unghváry, The Hungarian Protestant Reformation in the Sixteenth Century under the Ottoman Impact (Ann Arbor: University of Michigan Press 1989). Katalin, ‘Tolerance and Intolerance’, 255, describes typical practice as cuius patronatus, eius religio.

48

Democratic toleration?

Sometimes moves to relax restrictions on religious freedom met with a highly ambiguous response. In 1598, following the French wars of religion, Henri IV promulgated the Edict of Nantes, allowing freedom of worship to Huguenots. Even the registration of the edict provoked opposition; the Parlement at Rouen did not register it until 1609.22 Though the edict is often hailed as a landmark in the movement towards religious toleration, the decree was exploited by the Catholic clergy on the visitation commissions set up to administer the edict. In areas such as the Midi, where Protestantism had become dominant, the commissioners strove to reassert Roman Catholic rites, to retrieve assets, and to regain access to benefices of which the spread of the reformed religion had deprived them.23 More generally, the edict aimed to revitalise Catholicism across France – part of Henri’s efforts, as a convert to the Roman religion, to reassure his new co-religionists of his good faith.24 In areas where Protestants formed a minority, the episcopal visitation commission sought to exclude Huguenots from burial in ground consecrated by the Catholic church.25 Henri’s ‘perpetual and irrevocable’ edict, revoked by Louis XIV at Fontainebleau in 1685, caused an exodus of some 300,000 Huguenots, mainly to England.26 In Britain, the 1689 Toleration Act was passed by a reluctant Parliament at the instance of William of Orange, who indeed wished the act to go further and extend freedom of religion to Roman Catholics:27 as one historian has written, ‘what is clear is that the passage of the Toleration Act cannot be explained in terms of the general acceptance by either the laity or the Anglican clergy of the reasonableness of toleration as a political policy’.28 The same went for the declarations of indulgence issued by James II in 1687 and 1688, where the King was taken to be 22

23 24 25 26 27

28

Gabriel Audisio, ‘La réception de l’édit de Nantes en Provence (1598–1602)’, in Michel Grandjean and Bernard Roussel (eds.), Coexister dans l’intolérance: l’édit de Nantes (1598) (Geneva: Labor et Fides 1998), 268, 274. See also Jill Fehleison, Boundaries of Faith: Catholics and Protestants in the Diocese of Geneva (Kirksville, MO: Truman State University Press 2010), 102ff. Marc Venard, ‘L’Église catholique bénéficiaire de l’édit de Nantes: le témoignage des visites épiscopales’, in Grandjean and Roussel (eds.), Coexister dans l’intolérance, 291. Venard, ‘L’Église catholique bénéficiaire de l’édit de Nantes’, 292, notes that some clerical expropriations were made by Catholic landowners. Venard, ‘L’Église catholique bénéficiaire de l’édit de Nantes’, 294–5. Geoffrey Adams, The Huguenots and Public Opinion 1685–1787 (Waterloo, Ontario: Wilfrid Laurier Press 1991), 97. Jonathan Israel, ‘William III and Toleration’, in Ole Peter Grell, Jonathan I. Israel and Nicholas Tyacke (eds.), From Persecution to Tolerance (Oxford: Clarendon Press, 1991), 129–70, 154. Richard Ashcraft, ‘Latitudinarianism and Toleration: historical myth versus political history’, in Perez Zagorin, Richard Ashcraft and Richard W. F. Kroll (eds.), Philosophy, Science and Religion in England, 1640–1700 (Cambridge University Press 1992), 151.

Toleration and democratic politics

49

relaxing anti-Catholic laws as an expression of his own crypto-Popery. Earlier Anthony Ashley Cooper, Lord Shaftesbury, had argued in the House of Lords that ‘popery and slavery like two sisters go hand in hand’.29 Conversely, when James was deposed in 1688, his successor wished to enact more extensive toleration than Parliament was prepared to concede.30 The firebrand anti-Dissenter preaching of Henry Sacheverell stirred up popular feeling against ‘moderation’ into the reign of Queen Anne.31 William had strategic as well as conscientious reasons for including Catholics in a religious settlement.32 The fact that William happily presided over a much laxer regime in the 1690s than the Toleration Act envisaged helped to fuel popular High Tory revanchisme.33 In each of these cases, with the possible exception of the Edict of Nantes, rulers tried to push through a more permissive religious regime than the politically effective population wished to accept. Then as now, this questions the political legitimacy of the move towards toleration. However, the terms of trade between political decision-makers and the people have shifted as a result of the move to mass democracy. Now rulers wield power as representatives of the people, viewed as the democratic sovereign. This creates two parties, or at least aspects, of political agency, namely the represented and its representative. A notable feature of Hobbes’s account of sovereignty is that it reduces the interests of the former almost entirely to the expressed political will of the latter. On most democratic views, however, public policy sometimes runs afoul of sectional interests. That may be because the authorities take a view that runs directly counter to some specific interest-group: for instance, career criminals. Or, as in neutralist theory and often in political practice, the representative finds that different interest-groups oppose one another, and has to decide what if anything to do. Suppose that there is some practice, and that one group supports it, while another opposes it. Of course, the latter may choose to suffer in silence, or adopt low-level methods of expressing its disapproval, to which the first group takes no exception. Then, it seems, there is no 29

30

31 32 33

Shaftesbury, speech to the House of Lords, 24 March 1679; quoted in Andrew R. Murphy, Conscience and Community, 139. On the wider point, see Murphy, Conscience and Community, 135ff.; for the extent of Anglican clerical opposition to James, see John Spurr, ‘The Church of England, Comprehension, and the Toleration Act of 1689’. English Historical Review 104 (1989), 927–46. Hugh Trevor-Roper, ‘Toleration and Religion after 1688’, in Grell, Israel and Tyacke (eds.), From Persecution to Tolerance, 391. Cf. Israel, ‘William III and Toleration’ in the same volume, e.g. 139, 144. Trevor-Roper, ‘Toleration and Religion after 1688’, 393. Israel, ‘William III and Toleration’, 151ff. Israel, ‘William III and Toleration’, 162–3.

50

Democratic toleration?

political question to address: the two groups have arrived at a peaceful, if not wholly satisfactory, modus vivendi. But now suppose that matters go further than this, and the members of the second group begin to call for action against the first, unless the latter desists from the practice. These may initially be demands for political action, but may turn to incitement or direct violent action. If conflict develops, each side is proposing to do something of which the other disapproves.34 Of course, the authorities can let matters lie like this, as long as public order is not under threat, though there is no obvious reason why this course of action is more tolerant than any other. But in any case, this is not really the point. The issue has already become a political one by this stage, and those in power have to respond to first-order expressions of disapproval. In the circumstances of toleration, democratic politicians have to respond to those who argue, not that the practice should be permitted despite their own disapproval, but that the practice should be suppressed because they disapprove of it. Neither side may be inclined to tolerate the other’s behaviour in spite of this disapproval: if they were, the members of the first group might have been prepared to discontinue or modify the practice to meet the concerns of the second, while members of the latter might have thought that the practice should be permitted despite their own feelings of disapproval. Very often in conflicts, one party’s disapproval of another leads the latter to disapprove of this very disapproval. This is not very conducive to a peaceful resolution of the conflict. Dispute is liable to arise where one party sees toleration as the path of submission and as a result asserts what it takes to be its entitlements – a move which unsurprisingly leads the other to assert its own entitlements. It is an important, if obvious, empirical fact about political conflicts in democratic polities that they simplify the structure of single-person deliberative conflicts in certain respects.35 We can bring out this point by considering the claims made by the groups involved on the politicians who are accountable to them. In conflicts like the one described, it is not tolerant dispositions which are engaged. Neither party calls for a certain outcome (toleration) despite its disapproval of the other group; rather, each group expresses itself by demanding that what it disapproves of should be suppressed. Insofar as the confrontation is liable to come down to logically incompatible demands – that the practice be prohibited, and that it be permitted – each course of action is prone to excite charges 34 35

Here I adapt remarks from my Virtue, Reason and Toleration, 161. This is, incidentally, a reason for doubting the reducibility of public to private deliberation, or vice versa, as argued in Hampshire, Justice Is Conflict, 11ff.

Toleration and democratic politics

51

of intolerance. In either case, the course of action favoured will be one which the other side has branded as intolerant. The political authorities are caught in the middle of the dispute, and lie open to the charge of tolerating what is really intolerable whichever side they favour. The problem of ‘tolerating the intolerant’, often treated as an anomalous or marginal case in theories of toleration, such as that of A Theory of Justice,36 is often not a marginal one, but arises at the very point when toleration becomes a political issue.37 The structure of conflict in these situations is symmetrical. The courses of action determined upon by the two groups are incompatible. Moreover, each group strongly objects to the other’s (actual or projected) action, and is apt to brand the other side as intolerant, while calling for toleration for its own favoured course of action. We can note here in passing the need for a third type of reason in analysing toleration, since in distinguishing its own actions from those of the opposing side, each group has to be able to contrast disapproval (which can be overridden) from judgements that certain practices are intolerable. That is, there must be reasons justifying intolerance, as well as for toleration in the face of disapproval.38 And each side, ex hypothesi, lacks the means to enforce its views. The symmetrical structure limits the practical options available to those in power. I mean this not only in the sense that there are certain actions which may not be open to them (or, what comes to the same, may not be possible without unacceptable cost); it limits the descriptions, and hence justifications, available to the political decision-makers. Few, apart from some anarchists, are likely to deny that there are legitimate occasions for intolerance. If there are such occasions, then the only question is when it is legitimate to act intolerantly – when, in other words, to act against practices which really are intolerable. Someone may object that the putative symmetry in the political circumstances of toleration has been exaggerated. Surely there may be circumstances of toleration in which there is asymmetry. For example, if someone objects to my reading a book, that person may be charged with acting intolerantly. But it sounds implausible to say that if I go on reading, I balance up the symmetrical structure by also acting intolerantly: I am just sitting there and reading. So in cases like this one, at least, the symmetrical structure fails to apply. Reading a book is neither 36

37 38

See John Rawls, Theory of Justice, §35: ‘Toleration of the Intolerant’, 216–21. It is a marginal case for Rawls both because the intolerant are thought to lack dispositions characteristic of most citizens in the well-ordered society, and because they are thought to justify derogations from the equal liberty principle (217–18). See also my Virtue, Reason and Toleration, ch. 5. See my Virtue, Reason and Toleration, ch. 1.

52

Democratic toleration?

tolerant nor intolerant. By contrast, someone who tries to interfere with my reading may act intolerantly. The mark of doing so, on one view, lies in attempting to interfere with or curtail another’s agency.39 One problem in this notion lies in the fact that many intolerant acts seem to have purely symbolic content, and in some cases there is a choice to be made between curtailing the agency of the intolerant, and that of their victims. The notion of curtailment has normative content, and readings of this content, for the purposes of judging what is unacceptably intolerant, call for justification. A further response is to try to understand why the objection seems plausible. What action could be less objectionable than somebody’s reading a book? Certainly, one can imagine situations where the ‘mere’ act of reading might be objectionable. (‘What’s the matter with you? I’m just sitting quietly in this synagogue, reading Mein Kampf.’) Usually, of course, this private activity goes ahead without exciting objections. But there are exceptions that may raise the question of toleration. One example is the legal regulation of public, and indeed sometimes private, consumption of pictorial pornography; again, religious sects such as the Roman Catholic church have long tried to limit their adherents’ access to heterodox texts; alarm about terrorism has also made not merely the reading, but possession, of certain texts legally actionable.40 The mistake, however, is to place faith in a naturalism of acts that tries to mark off the intolerant ones from the rest, before considering how far a given act attracts political dispute. None of this entails relativism: nothing that has been said is inconsistent with the view that some value or values are the right or true ones. All that has been said is that there is a vantage point from which one can see the protagonists’ dispositions as symmetrical. Perhaps, when the circumstantial information is filled in, it seems that the person who wants to interfere with my reading is unjustified. But this does not mean that my reading the book must therefore not be intolerant. Justified intolerance is not to be identified with not behaving intolerantly. Maybe this is just a case of justified intolerance. It is certainly the case that others

39

40

For this suggestion, see Jones, ‘Making Sense of Political Toleration’. The curtailment idea has difficulties with purely expressive acts of intolerance, as I argue in ‘Political Toleration: a reply to Jones’; see also Jones’s ‘Political Toleration: a reply to Newey’. I first considered this example in ‘Is Democratic Toleration a Rubber Duck?’, Res Publica 7:3 (2001), 315–36 at 323. My remarks have been overtaken by events: compare the arrests under the Terrorism Act 2000 at Nottingham University of Hicham Yezza and Rizwaan Sabir, an employee and student of the university respectively, for possessing computer files of ‘The Al Qaeda Training Manual’ in 2008. Both were eventually released without charge.

Toleration and democratic politics

53

may object to reading material – for example, on the grounds of ‘sedition’. I am certainly likely to disapprove of this person’s attempts to interfere with my reading, and think of my attitude as justified. Not to be intolerant of one’s own dispositions is just part of what it is to have an orientation to value. But there is always in principle a question about the relation that it bears to others, and particularly to their responses to actions of mine that express it. As regards the symmetry claim more generally, I now accept that there are cases when describing the parties as symmetrically situated is strained, and the wider point is better made in another way.41 This is by drawing attention to the difference between theoretical models of toleration, and the circumstances in which toleration becomes politically contentious, which I gave in Chapter 2. Part of what is owed in the circumstances of toleration is an account of why the reader is behaving in this way, and what is held to be objectionable about it. Clearly, gratuitous objections may be made to reading, or any other activity. But that tells us little about cases where the objection is not gratuitous, and in political disputes the putative gratuitousness is no longer the point. The situation is one of normative conflict, and there is something self-defeating in attempts to resolve it by asserting a further norm. Either that norm itself will be controversial between the parties in the abstract, or they will dispute its priority, or its application to the case at hand. In this setting, appeals to supposedly naturalistic norms such as those of agency, far from putting paid to conflict, tend to perpetuate it. Toleration becomes practically urgent when individuals form conflicting normative judgements. In deciding which judgements may be justifiable, one may cite ‘negative’ liberal values which hold that the sole legitimate basis for interference itself lies in interference. But merely pointing to a certain value will not address the political problems which arise when that value confronts something alien to it. It is easy enough to describe a situation in which one can make no sense of the objector, and then pronounce that there is nothing here to object to. By contrast, in filling out the real details of practical conflict, one may come to appreciate why the objector is acting like this even if the values concerned are quite alien ones. In moving beyond one’s own norms, all that remains is the fact of confrontation, and a description of why, given the value-orientations of the protagonists, it has arisen. So, whether or not political conflicts over toleration invariably display symmetry, the basic structure of toleration, as reasons for disapproval 41

I am grateful to John Horton and Peter Jones for discussions on this point. See Jones, ‘Making Sense of Political Toleration’.

54

Democratic toleration?

checked by reasons for self-restraint, allows for indefinite applicability to situations where one person’s conduct excites another’s opprobrium. In this sense, these situations always make for symmetrical conflict. Attempts to ground asymmetry in general features of agency fail, not least because a person can be intolerant either through an action of their own, or by impeding the actions of someone else.42 If every perspective on agency makes normative commitments, then for each person there will be cases warranting ‘zero tolerance’. But to acknowledge this does not by itself transcend politics: rather it expresses one of its conditions. This is what makes politics, in a certain sense, transcendental: like any ideology, particular systems of value aim at an encompassing worldview, while the alienness of values makes the scope for clashes between perspectives inexhaustible. To point beyond the pale is not to point beyond politics, but to its main arena. Toleration, politics and political philosophy Modern philosophical liberalism seeks a basis for agreement on fundamental political institutions and procedures in general normative considerations. The interpretation of these considerations varies, of course, from one theory to the next. The mention of a ‘basis for agreement’ is not intended to imply that contractarianism is the sole or best version of modern liberalism, but it does make explicit a move typical of liberal theory in general: it aims to argue that the terms of the agreement are such that persons should accept them, whether or not they in fact would do so. It is a further question whether agreement itself is taken as basic, or whether a disposition to agree is taken to rest on some more basic kind of consideration, such as a natural duty. Toleration exposes a blind spot in this approach to justification. The approach tends to assume that there is some right answer to the question of whether or not a practice should be tolerated, and a political authority acts tolerantly if and only if its action in respect of a practice conforms to this answer, to which reasonable people would agree. It adds that the practice of toleration is justified only if it is justified by an agreement to which those involved (would, or should) have consented.43 As the parentheses suggest, there are problems in formulating this notion of 42 43

For this reason, Jones’s attempt to import asymmetry via a general distinction between doers (‘agents’) and sufferers (‘patients’) of action seems to me unpersuasive. As I argue in Virtue, Reason and Toleration, esp. ch. 5, philosophical argument about toleration’s structure cannot provide a definitive account of what should be tolerated. For related reasons, it is also very difficult to furnish an unambiguous content for toleration from within liberalism.

Toleration, politics and political philosophy

55

justification while keeping the intuitive appeal of an agreement-based approach. Without actual agreement, the mooted justification will slip into talk about the agreements by which the parties, were they rational or moral, would or should have bound themselves. Then the ‘agreement’ begins to look idle, since what then matters is that the parties ought to agree anyway. But once the political circumstances of toleration have been reached, what the protagonists ought to do is part of what is in dispute.44 If so, the approach risks ignoring or misconstruing the facts which made toleration politically urgent to start with. For the political circumstances of toleration often arise precisely when action is either unstructured by obligations, or the obligations themselves are disputed. There is in fact a further possibility: that one party fails to act on an obligation which they have already accepted. Then, something along the lines of the ‘normative agreement’ approach may be appropriate. But this situation is rather rare, and the reason why is fairly obvious. Either the party climbs down – accepts that this is the case – and the political impasse is cleared; or it does not because it denies that this action is required by the obligation. We can now see the underlying truth in the remarks of Kant and Paine quoted earlier. The practice of toleration essentially requires the use of power, since the political circumstances only arise when one party or the other, or both, have failed to act tolerantly, and so have to be compelled to adopt a different form of action. This is true even in the ‘free vote’ case discussed earlier, where legislators seen under one aspect are in a position analogous to that of interpersonal toleration. For even here the position is at best analogous. Using powers of restraint over one’s own agency is different from using political power to control the actions of others. Power is mobilised to enforce public policy against those who oppose it. There is a further problem with the normative agreement approach, even given its commitment to idealisation. For Rawls, for instance, political action must be orientated towards agreement between persons who see each other and themselves as free and equal.45 However, what gives rise to political disputes, including ones over toleration, will presumably be a form of failure, whether of reason or motivation, by 44

45

Not that this need doom attempts to provide an agreement-based justification for morality itself. There are, in any case, different conceptions of what such a justification should aim to do. One is simply that it should replicate the content of morality. Another is to tell us what the bindingness of morality consists in, i.e. to provide an answer to the question, ‘Why be moral?’ Yet another is to provide a perspicuous procedural model for alleged formal properties of morality (e.g. its applying categorically) which would otherwise be obscure. Rawls, Theory of Justice, 19, 505ff.; Political Liberalism, 19; also his ‘Fairness to Goodness’ and ‘Justice As Fairness: political not metaphysical’, both in Rawls, Collected Papers, at 278 and 397 respectively; see also Chapter 4.

56

Democratic toleration?

those involved in them: the theory prescribes a certain form of action (e.g. conformity to the principle of maximum equal liberty for all) and the political problem arises from some person’s or group’s failure to adhere to it.46 As I argued in Chapter 2, the situation becomes political when one party or another fails to recognise, or act as if it recognised, that there is reason to agree. Seen like this, describing the attempt to derive such principles as the ‘political’ branch of philosophy involves a category mistake.47 To see the limits of the normative agreement approach, as undergirded by the notion of reasonableness, we could take the case of Paul Donnachie, who was convicted in August 2011 of racially aggravated breach of the peace. His defence accepted in court that during the incident, which occurred on 12 March 2011 in a student hall of residence at St Andrews University, Donnachie, then a history student at the university and a member of the Scottish Palestine Solidarity Campaign (SPSC), had plucked one of his own pubic hairs before wiping it on an Israeli flag owned by Chanan Reitblat, a Jewish fellow student. Reitblat was on a year’s exchange visit to the university from New York; the flag was a gift from Reitblat’s brother, a member of the Israeli Defence Force. The incident occurred at around 1.30 a.m., after Donnachie had been drinking. Reitblat testified that after the episode he was unable to eat or sleep, and he ended his studies at St Andrews early. Donnachie was sentenced to 150 hours of unpaid community work and fined £300 by Cupar Sheriff Court; the university also expelled him. After the case, a spokesman for St Andrews University said that it had ‘a long tradition of tolerance, respect and the right to freedom of expression, but [it] was also a community which “abhors racial intolerance”’.48 In the background lay the political conflict in Palestine. Donnachie enjoyed strong support during the court proceedings from the SPSC,49 while groups such as the Scottish Friends of Israel backed Reitblat and condemned both Donnachie and the SPSC.50 Donnachie’s supporters spattered an Israeli flag with red paint, and some Jewish activists such as 46

47

48 49 50

The principles or prescriptions may operate at a very high level of generality. How much room this leaves for disagreement varies directly with the level of generality of the principles, and so with the latter’s ability to attract general agreement. Rawls’s description of his later theory as ‘political liberalism’ is a typical example, since it deals with the basis for agreement among the contractors, and even this of course has to draw not just on what they will agree to, but what they would agree to under idealised conditions. BBC news online, 23 August 2011: www.bbc.co.uk/news/uk-scotland-edinburgh-eastfife-14638515. Marcus Dysch, ‘Anti-Zionists Attack Scots Sheriff’, Jewish Chronicle, 1 September 2011. www.scottishfriendsofisrael.org/paul_donnachie.html.

Toleration, politics and political philosophy

57

Leehee Rothschild, a member of the group Boycott from Within, endorsed his right to criticise Israel. Donnachie also won backing from other Jewish groups, such as Scottish Jews for a Just Peace (SJJP)51 and the anti-Zionist Jewish Orthodox group Neturei Karta. Rabbi Aaron Cohen, a spokesman for Neturei Karta, said that Zionism and Judaism are two opposite and incompatible concepts. Zionism with its aim of a state for the benefit of a specific group in someone else’s land is by definition a sectarian and racist philosophy. For the Sheriff to mistakenly conflate Jewishness with being a ‘member of Israel’ is simply wrong, and this conviction is clearly ridiculous.52

The SJJP also complained that the Sheriff hearing the case had failed to call Jewish witnesses for the defence.53 It is not hard to see why Donnachie’s actions attracted charges of intolerance. Donnachie accepted that he knew Reitblat was Jewish, and his actions, on any reasonable interpretation, were calculated to defile the Israeli national flag. This might be construed as a crime against property, though he was not charged with an offence of this kind. The question before the court was whether the act of defiling the Israeli flag, which Donnachie did not deny, constituted a racially aggravated breach of the peace. An important element of the defence case was that a distinction should be drawn between objects symbolising the state of Israel, and symbols of Jewish ethnicity or religiosity. In its statement regretting the Sheriff’s failure to call Jewish defence witnesses, the SJJP expressed concern that the proceedings had ignored ‘the crucial distinction between Zionism – a political position of support for a Jewish state in Israel/Palestine – and Jewishness – a person’s religion or ethnicity’. The question here is not whether Donnachie’s actions were reasonable in themselves. Most would probably agree that they were drunken and boorish. The question before the court was whether the defiling of the Israeli flag was racially aggravated because it constituted an attack on Reitblat’s Jewishness and so constituted an anti-Semitic act. On this question room certainly exists for disagreement, as we have seen, and it is hard to know on what basis one party to it could be shown to be reasonable and the other unreasonable. Indeed, it is a staple bone of contention between defenders and critics of Israel whether criticism of Israel constitutes anti-Semitism in view of the fact that the state is an avowedly Jewish one. This becomes harder still if one endorses, as a 51 52 53

Scottish Jews for a Just Peace (SJJP), statement on the Donnachie case, 4 September 2011. Dysch, ‘Anti-Zionists Attack Scots Sheriff’. SJJP, statement on the Donnachie case.

58

Democratic toleration?

general position, the claim that people may reasonably disagree about ideas of the good, controversial questions of morality, politics, and so on. Yet the incident and ensuing court case surely involve toleration. The most obvious point is that whether Donnachie’s action counts as racial harassment is one on which opinions may differ. It is not even clear on what basis the judgement should be made. For instance, should Donnachie’s intentions count as decisive? But one may disagree about what those intentions were: for instance, whether those intentions are determined by Donnachie’s beliefs. Should a person who defiled, say, the Egyptian flag, falsely believing it to be the Israeli one, be convicted of anti-Semitic racial harassment on this basis? If Donnachie’s own beliefs play a key part in recovering his intentions, his professed belief in the distinctness of the Israeli state and the Jewish people might suggest lack of anti-Semitic intent. Of course, as far as this goes, the question is not what it would be reasonable for Donnachie to believe, but what he did believe – or, in the absence of firm evidence about that, a matter made more uncertain by his drunken state at the time – what it would be reasonable to believe that he believed. Again, one might try to represent this as a strict-liability offence, but this does not by itself decide how Donnachie’s act should be interpreted. Nor does it help to move the discussion to what an impartial observer, rather than Donnachie, might regard as reasonable. Symbols require interpretation, and reasonable people often disagree about that. Whatever one may think about the thesis of reasonable disagreement,54 it is hard to avoid the conclusion that this case exemplifies the very disagreement that Rawls’s political construction aims to overcome. The case also exemplifies the structure of modern conflicts over toleration. Each side claimed that its own actions were legitimate and the other’s illegitimate. These could readily be phrased as claims that their actions were tolerable, while the other was behaving intolerantly and intolerably. Clearly, neither Donnachie nor Reitblat was urging that the other party should be allowed to get its way despite his own reservations. Donnachie objected to Reitblat’s recourse to law, while Reitblat relied on the force of law to bring Donnachie to account. Meanwhile the state, in the person of Fife’s Sheriff, pronounced on the merits of these claims. But, in finding Donnachie guilty, the authorities did not make him act tolerantly: rather they judged his action intolerable. It acts as a third party, coercing one of the two parties to the conflict. In judging Donnachie’s actions intolerant, it draws a limit to what it, the state, itself

54

See chapters 4 and 5.

Conclusion

59

regards as tolerable.55 This is true regardless of whether one agrees with the judgement, or with the basis on which it was reached. Conclusion Symbols matter politically, in part, because of their ambiguity and openness to interpretation. Ambiguities of this sort set limits to how far politics, including the politics of toleration, can be circumscribed by notions of the reasonable. In this chapter I have argued that the move from two-party to three-party agency in today’s politics has transformed the interpersonal structures that marked the politics of toleration in the early modern period. Personal, prerogative or discretionary reasons for toleration give way to a politics of the public interest. Since politics tries to answer the question What do we do? its conditions are those of joint action. While it need not always require the use of power – it may not, for instance, if unanimity reigns56 – joint action often does require it. This remains true whether or not public justification is held subject to constraints on its content, as public-reason theorists argue.57 Modern political pressures push in the direction of intolerance, justifiable or otherwise, rather than toleration. They open to question how far toleration can, in these conditions, remain a libertarian value.

55 56 57

Though, as argued in Chapter 1, intolerance does not entail intolerability. It still does not follow that coercion is not required even in the face of unanimity, since people sometimes do what they have earlier agreed not to do, or the converse. E.g. Rawls, ‘The Idea of Public Reason Revisited’, repr. in Rawls, Collected Papers; also Gerald Gaus, The Order of Public Reason: a theory of freedom and morality in a diverse and bounded world (Cambridge University Press 2011); for an overview see also Gaus, Contemporary Theories of Liberalism: public reason as a post-Enlightenment project (London: Sage 2003).

4

Toleration as sedition

Although toleration is politically embattled, it refuses to go away. The rhetoric of ‘zero tolerance’1 testifies to its beleaguered state, while the rhetoric, despite itself, invokes the concept of toleration. The talk, after all, could merely be of not allowing these nefarious practices, but the rhetoric inverts toleration by pitting the concept against itself. It puts the political value of toleration to the service of an authoritarian tendency, and tends to give toleration itself a bad name. But toleration faces other challenges, not least from those who think of themselves as its friends. As this chapter will argue, the cause of retrieving toleration as a political value is not helped much by some recent liberal justificatory strategies – specifically, those which begin from the fact of moral or ideological disagreement between individuals, and try to turn this fact to toleration’s advantage. I shall suggest that private judgements are less congenial to the political conditions needed for toleration than modern liberals often think.

Reasonable pluralism Undeniably, the call for toleration arises from the fact that people have different evaluative beliefs and practices. Often the differences amount not just to divergence, but to disagreement, whereby a conception of the good is also a conception of the badness of others’ conceptions of the good. Accordingly, recent liberal attempts to provide foundations for toleration start from the idea that there are many such conceptions. They pursue the wider project of political design, which structures political processes and institutions from general normative considerations such as justice. On one widely favoured liberal justificatory strategy, none of the conceptions can be shown to be more reasonable than each of the others. 1

See also Chapter 1.

60

Reasonable pluralism

61

In Political Liberalism, John Rawls formulates ‘reasonable pluralism’ thus: the society in question [i.e. the one for which principles of political justice are to be decided] is one in which there is diversity of comprehensive doctrines, all perfectly reasonable. This is the fact of reasonable pluralism, as opposed to the fact of pluralism as such.2

Reasonable pluralism results from free institutions: [i]t is the fact that free institutions tend to generate not simply a variety of doctrines and views, as one might expect from people’s various interests and their tendency to focus on narrow points of view. Rather it is the fact that among the views that develop are a diversity of reasonable comprehensive doctrines (that is, ideals of the good life, or worldviews, as I shall call them).3

Rawls thinks that reasonable pluralism results from the ‘burdens of judgement’. These burdens arise from the complexity of evidence, the vagueness of key concepts used in articulating and justifying the component beliefs, the differential weighting of relevant factors, cultural influences on belief formation, and conflicts between values.4 For Rawls, the burdens of judgement show why reasonable persons can disagree, because reason is compatible with different worldviews. Nonetheless, there is disagreement – otherwise one could hold several of these views simultaneously. Rawls concludes that reasonable pluralism shows that stability is impossible if the basic structure rests on any single worldview: the idea of a well-ordered society of justice as fairness [set out in Rawls’s earlier A Theory of Justice] is unrealistic. This is because it is inconsistent with realizing its own principles under the best of foreseeable conditions. The account of the stability of a well-ordered society is therefore also unrealistic and must be recast.5

The burdens of judgement face both ways. On the one hand, they make pluralism ‘reasonable’ at the collective level: they show why no single doctrine can win a decisive victory over the others. On the other hand, Rawls represents individuals as holding particular worldviews despite the burdens of judgement. At the private level, each person thinks that the 2

3 4 5

Rawls, Political Liberalism, 24 fn. 27. On the labile role of the ‘reasonable’ in Political Liberalism, see Leif Wenar, ‘Political Liberalism: an internal critique’, Ethics 106:1 (1995), 32–62. Rawls, Political Liberalism, 36. See also ‘The Idea of Public Reason Revisited’, repr. in Rawls, Collected Papers, 573. Rawls, Political Liberalism, 58–9. Rawls, Political Liberalism, xix. See also my ‘Rawlsian Liberalism at the Limits of Intolerance’, in Shaun P. Young (ed.), The Legacy of John Rawls (Aldershot: Ashgate 2009).

62

Toleration as sedition

burdens, which explain reasonable pluralism, do not apply to her own worldview. A person can believe both that some of her beliefs are wrong, and believe, with regard to each specific belief, that the belief is true. As I shall argue, however, to believe, of a specific belief or set of beliefs such as a worldview, that it is contradicted by other beliefs no less reasonable than it, means that one no longer has sufficient reason to continue holding the belief. So one cannot go the whole epistemic hog at the private level, on pain of abandoning the very beliefs that need to be reconciled by the public political project.

Burdensome judgements The need for the burdens of judgement to face in two directions at once poses problems. Do the burdens show that reason under-determines belief in the worldviews? Suppose the answer is ‘Yes’. If so, reason alone is insufficient to judge which worldview one should believe, given that numerous doctrines are ‘perfectly reasonable’. While reason may exclude some worldviews, it will leave others still in play.6 Yet people do end up holding worldviews. The burdens of judgement are supposed to explain how this arises under free institutions. They explain how one person can reasonably affirm a doctrine, while another can reasonably deny it. But then, insofar as the burdens do explain this situation, it seems that they can only constitute an error theory – an explanation of why people believe more than reason warrants. But then how can such beliefs be called ‘reasonable’? This suggests that the answer to the question must be ‘No’: the burdens of judgement do not show that reason under-determines belief in the worldviews. But this throws up a puzzle. Suppose that reason tells one to believe p, the belief or beliefs constitutive of some particular doctrine. But there must be some p of which reasonable pluralism says that belief in not-p is reasonable too; otherwise there is no disagreement, reasonable or otherwise. This entails that whatever makes a belief reasonable does not necessarily make its negation unreasonable. But reasonableness is surely closed under entailment: in particular, to hold a belief reasonably is reasonably to hold a belief in the falsity of its negation. If the belief that p (say, a belief in theism) is reasonable, then it seems the belief, which it entails, that the belief that not-p (say, atheism) should be rejected is reasonable. To say, then, that it is 6

Rawls has little to say about the unreasonable, but it is exemplified by those unwilling to accept the burdens of judgement and who fail to abide by standards of fair cooperation; cf. e.g. his Political Liberalism, 52.

Burdensome judgements

63

reasonable to believe that p, and that it is reasonable to believe that not-p, seems to be no less than to say that reason cannot decide whether to believe one or the other. So the answer to our general question seems to be ‘Yes’ after all: reason does under-determine belief in the worldviews. The consequence of the burdens of judgement, on this showing, is less aptly described as ‘reasonable pluralism’ than as ‘reasonable scepticism’.7 If reason cannot decide between the conflicting doctrines, then there are insufficient grounds for believing in any one doctrine rather than any of the others. If the explanation for the disagreement is traced to the burdens of judgement, the explanation seems to undermine itself. Those who are in ‘disagreement’, if presented with the truth about the ‘burdens’ should agree that these showed that they would be unreasonable not to agree: unreasonable, that is, not to agree that the burdens show that nobody’s worldview can be shown to be more reasonable than everyone else’s. Each person in ‘disagreement’ has to accept that this applies to him, that his belief in the doctrine is no longer supported by sufficient reasons. This amounts to saying that reason requires that he abandon or qualify that belief.8 In other words, to remain coherent, reasonable pluralism must be an esoteric doctrine. But that breaches the requirement that political design be publicly justified.9 Belief in reasonable disagreement sometimes rests on the idea that persons constitute distinct subjects of experience and, to this extent, distinct loci of evidence.10 Your sensations, perspectives, et cetera are not mine, and conversely. Insofar as reasonable belief is evidentially based, and perception provides one source of evidence (for example, about the external world), the evidence available to you may ground your reasonable belief that p, whereas the evidence at my disposal reasonably leads me to believe that not-p. This, surely, is a familiar enough situation. The contrast in our dispositions to believe reflects the fact that our epistemic situations are distinct. Since reason can work 7 8

9

10

Reason is minimally characterised here as adherence to the principle of non-contradiction. For the claim that disagreement is traceable to the free operation of reason itself, see e.g. Joshua Cohen, ‘Moral Pluralism and Political Consensus’, in David Copp, Jean Hampton and John Roemer (eds.), The Idea of Democracy (Cambridge University Press 1993). Rawls, ‘Kantian Constructivism in Moral Theory’; ‘The Idea of Public Reason Revisited’; cf. ‘The Priority of Right and Ideas of the Good’; all repr. in Rawls, Collected Papers. E.g. Rawls cites among the ‘burdens of judgment’ the fact that ‘citizens’ total experiences are disparate enough for their judgments to diverge . . . on many if not most cases of any significant complexity’ (Political Liberalism, 57); cf. ‘The Domain of the Political and Overlapping Consensus’, in Rawls, Collected Papers, 477.

64

Toleration as sedition

only on the evidence at its disposal, it may be said, perspectival differences can ground differences in reasonable belief. However, in conditions of confrontation the extent of our ‘disagreement’ varies inversely with our distinctness as loci of experience. This itself imposes a limit on the phenomena of confrontation. You do not disagree with me if you find the taste of anchovy pleasant and I find it revolting. De gustibus non est disputandum, and the reason why tastes are not to be disputed is because, given that there is no evidentiary locus that could serve to put our different responses to trial, the two judgements of taste can be conjoined without contradiction. Conversely, when different reactions can be coherently explained, there is no disagreement. To use one of the ancient sceptics’ examples: if your hand has been immersed for a while in hot water, and mine in cold, lukewarm water will feel cold to you and hot to me. Indeed, one person could achieve this effect alone. Does she ‘disagree’ with herself? Disagreement does not subsist if reason can explain adequately, as with the tepid water, why perceptions can differ. Emphasis on persons as distinct loci of evidence also locks in the phenomena, as it were, at the pre-confrontation stage. Of course one may say that when one person infers that p on the strength of evidence available to her, while another infers that not-p given the evidence he has, they are in disagreement, even if they remain unaware of each other’s existence. But if so the confrontation between them is at most notional. Once it becomes real, a process of negotiating disagreement begins. That may lead to consensus, when one party concludes that his belief cannot stand in the light of new evidence. But it may lead to a hardening of stances whereby each side insists on its own view. If a confrontation arises that may engender political problems, however, it is no longer enough simply to say that because each side had reasons for its views in the pre-political phase, each side must remain reasonable once that phase is past. Confrontation demands an engagement of reasons. When the disagreement persists after confrontation, it is always open to doubt whether those involved continue to disagree because of, or despite, a shared commitment to reason. Sometimes Rawls seems to think that ‘reasonable’ is predicable of believers, rather than beliefs. A normative conception of reasonableness frames the contract.11 Citizens are reasonable when, viewing one another as free and equal in a system of social cooperation over generations, they are prepared to offer one another fair terms of cooperation according to what they consider the most reasonable

11

Rawls, ‘The Idea of Public Reason Revisited’, in Rawls, Collected Papers.

Burdensome judgements

65

conception of political justice; and when they agree to act on those terms, even at the cost of their own interests in particular situations, provided that other citizens also accept those terms.12

Rawls seems to accept a relativised notion of reasonableness, since the latter depends on what the advocates of particular ‘comprehensive’ ideas of the good regard as reasonable. But Rawls was not endorsing out-andout relativisation, as the reference to the need on the part of all citizens to regard one another ‘as free and equal’ indicates. He perhaps believed that overlapping consensus ultimately effaces this difference, since holders of worldviews can each endorse the basis of political agreement from their own perspective. But there is no good reason to think that reasonable worldviews will converge on any one set of principles. The danger is that the reasonableness of those holding them will be held to depend on their assent to the principles that the consensus is meant to justify.13 A person’s response to the burdens of judgement determines whether she is reasonable: to be reasonable is to recognise that one’s own worldview cannot justify the use of coercion in its own name. Reasonableness in this sense is tied to accepting one of the key features of political liberalism itself, or it is derived from dispositional scepticism towards one’s own first-order beliefs. But if this is the foundational sense of ‘reasonable’, we end up with a circle: political liberalism is justified because reasonable persons endorse it, but their endorsement of political liberalism is what makes persons reasonable. Rawls moves to talking of reasonable believers, rather than reasonable beliefs, when he argues that support for the doctrine of liberty of conscience can be derived directly from the burdens of judgement. Suppose you hold a reasonable worldview. This is held to require, for reasons examined more closely in the next section, that the view cannot be suppressed, but equally cannot demand the suppression of other reasonable worldviews. Reasonable persons see that the burdens of judgement set limits on what can reasonably be justified to others, and so they endorse some form of liberty of conscience and freedom of thought. It is unreasonable for us to use political power, should we possess it, or share it with others, to repress comprehensive views that are not unreasonable . . . [C]itizens as free and equal have an equal share in the corporate political and coercive power of society . . . [The authority

12 13

Rawls, ‘The Idea of Public Reason Revisited’, 578. Similar remarks apply, e.g., to Brian Barry’s contractualism, which attempts to produce agreement on terms ‘which nobody can reasonably reject’. See Barry, Justice as Impartiality (Oxford University Press 1995). Barry explicitly acknowledges at 8 that this relies on a normative conception of what is reasonable.

66

Toleration as sedition

to decide constitutional essentials on the basis of particular worldviews] is without grounds in public reason. What would be proposed instead is a form of toleration.14

What is it that is thought to justify the ‘so’ in the first sentence quoted? Its ground is distinct from that of the ‘reasonable pluralism’ doctrine discussed so far. The burdens of judgement ‘limit what can reasonably be justified to others’ because partisans of any particular worldview lack sufficient reason to force it upon supporters of other reasonable views. Because the burdens of judgement limit what can reasonably be justified to others, you can reasonably reject my worldview; therefore you are justified in rejecting any use of political power designed to impose my worldview on you. This argument does not require ‘reasonable pluralism’. All that it requires is scepticism about the relative standing of the worldviews: the ‘liberty of conscience’ doctrine is inferred on the grounds that the one thing which could provide sufficient reason to favour one over the others is not demonstrable. This in turn provides a respect in which persons endowed with the ‘moral powers’ can acknowledge how the burdens of judgement limit what is publicly justifiable. But this is not to say that scepticism does justify impartiality or neutrality, whether in its Rawlsian or any other form.15 Of course, a sceptical view would equally have to take a different view of the status of beliefs from any taken by the believers themselves, but this is less embarrassing to scepticism than to other interpretations of the burdens of judgement. The problem for the other interpretations is that they retract with one hand what they offer with the other. Such are the pitfalls in trying to separate believers from their worldviews in political design. The strength of the private judgements is what makes diversity matter, but holders of the judgements have to soft-pedal on them. The neutral state demands a certain double-mindedness about worldviews. The doctrines must matter enough to set the terms of the problem, but they cannot matter so much that those holding them can justifiably impose them on others. This is not a contradiction. However, it is very difficult to elicit a univocal answer to the question whether, given the burdens, reason under-determines belief in particular worldviews. The result is a bifurcation of reason into public and private spheres. 14 15

Rawls, Political Liberalism, 61–2. Emphasis added. Some might question whether Rawls espouses neutrality in Political Liberalism. As understood here, however, ‘neutrality’ does not connote a lack of moral commitment. Rather it embodies a lack of commitment between a set of worldviews whose eligibility for this treatment is based in part on their espousal of certain common moral values. Hence ‘neutrality’ is not to be identified with amorality. On other forms of neutrality, cf. ch. 6 of Newey, After Politics; Newey, Virtue, Reason and Toleration, ch. 5.

Toleration as sedition

67

Pluralism, reasonable rejectability and toleration Modern liberals try to find a basis for political design by requiring of justification that nobody could reasonably reject it.16 There is an obvious way in which the doctrines of reasonable rejectability and reasonable pluralism might be made to dovetail theoretically. Reasonable pluralism picks out the conceptions of the good relevant to political design, and these conceptions must be respected by the reasonable rejectability criterion. So if reasonable pluralism applies to a given conception, it will be reasonable (that is, not unreasonable) in itself, and therefore any political design which fails to accommodate it will by this fact be unreasonable: it will fail the criterion, since the conception’s supporters can reasonably reject it. The burdens of judgement mean that others can reasonably reject any particular worldview. So it would not be justifiable to enforce policies based on one. On the other hand, when a worldview is reasonable, it would also be unjustifiable for those who reject it because they adhere to an alternative (which will also be reasonably rejectable) to outlaw or otherwise disadvantage it. Insufficient-reason arguments justify some political outcome or process on the grounds that there is insufficient reason to justify any alternative to it. Thus if no conception of the good life is such that nobody could reasonably reject it, then no justified political design (on a stronger view, no justified political outcome at all) could be based on it. For example, it is unjustified to outlaw adultery or blasphemy merely because some reasonably rejectable ideas of the good, such as religious fundamentalisms, demand it. Reasonable rejectability thus works as a filter, removing certain kinds of reason from political justification. Some more demanding accounts of the reasonable may leave a justificatory impasse. From the fact that no alternative to a policy is justified, it does not follow that the policy must be justified faute de mieux. Perhaps no policy is justified on the insufficient-reason form of argument. Those conceptions which call for laws against adultery are deemed reasonably rejectable; but why are not conceptions which do not call for this – or

16

Apart from Rawls’s Political Liberalism itself, versions of this argument can, for instance, be found in T. M. Scanlon, ‘Contractualism and Utilitarianism’, in Amartya K. Sen and Bernard Williams (eds.), Utilitarianism and Beyond (Cambridge University Press 1982) and What We Owe to Each Other (Cambridge, MA: Harvard University Press 1999); Simone Chambers, Reasonable Democracy: Jürgen Habermas and the politics of discourse (Cambridge University Press 1996); Thomas Nagel, Equality and Partiality (Oxford University Press 1991); Jürgen Habermas, Between Facts and Norms, tr. William Rehg (Cambridge: Polity 1996); Charles Beitz, Political Equality: an essay in democratic theory (Princeton University Press 1989); Brian Barry, Justice as Impartiality.

68

Toleration as sedition

which specifically call for adultery not to be illegal – also reasonably rejectable? Perhaps because they are permissive rather than prohibitive. They leave those so inclined free to pursue the adulterous form of the good life, while leaving unmolested those who do not. But, of course, there is no suggestion that legal prohibition or compulsion is always reasonably rejectable: everybody except anarchists will think it reasonable to put certain actions beyond the pale both of toleration and legality. Even so, it will seldom be hard, for any presented policy, to find ideas of the good which deem the policy reasonably rejectable. Where policy is controversial between reasonably rejectable ideas of the good, reason either defeats all justifications, or supports incompatible policies. The way out of the impasse is to drop the requirement on the justification of political outcomes that the outcomes must not be reasonably rejectable, or qualify the burdens of judgement. Either way, certain private notions of the reasonable – that is, notions specific to certain private conceptions of the good – will find themselves translated into the public sphere, on pain of justificatory stasis. Some patterns of prohibition and permission then come to be sanctioned as ‘not reasonably rejectable’; but if this description is not to beg the question, an independent argument is needed to show that their rejection is unreasonable. What makes a doctrine reasonable at the level of private conviction – which for Rawls centrally includes religious worldviews – need not make it not reasonably rejectable at the public level. So, the double-mindedness which beset the burdens of judgement seems to recur with reasonable rejectability. As a private citizen I may reasonably hold a conception of the good which others reasonably reject. As already indicated, it is doubtful whether this is coherent from the perspective of private believers: insofar as each conception is reasonably rejectable, there is no reasonable basis for endorsing it. However this may be, there is a conflict between affirming the reasonableness of private believers’ grounds for holding their conceptions of the good life, and the rejection of these grounds at the public level. The problems facing reasonable pluralist attempts to justify toleration are amplified by toleration’s uncertain place in modern democracy, discussed in Chapter 3. There I argued that nowadays states are usually third-party actors in the political circumstances of toleration, and that the latter often involve mutual accusations of intolerance by the other two parties. Consider Rawls’s attempt to justify toleration via the burdens of judgement.17 The crux of this attempt is that in recognising reasonable

17

Rawls, Political Liberalism, 52–60.

Pluralism, reasonable rejectability and toleration

69

pluralism, I see that others may reasonably reject my favoured worldview. In seeing this, I also realise that there is no warrant for imposing my beliefs on others who reasonably reject them; that is, I see there is reason to tolerate them. Circularity threatens if the scope of the ‘reasonable’ is tailored to what I will accept – that is, I judge to be reasonable what I have already decided I am prepared to live with. Many of these things I will not disapprove of at all, and so will not qualify for toleration. Of those that I do disapprove of, my reasons for not suppressing them – supposing it lies within my power to do so – are not adequately grounded by the thought that my own beliefs may reasonably be rejected, or are subject to the burdens of judgement. I will disapprove insofar as I believe I have reasonable grounds for thinking of the practice as objectionable. My reasons for noninterference, if they exist, are likely to lie elsewhere, particularly in pragmatic considerations.18 As this suggests, the problems facing political toleration in modern states stem from conflicting norms, which come out inter alia in the incoherent role assigned to reason by the burdens of judgement and reasonable rejectability. These conflicting norms emerge in ambiguities over the scope of judgements and the distinction between public and private spheres. The private judgements include controversial claims about their own scope – that is, about whether or not they are intrinsically private. So they offer a specially unpromising basis for resolving political conflicts, at least if the resolution has to be uncontroversial in the sense of being not reasonably rejectable. The scope of ‘reasonable’ conviction, in particular, has to be limited to the private sphere, on pain of contradiction with the diagnostic and prognostic account of what is reasonable at the public level. The argument from reasonable pluralism wavers between construing each doctrine as reasonable, and endorsing epistemic criteria which preclude belief in any one of the doctrines. The first alternative, it is thought, can vindicate the private judgements by showing that the burdens of judgement, or reasonable pluralism, which generate a theory of public justification, will also show how individuals can reasonably uphold their different worldviews. However, the problem with construing each doctrine as reasonable at the public level is trenchantly stated by Hobbes in Leviathan: 18

My earlier discussion in Virtue, Reason and Toleration, ch. 1, suggested that a satisfactory distinction could be drawn between ‘principled’ reasons for non-interference, which ground toleration, and pragmatic reasons, which do not, and that such a distinction was relevant to determining whether a given policy (etc.) was tolerant.

70

Toleration as sedition

when men that think themselves wiser than all others, clamour and demand right reason for judge; yet seek no more, but that things should be determined by no other men’s reason but their own; it is as intolerable in the society of men, as it is in play after trump is turned, to use for trump on every occasion, that suit whereof they have most in their hand . . . bewraying their want of right reason by the claim they lay to it.19

The second alternative is to adopt scepticism at the public level towards private judgements. On the assumption that this scepticism is based on arguments from the ethics of belief,20 scepticism is liable to destroy reasonable grounds for belief in these judgements. Otherwise, a regime of epistemic austerity at the public level must rest on non-epistemic grounds, such as the claim that it is wrong to impose one’s beliefs on others, irrespective of the beliefs’ truth value. If these remarks apply generally to the role of private judgements in political justification, then they will apply to justifying political toleration. This is clearly true of the judgements of disapproval, but also the grounds for restraint, which are equally extruded from private judgements. It is not that there could not in principle be any method of sifting the competing claims which are traded in these situations. But if one starts by ascribing a peculiar importance or authority to private judgements, one will have to look elsewhere for a resolution of the conflicts between them. Beliefs about religion illustrate the problem.21 Differences between religions, or between them and non-believers, exemplify reasonable pluralism if anything does. So theists may reasonably disagree with atheists, or agnostics. Clearly their differences may include, but will also go beyond, assenting to or demurring from propositions such as ‘God exists’. One way in which the disagreement goes beyond propositional attitudes is that the nature of what is asserted or denied, like the basis for doing so, is itself in dispute.22 To that extent belief in God is not like belief in the existence of the Loch Ness monster. In that case, presumably, a method exists in principle for deciding the question – such as draining the loch. Then one might define ‘reasonable’, at a minimum, 19 20

21 22

See Hobbes, Leviathan, 33. The test of public reasonableness could simply be treated as a rule of public justification that where individuals hold conflicting views on certain questions, no view will be taken at the public level as to which is correct, or more reasonable. Rawls’s doctrine of the burdens of judgement is, however, squarely based on the ethics of belief. The attraction of this approach, and others like it, is of course that it promises a minimally controversial basis for justification. This is an advantage which is liable to vanish if the rule is motivated on other grounds, such as some doctrine of equal respect for persons. See also Chapter 6. For a classic statement, see Thomas Kuhn, The Structure of Scientific Revolutions, 3rd edn (University of Chicago Press 1996).

Pluralism, reasonable rejectability and toleration

71

by reference to whether or not someone would regard the outcome of such a procedure as an authoritative way of deciding the question. But nothing similar exists in the case of God, one of the bones of contention being, precisely, whether reason itself can decide the issue. While it seems uncontentious that theists and atheists disagree, their disagreement extends to what exactly they are disagreeing about. They may well disagree about whether their disagreement is a reasonable one. Some theists may regard some of those on the other side as reasonable, while some atheists may take the same view of some religious believers. But often people also disagree about whether those who disagree with them are reasonable. Can this itself be a reasonable disagreement? Of course, disagreement is not reasonable merely because the disputants disagree about whether their opponents are reasonable. But it is hard to see how disagreements about whether one’s opponents are reasonable can be fully insulated from the burdens of judgement. Exactly the same factors – concepts’ vagueness, the differential weighting of relevant factors, experiential and cultural biases in belief, and value-conflicts – arise with these second-storey disagreements as with the first-floor ones. With religion, it is not even a case of applying the same criteria to different effect. People also differ about whether the beliefs are assessable by reason,23 and indeed on whether the issue is about belief at all. If there can be reasonable disagreement, it is possible that one can reasonably disagree about whether a disagreement is a matter of reason.24 Someone may object that reasonable people will be united in deciding which norms of inference count as reasonable. Call this set of norms R. Reasonable people will concur, substantive disagreements notwithstanding, in accepting the norms contained in R. But then this will form the object of reasonable agreement about the epistemic norms on which judgements are to be reached. So, contrary to the previous argument, there can be reasonable disagreement without this extending to disagreement over which disagreements are reasonable. Note that convergence on norms must take the strong form whereby reasonable people agree that all and only those norms included in R will provide a basis for reasonable disagreement. If some reasonable people could endorse a norm outside R, they could reasonably disagree with those reasonable 23

24

‘[R]eason, in its one-sided concern to investigate human subjectivity, seems to have forgotten that men and women are always called to direct their steps towards a truth which transcends them’. Pope John Paul II, papal encyclical Fides et Ratio, 5, 14 September 1998. This can happen intra- as well as interpersonally. See David Hume, A Treatise of Human Nature, ed. L. A. Selby-Bigge, rev. Peter Nidditch (Oxford: Clarendon Press 1978), Bk I Part IV § 7, e.g. 270.

72

Toleration as sedition

people who confined themselves to norms in R; though the latter would see this disagreement itself as unreasonable, this divergence on norms of inference would itself exemplify reasonable disagreement. One such agreement could be over whether or not a certain source should be treated as authoritative. It would need to be shown, without begging the question, that irremediable disagreement about epistemic authority could not arise between reasonable people. But if it could not, the question would be whether they were any longer disagreeing with one another. For the set of uniquely reasonable norms would indifferently warrant inferring some specific judgement and not doing so. And that would mean that a reasonable person would demur at endorsing either. If there is reasonable disagreement at all, there is no reason why it should not extend to disagreement over the grounds on which judgements are reached. What about cases of asymmetry? Perhaps toleration applies just in those cases where one party rejects a presented norm, because the norm really is unreasonable. But it does not follow that those who reject the norm must themselves endorse one (say, the negation of the presented norm) that is also reasonably rejectable. Rather they reasonably reject the norm because they endorse a norm that is not reasonably rejectable. Modern liberals might wish to say that the norm proposed by the anti-gay Christian fits into this mould. It is reasonably rejectable by those who uphold norms that are not so rejectable. This response is coherent. Nothing about judgements of unreasonableness requires that they hold symmetrically. Round-earthers reasonably disagree with flat-earthers, but not conversely. Flat-earthers may think round-earthers are reasonable in disagreeing with them, but the compliment is unlikely to be returned. The flat-earthers will presumably think that their own disagreement with the round-earthers is reasonable, but that belief may itself testify to their lack of reasonableness. So ‘reasonably disagrees with’ is not a symmetrical predicate.25 However, things stand differently for modern liberals who accept the doctrine of reasonable pluralism, since it holds that no idea of the good cannot be reasonably rejected, and that goes too for any idea of the good that rejects anti-gay Christian fundamentalism. So the asymmetry between tolerator and tolerated cannot be marked by claiming that the latter’s idea of the good, but not the former’s, is reasonably rejectable. The claim is rather that certain attitudes towards one’s beliefs are more reasonable than others. 25

There is also the under-explored phenomenon of unreasonable agreement, where people accept the same proposition for inadequate or sometimes contradictory reasons.

Pluralism, reasonable rejectability and toleration

73

The epistemic reasonableness of beliefs is not what matters for toleration. The belief that the earth rests on a giant turtle may be unreasonable as well as false, but I may nonetheless act in an unjustifiably intolerant way if I try to stop people from holding it, from acting in ways that express it (such as turtle-worship) or persuading others of it.26 What matters is my attitude towards those who disagree with me or act in ways of which I disapprove. In summary: reasonable pluralism undermines itself, as reason displaces disagreement, or vice versa; if true, reasonable pluralism derails the political construction that Rawls tries to base on it; and, by substituting the ethics of belief for the ethics of conduct, it largely misses the point. It is also worth noting that questions of toleration often do not involve anything as grand as ‘conceptions of the good life’. Obese people, for instance, often may not be living out their conception of the goodness of the fat life. But the conflicts of resources that involve them can raise questions of political toleration – for instance, when it comes to their entitlement to public medical care, or benefits such as disability allowance. In 2010 the film director Kevin Smith was ejected from a Southwest Airlines flight after being required to purchase a second seat on the aeroplane. Opinions differ about Smith’s treatment. Southwest maintained that he was in breach of its ‘Customers of Size’ policy, which requires that ‘[c]ustomers who encroach upon any part of the neighboring seat(s) should proactively book the needed number of seats prior to travel’.27 The pilot held that Smith also constituted a ‘safety risk’.28 But however one views the case, it clearly raises questions of toleration – not just of the overweight, but also of those who deride them. The obese argue that they suffer unjust discrimination, while detractors accuse them of trying to impose an ideological or ‘politically correct’ straitjacket on discourse. The attraction of reasonable pluralism lies in its offering difference and consensus at once. One can accept disagreement in private beliefs, which hedges what believers can justifiably impose in public, but also what they can suffer at the hands of others. However, as I argued in Chapter 3, in the political circumstances of toleration the judgements usually operate only at the level of disapproval, rather than restraint despite disapproval. This is not to deny that people – even those involved in the circumstances of toleration – may sincerely profess principles 26 27 28

An argument famously made by Mill: see On Liberty, ch. 2, in Utilitarianism and On Liberty. www.southwest.com/html/customer-service/extra-seat/index-pol.html. www.guardian.co.uk/film/2010/feb/15/overweight-filmmaker-banned-southwest-airlines.

74

Toleration as sedition

supporting restraint. But these principles are usually absent from the political circumstances of toleration. The claims of each side amount to the demand that the public sphere implement private judgements of disapproval. Just as toleration tends to disappear when we translate the judgements from the private to the public sphere, so do the grounds for asserting these judgements when they are under the constraints of public justification, rather than confined to the sphere of private belief. The starting point of liberal neutrality lies in a world which is cognitively Hobbesian, a world where equality prevails between the claims of rival believers. The question is not whether it is justifiable to start there, given worries over the justification of normative judgements. Instead it is how sense can be made of toleration as a distinct political value, when the purpose of the political construction is to distance individuals from the private judgements whose very force triggers the circumstances of toleration. Of course such distancing may be cause for celebration, a welcome sloughing of credulity and fanaticism. But the further one gets from zealotry, the further one moves towards a commonwealth of apathy in which toleration is no longer an issue.29 The strength of people’s reasons to care about toleration varies with the strength of their reasons to care, period. Conscience and the sovereign state The thread linking my remarks on reasonable disagreement and toleration is the untranslatability of private judgements into an adequate theory of political justification. This fails to justify neutrality, if it is a condition of political justification that nobody can reasonably reject it. Rawls also gives insufficient reason to endorse a regime of toleration: the bare thought that no form of political design can be justified from within any particular worldview is compatible with any political dispensation, if the latter lacks warrant. Scepticism leaves a deficit in political justification30 and in the justification of toleration.31 Its real consequence is not an even minimal or metaphysically trim liberalism, but a pathology 29 30

31

A point made by Williams: see ‘Toleration: an impossible virtue?’ 26. Though it may, in the terminology introduced earlier, leave it possible to justifice political outcomes, since justificing requires only that the outcome is not such that some other outcome is demonstrably better than it. See my Virtue, Reason and Toleration, ch. 4, esp. 135–8. The basic problem is that scepticism can be applied to the justification of toleration itself. The similar claim that pragmatism justifies toleration, because of the difficulty in establishing which views should or should not be tolerated again falls prey to the objection that a free-for-all may be less pragmatically salient than a more restrictive regime. In many cases the state’s pragmatic purposes (e.g. stifling dissent) will be better served by restriction.

Conscience and the sovereign state

75

of the political. That pathology needs, however, to start closer to home than it might appear, in foro interno.32 The question is why what is thought to matter, at the foundational moment, really matters. What is the compelling reason for thinking that what matters are individuals’ differing private judgements of the good? If there is no scepticismvanquishing answer to that question, we are left with a political version of the myth of the given,33 with first-order normative beliefs or judgements playing the role of ‘raw feels’. The closer they approximate to the latter, the less amenable they are to rational interrogation. One answer to the question goes as follows: the doctrines are, as a brute fact, diverse, and each doctrine matters to the persons who hold it. Of course, this only shows that the diversity itself matters if it matters that the doctrines matter to those who hold them. And if those to whom they matter themselves matter, so that the doctrines have at least vicarious importance. There are various ways in which this claim can be made. One way to do so is by relying not on considerations of reasonableness, but on a substantive moral ideal, such as sanctity of conscience. Then we need only ask whether there is this sincerely held belief. While this provides an argument congenial to religious and other enthusiasts, the sanctity of conscience claim carries a heavy theoretical price. A gap opens up between what individuals hold to be justifiable given their first-order moral beliefs – of the general form ‘X is bad/good, so ban/allow it’ – and what the preferred account of political justification says it is justifiable for the sovereign state to do. This must be of the following form: that someone holds a private judgement which says: ‘Do this’ is not by itself sufficient to justify the state’s doing it. The sovereign state cannot simply translate private judgements of right and wrong into political practice. The veneration of private opinion, particularly in its conscientious form, is distinctive of liberalism: its most conspicuous hangover from solefidian Protestantism. This commitment is, however, double-edged. Despite his rather incongruous makeover in the past few decades as a prophet of toleration,34 Hobbes did not think that opinions or 32 33

34

This is of course a Hobbesian phrase (Leviathan, 110). Rawls picks it up at Political Liberalism, 54: ‘the reasonable always binds in foro interno’. Wilfrid Sellars’s phrase, in his 1955–6 lecture series ‘Empiricism and the Myth of the Given’, published as Empiricism and the Philosophy of Mind. Originally published in Herbert Feigl and Michael Scriven (eds.), Minnesota Studies in the Philosophy of Science, Vol. I (Minneapolis: University of Minnesota Press 1956), 253–329. Examples of this tendency include Alan Ryan, ‘A More Tolerant Hobbes?’ in Mendus (ed.), Justifying Toleration, and Richard Tuck, ‘Hobbes and Locke on Toleration’, in Mary Dietz (ed.), Thomas Hobbes and Political Theory (Lawrence: University Press of Kansas 1990). The main reasons why Hobbes can be mistaken for a defender of

76

Toleration as sedition

private judgements were, in themselves, worthy of respect. A strikingly caustic passage in Leviathan talks of men [who], vehemently in love with their new opinions (though never so absurd) and obstinately bent to maintain them, gave those their opinions that reverenced name of conscience, as if they would have it unlawful to change or speak against them; and so pretend to know they are true, when they know at most, but that they think so.35

The interesting point about the gibe against those ‘in love with their new opinions’ is that it is, after a fashion, Hobbes’s anticipation of Political Liberalism. On closer inspection it turns out, as I argued earlier, that the burdens of judgement show that people who ‘pretend to know’ that their opinions are true ‘know at most, but that they think so’, and this is necessary if Rawls’s political construction is to work. However, Hobbes seems to have a clearer sight of the landscape here than does Rawls. Hobbes sees that if the proliferation of private opinions, ‘though never so absurd’, sets the political problem, it is quixotic to look to those opinions for its solution. The tensions which we detected earlier in the idea of reasonable pluralism and the burdens of judgement arise from Rawls’s efforts to blend two unblendable thoughts, the first being an unvarnished sanctity of conscience doctrine, and the second, the thought that if conscience is allowed to let rip, it will make political life difficult or impossible. The latter thought offers little succour to toleration (which Hobbes in fact seldom mentions, and then unflatteringly,36 in Leviathan). A condition of political toleration was that the private judgements could be brought into the public sphere. But for Hobbes they can do so only if they do not disrupt the basis on which collective existence, in the form of the state, rests. In Hobbes’s ecclesiology, private judgements get what might be thought of as condign treatment. He turns around the ethics-of-belief argument37 which Locke was later to use in the Letter Concerning Toleration to make it support a monolithic, albeit Anglican and deist, confessional state. The claim, in other words, that belief is immune to external coercion is

35 36

37

toleration are that he took a minimalist view of the doctrinal formularies requisite to salvation, and in any case included a get-out clause in the form of a proto-Lockean ethics-of-belief argument. I say more about this in ch. 9 of Hobbes and ‘Leviathan’. Hobbes, Leviathan, 48. For a quasi-Hobbesian argument for a modus vivendi conception of toleration, see John Gray, Two Faces of Liberalism (Cambridge: Polity 2000), esp. ch. 1. In the ‘Review and Conclusion’ of Leviathan, Hobbes berates the ‘toleration of a professed hatred of tyranny’ as ‘a toleration of hatred to commonwealth in general, and another evil seed’. Murphy, Conscience and Community, 113, 145, notes the currency of this argument from the 1640s.

Conscience and the sovereign state

77

used to show that it does not matter that public doctrine is prescribed uniformly; so why not have uniformity? As Hobbes saw, the very privacy of private judgement, the fact that it lies beyond the ‘sword’ of civil authority, makes it ripe for this treatment. The wider point, which is independent of the ethics-of-belief argument, is that the more private the private becomes, the less heed the public authority needs to – and in fact can – pay to it. The immunity of private belief also explains why Hobbes can look like an advocate of toleration, while espousing a form of Erastianism consistent with strict uniformity in public worship. The fact that the privacy of judgement can be invoked to support both latitude and restriction shows the double edge of these judgements. The tension is heightened by sanctifying private judgements as delivered by conscience. Belief in the sanctity of conscience became current in the seventeenth century, when the doctrine of first-person epistemic authority was at its zenith.38 More surprisingly, the doctrine persists among liberals when its epistemic counterpart has come under heavy fire. The doctrine must hold that justification imposes more stringent conditions on public doctrine than those which bear on private judgements. But if the earlier critique of Rawls was right, and it is not merely legislated that different epistemic standards apply to private and public reason, one should doubt whether this line can be held very easily.39 If persons who are reasonable in respect of these beliefs reject them, reason counsels against accepting these beliefs (though of course not necessarily in favour of accepting their negation). What I can justify publicly is, in this sense, what I can justify, full stop. It is important to remember that the origins of the doctrine of conscience lie in religious revelation. That idea, crudely put, is that individuals may get the truth via a text from the Almighty. Modern secular conceptions of the good life lack a credible analogue of this idea.40 The very idea that inner revelation can provide privileged access to outward truth is one which for many of us now lacks content, except as part of certain highly contentious conceptions of the good life.41 What then justifies its persisting role in liberalism? 38 39 40 41

For a powerful argument against the claims of conscience to determine public justification, see Gaus, The Order of Public Reason, 230ff. The most sophisticated attempt yet to do so is Gerald Gaus’s Justificatory Liberalism: an essay on epistemology and political theory (Oxford University Press 1996). Moorean intuitionism is a prime example, as are the various brands of non-theistic superstition. Including liberal ones: for an example see Chandran Kukathas, The Liberal Archipelago: a theory of diversity and freedom (Oxford University Press 2003), e.g. 70–1. One question facing the theory is how to set the terms of argument without begging the question: the dealbreaker in justifying political design is what individuals can accept, which is tantamount to handing conscience a veto. But there is a question about what justifies that demand.

78

Toleration as sedition

The sovereignty of conscience doctrine is a doctrine of sovereignty within sovereignty. Neutrality seeks to solve this problem by accommodating private judgements at the level of political design. But as I have argued, this is of little help in attempts to justify the sovereign state. Without the firm smack of epistemic abstinence, the circumstances of toleration are liable to re-appear when individuals reassert their right of private judgement. And indeed in the political culture of contemporary liberal-democratic states, the call of conscience, insisting on the right of private judgement, has become a wild card played against acts of state that are held to be ‘intolerable’. There are good reasons to jettison unrestricted licence for private opinion. The objection from fanaticism is only the most obvious one. There is indeed a respect in which the sanctity of conscience makes of everyone a fanatic. Not everyone will exhibit zealotry. But the sanctity of conscience offers to all the privilege demanded by the fanatic, an exemption from the burdens of judgement. The importance assigned to the doxastic category, if that is what it is, of the conscientious itself is perplexing. It would be odd to think that the state is obligated to uphold or protect certain beliefs as such; would this mean, for example, that the state should suppress evidence which tended to undermine those beliefs? A belief is as good as its grounds, and this applies to beliefs content-neutrally. A wise man, as Hume said, proportions his beliefs to the evidence,42 and this institutes a kind of democracy among beliefs. By contrast, it may be said that the conscientious is a particular mode of holding beliefs. The most obvious version of this thought is that those who hold them regard conscientious beliefs as particularly important, or as fundamental to their own ‘identity’: the beliefs are held ‘deeply’. But again, the fact that beliefs are said to be held deeply gives no obvious reason why they should enjoy special protection. Creationism and astrology are beliefs whose absurdity in no way mitigates the weight which their adherents attach to them. They may be expressed with unusual vehemence, or prove pertinacious in the face of compelling reasons for abandoning them, and therefore in some cases invite nonrational responses (such as the use of power). But it is not obvious that a white supremacist, for example, deserves protection just because his beliefs are deep ones, or express his sense of his own ‘identity’. Such claims about identity are a modern, secular version of the old arguments from the sanctity of conscience. This is true even when beliefs are not primarily at issue: nowadays ‘identity’ is taken to embrace not 42

David Hume, Enquiry Concerning Human Understanding, ed. L. A. Selby-Bigge, rev. Peter Nidditch (Oxford: Clarendon Press 1975), 110.

Conscience and the sovereign state

79

only beliefs, but also facts about ethnicity, sexual orientation, language, and so on. Like the old claims of conscience, ‘identity’ is used to mount weighty political claims, e.g. by justifying rights against the state. The belief that these claims are warranted is itself a private judgement, not a self-evident truth. The validating ground may have changed – where previously the ground was divine revelation, it is now inwardness itself, sometimes explicated as ‘feeling’ or ‘personal experience’. But the political role these claims play remains similar: to create an enclave within which the state’s writ does not run. By the same token, conscience is invoked against uncongenial political outcomes. The view that conscience forms a legitimate no-go area within politics arises from its depiction as the sanctum sanctorum within the self, or even as what defines the self. This representation may correspond to nothing at all. Even if it does, however, conscience can be seen as a residuum, part of the remainders of politics. A contrasting approach would start not by collating the private judgements, and then seeing what is left of politics, but by assuming that public justification takes precedence, and then asking what space remains for private judgements. Conscience as a no-go zone for politics cedes, on this approach, to politics as a no-go zone for private judgements. The more private the judgements, the greater their capacity to infiltrate and survive in the remainders of politics, and the less heed politics needs to take of them.43 The problems facing reasonable pluralism as a defence of liberal political design, particularly neutrality, and the problems facing political toleration in democracies, both lie in an over-valuation of private judgements. It might be thought surprising that this common root exists: I explore the difficulties in accommodating toleration with neutrality in Chapter 5. However, if a liberal neutralist state could be established accommodating private judgements satisfactorily, there would be no public basis on which toleration could be justified. Moreover the force of the private judgements which set off the circumstances of toleration would have been sapped in the construction of the state itself. This does not mean that there is no possibility that the state could act tolerantly. What has to go is the idea that private judgements demand state neutrality. But as I have argued this is no great sacrifice, and is a sacrifice which has to be made anyway. Above all this requires an adequate understanding of corporate agency, as embodied in the democratic state – of how the state can be a subject for the attribution of actions, with the ancillary notions of responsibility, intention, and so on. And it 43

This is the privacy of personal inwardness, not that of (say) civil society, as it is often contrasted with the state.

80

Toleration as sedition

requires a valid account of political authorisation, which gives the state the go-ahead to act on its citizens’ behalf. These are large requirements. The absence of a credible theory of political authorisation which incorporates the citizenry as political agents is itself one of the prime sources of the view that private judgements can bypass the political process. However, to avoid the problems discussed earlier one has to see the state as a second party, not as a third neutrally holding the ring between the votaries of two rival private judgements. This in a sense moves the state back towards being an organisation of the personal and prerogative sort prevalent when toleration was first mooted as a political ideal in the seventeenth century. This is no coincidence, since if the arguments earlier are right, this is the only way in which the interpersonal structure which models toleration can be transferred without remainder to the political sphere. Conclusion The unitary state expresses the combined will of the demos for a more perfect union – e pluribus unum. In that sense it might be thought there was nothing for toleration to be exercised on. But this is not quite right. It can only tolerate people who have exempted themselves from the state – in Hobbesian or Schmittian terms, declared themselves ‘enemies’ of it, perhaps by exercising private judgement. Then, to the extent that the state incorporates the combined will of its citizens, it is possible to make sense of its acting tolerantly in respect of these enemies. The backdrop to this possibility is that the state’s preparedness to do so is likely to increase with its hold on political power. This chapter has questioned the use of private judgements to provide liberal foundations for toleration. Deference to private judgements in political justification, as Hobbes noted, tends to promote sedition, and this applies also to modern liberal attempts such as Rawls’s to ground toleration. If the argument for toleration appeals to private judgements, it will be hard to argue against those judgements when the time comes – as it will when the state enforces the toleration of one set of beliefs against the dictates of another, or acts in other ways uncongenial to the urgings of the inner voice. The epistemic divide between public abstinence and private licence cannot hold, not least because the inner voice does not confine its demands to the private realm. It also makes demands on public policy, and political design in general. Neutralist liberalism resists this, its general over-optimism about the justificatory power of reason notwithstanding. The tendency, not of toleration itself, but of the justification for it discussed in this chapter, is to subvert the state.

5

The trouble with respect

Liberals now embrace equal respect for persons as a foundational value or principle, and this poses problems for toleration. As once-restricted groups find their civil and legal disabilities relaxed, and become not full, but partial members of the polity, pressure grows to translate this marginal status into full civic and legal equality. The previously sidelined group may explicitly reject toleration as second best.1 Some neutralists and civic egalitarians also reject toleration. This suggests that neutrality arises from a familiar democratic demand, positively expressed as the claim for equality of status or treatment, and negatively as the repudiation of ‘elitism’. Some writers, however, seek to ground toleration on a theory of respect. Meanwhile liberals are often sceptical as to whether respect can accommodate toleration. I shall suggest that the latter’s worries are well founded, at least as regards an act- rather than agent-based conception of toleration. Indeed, a sharp distinction between acts and agents proves hard to sustain: this makes a practically robust conception of respect vulnerable in liberal societies committed to such principles as freedom of expression. I then consider problems facing respect-based accounts of toleration more generally. The respect conception provides a theoretical resolution of the seeming incompatibility between toleration and neutrality. However, the price of this resolution is that judgements of toleration are expressible neither at the political nor, on some conceptions of public reason, at the public level. Respect and equality In liberal writings the value of respect usually takes the form of a commitment to a principle of equal respect for persons, often as a foundational principle of liberal societies. This poses an immediate problem for 1

A superficially contrasting but ultimately similar move is to reverse the negative connotations of ‘toleration’: see Galeotti, Toleration as Recognition, e.g. 226–7.

81

82

The trouble with respect

toleration. One can fail to tolerate by coercing or otherwise censuring those of whom one disapproves, or by failing to disapprove of them to start with. However, I cannot tolerate what I do not disapprove of.2 But then toleration seems at odds with respect. Can one tolerate a person while disapproving of them? More pointedly, even if one can, is such an attitude consistent with commitment to equal respect? The principle of equal respect for persons faces a familiar problem. It claims either too much or too little.3 On the one hand, equal respect can appear merely platitudinous or practically toothless. Thus the principle may be thought to have a purely ex ante significance. It is accepted that some circumstance needs to be cited to justify differential treatment of persons; but without some account of what makes a condition eligible to justify such treatment, the principle is at risk of vacuity. The limit is reached when formal equality can be upheld along with conditions – for example, ‘He is black’, or ‘She is a woman’ – to vindicate what would by general consent qualify as grossly inegalitarian treatment. On the other hand, equal respect can look impossibly demanding. It may be thought to demand not just equal antecedent consideration, but equal treatment – that is, treating everybody the same. These problems are neatly encapsulated by Ronald Dworkin’s distinction between two rights, namely the right to treatment as an equal, and the right to equal treatment.4 In accepting that equal treatment – if that means that everyone must be treated in the same way – is neither feasible nor desirable, we need a non-arbitrary point at which to alight before reaching vacuity. In Dworkin’s case the distinction is cited to justify opposing racially segregated education, while also licensing affirmative-action discrimination in the name of treating persons as equal. The example is significant for the light it sheds on the role of equal respect in political debate. There are various goal-orientated arguments to promote equal respect that can be put in support of affirmative action policies. Among these arguments is that the policies promote the outcome of reducing inequality between racial groups, which may be thought good in itself. However, it clearly does not follow that those individuals who get differentially treated as a result of the policy, for example because of race-specific entrance standards, are themselves receiving equal treatment. Equal respect may be identified with Dworkin’s right to ‘treatment

2 3 4

See Chapter 1. A point made by Bernard Williams about the value of equality; see, ‘The Idea of Equality’ in Williams, Problems of the Self (Cambridge University Press 1973). Ronald Dworkin, ‘Defunis v. Sweatt’ in Marshall Cohen, Thomas Nagel and Thomas Scanlon (eds.), Equality and Preferential Treatment (Princeton University Press 1977).

Respect and equality

83

as an equal’. But if differential admissions standards based on race are consistent with treatment as equals, the point of vacuity seems to have been reached. What commends the policy, if anything, is the egalitarian goal’s desirability, not that the process taken in isolation conforms to egalitarian principle. Of course, various non-arbitrary alighting points have been proposed in the extensive literature on moral egalitarianism.5 Some ideal of reciprocity, for instance, might be thought to do the job,6 or impartiality,7 or a conception of what a person would be prepared to accept as a regulative principle given her desire to seek agreement with others on terms that nobody could reasonably reject.8 Each of these proposals has its problems. The question here however is neither where these alighting points lie, nor what is thought to justify them, but how their justification relates to toleration. And it seems that, whatever the justification, and however the non-arbitrary conception of equality plays out practically, it will pose difficulties. The problem – and the reason why it arises irrespective of the specific justification for equality – lies in the very fact that toleration and equality seem at loggerheads. Consider the numerous public policy controversies that arise over religious dress. Battle has been joined over whether Sikhs should be exempted from compulsory motorcycle helmet legislation, over the admissibility of wearing the burqa or the niqab in various settings, including the classroom, over the wearing of crucifixes by non-publicsector workers, and over the wearing of the Sikh ceremonial dagger, the kirpan, given laws banning the carrying of offensive weapons.9 Often such controversies are thought to involve questions of religious toleration. Again, plainly, the question is not what should be legal in such cases. Where the wearing of items of religious dress is prohibited by the law, toleration is absent. But where, conversely, it is not made illegal, is toleration present? Not necessarily. After all, if Sikhs benefit from exemptions otherwise requiring the wearing of helmets by motorcyclists, or prohibiting the carrying of knives in public places, this might be 5

6 7 8 9

See e.g. Larry Temkin, Inequality (Oxford University Press 1996); Harry Frankfurt, ‘Equality as a Moral Ideal’, Ethics 98:1 (1987), 21–42; Elizabeth Anderson, ‘What is the Point of Equality?’ Ethics 109:2 (1999), 287–337; Gerald A. Cohen, Rescuing Justice and Equality (Cambridge, MA: Harvard University Press 2008). See e.g. Lawrence Kohlberg, ‘The Claim to Moral Adequacy of a Highest Stage of Moral Judgement’, Journal of Philosophy 70:18 (1973), 630–46. E.g. Brian Barry, Justice as Impartiality. Thomas Scanlon’s formulation. See his What We Owe to Each Other, e.g. 153. In 2006 a Sikh, Ripudaman Singh, was convicted by a Danish court for wearing the kirpan. Article §25 of the Indian Constitution protects the rights of Sikhs to wear the kirpan.

84

The trouble with respect

thought of not as ‘mere’ toleration, but as an affirmation of the special value of Sikhs’ religious convictions and practices. In such cases the exemption could be thought justified on grounds similar to those supporting the ‘free exercise’ clause of the First Amendment to the US Constitution, which prevents Congress from prohibiting ‘the free exercise’ of religious faith. In these cases, one might think, there is a question not of toleration but the permissibility of certain forms of dress, and whether or not these fall within laws enacted for particular purposes. The judgement may go one way or the other. But if wearers of the turban or bearers of the kirpan are granted exemption from the relevant law, this only counts as an instance of toleration on the understanding defended in Chapter 1 if the practice is disapproved of, despite its being permitted. In itself, however, the practice is not disapproved of at all – it is rather that the practice faces collision with the law in certain contexts but not others. In ruling that Sikhs should enjoy an exemption denied to other citizens, legislators affirm the practice’s value, or that it embodies universal civic entitlements. The respect owed to the practice or its practitioners militates against its being merely tolerated. Some writers make such examples the basis for an avowedly celebratory account of toleration.10 The converse point also holds: just as equal respect precludes mere toleration, so toleration seems to preclude equal respect. Society can mark a practice with opprobrium without banning it. It can make the practice relatively costly, or difficult, subject it to legal restriction short of criminalisation,11 or shame those engaging in it.12 Thus attitudes towards cigarette smoking, for example, have moved from simple acceptance to one of toleration in recent years. Increased pricing through excise duty generates revenue, of course, but this together with official health warnings on the packet also ‘send a signal’ that smoking is disapproved of; this is consistent with the frequent observation that, as smoking generates revenue, the public authorities do not view it with unequivocal disfavour. In these respects smokers have, so to speak, been met by gamblers travelling in the opposite direction. Off-course betting was illegal in the UK until 1961, and is now subject to a regulatory regime that levies taxes, issues licences to off-course bookmakers, and debars entry to their premises to minors. 10 11

12

E.g. Galeotti, Toleration as Recognition. From 2008, when cannabis was again reclassified in the UK, offenders were subject to a ‘three strikes’ policy by which they would not gain a criminal record before their third offence. It has for example been suggested that customers purchasing alcohol in supermarkets should be required to use a dedicated ‘drinkers’ checkout’.

Respect and equality

85

Do these examples show that smokers or gamblers, who are tolerated within a regulatory regime, enjoy less respect than joggers or stamp collectors, whose pastimes are simply accepted? Several questions converge here. One concerns the relation between respect and equality of respect. A second concerns civic status, and the bearing upon it of activities that incur official disaproval. I deal with these below. A third question, which I consider in the next section, is the justificatory relation between the basis for the disapproval, and that for granting citizens (equal) respect. A quick argument to support the claim that toleration and respect are incompatible goes as follows. There is no more than a verbal difference between respect and equal respect, because to respect somebody entails regarding that person as an equal; to fail to regard, and treat, someone as an equal is to fail to respect them. So, to regard a person as unworthy of equal respect means, to that extent, that one does not respect that person. But to tolerate someone means not regarding them as worthy of equal respect. So to tolerate them means not respecting them, full stop. The quick argument is not very persuasive, at least if its being persuasive means that it compels assent to its conclusion, rather than recognition of its validity. Its premise is dubious: it is not credible that there are just the alternatives of equal respect, and no respect. There can be gradations of disrespect, from using some mildly patronising form of address, to gross denials of a person’s humanity. To derogate from equality of respect in some minor way need not entail a lack of all respect. If disrespect – the lack or effacing of respect – admits of degree, then so apparently does respect as well, and this goes for above-equal as well as below-equal respect. Even liberal societies remain residually timocratic:13 some citizens are accorded more than equal respect, for example by the awarding of public honours. Such practices either do not, despite appearances, mark a departure from equality of respect, or else such departures need not inflict a wholesale loss of respect for everyone else. Hence the strong claim, that toleration must be incompatible with respect because it breaches the only form of respect that there is – namely equal respect – fails. That still leaves the contention that, while toleration need not rule out respect, it is nonetheless incompatible with equal respect. Insofar as someone acts in a way that attracts official disapproval, he cannot be seen as enjoying the same measure of respect as others. However, a second 13

In Republic (tr. Paul Shorey (Cambridge, MA: Harvard University Press 1930)) 543a– 550c Plato identifies timocratic rule as the first stage of political degeneracy, though for him the problem with timocracy is not that it departs from equality, but that it departs in the name of the wrong sort of goods.

86

The trouble with respect

argument can be put to defuse the incompatibility of toleration with respect. This distinguishes between the act and the agent who performs it. In tolerating an activity such as smoking, the authorities signal their disapproval of the activity; but the disapproval need not extend to those who engage in it. Indeed, one could argue that the practice of toleration, far from excluding respect, can rest upon it. The very fact that persons are not disbarred from performing the activity, public disapproval notwithstanding, shows that they enjoy civic respect. One may add that this case is immune to the objection against the quick argument just discussed, since it can allow not merely respect, but equal respect, to those who are tolerated. The same case could indeed be made with regard to acts that are not merely tolerated, but criminal: theft is legally prohibited, but this does not, in itself, derogate from the civic standing of the thief qua citizen, and this has been the legal basis for example of European Court rulings that prisoners should have the right to vote.14 How persuasive this argument is depends, clearly, on distinguishing the status of agents and acts. However, arguments over toleration often insist on the inseparability of agent and act: the act is held to be integral to the agent’s ‘identity’, so that the latter’s manifestation in the controversial behaviour cannot simply be peeled away to leave a ‘radically disembodied’ but equal person beneath.15 But that calls for case-by-case argument, and on empirical grounds. Actor and act are not always as indissoluble as some theories of socially constituted identity maintain.16 It stretches credulity to argue that respect and toleration must be incompatible because the latter calls on people to deprecate acts with which the actors identify themselves. The very fact of disapproval may mitigate identification, as with guilty smokers, furtive consumers of pornography, and so forth. There is no problem in squaring equal respect in the abstract with a practice of toleration. Apart from the act/actor distinction, one can assign everyone civic rights which persist despite inevitable differences in how far a policy burdens one group as against another. But this misses the point. Policy favours some lives over others, such as savers over

14

15 16

See Hirst v. the United Kingdom (No 2) [2005] ECHR 681, where it was found that denying a UK prisoner the right to vote violated his Article 1, Protocol 3 rights under the European Convention on Human Rights. Michael Sandel, Liberalism and the Limits of Justice (Cambridge University Press 1982). This is true of some neo-Hegelian accounts of ‘socially constituted’ identity, such as Charles Taylor’s: see Taylor, Sources of the Self (Cambridge University Press 1989). The same goes for Alasdair MacIntyre’s notion of role-constituted selfhood; see his After Virtue (London: Duckworth 1981); and Whose Justice? Which Rationality? (London: Duckworth 1988).

Respect and equality

87

gamblers, or charitable donors over smokers. Those who claim that the losers forfeit respect may be told that they share equal rights with others, or could choose not to perform the disadvantaged acts. Insofar as smokers smoke, they enjoy less respect. They can quit. But a similar claim about, say, their religion would generally be thought uncompelling. Ad hoc social theory may make its entry at this point, and say that religion is less easily sloughed off than nicotine addiction, but it is unclear that this is generally, still less exceptionlessly, true. Few would think that religious people, let alone religion itself, were respected if their practices attracted severe fiscal penalties, or government warnings of religion’s harmfulness were placed over the entrance to temples. Of course paternalism can be cited as grounds for imposing on smokers in these ways, but paternalism, when applied to some kinds of life but not others, sits ill with equal respect, as J. S. Mill among others saw.17 For egalitarian liberals, equal respect is prior to or trumps toleration, so toleration should be discarded if they conflict. Perhaps they do conflict. However, familiar questions of toleration, say of religious practices such as halal or kosher slaughter methods, may not readily be answered by assuming that equality of respect is the issue. This risks raising the replication problem again:18 the parties to the conflict, whether or not they can be accounted beneficiaries of toleration, may not regard the justification as acceptable, even if toleration may, from the standpoint of the tolerated, be better than nothing. Possibilities like this suggest that the gap between moralised and more pragmatic views of political justification may be narrower than is sometimes supposed. Those who take a moralised view of equal respect as a justification face a peculiar problem. Equality is thought of as a moral debt specifically owed to those subject to political authority,19 and equal respect demands that this authority be justified to them: it is not enough to leave people better off by being tolerated than they would be were they not tolerated. Doing better than nothing may still be a lot worse than being equal. Indeed, once the fact of inequality becomes contentious, those in the disadvantaged group may well come to identify more closely with it than hitherto, and activists may seek to raise consciousness of the issue.20

17 18 19 20

Compare Mill’s attitude towards legal controls on the sale of poisons in On Liberty, ch. 5, in his Utilitarianism and On Liberty. See Chapter 1. For a strong statement of this view, see e.g. Ronald Dworkin, Sovereign Virtue (Cambridge, MA: Harvard University Press 2000), esp. ch. 3. Activism in support of homosexual equality provides a good illustration.

88

The trouble with respect

Respect as civic standing remains a nebulous idea. But it can gain content by concrete markers of standing: for example, the ascription to citizens of certain kinds of entitlement or right, or, more broadly, normative expectations.21 Co-equal standing could be seen as the normative basis for the legal entitlements from which those who are tolerated benefit. Unlike Sikh religious dress, say, the activities of smokers and gamblers are disapproved of, and hence eligible for toleration. But the basis on which the latter are tolerated reaffirms their equal status with others. In political conflicts over toleration the impact of treatment on civic status is often itself a bone of contention, as in the identity card controversy in Egypt during the 1990s. Egyptian government procedures allowed only adherents of the Abrahamic religions to obtain identity cards, with the civil benefits that followed concerning rights to marriage, divorce, and property transfer. Members of the Baha’i faith and others such as atheists, Seventh-Day Adventists and agnostics were disenfranchised and denied other basic civic entitlements.22 The judicial resolution of 2009 granted minority faiths access to these rights, but without granting official recognition to Baha’is and non-Abrahamic faiths.23 Coptic Christians also suffer significant discrimination. The gradations of civic status that put paid to the quick argument show also that toleration can accompany less than equal respect, as reflected in differences of civic standing. So far, then, it seems that respect does not entail equality of respect; but, even if it did, it would not follow that respect precludes toleration, because even equal respect does not clearly preclude it. Nor must equal respect be annulled by the disapproval that toleration implies. But neither does toleration guarantee equal respect, as is seen in the varieties of adaptive behaviour that regimes of toleration elicit, whether furtive or blatant. This is not to pass comment on the adaptive strategies. It testifies, however, to the fact that those punished by opinion themselves may experience a loss of civic parity. To insist that it remains untarnished because civic standing and disapprovedof acts remain distinct may strike those involved as unpersuasive, even disingenuous.

21 22 23

Duties may also be seen as markers of status, as in jurisdictions such as Austria and Australia which require their citizens to vote. See US State Department Bureau of Democracy, Human Rights, and Labor, 2004 Religious Freedom Report, www.state.gov/g/drl/rls/irf/2004/35496.htm. Baha’i World News Service, 14 August 2009, ‘First Identification Cards Issued to Egyptian Baha’is’, http://news.bahai.org/story/726.

The trouble with respect

89

Neutral toleration? Regimes that institute political toleration may, therefore, fall short of equal respect. Perhaps this lack can be made good by honouring equal respect via the notion of a title to justification. Recent writers who have tried to give a justification for toleration that incorporates the value of equal respect, adopt a version of liberal neutrality.24 This is not coincidental, since it seems that the normative motivation for neutrality itself lies in a commitment to equal respect, or the moral equality of persons. Neutrality meets the demand for justification via a regime of toleration respecting each citizen equally. So Egypt’s toleration of the Baha’i faith, though better than civic disenfranchisement, still fails to respect its Baha’i citizens, because they lack the official recognition that Abrahamic faiths enjoy. Neutrality bears a close relation to the ideal of equal respect. People disagree about the good life;25 maybe one can reasonably disagree about ideas of the good. But that only matters to political design, presumably, if it constrains how political arrangements are justified. The relevant constraint is usually put like this: no version of political design is justified if somebody – that is, anybody – could reasonably reject it. That constraint is a version of neutrality: it filters out anything favouring one idea of the good over others. For neutralists, political arrangements are unjustifiable unless they are non-partisan between different conceptions of the good.26 Since it is only plausible to claim that the arrangements are non-partisan if their justification is too, neutrality imposes a condition on justification: no set of political arrangements is justifiable if its justification rests on insufficient reason, and there is insufficient reason to base political arrangements on any one idea of the good. No such idea provides a basis for political design that nobody could reasonably reject.27 But why does it matter if one idea is favoured? Neutrality between ideas of the good which nobody holds appears politically unimportant. The obvious reason is that the beliefs’ value lies in their being held by persons. This makes the normative basis for neutrality clear – commitment to the equality of persons.

24

25 26 27

See, in particular, Rainer Forst, Toleranz im Konflikt: Geschichte, Gehalt und Gegenwart eines umstrittenen Begriffs (Frankfurt am Main: Suhrkamp 2003); and ‘The Limits of Toleration’, Constellations 11 (2004), 312–25. Also Peter Jones, ‘Making Sense of Political Toleration’. Galeotti’s Toleration as Recognition also appeals to equality-ofstatus arguments in grounding its critique of orthodox conceptions of toleration. See e.g. John Rawls, Political Liberalism; Brian Barry, Justice as Impartiality. Clearly only a small subset of all the possible conceptions will sustain this condition: conceptions outside this subset will include, for instance, non-liberal ones. See my After Politics, ch. 6, and Chapter 4 above.

90

The trouble with respect

Thus the norm of reasonable rejectability, as a constraint on justification, seems to derive from a commitment to the equality of persons. Suppose that persons were regarded as unequal, so that members of some group were deemed less worthy of respect than others. Suppose also that the group upheld an idea of the good G not endorsed by anybody else. Then a case could be made for treating G as less eligible than others in competition over justifications of public policy: the lower status of the group’s members could be taken to mean that G could be treated nonneutrally without violating the constraint on political justification that required civic equals to be subject to neutral treatment. But no neutralist seeks to make such an allowance. The underlying commitment is to equality of persons, not that of their beliefs. Compare established religions, such as the Abrahamic faiths, with a newly invented religion that worships, say, bananas. Nobody in fact subscribes to such a religion. But the reason why it is not granted parity with established faiths is not because the beliefs that support banana-worship must be more absurd than established religions’ doctrines.28 It is because there is nobody on whose behalf its claims can be lodged in the public sphere. If so, neutrality is grounded in equality, an ideal expressed by equal respect for persons. It is in that light unsurprising that the apparent incompatibility between equal respect and toleration surfaces also with neutrality. For political authority to remain neutral between ideas of the good seems to be at odds with expressing disapproval of certain ideas, or their associated practices, but not others. Suppose the gambling way of life expresses an attitude of risk-friendliness that is not obviously more unreasonable than one of risk aversion. Then perhaps the state unwarrantably departs from neutrality in signalling its disapproval of gambling by subjecting it to taxation and entry barriers (but tolerating it by not making it illegal), whereas savers benefit from tax breaks designed to reward thrift. If so, and the earlier explanation of neutrality’s relation to equal respect was right, the incompatibility between toleration and neutrality stems from the latter’s justificatory basis in a commitment to equality. Differential treatment of gambling would be unjustifiable because, in penalising it, the state breaches equality by favouring others over gamblers, whose beliefs are not obviously less reasonable than theirs.29 28

29

This is not to say that established religions are as absurd as banana-worship. It is to say only that the relative absurdity of a religious belief is not the main factor that decides whether it is given consideration in formulating policy. It is not obvious that any particular stance towards risk (such as maximin) is more rational than any other. For instance, on expected-utility theory, rationality is only predicable of behaviour in relation to a pre-established schedule of expected utilities that already incorporate the associated risks.

Neutral toleration?

91

The previous section set out reasons for doubting that respect in general, and even equal respect, are incompatible with toleration. But there are also grounds for doubting neutrality’s compatibility with toleration. Neutrality and toleration are both liberal ideals, however. This does not show that toleration is consistent with neutrality: perhaps the recent neutralist turn is an aberration within liberalism, precisely because it cannot accommodate toleration.30 The reasons that tolerators have for disapproval do not lose their force simply by the fact of permission: the reasons are generally outweighed rather than nullified. Then it is hard to see how the neutralist constraint on justification can allow for toleration, since attitudes of disapproval are liable to express a particular idea of the good. With equal respect, the counter-argument lay in arguing that equal respect was consistent with granting individuals the entitlement to pursue activities even when these were the subject of official disapproval. But a gap seems to open up between equal respect and neutrality. On the assumption that acts and agents are distinguishable, and if equality of respect consists in equal civic entitlements, it can be argued that the tolerant state upholds equal respect for, say, its gambling and nongambling citizens, despite its different fiscal treatment of them. On the other hand, the state, in taxing gamblers and subsidising savers, cannot pretend to neutrality, at least of outcome: it rewards one attitude towards risk and penalises another. But the gap closes again when a higher-order version of neutrality such as Barry’s is adopted. Neutralists usually accept nowadays that calling for neutrality in all political outcomes is a counsel of impossibility. What matters is that the outcomes are reached by procedures that do not breach neutrality. A method of reaching political outcomes can be neutrally justified only if nobody’s idea of the good could lead one reasonably to reject it. This will rule out social decision functions that allow dictatorship over outcomes. But it will not preclude outcomes simply because they are favoured by some reasonable ideas of the good but not others. Those who uphold reasonable ideas of the good that others reasonably reject will recognise that these ideas cannot simply be translated into public policy, and equally will accept that their own views may lose out to others that are not unreasonable. So differential treatment is consistent with procedural neutrality. The gap between neutrality and toleration is thus closed in a way that parallels the resolution of the earlier conflict between toleration and

30

Newey, Virtue, Reason and Toleration, ch. 1.

92

The trouble with respect

equal respect. The disapproval implied by toleration seems to preclude both equal respect and neutrality towards ideas of the good. With equal respect, the focus moved from acts to agents; with neutrality, the justificatory focus shifts from outcomes to how the outcomes are generated. In each case what does the normative work is the proviso that a justification only passes muster if it consists with equal respect. But the part played by politics in the process needs to be understood. It is not that people first commit to equal respect, decide whether it precludes toleration, and then pronounce on whether a certain practice should be tolerated. The immediate issue is between those who want the practice to go ahead and those who don’t, and then one decides how to ground these positions in values like toleration or equality. Some writers, such as Peter Jones,31 argue that neutrality is consistent with toleration. Jones invokes a distinction of levels between reasons in support of a neutralist conception of toleration. So-called negative exclusionary reasons32 are reasons for disregarding certain kinds of reason, such as those based on the supposed authority of scripture. Jones argues that toleration can be based on reasons which respect the neutralist condition on public justification by excluding the reasons for intolerance supported by controversial ideas of the good. Private ideas of the good must pass a justificatory test. This will demand neutrality if none is not reasonably rejectable.33 So, Jones argues, the justificatory condition which supports neutrality also supports toleration. Since any individual’s private idea of the good is reasonably rejectable, he must acknowledge that banning practices merely because that idea judges them to be bad is unjustifiable in the public realm; this amounts to tolerating those practices. A religious fundamentalist, for instance, may have to tolerate a society where women have entitlements he wishes to deny them. His belief in a divine warrant for imposing gender inequalities is debarred by negative exclusionary reasons: they are reasonably rejectable, and so cannot justify public policy. Jones argues that people’s commitment to toleration impels them to seek neutrality. To deny that neutrality supports toleration gets ‘things the wrong way round’.34 However, the most pressing political questions about toleration seldom arise between tolerant neutralist liberals. They tend to arise when people, whether liberals or not, believe there is no 31

32 33

Peter Jones, ‘Toleration and Neutrality: compatible ideals?’ in Castiglione and McKinnon (eds.), Toleration, Neutrality and Democracy. Also Jones, ‘Making Sense of Political Toleration’. Jones adopts Joseph Raz’s notion of ‘exclusionary’ reasons (Raz, The Authority of Law (Oxford: Clarendon Press 1979), 16–17). 34 Jones, ‘Toleration and Neutrality’. Jones, ‘Toleration and Neutrality’, 99.

Forst’s respect conception

93

good reason to tolerate something – they think that a certain practice is intolerable. Unless this attitude alters – or is outweighed by other considerations which lead the objector not to call for the practice’s prohibition – he is not acting tolerantly. There is little to be gained by arguing that those who are behaving intolerantly are, at any rate, obligated to behave tolerantly. This may occur even when the intolerant subscribe to a neutralist theory of public justification. They may sincerely believe that the activity to which they object really is intolerable, i.e. would be prohibited by laws which nobody could reasonably reject. In this setting the argument for neutrality will at most show that there is reason for the state to restrain them from behaving intolerantly. In brief, the theory risks operating in a political vacuum. No doubt parties to a hypothetical contract could opt for neutrality out of concern for toleration, but the contract would involve repressing disapproval politically, on neutralist grounds; toleration would not be a public stance. The public stance of the neutral state is not one of disapproval. It has no attitude. These points are in fact opposite sides of the same coin. The tolerator is to be found neither in the figure of the intolerant person, nor in the fictive person of the state. The fundamentalist believes that many practices permitted in modern secular societies, such as homosexual behaviour, are intolerable. Then he either believes that public reason sanctions this position – since it is not reasonably rejectable – or simply rejects this condition on public justification itself. In either case, he rejects the supposedly tolerant policy. Of course, a liberal can reaffirm public reason in support of the policy. But who acts tolerantly, if the policy is then implemented? Not the state, since it is neutral, and therefore does not regard homosexuality as fit for legal prohibition; nor the fundamentalist who would readily see it outlawed. The way to preserve formal neutrality remains by making a distinction between discriminated-against acts, and respect for agents.

Forst’s respect conception Rainer Forst has also attempted to reconcile modern liberal neutrality with toleration.35 Forst assigns a key justificatory role to public reason in order to guarantee neutrality in political justification. Forst relies on a Habermasian principle of discursive justification. According to the principle, whose origins Forst traces back via Habermas to 35

Forst, Toleranz im Konflikt. Also Forst, ‘Toleration, Justice and Reason’.

94

The trouble with respect

Pierre Bayle,36 every norm that is to be binding on persons, especially norms that are the basis of legal coercion, must be justifiable with reasons that are reciprocally and generally acceptable to all affected as free and equal persons. It is this undergirding moral entitlement to respect that constrains what is politically justifiable, since the demand for acceptability arises from the fact that justification is owed to beings who merit respect.37 This requirement of reciprocity and generality bears on toleration as follows. Suppose that it is proposed to introduce some norm that is endorsed by some ideas of the good but not all: perhaps the norm is local to the holy book of some religion, such as the reprobation of homosexuality in parts of the Bible. Fundamentalist Christians who regard the Bible as the word of God sometimes take such passages to justify laws outlawing homosexual acts between consenting adults. But a policy based on such claims would infringe reciprocity and generality. The claims cannot form the basis of a binding norm, since they are not justifiable by reasons that are reciprocally acceptable to all affected as free and equal persons. So for Forst the curbs on public justification yield the non-prohibition of homosexuality. More precisely, its prohibition cannot be warranted on the fundamentalist’s chosen grounds. Permissiveness itself is often identified as a ‘tolerant’ policy. But the policy of refusing to prohibit homosexual acts on the basis of biblical texts counts as tolerant because the justification principle rules out the presented basis of disapproval. In submitting to the principle of justification, each citizen accepts that her moral and other attitudes, if presented in the public sphere as a basis for policy or coercive legislation, have to pass the test of public justification. Attitudes of disapproval pass the test only if they meet the demand for reciprocity and generality. That demand is rooted in a more fundamental moral claim that each person should be respected, and as such is owed a duty of justification.38 Forst’s theory makes an imaginative attempt to show how a neutralist liberal can endorse toleration, despite the apparent incompatibility 36

37

38

Pierre Bayle, Commentaire philosophique sur ces paroles de Jésus-Christ, ‘Contrains-les d’entrer’, ed. Jean-Michel Gros (Paris: Presses Pocket 1992); cf. John Kilcullen, Sincerity and Truth: essays on Arnauld, Bayle and toleration (Oxford University Press 1988), Essay III, and Forst, Toleranz im Konflikt, 327, 330ff. Subsequently Forst has developed this into a general theory of a ‘right to justification’ owed to citizens as such. See Forst, Recht auf Rechtfertigung. Elemente einer konstruktivistischen Theorie der Gerechtigkeit (Frankfurt am Main: Suhrkamp 2007), translated by Jeffrey Flynn as The Right to Justification: elements of a constructivist theory of justice (New York: Columbia University Press 2011). Forst, Recht auf Rechtfertigung, e.g. 36.

Forst’s respect conception

95

between toleration and neutrality. Does it succeed? There is, first, a question of the theory’s scope, in the light of more general concerns raised by neutrality. This question could be thought to pose a dilemma for a neutralist theory of toleration. On the one hand, it seems that any theory which allows for reasonable disagreement between people, as do modern liberal theories such as Forst’s, will have to accept that there can be reasonable disagreements over political outcomes, viewed in isolation from their procedural origins. Before the political process takes its course, it is possible and indeed expected that people will disagree over public policy.39 That disagreement may well extend to the procedures. But even if it does not, there is a question about how justification within them unfolds. Argument often leads to stalemate. That is why political procedures embody decisionistic devices, like the parliamentary guillotine, which rely, directly or indirectly, on a vote. I return to this point later. But first, the theory needs to say something about how it will deal with dissensus. Like other liberals, including Rawls, Forst seems to underestimate the justificatory problems posed by reasonable disagreement.40 It is hard to see how the conditions that make for reasonable disagreement can be kept from undercutting the basis on which others’ judgements are held to be reasonable. People in disagreement may make different claims about what inferences agreed facts license, or the facts themselves, their relative salience, and so on – Rawls’s ‘burdens of judgement’.41 They may disagree about whether judgements of disapproval justify a certain public policy. The burdens lead one person to assert a judgement, and another to deny it (or demur from asserting it), though each is reasonable. But the people will remain in disagreement insofar as they regard the basis for arriving at their own judgement as more reasonable than the other’s. If people can reasonably disagree about first-order judgements like ‘homosexuality is sinful’, they can reasonably disagree also about the basis on which judgements like these are formed. A predictable, empirically well-attested outcome of disagreement is that people disagree about whether those who disagree with them are reasonable.42 If there can be reasonable disagreement at all, it is possible 39 40

41 42

E.g. because reasonable disagreement is held to be a foreseeable outcome of the exercise of human judgement in conditions of freedom. On the problems posed by reasonable disagreement for Forst’s constructivist account of justification, see Thomas Besch, ‘Diversity and the Limits of Liberal Toleration’, in Duncan Ivison (ed.), The Ashgate Research Companion to Multiculturalism (London: Ashgate 2010). See also Chapter 4. Rawls, Political Liberalism, 53ff. For a celebrated example, see Voltaire, Treatise on Tolerance, 132.

96

The trouble with respect

that there are reasonable disagreements about which disagreements are reasonable.43 If reasonable disagreement affects the bases of judgements of reasonableness, it will also yield substantive disagreement over which disagreements are reasonable, as I argued in Chapter 4. If there can be reasonable disagreement on the grounds Rawls cites, the burdens of judgement will apply also to disagreement over the grounds that support specific practical claims. This is particularly obvious with authority heuristics – that is, the very common method of reaching judgements by citing a source regarded as authoritative, such as the Royal Society, or the Bible. One response would be to deny that the disagreement between such putative authorities is reasonable. But it will be difficult to show that this is so without ruling out reasonable disagreement entirely. Consider the procedures that yield political outcomes. One might say that the procedures will rule out, as Forst’s discursive principle suggests, any judgement that a reasonable person can reject. It might be thought that this would be immune to the previous argument, as no judgement that could be reasonably rejected could justify a political norm. Then toleration would presumably consist in recognising that the procedures disqualified one’s own judgements of disapproval in this way. For example, I recognise, as a fundamentalist Christian, that my disapproval of homosexuality rests on grounds with which others who are not unreasonable disagree. In recognising this, I accept that my own disapproval is insufficient to justify public policy such as legislation banning homosexual acts. But if there can be reasonable disagreements about what is reasonable, the problem which the discursive principle aimed to solve afflicts the proposed ‘solution’ too. The problem here is not just the specific doctrine of reasonable disagreement. It is that a demand for justification rooted in the norm of equal respect cannot be met in the way the theory demands. For Forst, political justification requires that judgements of disapproval are eliminated as grounds for public policy if they are not reciprocally and generally acceptable. What does it mean for something to be ‘acceptable’? It seems to mean that it is capable of being accepted, or such that it can be accepted. But it is not clear what it means for something to be capable of being accepted.44 To say that a person can accept something – that she is capable of so doing – is not to say that she in fact accepts it. .

43 44

See also Chapter 4. There is a further question, whether conditional statements are entailed by those in which the modal verb ‘can’ appears. For doubts about whether they are, see J. L. Austin, ‘Ifs and Cans’, in Austin, Philosophical Papers, 3rd edn, ed. Joel Urmson and Geoffrey Warnock (Oxford University Press 1979), 206–32.

Forst’s respect conception

97

Perhaps saying that means only that there is some possible world in which she accepts it. More plausibly, the notion of acceptability makes an irreducibly normative claim: a proposition is acceptable, say, if a person would accept it, insofar as she is reasonable. But ‘reasonable’ here is a place-holder for the upshot of some real or idealised process of negotiation. The problem, summarily put, is that either there is some independent standard of reasonableness that a real process may fail to meet – and then dispute arises over that; or the process itself determines what is reasonable, and that will then perforce incorporate decisionistic elements.45 Suppose toleration is predicable of citizens who accept the principle of justification. If the Christian fundamentalist accepts the principle, he may also think that homosexual practices should not be prohibited, despite his disapproval of them. But this underlines the fact that judgements of toleration are privatised in Forst’s respect theory. The fundamentalist is meant to accept a requirement of public justification that calls on him to rein in his disapproval. However, he may accept the principle but think that his disapproval meets it. The public judgement licensing homosexuality will not be one that he ‘can’ accept. In what sense, then, has his right to justification been met? Then, schematically, the response seems to be to arrange or construe the procedures so that, despite what he thinks, his right has been met, and the notion of acceptability is parasitic on having the right grounds for acceptance. There is however no reason to think that the notion of reciprocal and general acceptability will indicate what acts can be justifiably tolerated. Compare homosexuality and paedophilia. Historically the line between male homosexuality and paedophilia has sometimes been blurred, as with the ancient Greek practice of pederasty.46 The line was blurred in more ways than one. Those whom older men pursued, the desired or eromonoi, were typically adolescent boys whose age was around that of the current age of consent for sexual intercourse (now sixteen in the UK), while the sexual practices themselves fell short of – or less tendentiously, differed from – penetrative homosexual acts.47 Yet among conventional opinion in western liberal states, the one practice is now taken as paradigmatic of tolerability while the other is seen as exemplifying its opposite. Was pederasty reciprocally and generally acceptable to ancient 45 46 47

I adapt here a parallel argument from my ‘Just Politics’, Critical Review of International Social and Political Philosophy 15:2 (2011), 165–82. See e.g. Kenneth Dover, Greek Homosexuality (Cambridge, MA: Harvard University Press 1978). Ancient Greek depictions of pederasty usually show intercrural rather than anal intercourse. Dover, Greek Homosexuality, e.g. 99ff.

98

The trouble with respect

Greek citizens? It is hard to credit that such distinctions can be drawn by reference to a pure, pre-political conception of the reasonable, rather than to the kinds of activity that people will, as a matter of fact, accept, and such dispositions embody cultural residues as well as the contingent effects of political processes. This points up the wider truth that judgements of reasonableness often do not reveal themselves at the start of a political process. What looked like a cantilever structure with the stress borne by reason proves to rest on political processes: judgements of acceptability, at the end of the process, bear the weight supposedly borne by independent judgements of reasonableness. Liberals who deplore the Egyptian state’s treatment of Baha’is or Copts make an independent judgement to this effect, which then yields a basis for political criticism. Justifiability to a group of people, nominally resting on what they can accept, appeals to a prescriptive ideal of the reasonable that, as far as the extra-political justification goes, rests on the claim to be proved. What someone finds to be reasonable in a given situation, will not in general lie far from what that person can live with.48 The obvious conclusion is that, in the presence of disagreement, justifiable bases for public policy are not determinable ahead of the political process. The case of paedophilia also illustrates the fact that arguments for respect are double-edged. Earlier I examined attempts to square toleration with equal respect via a distinction between agents and acts. It is at least doubtful whether that distinction can be made as robustly as it needs to be for equal civic status to survive the discrimination implicit not just in intolerance, but in toleration as well. But, in any case, the act– agent distinction under-determines toleration. The case for toleration in Forst’s theory appeals to an ideal of equal respect as equal civic status. Its real work is done by a baseline moral claim about the equal respect due to those involved, realised politically in an ideal of civic equality. Some may say that equal respect requires the toleration of, say, homosexual acts. Meanwhile others can say that they should not be tolerated, precisely because intolerance does not annul homosexuals’ equal civic status – this being distinguishable from the acts objected to. The acts–agents distinction that reconciled equal respect with toleration allows also for its opposite. It seems that the respect conception, like other modern liberal theories, is forked between its commitment to norms of public justification and its commitment to a doctrine of reasonable disagreement that makes 48

How far a given manifestation of the disposition to accept a judgement confers justification on the accepted arrangements is an irreducibly political question.

Forst’s respect conception

99

those norms unfulfillable. This criticism does not assume that the doctrine itself is incoherent. If there can be reasonable disagreements about whether or not a disagreement is reasonable, the latter fails as a basis for public justification. In applying the public justification filter, the respect conception purportedly sieves out reasonable from unreasonable ideas of the good. The sieve applies judgements about what people find acceptable: what, in other words, they could accept. I have suggested that the truth-conditions of this claim are unclear. To translate them into claims about what is reasonable risks begging the question. Does a society governed by Forst’s discursive norms act tolerantly? The discourse principle in effect privatises judgements of disapproval, so the political authority does not practise toleration in its own right. Certain judgements of disapproval, such as Christian fundamentalists’, will not pass the discursive test: they will be screened out, because reasonable people may reject them. As a result, the state or political authority, as an executive body, cannot be said to act tolerantly. There is no balancing, from the standpoint of a single agent, between competing reasons for action – ones that warrant disapproval on the one side, but self-restraint on the other. There are only reasons why state power should restrain another from trying to enforce his disapproval through political institutions. There would be a political issue to address even if politics could be defined uncontroversially as the domain of the reasonable. If the political construction at which Forst aims worked, we would then have a univocal account of tolerable and intolerable action. It must be common ground that, relative to this theory, some people would still act in an unjustifiably intolerant way. Like other liberals, Forst would presumably say that unreasonable action can be a fit object of coercion: the notion of reasonable acceptability tells us when the use of power is or is not justifiable. But even where this concept gives a clear verdict as to whether a given action is intolerable, it is an open political question how to deal with it, given that the action may legitimately be prevented. Whether it is a good idea to do so will depend on irreducibly political considerations, such as the threat, or lack of it, to public order.49 The circumstances of toleration exemplify the wider functional point about politics already made.50 We risk misunderstanding politics if we start with a stipulative definition of it, and then classify acts or 49

50

A further difficulty: presumably this prevention counts as ‘curbing agency’ if anything does. If so, this form of response to the intolerable will always take the form (albeit, by hypothesis, justifiable) of intolerance. See my After Politics, especially ch. 1, and Chapter 1 above.

100

The trouble with respect

concerns as political according to whether they fit in with the definition. The content and scope of politics is largely given independently of theoretical attempts to delimit it. It is a public arena in which concerns or interests contend with one another and make different claims on the direction of policy. The fact that these concerns surface in this way is not some contingent or marginal feature of the political scene: it largely is the political scene.51 To this extent, ‘politics’ simply is the name given to the clearing-house of proposed policies whose merits cannot be decisively judged in the abstract. There is still a political question to answer, namely, What do we do about these people? – even if we have decided that they are acting in breach of the ideal. To define politics as ‘agonistic’, risks putting the cart before the horse.52 In other cases, plausibly, neither side need be acting unreasonably, and then reason fails, even in theory, to determine policy. Clearly this failure will often be made good by the exercise of power. The circumstances of toleration exemplify this functional aspect of politics, because they arise precisely where the ideal has failed to forestall conflict between the actors and those who wish to impede them. I have argued that, since any theoretical basis for reasonable disagreement will have to allow for reasonable disagreement about the reasonable itself, public reason will not be able to determine in all such cases who is acting reasonably. But even when it can, the political question remains open.

Conclusion I said at the start of this chapter that equal respect, like equality, finds itself forked between claiming too little and too much. Trivially, differential treatment has to point to a difference, but as this is always possible given that cases are being distinguished, there is ample scope for differential treatment. But once equality of respect has more content, for example as equal civic status, it seems to preclude toleration. It also begs important political questions. Reasonableness cannot definitively answer questions of political action, since these questions remain even if either side demonstrably acts unreasonably. Public reason is much less dispositive than Forst supposes: each side can cite not its own interests, but principles purportedly promoting the public good.53 51 52 53

See my ‘Gray’s Blues’. For an example of such an approach, see e.g. Chantal Mouffe, The Return of the Political. A point well made by John Horton with regard to the Satanic Verses affair: see Horton, ‘Rawls, Public Reason and the Limits of Liberal Justification’, Contemporary Political Theory 2:1 (2003), 5–23, 17–18.

Conclusion

101

The best to hope for is to privatise judgements of disapproval at the public level. Anyway, since the content of reciprocal and general acceptability is yielded by political encounter, the notion of reciprocal and general acceptability as a criterion of validity for public reason fails effectively to filter competing claims. One can salvage equal status by insisting on a robust notion of civic equality that survives the disapproval meted out to particular acts; but that leaves the way open for intolerance, precisely because it does not prejudice equal status. It is not predetermined, in any situation, whether people can detach their civic status from acts that may attract public disapproval. Here, again, one-size-fits-all social theory helps little. Disapproval can serve either to concentrate or to dissolve identities: toleration may be self-stabilising, but also self-undermining. It stabilises itself by confirming the stigma attached to certain acts; this makes autonomy-based justifications of toleration problematic, since Mill’s ‘punishment by opinion’ makes the acts less socially available to agents. If that goes far enough, opinion achieves what legislation or policy baulked at doing, and the acts become effectively unavailable to anybody wishing to retain the social bases of respect. Meanwhile, where stigma provokes reactive identification, the perceived margin between full respect and toleration may eclipse that between toleration and no respect. Some public policy, and a good deal of political theory, effaces or foreshortens these dimensions of difficulty. The practical questions they provoke remain firmly on the table. The role of privatising disapproval – more inimical to toleration than it may appear – will be to yield a temporary armistice rather than a final peace. A world in which this is a reigning political dispensation is not one in which the political problem of toleration arises – such problems have been resolved in the political construction itself. Political philosophy can however focus on those worlds where the political problem of toleration does arise. One such is – for want of a better word – ours.54

54

The idea of a common or shared world is sometimes used to denote a matrix of shared normative agreement: see Creppell, ‘Toleration, Politics and the Role of Mutuality’. But at the most basic level, we can share a world without having any basis for agreement about what to do, as I argue in ‘Toleration, Politics, and the Role of Murality’, also in Williams and Waldron (eds.), Toleration and Its Limits.

6

How not to tolerate religion

Liberal orthodoxy, supported in the US by constitutional provisions such as the Equal Protection clause (Amendment 14) and Title VII of the 1964 Civil Rights Act, extends toleration to religion on grounds of justificatory equality. I refer to this as the orthodox position. I shall argue that the orthodox position is unstable, since religious toleration rests on reason of state, rather than, for example, on an abstract justification of political authority. The widespread acceptance of religious pluralism coexists, at least at the level of public rhetoric, with a marked distaste for multiculturalism. In western democracies, multi-faith societies are largely accepted as an established fact. On the other hand, many remain hostile to the notion that immigrant or other minority communities might exist largely in isolation from the host culture, or that a real or imagined indigenous monoculture might give way to pluralism at the public level. However, this conjunction of attitudes is surprising, not least because religion lies deeply embedded in the nexus of attitudes and practices which comprise ‘culture’. It is the more surprising, given that religion has often strongly challenged the bases of political order, and dominated early modern debates over toleration. With particular reference to Locke, I shall argue that this centrality has helped to obscure the real stakes in thinking about toleration – particularly the toleration of religion. Multiculturalism and religion In early 2008, the reported statement of the then Anglican Primate Rowan Williams that the introduction of Sharia law was ‘unavoidable’ in the UK was greeted with dismay and widespread condemnation. The Archbishop’s opponents tended to see themselves not as advocating curbs on religious liberty but as standing against the commandeering by cultural minorities of the public realm, and raised the spectre of legal 102

Multiculturalism and religion

103

(and social) balkanisation.1 No doubt it was the source of the remarks, the head of the Anglican Communion, as well as their content, which provoked reaction. Meanwhile the few liberal multiculturalists who defended the Archbishop argued that, if adopted at all, Sharia law would only apply in very limited areas and that there were precedents for this in, for example, halakha.2 It was also pointed out subsequently that the precedent in Jewish family law did not establish a separate domain of legal jurisdiction.3 Of course, religious toleration had already long been an accepted fact in the United Kingdom, where the Test and Corporation Acts were repealed in 1828; the Roman Catholic Relief Act was passed in the following year. Many other modern liberal democracies have institutionalised religious toleration constitutionally. For example, the United States Constitution enshrines religious toleration via the so-called ‘Free Exercise’ clause of the First Amendment, which provides that ‘Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof ’. Meanwhile, Article 9 of the European Convention on Human Rights, incorporated into UK statute law via the Human Rights Act 1998, provides that Everyone has the right to freedom of thought, conscience and religion; this right includes freedom to change his religion or belief and freedom, either alone or in community with others and in public or private, to manifest his religion or belief, in worship, teaching, practice and observance.

It is worth noting that this clause of Article 9 is immediately followed by a second, which provides that: ‘Freedom to manifest one’s religion or beliefs shall be subject only to such limitations as are prescribed by law and are necessary in a democratic society in the interests of public safety, for the protection of public order, health or morals, or for the protection of the rights and freedoms of others.’ While the jurisprudence of the UK Human Rights Act remains in its infancy, the US Supreme Court has long interpreted the Free Exercise clause in an expansive way, as in landmark judgements like Yoder.4 In this case, the Burger Supreme 1

2 3

4

For example, the then Leader of the Opposition David Cameron described the Archbishop’s remarks as ‘dangerous and illiberal’. A former Archbishop of Canterbury, Lord Carey, remarked that the extension of Muslim law would be ‘disastrous’ and the former Home Secretary David Blunkett said such a step would be ‘catastrophic’. See e.g. the 2009 report by the Centre for Social Cohesion, Beth Din: Jewish law in the UK; www.socialcohesion.co.uk/files/1236789702_1.pdf. Beth Din, 21, which notes that the ‘Beth Din is neither a legal court nor a parallel legal system’, whose ‘rulings and advice can only be reflected in UK law if both parties [i.e. to a marriage] freely agree’. Wisconsin v. Yoder, 406 US 205 (1972).

104

How not to tolerate religion

Court unanimously ruled that the attempt by Wisconsin state to compel the children of Amish parents to attend state school beyond the eighth grade conflicted with the parents’ interest in free exercise. The rise of ‘Pastafarianism’,5 and the fact that some 176,000 respondents to the 2011 Census in the UK identified themselves as adherents of the fictitious Jedi religion,6 indicate one of the difficulties facing a vigorous commitment to imposing freedom of religion legally: namely, that it licenses claims to special treatment on the grounds of real or feigned religious scruple. The ‘Pastafarian’ religion professes belief in a Flying Spaghetti Monster as the creator of the universe. Pastafarianism was devised in protest against the Kansas State Board of Education, which had decided to teach so-called ‘intelligent design’ as an alternative to evolutionary theory. An Austrian Pastafarian, Niko Alm, posed for a driving-licence photograph with a pasta-strainer on his head – the Austrian authorities require that applicants should not wear headgear for their licence photograph, with exemptions for religious dress such as the veil or turban – and successfully claimed the religious-dress exemption.7 However opportunistic the invention of Pastafarianism may seem, the serious intent behind the ‘religion’ is plain: to discredit appeals to religion in justifying derogations from the public education curriculum which rest, for example, on creationist objections to the teaching of evolutionary theory. But the fact that secularists perceived a need to respond to religion in this way shows the strength of religion, at least in the US. Wide-ranging protection for religion is a legal fact. In this case, at least, the fact of diversity between acculturated beliefs is not seen as threatening the legal order – on the contrary. In the US, the diversity promoted by religious toleration, far from being regarded as a threat to social cohesion, is seen as promoting it, or even as its necessary condition. As the response to the Archbishop showed, matters are different in the case of multiculturalism, often seen as a symptom or catalyst of social disintegration. Hostility to multiculturalism – conflated with the rejection of toleration tout court – is widespread, and its expression was triggered by his remarks. In a comment on Williams’s remarks, the journalist Ross Clark noted that they had coincided with the fly-posting of notices by Islamic radical activists that proclaimed the outlawing of gambling, alcohol, tobacco, music and pornography. 5 6 7

See Bobby Henderson, The Gospel of the Flying Spaghetti Monster (New York: Villard Books 2006). www.guardian.co.uk/uk/datablog/interactive/2012/dec/11/census-religion. See e.g. www.bbc.co.uk/news/world-europe-14135523.

Multiculturalism and religion

105

The idea of gangs of self-appointed moral guardians roaming the streets enforcing their own religious code should disgust anyone who cherishes our freedom and democracy. The police have said they will investigate who put up the posters and we should hold them to it. When the culprits have been found they should be charged with threatening behaviour. Previously the Metropolitan police has shown a worrying tolerance of Islamicists trying to establish sharia law in Britain.8

The hostility which the Archbishop’s remarks attracted largely stemmed from their apparent renunciation of a unitary legal order. His many critics charged him with foreshadowing a society disaggregated into legally autonomous enclaves. Few of the Archbishop’s defenders were prepared to argue that such a form of legal fragmentation would be desirable. Both sides took the view that society must protect itself against subversion, and that doing so requires a unitary legal order. In the article just quoted, Clark went on to argue that the Government should not ‘sit back while extremists try to claim parts of the country as their own fiefdoms, subject to their own arbitrary laws’.9 The spectre of balkanisation arises when civil society diversity extends to subverting the very basis of unitary legal and political authority.10 To this extent, the reaction to the Archbishop’s remarks expressed what is at root a Hobbesian worry: that a realm divided within itself cannot stand, and to juxtapose a clerical source of jurisdiction with the secular one is to import just such a division into the body politic.11 Any toleration extended to religious groups presupposes that these groups unite under laws to which they are all equally subject. One striking feature of the reaction against Williams’s remarks is that it subjected religion to charges similar to those levelled against alien cultures in debates over such matters as language, dress, and the public education curriculum. The main concern which such charges expressed was that there are limits to the amount of diversity that any cohesive society can accommodate. This hostility to multiculturalism is somewhat puzzling, when set alongside the generally strong commitment, and not 8 9 10 11

Ross Clark, ‘There Is No Place for Sharia Law in British Society’, Daily Express 29 July 2011, www.express.co.uk/posts/view/261698. Clark, ‘There Is No Place for Sharia Law in British Society’. In the academic literature this is a point which was argued strongly by Brian Barry: see his Culture and Equality (Cambridge: Polity 2000). Needless to say, the Archbishop’s position was more subtle than it was depicted as being by hostile commentators and politicians. The lecture on which hostile reaction focused was in fact an exercise in political philosophy, which argued on communitarian grounds that the entitlements of religious groups such as British Muslims could be met only by offering them a choice of jurisdictions within carefully circumscribed areas of law (such as family law).

106

How not to tolerate religion

only among liberals, to religious toleration. On the one hand, the current orthodoxy does not merely permit religion on sufferance, but actively incorporates it within the legal framework of the body politic. On the other hand, multiculturalism – at least in the form of the proposed introduction of Sharia in the UK – is seen as threatening this framework, and this indicates the flimsiness of the distinction on which the differing rhetorical treatment of multi-faith and multicultural societies rests.12 The divide between hegemonic and multiculturalist liberal positions is then defined by contrasting views of the conditions needed for socio-legal cohesion. Some, like Barry, argue that cohesion can be achieved only by a legal order committed to universal norms,13 while others maintain either that society can withstand the interpolation of diversity in norms, or that in modern conditions society may not survive without it. In response to the hegemonic view, it can be said that exemption is integral to the law. Well-known examples to benefit cultural minorities, such as the exemption of Sikhs from the legal requirement to wear helmets while riding motorcycles, form only one limited example of a much broader legal phenomenon – the differential application of legal incidents as between different groups. The entire tax and benefit system administered by modern states offers a clear example of this phenomenon. The system treats individuals differentially depending on whether or not they are high-earners, drinkers, smokers, parents, drivers, car-owners, shareholders, disabled, elderly, and so on – presumably, in neutralist liberals’ view, without derogating from equal respect for persons.14 Whether or not a given cultural practice such as wearing the turban should benefit from legal exemption can be debated, but the decision cannot be based on a pseudo-principle which holds that exemptions, merely as exemptions, must negate the force of law. It may be said in response that the difference with the Sharia case is that the tax-benefit system issues from a unitary legal and political authority sanctioned by Parliament. But the devolving of law-making powers is equally an established feature of the legal system. For example, statutory instruments account for several thousand items of legislation in each UK Parliamentary session, compared with only a few dozen Acts of Parliament. Those thereby endowed with legislative powers under the parent act include government ministers, local councils, devolved government, quangos, and the church. Hence there is ample precedent 12

13

In the UK, while the civic disabilities attached to membership of other churches have long been abolished, Anglicanism remains the beneficiary of state sponsorship. As such, it offers a good illustration of how toleration need not entail equality of treatment. 14 Barry, Culture and Equality. See Chapter 5.

Multiculturalism and religion

107

for creating areas of devolved jurisdiction via primary legislation. Of course the possibility arises that the exercise of powers conferred by a statutory instrument may conflict with extant primary legislation, by the conferee’s acting ultra vires. But this risk inheres in the very idea of delegated legislation. It is not clear why, in line with current practice regarding the administration of Jewish family law, for example, it would undermine the integrity of UK law to create a parallel cooperative framework for Muslims. Law is a flexible instrument for regulating society. I have suggested that this makes puzzling the belief that religion on the one side and multiculturalism on the other should get different treatment – at least if the reason for thinking this is that multiculturalism undermines the rule of law.15 It is mistaken to think that there is something, law, which can coexist with diversity in religious belief and practice, but not with other manifestations of ‘culture’. The very arbitrariness of such a position, once brought to light, suggests that other concerns may underlie it. Arguments ostensibly about the law express deeper misgivings about the integrity of the polity.16 They articulate conflicting views about what forms of diversity the polity can and cannot accommodate. In saying this, I do not mean to side with multiculturalism against a more ‘robust’ or universalist defence of liberal hegemony. The point is that the law can be made more or less encompassing, depending on whether the arguer aims to make a case for the toleration or exclusion of a certain group. Of course, part of the problem raised by the term ‘Muslim’ as currently used is that it is habitually applied not simply as a religious label but as a badge of cultural, even of ethnic, identity. Thus ‘Muslims’ benefit from the liberal norm of religious freedom insofar as this results from undergirding principles such as free exercise or freedom of association. Meanwhile the assimilation of Islamic culture in other areas of social life provokes disquiet, even though this is consonant not only with Islamic religious teaching but the teaching of other tolerated religions as well. These debates are not new. They raged in England during the early modern period, when political leaders disputed the civic inclusion or otherwise of non-Anglican Protestant sects, including freedom of worship. In markedly similar vein, those who proposed the ‘comprehension’ of Dissenters within the Anglican Church in England after the Restoration were opposed by a narrower High Anglicanism, which 15 16

A simpler route to the same conclusion would note that arguments over multiculturalism often take the form of disagreements about religious differences. See e.g. Will Kymlicka, Multicultural Citizenship (Oxford University Press 1995).

108

How not to tolerate religion

sought the exclusion of Nonconformists.17 In other words, judgement is driven by asking, first, whether or not a certain group or its credo is acceptable. A legal or theological rationale is then provided to underwrite it. In each case the motivating question is: can we live with these people and their beliefs? In this respect the decisions facing modern liberal states in addressing religious fundamentalism parallel those which early modern rulers had to take in mapping the due limits of religious dissent. The dispute over Sharia law in the UK broached the question of the limits of the law – whether toleration can be extended to those religions which challenge, or are thought to challenge, the basis of a unitary legal framework. I have suggested so far that the surface arguments fail to convince, because their normative content can be made to point either way. In the next section I address similar questions raised by Locke’s account of toleration. Locke on toleration Locke argues in the Letter that religion lies outside the secular ruler’s jurisdiction. In line with such critics as Locke’s contemporary Jonas Proast,18 and with the earlier account of religious toleration by Hobbes, I shall suggest that it is not the secular ruler’s claims to spiritual authority, but concern about the threat posed by spiritual leaders’ claims to secular authority, which provides the best argument for intolerance of religion – and hence the argument which defenders of religious toleration have to defeat. Locke’s arguments for toleration, both in the 1667 Essay and the later Letter, are well known. Equally familiar is Locke’s withholding of toleration from Roman Catholics and atheists. His best-known argument for toleration focuses on the alleged irrationality of coercing belief: since, according to Locke and others, belief cannot be induced by force, the ruler who uses physical coercion to make his subjects believe in religious orthodoxy is attempting the impossible.19 For laws are of no force at all without penalties, and penalties in this case are absolutely impertinent, because they are not proper to convince the mind. Neither the profession of any articles of faith, nor the conformity to any 17 18 19

See e.g. John Marshall, John Locke: resistance, religion and responsibility (Cambridge University Press 1994); also Andrew Murphy, Conscience and Community. See Jonas Proast, The Letter Concerning Toleration Briefly Considered and Answer’d (Oxford 1690). Of course, this would still not mean that the ruler must be acting irrationally, so long as he (the ruler) thought it was possible to induce belief in this way. Below I give independent reasons for thinking it is not irrational.

Locke on toleration

109

outward form of worship, can be available to the salvation of souls, unless the truth of the one, and the acceptableness of the other unto God, be thoroughly believed by those that so profess and practise. But penalties are no ways capable to produce such belief. It is only light and evidence that can work a change in men’s opinions.20

This position was reiterated by a number of defenders of toleration among Locke’s contemporaries.21 Critics have frequently pointed out examples of this supposedly impossible feat.22 In some cases it is indeed impossible forcibly to induce belief: for example, it is plausible to think that I cannot force myself to believe that p by sheer effort of will, if I in fact believe that not-p.23 But it does not follow that one person cannot forcibly bring another to form a belief. This is true a fortiori if force includes the use of propagandistic, subliminal and other non-rational methods of inducing beliefs. Along with advertising and propaganda, education would be pointless if the intentional induction of a belief that p in someone currently disposed to believe that not-p were impossible. Insofar as Locke’s argument rests on an empirical generalisation about the ontogenesis of belief, it is unsound. The argument, moreover, can be made to cut the other way, as Hobbes observed long before Locke made it.24 Suppose that belief cannot be induced forcibly, and that salvation requires the sincere inward

20 21

22

23

24

Locke, A Letter Concerning Toleration, in Locke, Letters on Toleration (Bombay: Education Society 1867), 7. Other exponents of this view included writers such as Philipp van Limborch. See John Marshall, John Locke, Toleration and Early Enlightenment Culture (Cambridge University Press 2006), 647ff. See e.g. Jeremy Waldron, ‘Locke: toleration and the rationality of persecution’, in Susan Mendus (ed.), Justifying Toleration: conceptual and historical approaches (Oxford University Press 1988); Paul Bou-Habib, ‘Locke, Sincerity and the Rationality of Persecution’, Political Studies 51:4 (2003), 611–26; Micah Schwartzmann, ‘The Relevance of Locke’s Religious Arguments for Toleration’, Political Theory 33:4 (2005), 678–705. Even this claim needs some qualification, since it is not impossible that someone could set herself, as a long-term goal, the acquisition of the belief that p when she currently believes that not-p. An example is the regimen which Pascal envisaged for the unbeliever who recognises that the wager on salvation requires religious belief (which is then acquired through developing habits of worship, etc.). See Hobbes, Leviathan, ch. 42, 343–4, discussing the case of Naaman the Syrian. As I interpret him, Hobbes makes of the inviolability of private conviction a ground for religious uniformity, since given this inviolability, nothing the sovereign can do can jeopardise the subject’s chances of salvation insofar as the latter depends on private conviction. See my Hobbes and ‘Leviathan’, 151ff. A contrasting view is given by Alan Ryan, ‘Hobbes, Toleration, and the Inner Life’, in David Miller and Larry Siedentop (eds.), The Nature of Political Theory (Oxford: Clarendon Press 1983), and Ryan, ‘A More Tolerant Hobbes?’ in Mendus (ed.), Justifying Toleration; see also Tuck, ‘Hobbes and Locke on Toleration’.

110

How not to tolerate religion

conviction of the believer. Then the believer’s salvation is immune from whatever earthly penalties may be inflicted by the ruler in order to induce orthodox belief. So, just for that reason, the defender of toleration cannot argue that a ruler who does inflict penalties on the heterodox must be endangering their immortal souls. If doxastic impossibility refutes the argument that rulers may attempt to procure salvation through intolerance, it must also, according to Hobbes, refute the argument that trying to change people’s minds at the point of a sword will tend to their damnation. The mere fact of a grave external threat cannot change what those subject to it actually believe, so insofar as salvation depends on belief, it will not be jeopardised by external compulsion. However, an act may be irrational not because it is impossible: it may be irrational because the only available means of doing it removes the point of the action, for example by destroying the basis for its having value. On this reading, Locke’s argument relies on the claim that the value of religious belief depends on its not being forcibly induced. The religious believer can only attain salvation through an inward movement of conscientious conviction, as a free assent of the believer’s spirit. So, whatever beliefs the subjects may be frightened into accepting, a belief that merits salvation cannot be achieved by the secular ruler’s threats of persecution. Sometimes Locke seems to have this argument in mind. The power of using force to bring men to believe in faith and opinions and uniformity in worship could not serve to secure men’s salvation, even though that power were in itself infallible, because no compulsion can make a man believe against his present light and persuasion, be it what it will, though it may make him profess indeed. But profession without sincerity will not set a man forwards in his way to any place.25

Thus glossed, as a case for toleration, the argument still fails to convince. Even if it were impossible forcibly to induce belief that had the value attaching to unforced conviction, it would not follow that the ruler’s attempts to secure conformity in the outward tokens of belief must be in vain. To this extent Locke’s argument misses the point. He assumes that the ruler’s uppermost concern must be the actual credal states of the subjects, rather than the behaviour which expresses those states. But the would-be intolerant ruler may care less about inward belief than outward conformity. The motive for intolerance may be not that the ruler wishes to engineer subjects’ salvation, by ensuring that they enter the afterlife equipped with true religious beliefs, but to ensure that subjects observe 25

Locke, ‘Toleration D’, in Political Essays, ed. Mark Goldie (Cambridge University Press 1997), 276.

Locke on toleration

111

the official religion – and few would dispute that outward behaviour, as distinct from belief, is coercively modifiable. So the secular ruler may well have business other than subjects’ salvation in mind. The rulers themselves might not even believe in the official religion but enforce it anyway, as with the Roman auguries discussed by Machiavelli, to win acceptance of the political regime.26 If so, the case for toleration needs a further claim of Locke’s, often treated as a separate argument – that religion and politics are fundamentally distinct, and the methods of one cannot be used to promote the goods of the other. But this would still not give the persecuting ruler reason to tolerate religious beliefs that threaten the state with sedition. In assuming that the ruler cares about salvation rather than security, Locke blurs the very distinction which he takes pains to draw elsewhere in the Letter. I have set out the familiar version of Locke’s doxastic argument, and the less familiar version from the value of religious conviction, to show how clearly they rely on an appeal to religious rather than political goods. The same goes for the argument that if forcible indoctrination were possible, it would make salvation a lottery. This is not to say that Locke’s arguments for toleration always appeal to religious goods. The major political consideration in support of toleration is that persecution proves counterproductive: the victims of intolerance will be embittered against a regime which denies them religious freedom.27 If so, a would-be persecutor has prudential reasons for tolerating religious heterodoxy.28 This is a political argument. But it can point either way. The claim that intolerance will jeopardise security by embittering its victims against the regime can always be countered by the argument that tolerating sedition cuts at the foundations of the state. The security-based case for intolerance is often spurious or at least unpersuasive, but it is hard to think of a general argument that will show that it can never succeed. Perhaps commentators have been set on a false trail by failing consistently to adopt the standpoint of the ruler, for whom the value of belief, rather than the fact of it and its behavioural consequences, may be of little account. At the same time, the doxastic argument vacillates between the perspective of ruler and ruled: Locke seems to adopt the ruler’s perspective in arguing that enforcing doctrinal orthodoxy is bad in purely 26 27

28

Niccolò Machiavelli, Discourses on Livy, tr. Harvey Mansfield and Nathan Tarcov (University of Chicago Press 1996). Book I, xiv. The converse form of the argument – that tolerant regimes will be self-reinforcing – figures prominently in Rawls’s account of stability in the theory of toleration he presents in Theory of Justice, §34, §35. A modern version of this argument is put by John Rawls in his argument for the stability of the principles of justice in Theory of Justice, ch. 8.

112

How not to tolerate religion

self-interested terms. He tries to present the coercive induction of belief as a practical impossibility for the ruler. But, as I have suggested, this matters only if inward conviction rather than outward conformity is what the ruler aims to achieve. And that is unlikely to be true if the main concern is not with the individual believer, but the effects which belief, or its profession, will have in society at large.29 It is, then, not surprising that commentators should have failed consistently to take a ruler’s view of the argument, since it is a standpoint which Locke himself only sporadically adopts. The Letter, in particular, wavers between the standpoints of Realpolitik and soteriology. This is clear when Locke discusses his famous exclusions from the scope of toleration. With regard to atheists, Locke could be seen as consistent: the atheist has no soteriological beliefs, and as such the value attaching to religious conviction could be thought to be absent. But in fact Locke makes little of this point and rests his case on an appeal to what he regards as the atheist’s self-evident perfidy30 – even though, as has often been pointed out, one could argue that only atheists’ motives for moral action are untainted by the self-interest of seeking other-worldly salvation, and indeed they alone are capable of undertaking moral action for its own sake. With Roman Catholics, however, the case is different. Presumably the argument which rests on an appeal to the value of uncoerced religious belief will apply as much to Catholics as to, say, Dissenters. So Locke should be able to argue that no value will attach to the enforced orthodoxy of would-be recusants. But his argument appeals only to the threat to the security of the state which is posed by Catholic fifth-columnists. At this point, Locke’s arguments – his appeal to the distinctness of political and religious goods, to the value of uncoerced conviction, to the arbitrariness of resigning one’s eternal destiny to the doctrinal whims of the secular ruler, to the counterproductiveness of efforts to root out ‘seditious’ religions, and so on – cede to raison d’état. This is despite the fact that Locke notes that Roman rituals are, in themselves, no more deleterious to the standing of the state than are Protestant ones.31 29

30

31

Alex Tuckness argues in his ‘Locke’s Main Argument for Toleration’, in Williams and Waldron (eds.), Toleration and Its Limits, that Locke came, especially in the Third Letter, to adopt a ‘universalisation’ argument for toleration: since all humans are fallible, there is insufficient ground to believe that God would have authorised earthly rulers to impose what they regard as the true religion by force. ‘Promises, covenants, and oaths, which are the bonds of human society, can have no hold upon an atheist.’ Locke, Letter, in John Locke, A Letter Concerning Toleration: in focus, eds. John Horton and Susan Mendus (London: Routledge 1991), 33. ‘Is it permitted to worship God in the Roman manner? Let it be permitted to do it in the Geneva form also. Is it permitted to speak Latin in the market-place? Let those that have a mind to it, be permitted to do it also in the church.’ Locke, Letter, 35.

Locke on toleration

113

A jaundiced reading of Locke might represent the Letter as putting the ideological cart before the justificatory horse. It would accuse him of first lighting on political positions – such as the desirability of ‘comprehension’ for Nonconformists, and the intolerability of Catholicism – and then casting about for philosophical vindications of them. Locke’s defence of the status quo is prior, on this reading, to the justification cited in its support.32 But it is more interesting, as well as hermeneutically charitable, to seek underlying consistency beneath the appearance of contradiction. And in fact a consistent position is not hard to elicit. The consistency lies in Locke’s insistence on reason of state as a trump card in arguments over toleration. Locke gives an example of such an argument in his exclusion of atheists from the scope of religious toleration. Modern readers are likely to be struck by the implausibility of Locke’s case for excluding atheists. It is indeed hard to believe that atheists’ failure to be cowed by the prospect of punishment in the hereafter will make them generally untrustworthy here and now.33 The reason why the grounds for this exclusion are faulty lies not in the invalidity of the argument, but in its reliance on a false factual premise, the claim that atheists should rationally be treated as dangerous to the state. In other words, though unsound, the argument is valid, having the following form: 1. If the state rationally judges that some person(s), group, or doctrine will prevent or gravely threaten its survival, it is justified in acting intolerantly towards them. 2. The state rationally judges that person(s), etc. P (such as atheists) will prevent or gravely threaten its survival. So 3. The state is justified in acting intolerantly towards P. The major premise of Locke’s implicit argument – that if someone poses a grave threat to the survival of the state, he cannot be tolerated – remains persuasive. If atheists really were perfidious, and as such a threat to security, the case for withholding toleration from them would be strong. It is not the denial of toleration to the seditious which undermines 32

33

For a parallel argument with regard to western discourses on toleration, see Wendy Brown, ‘Tolerance as/in Civilizational Discourse’, in Williams and Waldron (eds.), Toleration and Its Limits; see also Chapter 9. In the course of their polemics Proast uses Locke’s opposition to tolerating atheism in order to gain leverage against his toleration of Dissenters, arguing that the latter would lead willy-nilly to the proliferation of atheism. See Locke, A Third Letter Concerning Toleration, in Locke, Letters on Toleration, 386–7.

114

How not to tolerate religion

Locke’s argument, but the fact that no good ground exists for thinking that atheists will inevitably foment sedition, as his contemporary Bayle argued, contrary to the donnée of the time.34 The widespread acknowledgement that Locke’s justification for refusing toleration to atheists fails has obscured the merits of his argument for doing so. The argument above is plainly valid. The false factual premise mentioned earlier arises when ‘Atheists’ is substituted for ‘P’ in premise 2. Political power can be reasonably exercised against atheists, in Locke’s view, because not doing so will undermine the state, and therewith subvert the very basis for exercising power.35 Similar remarks apply to Roman Catholics, as we shall see. Nor is Locke’s position here merely the expression of some clannish Protestant prejudice against the unreformed religion. For ‘perhaps the Quakers, were they numerous enough to become dangerous to the state, would deserve the ruler’s care and watchfulness to suppress them’.36 As we have seen, Locke does not think that the purely doctrinal content of Roman Catholicism (such as belief in the Real Presence) is seditious, insofar as this is separable from a particular view about the due objects of secular allegiance. The more general point is that toleration lasts only for as long as the tolerated pose no danger to political stability. When any such distinct party is grown or growing so numerous as to appear dangerous to the ruler and seem visibly to threaten the peace of the state, the ruler may and ought to use all ways, either of policy or power, that shall be convenient, to lessen, break and suppress the party, and prevent the mischief.37

Unlike some of his contemporaries,38 Locke accepts that in principle the Jewish religion should be tolerated. ‘What hinders but that a Christian ruler may have subjects that are Jews?’ Locke asks rhetorically;39 if ‘a Jew do not believe the New Testament to be the word of God, he does not thereby alter anything in men’s civil rights’.40 Indeed, ‘Not even [native] Americans, subjected unto a Christian prince, are to be punished, either

34 35 36 37 38 39 40

Pierre Bayle, Pensées diverses sur la Comète, ed. A. Prat (Paris: facsimile reprint Société des Textes Français Modernes 1912), e.g. 474; cf. Forst, Toleranz im Konflikt, 325ff. For reflections on the treatment of atheists in contemporary writers, see Forst, Toleranz im Konflikt, 311, 323f. Locke, Essay Concerning Toleration, in his Political Essays, ed. Mark Goldie (Cambridge University Press 1997), 148. Locke, Essay Concerning Toleration, 147. See Marshall, John Locke, Toleration and Early Enlightenment Culture, ch. 19. Locke, Letter, 26. Locke, Letter, 35. Locke argues in the Essay that the variation of the Sabbath between Christianity and the Muslim and Jewish religions should not be the subject of civil penalties. Locke, Essay Concerning Toleration, 139.

Locke on toleration

115

in body or goods, for not embracing our faith and worship.’41 In sum, ‘neither pagan, nor Mohametan, nor Jew ought to be excluded from the civil rights of the commonwealth because of his religion’.42 But this toleration has a clear limit. That limit is reached when the doctrine or practice of a religion delivers its adherents into the hands of a foreign prince, or otherwise threatens the commonwealth. Locke reserves the clearest statement of his position in the Letter when arguing for the toleration of Muslims: It is ridiculous for any one to profess himself to be a Mahometan only in religion, but in everything else a faithful subject to a Christian ruler, whilst at the same time he acknowledges himself bound to yield blind obedience to the Mufti of Constantinople, who himself is entirely obedient to the Ottoman emperor.43

Similar justified grounds for denying toleration hold, in Locke’s view, with regard to Roman Catholics. Like Muslims, Romanists did indeed acknowledge the dominion of a foreign prince, the Pope. In the 1667 Essay, Locke considers the relation between secular and religious doctrines. He argues that, at least in the case of Catholics, these cannot be kept apart. Accordingly, Catholicism must be intolerable: not because of its doctrinal content, but because of Catholics’ divided allegiances. Here, at least, the fact that belief cannot be induced forcibly, or that its value depends on not having been elicited by force, no longer matters. Nor, it seems, does it matter here that the forcing of men’s consciences makes their salvation depend on an accident of birth. It often happens that [people] mix with their religious worship and speculative opinions other doctrines absolutely destructive to the society wherein they live, as is evident in the Roman Catholics that are subjects of any prince but the pope. These, therefore . . . ought not to be tolerated by the ruler in the exercise of their religion unless he can be secured that he can allow one part without the spreading of the other, and that those [i.e. relating to secular matters] opinions will not be imbibed and espoused by all those who communicate with them in their religious worship, which, I suppose, is very hard to be done.44

The problem Catholicism posed for Locke’s official position is clear. That position relies heavily on a sharp distinction between secular allegiance and religious belonging. The distinction supports Locke’s view 41 42 43 44

Locke, Letter, 30. Locke, Letter, 46. For similar remarks on Muslims, see A Third Letter Concerning Toleration, e.g. 275. Locke, Letter, 41. Locke, Essay Concerning Toleration, 146. The passage ‘and that those . . . religious worship’ replaced the manuscript version, which at this point has ‘and that the propagation of these opinions may be separated from their religious worship’.

116

How not to tolerate religion

that Jews, Muslims and pagans may be tolerated when they pose no threat to political authority. But, of course, certain religious doctrines themselves make claims about who should hold secular power, as illustrated by the modern Catholic doctrine of papal infallibility and that secular sovereigns may lawfully be deposed.45 So the distinction cannot justify the toleration of Catholics, to the extent that they subscribe to seditious religious doctrines. Rather the distinction overlays a more fundamental one in Locke’s argument, between those who do and those who do not threaten the survival of the state. In fact Locke gives yet a further justification for toleration in the Essay,46 in a passage referenced in Locke’s index as ‘Toleration: necessary in our state of knowledge’,47 although the term ‘toleration’ itself does not appear in the text at that point. The argument appears in a chapter on ‘degrees of assent’ to propositions. Locke argues that people’s beliefs, most of the time, rest on reasons which remain implicit and cannot be fully spelt out under challenge. It is unreasonable to expect that anyone should ‘obsequiously quit his own opinion, and embrace ours with a blind resignation to an authority, which the understanding of man acknowledges not’.48 Insofar as this differs from the doxastic impossibility argument, it rests on a normative claim: it is unjustifiable to impose epistemic authority on another, given that everyone has beliefs that he cannot fully justify. But if applied generally, this argument licenses much wider freedom of opinion than Locke countenances in his political writings. For Locke, religious and political goods are indeed distinct, despite his failure (as I have contended) to maintain a clear argumentative distinction between them. But as often with arguments from plural goods, the practical decision can go one way or the other in cases of conflict. Locke’s argument is that unless there is a clear and present threat to the security of the state, the good of uncoerced religious conviction should be allowed to flourish. For atheists, of course, there is no such good to be had, in Locke’s view, while the threat remains, as in their own minds atheists can face no eschatological sanction for breaches of faith.

45

46 47 48

Locke seems to have grappled with the practicalities of enforcing this distinction in respect of Catholics. He is the likely author of The Particular Test for Priests, a manuscript (Locke, Political Essays, 222–4; cf. Marshall, John Locke, Toleration and Early Enlightenment Culture, 689) devoted to setting out a test which Roman Catholic priests might take to demonstrate their allegiance to a Protestant sovereign. Locke, An Essay Concerning Human Understanding, ed. Peter Nidditch (Oxford: Clarendon Press 1975), Book IV Ch. XVI §4, 659f. Essay Concerning Human Understanding, 746. Essay Concerning Human Understanding, 660.

Locke on toleration

117

Catholics, on the other hand, owe their allegiance to a foreign temporal prince – not merely to one whose kingdom is not of this world. Thus Locke equivocates over whether or not to tolerate religious deviance. The arguments he gives prove either more or less than he needs to support his political position. The doxastic impossibility argument seemingly extends toleration to some from whom Locke wishes to withhold it, such as atheists, since they cannot be coerced into theistic belief. Meanwhile, his argument from a strong distinction between church and state threatens to deny toleration to those, such as Roman Catholics, to whom Locke, at least in certain moods, wishes to grant it. The uncertain status of his arguments for toleration is exemplified by his wavering between the standpoint of the religious believer and that of the ruler. When he unambiguously adopts the latter’s perspective, Locke’s arguments – whether for toleration or against it – are rooted firmly in reason of state. The arguments then face a constitutive tension between their political point and what they show in the abstract. Compare Spinoza’s arguments for free speech in the Tractatus Theologico-Politicus. There, like Locke, he puts an epistemic argument for free speech.49 But Spinoza also argues for free speech via what might be called the Bootleg Argument: that legal prohibition is pointless since it will drive heterodox speech underground.50 Opinion will out. However, insofar as the Bootleg Argument succeeds, it fails. If prohibition will fail anyway, there is to that extent no reason to oppose it. Otherwise, prohibition proves more or less successful in stamping out heterodoxy, and to that extent the policy works. In other words, the Bootleg Argument fails to justify the policy it aims to support. At stake is a partly symbolic argument about what can be integrated within the polity, rather than a causal dispute about what will result from a given policy. I argued above that a similar flip-flop marks the recent debate over the introduction of Sharia law in the UK. Legal arguments can be run either way, to explain how incorporating Sharia either would wreck the integrity of the legal system, or could be fitted with the devolved structures already embodied in legal practice. I suggested that these arguments about law, just because they can point either way, conceal a more basic disagreement over what is tolerable. Ultimately this 49

50

Spinoza, Tractatus Theologico-Politicus, in Spinoza, The Political Works, ed. A. G. Wernham (Oxford: Clarendon Press 1958), 227. But cf. ch. XVII, where Spinoza accepts that the sovereign can in fact induce mental as well as verbal assent, e.g. 151: ‘sunt tamen animi aliqua ratione sub imperio summae potestatis, quae multis modis efficere potest ut permagna hominum pars quicquid vult credat’. Spinoza, Political Works, e.g. 237: ‘tales leges inutiles omnino sunt; nam qui opiniones quae legibus damnatae sunt sanas esse credent, legibus parere non poterunt’.

118

How not to tolerate religion

disagreement concerns which persons, groups or doctrines can be encompassed within a unitary polity. Here again, the argument is driven by reason of state: how far should we, the majority or the powerful, abide the presence of these others in our midst?

Liberalism after Locke In some ways despite himself, Locke is strongly drawn to the idea of salvation as the basic good. That is why the doxastic argument attracts him. However, the argument relies on the assumption that the political and religious Ur-goods do not conflict with one another. When they do – that is, when the absolute separation between political and religious goods proposed in the Letter proves unsustainable – Locke reasserts the primacy of the state over the individual conscience. The political arguments from public order then come to the fore, that intolerance is counterproductive even when judged purely on a political basis. Of course, even here the argument risks doing more than Locke wants, given that he aims to exclude those whose allegiance is directed either to foreign princes, such as Catholics, or to no prince at all, namely atheists. Intolerance is held preferable to risking the dissolution of political society.51 One implication of this reading is that Locke cannot as readily eschew the making of specific religious truth-claims as he would like. Some commentators discern in the Letter and his subsequent polemics with Proast an argument of the following form:52 we cannot know what true religion is; since there is only one true religion and many false ones, it is more likely that an earthly prince will impose a false religion rather than the true one; but imposing a false religion on his subjects will ensure their damnation;53 therefore the ruler should impose no religion at all.54 It may be that Locke held that – at least at a suitably high level of doctrinal resolution – religious truth was not to be had by fallible human beings. This seems to be his position, for example, on variations in forms of 51

52 53 54

This conclusion is of course the more striking given that Locke – at least the Locke of the Second Treatise of Civil Government, which was composed five or six years before the Letter – did not believe that civil society should be preserved at all costs. Locke seems drawn to this argument in some passages in the lengthy Third Letter. As already noted, of course, this premise is at odds with the doxastic argument for toleration. Tuckness sees this as the strongest of Locke’s arguments for toleration. In general, it faces problems familiar from other attempts to justify toleration on sceptical or fallibilist grounds (see my Virtue, Reason and Toleration, ch. 5); the point here, however, is that these grounds are not available to Locke, since what drives his arguments about toleration are specific truth-claims about what doctrines, and what groups, pose an unacceptable threat to the state.

Liberalism after Locke

119

worship between different churches.55 But the argument from religious scepticism or fallibilism is overshadowed by Locke’s concern with the stability of the polity. He has to decide between the value of free exercise – where this may raise the spectre of sedition – and that of political security. Of course, if the religion thus suppressed were the true one, Locke would presumably regard this as the ultimate bad. It is because he does not regard as true those religions which suffer civil penalties because they threaten the basis of political order, that he thinks suppressing them is justified. I have suggested that Locke’s ambiguity about the toleration of minority religions parallels latter-day wavering on the same subject. In each case, the ostensible rationale for inclusion or exclusion proves to rest on claims about security – about whether or not the polity can absorb the inclusion of marginal groups or doctrines. This is not to say that the argument should not rest on this ground. On the contrary, the absolute limit of the tolerable must lie where failing to exercise intolerance would undermine the basis for exercising tolerance itself. The tolerance frontier, defined in terms of what the state can accept without ceasing to exist as an executive agency, trumps other arguments about toleration. The absolute political value of security outweighs appeals to other values when the survival of the polity really is in question. The peculiar challenge posed by religion now becomes clear. The arguments rehearsed above about tolerating religious minorities – be they modern-day Muslims, or early modern Catholics, atheists, Jews and pagans – are double-edged. They can be pressed into service to support the inclusion of one group or, handy-dandy, the exclusion of another. This is true of arguments about toleration generally, as I have argued elsewhere.56 However, the case of religion poses special problems of its own. This is because religion concerns itself with ultimate goods – with what has absolute value, overriding and remaining independent of other valuable things. In general terms, the threat posed by religion to the state is clear: it has the resources to challenge the very basis of the political order. This fact both underlies Locke’s insistence on separating politics and religion, and makes that separation untenable. For the state, security constitutes the basic good – the summum bonum or at least the sine qua non of political life. That means that the state has to give priority to underwriting the conditions of public order. 55

56

‘If we consider right, we shall find that for the most part they are such frivolous things as . . . without any prejudice to religion or the salvation of souls, if not accompanied with superstition or hypocrisy, might either be observed or omitted.’ Locke, Letter, 20. Newey, Virtue, Reason and Toleration, ch. 5; also Chapter 3.

120

How not to tolerate religion

For (Christian) religious believers, however, security ultimately consists in salvation. It is this rival notion of security which makes religion such a problem for the state – and conversely: each threatens the other’s summum bonum. Schematically put, the state can be seen as threatening salvation, i.e. religious security, while religious goods may threaten state security. So religious toleration is double-edged: on the Lockean argument it should be extended to religions to buttress secular authority, even while those for whom salvation is the highest good cannot offer wholehearted allegiance to it. This was true, or thought to be true, of Roman Catholics in the late seventeenth century and is true of some Islamic groups today. Religion casts a long shadow over the modern liberal project of political justification. This is not because some specific religion is true. It is rather because religious people think that their favoured religious doctrine is true, and draw the consequences regarding ultimate goods. I may believe some religion and as a result believe that I have a duty of proselytism towards infidels, even of converting them at the point of a sword. Of course, most people in liberal democracies do not take this view. But that is not because some general or abstract notion of the reasonable shows that doing so is contrary to reason.57 It is because such a view runs counter to the embodied standards of reasonableness in public institutions and opinion. That is, there is no general or abstract standard by which the person who regards salvation in the hereafter as more important than security in the here and now can be judged unreasonable. The argument becomes more pointed still given modern liberals’ willingness to accept that religious belief, with its attendant eschatological or soteriological content, is reasonable in itself.58 This has serious implications for modern liberal defences of religious toleration. For modern liberals such as Rawls, Barry and Scanlon,59 political power, including power in matters of religion, may justifiably be exercised only on grounds which nobody could reasonably reject.60 This constraint on justification extends to the enforcement of the legal 57 58

59 60

Liberal theories such as Rawls’s which purportedly work from such a conception make the disposition to accept liberal norms criterial of reasonableness. As Rawls explicitly argues; see his Political Liberalism, 52ff. Rawls reiterates the position elsewhere in his later writings, e.g. The Law of Peoples (Oxford University Press 1999), 16 and fn. 8; also ‘The Idea of Public Reason’ and ‘The Idea of Public Reason Revisited’ in Rawls, Collected Papers. See Barry, Justice as Impartiality; Scanlon, What We Owe to Each Other. Similar remarks apply to Rawls, whose argument from the burdens of judgement in Political Liberalism is designed to show that it would be unjustifiable to institute a political regime based on any single ‘comprehensive doctrine’, since it is in the nature of such doctrines that they may reasonably be rejected.

Liberalism after Locke

121

order via the exercise of political power. Reasonable rejectability is taken to license the orthodox position. But it seems that religious doctrine may lead individuals reasonably or, at least, not unreasonably, to reject the basis on which political power is justified, simply because that basis ignores the goods which they regard as ultimate. The reasonable rejectability basis for justification cannot do its job of justifying liberal politics without relying on a conception of the reasonable which has already been filtered for that very purpose. If so, security either trumps the reasonable rejectability criterion of justification, or else it constrains the content of the reasonable in such a way that it justifies the repression of seditious doctrine.61 The fact that some religious doctrine is such that all other doctrines may reasonably be rejected in favour of it must be consistent with its being reasonable to reject that doctrine itself, since on the standard view all such doctrines are reasonably rejectable. And, in censoring such a doctrine in the name of security, the state is not asserting the truth of any specific competing religious doctrine, which would be contrary to the dictates of neutrality. If it is held to be unreasonable to use state power to suppress any doctrine which is reasonable (not unreasonable), there is nothing to preclude the judgement that seditious doctrine is, in its nature, unreasonable. The rejection of the political order, for example by excluded religions, undermines the justification of their exclusion – unless that rejection is unreasonable, as would widely be thought to be true of fundamentalist imams in the UK. However, what is reasonably rejectable varies circumstantially. Often security is seen as the sine qua non of political action. But what ‘security’ encompasses is itself an intensely political question. That this is so provides further reasons for thinking that the project of political design, conceived as a ground-up philosophical attempt to shape political institutions and procedures, must be misguided. The task of deciding what limits to impose on political inclusion must begin in medias res. Any attempt to foreclose the issue, for example by pronouncing certain agents or attitudes as reasonable and others as not, will serve only to politicise the values or concerns underlying the conception of the reasonable. As a result, political philosophy is, more than is usually realised (for example, by political philosophers), an act of bricolage. It is plausible to think that security involves the protection of fundamental goods. But, again, the content of those goods is not given. The difficulty posed by the circumstances of religious toleration is that they

61

See Chapter 4.

122

How not to tolerate religion

show both that the fundamental goods are not given, and that basic political conflicts may arise from attempts to specify them. Insisting on a firm separation between the goods of politics and religion, as Locke does, does not advance matters much. It is likely merely to repeat the original conflict. That is perhaps why Locke finds himself tempted by the thought, offered by the doxastic argument, that ultimate religious goods can coexist with political ones: it offers the comfort of thinking that the conflict can be defused. Conclusion To this extent, the twenty-first-century state finds itself in a bind curiously familiar from the early modern period, between ‘comprehension’, or inclusiveness, and suppression. At a purely descriptive level, this could be seen as having to decide whether security is better served by religious tolerance or intolerance. But it would be mistaken to think that such decisions can be made without normative commitments. The state has to decide when security demands the proscription of seditious malcontents, rather than according them civic equality. Since judgements about security always involve deciding what is valuable, the state has to decide when, for example, justice or equality has to be sacrificed or compromised in order to protect other goods. The ‘fundamental’ principle of civic equality is thus contingent on safeguarding civic peace. For us now as for Locke, the task of the secular ruler is to determine how – and on what terms – this is to be done.

7

Liberty, toleration, security

It is often assumed that the values of toleration and security must be inimical to one another, and it is not hard to see why. Toleration is a form of freedom, in the sense that those who are tolerated enjoy freedom that they would not enjoy were they not to be tolerated. On a widely held view, freedom (or liberty)1 and security are inimical to one another, or at least have to be ‘balanced’.2 As a result, there is a prima facie conflict between security and toleration. To tolerate sedition is often thought to cut away the very foundations of political order. Hobbes famously wrote in the ‘Review and Conclusion’ to Leviathan that ‘the toleration of a professed hatred of Tyranny, is a Toleration of hatred to Commonwealth in general’.3 Indeed, some thinkers have seen security as setting the limit to the tolerable. Rawls argued that although ‘an intolerant sect does 1 2

3

I use these terms interchangeably. E.g. Stuart Taylor, ‘Rights, Liberties, and Security: recalibrating the balance after September 11’, www.brookings.edu/articles/2003/winter_terrorism_taylor.aspx (Brookings Institution); Eric A. Posner and Adrian Vermeule, Terror in the Balance: security, liberty, and the courts (Oxford University Press 2006); Tony Smith, ‘Balancing Liberty and Security? A Legal Analysis of United Kingdom Anti-Terrorist Legislation’, European Journal of Criminal Policy and Research 13:1/2 (2007), 73–83; Ben Golder and George Williams, ‘Balancing National Security and Human Rights: assessing the legal response of common law nations to the threat of terrorism’, Journal of Comparative Policy Analysis 8:1 (2006), 43–62; Paul De Hert, ‘Balancing Security and Liberty within the European Human Rights Framework’, Utrecht Law Review 1:1 (2005), 68–96; Darren W. Davis and Brian D. Silver, ‘Civil Liberties vs. Security: public opinion in the context of the terrorist attacks on America’, American Journal of Political Science 48:1 (2003), 28–46; Richard Posner, ‘Security versus Civil Liberties’, Atlantic Monthly (1 December 2001), 46–7. For more sceptical views on the notion of balance, see Jeremy Waldron, ‘Security and Liberty: the image of balance’, Journal of Political Philosophy 11:2 (2003), 191–210; Mark Neocleous, ‘Security, Liberty and the Myth of Balance: towards a critique of security politics’, Contemporary Political Theory 6:2 (2007), 131–49; Andrew Ashworth (ed.), The Criminal Process: an evaluative study, 2nd edn (Oxford University Press 1998); Christopher Michaelsen, ‘Balancing Civil Liberties Against National Security?’ University of New South Wales Law Journal 29:2 (2006), 1–21, and his ‘Liberty versus Security: rightsbased objections to a simple balancing exercise’, in Christian Enemark (ed.), Ethics of War in a Time of Terror (Canberra: Australian National University Press 2006), 41–54. Hobbes, Leviathan, 486.

123

124

Liberty, toleration, security

not itself have title to complain of intolerance, its freedom should be restricted only when the tolerant sincerely and with reason believe that their own security and that of the institutions of liberty are in danger’.4 It is clear both how such a case can be made, and that the case can readily be used to serve ideological ends.5 As with other abridgements of freedom, toleration can be curbed on the plea that the republic is in danger. Talk of ‘zero tolerance’ often relies on the claim that intolerance is needed because toleration will destroy fundamental goods, especially security; much of the talk comes from those with a commercial stake in security ‘solutions’ in public places such as schoolyards and airports. I shall argue, however, that identifying more security with less toleration is misconceived.

Liberty or security, or security with liberty? Talk of ‘balance’ is long-established in discussions of security, despite the more nuanced account of the relation between liberty and security in historic texts by such authors as Hobbes and Spinoza, and The Federalist Papers by Madison, Jay and Hamilton. But the metaphor of balance is odd in itself.6 To strike a balance between two things is to place them in equilibrium. If one thinks literally of a set of scales, it is balanced when whatever is in the one scale weighs the same as whatever is in the other. That means that if the scales are initially in equilibrium, and then more is put into one scale, an equal extra weight has to be put into the other scale in order to restore equilibrium.7 But in discussions of public policy, say in relation to concerns over terrorism, talk of ‘balance’ generally accompanies the proposal that a greater measure of security must be bought at the cost of a loss in liberty or – less often – the converse.8 If more security is needed to be in ‘balance’ with liberty, then they must have been out of balance to start with. Sometimes, incoherently, the idea is advanced that one ‘balance’ between security

4 5 6 7

8

Rawls, Theory of Justice, 220. Mark Neocleous, Critique of Security (Edinburgh University Press 2008). See Waldron, ‘Security and Liberty: the image of balance’. If security and liberty are incommensurable goods, there is no measure to weigh security against liberty, so talk of balancing is empty. See Joseph Raz, The Morality of Freedom, ch. 13. In a speech to the Security Commission of the Institute for Public Policy Research in June 2008 the UK Prime Minister endorsed the House of Commons Home Affairs Select Committee’s call ‘for both practical measures and principled commitments from the Government to ensure the balance of liberty and security is maintained’; www.number10.gov.uk/Page15785.

Liberty or security, or security with liberty?

125

and liberty needs to be replaced with a different one, for example in response to a perceived rise in security threats.9 Accordingly, the notion of ‘balance’ needs to be reinterpreted or jettisoned. On the underlying idea of a trade-off between liberty and security, increasing security will demand some compensating loss in liberty. For example, granting the police additional powers to arrest and detain terrorism suspects without charge must diminish civil liberties, such as the liberty not to be held without charge; similarly, restricting certain kinds of ‘hate’ speech may be justified on the ground that the speech may incite acts of violence. Debate then follows about whether the extension of police powers really will serve to enhance security. It is accepted on both sides that this extension reduces the civil liberties of terrorism suspects. But, rather than enhancing security, it may be that measures of this kind diminish it. The point is not that the measure may fail to achieve its objective – perhaps because extending detention times is not an effective way of screening for terrorists, or because it may prove counterproductive by radicalising those who would otherwise not become terrorists. These are possible empirical consequences of the policy, but the policy may be thought objectionable in itself on security grounds. The point is that the diminution of liberty could be seen, by the same token, as constituting a diminution in security. Why? Summarily put, because if liberty itself is a good (as proponents of the balancing metaphor seldom deny), then it is something which people have reason to value and hence to pursue; but also, if liberty is a good, it is itself something in relation to which people can be secure or insecure, and which they can enjoy or fail to enjoy. A further respect in which it can seem that there is a necessary trade-off between liberty and security is by setting against one person’s or group’s gain in security the loss in liberty of another, or conversely. For example, when debating public policy, what is often in prospect is to enhance the security of one group, such as airline passengers, while diminishing the liberty of another, such as those suspected of terrorism offences. In these contexts it may be appropriate to talk of trading off a gain in security for the first group against a reduction in liberty for the second. But this does not show that there is some general incompatibility between security and liberty, such that a unit increase in security must demand some equal or proportionate cut in liberty. The mere fact that some 9

E.g. Paul Rosenzweig, ‘Balancing Security and Liberty’, paper published by the Heritage Foundation 14 May 2003: ‘Significant events like Sept. 11 alter the balance between liberty and security, but the pendulum always returns to center as the threat diminishes’; www.heritage.org/press/commentary/ed051403a.cfm.

126

Liberty, toleration, security

situations can be constructed in which a pair of values, such as liberty and equality, or democracy and justice, conflict with one another, does not show that an inverse relationship of this kind holds between the two values in general. If the values really are distinct, one would expect that incompatibilities between them will sometimes arise, but this is not to say that the values must be opposed to each other. Consider for example the long-running debate in the UK over the inclusion of innocent persons’ DNA on the national database maintained for the purpose of detecting and convicting criminals. DNA samples have been collected from crime scenes and from persons held in police custody. The number of profiles held on the database has grown fast in the early years of the twenty-first century.10 Its security use – to identify recidivists and other repeat offenders and match evidence at new crime scenes with a list of those known to have committed similar crimes previously – is obvious, and explains the UK government’s heavy investment in the supporting technology and infrastructure.11 However, civil libertarians have opposed the database, or at least the database profiling of all persons whose DNA is found by the police at crime scenes, irrespective of whether or not they are otherwise implicated in the crime; DNA from persons arrested but not convicted also remains on the database.12 Clearly it could be argued that, whatever the security benefits, the DNA database policy infringes individual liberties. As already argued, liberty itself is a good in respect of which one may be more or less secure. However, as regards each of the points raised above, a similar claim could be made about individuals’ security. The main point is that those on the database could be thought of as suffering reduced security because their presence on it pushed them up the list of suspects for the offences in relation to which their DNA profile had been entered on the database. For instance, insofar as inclusion made it more likely that someone would first be a suspect, then charged, and then tried for the offences in question, they would also be more likely to suffer a wrongful conviction – just because one can only be wrongly convicted if one is first tried, and

10

11 12

www.homeoffice.gov.uk/science-research/using-science/dna- database/. By March 2013, the figure for profiles held on the National DNA Database had risen to some 5.93 million, or about 9.6 per cent of the total UK population: www.gov.uk/government/ publications/national-dna-database-statistics. Estimated at £300 million between 2004 and 2009: www.homeoffice.gov.uk/scienceresearch/using-science/dna-database/. The profiling of innocent persons occurs in England and Wales, but not in Scotland, where the DNA of acquitted persons without an existing criminal record is destroyed.

Liberty or security, or security with liberty?

127

database inclusion made the latter, necessary condition for that more likely.13 That is, these persons suffer reduced security against miscarriages of justice. This is not to argue for or against the database either in principle or as regards the specific policies adopted towards it by the UK government. It is only to argue that the language of ‘liberty’ can often be translated into that of ‘security’ when discussing the effects of some government initiative, and this makes the notion of ‘balancing’ or trading one off against the other seem questionable. This is not surprising given that liberty itself is a valuable thing, which one can hold more or less securely.14 The upshot is that concerns over liberty can often be translated into the language of security. That one can be more or less secure as regards one’s holding of certain liberties is often urged by republican theorists against proponents of unitary sovereignty.15 Not that claims about freedom and security must always be interchangeable. For one thing, nothing has been said so far about the translatability of security claims into talk about liberty: so in some situations where a person has or enjoys security it may be inappropriate to talk about their having or enjoying liberty. For example, in an efficiently run jail the prisoners may have and enjoy security (e.g. from assault by other prisoners), but not liberty. It seems implausible to say that in this case the prisoners have liberty by dint of having security. Security can be understood in abstraction from the specific goods with regard to which security is enjoyed: freedom is one of these, but there are many others, such as physical integrity, the well-being of others such as family members, or access to resources. However, it is a categorial confusion to think that liberty is merely one more among a range of goods with respect to which one can have or lack security. Liberty could be thought of as an adverbial good of security: one can have security with liberty: that is, in being secure – having the good of security – one can do so with liberty, or not. What has gone wrong when the latter holds is a matter for further investigation, and a bone of contention, for example, between Hobbes’s critics and admirers.16 13

14

15 16

This is not to say that a person whose DNA was on the database when tried would be more likely to be convicted than another person on trial whose DNA was not. The base rate for comparison purposes is with the population generally, not with other defendants. Cf. Article 5 of the European Convention on Human Rights, ‘Right to Liberty and Security: 1. Everyone has the right to liberty and security of [their] person’. The article sets out liberties that convention signatories undertake to secure for their citizens. See e.g. Philip Pettit, Republicanism: a theory of freedom and government (Oxford University Press 1997). See Quentin Skinner, ‘Hobbes on the Proper Signification of Liberty’, repr. in Skinner, Visions of Politics, Vol. II: Hobbes and Civil Science (Cambridge University Press 2002).

128

Liberty, toleration, security

Similarly, security can be seen as an adverbial good of liberty. One can have the good of liberty with security, or not, as the case may be. Again, these possibilities go to suggest that the alleged incompatibility between security and liberty is a merely contingent matter. Security v. liberty qua toleration Still, toleration and security seem to stand on an uneasy footing with respect to one another. The history of toleration in the early modern period in Europe bears this out. For example, the limited freedom of religious worship granted to the Huguenots in 1598 under the Edict of Nantes by Henri IV following the French wars of religion marked an increase in freedom vis-à-vis a persecutory confessional Roman Catholic state. This became apparent eighty-seven years later, when Henri’s successor Louis XIV revoked the Edict, causing many Huguenots to flee into exile. The reasserted Anglican ascendancy of the early 1660s following the restoration of Charles II curtailed liberties enjoyed by Nonconformists following the victory in the civil wars and the wide religious toleration for Protestant sects during the Commonwealth regime. These restrictions were then eased by the Toleration Act in 1689 and, much later, by the repeal of the bar on Nonconformists holding public office in 1828. In each case, shifts in policy on toleration can be seen as marking a gain or diminution in freedom for the religious groups concerned. Similar remarks can be made about the extension of toleration to groups such as homosexuals in more recent times.17 As I argue in more detail in Chapter 8, judgements of toleration are vulnerable to the objection that, in the tolerator’s ideal world, the act, person or state of affairs that is tolerated would not exist. For example, if toleration rests on a principle of respect for persons, the tolerator may be pressed, as follows: whatever may be the case in this world, would you not prefer a world in which those whose acts you disapprove of refrained from performing them? If so, would you not choose, other things being equal, to create a world of this latter kind, rather than the one we are in? If the principle undergirding toleration is analysed with respect to counterfactuals that may void judgements of toleration, freedom or autonomy seems to be the only principle that can immunise such judgements 17

Of course, in each of these cases it can be argued that the greater liberty enjoyed by the previously persecuted minority was offset by reducing the liberty of the persecutors. I do not think that such claims can be dismissed out of hand, whatever the justifiability of the policy itself. But the point is that insofar as the claims are correct, they indicate not that toleration and liberty are out of step, but that liberty is a sufficiently complex value to be in conflict with itself.

Security v. liberty qua toleration

129

against counterfactual refutation. Otherwise the tolerator is always open to the claim that her ideal world would not contain the tolerated action. Only if the action is tolerated because of a principle of liberty, freedom, or autonomy will judgements of toleration be secured against counterfactual idealisation (and perhaps not even then). A further common line of thinking seems to be corroborated by this – the mutually inimical relation between security and toleration. Freedom and security are often thought to be at odds, both in the work of theorists and in the public political sphere.18 If the relation between toleration and freedom is as close as I have just argued, the tension between toleration and security is not surprising. To extend freedom to hitherto marginal or outlawed groups risks fostering subversion, since these groups have often encompassed those whose opposition to the established order, whether or not it tolerates them, extends to seeking its overthrow, if necessary by unconstitutional means. A clear example of this line of thinking was expressed by Patrick Devlin during the debate in the UK in the 1960s over legalising homosexuality. In The Enforcement of Morals,19 Devlin argued that the effect of legalisation would be to dissolve society itself. By this he did not mean that society would no longer exist in any form, if homosexuality were tolerated. He meant that the policy would undermine the constitutive values of society as it had existed until that time. It is not always clear whether Devlin thought only that the Christian values that in his view showed homosexual acts to be immoral would be flouted by legalising them, or whether he also thought that society could not continue to exist in any worthwhile form by doing so. What is clear, however, is that Devlin thought that the security of a way of life was at stake in the debate over tolerating homosexuality, and that this told decisively against it. The prima facie opposition or tension between toleration and security can be thought to stem from that between liberty and security. That such an opposition exists is, again, a staple of political discourse. The notion that there is a trade-off to be negotiated between liberty and security has a considerable lineage. In his contributions as ‘Publius’ to The Federalist Papers, Alexander Hamilton argued, albeit without recourse to the ‘balancing’ metaphor, that liberty and security must come into conflict. When they do, citizens will inevitably forgo liberty for the sake of security:

18 19

See fn. 2. Patrick Devlin, The Enforcement of Morals (Oxford University Press 1965).

130

Liberty, toleration, security

Safety from external danger is the most powerful director of national conduct. Even the ardent love of liberty will, after a time, give way to its dictates. The violent destruction of life and property incident to war, the continual effort and alarm attendant on a state of continual danger, will compel nations the most attached to liberty to resort for repose and security to institutions which have a tendency to destroy their civil and political rights. To be more safe, they at length become willing to run the risk of being less free.20

Hamilton found himself raising the spectre of authoritarianism even as the Continental Congress was meeting to shape the US Constitution, a document that strongly influenced The Federalist Papers, of which Hamilton was one of the principal authors. Of course, part of the point is to institutionalise checks on executive power so that the pressures towards repressiveness are held back. But Hamilton’s final sentence discloses an underlying problem with the very idea of a trade-off, or conflict, or balance, between liberty and security: in downgrading liberty for the sake of security, the executive jeopardises security itself.21

A purported tension It is often supposed that toleration was born during the early modern period in Europe, but arguably religious toleration in the Islamic world predated this by a millennium. The Constitution of Medina,22 drawn up by Muhammad in about the year 622, shortly after the hijrah, aimed to achieve reconciliation between warring tribal groups divided by religion.23 The provisions of the constitution allowed for the toleration of Muslims, Christians, Jews in the name of securing Medina (Yathrib) against clans from the Quraysh, the dominant tribe in Mecca. Not only the ‘people of the book’ but also non-believers were drawn into an agreement to create and protect an ummah or community. The constitution probably preceded Muhammad’s victory at the battle of Badr in 624, which led eventually to the formation of the first caliphate.

20

21

22 23

Alexander Hamilton, ‘The Consequence of Hostilities between the States’, The Federalist No. 8 (20 November 1787) in The Writings of Alexander Hamilton, ed. J. B. Freeman (New York: Library of America 2001), 191. In the words of Hamilton’s fellow American Revolutionary Benjamin Franklin: ‘They that can give up essential liberty to obtain a little safety deserve neither liberty nor safety.’ The quotation is inscribed on the pedestal of the Statue of Liberty. A version is reproduced in W. Montgomery Watt, Muhammad at Medina (Oxford: Clarendon Press 1956), 221–5. See Saïd Amir Arjoman, ‘The Constitution of Medina: a sociolegal interpretation of Muhammad’s acts of foundation of the umma’, International Journal of Middle East Studies 41 (2009), 555–75.

A purported tension

131

The document cites the parties to it as ‘the believers and Muslims of Quraysh and Yathrib and those who follow them and are attached to them and who crusade (jahadu) along with them’.24 As Watt notes, the Koranic verse ‘permitting Muslims to eat the food of the People of the Book (5. 5–7) takes on a different colour if it is regulating the relations of two religious groups within a single political community’,25 rather than regulating a Jewish ummah distinct from the Muslim one. In some respects the constitution might be seen as an early exercise in creating an ‘overlapping consensus’, but the document as it now exists seems to reflect power imbalances in contemporary Medina and does not extend equal rights to all. Muhammad’s political strategy appears to have aimed in part at propitiating Medina’s Jewish community, for example, not only because Jewish clans dominated the city but because at that point he still hoped to reconcile Jews to the Muslim faith. The Constitution of Medina illustrates the circumstances in which toleration becomes politically contentious. First, the motives which lead to a regime of self-restraint, and so putatively of toleration, are often mixed. They may be sufficiently mixed not to look distinct to politicians who act on them. Even actors who are clear-sighted enough to separate them in thought will probably not pause to consider whether each is severally necessary, and jointly sufficient, to justify toleration in the circumstances. Second, and more important, the conditions in which toleration may emerge seldom display an equality of power between all concerned parties. The facts of power may prompt toleration via the qualifying thought that what seemed intolerable may be, after all, tolerable. Security, far from stifling toleration, may bring it to birth. It seems plain that the Medina Constitution reflects facts of power on the ground in the early seventh-century Arabian peninsula. Those who, with Rawls, dismiss ‘mere’ modus vivendi26 toleration as at best the prelude to the advent of overlapping consensus perhaps underplay how far political settlements superpose modi vivendi with principle, how far this ambiguity has a motivational analogue, and how far any settlement that goes beyond it will rely just as heavily on facts of power as does modus vivendi itself. Since security and toleration are not the same value, circumstances can arise or be devised where they pull in opposite directions. Notoriously, also, alarms over supposed danger to the state can be used as the pretext for infringements of liberty, including toleration. Daniel Kahneman notes that in Israel the annual death toll from terrorist attacks rarely even 24 26

25 Watt, Muhammad at Medina, 221. Watt, Muhammad at Medina, 201. Rawls, Political Liberalism, xxxix, 147; cf. Rawls, Justice as Fairness: a restatement, 194–5.

132

Liberty, toleration, security

approaches that from road traffic accidents, which are seldom treated as threats to the foundations of public safety.27 But there is no greater conflict between security and toleration than there is between other pairs of political goods or values, such as liberty and equality, democracy and justice, the public interest and transparency, solidarity and efficiency. As with some of these other pairings, the idea of a constitutive opposition can be dispelled by pointing out ways in which the values can be mutually reinforcing rather than destructive. An argument for toleration can often be made from security, just as the argument for security can be made from toleration. However, the case for symbiosis needs to be made with care. There has long been a tendency in political thinking about the relation of security to liberty that contrasts with the one articulated by Hamilton. Spinoza seems to suggest as much early in the Tractatus Politicus, where he argues that ‘liberty of the spirit, or courage, is a private virtue; but security is the virtue of the commonwealth’.28 At first glance, this may seem to point up a demarcation of spheres between security and liberty.29 But elsewhere, including in the earlier Tractatus TheologicoPoliticus, Spinoza seems to express a different view: from the explanation of the foundation of the commonwealth which was given earlier, it very clearly follows that its final purpose is not that of domination, nor is it to keep men cowed through fear and to subject them to the rule of another, but on the contrary to liberate each and all of them from fear so that each one may enjoy security insofar as he is able to do so, that is, so that he can best retain his own natural right of living and acting, without injury either to himself or to other people. The commonwealth’s purpose is not, I say, to change men from rational beings into brute beasts or robots, but on the contrary to allow them to develop the faculties of their minds and bodies in safety, so that they may make free use of their reason and may avoid the conflict and the unjust mutual abuse that result from enmity, anger or deception. Therefore, the purpose of the commonwealth is, in fact, liberty.30

27 28

29

30

Daniel Kahneman, Thinking Fast and Slow (London: Allen Lane 2011), 142–3; 322. Animi enim libertas seu fortitudo privata virtus est; at imperii virtus securitas. Spinoza, Tractatus Politicus, §1.6, in Spinoza, The Political Works ed. A. G. Wernham (Oxford: Clarendon Press 1958), 264. For commentary on this passage see Susan James, ‘Democracy and the Good Life in Spinoza’s Philosophy’, in Charlie Huenemann (ed.), Interpreting Spinoza: critical essays (Cambridge University Press 2008), 128–46. Étienne Balibar, Spinoza and Politics (London: Verso 1998), e.g. 116, argues that the incomplete Tractatus Politicus marks a shift in Spinoza’s thought towards a belief in security as the political summum bonum. Compare V ii of the Tractatus. For a contrasting view, see Lewis S. Feuer, Spinoza and the Rise of Liberalism, 2nd edn (New Brunswick, NJ: Transaction Books 1987), 193. Spinoza, Theologico-Political Treatise, ch. 20, §6, in Spinoza, The Political Works; my translation.

A purported tension

133

However, the contrast is more apparent than real. What Spinoza upholds here is in effect the merging of security and liberty as ends or purposes of civil association. The commonwealth comes into existence in order to free people from fear so that they can enjoy security. However, as Spinoza goes on to explain, this end (while presumably good in itself ) serves the ulterior purpose of creating the conditions in which people can develop their mental and physical faculties and thereby ‘make free use of their reason’ – which Spinoza identifies, for human beings, with liberty. It is unsurprising that Spinoza should see political liberty and security as having this close relationship. A being’s overriding drive towards selfpreservation – which one could see as the valuable good sine qua non – is identified with that being’s freedom. Hence for Spinoza it would be wrong to think of freedom as something which is traded for self-preservation. Instead, freedom consists in the conatus by which all creatures strive to preserve their own nature.31 Even with Hobbes, often cited as a thinker who gives security absolute priority over liberty,32 matters are not straightforward. In Hobbes’s view the decision that people will rationally take in the state of nature does not reduce simply to abandoning liberty in favour of security. Instead, in realising the rational necessity of obedience to a sovereign, they recognise that the freedom that consists in the right of each person to everything necessary for self-preservation needs to be given up for the condition of civil liberty. In that condition, as Hobbes says, each citizen’s freedom depends on ‘the silence of the laws’.33 The point is that the apparently limitless freedoms embodied in the natural rights of the state of nature, whose rationale is self-preservation, are of little value without security; liberties in the civil state, however, though circumscribed by the law, are underwritten by the sovereign power.34 It is not that the imagined agreement to institute political authority involves a trading of liberty for security. Rather, by the fact of agreeing, people are seen as creating the security which is needed for their liberties to be effective.35 31

32 33 34

35

Spinoza, Ethics (ed. and tr. Edwin Curley, in The Collected Works of Spinoza (Princeton University Press 1985), Vol. 1), Part III, Proposition VI: Unaquaeque res, quantum in se est, in suo esse perseverare conatur. Cf. Colin Bird, The Myth of Liberal Individualism (Cambridge University Press 1999), 125. Hobbes, Leviathan, 152. This point is well made by Quentin Skinner, Hobbes and Republican Liberty (Cambridge University Press 2008), 52–3: ‘if we covenant, we lose only those elements of our natural liberty that, if retained, would undermine our own safety’. This is not to ignore Locke’s point against statism, that in instituting a monolithic power, it creates a remedy worse than the disease; Second Treatise of Civil Government, §93. But the objection rests not on an opposition between liberty and security, but the thought

134

Liberty, toleration, security

I suggest, then, that the apparent incompatibility between toleration and security can be dispelled in a way similar to that between toleration and liberty. As with the earlier argument by Devlin, it is important to distinguish between constitutive and contingent incompatibility. Sometimes the supposed incompatibility rests on the claim that tolerating some practice is intrinsically injurious to security – for instance the security of the values or goods that counsel intolerance. Even this contention can be challenged. If, say, the claim is that tolerating homosexual marriage will undermine traditional family structures, this may be true in the trivial sense that the structures have excluded same-sex unions; so tolerating it will bring that exclusion to an end. But it may be untrue, to the extent that the institution of monogamy in general and its associated family structures may not be harmed – they may even be strengthened, as proponents of homosexual marriage have argued – by such a step. Nor does it follow that security in any wider sense must be undermined by it. Further, the idea that toleration, in such circumstances, threatens security is elliptical or metonymic. Tolerating same-sex unions does not undermine security itself, but it may undermine the security of certain goods or supposed goods, such as the heterosexual nuclear family. One enjoys or lacks security as regards other goods – that is, goods other than security itself – but security is not to be identified with any subset of these goods. So a policy of toleration – even assuming that one accepts the claims about traditional family structures – will only be inimical to specific goods, and thus the security with which they are held, rather than to security itself. To this extent, whether something counts as a threat to security depends not only, and obviously, on whether there really is a threat, but also on whether the thing threatened really is a good. Security, toleration, terrorism More broadly, if security is identified with some specific set of goods, it is easy to construct examples that show security is incompatible with itself. So, suppose that the goods with respect to which one is secure include freedom of religion: secure in the sense that, as in the US Constitution, the right to religious freedom is protected by law. Nothing in that account rules out that someone may regard his security, say in the afterlife, as depending on his being able to engage in proselytism. Or, again, suppose that security is held to include protection against crimes that any regime powerful enough to retrench on private liberties will also be a bad bet as regards security.

Security, toleration, terrorism

135

such as burglary. Securing neighbourhoods against crime by surveillance methods may serve to export crime to other areas. Here, as often, one group’s security is enhanced at the cost of another’s.36 This is to be distinguished from the fact that arguments from security often point in opposite directions regarding one and the same policy; for example, between those who claim that possessing a nuclear arsenal enhances security by deterring aggressors as against those who argue that it diminishes security by making the country a target for a pre-emptive strike.37 Nor is this point confined to security as a value: the same argument can be made, for example, with liberty, or with equality. Suppose someone identifies liberty with particular acts with respect to which one is or is not free. Then clearly liberty can be represented as being in conflict with itself: the liberty, say, to have one’s personal integrity respected, versus the liberty to treat other people as one likes. Similarly, equality in one respect can only be achieved at the cost of inequality in another: given inequalities in the initial distribution, this is a necessary truth. One can achieve equality in disposable incomes, say, only by differentially – that is, unequally – treating those who start off with different levels of income. Or, if the aim is to equalise preference-satisfaction at the margin, the regime will need to treat people unequally to take account of interpersonal differences in efficiency. And, indeed, the same can be said of political toleration. To enforce toleration of one side, by preventing the other from acting on its disapproval of the first, is to be intolerant, justifiably or otherwise, of the second.38 Here, again, it might be said that toleration is divided against itself. All these considerations tend to weaken the claim that toleration and security must oppose one another whereas other values do not. It is often said that tolerating would-be subversives will prejudice the security of society at large, by for example affording terrorists the space in which to organise and carry out terrorist attacks such as those in London on 7 July 2005, which killed fifty-two people besides the terrorists themselves. It would be implausible to deny that acts like the 7/7 bombings are attacks on security. However, on a comparative statistical view, any individual UK citizen’s chance of dying in a terrorist attack is very slight – far lower, say, than that of dying on the roads, in domestic accidents, or from hospital-borne infections. The aim behind the bombings, as

36 37 38

This is not to say that security must always be zero-sum. An effective system of contributions is likely to enhance the security of all, as Hobbes supposed. See Jeff McMahan, British Nuclear Weapons: for and against (London: Junction Books 1981). Newey, Virtue, Reason and Toleration, ch. 5.

136

Liberty, toleration, security

with other acts of terrorism, is not merely to kill people, but to create a climate of fear, and thereby to attack the basis for confidence on which the good of security rests.39 Terrorism in general should be seen as a symbolic attack on security itself, which aims to undermine citizens’ enjoyment of security. Consider first what may be called the basis of security as a good, namely the reason to value and hence pursue it for its own sake: so the basis of the good of security is the reason to value and hence pursue security for its own sake. The claim that security has a basis, so that there is reason to value and pursue it for its own sake, is consistent with thinking that there are also other reasons for pursuing security, given perhaps by the value of the goods that are secured. The basis of security as a good consists in the settled mental state resulting from justified belief that one holds things of value. That is, the basis of security as a good lies in the settled mental state that results when the believer overcomes the hurly-burly of conflicting grounds for belief, and succeeds in forming a justified belief.40 Such a person benefits from holding the belief, with the arrest in judgement resulting from it. When someone attains such a belief, he achieves the mental equilibrium that consists in believing that he holds the goods, and is aware this belief is justified. To enjoy a good is to benefit from possessing it, which usually (but not always)41 involves a reflexive awareness that one possesses it. Thus, to enjoy the good of security is to benefit from possessing it by having grounds that justify the belief that one holds the goods. Terrorism attempts to prevent citizens from benefiting from possessing grounds justifying this belief. In this sense, the intended victims of terrorism constitute a much larger group than those directly injured by it. However, if it is true to say that terrorism does not for most people attack security itself, except by affecting their ability to enjoy the good of security, then the relation between terrorism and toleration also needs to be reassessed. For instance, most people would accept it as obvious that terrorist violence is intolerable as an attack on persons and other things that have value. But it is less obvious that attempts to undermine the basis for justified belief that one holds these goods must always be intolerable. For instance, those who promulgate conspiracy theories may aim to shake their fellow citizens’ confidence that the state is acting 39 40 41

See e.g. Michael Ignatieff, The Lesser Evil: political ethics in an age of terror (Princeton University Press 2005). This is an external standard: that is, one cannot hold a justified belief that p merely by believing that one does so. For instance, oblivion may be a good in certain situations. But having it – being in a state of oblivion – precludes reflexive awareness that one has it.

Security, toleration, terrorism

137

in their best interests. Often, no doubt, such activities are vexatious or express some personal obsession of the conspiracy theorist, but even here it is far from obvious that the activity merits intolerance – at least, if that involves suppressing the activity. In some cases, indeed, such theories may perform a public service by pointing out ways in which the public interest is damaged by those in power. Public comment often fails to distinguish real from symbolic consequences of terrorist attacks. Sometimes the political response to such attacks also confuses these types of consequence. Symbolic attacks on state security are essentially attempts to destroy confidence in the state’s ability to underwrite fundamental goods, and thereby to wreck or diminish citizens’ enjoyment of security as a good. One way, accordingly, in which the state can respond is by trying to reinforce citizens’ confidence in its ability to safeguard those goods, or at least by not actively conniving at terrorists’ attempt to undermine it. In part, of course, this demands a visible response to the risk of terrorism, but it also requires confidence-building measures through education, by for example drawing attention to the statistically nugatory scale of the ‘threat’ that terrorists pose.42 In this way, terrorism’s potency lies in attacking the basis for enjoying the good of security rather than, for most people, security itself. Short of acquiring nuclear or biological weapons, terrorists cannot destroy life, as a fundamental good, on a sufficiently large scale to make death from terrorist attack as statistically significant as deaths from cancer, road accidents, murder and many other causes. However, by threatening mass death, they can help to undermine people’s enjoyment of the good of security if they succeed, as sometimes they do, in making people believe wrongly that the threat is significant; this process is often aided by the ‘availability’ effect, whereby people give certain information undue weight because it is made cognitively salient.43 As security is a reflexive good, enjoying security requires awareness of having it. Someone who believes, even if falsely, that he lacks security, lacks this awareness. Hence the power of terrorists stems from their ability to induce people to form a false belief and then draw a correct inference from it. In this respect terrorists exploit what Kuran and Sunstein call the ‘availability cascade’ – that is, the tendency of certain beliefs to 42

43

Compare, for example, the 2,752 deaths recorded as resulting from the 11 September 2001 attacks, with the annual US death rate of some 18,000 from domestic accidents, or the figure of 42,196 deaths in road accidents in the US in the same year (National Highway Traffic Safety Administration figures; www-fars.nhtsa.dot.gov/Main/index. aspx (accessed 18 November 2011)). Daniel Kahneman and Amos Tversky, ‘Availability: a heuristic for judging frequency and probability’, Cognitive Psychology 5:2 (1973), 207–32.

138

Liberty, toleration, security

self-propagate among a populace prone to the availability error.44 However, it is not enough simply to point out that the judgements on which the salience of terrorist ‘threats’ rests reflect cognitive error. There are at least two reasons why misperceptions of threats cannot simply be brushed aside. One is that judgements that are widely made in democracies raise a question about the legitimation of policies that run counter to them. Maybe religion – all religion – is false, insofar as it rests on claims about the nature of the afterlife and putative supreme beings, but many who believe this to be true would nonetheless accept that religious belief has a legitimate claim on public policy, for example in the form of protection for religious beliefs and practices. Second, the measurement of risk cannot be put on an objective footing because risk itself is not objectively measurable, as Paul Slovic notes. Risk does not exist ‘out there’, independent of our minds and culture, waiting to be measured. Human beings have invented the concept risk to help them understand and cope with the dangers and uncertainties of life. Although these dangers are real, there is no such thing as ‘real risk’ or ‘objective risk’.45

So, for instance, the statistical incidence of deaths from terrorism in the UK and elsewhere is far lower than from causes such as road accidents. But people may decide, even when provided with accurate comparative statistics, that certain kinds of death are more undesirable than others; if so, the ‘risk’ of death from terrorism may be judged greater than that from a car accident, just because its negative weighting may be so much greater. One may prefer to risk a bad outcome with a higher probability than a worse outcome with a lower probability, even leaving aside nonlinearity effects created by deviations from expected-utility theory.46 These brute facts about preferences may be compounded in availability cascades by a catastrophic loss of public confidence. So far I have argued that security and toleration are not incompatible but different values, and have pointed out ways in which terrorist attacks, while seemingly threatening security, in fact do not immediately threaten the good of security itself, but rather threaten people’s enjoyment of the good insofar as the amplification effect succeeds. Nonetheless, since the state’s ability to concert security for its citizens depends partly on its ability to generate confidence, a de facto loss of confidence may 44 45 46

Timur Kuran and Cass R. Sunstein, ‘Availability Cascades and Risk Regulation’, Stanford Law Review 51:4 (1999), 683–768. Paul Slovic, The Perception of Risk (Sterling, VA: Earthscan 2000), xxxvi; cf. Kahneman, Thinking Fast and Slow, 141. That is, where a subject prefers (ex ante) expected utility U1 over U2, even though U1 < U2, e.g. because U2 involves a ‘sure-thing’ loss.

Security, toleration, terrorism

139

undermine the state’s capacity to provide security: that is, in some situations, alarmism about security can become self-justifying. However, the level of security enjoyed by a population at a given time depends on its vulnerability to loss – which is not limited to particular modes of loss, but in principle includes any way of losing a good. It overstates matters to say that there is no threat to security from terrorism, although the extent of the threat has been exaggerated. But it would also be too hasty to infer that security measures applied to the threat incur no prospective loss in liberty or toleration. The very fact that terrorism can be thought to pose a security threat, and that it can be believed that dealing with the threat demands sacrifices of liberty, can help to make it true that these losses are incurred. As I have stressed, even if ex ante expected utilities are significantly lower than some other kinds of risk, amplification effects at the symbolic level may weaken the state’s ability to offer security in the face of public alarm. That rests on the converse of the Hobbesian idea that, since the reputation of power is power,47 an agency that is reputed to be powerful enough to provide security can do so. The very fact that securitising measures can infringe the liberty and toleration enjoyed by citizens makes the opposition between them and security overdrawn. This may seem a perverse claim. I have indicated a way in which amplification effects can create such an opposition. But, as argued earlier, toleration and other forms of liberty are themselves goods which can be held with more or less security: the good of security depends on there being other goods that are securely held. Moreover, since the state, as well as underwriting security, poses a threat to it – that is, it can threaten the security of its citizens – enhancing the power of the state is always double-edged for security. For example, the repression of supposed subversives in the US during the ‘First Red Scare’ of 1919–20 led to large-scale deportations of communists and anarchists instigated by the Attorney General Mitchell Palmer.48 Palmer tried with the backing of President Wilson to perpetuate the Sedition Act 1918, devised to protect the US war effort, which outlawed general strikes. Here the ‘climate of fear’ used to legitimate state repression – in this case, mainly of immigrants – can be reinterpreted as justified fear of state power, and can readily spread to groups other than those directly targeted. The activities in the 1940s 47 48

Hobbes, Leviathan, ch. X., 62. On the Palmer raids, see e.g. Kenneth D. Ackerman, Young J. Edgar: Hoover, the Red Scare, and the assault on civil liberties (New York: Carroll & Graf 2007); also Bonnie Honig, Emergency Politics: paradox, law, democracy (Princeton University Press 2009).

140

Liberty, toleration, security

of the US House of Representatives Un-American Activities Committee, later renamed the House Committee on Internal Security, provides a further illustration of the threat to citizens’ security posed by state power. The relation between security and toleration is ambiguous because security itself is not a monolithic good or value. Measures to promote security always involve protecting some goods – and the goods of some groups – at the expense of others. Of course, it may be said that terrorism is intolerable – that it should not be tolerated. The wanton killing of innocent civilians is surely intolerable, in the sense of being unjustifiable. It is also clearly not tolerated, in the sense that the criminal law prohibits acts of terrorism. But official intolerance of terrorism need not rule out its sanctioning principles, observance of which can make terrorism more likely. The toleration that is extended by free speech, freedom of movement, of association, proselytism, and so on, helps to create an environment in which terrorist acts can be organised and carried out. The political authority may exercise toleration because this is the consequence of legal and other institutions that embody the ideal of freedom, even while those running the institutions know that it has this consequence. A further result of free institutions is that notions of security themselves can be aired and disputed. Attempted ‘decontestations’ of security – for example, those that try to set the terms of trade between security and freedom – can be called into question. This is not surprising, given that any understanding of security will give primacy to certain types of goods and certain ways of holding them, as well as prioritising some threats to them over others. An indeterminate role for the concept of freedom is itself a result of the working of free institutions. For example, those who favour a laissez-faire or market-based conception of freedom may argue that the freedom to appropriate and use private property is the paradigmatic form of freedom, but this is only one decontestation among others. Conceptions of security may be more or less inclusive, and the rosters of goods that are secured, and their relative priority, are all matters for further determination.49 Conclusion Received political wisdom has it that toleration is and should be inimical to terrorism, that terror is ‘intolerable’. At the same time, toleration and security are often thought to be in tension with each other. This creates a 49

See Ian Loader and Neil Walker, Civilizing Security (Cambridge University Press 2007), ch. 1.

Conclusion

141

puzzle: how can security policy, as exemplified by the ‘war on terror’, be opposed both to terrorism and to toleration, if the latter oppose each other? However, contrary to the received wisdom, toleration and security may in fact prove to be mutually reinforcing rather than internecine political goods. An effective regime of toleration requires the securing of certain fundamental liberties of the person, such as freedom from arbitrary arrest and detention. More surprising, perhaps, toleration also does not preclude terrorism, in the following respect: liberties such as those just mentioned foreseeably create an environment in which terrorist acts are more likely.50 To this extent, the original tensions are in fact not between toleration and security on one side, and terrorism on the other, but within security itself. Governments dealing with the perceived threat of Islamic terrorism have to choose whether to secure liberties, or to secure citizens against the threat. However, since the threat itself tends to be exaggerated by liberal societies, the choice itself is overdrawn. That goes also for the partly rhetorical opposition between security, or ‘securitisation’, and toleration. In order to remain recognisably liberal, these societies have to remain in some degree tolerant of terrorist activity. Western governments risk engineering a situation in which citizens are secure, but are unable to enjoy the good of security.

50

I discuss a further respect in which liberal democracies may be thought to tolerate terrorism in Chapter 10.

8

Toleration and power

In this chapter I shall use a largely unanalysed account of power to try to get clearer about the landscape in which toleration operates, and suggest that that landscape is in some ways more confined than it may look. When we consider what is given and what is counterfactually variable in thinking about toleration, particularly as regards power, we encounter limits to idealisation in the face of power. This idealisation surfaces when we ask what makes a given instance of self-restraint tolerant. In rejecting modus vivendi accounts of toleration, we also have to accept that the principles which support toleration are limited in their application. Toleration, on this view, is inextricably embedded in the less than ideal conditions of the here and now. No doubt this helps to explain why so many commentators find it awkward. The approach I use in discussing the power condition will be more analytical than that adopted elsewhere in this book. That is because, while this approach can advance understanding, its limits become apparent precisely when trying to make the condition analytically exact. Power and the power condition This chapter has more to say about toleration than about power. That is not, clearly, because the notion of power demands no explication. In fact, ‘power’ here serves simply as a place-holder for an agent’s capacities: it is within an agent’s power to do something if, in the circumstances,1 she has the capacity to do it. So, to lack power is to suffer a lack of capacity with respect to an action. Having the capacity to do something can be understood in an occurrent or in a dispositional way: a person who has the capacity to perform an action, such as piloting a passenger aircraft, may find herself circumstantially unable to do so, for example through illness or a sudden loss of resources (such as the unavailability of an 1

This qualification is not redundant. One’s capacities depend not only on skills, knowledge, etc., but also on whether, in the circumstances, one can exercise these skills.

142

Power and the power condition

143

airworthy plane); however, the distinction between occurrent and dispositional capacity does not much matter for our purposes. When toleration is at issue politically, the question is usually whether to tolerate something, and that will mean that the capacities are engaged here and now. What capacities one has depends importantly on the resources at one’s command. That may in turn depend on luck. Insofar as toleration depends on capacity, then, it will depend on these other things on which capacity itself depends. The lack of capacity suffered by a powerless agent may result either from the agent’s own inability, or from limits in the practical environment where she is situated. I currently lack the power to jump fifty feet in the air. I may acquire the power to do so if provided with appropriate mechanical aids. This shows that the level of power enjoyed by an agent can be varied, by imagining that she could be endowed with capacities which she currently lacks, or conversely. Then we can ask what this person would do had she power, that is, a capacity, which she in fact lacks. But even if I can jump fifty feet in the air with mechanical aids it will still not be within my power to do so unaided, the laws of nature being as they are. This implies that some things are not within my power as things stand, while others are simply not within my power, period – even if, in some imaginable other state of the world (such as one with different physical laws) they might be possible for me. It is a significant feature of thinking about power that it both admits of counterfactual variation, and that the scope for this variability can be limited, for example by imposing further conditions. These conditions minimally include that of non-contradiction. In a strong sense, it is not within my power both to be here and not to be here, now. Leaving logical or metaphysical impossibility aside, there is still ample room to imagine that the world might be sufficiently different to give an agent powers which he currently lacks, or the converse. In some cases, further, the counterfactual names a possibility which is or was accessible via the person’s own agency. Fairly obviously, this will be a small subset of the set of counterfactual possibilities. There is a consensus among writers on toleration that if an agent A tolerates the performance of some action P, then A has the power to prevent or otherwise sanction P.2 I shall refer to this as the Power Condition on toleration.

2

E.g. Andrew J. Cohen, ‘What Toleration Is’, Ethics 115 (2004), 68–95; Catriona McKinnon, Toleration: a critical introduction (London: Routledge 2005); Thomas M. Scanlon, ‘The Difficulty of Tolerance’, repr. in Scanlon, The Difficulty of Tolerance (Cambridge University Press 2003), 187; Newey, Virtue, Reason and Toleration, 31ff.; King, Toleration, 21.

144

Toleration and power

Power Condition 1 (PC1): A can only tolerate a disapproved-of action or state of affairs P if A believes that it is within his power to prevent (etc.) P. Non-toleration may of course take the form not of outright prevention but of censure, condemnation, or the imposition of other sanctions or penalties on those who are disapproved of, hence the ‘etc.’3 These remarks bypass a recent trend in philosophical writing on toleration, which demurs at seeing toleration in terms of prevention and permission.4 This approach distinguishes the practice of toleration in, particularly, the early modern period, where toleration was often granted on the basis of royal prerogative, from the form which toleration should take in liberal democracies, where those subject to power are also those who ultimately wield it. This ‘respect’ conception of toleration5 works with a conception of public justification according to which citizens acknowledge that disapproval arising from reasonably rejectable conceptions of the good life cannot reasonably be imposed on others.6 Thus, for instance, even if I subscribe to a form of Christian fundamentalism which reprobates bestiality or incest on biblical grounds, I acknowledge that it is unjustifiable to use the power of the state to impose my views by force of law, since my Christian outlook is one which others may reasonably reject. On some models, toleration is exercised not from the top downwards, but via a democratic political will in which power is divided equally among those subject to law. I shall not comment in any detail here on this approach to toleration.7 However, the respect conception still enables us to talk of prevention and permission. It can talk of legalising narcotics, for example, where the public justification test rules that outlawing them would rest on reasonably rejectable ideas of the good. Moreover, it will still be appropriate to think in terms of the exercise of power, even on this revised view of its legitimate use. A legitimate state has the capacity to use the legal and judicial means to enforce the criminal law. To that extent the legalising of narcotics would mark the state’s eschewal of coercion. It may however be pointed out that political capacity in a democratic state is more complex than this, as it depends not simply on brute force but on the wider means available to political leaders in canvassing acceptance for public policy. This includes arguments of principle, examined further below. Before this elaboration, however, let us first consider the intuitive basis for endorsing the Power Condition. If PC1 is seen as necessary 3 4 5

From now on I take the ‘etc.’ as read. See Chapter 5. Cf. Forst, Toleranz im Konflikt; and Jones, ‘Making Sense of Political Toleration’. 6 7 Forst’s term: Toleranz im Konflikt, e.g. 127. See Chapter 5. See Chapter 3.

Power and the power condition

145

for toleration, it will void toleration where the non-prevention of a disapproved-of action or state of affairs results solely from the agent’s lack of power to prevent it. Suppose, for instance, that a racist would gladly repatriate immigrants if he had the power to do so, but in fact lacks that power: on this question, he finds himself in a minority which does not control political outcomes. Such a person could not be described as tolerant. This reveals the intuitive basis for PC1: I cannot be said to tolerate something which I disapprove of or dislike, if my failure to act against it results solely from the fact that I cannot do anything about it, and I would act against it if I thought it lay within my power to do so. In such circumstances, I should be said to put up with the activity, rather than to tolerate it. Can I be said to tolerate an activity I dislike or disapprove of, despite lacking the power to act against it, if I would not prevent the activity even if I (believed that I) had the power to do so? It could be argued that in this case I do tolerate the activity. If I wrongly believe that I have the power to stop the activity, but decide on principle not to, it seems that I act tolerantly: what matters is my stance towards the activity, which is not to use the power I believe that I have to prevent it. But if so, it seems hard to see what difference it makes if I correctly believe that I lack the power to prevent it. The stance I take is the same, a principled eschewal of the use of power. If so, one could conclude that what is necessary for toleration is not to have the power of prevention, but not to be disposed to use the power if one believes one has it. This suggests that the PC1 should be modified along the following lines. Power Condition 2 (PC2): A can only tolerate a disapproved-of action or state of affairs P if A believes that it is within his power to prevent (etc.) P. A does not tolerate P if A would seek to prevent P, if A believed he had the power to do so.8 This is a necessary and not a sufficient condition of toleration. On any view, further conditions need to be met. Yet someone who meets PC2, and the supplementary conditions, may still fail to tolerate. This is because the capacities relevant to assessing toleration can in turn depend on whether a particular option is deliberable. Someone may decide that because an option incurs unacceptable costs, it cannot be taken. For instance, I could perhaps just afford (i.e. I have just enough money) 8

An important question concerns the relevant counterfactual set, the truth-conditions of whose members bear on the truth of the judgement that A tolerates P. What counterfactually varies is the power A exercises. But sometimes the fact that A has additional power alters the practical environment for others (the putatively tolerated), and with it the relevant counterfactual set. A’s greater power may work proleptically, deterring people from doing P, so no question of toleration arises.

146

Toleration and power

to buy some luxury item, but given that no money would be left for more important purchases, regard the cost as unacceptable. Again, in the circumstances of toleration, the government or some other agency may decide that suppressing some group, though possible, would incur undue costs. As this example suggests, the capacities relevant to power are scalar properties. This is unsurprising, inasmuch as the resources on which capacities often depend likewise admit of degree.9 More significant, when an agent fails to prevent a disapproved-of action because doing so will incur certain kinds of cost, this will void the claim that she acts tolerantly. To see this, consider the following example. Martha has it within her power to stop her daughter Lois from going out one evening, but fails to do so. However, she fails to do so only because the costs of prevention are unacceptably high to her: Lois has convincingly threatened that she will hurl a chair through the window unless she gets her own way. Martha decides not to use the power she has to thwart Lois’s plans. Here, although Martha fails to use her power to stop Lois from going out, her failure to use her power results only from the fact that to do so would inflict unacceptable costs on her, Martha. Here, it seems implausible to say that she tolerates Lois’s going out. Of course, the fact that Lois can use the threats itself restricts Martha’s power, but it does not simply make Martha unable to prevent Lois from going out. Lois makes Martha’s preferred combination of outcomes, that Lois stays at home and the chair does not go through the window, unavailable. However, since Martha could stop Lois if she were prepared to bear the cost of doing so, it is mistaken to say that she cannot prevent what she disapproves of. Someone may say that Martha cannot now stop Lois, since she cannot bear to take the cost of stopping her. In that case, of course, it would then follow with added force that she does not tolerate Lois’s going out. But in fact practical modality is more nuanced than this. It would not be thought a plausible line of argument to reason as follows. 1. If one cannot φ, then it follows that it cannot be the case that one ought to φ. 2. If one cannot be bothered to φ, then one cannot φ. Suppose further, for φ ¼ ‘tidy my bedroom’, 3. I cannot be bothered to φ. Therefore 4. It cannot be the case that I ought to tidy my bedroom. 9

For this reason, on my analysis it is sometimes indeterminate whether someone tolerates a practice.

Power and the power condition

147

Presumably few, except perhaps some teenagers, would find this argument persuasive. Apart from the dubiousness of the ‘ought implies can’ premise 1, accepting 2 in its unrestricted form exacts a heavy price: for one, it has to insist that I can φ only if I can be bothered to. It is a feature of costs that, short of outright unaffordability, one may incur more of them than one would ideally like, and a change in circumstances (such as what other people are prepared to accept) may change the schedule of costs attached to one’s options. Even so, the options often remain options even when their relative costs have shifted. Martha finds that she has to put up with something that she finds very disagreeable, for fear of suffering something yet worse. The outcome is induced by making the alternative to it unsustainably costly. If so, but one aims to keep hold of the intuition that in the story Martha fails to act tolerantly, the power condition needs to be interpreted in such a way that an agent’s power can be removed or weakened by imposing costs on prevention which the agent, the putative tolerator, regards as unbearable. Hence, though Lois’s threat does not make it impossible for Martha to prevent her from going out, it dictates reasons for Martha such that her non-prevention fails to qualify as tolerant. One might think that it does not matter whether we say that Martha lacks the power of prevention here, and so doesn’t tolerate her daughter’s going out; or that she has the power, but fails to exercise it because of the costs, and so still doesn’t tolerate it. These are, to be sure, not simply two descriptions of the same situation, because on the one account Martha has the power of prevention, and on the other she does not. However, what makes the matter indifferent as regards toleration is that Martha’s attitude remains the same in each case. As Martha sees things, it would be better if she had the power to stymie her daughter’s plans without the attendant costs. As far as this goes, what matters for toleration is not so much the extent of Martha’s actual power but her attitude in response to conduct of which she disapproves. Two implications follow from this discussion. First, the example suggests that PC2 should be elaborated to respect the intuition that an agent does not tolerate a practice if the failure to prevent it is due only to the fact that prevention incurs unacceptable costs. Martha would prevent Lois’s going out if doing so did not incur costs which she does not wish to bear. Then the limits on her power, duly augmented to allow for the unacceptable costs of prevention, will refute the claim that she acts tolerantly. The situation would be different if Martha, while disapproving of Lois’s nocturnal excursions, permits her to go out, not because of some threat Lois has made, but because she, Martha, subscribes to a principle of autonomy.

148

Toleration and power

Second, as regards the prospective costs of prevention, it will often be unclear whether a given action is tolerant. Self-restraint frequently rests on mixed motives of pragmatism and principle. So there can be reasonable uncertainty as to whether an agent restrains himself because he regards the alternative as unduly costly, or for reasons of principle. But the distinction between ‘costs’ and ‘principles’ itself is far from clearcut. With certain considerations, it may be doubtful how far an agent mindful of them in not preventing an action of which he disapproves can be said to tolerate that action. This is particularly true in political decision-making. For example, military intervention to redress human rights abuses in contemporary Zimbabwe would run various kinds of risk. Reasons of principle as well as cost may deter foreign governments or the international community from military intervention. But given that governments legitimately curate the national interest, even considerations of cost might be thought a principled reason for (in)action.10 Certainly, a pragmatic compromise may be the best outcome available in the circumstances. But are the parties to it acting tolerantly? A policy which some might describe as ‘tolerant’ in its effect – which may mean simply that it involves a compromise – need not be accompanied by a tolerant attitude from its authors. In fact, they may be only too willing to put their adversaries to the sword if granted the opportunity. Such is the stance taken by the racist discussed earlier. He would cheerfully repatriate the immigrants given the chance. But in no sense does his behaviour strike one as tolerant. In his favoured world, those who share his views would have the power to repatriate those of whom he disapproves. Hence a policy can be tolerant in its effects, even though none of the parties to it has a tolerant attitude. Where, as in modus vivendi, non-suppression rests purely on stalemate, the argument so far suggests that the parties to it should not be seen as tolerant.11 The power condition should take account of the fact that prevention, though not impossible for the agent, nonetheless may incur costs which she judges too heavy to bear. So we need to modify PC2 in turn, and arrive at: Power Condition 3 (PC3): A can only tolerate a disapproved-of action or state of affairs P if A believes that it is within A’s power to prevent (etc.) P. A does not tolerate P if A would seek to prevent P if he believed he had the power to do so, including the possibility that A would prevent P if doing so did not, in A’s judgement, incur undue costs for himself. 10 11

Political principle itself, such as commitment to an ideal of reciprocity, is also often based on self-interest. The further claim may be made that instituting a modus vivendi now may encourage the formation of genuinely tolerant attitudes over time.

Power and the power condition

149

It might be thought that what underlies this newly revised version of the test is the fact that judgements of toleration are irreducibly normative – that one can judge that A acts tolerantly only if one thinks that the reasons for A’s self-restraint are good ones. However, this is mistaken. One may judge that someone’s self-restraint is tolerant, even if the principle on which it is based, or the belief that the principle applies in a given case, seems misguided, as with a parent whose reason for tolerating his child’s uncouth antics lies in a misguided belief in self-expression; here, as elsewhere, there is no supposition that toleration must be a good thing. It is for this reason that a normative conception of toleration that claims only justified disapproval, or justified responses to disapproval, can count as tolerant, does not adequately answer the censorious tolerator problem: the possibility of someone who, keeping his behaviour constant, extends the range of activities of which he disapproves.12 This criticism could perhaps be met by qualifying the normative judgement, so that it makes not an all-in assertion about justification, but claims only that what conforms to the structure of toleration is prima facie justifiable. Conversely, someone who restrains himself from condemning vandalism carried out by a mob of youths, because to act otherwise would risk serious injury or death, is not acting tolerantly, even though one may accept that these reasons for selfrestraint are good ones. Not all good reasons ground toleration, just as toleration can be grounded on reasons that are not good. However, PC3 still fails to give an adequate representation of the role of power in judgements about toleration. This is implicit in the commentary on the Zimbabwe example given earlier. For the costs themselves may be measured in terms of principles which A upholds. In this case it seems wrong to treat acting on reasons of cost as voiding toleration. I may be unable to bring myself to do something because of some principle I hold, rather than because of simple lack of power, or pragmatic costs. For example, I may be unable to bring myself to torture a baby, not because I lack the physical, pecuniary, etc., resources to do so, but because on grounds of repugnance, I cannot bring myself to do it. I cannot do it, and this inability results from my principled objection to torture. This kind of inability should fall outside the scope of PC3. Someone who subscribes to the principle of respecting autonomy, for instance, may decide that she cannot breach the principle even though doing 12

This problem was pointed out by John Horton in ‘Three (Apparent) Paradoxes of Toleration’, 16ff. Horton proposes a solution along the lines mentioned and criticised in the text. For further discussion see also my Virtue, Reason and Toleration, ch. 4.

150

Toleration and power

so will license behaviour of which she conscientiously disapproves. Such might be the case, for instance, if Martha were more liberal-minded. Clearly this will become more likely, the heavier the moral costs (as judged by the agent) incurred by breaching the principle of autonomy through coercive intervention. Here the prospective costs, reckoned in terms of autonomy, may indeed counsel self-restraint. But if A permits another person to act in a way of which A disapproves, because A subscribes to a principle of autonomy, there seems no reason not to describe her behaviour as tolerant.13 Unlike costs or the material resources needed to wield greater power, I cannot coherently see my principled incapacity as cause for regret, an incapacity from which I would not suffer in some other and more ideal world. We can make clear the contrast between costs of ideals and other costs by considering a case in which the agents are restrained only by a pseudo-ideal of efficiency. In this connection, consider the example of the Holocaust. During World War Two, the Nazis could presumably have diverted further resources from their war effort to track down and kill yet more Jews. The costs of detecting and apprehending remaining Jewish fugitives were presumably subject to rising marginal costs. The fact that the Nazi regime failed to track them down does not show that it was tolerant of the remaining Jews’ existence. We can suppose that they would have killed them had it not been unacceptably costly to do so.14 It is plausible to bring such costs within the scope of the power condition, precisely because they would not have to be borne in the Nazis’ ideal world. Of course, that example is purely illustrative. It may be said that the distinction between costs and ideals is a purely ad hoc one. But the argument accepts that the content of the ideal world can be varied at will by the putative tolerator. It can therefore accommodate even grotesque ideal-world contents of the sort just mentioned. The point is that getting a fix on that content helps to decide whether or not someone is acting tolerantly, by asking about their attitudes in the real world. What if the principles themselves could be thought sub-ideal?15 Perhaps in extremis even they may be idealised away. For example, I may come to think that I arrived at the principles I now hold through some cognitive malfunction or misadventure, such as brainwashing (or simply strong 13 14 15

See Raz, ‘Autonomy, Toleration and the Harm Principle’, and The Morality of Freedom, Part V. I owe this example to Graham Long. One version of this thought is the ‘dominant ideology’ critique of liberal capitalism. However, for the agent to think that this vitiates his own toleration would, on this critique, require that he has unmasked his own ideological commitments, i.e. has become, in a manner of speaking, a critical theorist.

Power and the power condition

151

social conditioning) in early life.16 I might then think that though I couldn’t help believing in and acting on what I believe, in a better world I wouldn’t adhere to these principles. The concern is not about backsliding – whether I manage to remain true in my actions to the principles I consciously endorse – but rather about my grip on the principles themselves. I may come to think that in my ideal world, I would adhere to some different set of principles from those I actually hold. Perhaps, if a person were to exercise self-restraint because of a principle in which he believes, while also regretting his adhesion to this principle, it would be mistaken to affirm that he acts tolerantly.17 Perhaps, by the same token, someone who restrains herself from acting on a principle which licenses disapproval of some practice, does act tolerantly. This situation is fairly common. Someone who has been brought up to believe that eating pork is bad, unclean, and so on, may come to think that these attitudes result from conditioning in early life, and lack any objective justification. Nonetheless, she may still feel residual distaste – or more than that – at eating pork. To overrule one’s own long-established attitudes of disapproval after coming to regard them as prejudices might be thought well-nigh paradigmatic of toleration. It is not obvious, at least, that someone who struggles against such attitudes of disapproval which she finds inexpugnable counts as less tolerant than someone who, never having been inculcated with the underlying beliefs, sees the practice as unexceptionable. The idealising disowning of one’s actual attitudes can thus work in either direction: towards undermining either the principle which supports toleration, or the attitude of disapproval. There is little gain in trying to regiment phenomena which resist regimentation, not least because conflicts of attitudes are both a psychological commonplace and hard to pin down, as the presence of each attitude constitutes evidence for the absence of the other. Judgements about toleration have to take cognisance both of agents’ actual attitudes and of their resultant idealisations – including, in the range of cases just discussed, attitudes about these very attitudes.18 16 17

18

For an example of this process, see G. A. Cohen, If You’re an Egalitarian, How Come You’re So Rich? (Cambridge, MA: Harvard University Press 2002), 8ff. This case is also different from one where I imagine that it is better that my principles are not generally held. I could think that the world is better for the fact that I hold them, although it’s also better that others don’t. It may be said that if I see my principled beliefs as formed through brainwashing, etc., I to that extent don’t hold the principle at all. I may believe that children may be subjected to corporal punishment, but also think that this belief is one I hold only because of my own peculiar upbringing. But if I regularly flog my offspring despite holding this meta-belief, I still act on the belief that flogging them is appropriate.

152

Toleration and power

Power does play a role in judgements about toleration. But its significance lies in the underlying attitude of the putative tolerator, which can be brought out by examining the counterfactuals which apply. Someone who fails to exercise her power of intervention will nonetheless fail to act tolerantly if her failure results solely from an aversion to nonprincipled costs, though the line between principled and other costs is often difficult to draw. Toleration and the limits of idealisation The counterfactual reasoning allowed for so far permits an element of idealisation to enter in forming judgements about when agents act or fail to act tolerantly. The idealisation sifts correct and mistaken judgements by asking how agents would act in similar circumstances, adjusted to take account of their relevant capacities. The guiding intuition has been that an agent who fails to exercise her capacity to prevent activity of which she disapproves can only be judged to behave tolerantly if her self-restraint rests on certain kinds of reason. Only actions duly informed by principle will be eligible for description as tolerant. The analysis thus involves a modest level of idealisation. However, some could say that the idealisation does not go far enough. It may be argued that there has to be a non-contingent link between the principle that supports self-restraint and the fact that the action tolerated provokes disapproval. Otherwise, it seems, one can always construct counterfactual examples which pull apart the grounds for self-restraint, on the one hand, and the possibility of disapproval, on the other, and then use an idealisation argument to show that the agent has not, after all, acted tolerantly. In the agent’s ideal world, the empirical conflict between the reasons which support prevention on the one side, and permission on the other, would not exist. Consider the example of civil liberties embodied in safeguards to which those suspected of or charged with criminal offences may lay claim. In liberal polities the relevant entitlements include limits on detention without charge, freedom from torture, rights to due process, access to legal representation, and so on. A foreseeable result of these safeguards is that some people who are guilty of an offence are either not charged or are acquitted at trial. To this extent, the system of legal entitlements for detainees could be thought of as tolerant. The state, as embodied in the judiciary, permits that of which it disapproves, namely the de facto release or acquittal of those who are really guilty, in the name of a principle, or set of principles, of personal liberty. Moreover, we may assume that the state has the power to

Toleration and the limits of idealisation

153

remove these entitlements, for example by primary legislation, as has in fact happened in some countries.19 So far, then, it would seem, so tolerant. But is it? The objector may say: to be sure, we lack a system of what Rawls called perfect procedural justice20 in all actually existing criminal jurisdictions. That is, all systems of criminal justice are fallible as methods for determining criminal guilt, and so are only imperfectly procedurally just. But one could idealise away these shortcomings by imagining a system of criminal justice in which there is a perfect correspondence between substantive guilt under the law and the verdicts reached in criminal trials. Then, the objector continues, the rationale for the safeguards, which rests on fallibility, would be removed. In this case, the liberal state’s commitment to civil liberties would be seen as contingent on the fallibility of its own legal system. In its ideal world, in which ‘Guilty’ verdicts perfectly track substantive guilt, the state would dispense with the liberties, as their rationale would have been exhausted. So the liberal state’s claim to toleration turns out to be as flimsy as Martha’s. One response is that the rationale for civil liberties is not exhausted by the fallibility of judicial systems. However, the underlying point remains when toleration rests on contingencies that can be imagined away with augmented capacities. Then the idealisation threatens to call into question the commitment to non-prevention itself, since whatever gives rise to disapproval can be imagined away. The basic version of this thought is that in a better world the activity disapproved of would not exist, and so no need for toleration would arise. Surely anyone who believes that the activity is a fit object of disapproval is committed to thinking that it is pro tanto bad. That means, it may be said, that it would be better that the activity did not exist. The objector in effect says the following to defenders of toleration. Imagine yourself to be omnipotent; then, if you think that the activity really merits disapproval, why would you use your power to create a world which contained that activity? Surely you wouldn’t. But if not, you are really no different from the pragmatic ‘tolerator’ who permits what she disapproves of solely because of the costs attendant on not doing so. For in each case the reasons which counsel non-prevention result merely from a deficit of power. Martha’s claims to act tolerantly were refuted by counterfactual idealisation: in her ideal world, she would have no need 19 20

In the UK, for instance, the Terrorism Act 2006 extends the period for which suspects may be detained without charge to 28 days. Rawls, Theory of Justice, 85. On that notion, see my ‘Just Politics’, Critical Review of International Social & Political Philosophy 15:2 (2012), 165–82.

154

Toleration and power

to restrain herself from preventing her daughter from going out – since in that world no costs would be incurred by prevention; or, more ideally still, her daughter would not wish to go out anyway. So, by the same token, in this world you lack the power to prevent the activity you disapprove of; but in your ideal world, you could ensure that it never occurred. In this sense, the idea that you are committed to ‘toleration’ is self-deluding. This line of objection may be compared with critiques of toleration offered by Herbert Marcuse and Wendy Brown.21 These critiques, broadly speaking, level at liberalism the charge that its profession of toleration is at best compromised by its exercise of power, which differentially benefits certain groups or interests. The present objection holds, by contrast, that the appearance of toleration actually results from a deficiency of power on the part of the soi-disant tolerators. Their ideal world is one in which the activities tolerated fail to occur, and thus the call for toleration exists only where liberals’ shortage of power means that they have to put up with the activities willy-nilly, because the costs of suppression are still greater. Then the question is whether anything can rightly be labelled ‘toleration’ at all. We began by observing that tolerating a practice seems to require having the power to act against it; and that sometimes when an agent has this power but fails to use it, intuition baulks at saying that she behaves tolerantly. The question immediately follows, how genuine acts of toleration are to be distinguished from mere non-prevention. The initial answer to this question drew a distinction between ‘costs’ and ‘principles’ as constraints on the agent’s powers of prevention. However, this distinction seemed ad hoc, and in need of amplification. We sought this from the notion of counterfactual idealisation – in other words, from considering what the agent would allow, given the power to create her ideal world. And this seems to make more sense of the distinction between principles and adventitious costs. For principles are such that they are largely immune from counterfactual idealisation, since the principles themselves go to determine what the counterfactual ideal is. Costs can be counterfactually varied between this and the ideal world, but principles cannot, because they serve to frame the ideal world itself. Again, principles largely operate to fix the content of the ideal. Most people do not see their principled beliefs themselves as less than ideal. But now the challenge is to explain why, not the principles themselves, 21

See Herbert Marcuse, ‘Repressive Tolerance’, in Robert Paul Wolff (ed.), A Critique of Pure Tolerance (Boston, MA: Beacon Books 1969); Wendy Brown, Regulating Aversion: tolerance in the age of identity and empire (Princeton University Press 2008).

Toleration and the limits of idealisation

155

but the activities to which tolerators apply them, should not be able to be counterfactually idealised away in the name of moral maximisation. The explanation should not be vulnerable to the charge that permission is revocable counterfactually, given the power to create an ideal world. The range of candidates for such a non-contingent basis is likely to be small. For it will have to show both why permission is a good idea even though it countenances the doing of sub-optimal acts, while also showing why it would not be better, had one the power, to bring about a world in which these acts did not occur. Since by hypothesis the acts are sub-optimal, the explanation will need to point out some morally significant concomitant of the acts which makes it better not to bring that world about. An obvious response suggests itself. We do not, alas, live in anybody’s ideal world. That is what gives rise to the circumstances of toleration. At both the personal and political levels people are confronted with the decision whether or not to prevent disapproved-of activities. Given that this is so, and no power on earth can change it, persons and political actors have to decide what to do for the best. A theory of political toleration, in particular, should be firmly rooted in the real circumstances in which decisions have to be made. The demand for idealisation repeats an error which ‘realist’ critics of liberalism have recently claimed to expose.22 At first sight, this response misses the point. The challenger did not ask whether prevention was a good thing in the circumstances – which must presumably be judged on a case-by-case basis. Instead, his question, even when permission is deemed to be preferable to prevention, is whether this action can be described as tolerant. But, it may be said, there is still a difference between the truly objectionable cases and those in which the agent acts tolerantly. With Martha, her aim is to prevent a real person, her daughter, from acting in a certain way. With genuine toleration, by contrast, it is not that the agent’s ideal world will adjust things so that someone who is here permitted to act will there be prevented from acting. Instead the activity will not arise in the first place. But this hardly seems to dispose of the problem. The position now is that the person who tolerates a practice, rather than merely putting up with it, does so in virtue of holding to some principle: one which says that, though permission is better than prevention given that the practice is going on, it would be better still if the world was such that it didn’t go on at all. But this stance hardly seems characteristic of toleration. If anything, it seems to verge on megalomania. 22

See for example Raymond Geuss, Philosophy and Real Politics (Princeton University Press 2008). For a survey of the realism v. moralism debate, see Marc Stears, ‘The Politics of Compulsion’, British Journal of Political Science 37:3 (2007), 505–31.

156

Toleration and power

However, reflecting on why this response is inadequate will, I believe, help us to become clearer about the relation of toleration and power. Take a still more extreme case: suppose I imagine a world in which everyone acts as I think best, and none acts in ways of which I disapprove. One version of this idea would be a world over which I enjoy absolute control, so that nobody gets the opportunity to incur my disapproval. But what about a second version, in which my disapproval is never provoked – not because I exercise control over people, but because each person freely chooses to act in ways of which I approve (or at least don’t disapprove)? How could anyone object to this on the grounds of freedom? Whether this is coherent depends on whether it is coherent to imagine a world in which everyone freely avoids doing anything I think they should not do. This brings the discussion, fairly obviously, into the venerable debate over theodicy. The world contains bad things, some arising from human agency. On the standard Christian picture, these bad things come about as a necessary consequence of the fact that human beings have free will. I do not aim to revisit, still less add to, these discussions. The point is that we have now arrived at a specification of the basis for non-interference, as both knowingly licensing sub-optimal acts, and not revocable in an idealised world. Earlier I noted that the range of candidates for such a non-contingent basis must be small. In fact, it seems that the basis must consist in a commitment to the value of freedom. I foresee that the activity will take place; I must think that a world containing the disapproved-of act will, pro tanto, be worse; and also that, nonetheless, there is reason to permit it. Even if I had unrestricted power to implement my ideals it would still not be better to prevent the acts in question. I must restrain my powers of prevention, or at least the extent to which I exercise them,23 in the name of creating a better world. This means letting the doers of the acts get on with doing them.24 In other words, the principle must involve granting the other person the freedom to act. Counterfactual imaginability can even be applied to freedom, provided that someone’s being free is compatible with their always doing what

23

24

These alternatives are, of course, distinct: they encapsulate the main point of difference between, respectively, proponents of ‘republican’ and ‘negative’ freedom. But since the question here is what powers I grant myself in my ideal world, they come in this case to the same thing. For republicans, in particular, there will be something objectionable even in the supposed ideal insofar as it grants a discretionary power: so here I need (obscurely enough) the power even to grant myself the power to imagine such a situation. It is far from clear that I myself cannot figure among the doers. If I can, I can tolerate myself. That is, I can act in ways of which I disapprove (pro tanto) in the name of bringing about a morally maximal world. See Chapter 10.

Toleration and the limits of idealisation

157

I want. But it is not clear why God could not have created a world in which people freely choose good over evil all the time.25 Alvin Plantinga and others attempt to show that though it is possible that a world exists in which people always freely choose to avoid evil, this is not a world which God could have created.26 This defence has in turn to ward off the objection that, given ‘middle knowledge’ (which Plantinga’s defence seems to demand), God could have confined himself to creating only those beings who would, at each point where they are faced with a choice between doing and avoiding evil, actualise the truth of a conditional that, if faced with such a choice, they will choose to avoid evil. But, if granting them freedom were conditional on their always doing what I wanted, that would not count as an attitude which respected the other’s freedom. In other words, the freedom on whose basis the tolerator extends toleration must be one consistent with the lack of ‘middle’ knowledge – the knowledge of freely made choices – or with knowledge that the tolerated person will choose to do what she, the tolerator, believes to be wrong. My commitment to freedom must be such that I am prepared to accept that the other person will act in ways of which I disapprove, but let them do so anyway. The one basis for permission which resists the challenge posed by counterfactual idealisation is freedom. Looked at from another angle, there is an obvious reason why the power condition should have this effect. There is a clear inverse relation between power and freedom. If, for example in the circumstances of toleration, I use my power to prevent some act which you wish to perform, I have prima facie limited your freedom.27 If counterfactually I would use power I in fact lack to forestall your doing something of which I disapprove, I would limit your freedom, even if, things being as they are, I lack this power.28 The counterfactual here is used as a way of getting clear about my attitude here and now.

25 26

27

28

J. L. Mackie, ‘Evil and Omnipotence’, Mind 64:254 (1955), 200–12. Or, more weakly, God might not have created such a world, so it is possible that God might not have created a world which both contains free will and from which evil actions are absent. See Plantinga, God, Freedom, and Evil (Grand Rapids, MI: Wm Eerdmans 2002). Only prima facie, because your capacity so to limit me may be something to which I have freely assented. See Williams, ‘From Freedom to Liberty’ in Williams, In the Beginning Was the Deed. If your merely prima facie freedom would be counterfactually limited, it can be argued that I act tolerantly, since I have warrant sanctioned by you for limiting your freedom in those circumstances.

158

Toleration and power

The picture further complicated Several conclusions may seem to follow. First, toleration has a dual linkage to freedom. On the one side, the tolerator must have the freedom to prevent the disapproved-of act which is conferred by having the power to do so, while also having the freedom to permit the act. The same, it seems, goes for the doer of the tolerated act: she certainly has ex ante the power to do it, if not the power not to do it.29 Hence toleration belongs in a structure of power which is, at least to a minimal degree, dispersed among different agents, though of course this in no way demands an equal distribution of power. Second, toleration depends on attitudes of the agent. These attitudes are explicated dispositionally. It depends on dispositions whose truth is partly secured on that of counterfactuals. Since virtues are dispositions of a certain kind,30 the above account furnishes materials from which an account of toleration as a virtue may be constructed. Third, the discussion suggests that justifications of toleration based on modus vivendi may be faulty. According to such justifications, warring factions have good reason to tolerate one another when, as a matter of fact, none has the power to enforce its views on the other(s). So modus vivendi justifications deny that self-restraint can only count as tolerant when it rests on principle rather than on a pragmatic compromise: political deals reached because neither side enjoys a monopoly of power may nonetheless merit being called tolerant.31 As I have suggested, however, the way in which we describe or understand their actions here and now depends in part on how they would act counterfactually.32 No doubt modus vivendi is often a real political achievement. But the upshot is that the parties to it will not count as individually tolerant if they accept the arrangement only because they lack the power to get something better. However, matters are less clear-cut than this summary makes them look. Politics has to work with a notion of embodied freedom rather than of freedom as mere indeterminacy. Political action addresses a more or less limited menu of options, and this clearly sets practical limits to idealisation. A regime of non-prevention may result because the limits 29

30 32

That is, some such acts may qualify as practically necessitated. This covers not just compulsive or physically constrained behaviour, but principled necessity too: compare Luther’s Here I stand; I can do no other. Luther’s profession would have had a different resonance had it been prompted by, say, lumbago. 31 Aristotle, Nicomachean Ethics, 1105b25ff. See Gray, Two Faces of Liberalism. On some views, these counterfactuals may lack determinate truth-conditions. If so, so may the corresponding judgements about whether A acts tolerantly in failing to stop P.

The picture further complicated

159

on power mean that there is no bearable alternative to it; the idea that non-prevention through sheer lack of power was not toleration motivated the original power condition. Perhaps less obviously, toleration may be limited, not because of intolerant attitudes, but because political power itself is limited. It can be a real political question, whether something is tolerable – whether, that is, in the circumstances, a certain action can be so described. In fact the objections cut deeper than this. It is not easy to sustain any hard-and-fast line between ‘costs’ and ‘principles’. Political concerns typically straddle this line: for instance, does a government motivated by desire to secure public order, or to achieve civic peace without jeopardising other public services, act from a ‘pragmatic’ concern for costs, or from principle? This provides an illustration of the wider point that understanding in political theory is ill-served by a blank dichotomy between moral ‘principle’ and self-interested ‘prudence’. When political authorities act on behalf of the public interest, they typically take account of diverse interests and seek to arrive at an all-in decision about what to do. Political action is usually an amalgam of principle and prudence, telescoped in the notion of the public interest. The further one goes in the direction of idealising, the greater the risk of inconsequentiality. I suggested that the only final basis for toleration that could survive counterfactual interrogation is a commitment to the tolerated party’s freedom, and perhaps not even that. The point is not that nobody can ever wield enough power to forestall every action of which they disapprove. Nor is it only that the counterfactual’s remoteness disqualifies it as a basis for judging actual motivations, though it is certainly open to question whether claims as to how someone would act in the imagined situation shows how to evaluate their conduct here and now. The problem is that there is no good reason to think that the counterfactual has the determinate truth-conditions it needs to do the job that theory assigns it. Suppose I had the supra-godly power that the imaginary case grants me; perhaps I would turn megalomaniac and not bother much about principles. It may be said that this introduces a gratuitous element into the fantasy. The intended response may be that, at any rate, I should bother about the principles. But then the task is to show why one way of specifying the counterfactual is authoritative over others, without leaving us back at the foot of the justificatory hill. Indeterminacy is integral to political toleration. Given what was said earlier, it will also often be unclear whether the reasons for nonprevention make an act one of toleration. When confronted with the brute fact of a disapproved-of action, the putative tolerator may still think counterfactually that it would have been better that the action had not

160

Toleration and power

been performed, and that she would have prevented it, had she had the power to do so. This goes beyond the abstract reflection that a world without the action would be better, pro tanto, than one with it. The thought is that of an agent. I argued that the only kind of reason for toleration which stands up to this counterfactual thinking is one which is based on commitment to the other party’s freedom. Beyond this, however, lies a more basic contrast between those who accept toleration as a principled strategy in response to the reality of power – including the reality that everybody’s power is limited – and those who reject this. One version of the rejection lies in the attitude of those who merely restrain themselves, and accept a modus vivendi, because they lack sufficient power to enforce their will. Another is the idealism of theorists who imagine a world in which principles do not conflict, because there is, ultimately, only one principle. Such a stance is detectable in the work of some but not all liberal writers past and present. Even if it is accepted that the principles can conflict in theory, liberal idealists imagine that empirical conditions can be counterfactually reinterpreted so that the conflict fails to arise in practice.33 Ideals or principles serve to constrain what one can do. They resist revision via the counterfactual idealising that underlies the power condition, since they themselves fix the content of the ideal. This prompts the broadly anti-Kantian thought that the ideal of freedom at which the counterfactual test arrives is too thin to gain much political purchase. This is not because an abstract ideal of freedom is politically irrelevant.34 It is that that ideal requires a normatively thickened account of circumstantial freedom, to which appeal is made, for example, in judgements about whether a policy or act is tolerant. That a decision – for instance, about whether to tolerate some practice or group – is political, stems from its relying on a specific notion of freedom, even while the abstract notion remains in the background as a challenge to it. Its being political requires, but at the same time cannot rest at, the idea of freedom as simple indeterminacy.35 In fact, the emphasis on radically contrary-to-fact possibilities also ignores the resources afforded by more modest forms of idealisation. One form that these more modest forms of idealisation can take is to ask how far an agent’s actions are those of a tolerant person – that is,

33

34

This stance can also be seen in apparently less extreme forms of idealism, where principles themselves are given, and come into conflict, but theorists devise strategies for eliminating conflict. The doctrine of double effect constitutes one casuistical strategy along these lines. 35 See Chapter 10. Or ‘undetermination’, as I refer to it.

The picture further complicated

161

someone with the virtue of toleration.36 Once it is granted that describing an act as tolerant need not commend it, there need be less puzzlement that coexisting after mutual attrition may be tolerant in that it embodies a certain form of self-restraint, without expressing tolerance as a virtue. That is true – and this was the initial plausibility in the power condition and the underlying notion of an ideal world – of those who settle on peace or modus vivendi while heartily wishing that the other side did not exist, would convert, or go away. Lack of power may also impose limits both on what sort of agents the parties to a modus vivendi can be, and on what normative descriptions of them and their actions are appropriate. The radically counterfactual notion of freedom explored above with regard to those tolerated has its counterpart in a similarly radical understanding of the tolerator’s freedom. This surfaced earlier in the idea that the tolerator might renounce even the principles by which counterfactual ideals are framed. Perhaps, one may say, the tolerator’s evaluative dispositions could and should have been different. Here again, the problem is not in filling out the pertinent counterfactual ‘in principle’. It is that there is no good reason in general to think that a shift from this world to the relevant counterfactual one lies within the tolerator’s power. Ethical dispositions, including evaluative ones, are not to be put on and taken off like a coat: it is plausible to think that their not being like this goes towards making them ethical dispositions. The picture I am resisting is of a pared-down notion of agency in which the supposed toleration turns out to be a misnomer, on the grounds that there is some counterfactual world where the tolerator lacks the ethical dispositions that she in fact has. The dispositions themselves constrain the power the agent has. One is invited to infer that because it is conceivable in principle that the tolerator might have been free of the dispositions engaged in toleration, she should not be accounted tolerant in the case under review. If however it does not lie within her power to slough the dispositions off, this is not something she can do. The lack of power to change one’s dispositions does not efface toleration. This is very different from a situation where someone backs off from persecution at the point of a gun. The limits on idealisation suggest limits on the scope for normative correction within judgements of toleration. An agent does not fail to act tolerantly merely because she acts from principles that she cannot slough off. How far principle bears on modus vivendi arrangements is a matter of judgement for actors and observers. Of course, pragmatic motivations 36

I sketch an account along these lines using Aristotelian materials in my Virtue, Reason and Toleration, ch. 3.

162

Toleration and power

may be as hard to disown as principled ones. In such cases the power condition may be met by default, so that the acts qualify as tolerant, or not intolerant: the agent simply cannot disown her commitment to avoiding the pragmatic cost. But those who forbear may still fail to express the virtue of toleration: the actors lack, for instance, the attitudes that characterise the tolerant person. It is a good question how far the circumstances of modern politics work systematically to stifle toleration as a virtue, by selecting against the attitudes that characterise it.37 ‘Power’, as a place-holder for agents’ capacities, may look like a purely descriptive notion: it is either a fact, say, that one can run a four-minute mile, or a fact that one cannot. But this pure fact-of-the-matter reading comes under growing strain when, as above, counterfactual idealisation is pressed. Insofar as agents are held to the ideal-world test, they are subject to an inquiry not about whether they are descriptively tolerant, but about whether they have acted in a way that makes their action an achievement, something that admits of commendation. The point if any of pressing the test lies not in debunking agents’ claims to tolerate, via wildly counterfactual hypotheses, but in seeing how far their acts square with their ethical dispositions. An observer can still take this view even if he thinks that those dispositions or their expression are misconceived. Conclusion This chapter has asked what is fixed and what is variable in thinking about toleration. On the view advanced here, toleration is located between pure powerplay and idealism. The space for toleration is thus limited, but not vanishingly small. The point, however, is that toleration is not in the place where we ideally would like it to be, since that is fixed by the very principles which toleration compromises. As such it is uncomfortable and often unstable. This makes it doubly unwelcome to ideal theory. The ironic movement traced by the argument of this chapter is that as we attempt to strain and refine toleration, purifying it of circumstantial sediments, they crop up again as aporias of theory. It is hard to stand toleration. It proceeds from accepting that the sub-ideal is given, and we have to do the best we can with it. That means accepting the fact of power as an ineradicable feature of political life, even while acknowledging the limits on what we have the power to do, in theory as in practice.

37

See also Chapter 3.

9

Tolerating ourselves, tolerating terror

It is often said that part of what it means to call a state ‘liberal’ is that it is committed to the ideal of toleration.1 Whatever that means, however, it cannot mean that a liberal state cannot classify certain acts as intolerable. Toleration is doubtless a good, but a pro tanto rather than an all-in good. That policies permitting citizens to speak their mind, or to form sexual liaisons free from state diktat, are tolerant, counts in favour of them, but falls short of trumping all other considerations. With some acts it is better, all things considered, to be intolerant – to treat certain acts as being intolerable – rather than tolerant.2 The possibility of misplaced toleration, of tolerating acts that are properly regarded as intolerable, is reflected in the modern political rhetoric of ‘zero tolerance’ towards, for example, certain kinds of public nuisance. Some acts lie further still beyond the pale of toleration, in the sense that their toleration would widely be regarded as not even good pro tanto. Acts of terrorism are widely regarded as intolerable in this way. They are often seen as epitomising what liberal societies cannot and should not tolerate. The political consensus on the intolerability of terrorism is impressive. Here are two examples from recent United Nations proceedings: VICTOR TCHATCHOUWO (Cameroon) said Member States had become aware that the tentacles of terrorism reached everywhere. Thirteen instruments against terrorism had been elaborated, but terrorism remained one of the most serious situations faced by humanity. It was intolerable.3

1 2

3

See e.g. Steven Kautz, ‘Liberalism and the Idea of Toleration’, American Journal of Political Science 37:2 (1993), 610–32. For an exemplary treatment of these issues, see Peter Jones, ‘Can Speech Be Intolerant?’, in Glen Newey (ed.), Freedom of Expression: counting the costs (Newcastle upon Tyne: Cambridge Scholars Press 2007). UN General Assembly debate, 9 October 2008, GA/L/3341. Debate on CounterTerrorism. www.un.org/News/Press/docs/2008/gal3341.doc.htm.

163

164

Tolerating ourselves, tolerating terror

Another statement to the United Nations was made by the French envoy Emmanuel Bonne: We know the threat that terrorism poses to states and peoples. It affects the civilian population, who are the primary victims of terrorism. It is indeed an intolerable threat against which we have no choice but to unite and act together.4

Consider finally this press release by the Justice and Home Affairs Committee of the Council of the European Union. The European Union is resolved to keep our society free, open and democratic. Because its founding values respect individual rights and point the way forward, it means to protect them stalwartly from all forms of extremism. Not only is terrorism intolerable, it is also unjustifiable.5

This last statement seems to distinguish between the intolerable and the unjustifiable. On most views, what is intolerable would be taken as necessarily unjustifiable – that is, to say that it is intolerable simply means that tolerating it cannot be justified. In some cases, however, there may be a discretionary right to tolerate acts that are unjustifiable, and which one would therefore not be unjustified in suppressing; or more simply, there may be pragmatic reasons for not suppressing an act even though that act itself is unjustifiable.6 Within those acts that may be tolerated, one can distinguish between those acts whose toleration creates a supererogatory good and those whose toleration does not: while on some views toleration is always required when it is permissible, non-deontic theories can allow that toleration can realise a good that is not morally required.7 At the same time, within intolerable acts, some may be tolerated while others must be treated as intolerable: there is a duty not to tolerate them. Relative to this spectrum, terrorism is usually seen as intolerable in the strong sense that intolerance of it is not just morally permissible, but required. However, as I shall argue, the toleration of terrorism, or something morally equivalent to it, comes closer to home than that. The notion of toleration should not be used selectively to flatten out moral complexities which modern liberal societies can hardly

4 5 6

7

Statement by Bonne, Political Counsellor of the Permanent Mission of France to the UN, 13 November 2009. www.franceonu.org/spip.php?article4236. European Council press release 6228/05, 24 February 2005. I argue this in ‘Against Thin-Property Reductivism: toleration as supererogatory’, Journal of Value Inquiry 31:2 (1997), 231–49. For a similar view, see Hagit Benbaji and David Heyd, ‘Forgiveness and Toleration as Supererogatory’, Canadian Journal of Philosophy 31:4 (2001), 567–86. Opinions differ as to whether there is a class of ‘suberogatory’ actions. The notion is incoherent if parity is thought of as being set by moral norms’ requirements, which supererogation may permissibly exceed.

The wider stakes

165

escape. Those societies often assert their identity by appealing to the value of toleration, but toleration should be seen instead as expressing their morally compromised nature.

The wider stakes Political toleration, when it involves only two parties, is transacted between the relatively powerful and the relatively powerless.8 Liberals and egalitarians embarrassed about that fact adopt various strategies for dealing with it, as has already been noted, the polar positions being either to accept that toleration involves these imbalances of power, and reject it accordingly, or to try to rescue it by transforming it into a value from which the imbalances have been purged.9 Some of these exercises in rethinking toleration effectively replace it with something less perturbing.10 Others do more to retain toleration’s distinctiveness, but make it hard to see how toleration keeps its shape as a political value. This seems to me true of attempts to blend liberal neutrality with toleration. In the sophisticated and illuminating work of Forst and Jones,11 the theory of political toleration sets out justificatory conditions under which normative consensus prevents problems of political toleration from arising. Rather, the conditions describe an ordered society where toleration already prevails. That account seems too comfortable. The danger lies in a bland assumption that a principle of public justification can be relied on to decide which private judgements of disapproval are fit or unfit for translation into public policy. Schematically, the principle will be either drawn narrowly, and will stir up again the normative disputes it was meant to resolve; or it will be broader, and then command broad acceptance in the abstract, but just for this reason be of little help when there are sharp differences over the acceptability or otherwise of some practice. Take the case of Holocaust denial. Some western liberal jurisdictions such as Austria ban it, while others, like the United Kingdom, do not. One can take different views on the substantive issue, but also on which policy is the more tolerant, or the less intolerant. If the public justification principle talks of ‘reasonableness’, it may remain unclear which policy is the more reasonable, and a familiar ping-pong of intuitions follows. Or something more specific-seeming, such as ‘reciprocal acceptability 8 9 10

A point brought out differently by Herbert Marcuse, ‘Repressive Tolerance’, and Wendy Brown, Regulating Aversion. See chapters 1 and 2. 11 E.g. Galeotti, Toleration as Recognition. See Chapter 5.

166

Tolerating ourselves, tolerating terror

by all affected’, may be adduced. But then either banners or permitters may plead the principle in aid, or the principle may prove to support a veto in both directions at once. Here as before, my main concern is with how this limits certain forms of idealised theorising about toleration. We have, at one end, groundlevel political judgements, such as those that affirm that ‘we’, as doers of politics, are the tolerators rather than the tolerated. On the other hand, in trying to elaborate conditions for toleration, such as the power condition,12 one is pushed towards idealisation so stringent as to void most judgements that political agents act tolerantly. Taking idealisation seriously runs afoul of the conditions that make judgements of toleration political. I set out what I see as a further extended illustration of this below. One can cut through the knot and judge that nothing answers normatively to the value of toleration – that its value is ‘provisional’ or illusory. But it is a perfectibilitarian fallacy to think that no loss of value need result if the knot is cut. What is lost is, precisely, an awareness of value as partial – that imperfection is the common lot. Such a claim need not rest on illusions about becoming ennobled through conflict, or some updated story about the white man’s burden or la mission civilisatrice. The value consists, at the limit, in awareness that value itself is partial. As classical liberals like Mill and modern thinkers like Isaiah Berlin and John Gray have recognised,13 there can be a loss of value in thinking – and attempting to put into practice the thought – that value itself can be maximally realised. If that is right, the doctrine of perfectibilitarianism is not merely flawed because of empirical difficulties in motivating people to do the right thing, or in coordinating their actions.14 The doctrine is flawed because it is threatened with incoherence – a threat that those who hold to it can fend off only by taking a depleted view of value itself. The perspective on value that someone would need to have in order to believe in the doctrine would be such that he had already obscured from his view some dimensions of value. It would be a view (to that extent) inferior to one that acknowledged its own constitutive limitations. If that is right, there can be value in toleration because there is value in seeing that there are limits on how far value can be realised in 12 13

14

See Chapter 8. See Berlin, e.g. The Crooked Timber of Humanity, ed. Henry Hardy (Princeton University Press 1998); John Gray, Enlightenment’s Wake (London: Routledge 1995), Isaiah Berlin (Princeton University Press 1997), and Liberalisms (London: Routledge 2009). Perfectibilitarianism is not, of course, to be confused with ‘perfectionism’ as a label used by some (e.g. Charles Larmore) for the liberal position opposed to neutrality. See Larmore, Patterns of Moral Complexity.

The wider stakes

167

practice, including political practice. This is only in part because a disposition to distrust one’s own evaluative judgements has value, as fallibilist arguments for toleration have maintained. It is also because there is value in recognising what is anyway inescapable, the blunt imperfection of political processes. Again, this is not only because meliorist projects often prove pernicious in practice. The value lies in recognising the fact that agency figures as part of the political process, with the compromises that that requires. One can uphold a picture of toleration as an achieved state, where citizens united on fundamental political norms accept restraints on their private beliefs, and this may serve as a regulative ideal of a tolerant political order.15 It is, to put it no stronger, optimistic to think that the diverse private beliefs that this ideal addresses can coexist with full agreement on the political norms devised to filter them. It is conceivable perhaps that they could coexist, though (to reiterate a point familiar from discussions of toleration)16 the stronger the private beliefs are, the harder it will be to come up with political principles, even of the ‘overlapping consensus’ form, to do the job of regimenting them. For the reasons already given, however, it is not obvious that political consensus is necessarily more desirable than consensus in private beliefs. This point need not rest on an ‘agonistic’ celebration of conflict as valuable in itself, but on the fact that value – even on a monistic view – generates conflict.17 Beyond this, even if it were desirable, there would still be problems of toleration raised by those who failed to abide by the political norms, and then sought to impose their private beliefs. It does not follow from the neutralist or respect notions of toleration that such persons (or their beliefs) ought to be suppressed. When these problems do arise, those in power – the representatives of the dominant consensus – will have to decide whether and how to use it against diehards.18 Things change. As regards current political données, the beneficiaries of reasonable judgements – variously, homosexuals, some ethnic minorities, anti-smokers, practitioners of male circumcision – either have not always so benefited, or may not continue to do so. Similarly for those deemed unreasonable. It is not that judgements of reasonableness in the abstract wag the political dog. Instead, the dog itself is in part fleshed 15 16 17 18

I have in mind particularly here Forst’s theory of toleration as ‘respect’ and the neutralist toleration theory of Jones; see Chapter 5. It is made, e.g., in Williams, ‘Toleration: an impossible virtue?’ I argue this further in ‘Value Pluralism in Contemporary Liberalism’, Dialogue: the Canadian Philosophical Review 37:3 (1998), 493–522. It does not follow merely because someone acts unreasonably that it is reasonable to use power against them, so there is also a question about what would justify using it.

168

Tolerating ourselves, tolerating terror

out by concrete judgements of what is reasonable. Someone might ask whether, and why, this matters. Maybe what should pass for ‘reasonable’ simply depends, acceptably, on what is taken as reasonable here and now.19 The main problem here is not that the consensual judgements may prove frail or transitory. It is that if the basis for the judgements is the brute fact of consensus, there will be nothing beyond that brute fact when persons and their practices excite the disapproval of parties to the consensus: which form the basic circumstances of toleration. The critical impulse behind both neutrality and the respect theory was to subject power to the test of justification. But the test, to have practical effect, needs to incorporate concrete judgements, ta endoxa, and in doing so the threat returns that power acts beyond justification. The problem faces any attempt to promote liberalism by socially coercive means, such as Mill’s ‘punishment by opinion’. It is not that the problem is worse for neutralist and respect theories than for anyone else: but it is no better for them, either. The problem is there, and is a constitutive one. Some things attract and merit disapproval, even though their suppression is unjustified. But when those in power express disapproval, they make them less eligible. That is one reason why the pure autonomy justification for toleration, as a way of handling counterfactual idealisation in the face of power,20 remains unsatisfactory. It may suffer, though, not only from a short-sighted view of those who are tolerated, but also from an insufficiently critical stance towards the judgements of the soi-disant tolerators. One way to bring this out is to notice judgements of the sort that are not made when one talks of toleration. No doubt this is often because the judgements are too outlandish to carry conviction. But the highly localised provenance of judgements about what is tolerable or not, reasonable or not, and so on, should sound an alert. Unduly swift judgements may disclose a failure of sensibility. Here the risk is not just of not tolerating what should be tolerated, but of ignoring biases in deciding what is, as it were, raw material for toleration. Insofar as we give the données to ourselves, it is up to us whether we take them at face value, but also whether to call up others for inspection. Perhaps the local judgements have such a grip that they cannot be disowned.21 But one job that theory can do is to note that a judgement is being made 19

20 21

For one version of such an argument, see Richard Rorty, ‘Postmodernist Bourgeois Liberalism’, Journal of Philosophy 80:10 (1983), 583–9; also his Contingency, Irony, and Solidarity (Cambridge University Press 1989). See Chapter 5. This is not to say that they can simply be discarded. They may be like synthetic a priori judgements in this respect.

Tolerating oneself

169

even when thoughts are dismissed in the most peremptory way – by not entertaining them in the first place.

Tolerating oneself I shall now suggest that one area where this may occur is in relation to self-toleration. My aim here is only to show that the notion of selftoleration makes sense, and that it can apply in real situations. Can one tolerate oneself? That may seem an odd idea. It appears to involve the kind of ad hoc psychological bipartition familiar from attempts to explain phenomena such as self-deception or akrasia. I have not tried to set out a formal set of necessary and sufficient conditions for toleration in this book, and it has been part of the argument I have put that no such set will suffice to understand the politics of toleration.22 Those conditions would however generally be thought to include the following. With respect to some act, X, there is an objector B who dislikes or disapproves of X; B has the power to prevent or otherwise sanction it; B does not exercise this power; B’s failure to exercise the power rests on a due reason, such as B’s commitment to a principle of autonomy.23 What is to count as a due reason here clearly demands further inquiry, and that is one reason why the conditions here fail adequately to represent political considerations. Some reasons may be ineligible, such as where B’s failure to exercise her power of intervention rests only on reasons of cost or inconvenience.24 Other kinds of reasons may get blotted out through the screening processes described in the last section. Take the reasons that debarred large sections of the population, until recently in human history, from participating in politics.25 In general, the 22 23 24 25

I myself tried to provide such a set in Virtue, Reason and Toleration, ch. 1. Newey, Virtue, Reason and Toleration, ch. 1; also Andrew J. Cohen, ‘What Toleration Is’. See Chapter 8. Cf. e.g. Spinoza, Tractatus Politicus 11§4: ‘But maybe someone will ask, if women are under men’s power by nature or convention. For if it’s by mere convention, then no reason forced us to exclude women from government. But if we consult experience itself, we shall see that it stems from their weakness. For nowhere do men and women rule together, but wherever on earth men and women are found, we see that men rule and women are ruled, and that by this arrangement both sexes coexist peaceably. However, the Amazons, famed as having ruled long ago, did not permit men to stay in their country, but reared only their female children, killing the males to whom they gave birth. But if women by nature were men’s equal, and were equally distinguished by strength and wit, in which human power, and so right, mainly consist; surely there would be some nations among so many different ones, where both sexes rule equally, and others where men were ruled by women, and educated so that they are less capable. As this happens nowhere, it can be absolutely affirmed that by nature women lack equal rights with men’ (my translation).

170

Tolerating ourselves, tolerating terror

project of amplifying the conditions so as to filter eligible reasons from the rest is a forlorn one.26 But the following argument does not depend on the contingency of reasons that enter into judgements of toleration. It aims instead to bring those judgements closer to home, and to suggest that certain forms of judgement get made at the expense of others. In particular, the transitive use of toleration is highlighted, while its reflexive use is largely ignored. The effect of this is to turn attention on one aspect of political agency while ignoring another. It displaces the bad to an external source, on which power then sets to work.27 Though the point I aim to make is not mainly one about the conceptual analysis of toleration, that point is advanced by means of such an analysis. As far as traditional theories go, the argument issues a challenge. If traditional theories deny that their analysis of the conditions for toleration makes self-toleration possible, they need to show why the conditions rule this possibility out. Of course, the idea that one tolerates oneself is not part of the political discourse of toleration. The tolerant are, in this discourse, those who extend toleration to an unsavoury or otherwise deficient other, who have nonetheless done enough to be tolerated. It belongs to the grammar, in Wittgenstein’s sense,28 of talk about toleration, that it effects these displacements. Suppose, then, that we set out the conditions already given like this, so that the individuals previously distinguished are now one and the same individual, A. Then the situation is this: 1. 2. 3. 4. 5.

A performs X. A disapproves of or dislikes X. A has the power to prevent or otherwise sanction the doing of X. A fails to prevent, etc., the doing of X. A’s failure to prevent, etc., the doing of X rests on a due reason.

The distinctive feature of this situation is that A himself is the one who has it in mind to perform X, while at the same time being the person who disapproves of, or dislikes, this action. Suppose A goes ahead and does X. Then it seems we can say that, as regards the doing of X, A tolerates himself. Two objections could be made at this point. The first tries to open up an asymmetry between the one- and the more familiar two-party cases of putative toleration. In the one-party case, it may be said, the action that is supposedly tolerated 26 27 28

See chapters 4 and 5. On this point, see Wendy Brown, ‘Tolerance as/in Civilizational Discourse’. See Ludwig Wittgenstein, Philosophical Investigations, tr. and ed. G. E. M. Anscombe (Oxford: Blackwell 1951), e.g. §371, §373.

Tolerating oneself

171

is either not eligible for toleration or entirely gratuitous. On the one hand, if A lacks the power of prevention, then clearly condition 3 is not met, and as the power condition is necessary for the exercise of toleration, A fails to exercise toleration. Suppose on the other hand that A has the power to decide whether to perform the action or not. The putative tolerator can, unlike in the familiar two-party cases, easily avoid creating a problem of toleration by the simple exercise of self-restraint: he can prevent the action merely by failing to do it. Then the question of toleration will not arise – there will be nothing on which to exercise toleration. So either way, it may be said, toleration fails to enter the picture. For pure deontological accounts, the only question is whether the act of toleration conforms to the supposed duty of the tolerator. It is a good question for such accounts whether they give any basis for preferring a world in which the occasion for exercising the duty arises, to one in which it fails to do so. But, putting that question to one side, if self-prevention conforms with an assignable duty, notwithstanding reasons for disapproving of the action in question, I can again be said to tolerate myself. In this case, then, it seems unobjectionable to say that I tolerate myself. More generally, where a disapproved-of action is necessary in order to achieve some other worthwhile end, we can speak of the agent as tolerating his own act. Certainly it seems that if the agent cannot in fact prevent the doing of X, despite its being his own action – if, as it is sometimes said, he cannot help himself, and performs the action, his disapproval of it notwithstanding – then there is no reason to think of his doing so as tolerant. That is, he fails to meet the power condition, even though X originates with him. This might be so if for instance the action was some form of compulsive behaviour. Then he would have no power to prevent the action, and so its going ahead would not count as tolerant. But compulsive behaviour is something of a marginal case, even within the modality of practical necessity.29 Not every action that conforms to the structure counts as compulsive behaviour. If the agent can prevent the action – that is, he can prevent himself from performing it if he chooses, but fails for due reasons to exercise this power – then there seems no reason to deny that he acts tolerantly in failing to prevent it. Moreover, there are cases in which, although the agent cannot act otherwise, his inability to do so does not void the judgement that he acts tolerantly, because this inability expresses his own evaluative commitments.30 29 30

See Chapter 8. See Chapter 8. My inability to prevent myself from performing the disapproved-of act may stem from my own evaluative dispositions. Here, my inability to act otherwise fails to breach the power condition. The fact that I find I cannot simply disown the relevant evaluations should not disbar me from exercising toleration here, just as it should not do so in two-party cases.

172

Tolerating ourselves, tolerating terror

A second line of objection urges that it would be better if the action were not performed. This confronts a more general problem for toleration, discussed in Chapter 8: surely an ideal world would not contain the tolerated act at all? Toleration may be better, given that the action has been or will otherwise be performed, but this is not to say that action-together-with-toleration is better, all things considered, than the absence of the action coupled with the absence of toleration. The fact that it is better to tolerate X, given that it is performed, does not show that the world is better off with X than without it. But in the self-toleration case, A must judge it better to perform it than not. But acts of toleration must be good on balance even to count as such acts. So, if one performs an act which it would be better, all things considered, not to perform, one fails to act tolerantly. However, this objection is mistaken, for more than one reason. Sometimes a tolerated action is so bad that it would be better were it not performed. It is an interesting question whether any action not so bad as to be intolerable is bad enough to make a world in which it is tolerated necessarily worse than one in which it was not done at all.31 But there is no reason to think that this ranking – best, no objectionable act X; second-best, X tolerated; worst, X not tolerated (with censure short of prevention) – fits all the cases in which toleration is at issue. For one thing, X often cannot be considered in isolation. The objectionable action, and the response to it, may be unavoidable if some other yet worse bads are to be avoided. In addition, the claim that any act, such as one of non-prevention, must be better all things considered if it is to count as tolerant, is mistaken in itself. There can be misplaced acts of toleration, and presumably the world might have been better had those acts, and the practices they license, not been performed.32 Maybe someone tolerates an act of religious bigotry through a misplaced commitment to freedom of religious expression: they tolerate

31 32

That is, not done because nobody cares to do it anyway, irrespective of others’ wish to suppress it. For this reason I demur from John Horton’s proposed resolution of (what I have elsewhere called) the ‘censorious tolerator’ problem, which arises when somebody extends the range of acts of which she disapproves, with her own actions held constant, so that it seems, contrary to intuition, that one can become more tolerant simply by spreading one’s disapproval more widely. Horton argues (‘Three (Apparent) Paradoxes of Toleration’) that the problem ignores the normative dimension of toleration, so that extending one’s disapproval would not count as tolerant unless the judgements underlying the disapproval were sound. On my contrasting view, one can argue that the dispositions constitutive of the virtue of toleration must have a sound judgemental base, but it is not definitive of tolerant acts that the actor’s motivating judgements be well grounded, or endorsed by an observer who describes her as tolerant.

Tolerating oneself

173

what should rightly be treated as intolerable. To describe an act as tolerant need not be to commend it. Still, it may be said that from the tolerator’s own point of view, it must seem better to perform the act of toleration than not to do so, and this is just what is missing in self-toleration. The tolerator’s decision is simply whether to tolerate, given that the tolerated act will otherwise go ahead. But the judgement that the tolerated act is bad in itself can be overborne in the first-person case by the same considerations that apply with two-party toleration. The reasons that make the tolerated act bad can be outweighed, or silenced, by the reasons for performing it. For instance, the objected-to act may be required in order to realise some other benefit. This may be one extrinsic to the act, such as promoting civic peace. The more contingently related the bad of the tolerated act is to the good of tolerating, the easier it will be to press the point that an ideal world will lack it. But, as we saw in Chapter 8, that goes for interpersonal toleration as well. The one difference is that, unlike the interpersonal case, the tolerated act can, apparently, be cancelled without interfering with the freedom of the person who would perform it. This may help to explain why self-toleration seems counter-intuitive. It may seem to require that the tolerator takes an external view of his own agency. In two-party cases, the circumstances of toleration extend to the notion of what will happen anyway: this objectionable act is going to be performed, and then the putative tolerator A has to decide what if anything to do about it. But, with self-toleration, it seems disingenuous or deluded for A to think that the act – his own act – is going to happen anyway. To think this way abstracts from the picture of A’s own status as an agent, downgrading it to being a par with clumsiness. Again, this seems to make the supposed tolerator one who suffers his own action, as in cases of compulsive behaviour.33 However, the self-tolerating agent need not think of the tolerated act as one that is going to happen anyway. He may take as deliberated or reflective a view of it as of any other prospective act. The deliberative choice is between doing the act and not doing it, as with any other act. What matters is the agent’s attitude towards it. The point can be put, indeed, using the counterfactual reasoning set out in Chapter 8: the reason for non-prevention or non-interference is concern for the agent’s own freedom, even though in an ideal world the action would not occur. Surely an agent has maximal power over his own action – that is what it is to be an agent. So, in performing an act which, at least in some respect, is sub-ideal – would

33

For an argument from similar assumptions, see Cohen, Rescuing Justice and Equality, 41ff.

174

Tolerating ourselves, tolerating terror

not occur in an ideal world – the agent respects the judgement that, despite this, a free agent can choose to perform it, this less-than-ideal aspect notwithstanding. The agent-independence lies in the fact that, even though the capacity for judgement is idiosyncratic to an agent, the bases for reaching judgement seem to the agent himself to lie outside him. In fact, the objector here relies on the same thought that was mentioned earlier, when discussing the displacements made by the grammar of political toleration. I argued there that the latter tends to foreground some kinds of agency at the expense of others: tolerators take themselves to be transitive agents acting on the tolerated, while their own conduct does not itself raise issues of toleration. In this respect there is only a partial view of the tolerator’s agency. Now it is objected that selftoleration fails for precisely this reason, that it treats as a non-agent what is in fact an agent – the tolerator, insofar as he acts in a way of which he himself disapproves. As I have argued elsewhere,34 two-way disapproval is in fact common in political toleration. The wider point however remains that the possibility of self-toleration inheres in the possibility that reasons conflict. As I have suggested, that creates deliberative indeterminacy, and with it the possibility of politics. In this sense, the possibility of toleration – of oneself or of others – is implicit in the possibility-conditions of politics itself. I come back to this point at the end. To take a specific illustration: I may in general disapprove of lying, yet lie in order to protect the confidence of a friend. Here it could be said that I fail to prevent myself from lying, despite my pro tanto disapproval of my own act.35 In this case: 1. 2. 3. 4. 5.

I lie. I disapprove of/dislike lying. I have the power to prevent myself from lying. I fail to prevent myself from lying. My failure to prevent myself from lying rests on a due reason (e.g. respecting confidentiality).

There is no suggestion here that lying is good in itself.36 Thus an ultimately ideal world would or at least could still be one in which the 34 35

36

See my Virtue, Reason and Toleration, ch. 5; also Chapter 3 above. This need not be an act I cannot help but perform – compulsive mendacity differs from altruistic lying. Even where I cannot help it, the grounding of the act on certain kinds of motivating valuations will still not mean that it cannot be regarded as tolerant, for reasons already given. It is, however, assumed, against some readings of Kant, that it is possible that lying can sometimes be morally justified. Whether Kant himself thought this is a vexed hermeneutic question.

Tolerating oneself

175

lie was not told, or indeed in which no lies were ever told. I may recognise this at the time of telling the lie. However, in the situation I have the choice of telling the lie and realising the benefit, or not telling it and not realising it – or, in some cases, I may find myself under a countervailing obligation to tell the lie.37 To accept the moral possibility of doing so is not to endorse the doctrine of double effect. I understand that as the claim that an action may admit of two descriptions, under one of which it is intended and the other not; and there may be a description, true of the action or its consequences, such that it would be morally unjustified if intended under that description, but morally justified if not intended under it.38 It is uncontroversial that in some situations a countervailing benefit may be realised only by incurring certain moral costs. Lying cannot in fact be fitted into the standard double effect scheme, since it is a necessarily intentional act: the liar must intend to deceive her victim. So it cannot be that the lying is a foreseen but unintended means to some ulterior good end. This is also true on those deontological accounts which treat toleration as simply required by certain kinds of moral consideration.39 In fact, the above analysis does not rely on claims about the justifiability of the lie at all. Perhaps someone who tells a lie that conforms to the structure above fails to act justifiably. But it does not follow that she does not tolerate her own act of lying. To do that, she has to believe that there is a due reason which warrants lying in the circumstances, and people often lie while holding such a belief. They often also think that lying is pro tanto wrong: to that extent there is a reason to disapprove of their own act. They could simply keep quiet and not lie. But, as I have already argued, it is not a condition of toleration that a world in which the objectionable act is performed and is tolerated must be seen as worse than one where the act is not performed. So even if lying is circumstantially unjustified, or unjustifiable across the board, it may still be tolerable – that is, capable of being tolerated, regardless of whether the world that contains a certain act of lying is better or worse than one that does not. It is a further claim, that even if lying is always unjustifiable, it may sometimes be justified to tolerate it. The conclusion

37 38 39

On this point, see my ‘Political Lying: a defense’, Public Affairs Quarterly 11:2 (1997), 93–116. As this shows, it is hard to formulate the doctrine clearly. For a powerful critique, see Alison McIntyre, ‘Doing Away with Double Effect’, Ethics 111:2 (2001), 219–55. Jones perhaps takes this view, e.g. in ‘Making Sense of Political Toleration’, but as already argued in this chapter, this is not internal to the neutralist theory of toleration.

176

Tolerating ourselves, tolerating terror

that one can tolerate one’s own lying does not rely on normative claims at all, only on claims about the liar’s normative beliefs.

Toleration and terror I conclude that self-toleration is coherent and can be elaborated from the conditions set out in standard accounts of toleration, such as those given above. It is a separate question whether the conditions in which ‘toleration’ and its cognates are used in political discourse are receptive to the idea of self-toleration. Received discursive oppositions between tolerators and tolerated tend to displace the idea that political agents tolerate themselves rather than others, but presumably part of the aim of conceptual analysis in political theory is to appraise rhetorical données critically rather than merely take them at face value. As Wittgenstein said, philosophical questions arise when language goes on holiday.40 A similar rhetorical dysplasia attends the term ‘terrorism’ and its cognates. As with those who are tolerated, it belongs to the political endoxa that those spoken of, but not those who speak, may be terrorists. Consider the following argument. 1. States such as the UK and US engage in acts of war, e.g. invading Iraq in 2003. 2. In modern conditions, a foreseeable consequence of engaging in acts of war is that civilians are killed. 3. In the context of war, civilians are innocent persons. So (from 2 and 3) 4. In modern conditions, a foreseeable consequence of engaging in acts of war is that innocent persons are killed. So (from 1 and 4) 5. In modern conditions, states such as the UK and US engage in acts of war, whose foreseeable consequence is that innocent persons are killed. Furthermore, 6. The acts of war mentioned in 5 are intentional acts; and 7. For all acts X, if an agent A does X intentionally, and X-ing has some foreseen consequence C, then in X-ing, A intentionally brings it about that C. 40

Wittgenstein, Philosophical Investigations, 19, §38.

Toleration and terror

177

So (from 5, 6 and 7), 8. In engaging in acts of war, states such as the UK and US intentionally bring about the death of innocent persons. And, 9. For all Z, to bring it about that Z dies is to kill Z; 10. States such as the UK and US engage in acts of war for political objectives. So (from 8, 9 and 10), 11. In engaging in acts of war, states intentionally kill innocent persons for political objectives. 12. Any act which involves the intentional killing of innocent persons for political objectives is an act of terrorism. So (from 11 and 12), 13. States that perform the acts mentioned in 5 engage in acts of terrorism. On what grounds could the conclusion be resisted? As empirical claims, premises 1 and 2 are beyond serious dispute. In some cases it might be argued that 3 was false to the extent that civilians may voluntarily support or perpetuate the regimes involved in armed conflict. But this is harder to sustain with those civilians, such as those Iraqi civilians killed during the 2003 invasion, who are subject to non-democratic and highly repressive regimes. Nor is 6 seriously to be disputed in general, though of course some killings in war – such as some cases of ‘friendly fire’ or ‘collateral damage’ – may not be intended under some morally relevant true descriptions of them. An obvious stage at which to reject the argument is at 7, and hence the inference of 11 from 8, 9 and 10. The premise is certainly controversial. It amounts to the denial of the doctrine of ‘double effect’, discussed earlier in relation to lying. My aim here is not to refute the doctrine, which would require much fuller discussion.41 One can certainly affirm that intentionally doing something X which has a foreseen effect C differs from intending to produce C: for instance, one can argue that what might be called the success-conditions of X, the conditions in which the 41

See e.g. McIntyre, ‘Doing Away With Double Effect’; also Warren Quinn, ‘Actions, Intentions, and Consequences: the doctrine of double effect’, Philosophy and Public Affairs 18:4 (1989), 334–51; Frances Kamm, ‘Physician-Assisted Suicide, the Doctrine of Double Effect, and the Ground of Value’, Ethics 109:3 (1999), 586–605.

178

Tolerating ourselves, tolerating terror

act of X-ing succeeds, will exclude C if producing this effect is not intended. Thus the act of war’s success-conditions, it may be said, do not include the killing of civilians.42 Here we can put a version of Chapter 8’s idealisation test: the war-maker can be asked whether he would prevent the deaths of civilians in a world in which he had maximal power to do so. Certainly the killers could be said to have achieved all that they wanted to achieve in this counterfactual situation. This test might then be thought to partition those acts which involve the intentional killing of civilians from those where such killings, though foreseen by the killers, are unintended. The trouble is not that the distinction cannot be drawn between intending an effect and foreseeing that one will bring it about, but that it begs the question to assume that the distinction between intended and merely foreseen bads carries the moral weight that defenders of the doctrine claim for it. Robert Nozick gives the example of someone who enjoys swinging a baseball bat, when the only place where this can be done is in front of a cow, whose skull is then smashed. The fact that the bat-swinger does not want – even does not intend – to smash the cow’s skull does not seem to remove his moral culpability.43 Even if the casuistical distinction can be sustained between action that is intended as an essential means to some end and side effects, as it presumably can in the bat-swinging case, this does not show that it does decisive moral work. The tenability or otherwise of double effect is not really the point, unless its truth undermines the moral equivalence that the argument sets out. If the doctrine is false, the non-equivalence will have to come from somewhere other than a moral distinction between intentional and merely foreseen killing. This distinction is precisely that so often relied on by those who seek to draw a clear moral line between killing in war and terrorism, as assumed by the tag ‘the war on terror’. It is far from obvious that the doctrine does the work wanted from it in this instance. The terrorist might argue, using double effect, that the doctrine exonerated him as well. The argument would again be that the collateral costs of terrorist bombings, and so on, were a foreseeable but not intended consequence of the political struggle in which he was engaged. For example, an IRA bomber might argue, using the success-conditions 42

43

Modern discussion of the double-effect element in ius in bello conditions sometimes slips from its main earlier application, namely the wrongness of killing combatants, to civilians. See e.g. Michael Walzer, Just and Unjust Wars (New York: Basic Books, 1977); Judith Lichtenberg, ‘War, Innocence, and the Doctrine of Double Effect’, Philosophical Studies 74:3 (1994), 347–68. Nozick, Anarchy, State, and Utopia (Oxford: Blackwell 1974), 36–7.

Toleration and terror

179

notion set out earlier, that civilian casualties were a foreseen but unintended outcome of the struggle for a united Ireland. The IRA would have succeeded in its aims if it had induced Britain to withdraw from the island of Ireland, without killing civilians. If it is said that planting a bomb in a shopping centre, say, must be intended to kill civilians, the same goes for many acts of war, such as the bombing campaigns in the European and Pacific theatres in World War Two, where the large-scale killing of civilians was certainly part of the aim.44 It should be noted that double effect, which standardly figures in the conditions for just war-making, plays an ambiguous role. The condition was traditionally taken to justify only the foreseen but purportedly unintended killing of combatants: it was devised to handle the fact that all foreseen killing was taken to be wrong pro tanto, even in self-defence, and to furnish a justification for nonetheless killing enemy soldiers. But now it is cited to justify the foreseen but unintended wartime killing of civilians.45 Warren Quinn distinguishes cases where people are harmed by the agent’s intentionally ‘involving them in something in order to further his purpose precisely by way of their being so involved’ from ‘harmful agency in which either nothing is in that way intended for the victims or what is so intended does not contribute to their harm’.46 His point is not that there is anything necessarily morally objectionable in involving someone in one’s purposes by way of their being so involved – that is true, say, when one uses the services of a bus driver to reach one’s destination. It is rather that when one involves another in this way, one’s intentions regarding them differ from those when the other’s agency is adventitiously involved in one’s own purposes (as with hitching a lift somewhere, and thereby taking advantage of the fact that the driver is going in that direction anyway). Quinn’s point would then be that a pro tanto morally wrong form of treatment may be justifiable if it involves another adventitiously in this treatment, but not if it involves them by way of their being so involved. One might think that Quinn’s formula offered a way to distinguish the killing of civilians by the state from those by terrorists. Surely the terrorist involves civilians in killing precisely by way of their being so 44

45

46

See e.g. Donald Bloxham, ‘Dresden as a War Crime’, in Paul Addison and Jeremy Crang (eds.), Firestorm: the bombing of Dresden (London: Pimlico 2006); Michael Bess, in his Choices Under Fire: moral dimensions of World War II (New York: Knopf 2006). See e.g. the Mount Holyoke version of the just war principles: ‘The deaths of civilians are justified only if they are unavoidable victims of a deliberate attack on a military target’, which seems even to dispense with non-intendedness. www.mtholyoke.edu/acad/intrel/ pol116/justwar.htm. Quinn, ‘Actions, Intentions, and Consequences’, at 343.

180

Tolerating ourselves, tolerating terror

involved; whereas in the case of war, what is intended by belligerents may not, in intention or in fact, harm them? In pressing this line of argument, one might distinguish between what might be called the intrinsic and extrinsic features of someone’s purposes, where the former but not the latter features are essential to its being the kind of purpose that it is. Terrorists’ blowing up civilians is not, it may be said, extrinsic to their purposes: rather it is intrinsic to their activity as terrorists. The point of terrorist acts is, precisely, to destroy human life, and the terrorists’ purposes cannot be understood in isolation from this. Three things can be said in response. First, as noted before, if there is a moral distinction in this area, it seems to concern not intention but an evaluation of the deed irrespective of its success-conditions. As in Nozick’s example, the action may incur unacceptable moral costs regardless of the agent’s intentions, or whether the cow is involved in one’s bat-swinging by way of its being so involved: that returns us to the initial question, whether a certain moral assessment is right. Second, if that assessment can incorporate idealised success-conditions, that will work as well, in respect of ultimate objectives, for the terrorist as for the state. Few terrorists murder simply for its own sake, and insofar as the murders they commit form part of an ulterior strategy, such as one of liberation, they can avail themselves of double effect as well as state belligerents. Finally, the argument insofar as it bears on toleration is not directly about moral justification. It may be that Quinn’s formula offers a way of distinguishing morally between some state and some terrorist killings of civilians. On any plausible view, all these killings involve moral costs, and the decision to bear them, justifiable or not, instantiates the structure of toleration. Whether or not a particular instance of it is justified is a further question. The notion that we tolerate our acting in ways morally indistinguishable from terrorists seems to perpetrate a double solecism. I have argued that acts of war incur moral costs such as the foreseen killing of civilians, and that without something like Quinn’s reformulation of double effect (and perhaps even with it), it is hard to draw a sharp moral line between such acts and acts of terrorism. I have further argued that the conviction that there must be some distinction between ‘us’, as state actors or those they claim to represent, and ‘them’, the terrorists, prompts a quest for clear moral line-drawing. Even if one is convinced that some version of double effect will justify the line-drawing, the killing of large numbers of non-combatants in wartime,47 such as the large-scale bombing of 47

See Iraq Body Count data for detailed statistics about violent deaths in Iraq following the 2003 invasion: www.iraqbodycount.org/analysis/numbers/2011/. This put the early 2012 level of civilian deaths directly due to the US-led coalition at some 14,718.

Toleration and terror

181

civilian populations during World War Two, would still be plausibly thought to incur moral costs. That raises the question of toleration, even if the costs are thought justified. Do we then tolerate the killing of innocents by ourselves? In Chapter 8, I discussed the idealisation test, which set a high bar for acts of toleration: I tolerate some act only if I would accept its commission in a world where I had maximal power, including the power to prevent it pre-emptively. As I argued, this test has little force in the world as it is, where agents lack such power and political judgements usually address the range of options open to them. But if one accepts it as a test of toleration, it can also be used by terrorists to deny that they really tolerate the killing of civilians, just as it can by liberal states who kill civilians in acts of war: a world where they had maximal power would achieve the strategic end, without killing the civilians. Conversely, if that counterfactual thought is too thin – too remote from the world as it is – to undo the judgement that terrorists tolerate their own acts of terror, it can hardly perform a similar task for liberal states who plead on parallel grounds that they do not tolerate the killing of civilians in war. That is not the sort of thing that liberal states say about themselves. My ulterior aim is not to argue for a moral equivalence between war-making and terrorism. It is to argue that liberal democracies endorse – and I have suggested, tolerate – acts whose moral distinctness from those of terrorism is at any rate sufficiently dubious to make puzzling, on their purely analytical merits, the treatment of the latter as a paradigm of the intolerable. Liberal democracies tolerate acting in ways that, at least, need a good deal of casuistical apparatus to distinguish them from acts of terror. Politics largely takes grammar at face value, and currently in this grammar, toleration is usually a good thing that we do, whereas terrorism is a bad thing done to us. That creates a dual repugnance to the idea that we tolerate our own terrorist acts. There are structural reasons why politics works with norms of this sort, rather than operating critically on them, one being that politics as a deliberative process selects for conflict.48 Maybe all this is wrong. But the intuition that liberal states’ acts of war do not differ markedly from terrorists’ is likely to present itself much more clearly than the moral distinctions that purport to ground it; and similarly mutatis mutandis with the thought that we, as tolerators, are not those tolerated – least of all by ourselves.

48

See Chapter 2.

182

Tolerating ourselves, tolerating terror

Conclusion I have staked out the tension between the analysis of toleration, as it relates to terrorism, and the pragmatic conditions in which political judgements are made. How, if at all, that tension should be alleviated is a further question. The aim has not been to argue that liberal states tolerate their own terrorism, so much as to encourage questioning the impulse to rejection that the claim invites. One casts about for something to justify the idea that the claim must have something wrong with it. That perpetuates a one-sided view of toleration as transitive and irreflexive. The maximal-power test partially corrects this, insofar as it appraises reasons for toleration and promotes a view of the tolerated as free agents, rather than merely as transitive objects. That is a strength of the respect theory of toleration.49 But there are limits to how far that can go. Politics is a realm of determinate freedom, in that freedom is presupposed as a condition of joint action, while any specific application of it to policy – such as tolerating something rather than not – in turn raises a question about freedom. This is a constitutive tension within politics. The freedom that makes for politics is unstructured in the abstract, but demands content in the here and now of political encounter. That content depends not on free agreement to terms of association, as in classical contract theory, but on the manifest fact that no such agreement can be made, even though free association is the premise of collective action; at the most general level, toleration is implicated in political conflict because that is the condition of political freedom.

49

See Chapter 5.

10

Toleration, free speech and the right to lie

Liberal democracies uphold strong free speech norms. That is, indeed, an important aspect of their liberalism. The First Amendment to the US Constitution firmly proscribes congressional interference with free speech, and the amendment has generally been interpreted broadly by the judiciary.1 Article 10 of the European Convention on Human Rights similarly endorses the ‘freedom to hold opinions and to receive and impart information and ideas without interference by public authority and regardless of frontiers’.2 But speech remains contested, as recent events have shown. The murder of the Dutch film-maker Theo van Gogh and of Hitoshi Igarashi, the Japanese translator of The Satanic Verses, as well as the would-be murderous attack in 2009 on the Danish cartoonist Kurt Westergaard and the enforced exile from the Netherlands of van Gogh’s associate and erstwhile parliamentarian Ayaan Hirsi Ali, show that free speech remains central to public life. In 2012 a scurrilous film trailer mocking the Islamic religion released on the internet was blamed for a wave of violence which claimed among its victims the US envoy in Benghazi.3 Indeed, in law, free speech is not in fact as undiluted as liberal rhetoric sometimes suggests. An obvious example is legislation against ‘hate’ speech directed against racial groups, the disabled, homosexuals, religious sects, and so on. Certain jurisdictions, like the UK, have extended the law in these areas. Legislation outlaws Holocaust denial in a number of countries, including Germany, Israel, Austria, France, Portugal, the Netherlands, Belgium, Spain, Luxembourg and Poland. Meanwhile lobbies often move from disapproving of some piece of speech or some speakers, to arguing that it or they should be stifled. Such was the case, for instance, 1

2 3

Cf. e.g. Buckley v. Valeo, 424 US 1 (1976), which found that political candidates’ constitutional speech rights were violated by imposing limits on the funds which they could contribute towards their own election campaigns. See www.hri.org/docs/ECHR50.html#Convention. Subsequently US intelligence indicated that the attack had been planned before the film’s release.

183

184

Toleration, free speech and the right to lie

with the Satanic Verses dispute in the UK in the late 1980s, and latterly with attempts by Christian activists to cancel the television screening of Jerry Springer: the Opera, or by anti-fascists to prevent the leader of the far-right British National Party from appearing on a BBC television programme in 2009. Speech has always been central to the politics of toleration. The most obvious reason for this is that speech pervades most human activity. This is where discussion should begin, rather than by claiming that some use of speech, such as to express beliefs, is the most important or basic one. By keeping a broad view of speech, we can come to understand not only why its protection – denying attempts to deny speech, as I shall call it – matters, but also how free speech precedes and conditions the political context of toleration itself.

Tolerable and tolerant speech Toleration is clearly double-edged in relation to free speech, as the continuing ‘speech wars’ over such matters as ‘politically correct’ language show. Summarily put, does the use of epithets like ‘poof ’ or ‘wog’ exhibit intolerance, or is it intolerant to censure their use? Of course, it may be that using such language and censuring it both display intolerance; perhaps the difference is that, say, its censure, but not its use, is justifiably intolerant. But, in any case, if such a use of speech counts as intolerant, it seems an unrestricted principle of free speech will promote or permit intolerance. The idea that speech is ‘free’ often seems taken to suggest that it is costless. But it may incur externalities, as well as costs gauged in terms of speech itself – for instance, when someone gets a say at the expense of somebody else. We are left unclear about the relation between free speech and toleration. Nonetheless, cases where a speaker faces attempts to silence her seem at once to raise the question of toleration.4 Often, as in Mill’s On Liberty, freedom of speech is held to follow from a more general principle of freedom or toleration: people have a general right to freedom – in particular, to act freely when their conduct excites the disapproval of others, the basic circumstances of toleration.5 One form of conduct that often excites others’ disapproval is speech; so speech benefits from a more general principle that protects freedom of conduct in the face of 4 5

See Peter Jones, ‘Can Speech Be Intolerant?’ in Newey (ed.), Freedom of Expression: counting the costs. Of course, it may be that toleration licenses free speech, but some who exercise this freedom act intolerantly.

Tolerable and tolerant speech

185

others’ disapproval. Thus it is tempting to think that there is some general principle of toleration, and free speech is just the application of this principle to disapproved-of speech. In failing to prevent disapprovedof speech, one applies to it a broader principle of toleration, which accordingly is prior to it. But is this right? I shall suggest we should see the relationship as holding, more or less, the other way round. Some forms of toleration follow from a commitment to free speech. Other forms of toleration, which do not concern speech, are parasitic upon this commitment, given that toleration requires a discursive and transactional basis. The underlying picture is one in which free individuals are thought of as combining with one another on mutually agreed terms. Freedom of speech follows from a more basic commitment to freedom of association. I will illustrate the general claim via the discussion of a case that poses problems for my theory, involving an appeal to a right to do wrong. It might be thought that cases of toleration were always of this form, but in fact this is mistaken. First, the behaviour in question may be objectionable, say on aesthetic grounds, without being morally wrong; second, someone who elects to tolerate it may do so not because he recognises that the other person has a right to non-prevention, but because his own set of reasons includes either non-right-supporting obligations to do so, or grounds that make his own non-prevention supererogatory.6 At any rate, the example is problematic because it seems to frustrate the very purpose that, as I shall argue, gives free speech its rationale. Defending free speech may be described as a matter of denying denial – that is, denying attempts to deny people a say, or to deny the saying of certain things. The aim, in formulating the matter this way, is to move attention away from the basis for free speech to the real site of contentions over speech: attempts to stop speakers from having a say, or to stop something from being said. Defending free speech means resisting attempts to deny speakers a say, or a mode of expression, or a hearing. It may be said that these ways of denying denial must rest on some positive claim, for example about the sanctity of conscience, or the value of truth, or the role of free speech in promoting autonomy.7 But the most basic 6 7

On this last point see my ‘Against Thin-Property Reductivism’. The truth-based defence can be found in canonical works such as Milton’s Areopagitica and Mill’s On Liberty. Conscientious justifications are also central to the lexical priority accorded to the Liberty Principle in Rawls’s A Theory of Justice. For the autonomy justification, see David A. J. Richards, ‘Autonomy in Law’, in John Christman (ed.), The Inner Citadel: essays on individual autonomy (New York: Oxford University Press 1989), 246–59; for a critique, see Susan Brison, ‘The Autonomy Defense of Free Speech’, Ethics 108:2 (1998), 312–39.

186

Toleration, free speech and the right to lie

statement of the case for denying denial may simply be that attempts to stifle a speaker or speech seek to block communication, which may have any one (or more) of a number of aims – as many aims, in fact, as one person may have when engaging in communication with another. The reasons for engaging in communication are multifarious, and this undermines attempts to justify free speech via some supposedly single or overriding purpose of speech. I will argue that these reasons can be traced back to freedom of association: there is no general ground for restricting in advance the terms on which individuals (including corporate individuals) may seek to associate with one another. What, if anything, makes speech special? Consider first what it is about speech that makes it liable to censorship, and makes speech eligible for protection. Often, in political argument, the debate is over whether an acknowledged good justifies the acknowledged bad(s) allegedly needed to achieve it. For example, in debate over national security, the goal of preventing terrorist attacks is an acknowledged good, and loss of civil liberties an acknowledged bad; the dispute is over whether the latter must be the price of achieving the former.8 But speech is standardly treated as immune to such balance-of-goods reasoning, as in Rawls’s case for the lexical priority of basic liberties. Whether or not to permit it is viewed as peculiarly important by liberals and censors alike. It is thus unusual for those favouring censorship, in disputes over free speech, to argue that, compared with other goods, speech is not all that important. Its importance is enlisted as a reason for censoring it. Indeed, an oddity of political conflicts over censorship and free speech is that both sides often seem to focus on the selfsame features of speech. The fact that speech aims to get something across to an audience is what is held both to make it valuable and to make it objectionable. It is not that those favouring censorship spotlight one feature of speech, while defenders of free speech emphasise another, or even that one side stresses the costs or benefits of speech, while the other highlights external benefits or costs. Instead, the parties share a view of speech’s importance, but differ in how they evaluate its permission and prevention. On the one hand, features of speech that seem intrinsic to it ground calls for censorship, as when it communicates certain ideas. On the other hand, calls for free speech often also rest on claims about the nature of speech, such as its alleged purpose. This is not to deny that arguments over censorship sometimes turn on different aspects of speech. An important contrast, not always

8

See Chapter 7.

Tolerable and tolerant speech

187

acknowledged in the extensive literature on free speech, is between freeing the content of what is said, and freeing people so that they have the opportunity to speak their mind, a contrast bracingly captured by Meiklejohn’s proviso that in public debate ‘what is essential is not that everyone shall speak, but that everything worth saying shall be said’.9 Speech may be objected to, and thus its denial sought, on one or more of several counts. For instance, someone may object to the content of what is said (what); or to the way in which it is said (how); or to the person who says it (who); or to the context in which it is said (where/when).10 Take some controversial question of speech, such as whether to grant a platform to far-right speakers, a conflict over opportunities to speak; or some issue of offence, as with the screening of Jerry Springer, which concerned the musical’s content. Sometimes conflicts over speech fuse concerns about opportunity and content, as when battle is joined over whether school curricula may include the teaching of ‘creation theory’. Again, the objection may be to the mode of expression of content which is not prescribed when expressed in some other form, such as the famous case of Cohen during the Vietnam War.11 Terrestrial television broadcasters generally observe a ‘watershed’ that limits the presentation of content to certain contexts, and material deemed transmissible in one medium may be censored in another. But each of these modalities – that it can communicate a certain content, that the content can be imparted in different ways, by different speakers, and in different contexts – seems to be inherent in speech. Thus, while ‘free speech’ aims to remove restrictions on one or more of these modalities of speech, opponents of censorship also appeal to features of speech in making their case. Take for example the conscientious justification of free speech. Those who have defended free speech on these grounds have seen speech as the outward tokening of the inward voice.12 But those who aim to stifle speech do not usually deny that speech can have this function, or that it is a fundamental use of speech. Indeed, it is precisely because speech is a vehicle by which the bearer of one conscience communicates with another, that it has been held necessary to suppress it. Again, it has sometimes been argued that 9 10 11

12

Alexander Meiklejohn, Political Freedom (New York: Galaxy 1960), 26. See my ‘Denial Denied: freedom of speech’, Amsterdam Law Forum 2:1 (2009). http:// ojs.ubvu.vu.nl/alf/article/view/109/198. Cohen v. California, 403 US 15 (1971). The accused had entered a courtroom wearing a jacket that bore the slogan ‘Fuck the draft’, where legal action was initiated because of the offensiveness of the expletive, rather than the view expressed about conscription. For doubts about whether conscience can be used to ground a general free speech principle, see Alan Haworth, Free Speech (London: Routledge 1998), 185ff.

188

Toleration, free speech and the right to lie

censorship frustrates the prime purpose of speech, namely communicating truths.13 Those who oppose freedom of speech do not usually rest their case on the claim that this is not an important purpose of speech. It was precisely because speech is an important vehicle for expressing truths that it was deemed to be so important to subject it to official regulation, by for example a licensing system. For free speech advocates, it is the very fact that speech is used communicatively that explains why it needs protection (except when, as in what I called the ‘Bootleg Argument’ in Chapter 6, it is said that speech will prevail in any case). But that is also what leaves it open to censorship. Speech communicates thoughts deemed dangerous or bad in themselves, or gives a voice – and a hearing – to those whom others would silence. So both objectors to speech and its protectors hold that it is communicative. That is, speech is a two-party (or more) process that is typically transactional. Seeing free speech as the application to speech of a generic toleration principle fails to do justice to the way in which speech is special. Although the impulse behind censorship clearly often lies in disapproval of what is said, speech is not just one more thing one can disapprove of, like dress, diet, religion, sexual practices, and so on. Speech not only prefaces association, but also instantiates it – to engage in speech is almost always to engage in association.14 This feature of the importance of speech comes out when we consider the principle of freedom of association: the freedom individuals have to pool their agency on agreed terms. This involves the coming together of individuals to form some joint enterprise, which expresses the collective will of the participants. Contractual forms of association exemplify this process. Consider, for instance, the pooling of agency involved in a contract, such as one for the provision of a service. The premise of civil contractual arrangements – for example, during litigation arising in case of breach – is that those bound by the contract have freely agreed to accept its terms. That is, contractual validity assumes that those bound by it signed as free individuals. To forestall misunderstanding, this is not to say that political association should be seen as the result of a contract. But the account given here of free speech assumes that

13 14

The pro-censorship position need not deny that speech conveys truths, only that truth is best promoted by deregulating speech. The qualification ‘almost always’ is intended to exclude soliloquising, muttering in one’s sleep, etc. As I shall suggest, the controversial uses of speech are precisely those cases where, as is usual, it is used communicatively, i.e. to convey semantic and illocutionary content to an audience.

Tolerable and tolerant speech

189

participants, in order to pool their agency, proceed on the basis that they are aiming at agreed terms of association. The freedom involved should be understood as verbally dependent. The parties to a contract, or other form of civil association, have to conduct their negotiations verbally, as a condition of freely assenting to the final terms. The association, to be free, depends on speech as the medium of negotiation. To that extent, speech is prefatory to an act of free association. Someone may respond that the process by which people negotiate an agreement over a contract is a highly specialised use of language, and cannot be held up as the model for speech in general; speech does not usually involve formal negotiations of this kind. But in most speech, in humdrum as in more formal contexts, the parties’ assent to the interaction is assumed. Even where, as often, one party or the other would probably rather not be engaged – such as those on trial in criminal proceedings – they nonetheless may have good reason for seeing the process through rather than abandoning it, in that the only alternative is worse. Withdrawal from communication is usually a move that can be freely made. So the freedom of association that marks contractual agreements, and the negotiations that precede them, surfaces in communicative situations generally. To be sure, one can then ask about the general conditions in which people freely communicate. Often, as in the courtroom, communication occurs against a backdrop of coercion. Here I can only deal summarily with a matter that calls for much fuller treatment. Political argument aims to frame terms on which those involved and affected can concert joint action. At the most general level, those who lend their agency to joint action in this way need to think of themselves as free – otherwise it becomes unclear how the resulting action is theirs.15 But, in the absence of any real contract of association, political argument addresses different ways of presenting action as an object of free agreement. But there is no definitive resolution to this process, beyond the decisionistic elements that issue in action. Politics obtrudes when this decisionistic element supersedes the possibility of free accord, leaving behind other proposed or possible ways of framing agreement.16 Since freedom in the abstract is merely undetermination – the fact that an agent is not forced to a course of action by an external cause – there are always residues left by the winning argument. They can be made consistent with the abstract idea. 15 16

This underlies, for example, Hobbes’s painstaking account, in Leviathan Part II, of how individual wills combine to form a collective agent. This requires only that politics incorporate decisionistic devices, such as the parliamentary guillotine, not a more thoroughgoing Schmittian picture of, say, sovereignty. See Carl Schmitt, Political Theology: four chapters on the concept of sovereignty, tr. and ed. George Schwab (University of Chicago Press 2005).

190

Toleration, free speech and the right to lie

This suggests a reason why advocates and opponents of censorship can agree that speech is important: it is the communicative medium through which individuals negotiate the terms of their association. As free individuals, they have to see themselves as lending their agency to proposed courses of joint action, and this demands negotiation that can only be effectively carried on by verbal means. In this respect, politics is discursively constituted. Attempts to limit the content of speech seek to foreclose politics itself. The same may go for attempts to ration opportunities for speech, by excluding certain persons or groups from participation. Both, however, are required in more formal contexts where political negotiation aims at a decision, even while the attempt to limit content and opportunity attracts political objections: for instance, ruling speech out on grounds of irrelevance. As this suggests, speech has a further significance. Although political engagement seldom involves explicit negotiation over terms of association, the background as with any joint concern is one of tacit negotiation on the terms on which the participants engage. Much association is informal. It arises in an ad hoc way that involves a continuing renegotiation of its terms, sometimes explicitly, but more often at the level of implicature. This applies to everyday talk, as well as in commercial or institutional settings. A theme of discursive encounters, overtly or otherwise, is whether to continue with the very interaction that this question in part serves to constitute. Since the interaction is already there, negotiating its own terms must go on in medias res. Communication is a form of association in its own right. As I have already noted, freedom of association demands negotiation over its terms. But the negotiation itself clearly requires speech, or something that generates similar semantic and illocutionary content. This means that one form of association must exist prior to negotiation, namely speech itself, understood as a precondition for negotiation; there can be no significant negotiation without speech.17 Thus there are two respects in which speech can be seen as special. First, as a medium of communication, it is the precondition of negotiation and, through it, of association. Second, speech is itself a form of association and as such exemplifies the possibility of an association that exists without, or at least prior to, negotiation. These features help to explain the double-sided feature of speech discussed earlier. Free speech defenders can point to the fact that speech contributes to association as something yet to be settled, while censors essentially take association 17

In some cases interaction can occur in the absence of speech. See Michael Bratman, Faces of Intention (Cambridge University Press 1999).

Free speech and the right to lie

191

and its framing terms as given. In this way, despite their agreement about the importance of speech as communication, proponents and opponents of free speech take contrasting views of the role of speech in political association. But advocates of free speech and censorship alike see the centrality of speech to political inclusion and exclusion. The centrality of communicative speech to human interaction in turn suggests two conclusions about political association. First, thinking of the participants as engaging on contractual terms – that is, as parties to a prospective contract – though hard to avoid, may mislead, by overreading the content of political interaction. Much communicative speech is not concerned with agreeing a contract. Though communication must involve negotiation in the sense that it is an interaction aimed at least at understanding, it need not aim at agreement. Second, the terms of political association cannot themselves be thought of as literally contractual, because association already exists in the fact of negotiating over it. In this sense the facticity of political association precedes attempts to give it a normative grounding. Free speech and the right to lie The strong emphasis on speech as a conduit for association creates a particular problem with deception. As in contract, misrepresentation dissolves agreement. When individuals freely agree to combine their agency, they have to know the terms to which they are agreeing: otherwise they will not freely combine their agency with others in a way that constitutes a joint project. This seems to rule out lying and other intentional deception. A person cannot think of himself as free in relation to his actions if his assent to them rested on false or deficient information. So association, the merging of individuals in joint agency, requires non-deception. But on the view sketched above, the public sphere itself is not an association – or at least, it is not an association that results from terms agreed by prior negotiation. So, it may be said, deception does not violate one of its conditions, as it does with other forms of association. Nonetheless, deception seems peculiarly problematic, precisely because it has not been agreed to and makes agreement on terms of association invalid. This is not to say that agreed deception is impossible. On the contrary, some associative relations may call for just that, such as membership of a poker school. There is no contradiction in agreeing that one will enter a situation in which one will be deceived from time to time. Hence deception need not undermine joint agency.18 18

On this point, see my ‘Political Lying: a defense’.

192

Toleration, free speech and the right to lie

On one view the right to free speech amounts to a free-for-all; there should be no restrictions on speech, i.e. on what one can say. This gives rise to a puzzle that is inherent in defences of free speech. Suppose there is a right to free speech that requires limits neither on content nor on the speech-acts that may be performed by means of that content. However, lying is a form of speech-act; so, does the right to free speech entail a right to lie, even though most people would say that lying is, in general, wrong? As it stands, it may seem that the puzzle results simply from the collision of two intuitions, one in favour of free speech and the other against deception: the puzzle can be framed in terms of the theory of free speech already sketched. The rationale for free speech lies in freedom of association, but association fails in the presence of deception. Let us assume that lying is, in general, wrong, at least in the sense that lying is always wrong either pro tanto or prima facie. Nonetheless, lying may be regarded as the exercise of a right to free speech. This brings into the spotlight the more general issue mentioned earlier: can there be a right to do what is wrong? That is what toleration might be thought to be, after all, though on some accounts toleration depends on a discretionary right of the tolerator, rather than a claim-right wielded by those tolerated. Leaving this aside, is this as paradoxical as it sounds – as, often, toleration itself has been thought to be?19 What is the nature of the right to free speech, assuming it exists? If it is only a legal right, there is no problem in doing what is morally wrong. There can, unproblematically, be a legal right to do what is morally wrong, such as committing adultery, showing ingratitude, etc., except on strong versions of the doctrine that lex iniusta non est lex.20 The legal right to do what is morally wrong does however pose a problem for philosophers who try to rest legal rights on prior moral rights.21 Then either the legal right to free speech lacks a moral basis even though legal rights rest on moral rights; or there can be a moral right to do what is morally wrong, and this faces problems of coherence, though this is not to say that the incoherence cannot be dispelled.22 Whatever may be the case in general, on the theory offered above, the legal right to free speech does not have its basis in some foundational 19 20

21 22

See also Chapter 9. Usually attributed to Saint Augustine: see De libero arbitrio Bk. I, ch. 5, §11 xxxiii: Nam lex mihi esse non uidetur, quae iusta non fuerit; cf. Saint Thomas Aquinas, Summa Theologica I-II, qu. 96, art. 4. E.g. Ronald Dworkin, Taking Rights Seriously (London: Duckworth 1977), and A Matter of Principle (Oxford University Press 1986). One obvious method would be to argue for a moral right to autonomy, whose exercise might be held to include the doing of acts which were morally wrong in themselves.

Free speech and the right to lie

193

moral principle such as treating others as ends in themselves rather than as mere means. Instead, its basis lies in the absence of prior conditions on the terms of association between free individuals, and in the minimal conditions that need to be met for individuals to incorporate themselves in a form of collective agency. The main condition is that individuals must combine their agency freely, in order to act collectively. Otherwise they do not own the collective’s actions in a way that sustains the usual ascriptions of acts to agents, responsibility, and intention. That seems to preclude lying in negotiating the terms of association. This raises the question of the theory’s moral commitments. Is upholding freedom of association not itself a moral principle, or at least the insistence that individuals must be free in order to join in collective action? I cannot do full justice to this question here, but the gist of the answer is that collective action is not escapable for us: it is simply there. So the only question is how to regard oneself in relation to it. Insofar as I think of myself as free in relation to it, I accept it. But what is important for that is that I think of what is simply there as something that admits of the possibility of free acceptance. The commitment to freedom lies in accepting the possibility of accepting what is given as though it were an object of free choice. Hence freedom of association emerges as necessary to understanding the given as acceptable. But it is not directly as a moral argument, one that proceeds from the assumption that freedom of association has foundational moral status. It is rather a transcendental argument for the inescapability of freedom in relation to a primordial, pre-moral, attitude of acceptance. On the account just sketched, it is not that free speech follows from some wider principle of toleration, but that free speech is required as a condition of politics qua joint action.23 Politics, as I argued earlier, attempts to answer the question What do we do? The underlying idea is that no prior limits can be set to the terms on which people engage with each other – that is, on what content can go into negotiations about joint action. Since the theory aims to keep open in this way the basis for free association, some modalities of speech matter more than others at this level. It is obvious that the content of speech gets protected if free association means that no prior limits are set to the basis on which people negotiate over its terms. It is equally clear, however, that concrete examples of political talk, for example in formal settings such as parliamentary or congressional debate, will impose limits on grounds of relevance to what can be said. Here again Meiklejohn’s dictum that

23

See also Chapter 2.

194

Toleration, free speech and the right to lie

everything worth saying be said applies – but even that is subject to appeal against control of the agenda. This open-content position would exclude, for instance, legal prohibitions on Holocaust denial. It is not required that some plausible-sounding narrative be furnished to illustrate how Holocaust denial could form an article of association between people, or could enter into their negotiations over association. It would need to be shown how this was not possible – how no association could include discussion of the historicity of the Holocaust in its terms. However, some other speech modalities do not demand such stringent protection. By contrast with the content of speech, its expression does not gain support from reflecting, as above, on the necessary conditions of association. I will give as a further illustration the implications of the idea of free speech set out here for the protection or otherwise of pornography. In the US, in particular, pornography and its producers have benefited from First Amendment protection for speech, and philosophers have endorsed their right to do so.24 The right thus invoked bears as much on consumers who want pornography to be available to them as on producers who wish to provide it. To this extent it mirrors other kinds of speech protection, in that its proponents rely on a right to hear, as it were, as well as on a right to be heard. One might doubt whether rights should be asserted here, even if pornography is morally unobjectionable. There are some claims that, while morally unproblematic in themselves, do not form the basis of a right. There is nothing exceptionable about driving on the left, but the law may require road users to drive on the right. Some matters, just because they are morally indifferent, can be legislated for without violation of right. There is the further point that some moral claims resist formulation in terms of rights.25 The main point, however, is that the above account of free speech gives little support for the view that pornography should be made freely available as a matter of moral entitlement. First, it is not clear that visual pornography – overwhelmingly the most important part of the industry – qualifies as speech and, if not, the earlier arguments simply go past it. Producing pictorial material is no doubt expressive in some sense, but the very broad range of acts to which expressive content could be imputed include many that liberal and other societies will need to suppress. Suppose, however, that visual pornography does qualify as 24 25

E.g. Ronald Dworkin, ‘Is There a Right to Pornography?’ Oxford Journal of Legal Studies 1:2 (1981), 177–212. See Joseph Raz, ‘Can There be a Right-Based Moral Theory?’ in Jeremy Waldron (ed.), Theories of Rights (Oxford University Press 1984).

Free speech and the right to lie

195

speech. Then the question is whether pornography benefits from the communicative rationale for free speech.26 I suggest that it does not, for three reasons. First, most visual pornography is extremely repetitious. Its communicative content devolves on stereotypical scenarios which are repeated at length – indeed, in much pornography the repetition is part of the point. Repetitiousness need not warrant censorship, of course. But the point of protecting content is to ensure that no limits are set to the terms on which people can negotiate over their association. This need not extend to the indefinite replication of content. Indeed, it remains an open question whether content-protection extends to the purchase of TV advertising time to air particular messages, such as partisan commercials during elections – not merely because this will promote some content at the expense of others, but also because purchasing ad spots commodifies, and therefore indefinitely replicates, certain messages. Second, its commodification – the fact that it is purveyed and experienced as a straightforward commercial transaction between producer and consumer – moves pornography into a different category from non-commercial speech. This follows from the rationale that also supports curbing misrepresentation and other forms of manipulation (such as subliminal advertising) in commercial speech. This does not apply to pornography created, as is sometimes the case, for non-commercial purposes. But there is little reason to conflate with this, as courts have sometimes done, the production of pornography as a pure market transaction. I pursue the question of how to demarcate commercial from non-commercial speech below. Third, insofar as pornography encapsulates speech that might be thought to benefit from the rationale for its protection given earlier, it is to be seen both as embodying and as prefatory to association. Much pornography trades on purportedly offering sex to the consumer: the most widespread version of the stock scenarios already referred to presents women as sexually available to men. In other words, the content of the image standardly involves an invitation to a certain form of association, namely sex. This is of course literally false. Its function is as a basis for fantasy, typically as an aid to masturbation. Since the imposture is transparent, it cannot be seen as having deceptive intent, except to the extent that it reinforces the idea that women in general, or whoever is represented in it, are sexually available to consumers of pornography. Since the content involves only fantasised association, it 26

Cf. Ashcroft v. Free Speech Coalition, 535 US 234 (2002), where relaxations to laws against child pornography were sought on free speech grounds.

196

Toleration, free speech and the right to lie

does not fall under the rationale for protecting speech given by freedom of association. None of this shows that it must be prohibited, but it suggests that the above justification for freedom of speech does not extend to these forms of pornography.

The right to lie: Nike and Kasky So far I have argued that there is no free-speech-based right to lie in the public sphere because free speech is based on freedom of association, whereas lying about the terms of association precludes joining agency with others on mutually agreed terms. The dupe’s assent rests on a false prospectus. The case of Kasky v. Nike27 arose when a California-based political activist, Marc Kasky, sued the Nike corporation under the state’s Unfair Competition and False Advertising statutes.28 Kasky alleged that Nike had lied about its employment practices in China, Indonesia and Vietnam, particularly its alleged use of child labour in ‘sweatshop’ conditions. He further argued that Nike’s wilful misrepresentation of its labour conditions secured it an unfair commercial advantage, for example by misrepresenting facts that would have alienated ‘ethical’ consumers. Andrew Young, the former United States ambassador to the United Nations, visited twelve of Nike’s factories in China, Vietnam and Indonesia at the company’s request and gave them a largely clean bill of health.29 But a confidential audit of a Nike supplier factory in Vietnam by the accountants Ernst & Young in January 1997 alleged that a number of serious malpractices marked Nike’s operations in the region.30 Among the most serious of Ernst & Young’s findings were that Nike workers were exposed in the factory, the Tae Kwang Vina Industrial Ltd. Co., in Bien Hoa City, southern Vietnam, to toxic chemicals without protection, were forced to work excess overtime, and endured poor ventilation and lack of drinking water.31 The Ernst & Young report highlighted the lack of ventilation in the factory it visited, and the consequent respiratory problems suffered by the workforce, most of whom were young women. In addition, 27 28 29 30

31

Kasky v. Nike Inc., 27 Cal. 4th 939. California Business and Professions Code § 17200 et seq.; § 17500 et seq. Young’s report is online at www.calbaptist.edu/dskubik/young.htm#introduction. Nike did not in general own the factories that manufactured its products. As Young points out, the plants were often run not by locals but by a management staff foreign to the site of manufacture, such as Taiwanese in one of Nike’s large Vietnamese works. For the Ernst & Young report, see www.corpwatch.org/article.php?id=2488. Ernst & Young report.

The right to lie: Nike and Kasky

197

Ernst & Young found that one factory shift was made to work overtime against its will; that reduced training and probationary pay rates were paid to workers beyond the time of actual training; that a particular worker was consistently paid less than the minimum wage over an extended period of time; and that religious holidays and ‘days off ’ were regularly denied to certain workers. However, the Ernst & Young report found no basis for the most frequent allegation made against Nike’s working practices in Asia, namely its alleged use of child labour.32 During this period, in response to the Ernst & Young report but also others alleging malpractice in its Asian workplaces, Nike had issued press releases and written to newspapers and university authorities disputing the reports about its labour practices. Marc Kasky then sued Nike. In its defence, counsel for Nike did not deny that the firm had lied about these practices. Instead Nike’s counsel argued that the firm had a right under the First Amendment knowingly to make false statements. The nub of this contention was that in engaging with issues of labour practices in the developing world, Nike was being drawn into a debate that did not involve commercial speech. That speech was subject to advertising laws with their restrictions on commercially motivated misrepresentation. It instead involved speech in the public sphere, to which full First Amendment rights should therefore apply. For his part, Kasky argued that Nike’s misrepresentation of its labour conditions secured it an unfair commercial advantage among ‘ethical’ consumers. The California circuit court found for Nike, on First Amendment Grounds.33 The California Court of Appeals followed in finding for Nike and against the plaintiff,34 judging that Nike’s communications ‘form[ed] part of a public dialogue on a matter of public concern within the core area of expression protected by the First Amendment’.35 But that verdict was overturned on appeal to the California Supreme Court.36 The court found that ‘these messages [i.e. those by the Nike Corporation regarding conditions in its Asian factories] are commercial speech for purposes of applying state laws barring false and misleading commercial messages’. The majority held that ‘commercial speech that is false or misleading is not entitled to First Amendment protection’.37 32

33 34 35 36 37

Ernst & Young investigated a sample of fifty workers in the factory using payroll information for October 1996, determining that all were over the age of fifteen (the minimum working age under Vietnamese law). Nike, Inc. v. Kasky, 539 US 654 (2003). Kasky v. Nike, Inc., 79 Cal. App. 4th 165, 93 Cal. Rptr. 2d 854 (Cal. Ct. App. 2000). Kasky v. Nike, Inc., 79 Cal. App. 4th 165, 178 (Cal. Ct. App. 2000). Kasky v. Nike, Inc., 27 Cal. 4th 939, 45 P.3d 243 (Cal. 2002). Kasky v. Nike, Inc., 27 Cal. 4th 939, 951 at 953 (Cal. 2002).

198

Toleration, free speech and the right to lie

Nike appealed to the Federal Supreme Court, securing a writ of certiorari. Subsequently however, the court refused to reach a verdict, on the grounds that the writ had been ‘improvidently’ granted. Three judges dissented.38 So the California Supreme Court’s verdict in Kasky v. Nike stands, but is not dispositive federally. The Southern California branch of the American Civil Liberties Union (ACLU) argued that ‘Governmental regulation of false or misleading commercial speech does not chill constitutionally protected speech’.39 Nonetheless, in its deposition to the California Supreme Court, the Northern California branch of the ACLU based its support for Nike squarely on the First Amendment, and on urging that Nike’s interventions in the debate over its Asian factories should not be seen as commercial speech: ‘the Court should not allow First Amendment rights to be diluted through an expanded definition of commercial speech’. The Northern California ACLU continued: ‘The rigid distinction between commercial and non-commercial speech . . . distorts that debate by holding one side of the dispute more accountable than the other for the accuracy of its statements.’40 The ACLU brief thus argued for a wide-scope free speech principle, indifferently applicable both to Kasky and to Nike: to grant Kasky and other critics free rein in pillorying Nike for its working practices while denying the corporation similar scope in response was a breach of equity. These arguments were dismissed by Justice Kennard, for the majority. The majority verdict alluded to California’s unfair competition statutes under which Kasky had first filed the case. ‘To the extent that application of these laws may make Nike more cautious, and cause it to make greater efforts to verify the truth of its statements, these laws will serve the purpose of commercial speech protection by “insuring that the stream of commercial information flows cleanly as well as freely”.’41 Meanwhile, Justice Chin, for the minority, reiterated the ACLU’s point that an equitable field of play in debating the accusations demanded equally extensive speech rights on each side. ‘Nike’s overseas labor practices were undoubtedly a matter of public concern, and its speech on this issue was therefore entitled to special protection . . . Nike could hardly 38

39 40

41

Justices Sandra Day O’Connor and Stephen Breyer both dissented from the majority view that certiorari had been wrongly granted, but they also recorded opinions in support of Nike. www.reclaimdemocracy.org/nike/commercial_speech_socalaclu.html. ACLU Foundation of Northern California, amicus curiae deposition to US Supreme Court (24 February 2003), online at http://supreme.lp.findlaw.com/supreme_court/ briefs/02–575/02–575.mer.ami.aclu.pdf, 4. www.reclaimdemocracy.org/personhood/kasky_nike_decision.pdf, 12–13.

The right to lie: Nike and Kasky

199

engage in a general discussion on overseas labor exploitation and economic globalization without discussing its own labor practices.’42 So the question in Kasky was whether Nike’s claims about labour conditions in its Asian factories counted as commercial speech and therefore fell within the scope of laws prohibiting commercially motivated deception. The California Supreme Court agreed with Kasky that it was, although the inferior courts had sided with Nike. The dispute was not on the principle of acknowledging that the public had an interest in regulating commercial speech: it was agreed on all sides that it did, and nobody in the case disputed the statutory validity of California’s laws against unfair competition and commercial misrepresentation that Kasky had cited in bringing the action.43 As far as this goes, the legal issue appears clear-cut.44 As so often, however, what is at first sight a dispute on a point of law turns instead on meta-legal considerations. The question rather was what the due scope of commercial speech should be, and the complementary space of protected public speech. At this level, the point at issue was whether a variation in context was sufficient to move Nike’s speech from the legally more closely circumscribed terrain of pure commerce to the open space of public debate; and whether considerations of equity required that any asymmetry between the speech opportunities enjoyed by plaintiff and defendants be corrected. It was suggested in some quarters that Nike as a corporate person should not be eligible to enjoy First Amendment rights in the manner that Marc Kasky, as an individual US citizen, clearly was. But it is hard to see why the mere fact of being a corporate person should deprive Nike of a platform, any more than it should non-commercial bodies such as NGOs; and the opportunity restriction seems unwarranted. It might be said that the fact that Nike is already incorporated means that the free speech argument in the text fails to apply, since this is premised on the fact that the public sphere is a pre-association – that is, interaction not mediated by prior agreement. The Nike Corporation is certainly an association of a kind. But what matters in the legal case is the fact that as a commercial enterprise it was in the business of proposing new associations to prospective consumers. Thus the mere fact of being a 42 43

44

www.reclaimdemocracy.org/personhood/kasky_nike_decision.pdf, 10. Though it should be added that Thomas Goldstein, acting for Nike in the case, contends that the statutes themselves contravened California’s constitution. Goldstein, ‘Nike v. Kasky and the Definition of “Commercial Speech”’, Cato Supreme Court Review (2002–3), 63–79, at 66, fn. 11. For a discussion which views the issues in these terms, see Goldstein, ‘Nike v. Kasky and the Definition of “Commercial Speech”’. Goldstein advised Nike in the case.

200

Toleration, free speech and the right to lie

corporate body should not deprive Nike of a platform, any more than it should non-commercial bodies such as NGOs; and the opportunity restriction seems unwarranted. The theory restated The arguments on either side focused on aspects of speech. Kasky’s counsel emphasised the content communicated to the audience (what?); Nike’s the opportunity to speak freely enjoyed by commercial corporations (who?). Nobody made much play of the way (how?) in which Nike had expressed its claims. At the same time, it was of crucial significance that the talk allegedly involved deception. But there was an underlying bone of contention about the context (where?) in which Nike’s allegedly deceptive utterances had been made. Kasky’s case relied on a crosscontextual claim: that standards applicable to Nike in one context (advertising) should apply also in another (the wider public sphere). Or rather, the fact that Nike was making commercially sensitive claims about its working practices converted the context of the speech into a commercial one. The California Supreme Court’s case for non-protection relied less on the content of what was said than the context in which it was said. This made it important to determine whether the talk could be assigned to the public sphere. Is this right? If it is, firms like Nike would be in effect excluded from public debate on any matter that could be construed as impinging on their commercial interests. The defence of free speech outlined earlier holds that there are no restrictions, prior to association, on speech as a vehicle of communication. This means that there are no limits on the content that can be put into the public sphere. Nor can there be restrictions on the opportunities afforded individuals to combine with others on mutually agreed terms. Free speech is association that exists prior to negotiation, and so is minimally structured by jointly agreed conditions. As we saw, this theory of free speech gives lying a particularly problematic status. There is a clear free-speech case for permitting lying and other deceptive talk, since these undeniably count as speech. If free speech means that there should be no restrictions on speech whatsoever, then there should be no restrictions on mendacious speech. The basis for this argument can be derived from the theory. That says that there can be no limits set on the capacity of free individuals to associate on mutually agreed terms. Any communicative act may in principle play its part in negotiating over association, so no restrictions can be placed on speech, including what is communicated, in what way, by whom, and in what context.

The theory restated

201

So far the theory seems to yield an open-and-shut case in support of Nike’s position. But matters are not so simple, as becomes clear both when considering the theory in the abstract and in relation specifically to the Kasky case. Speech in the public realm is an instance of association. But speech has the peculiarity that it cannot be subject to negotiation, since negotiation itself requires speech. So no prior terms can be set to the association that is constituted by communication in the public sphere. But, as we saw, association is invalidated by deception, since this prevents the parties from giving a cohesive account of their joint agency. Speech, as a form of association, clearly cannot preclude lying, since mendacity is verbal. If the rationale for free speech is that no prior limits can be set to speech beyond the generic conditions for communication, then no embargo on lying can arise from these generic conditions. This is as much as to say that a wide-scope justification for free speech will extend to support for a right to lie. At the same time, lying and other forms of verbal deception negate any association to which they give rise.45 This ambivalence is reflected in the way in which the theory responds to Kasky, and captures the key contentions made on either side. On the one side, Nike contended that its mailshots about its South East Asian factories were intended to contribute to a debate in the public sphere regarding its labour practices. By trying to place its speech on the public plane, Nike hoped to benefit from a wide-scope principle of free speech, as enshrined in the First Amendment, which places no restriction on deceptive speech, including lies. On the other hand, Kasky was concerned to show that Nike’s speech belonged not within the generic public sphere, but in commercial speech that proposes a transaction between purveyors of goods or services, and consumers.46 As noted above, the theory distinguishes between the free exchange of speech in the generic public sphere, and the much more closely regimented speech of most discursive encounters. This apparent contradiction, between a strong free speech principle and the routine restriction of speech, is resolved by reference, again, to the notion of association: discursive curbs are justified by the fact that the parties have freely agreed to these curbs as part of the norms of association. The restriction on lying as a part of 45

46

This is one way of understanding the Kantian embargo on lying: the philanthropic motive for lying still fails to justify it, since it cannot form a basis for common purpose between the liar and the murderer. On this point see Christine Korsgaard, ‘The Right to Lie: Kant on dealing with evil’, Philosophy and Public Affairs 15:4 (1986), 325–49. Kasky’s demand was that Nike disgorge all profits from sales of its sports apparel in California in violation of the statutes against false advertising and unfair competition. See Goldstein, ‘Nike v. Kasky and the Definition of “Commercial Speech”’, 66.

202

Toleration, free speech and the right to lie

the regulation of commercial speech then follows as a speech-limiting norm of association – the association that arises between a producer and consumer who enter into a commercial transaction. This shows that the theory can make sense of either side’s position. But does the analysis have to rest there? Must it conclude merely that lying was justifiable if Nike’s speech occurred within the public sphere, but not if it was commercial discourse aimed at inducing consumers to buy Nike products (or at dispelling their moral scruples about doing so)? Consider again the notion of lying. To repeat an earlier point: if free speech means that no holds are barred, then lying cannot be barred. On the other hand, fraudulent commercial transactions – such as slimming pills or baldness remedies sold on false claims about their efficacy – provide as clear an example as anything of how lying can invalidate association. In other transactions and indeed other examples of civil association, the parties can provide a cohesive account of what they are doing, which provides a shared basis for agreement.47 But this is just what is lacking in the case of commercial misrepresentation. The level at which lying cannot be proscribed is that of what might be called pre-association, the generic public sphere. Insofar as constitutional principles on the lines of the First Amendment regulate this sphere, they cannot preclude the possibility of lying on the transcendental ground of pre-association, namely that there can be no prior limits, set by appealing to norms of communication, to the terms on which free individuals choose to associate. Despite this, deception clearly disrupts association, as we have seen. It means that the proposition to which agreement is being sought differs from that on which the proposer aims to act. The theory underlying this discussion suggests that the conditional shared premise on which both Nike and Kasky relied was correct: if Nike’s claims about its South East Asian factories can be assigned to the public sphere, then they enjoy protection. According to the theory, the public sphere as a pre-association lacks the background of negotiated agreement needed for lying to have been ruled out as a possible component of joint agency. Lying is indeed a form of communication, so cannot be excluded purely by reference to generic conditions on communicative acts. To this 47

It is notable that the 1998 Belfast Agreement between Unionist and Nationalist/ Republican politicians in Northern Ireland allowed a range of interpretations, leaving open how far all the parties were agreeing. See e.g. National Committee on American Foreign Policy, ‘Beyond Textual Ambiguity: the need to clarify the Belfast Agreement’, American Foreign Policy Interests: the journal of the National Committee on American Foreign Policy 23:4 (2001), 253–5. As Arthur Aughey notes, a key feature of the Agreement was ‘constructive ambiguity’: see his The Politics of Northern Ireland: beyond the Belfast Agreement (London: Routledge 2005), 148.

The theory restated

203

extent there is indeed a ‘right to lie’ in the public sphere, and it becomes clear both why Nike wished to locate its discourse there, and why Kasky wanted to assign it to the commercial realm, where non-deception is an explicit part of producer–purchaser agreements. In commercial transactions, lying is ruled out on mis-sale grounds: consumers are entitled to know what they are paying for. But this still seems unsatisfactory. I suggest that the source of the dissatisfaction lies in the ambivalent place of lying in the public sphere. Since it is also true that deception disrupts joint agency, there is no shared prospectus on which joint agency is based; this seems to undermine the public sphere’s standing as a forum from which subsequent associations are created. Surely, it may be said, even if the deception is annexed to the public sphere, this fails to dispel the notion that deceptive talk is deeply problematic, even if Nike’s claims are not classed as commercial discourse stricto sensu. A fairly obvious recourse lies open at this point. This is to suggest that while lying and other deception may not be legally actionable, they also cannot be protected by means of a constitutional right. Thus a plaintiff could not bring to court an action citing the First Amendment against a defendant who had prevented her from lying (for example, by denying her a platform from which to do so), or if the plaintiff asserted her First Amendment speech rights in bringing such an action. The demonstration that she had deceived, or intended to do so, by means of the speech would be a valid defence at law. In this case the public sphere ‘right to lie’ would be a liberty right, that an agent enjoys by dint of having no countervailing obligation not to lie. As such, however, it would also not entail any obligation on others to permit the speech, such as by making publicly accessible premises available for this purpose.48 Of course, this would not mean that the audience themselves possessed a legal right on which they could act against a party that had engaged in deceptive talk. So in the Kasky case, the fact that Nike lacked a claim right to make deceptive statements – that is, a right that holds in virtue of obligations on others to enable Nike to exercise the right – would not mean that consumers, or activists like Kasky, enjoyed a right that they could assert against Nike if the latter engaged in deceptive talk. The field of rights would, to that extent, lie open on all sides. That may make the final outcome look like a no-score draw between Kasky and Nike, or between those who favour, respectively, a more 48

In Pruneyard Shopping Center v. Robins, also heard in California (447 US 74 (1980)), a group of high school students sought the right to solicit signatures for a petition against the anti-Zionist UN Resolution 3379 in Pruneyard, a privately owned shopping mall.

204

Toleration, free speech and the right to lie

restrictive and a more inclusive view of corporations’ First Amendment entitlements. But the result is not quite as bland as that. This discussion has indeed indicated that there is some merit in Nike’s claims to be engaging in public-sphere discourse, rather than something akin to a sales pitch. Whether or not Nike’s media offensive should be seen as having been carried out purely for commercial gain, the putting of false information into the public sphere could be treated as a strict-liability matter if a complainant can argue that she has made a purchasing decision only because false information relevant to it has been put there by relevant commercial interests.49 It would then be open to complainants to file a class action on the basis that their purchasing decisions had been suborned by misinformation put out by those interests. The general underlying principle here is that free agents who exercise their legal rights are not thereby secured against all avenues of legal redress sought by those to whom their conduct may cause grievance. This can be seen in the case of speech itself, which despite being legally protected, can be exercised in such a way as to be liable to action under the common law tort of nuisance, over and above specific criminal offences such as incitement that may be committed by means of speech.50 In other words, liability to civil recovery on proof of tortious conduct can remain even if the behaviour concerned is a token of a type that is legally protected. So, in the case of a corporation such as Nike that may be exercising its First Amendment right to free speech in order to make false statements about its commercial activities, it should lie open to buyers of its merchandise to bring a civil action under legislation governing contractual misrepresentation.51 This would not prevent corporations from exercising their rights to free speech. They could comment freely on matters not relating to their commercial activities. They could also comment freely on those activities – but in so doing they would be as liable as are other speakers in exercising their speech right, to legal redress.

49

50

51

It would not be necessary to establish that the party knowingly put out false information (and therefore, presumably, acted with the intent to deceive). It would be sufficient to establish, first, that the information was false; and, second, that a reasonable person in the corporation’s position would have believed that it was false. There is in addition the criminal offence arising from nuisances to which speech may give rise. In Kovacs v. Cooper, 336 US 77 (1949), the US Supreme Court held that the amplification of speech by the defendant Charles Kovacs in Trenton, New Jersey, violated borough ordinance 430, prohibiting ‘loud and raucous noise’. Kovacs appealed to the federal Supreme Court. It held that the conviction did not violate the First Amendment. In England, for instance, section 2(1) of the Misrepresentation Act 1967 lays down that even non-fraudulent misrepresentation will be liable to recovery unless the statement in question was one the speaker could reasonably have believed true.

Conclusion

205

It would seem that there is an equitable basis for this approach. It is clearly possible that prospective consumers hear the messages that corporations project into the public sphere about their commercial operations and are influenced thereby to buy their products. It is plausible to think that this is part of the intention behind such interventions, as in the Nike campaign in 1996–7 to influence opinion about its suppliers in South East Asia. But even if not, it is open to consumers swayed by such media campaigns to argue that their purchasing decisions were taken only because the campaigns were such that a reasonable person would have been led to believe that health, safety, remuneration, unionisation or other aspects of the workforce’s terms and conditions were of a certain kind. Again, it seems reasonable to treat this as a strict liability tort rather than requiring that deceptive intent be established directly. It would also allow for a further quid pro quo, which is fully consistent with the theory of speech on which this chapter has drawn. While corporations would enjoy a liberty right to engage in misrepresentation, laws governing corporate libel should be as weak as they need to be in order to permit a full public debate over such matters as labour practices. The corollary of a public sphere with no bars on content is that powerful – wealthy – parties should not be able to use their financial muscle to gag those whose views they dislike.52 The obverse of requiring that corporations are not silenced in the face of critics enjoying wide speech rights, is that they should not be able to silence those same critics by the threat of costly, and often unaffordable, litigation. Lying is, in the sense used throughout this book, tolerable: it is not protected as a matter of legal right, but occupies a more legally negotiable sphere whereby it benefits vicariously from interests that there are jurisprudentially good reasons to protect.

Conclusion I have argued that free speech, including toleration of speech one regards as wrong, follows from the principle of freedom of association. Lying raises a special problem because it destroys joint agency – that is, association on agreed terms. This explains why lying in the public sphere is objectionable: it makes it impossible to associate on such terms. Association as embodied speech comes before any agreement on its own terms. That is 52

This is not to deny that the statute books should contain an offence of corporate defamation. But this should be subject to two conditions: the burden of proof should lie with the plaintiff; and a wide-scope ‘fair comment’ defence should be available in personal defamation actions.

206

Toleration, free speech and the right to lie

why toleration is, in general, subsequent and not prior to discursive principles such as freedom of speech. At the same time, since speech as association must precede agreement on its own terms, the stipulation not to lie cannot be there from the start as an article of association: attempts at this point to construct a transcendental argument simply fall to existential disproof. The reason one should tolerate saying what is wrong, bad speech, is that free speech is a precondition of association, joint agency, in general. Toleration qua the ‘right to do what is wrong’ follows from a condition of there being joint agency at all – that is, a precondition of there being politics. It then becomes clear that the form of bad speech which is lying poses a particular problem of toleration. As ‘bad’ speech it seems eligible for toleration, but to tolerate it risks undermining the public sphere itself. Nike’s misstatements can thus be seen as a toleration issue – as a liberty-right to do, or in this case say, what is wrong. If its claim to be engaging in public rather than commercial speech was credible, Nike could benefit from a wide-scope free speech principle. That principle rests, as I have argued, on freedom of association. That what is involved may be deceptive speech, such as lying, shows that the issue is one of toleration. I have tried to explain how that can be so – despite wellfounded reasons for finding such speech objectionable – given the undergirding theory of free speech. Here, as elsewhere, the remedy for the aggrieved in a tolerant society may be not to call upon constitutional law but to look to civil redress, with the uncertainties that legal processes, like politics, unavoidably face.

Epilogue

This book has explored a range of questions about toleration and its role in political and philosophical argument. It has not tried to put forward a theory of toleration. Nonetheless I have aimed to uphold a thesis about toleration, namely that though awkward, both in theory and practice, it is also politically unavoidable. Political philosophers for whom theoretical coherence is, if not everything, a condition of anything, are especially apt to be drawn by accounts that dispel seeming contradictions by either reconfiguring or abandoning toleration. It is not an accident, in modern democratic conditions, that toleration should be at once troublesome and hard to get away from. It is not very helpful to ask whether toleration’s political presence is explained by its awkwardness, or the converse. To understand why not, however, is also to understand why the two features tend to go together. On a functional view of politics, the basic political question is What do we do? That selects for disagreement not because of some supposedly deep truth about ‘agonism’ but, more prosaically, because of deliberative indeterminacy: it is seldom simply obvious what we should do, and indeed sometimes not obvious who ‘we’ are. Conflicts over toleration pose these questions in especially sharp form – they force reflection about what a polity is, about its ends, and what it can encompass. The awkwardness of toleration becomes manifest in the fact of being politicised. A theme that marks much of the preceding argument is the political situatedness of reasoning about toleration. To suggest that the air of contradiction that envelops arguments over toleration may be dispelled by a conceptual regimen risks legislating in a vacuum. This is particularly clear when one looks at the democratic culture of toleration. The main problem is how toleration survives when democracy’s institutions and often its ethos prove inimical to it. Again, theorising offers little help if it directs its energies either to a wholesale remoulding of toleration, or to laying down justificatory conditions that ideally rational citizens would meet. Political struggles over toleration enact a basic tension in liberal democracy, between the totalising impulse of the demos and the 207

208

Epilogue

urge to carve out a domain of private omnipotence as one, precisely, where the sovereign’s writ cannot run. As the normative background is unstable in this way, the justifications offered for toleration are likely to prove labile too. Indeed, justifications by theorists as well as political actors often prove surprisingly ad hoc. In modern liberal defences of toleration such as those that aim to heed neutralist side-constraints on justification, the conscientious grounding of toleration that surfaces in modern form in Rawls’s doctrine of the burdens of judgement is again menaced by the possibility of a rival and absolute order of value, since there is no guarantee that the dictates of conscience, secular or religious, will sit happily with liberal political norms; the notion of the reasonable, in the face of reasonable disagreement, is then brought in to head off this potential source of sedition. A parallel job is done by the ideal of public reason as a filter for presented demands on public policy and resources. Summarily put, either public reason is front-loaded to generate the desired outcomes, or a purely procedural notion of reason is insufficient to guarantee liberal outcomes. As such the account relies on an inextricably political justification whose effect is to obscure the politics behind the justification itself. A structurally similar pattern of argument surfaces in Locke’s defence of limited toleration. Overshadowing that defence is the thought that eschatological reasons can endanger the state’s own purposes as a distinct and, despite Locke’s strenuous efforts to keep them separate, rival association to that of the church. How much credence he gives to reasons for toleration such as the doxastic argument depends on the perceived political stakes. Locke in effect de-securitises religious conviction in the name of toleration – except when he does not. Security itself often plays this role in modern theory, as the ne plus ultra that any ruling idiom must respect. But security does not have to do this job. Of course, one can stipulatively define ‘security’ as whatever is regarded as needed for the polity’s continued existence. But even that clearly does not compel acceptance of specific claims about security – for instance, that in some situations it demands the sacrifice of liberties, including those extended by toleration. Often, given the ‘availability cascade’ effect in public thinking about security, the public can make it performatively true that the sacrifice is called for, and what can performatively be made true can also performatively be made false. Political deliberation always needs closure mechanisms, and security arguments transpose them from institutional procedure to the domain of reasons themselves.

Epilogue

209

That does not imply the triumph of a blank Realpolitik, as it calls in question the very notion of a pre-existing order of prevailing forces. At the same time, even within a partially normative account there are limits to idealisation when thinking about toleration and its relation to power. Once again the imagined paradigm is of a political subject deciding how to deal with a bothersome other. The limits in question are not pragmatic or ‘realist’, but those imposed by the ideal itself. Nor is it that idealisation proves, in the abstract, incoherent. It is that the level at which idealisation remains coherent is one that provides politics with a limiting condition rather than with any determinate content. Here, as elsewhere, theory has to reckon with the fact that the raw materials with which toleration works are situational and embodied. The mistaken inference to draw is that that fact itself ‘justifies’ the norms themselves, or the political use to which they are put, with conservative or ‘hegemonic’ results. These consequences do not follow. An awareness that the norms are embodied can stand with a wide range of attitudes towards them. The mistake made in idealisation is that the norms must be held to a standard imposed by a particular form of theorising – in this case, one reliant on a certain idea of freedom that cannot be embodied in the thoughts and deeds of political actors. Toleration is not just about living with a threatening other, but about living with ourselves. That often involves ignoring how the other lurks not without, but irremovably within. Such prolepsis makes certain judgements avowable, and others not: that we tolerate diversity, say, in religion or some forms of sexual practice, but not that we tolerate ourselves as perpetrators of torture, extrajudicial killing, extraordinary rendition, or acts of terror. One version of the proleptic response is to try to break up the political subject when things turn nasty, as in the slogan ‘Not in my name.’ Another is to face the fact that we, the civilised, do these things, even tolerate them. Which side of the line bounding self and other are we then on? What is at stake in keeping faith with the idea that there is such a line, and that we must lie just on one side of it? Toleration’s shape readily fits in with the functional view. It posits a first-person subject of action, the prospective tolerator; and a ‘they’ on which this actor gets to work, with a proposed course of action, such as letting them do their thing. Its awkwardness then arises from the fact that the political setting admits of indefinitely many subjects of action and their acts, even while a determinate idea of each is implicated in any course of political action, such as one of toleration. As I argued in Chapter 10, what lies in the background is freedom as undetermination,

210

Epilogue

by which it is not given ahead of political engagement (for example, in the form of transcendental conditions) how people will elect to pool their agency. Any determinate basis for doing so will, given the open-endedness of deliberation, leave remainders. The persistence of toleration lies in the inescapably political conditions of collective life that theorists still sometimes see as calling for remedy.

Bibliography

Ackerman, Kenneth D., Young J. Edgar: Hoover, the Red Scare, and the assault on civil liberties (New York: Carroll & Graf 2007). Adams, Geoffrey, The Huguenots and Public Opinion 1685–1787 (Waterloo, Ontario: Wilfrid Laurier Press 1991). Addison, Paul, and Jeremy Crang (eds.), Firestorm: the bombing of Dresden (London: Pimlico 2006). Ahmed, A., N. H. Mbibi, D. Dawam and G. D. Kalayi, ‘Complications of Traditional Male Circumcision’, Annals of Tropical Paediatrics 19:1 (1999), 113–17. Anderson, Elizabeth, ‘What Is the Point of Equality?’ Ethics 109:2 (1999), 287–337. Aristotle, Nicomachean Ethics, tr. H. Rackham (Cambridge, MA: Harvard University Press 1934). Arjoman, Saïd Amir, ‘The Constitution of Medina: a sociolegal interpretation of Muhammad’s acts of foundation of the umma’, International Journal of Middle East Studies 41 (2009), 555–75. Ashcraft, Richard, ‘Latitudinarianism and Toleration: historical myth versus political history’, in Zagorin, Ashcraft and Kroll (eds.), Philosophy, Science and Religion in England, 1640–1700. Ashworth, Andrew (ed.), The Criminal Process: an evaluative study, 2nd edn (Oxford University Press 1998). Audisio, Gabriel, ‘La réception de l’édit de Nantes en Provence (1598–1602)’, in Grandjean and Roussel (eds.), Coexister dans l’intolérance. Aughey, Arthur, The Politics of Northern Ireland: beyond the Belfast Agreement (London: Routledge 2005). Austin, J. L., ‘Ifs and Cans’, in Austin, Philosophical Papers, 3rd edn, ed. Joel Urmson and Geoffrey Warnock (Oxford University Press 1979). Baha’i World News Service, 14 August 2009, ‘First Identification Cards Issued to Egyptian Baha’is’, http://news.bahai.org/story/726. Balibar, Étienne, Spinoza and Politics (London: Verso 1998). Balint, Peter, ‘Not Yet Making Sense of Political Toleration’, Res Publica 18:3 (2012), 259–64. ‘Toleration as a Liberal Political Practice’ (unpublished paper). Baron, William, The Dutch Way of Toleration, Most Proper for Our English Dissenters Written at the Request of a Friend (1699) (Ann Arbor: University of Michigan Digital Library, Eebo editions 2010). 211

212

Bibliography

Barry, Brian, Culture and Equality (Cambridge: Polity 2000). Justice as Impartiality (Oxford University Press 1995). Bayle, Pierre, Commentaire philosophique sur ces paroles de Jésus-Christ, ‘Contrains-les d’entrer’, ed. Jean-Michel Gros (Paris: Presses Pocket 1992). Pensées diverses sur la Comète, ed. A. Prat (Paris: facsimile reprint Société des Textes Français Modernes 1912). Beitz, Charles, Political Equality: an essay in democratic theory (Princeton University Press 1989). Benbaji, Hajit, and David Heyd, ‘Forgiveness and Toleration as Supererogatory’, Canadian Journal of Philosophy 31:4 (2001), 567–86. Beneke, Chris, Beyond Toleration: the religious origins of American pluralism (Oxford University Press 2008). Berlin, Isaiah, The Crooked Timber of Humanity, ed. Henry Hardy (Princeton University Press 1998). Besch, Thomas, ‘Diversity and the Limits of Liberal Toleration’, in Ivison (ed.), The Ashgate Research Companion to Multiculturalism. Bess, Michael, Choices Under Fire: moral dimensions of World War II (New York: Knopf 2006). Bird, Colin, The Myth of Liberal Individualism (Cambridge University Press 1999). Bloxham, Donald, ‘Dresden as a War Crime’, in Addison and Crang (eds.), Firestorm. Bollinger, Dan, ‘Lost Boys: an estimate of US circumcision-related infant deaths’, THYMOS: journal of boyhood studies 4:1 (2010), 78–90. Bou-Habib, Paul, ‘Locke, Sincerity and the Rationality of Persecution’, Political Studies 51:4 (2003), 611–26. Bratman, Michael, Faces of Intention (Cambridge University Press 1999). Brison, Susan, ‘The Autonomy Defense of Free Speech’, Ethics 108:2 (1998), 312–39. Brown, Wendy, Regulating Aversion: tolerance in the age of identity and empire (Princeton University Press 2008). ‘Tolerance as/in Civilizational Discourse’, in Williams and Waldron (eds.), Toleration and Its Limits. Burge, Tyler, ‘Individualism and the Mental’, Midwest Studies in Philosophy 4:1 (1979), 73–121. Burke, Roger H. (ed.), Zero Tolerance Policing (Leicester: Perpetuity Press 1998). Buzan, Barry, Ole Waever and Jaap de Wilde, Security: a new framework for analysis (Boulder, CO: Lynne Rienner 1998). Castiglione, Dario, and Catriona McKinnon (eds.), The Culture of Toleration in Diverse Societies: reasonable tolerance (Manchester University Press 2003). Toleration, Neutrality and Democracy (Dordrecht: Kluwer 2003). Centre for Social Cohesion, The, Beth Din: Jewish Law in the UK, 2009; www.socialcohesion.co.uk/files/1236789702_1.pdf. Chambers, Simone, Reasonable Democracy: Jürgen Habermas and the politics of discourse (Cambridge University Press 1996). Christiano, Thomas, The Constitution of Equality: democratic authority and its limits (Oxford University Press 2008).

Bibliography

213

Christman, John (ed.), The Inner Citadel: essays on individual autonomy (New York: Oxford University Press 1989). Clark, Ross, ‘There Is No Place for Sharia Law in British Society’, Daily Express 29 July 2011; www.express.co.uk/posts/view/261698. Cohen, Andrew J., ‘What Toleration Is’, Ethics 115 (2004), 68–95. Cohen, Gerald A., If You’re an Egalitarian, How Come You’re So Rich? (Cambridge, MA: Harvard University Press 2002). Rescuing Justice and Equality (Cambridge, MA: Harvard University Press 2008). Cohen, Joshua, ‘Moral Pluralism and Political Consensus’, in Copp, Hampton and Roemer (eds.), The Idea of Democracy. Cohen, Marshall, Thomas Nagel and Thomas Scanlon (eds.), Equality and Preferential Treatment (Princeton University Press 1977). Copp, David, Jean Hampton and John Roemer (eds.), The Idea of Democracy (Cambridge University Press 1993). Creppell, Ingrid, ‘Toleration, Politics and the Role of Mutuality’, in Williams and Waldron (eds.), Toleration and Its Limits. Daniels, Norman (ed.), Reading Rawls: critical studies of ‘A Theory of Justice’ (Oxford: Blackwell 1975). Davis, Darren W., and Brian D. Silver, ‘Civil Liberties vs. Security: public opinion in the context of the terrorist attacks on America’, American Journal of Political Science 48:1 (2003), 28–46. The Debates in the Several State Conventions On the Adoption of the Federal Constitution, as Recommended by the General Convention at Philadelphia, in 1787, ed. Jonathan Elliot (Washington, DC: J. B. Lippincott & Co. 1866). De Hert, Paul, ‘Balancing Security and Liberty within the European Human Rights Framework’, Utrecht Law Review 1:1 (2005), 68–96. Devlin, Patrick, The Enforcement of Morals (Oxford University Press 1965). Dietz, Mary (ed.), Thomas Hobbes and Political Theory (Lawrence: University Press of Kansas 1990). Dover, Kenneth, Greek Homosexuality (Cambridge, MA: Harvard University Press 1978). Dworkin, Ronald, ‘Defunis v. Sweatt’, in Cohen, Nagel and Scanlon (eds.), Equality and Preferential Treatment. ‘Is There a Right to Pornography?’ Oxford Journal of Legal Studies 1:2 (1981), 177–212. A Matter of Principle (Oxford University Press 1986). ‘The Original Position’, in Daniels (ed.), Reading Rawls. Sovereign Virtue (Cambridge, MA: Harvard University Press 2000). Taking Rights Seriously (London: Duckworth 1977). Edwards, David, and Susan Mendus (eds.), On Toleration (Oxford: Clarendon Press 1987). Enemark, Christian (ed.), Ethics of War in a Time of Terror (Canberra: Australian National University Press 2006). Fehleison, Jill, Boundaries of Faith: Catholics and Protestants in the Diocese of Geneva (Kirksville, MO: Truman State University Press 2010).

214

Bibliography

Feigl, Herbert, and Michael Scriven (eds.), Minnesota Studies in the Philosophy of Science, Vol. I (Minneapolis: University of Minnesota Press 1956). Feuer, Lewis S., Spinoza and the Rise of Liberalism, 2nd edn (New Brunswick, NJ: Transaction Books 1987). Forst, Rainer, Contexts of Justice: political philosophy beyond liberalism and communitarianism, tr. John M. M. Farrell (Berkeley: University of California Press 2002). ‘The Limits of Toleration’, Constellations 11 (2004), 312–25. Recht auf Rechtfertigung. Elemente einer konstruktivistischen Theorie der Gerechtigkeit (Frankfurt am Main: Suhrkamp 2007); translated by Jeffrey Flynn as The Right to Justification: elements of a constructivist theory of justice (New York: Columbia University Press 2011). Toleranz im Konflikt: Geschichte, Gehalt und Gegenwart eines umstrittenen Begriffs (Frankfurt am Main: Suhrkamp 2003). ‘Toleration, Justice and Reason’, in Castiglione and McKinnon (eds.), The Culture of Toleration in Diverse Societies. Frankfurt, Harry, ‘Equality as a Moral Ideal’, Ethics 98:1 (1987), 21–42. Freeden, Michael, Ideologies and Political Theory (Oxford University Press 1998). Ideology: a very short introduction (Oxford University Press 2003). Galeotti, Anna Elisabetta, Toleration as Recognition (Cambridge University Press 2002). Gallie, W. G., ‘Essentially Contested Concepts’, Proceedings of the Aristotelian Society 56 (1956), 167–98. Gaus, Gerald, Contemporary Theories of Liberalism: public reason as a post-Enlightenment project (London: Sage 2003). Justificatory Liberalism: an essay on epistemology and political theory (Oxford University Press 1996). The Order of Public Reason: a theory of freedom and morality in a diverse and bounded world (Cambridge University Press 2011). Geuss, Raymond, Philosophy and Real Politics (Princeton University Press 2008). Golder, Ben, and George Williams, ‘Balancing National Security and Human Rights: assessing the legal response of common law nations to the threat of terrorism’, Journal of Comparative Policy Analysis 8:1 (2006), 43–62. Goldstein, Thomas, ‘Nike v. Kasky and the Definition of “Commercial Speech”’, Cato Supreme Court Review (2002–3), 63–79. Grandjean, Michel, and Bernard Roussel (eds.), Coexister dans l’intolérance: l’édit de Nantes (1598) (Geneva: Labor et Fides 1998). Gray, John, Al Qaeda and What It Means to Be Modern (London: Faber 2004). Enlightenment’s Wake (London: Routledge 1995). Isaiah Berlin (Princeton University Press 1997). Liberalisms (London: Routledge 2009). Two Faces of Liberalism (Cambridge: Polity 2000). Grayling, Anthony C., Liberty in the Age of Terror: a defence of civil liberties and Enlightenment values (London: Bloomsbury 2009). Green, Leslie, ‘On Being Tolerated’, in Kramer (ed.), The Legacy of H. L. A. Hart.

Bibliography

215

Grell, Ole Peter, Jonathan I. Israel, and Nicholas Tyacke (eds.), From Persecution to Tolerance (Oxford: Clarendon Press, 1991). Grell, Ole Peter, and Bob Scribner (eds.), Tolerance and Intolerance in the European Reformation (Cambridge University Press 1996). Gutmann, Amy, and Dennis Thompson, Deliberative Democracy and Disagreement (Cambridge, MA: Belknap Press 1996). ‘Reply’, in Macedo (ed.), Deliberative Politics. Habermas, Jürgen, Between Facts and Norms, tr. William Rehg (Cambridge: Polity 1996). Hamilton, Alexander, The Writings of Alexander Hamilton, ed. J. B. Freeman (New York: Library of America 2001). Hampshire, Stuart, Justice Is Conflict (Princeton University Press 1999). Haworth, Alan, Free Speech (London: Routledge 1998). Henderson, Bobby, The Gospel of the Flying Spaghetti Monster (New York: Villard Books 2006). Heyd, David (ed.), Toleration: an elusive virtue (Princeton University Press 1996). Hobbes, Thomas, Leviathan, ed. Richard Tuck (Cambridge University Press 1996). Honig, Bonnie, Emergency Politics: paradox, law, democracy (Princeton University Press 2009). Horton, John, ‘Rawls, Public Reason and the Limits of Liberal Justification’, Contemporary Political Theory 2:1 (2003), 5–23. ‘Three (Apparent) Paradoxes of Toleration’, Synthesis Philosophica 7:1 (1994), 7–20. ‘Why the Traditional Conception of Toleration Still Matters’, Critical Review of International Social and Political Philosophy 14:3 (2011), 289–305. Horton, John, and Glen Newey (eds.), The Political Theory of John Gray (London: Routledge 2007). Huenemann, Charlie (ed.), Interpreting Spinoza: critical essays (Cambridge University Press 2008). Hume, David, Enquiry Concerning Human Understanding, ed. L. A. Selby-Bigge, rev. Peter Nidditch (Oxford: Clarendon Press 1975). A Treatise of Human Nature, ed. L. A. Selby-Bigge, rev. Peter Nidditch (Oxford: Clarendon Press 1978). Ignatieff, Michael, The Lesser Evil: political ethics in an age of terror (Princeton University Press 2005). Israel, Jonathan, ‘William III and Toleration’, in Grell, Israel and Tyacke (eds.), From Persecution to Tolerance. Ivison, Duncan (ed.), The Ashgate Research Companion to Multiculturalism (London: Ashgate 2010). James, Susan, ‘Democracy and the Good Life in Spinoza’s Philosophy’, in Huenemann (ed.), Interpreting Spinoza. Jones, Peter, ‘Can Speech Be Intolerant?’ in Newey (ed.), Freedom of Expression. ‘Making Sense of Political Toleration’, British Journal of Political Science 37:3 (2007), 383–402.

216

Bibliography

‘Political Toleration: a reply to Newey’, British Journal of Political Science 42:2 (2011), 444–7. ‘Toleration and Neutrality: compatible ideals?’ in Castiglione and McKinnon (eds.), Toleration, Neutrality and Democracy. Kahneman, Daniel, Thinking Fast and Slow (London: Allen Lane 2011). Kahneman, Daniel, and Amos Tversky, ‘Availability: a heuristic for judging frequency and probability’, Cognitive Psychology 5:2 (1973), 207–32. Kamm, Frances, ‘Physician-Assisted Suicide, the Doctrine of Double Effect, and the Ground of Value’, Ethics 109:3 (1999), 586–605. Kant, Immanuel, Political Writings, rev. edn tr. and ed. Hans Reiss (Cambridge University Press 1991). Kaplan, George W., ‘Complications of Circumcision’, Urologic Clinics of North America 10:3 (1983), 543–9. Katalin, Péter, ‘Tolerance and Intolerance in Sixteenth Century Hungary’, in Grell and Scribner (eds.), Tolerance and Intolerance in the European Reformation. Kautz, Steven, ‘Liberalism and the Idea of Toleration’, American Journal of Political Science 37:2 (1993), 610–32. Kilcullen, John, Sincerity and Truth: essays on Arnauld, Bayle and toleration (Oxford University Press 1988). King, Preston, Toleration, 2nd edn (London: Frank Cass 1998). Kohlberg, Lawrence, ‘The Claim to Moral Adequacy of a Highest Stage of Moral Judgement’, Journal of Philosophy 70:18 (1973), 630–46. Korsgaard, Christine, ‘The Right to Lie: Kant on dealing with evil’, Philosophy and Public Affairs 15:4 (1986), 325–49. Kramer, Matthew H. (ed.), The Legacy of H. L. A. Hart: legal, political, and moral philosophy (Oxford University Press 2008). Kuhn, Thomas, The Structure of Scientific Revolutions, 3rd edn (University of Chicago Press 1996). Kukathas, Chandran, The Liberal Archipelago: a theory of diversity and freedom (Oxford University Press 2003). Kuran, Timur, and Cass R. Sunstein, ‘Availability Cascades and Risk Regulation’, Stanford Law Review 51:4 (1999), 683–768. Kymlicka, Will, Multicultural Citizenship (Oxford University Press 1995). Larmore, Charles, Patterns of Moral Complexity (Cambridge University Press 1987). Lasswell, Harold, Politics: who gets what, when, how (London: Peter Smith 1990). Lázár, István, Transylvania: a short history, tr. Thomas J. de Kornfeld (Safety Harbor, FL: Simon Publications 1996). Li, Xiaorong, ‘Tolerating the Intolerable: the case of female genital mutilation’, Philosophy and Public Policy Quarterly 21:1 (2001), 2–8. Lichtenberg, Judith, ‘War, Innocence, and the Doctrine of Double Effect’, Philosophical Studies 74:3 (1994), 347–68. Lijphart, Arend, The Politics of Accommodation: pluralism and democracy in the Netherlands (Berkeley: University of California Press 1968). Loader, Ian, and Neil Walker, Civilizing Security (Cambridge University Press 2007).

Bibliography

217

Locke, John, An Essay Concerning Human Understanding, ed. Peter Nidditch (Oxford: Clarendon Press 1975). A Letter Concerning Toleration: in focus, ed. John Horton and Susan Mendus (London: Routledge 1991). Letters on Toleration (Bombay: Education Society 1867). Political Essays, ed. Mark Goldie (Cambridge University Press 1997). Macedo, Stephen (ed.), Deliberative Politics: essays on democracy and disagreement (Oxford University Press 1999). Machiavelli, Niccolò, Discourses on Livy, tr. Harvey Mansfield and Nathan Tarcov (University of Chicago Press 1996). MacIntyre, Alasdair, After Virtue (London: Duckworth 1981). Whose Justice? Which Rationality? (London: Duckworth 1988). Mackie, John L., ‘Evil and Omnipotence’, Mind 64:254 (1955), 200–12. Marcuse, Herbert, ‘Repressive Tolerance’, in Wolff (ed.), A Critique of Pure Tolerance. Marshall, John, John Locke: Resistance, Religion and Responsibility (Cambridge University Press 1994). John Locke, Toleration and Early Enlightenment Culture (Cambridge University Press 2006). Mason, Andrew, Explaining Political Disagreement (Cambridge University Press 1993). McIntyre, Alison, ‘Doing Away with Double Effect’, Ethics 111:2 (2001), 219–55. McKinnon, Catriona, Toleration: a critical introduction (London: Routledge 2005). McMahan, Jeff, British Nuclear Weapons: for and against (London: Junction Books 1981). Meiklejohn, Alexander, Political Freedom (New York: Galaxy 1960). Mendus, Susan (ed.), Justifying Toleration: conceptual and historical approaches (Oxford University Press 1988). Michaelsen, Christopher, ‘Balancing Civil Liberties against National Security?’ University of New South Wales Law Journal 29:2 (2006), 1–21. ‘Liberty versus Security: rights-based objections to a simple balancing exercise’, in Enemark (ed.), Ethics of War in a Time of Terror. Mill, John Stuart, Utilitarianism and On Liberty, ed. Mary Warnock (Oxford: Blackwell 2003). Miller, David, and Larry Siedentop (eds.), The Nature of Political Theory (Oxford: Clarendon Press 1983). Morris, William, News from Nowhere, ed. David Leopold (Oxford University Press 2003). Mouffe, Chantal, The Return of the Political (London: Verso 2005). Murphy, Andrew, Conscience and Community: revisiting toleration and religious dissent in early modern England and America (University Park, PA: Penn State University Press 2001). Murphy, Jeffrey, Kant: The Philosophy of Right (Basingstoke: Macmillan 1970). Nagel, Thomas, Equality and Partiality (Oxford University Press 1991).

218

Bibliography

National Committee on American Foreign Policy, ‘Beyond Textual Ambiguity: the need to clarify the Belfast Agreement’, American Foreign Policy Interests 23:4 (2001), 253–5. Neocleous, Mark, Critique of Security (Edinburgh University Press 2008). ‘Security, Liberty and the Myth of Balance: towards a critique of security politics’, Contemporary Political Theory 6:2 (2007), 131–49. Newey, Glen, After Politics: the rejection of politics in contemporary liberal philosophy (London: Routledge 2001). ‘Against Thin-Property Reductivism: toleration as supererogatory’, Journal of Value Inquiry 31:2 (1997), 231–49. ‘Denial Denied: Freedom of Speech’, Amsterdam Law Forum 2:1 (2009). http://ojs.ubvu.vu.nl/alf/article/view/109/198. ‘Gray’s Blues: pessimism as a political project’, in Horton and Newey (eds.), The Political Theory of John Gray. Hobbes and ‘Leviathan’ (London: Routledge 2008). ‘Is Democratic Toleration a Rubber Duck?’ in Castiglione and McKinnon (eds.), Toleration, Neutrality and Democracy, repr. from Res Publica 7:3 (2001), 315–36. ‘Just Politics’, Critical Review of International Social & Political Philosophy 15:2 (2012), 165–82. ‘Metaphysics Postponed: liberalism, pluralism and neutrality’, Political Studies 45:2 (1997), 296–311. ‘Philosophy, Politics, and Contestability’, Journal of Political Ideologies, 6:3 (2001), 245–61. ‘Political Lying: a defense’, Public Affairs Quarterly 11:2 (1997), 93–116. ‘Political Toleration: a reply to Jones’, British Journal of Political Science 41:1 (2011), 223–7. ‘Rawlsian Liberalism at the Limits of Intolerance’, in Young (ed.), The Legacy of John Rawls. Review of Anna Elisabetta Galeotti, Toleration as Recognition, Utilitas 18:3 (2006), 310–12. ‘Toleration, Politics, and the Role of Murality’, in Williams and Waldron (eds.), Toleration and Its Limits. ‘Value Pluralism in Contemporary Liberalism’, Dialogue: the Canadian philosophical review 37:3 (1998), 493–522. Virtue, Reason and Toleration: the place of toleration in ethical and political philosophy (Edinburgh University Press 1999). (ed.), Freedom of Expression: counting the costs ( Newcastle upon Tyne: Cambridge Scholars Press 2007). Nozick, Robert, Anarchy, State, and Utopia (Oxford: Blackwell 1974). Paine, Thomas, Rights of Man, ed. Henry Collins (Harmondsworth: Penguin 1969). Peacocke, Christopher, A Study of Concepts (Cambridge, MA: MIT Press 1992). Pettit, Philip, Republicanism: a theory of freedom and government (Oxford University Press 1997). Plantinga, Alvin, God, Freedom, and Evil (Grand Rapids, MI: Wm Eerdmans 2002).

Bibliography

219

Plato, Republic, tr. Paul Shorey (Cambridge, MA: Harvard University Press 1930). Popper, Karl, The Open Society and Its Enemies (London: Routledge 1946). ‘Toleration and Intellectual Responsibility’, in Edwards and Mendus (eds.), On Toleration. Posner, Eric A., and Adrian Vermeule, Terror in the Balance: security, liberty and the courts (Oxford University Press 2006). Posner, Richard, ‘Security versus Civil Liberties’, Atlantic Monthly (1 December 2001), 46–7. Proast, Jonas, The Letter Concerning Toleration Briefly Considered and Answer’d (Oxford 1690). Quinn, Warren, ‘Actions, Intentions, and Consequences: the doctrine of double effect’, Philosophy and Public Affairs 18:4 (1989), 334–51. Raphael, David D., ‘Toleration, Choice and Liberty’, Government and Opposition 6:2 (1971), 229–34. Rawls, John, Collected Papers, ed. Samuel Freeman (Cambridge, MA: Harvard University Press 1999). Justice as Fairness: a restatement, ed. Erin Kelly (Cambridge, MA: Harvard University Press 2001). ‘Justice as Fairness: political, not metaphysical’, in Rawls, Collected Papers. The Law of Peoples (Oxford University Press 1999). Political Liberalism (New York: Columbia University Press 1996). A Theory of Justice (Oxford University Press 1971). Raz, Joseph, The Authority of Law (Oxford: Clarendon Press 1979). ‘Autonomy, Toleration and the Harm Principle’, in Mendus (ed.), Justifying Toleration. ‘Can There be a Right-Based Moral Theory?’ in Waldron (ed.), Theories of Rights. The Morality of Freedom (Oxford: Clarendon Press 1986). Richards, David A. J., ‘Autonomy in Law’, in Christman (ed.), The Inner Citadel. Richardson, Henry, Democratic Autonomy: public reasoning about the ends of policy (Oxford University Press 2002). Riker, William, Liberalism versus Populism (San Francisco, CA: W. H. Freeman 1980). Rorty, Richard, Contingency, Irony, and Solidarity (Cambridge University Press 1989). ‘Postmodernist Bourgeois Liberalism’, Journal of Philosophy 80:10 (1983), 583–9. Rosenzweig, Paul, ‘Balancing Security and Liberty’, paper published by the Heritage Foundation, 14 May 2003; www.heritage.org/press/commentary/ ed051403a.cfm. Rousseau, Jean-Jacques, The Social Contract, tr. G. D. H. Cole (London: Dent 1913). Ryan, Alan, ‘Hobbes, Toleration, and the Inner Life’, in Miller and Siedentop (eds.), The Nature of Political Theory. ‘A More Tolerant Hobbes?’ in Mendus (ed.), Justifying Toleration.

220

Bibliography

Sandel, Michael, Liberalism and the Limits of Justice (Cambridge University Press 1982). Scanlon, Thomas M., ‘Contractualism and Utilitarianism’, in Sen and Williams (eds.), Utilitarianism and Beyond. ‘The Difficulty of Tolerance’, repr. in Scanlon, The Difficulty of Tolerance (Cambridge University Press 2003). What We Owe to Each Other (Cambridge, MA: Harvard University Press 1999). Schmitt, Carl, Political Theology: four chapters on the concept of sovereignty, tr. and ed. George Schwab (University of Chicago Press 2005). Schwartzmann, Micah, ‘The Relevance of Locke’s Religious Arguments for Toleration’, Political Theory 33:4 (2005), 678–705. Sellars, Wilfrid, ‘Empiricism and the Myth of the Given’, in Feigl and Scriven (eds.), Minnesota Studies in the Philosophy of Science, Vol. I. Sen, Amartya K., and Bernard Williams (eds.), Utilitarianism and Beyond (Cambridge University Press 1982). Skinner, Quentin, Hobbes and Republican Liberty (Cambridge University Press 2008). Visions of Politics, Vol. I: Regarding Method (Cambridge University Press 2002). Visions of Politics, Vol. II: Hobbes and Civil Science (Cambridge University Press 2002). Slovic, Paul, The Perception of Risk (Sterling, VA: Earthscan 2000). Smith, Tony, ‘Balancing Liberty and Security? A Legal Analysis of United Kingdom Anti-Terrorist Legislation’, European Journal of Criminal Policy and Research 13:1/2 (2007), 73–83. Spinoza, Benedictus de/Baruch, Ethics, ed. and tr. Edwin Curley, in The Collected Works of Spinoza (Princeton University Press 1985), Vol. I. The Political Works, ed. A. G. Wernham (Oxford: Clarendon Press 1958). Spurr, John, ‘The Church of England, Comprehension, and the Toleration Act of 1689’, English Historical Review 104 (1989), 927–46. Stears, Marc, ‘The Politics of Compulsion’, British Journal of Political Science 37:3 (2007), 505–31. Taylor, Charles, Sources of the Self (Cambridge University Press 1989). Taylor, Stuart, ‘Rights, Liberties, and Security: recalibrating the balance after September 11’; www.brookings.edu/articles/2003/winter_terrorism_taylor. aspx. Temkin, Larry, Inequality (Oxford University Press 1996). Trevor-Roper, Hugh, ‘Toleration and Religion After 1688’, in Grell, Israel and Tyacke (eds.), From Persecution to Tolerance. Tuck, Richard, ‘Hobbes and Locke on Toleration’, in Dietz (ed.), Thomas Hobbes and Political Theory. Tuckness, Alex, ‘Locke’s Main Argument for Toleration’, in Williams and Waldron (eds.), Toleration and Its Limits. Unghváry, Alexander Sándor, The Hungarian Protestant Reformation in the Sixteenth Century under the Ottoman Impact (Ann Arbor: University of Michigan Press 1989).

Bibliography

221

US State Department Bureau of Democracy, Human Rights, and Labor, 2004 Religious Freedom Report; www.state.gov/g/drl/rls/irf/2004/35496.htm. Venard, Marc, ‘L’Église catholique bénéficiaire de l’édit de Nantes: le témoignage des visites épiscopales’, in Grandjean and Roussel (eds.), Coexister dans l’intolérance. Voltaire, Treatise on Tolerance and Other Writings, ed. Simon Harvey, tr. Brian Masters (Cambridge University Press 2000). Waldron, Jeremy, ‘Homelessness and the Issue of Freedom’, in Waldron, Liberal Rights. Liberal Rights: collected papers 1981–1991 (Cambridge University Press 1993). ‘Locke: Toleration and the Rationality of Persecution’, in Susan Mendus (ed.), Justifying Toleration. ‘Security and Liberty: the image of balance’, Journal of Political Philosophy 11:2 (2003), 191–210. (ed.), Theories of Rights (Oxford University Press 1984). Walzer, Michael, Just and Unjust Wars (New York: Basic Books, 1977). Watt, W. Montgomery, Muhammad at Medina (Oxford: Clarendon Press 1956). Wenar, Leif, ‘Political Liberalism: an internal critique’, Ethics 106:1 (1995), 32–62. Williams, Bernard, ‘The Idea of Equality’, in Williams, Problems of the Self (Cambridge University Press 1973). In the Beginning Was the Deed (Princeton University Press 2005). ‘Toleration: an impossible virtue?’ in Heyd (ed.), Toleration: an elusive virtue. Williams, Melissa S., and Jeremy Waldron (eds.), Toleration and Its Limits: NOMOS XLVIII: (New York University Press 2008). Willis, Andrew, Silent Accomplice: the untold story of France’s role in the Rwandan genocide (London: I. B. Tauris 2007). Wilson, James Q., and George L. Kelling, ‘Broken Windows’, Atlantic Monthly 249 (1982), 29–38. Wittgenstein, Ludwig, Philosophical Investigations, tr. and ed. G. E. M. Anscombe (Oxford: Blackwell 1951). Wolff, Robert Paul (ed.), A Critique of Pure Tolerance (Boston, MA: Beacon Books 1969). In Defense of Anarchism (New York: Harper & Row 1970). Young, Shaun P. (ed.), The Legacy of John Rawls (Aldershot: Ashgate 2009). Zagorin, Perez, Richard Ashcraft and Richard W. F. Kroll (eds.), Philosophy, Science and Religion in England, 1640–1700 (Cambridge University Press 1992).

Index

abortion 16, 43, 45 absence of action 172 acceptance 43, 45, 96 act–agent distinction 98 ad hoc social theory 87 adultery 67 advocacy 5 affirmative action 82 agency 45, 52, 54, 86, 161, 170 attitude of agent 158 collective 193 intentionality and 179 power and 142–3, 161–2 self-toleration and 170–1, 173 agonism 25, 27, 100, 167, 207 akrasia 169 Ali, Ayaan Hirsi 2, 183 Alm, Niko 104 amplification effects 138–9, 154 anarchism 24 Aristotle 10 Armenian genocide 26 arthritis 33 association 189, 191, 206 freedom of 182, 185, 188–91, 193, 196 astrology 78 asymmetry 37, 51, 54, 72 atheism 112–14, 116 authoritarianism 60, 130 authoritative sources 72 authority 18, 34 imposition of 15 autonomy 28, 41, 43, 149 toleration and 10, 28–9 availability effect 137, 208 awkwardness of tolerance as political ideal 3, 207 Baha’i faith 88–9 balance between liberty and security 123–8, 140–1, 186 purported tension 130–4

222

terrorism and 124–5, 131, 134–40 toleration and 128–30 balkanisation 103, 105 Barry, Brian 91, 106, 120 Bayle, Pierre 30, 40, 94, 114 Benedict XVI, Pope 1 Berlin, Isaiah 166 blasphemy 67 bombs 179–80 Bonne, Emmanuel 164 books 52–3 burning 2 Bootleg Argument 117, 188 Brown, Wendy 154 burdensome judgements 61–8, 95, 208 Burge, Tyler 33 capacity 143, 152 censorship 1, 11, 121, 186–8, 190–1 change, toleration and 14–19 child labour 196–205 circumcision female genital mutilation (FGM) 34–7 male 35–7 civic rights 12, 40, 86, 100–1, 122, 125, 152–3 civil society groups 15–16, 30 Clark, Ross 104–5 coercion 15, 25, 39, 43, 76, 82, 94, 99, 108, 112, 144, 189 coexistence 161, 167 Cohen, Aaron 57 compromise 45, 148 compulsive behaviour 171 confessional states 44 confrontation 64 conscience freedom of speech and 187 politics and 79 sovereign state and 74–80 consensus 165, 168 overlapping 131, 167

Index consequentialism 9 conspiracy theories 137 Constitution of Medina 130–1 constructivism 29 contractarianism 29, 54 contracts 33–4, 64, 93, 182, 188–9, 191 controversy, normative 5 convictions, private 68 corporations Nike and Kasky case 196–206 counterfactual variation 143, 153, 156, 158–61, 173, 181 courage 13 creationism 78, 187 criminal justice systems 153 culpability 178 Dávid, Ferenc 47 deception see lying decontestation 4 definitions politics 22–7, 59, 99, 193, 207 toleration 6, 19 deliberation 4, 24–5, 46, 208 deliberative forums 24 democratic toleration 40, 59, 144, 207 structural transformation of toleration 40–7 toleration and democratic politics 47–54 toleration and political philosophy 54–9 denial, denying of 185 descriptive accounts of toleration 38 Devlin, Patrick 129, 134 disagreement 32, 62, 73, 95 reasonable 63–4, 71–2, 96 religious differences 70 toleration and 6 disapproval 13, 28, 34, 49–50, 70, 73, 82, 94, 97, 101, 151, 174 of speech 184 toleration and 6–7, 9–10, 14, 27, 30–1, 45, 92 discursive principle 96, 99 disputes over toleration 12, 17 distancing 74 distaste, toleration and 6 distortions of toleration 3 diversity 75 DNA database 126–7 Donnachie, Paul 56–9 double-effect doctrine 175, 177–8, 180 doxastic argument 78, 110–11, 117–18, 122 drugs 28, 44 Dworkin, Ronald 82

223 Edict of Nantes 14, 48, 128 education 82, 109 Egypt identity card controversy 88 religious minorities 88–9, 98 embodied freedom 158 environmental issues 26 epistemic reasonableness of belief 73 equality 13–14, 38, 41, 74, 122, 126, 135, 165 equal treatment 82 neutrality and 90 respect and 81–9, 91, 100 ethics-of-belief argument 76 European Convention on Human Rights 103, 183 European Union, terrorism and 164 existence of God 70 expressivism 44 fairness 61 fallibilism 119, 167 fanaticism 78 female genital mutilation (FGM) 34–7 flags 56–9 flat-earthers 72 forbearance 42–3 force majeure 21 Forst, Rainer 93–100, 165 France banning of burqa 2 religious freedom in 14, 48, 128 freedom 34, 39, 123, 156, 158, 160, 182 of association 182, 185, 188–91, 193, 196 balance between liberty and security 123–8, 140–1 purported tension 130–4 terrorism and 124–5, 131, 134–40 toleration and 128–30 free will 156 power and 157 religious freedom 14, 47–9, 102–3, 122 Constitution of Medina and 130–1 France 14, 48, 128 liberalism after Locke 118–22 Locke on 108–18 United Kingdom 48–9, 103, 106, 128 United States of America 102–4, 134 of speech see speech, freedom of Frost, Robert 14 functional view of politics 32, 99–100, 207, 209 fundamentalism 1, 8, 17, 67, 92–4, 96–7, 99, 144

224

Index

Galeotti, Anna Elisabetta 14 gambling 84, 88, 90–1 generality 94 genocide 26 Giorgi, Loretta M. 36 Gogh, Theo van 1, 183 good life 73, 77 Gray, John 166 Grayling, Anthony C. 9 Greece, ancient 97 guilt 153 Gutmann, Amy 14 Habermas, Jürgen 94 Hamilton, Alexander 37, 129, 132 hegemony 106–7 history of toleration 5 Hobbes, Thomas 5, 24, 49, 69, 75–7, 80, 108–9, 123–4, 127, 133 Holocaust 150 denial 165, 183, 194 homosexuality 2, 11, 13, 31, 43, 93–8, 129 marriage and 134 Hume, David 78 Hungary 14, 47 ideals 150, 159–60, 162, 166, 178, 181, 209 ideal conditions for toleration 4 idealism 38 ideal-theory model 17 limits of idealisation 152–7, 161 politics and 26 identity 86 beliefs and 78–9 identity card controversy in Egypt 88 Igarashi, Hitoshi 183 imperfection 166 imposition of authority 15 impossibility of toleration 11 incapacity 150 inconsequentiality 159 indeterminacy 159, 174 indifference 7 indoctrination 111 institutions 61 insufficient-reason arguments 67 intelligent design 104 intention 178–9 interest groups 15–16, 30 interpersonal toleration 27, 42 intolerance 110 justified 52 liberalism and 163 mutual accusations of 68

security-based arguments for 19, 111, 113, 121–2 of terrorism 140, 163–4, 176–81 toleration of 8, 11, 15, 18, 51 zero tolerance 13, 46, 54, 60, 124, 163 intransitive toleration 7, 30 Iraq war 176–7 Ireland 179 irrationality 108, 110 Islam 2, 106, 115, 120, 183 Constitution of Medina 130–1 ‘Ground Zero Mosque’ 2 male circumcision and 35 religious clothing 2, 7, 83 religious toleration and 130–1 Sharia law 1, 102, 105–6, 108, 117 Israel desecration of Israeli flag 56–9 terrorist attacks on 131 János II Zsigmond Zápolya 47 Jedi religion 104 Jones, Peter 92, 165 Judaism 114 desecration of Israeli flag and accusations of anti-Semitism 56–9 family law 103, 107 Holocaust 150 denial 165, 183, 194 Islamic Constitution of Medina and 130–1 male circumcision and 35 judgements burdensome 61–8, 95, 208 privacy of 77, 80 vulnerability of judgements of toleration 128 see also disapproval justice 21, 61, 126, 153 justification 75, 94, 97–8, 120, 149, 164–5, 209 Kahneman, Daniel 131 Kant, Immanuel 41, 46, 55 Kasky, Marc 196–205 Kuran, Timur 137 Larmore, Charles 43 Lasswell, Harold 22–3 legislation 106–7 libel 205 liberalism 1, 8, 11, 14–15, 17–18, 21, 23, 39–40, 54, 65, 75, 80, 93, 182, 208 after Locke 118–22 conscience and 77, 79

Index equality and 87 hegemony 106–7 intolerance and 163 multiculturalism and 102–8 neutrality and 89, 94, 106 respect and 81, 85 war and 181 libertarianism 4 limits of toleration 8 Locke, John 18, 30, 40, 76, 108–19, 208 lying 174–6 freedom of speech and the right to lie 191–6, 202, 205 Nike and Kasky case 196–206 Machiavelli, Niccolò 111 male circumcision 35–7 Marcuse, Herbert 154 marriage, same-sex 134 Meiklejohn, Alexander 187, 193 meliorist projects 167 Mill, John Stuart 10, 87, 101, 166, 168, 184 Milton, John 30 misplaced toleration 163 misrepresentation 191, 205 see also lying modus vivendi 142, 158, 160–1 monarchy, prerogative power and 30, 41, 144 moralist approach to toleration 27 Morris, William 24 Mouffe, Chantal 25 multiculturalism 102–8 natural rights 133 naturalism 29 Nazi regime 150 negative exclusionary reasons 92 negotiation 189–90 Netherlands 6 Neturei Karta 57 neutrality 42, 44, 49, 74, 78–80, 167 equality and 90 liberalism and 89, 94, 106 respect and 81, 89–93 Nike Corporation 196–206 non-political toleration 27–31 non-prevention 158 norms 15, 209 conflicting 69 conflicting normative judgements 53 normative agreement approach 55–9 normative claims 23 normative controversy 5

225 reasonable 71 rejection of 72 Nozick, Robert 178, 180 obesity 73 objectivity 23 omnipotence 153, 156, 159 opinion, punishment by 101, 168 overlapping consensus 131, 167 paedophilia 97–8 Paine, Thomas 41, 46, 55 Palestine 56 Palmer, Mitchell 139 paradoxical nature of toleration 7–10, 12, 18 partial compliance theory 16 Pastafarianism 104 paternalism 87 perfectibilitarianism 166 perfectionism 44 permission 68, 144 philosophy: toleration and political philosophy 54–9 Plantinga, Alvin 157 pluralisation 12 pluralism reasonable 60–2, 66–7 reasonable rejectability and 67–74 religious 102 political correctness 13, 184 political toleration 22, 31–8, 155, 165, 174 pornography 52, 194–6 positive discrimination 82 power 20–1, 24, 26, 142, 162, 165 coercion 15, 25, 39, 43, 76, 82, 94, 99, 108, 112, 144, 189 complications 158–62 democracy and 41 omnipotence 153, 156, 159 power condition 142–52 prerogative 30, 41, 144 reputation of 139 pragmatism 38, 148, 153, 159 prejudice 151 prerogative power 30, 41, 144 prevention 144, 147, 155 costs of 146, 148 prison 127 privacy of judgement 77, 80 private convictions 68 private opinion 75–6, 78 Proast, Jonas 108, 118 prohibition 45, 68, 94 property rights 4, 140 proselytism 120

226

Index

public goods 45 public interest 33, 159 public order 39, 45 Quinn, Warren 179–80 race 26 racism 145, 148 desecration of Israeli flag and accusations of racism and anti-Semitism 56–9 paradox of ‘tolerant racist’ 9–10 Rawls, John 14, 21–3, 38, 55, 61, 64–5, 68, 74, 76–7, 80, 95–6, 120, 123, 131, 153, 186, 208 realism 33 Realpolitik 112, 209 reason 32 reasonableness 56, 59, 64, 98–100, 120, 165, 167, 208 epistemic reasonableness of belief 73 reasonable disagreement 63–4, 71–2, 96 reasonable norms 71 reasonable pluralism 60–2, 66–7 reasonable rejectability 67–74, 90, 121 reciprocity 94 recognition, toleration as 14 reflectiveness 38 Reitblat, Chanan 56–9 rejectability, reasonable 67–74, 90, 121 relativism 52, 65 religion 90 conscience and the sovereign state 74–80 differences in 70 freedom of 14, 47–9, 102–3, 122 Constitution of Medina and 130–1 France 14, 48, 128 liberalism after Locke 118–22 Locke on 108–18 United Kingdom 48–9, 103, 106, 128 United States of America 102–4, 134 fundamentalism 1, 8, 17, 67, 92–4, 96–7, 99, 144 identity card controversy in Egypt and 88 multiculturalism and 102–8 pluralism 102 religious clothing 2, 7, 11, 32, 83, 88, 104, 106 respect and 87 symbols 1, 11 replication problem 12, 14, 87 repugnance 149, 181 reputation of power 139 respect 38, 81, 100–1, 144, 157, 167, 182 equality and 81–9, 91, 100 Forst’s respect conception 93–100 neutrality and 81, 89–93

risk 138 Roman Catholic church 1 adoption agencies and same-sex couples 2, 11 Locke on toleration and 112–17, 120 romanticism 44 Rothschild, Leehee 57 Rousseau, Jean-Jacques 34 Rwanda 26 Sacheverell, Henry 49 Scanlon, Thomas M. 120 scepticism 10, 63, 66, 70, 74, 119 Scottish Jews for a Just Peace (SJJP) 57 Scottish Palestine Solidarity Campaign (SPSC) 56 security 39, 123, 208 arguments for intolerance and 19, 111, 113, 121–2 balance between liberty and security 123–8, 140–1 purported tension 130–4 terrorism and 124–5, 131, 134–40 toleration and 128–30 DNA database and 126–7 sedition 19, 53, 114, 121, 123 toleration as 60, 80 burdensome judgements 61–8, 95, 208 conscience and the sovereign state 74–80 pluralism, reasonable rejectability and toleration 67–74 reasonable pluralism 60–2 segregation 82 self-deception 169 self-limiting mechanisms 8 self-restraint 34, 44, 54, 99, 131, 142, 149–52, 160 self-toleration 169–76 Shaftesbury, Lord 49 Sharia law 1, 102, 105–6, 108, 117 Sikhism, religious clothing and 83–4, 88, 106 situatedness, political 207 slavery 33–4, 37 Slovic, Paul 138 Smith, Kevin 73 smoking 84, 88 ban 16, 32 social context of toleration 10 soteriology 112 sovereignty: conscience and the sovereign state 74–80 speech, freedom of 183–4, 205–6 censorship 1, 11, 121, 186–8, 190–1 hate speech 183

Index

227

right to lie 191–6, 202, 205 Nike v. Kasky case 196–206 tolerable and tolerant speech 184–91 Spinoza, Baruch 18, 117, 124, 132–3 stability 21 stigma 101 strains of commitment 12 structural transformation of toleration 40–7 Sunstein, Cass R. 137 surveillance 135 sweatshops 196–205 symbols 59 desecration of Israeli flag 56–9 religious 1, 11 terrorism and 137 symmetry 51, 53 asymmetry 37, 51, 54, 72

United Kingdom 14, 45 circumcision in 36 DNA database 126–7 gambling 84 legalisation of homosexuality 129 legislation 106–7 religious freedom 48–9, 103, 106, 128 Sharia law in 1, 102, 105–6, 108, 117 smoking ban 16 terrorism in 135 United Nations, terrorism and 163 United States of America circumcision in 36 freedom of speech in 183, 194, 201–2 Nike v. Kasky case 196–206 Red Scare 139 religious freedom in 102–4, 134 terrorist attack on 1–2

tastes 64 taxation 106 Tchatchouwo, Victor 163 television broadcasting 187 terrorism 1–2, 8, 52, 182, 186 balance between liberty and security and 124–5, 131, 134–40 intolerance of 140, 163–4, 176–81 war on terror 141, 178 theft 86 theodicy 156 theories of toleration 2–5, 28–9, 31, 53, 166, 170 Thompson, Dennis 14 transitive toleration 7, 170, 182 truisms 6 Turkey 26

volatility of toleration 31 war 21, 176–7, 181 on terror 141, 178 Watt, W. Montgomery 131 welfare benefits 106 Westergaard, Kurt 183 Williams, Rowan 104–5 Wittgenstein, Ludwig 170, 176 women: female genital mutilation (FGM) 34–7 Young, Andrew 196 zealotry 78 zero tolerance 13, 46, 54, 60, 124, 163 Zimbabwe 148–9