Liberty, Games and Contracts: Jan Narveson and the Defence of Libertarianism

LIBERTY, GAMES AND CONTRACTS

This page intentionally left blank

Liberty, Games and Contracts Jan Narveson and the Defence of Libertarianism

Edited by MALCOLM MURRAY University of Prince Edward Island, Canada

© Malcolm Murray 2007 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without the prior permission of the publisher. Malcolm Murray has asserted his moral right under the Copyright, Designs and Patents Act, 1988, to be identified as the editor of this work. Published by Ashgate Publishing Limited Gower House Croft Road Aldershot Hampshire GU11 3HR England

Ashgate Publishing Company Suite 420 101 Cherry Street Burlington, VT 05401-4405 USA

Ashgate website: http://www.ashgate.com

British Library Cataloguing in Publication Data Liberty, games and contracts : Narveson and the defence of libertarianism 1. Narveson, Jan, 1936– 2. Libertarianism 3. Contractarianism (Ethics) I. Murray, Malcolm 320.5'12 Library of Congress Cataloging-in-Publication Data Liberty, games, and contracts : Jan Narveson and the defence of libertarianism / edited by Malcolm Murray. p. cm. Includes bibliographical references and index. ISBN 0-7546-5681-0 (hardcover : alk. paper) 1. Libertarianism. 2. Social contract. 3. Contractarianism (Ethics) 4. Narveson, Jan, 1936– —Political and social views. I. Murray, R. Malcolm (Robert Malcolm), 1959– II. Narveson, Jan, 1936– JC585.L4245 2006 320.51'2—dc22 2006011547 ISBN 978-0-7546-5681-4

Typeset by IML Typographers, Birkenhead, Merseyside and printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall.

Contents Notes on the Contributors Acknowledgments Biographical Note on Jan Narveson

vii xi xiii

Introduction Malcolm Murray I

1

Contractarianism 1. Some Remarks on the Foundations of Libertarianism E. J. Bond 2. Contracting Justice John T. Sanders 3. Is Agreement Enough? Tibor R. Machan 4. Does Scepticism Beget Libertarianism? A Response to Narveson on Reason, Morality and Politics Leo Groarke 5. Contractarian Education Chris Tucker

II

11 19 33

47 65

From Contractarianism to Libertarianism 6. The Value of Values: The Importance of Autonomy in Contractarian Reasoning Susan Dimock 7. Simple Games and Complex Ethics Peter Danielson 8. Why Contractarians are not Libertarians ... Evolutionarily Speaking Malcolm Murray 9. Getting the Baseline Right Paul Viminitz

III

81 103 115 129

Property and Liberty

10. Liberty, Property and the Libertarian Idea Ann Levey

v

147

Liberty, Games and Contracts

vi

11. Who Owns Me: Me or my Mother? How to Escape Okin’s Problem for Nozick’s and Narveson’s Theory of Entitlement Duncan MacIntosh 12. On Original Appropriation Peter Vallentyne 13. Morality as an Evolutionarily Stable Strategy: A Naturalistic Account of Libertarianism Grant A. Brown 14. Reconciling Radicals: Market Contractarianism and Fundamentalist Utilitarianism Sheldon Wein IV

157 173

179

197

Response

15. Social Contract, Game Theory and Liberty: Responding to My Critics Jan Narveson Bibliography Index

217

241 251

Notes on the Contributors E. J. Bond, BA, MA (Queen’s University at Kingston, Ontario), PhD (Cornell), Professor Emeritus of Philosophy, Queen’s, is the author of Reason and Value (1983) and Ethics and Human Well-being (1996). He contributed four articles to Becker & Becker (eds) the Encyclopedia of Ethics, 2nd edn (2001), and has contributed many papers to professional journals. Grant A. Brown has a BA and MA in Philosophy (University of Waterloo, Ontario) and a DPhil in Political Philosophy (Oxford University). His thesis, Functional Libertarianism, explores the implications for political theory of accepting the neo-Darwinian conception of human nature. From 1990 to 1999, Dr Brown taught ethics and political philosophy at the University of Lethbridge, Alberta. In 2002, he obtained an LLB (University of Alberta) and has been in the private practice of law since then. Peter Danielson is Director of the W. Maurice Young Centre for Applied Ethics, University of British Columbia. His theoretical work deploys evolutionary game theory and modelling. His NERD (Norms Evolving in Response to Dilemmas) group, funded by Genome Canada, constructs new computational devices to explore how people evaluate new technology. Publications include Artificial Morality: Virtuous Robots for Virtual Games (1992), Modeling Rationality, Morality, and Evolution; Vancouver Studies in Cognitive Science, vol, 7 (ed., 1998) and ‘Competition among Cooperators: Altruism and Reciprocity’ (2002). Susan Dimock received her PhD from Dalhousie University, Halifax, Nova Scotia. She has been a member of the Philosophy Department at York University, Toronto, Ontario since 1991. Her research interests span topics in meta-ethics, normative theory, political philosophy and the philosophy of law, and the history of early modern philosophy. She has edited and co-edited numerous books, including a volume on liberalism with Jan Narveson. Leo Groarke is Professor of Philosophy and Dean of the Brantford Campus of Wilfrid Laurier University, Ontario. He has published widely on the history of ideas, argumentation, ethics, and social and political issues. His books include Greek Scepticism: Anti-Realist Trends in Ancient Times (1985), The Ethics of the New Economy (1989) and (with Michael Allen Fox) Nuclear War: Philosophical Perspectives (1985).

vii

viii

Liberty, Games and Contracts

Ann Levey is Associate Professor of Philosophy at the University of Calgary. Her research interests include property rights, feminism and Hume. She is currently working on feminism and contractarianism and is the author of ‘Liberalism, Adaptive Preferences and Gender Equality’ (Hypatia, 2005). Tibor R. Machan holds the R. C. Hoiles Chair in Business Ethics at Chapman University, California, and is a research fellow at Stanford’s Hoover Institution, and the Pacific Research Institute, San Francisco. Machan has written 34 books, including Human Rights and Human Liberties (1975), Individuals and Their Rights (1989), Capitalism and Individualism (1990), The Virtue of Liberty (1994), Private Rights and Public Illusions (1995), Generosity: Virtue in Civil Society (1998), and Libertarianism Defended (2006). Duncan MacIntosh is an Associate Professor of Philosophy at Dalhousie University. His papers have appeared in the collections of Christine Tappolet and Sarah Stroud (eds), Weakness of Will and Practical Irrationality (2003) and in Peter Danielson (ed.), Modeling Rationality, Morality and Evolution; Vancouver Studies in Cognitive Science (1998), as well as in various journals in Australia, Britain, Canada and the United States. Malcolm Murray is an Associate Professor of Philosophy at the University of Prince Edward Island, Canada. His publications include The Moral Wager: Evolution and Contract (2007), Critical Reflection, with Nebojsa Kujundzic (2005) and articles on contractarianism in Canadian, American, Australian and European philosophy journals. John T. Sanders is Professor of Philosophy at the Rochester Institute of Technology. He is the author of The Ethical Argument Against Government (1980), co-editor of volumes 1–5 of The Philosopher’s Annual, co-editor of Debating the State of Philosophy: Habermas, Rorty, and Kolakowski (1996) and co-editor (with Jan Narveson) of For and Against the State (1996). Chris Tucker received a PhD from the University of Waterloo in 2001. He has since taught at a variety of institutions, including Duke University (North Carolina), Queen’s University (Kingston, Ontario) and Saint Mary’s University (Halifax, Nova Scotia). He has published numerous articles on contractarian theory, and co-edited (with Susan Dimock) Applied Ethics: Reflective Moral Reasoning (2004). He is currently an LLB candidate at Osgoode Hall Law School, Toronto. Peter Vallentyne is Florence G. Kline Chair in Philosophy at the University of Missouri-Columbia. He is the editor of the Equality and Justice Series (2003), The Origins of Left Libertarianism: An Anthology of Historical Writings and Left Libertarianism and Its Critics: The Contemporary Debate (both with Hillel Steiner, 2000) and Contractarianism and Rational Choice: Essays on David

Notes on the Contributors

ix

Gauthier’s Morals by Agreement (1991), and has 60 published articles in journals and collections. Paul Viminitz is an Assistant Professor at the University of Lethbridge in Alberta and his publications include Artificial Prudence (forthcoming), ‘A Defence of Terrorism’, in Frederick R. Adams (ed.), Ethical Issues for the Twenty-First Century (2005), ‘Ameliorating Computational Exhaustion in Artificial Prudence’, in K. Peacock and A. Divine (eds), Mistakes of Reason: Essays in Honour of John Woods (2005) and ‘A Proof that Libertarianism is Either False or Banal’, Journal of Value Inquiry, 34/2–3, (2000). Sheldon Wein is a Professor of Philosophy and International Development Studies at Saint Mary’s University in Halifax, Nova Scotia. His articles on ethical issues in international development, philosophy of law and political theory have appeared in Journal of Social Philosophy, Simulation and Games, Journal of Value Inquiry, Philosophia, Dialogue, Studies in Social and Political Theory, International Journal of Social Economics and Journal of Philosophical Research, as well as collections by C. F. Delaney (The Libertarian–Communitarianism Debate) and John Hooker (Issues in International Corporate Responsibility).

This page intentionally left blank

Acknowledgments This collection of essays is dedicated to Jan Narveson in recognition of his significant and lasting contribution to moral and political philosophy. Certainly there is no philosopher to whom I owe more, and I was pleasantly surprised to discover the degree of appreciation I have toward Jan is equally shared by all the contributors gathered here. Despite being critical of at least some part of Narveson’s libertarian philosophy, every single contributor was thrilled that a collection in honour of Jan Narveson was being undertaken. It made my job so much easier. The original idea for a Festschrift for Jan Narveson was spearheaded by Grant Brown. The first step is always the most difficult and had he not initiated the project, it may well have never been undertaken. All contributors share with me my gratitude to Grant. Nor could I have undertaken this project without Louis Groarke’s support. I am greatly in debt to him for his advice, assistance and encouragement throughout the process. In the course of editing this collection, I have had the pleasure of receiving countless expressions of gratitude and encomiums toward Jan. Well-wishers included Gerry Cohen, David Gauthier, Paul Groarke, Michael Kubara, Chris MacDonald, Darryl Pullman and David Schmidtz. Gratefully appreciated financial support was provided by the Dean of Arts at the University of Waterloo and the Canadian Association for Publishing in Philosophy. I also wish to thank Paul Coulam at Ashgate Publishing, who shared – perhaps even exceeded if possible – our enthusiasm for the project. To Pat and Emma, to all of the above, and especially to Jan – thank you. MM

xi

This page intentionally left blank

Biographical Note on Jan Narveson Jan Narveson was born in Minnesota, USA, and educated at the University of Chicago (BA in Political Science, 1955, and in Philosophy, 1956). He earned his PhD at Harvard (1961) with a year at Oxford (1959–60) on a travelling Fellowship. He has taught at the University of New Hampshire (1961–63), and since then at the University of Waterloo. Professor Narveson was Visiting Professor at Johns Hopkins (1967), Stanford (1968) and Calgary (1976), and a Visiting Research Scholar at the Center for Philosophy and Public Affairs at Bowling Green State University, Ohio (1990). His publications include Morality and Utility (1967), The Libertarian Idea (1988; reprinted 2001), Moral Matters (1993; 2nd, expanded edition, 1999), Respecting Persons in Theory and Practice: Essays on Moral and Political Philosophy (2002), and he is the co-author, with Marilyn Friedman, of Political Correctness: For and Against (1995). Jan is also the editor of the anthology, Moral Issues (1983) and the co-editor, with John T. Sanders, of For and Against the State (1996). Over 200 of his articles and reviews have appeared in Analysis, Australasian Journal of Philosophy, Ethics, Mind, Journal of Philosophy, The Monist, Dialogue, Canadian Journal of Philosophy and others, and many of his papers have appeared in numerous collections. Jan Narveson has served on the editorial boards of Dialogue, Ethics, Philosophical Research Archives, The Journal of Social Philosophy, Social Philosophy and Policy, The Journal of Value Inquiry, The International Journal of Applied Philosophy, Canadian Journal of Philosophy and Public Affairs Quarterly. He serves or has served as referee for many other journals and for several university presses. He also frequently presents papers, talks and commentaries at workshops, conferences and colloquia in North America, the United Kingdom and occasionally elsewhere. Jan’s primary non-philosophical activity concerns classical music. He is the founder and president of the Kitchener-Waterloo Chamber Music Society, was on the boards of directors of several local musical organizations, including the famed Kitchener-Waterloo Symphony Orchestra, writes a weekly column of musical criticism for the University of Waterloo Gazette and presents the commentary on a weekly radio programme devoted to chamber music. He boasts of having one of Ontario’s largest record collections and has presented illustrated talks on music to local audiences. His many activities on behalf of music in the Kitchener-Waterloo community have brought him a Volunteer Award (1987) and an Honorary Doctorate (DLitt) from Wilfrid Laurier University, Ontario. He is listed in Who’s Who, Canadian Who’s Who, Contemporary Authors and Who’s Who in American Education. In 1989 he was elected to membership in the Royal Society of Canada (Canada’s highest recognition of scholarly achievement). In 2003 Jan Narveson was inducted as an Officer of the Order of Canada. xiii

This page intentionally left blank

Introduction Malcolm Murray

Jan Narveson When not briskly walking along the pavement to keep up with the bagpipers marching in the Oktoberfest parade, Jan Narveson would most often be found with chalk marks in his hair, without shoes on his feet, and in some nondescript, perhaps overworn, sweater. He would be intently typing something on his Apple Macintosh computer while towers of books leaned precariously over him from all angles. Despite Jan’s prolific writing, and despite his tireless second career as a classical music host to the Kitchener-Waterloo community, Jan always had time to discuss a wide range of topics with an interested student or colleague – even after being woken from his nap. (At such times, Jan would roll up his camping mat – his hair mussed, his eyes elsewhere – and I would warn myself, ‘Murray: you’d better make this intrusion worthwhile!’) He was always very generous that way and his interlocutors would leave – if not assuaged or converted – enriched. It is that generosity, I think, which has inspired so many to come forward to help with this present edition in honour of a truly remarkable and energetic thinker. Upon taking my first Jan Narveson course, I cannot say I awoke from a ‘dogmatic’ slumber, whatever that might mean, but I awoke from a slumber nonetheless. I was at the time actively engaged in what may be referred to as speculative philosophy: anything conceivable was open for consideration and reflection. The problem was I did not know that mere speculating was not doing philosophy. Speculation needs to be put to some sort of test, and I had never before witnessed the degree of rigour such a test might undergo – until I took Jan’s courses. He would stand in front of the class in his woollen socks, his face covered in chalkdust, and after writing what seemed an innocuous proposition on the board, he would spend the remaining part of the lecture analysing the various interpretations of each term, including – in fact often citing as the most crucial of terms – the little word ‘the’, or perhaps an ‘a’. Of course, mere interpretation was not the point. This was not an exercise in deconstruction: the point was to assess whether even the most charitable interpretation was something we ought to accept. Jan brought the same rigour to his marking of our papers, of course, and so I was forced to become more careful in what I could say. … I should be careful in saying that, of course, since we cannot infer from the above that I thereby have achieved clarity and

1

2

Liberty, Games and Contracts

precision. Although I might have received a few ‘nicely put’s, in the margins, more often it was ‘Are you sure you mean to say that?’ or even worse, ‘hmmm’, which in context I took to mean ‘This is such a wildly indefensible position that I just don’t know where to begin.’ At the time I did not realize how widely known Jan was until I started to go to international conferences. My being introduced as a student of Jan constituted an admission into an elite society of intellectuals to which I felt I did not belong, but from whom I benefited greatly. Not all of those who welcomed me for being a student of Jan agreed with Jan. One speaker at a conference in Syracuse, New York told me – rather bluntly in fact – that he thought Jan’s position – and by inference Jan himself – was evil. But even he recognized the difficulty in countering Jan’s libertarianism.

The Articles As far as Jan Narveson’s philosophical position goes, I feel myself very much like Plato’s Glaucon or Adeimantus when faced with a theory I hope is wrong but have not the wherewithal to know where it goes astray, and therefore appeal to better minds. Anyone who has met Jan in person will know a critique of libertarianism is certainly not a task one should attempt unprepared. Few are more prepared than the authors gathered in this collection, each having personally known Jan in one capacity or another – either as his student, colleague, co-editor, co-respondent, friend or, in many cases, all of the above. That they are well positioned to show what is wrong with the libertarian idea does not mean they succeed. In his commentary at the end of this collection, Jan will take pleasure in pointing out just how they do fail. Jan Narveson defends libertarian politics on the basis of a Hobbesian individualist-based contractarian ethic. Put simply, rational agents in a pre-moral choice situation will agree to do only what is necessary to achieve peace compatible with a maximum amount of liberty to pursue their individual aims. That is, they will recognize the need to endorse and adhere to a rule restricting harms. From such beginnings, they will be unable to endorse rules requiring helps. Given this structure, it is possible to criticize Narveson’s position in one of three ways: (i) to reject contractarian moral theory; (ii) to reject the link between contractarianism and libertarianism or (iii) to deny that negative rights alone – particularly those cashed out in terms of property rights – is the best protector of human liberties. The papers in this collection are sorted under these three headings. This collection ends with a commentary on the papers by Jan Narveson. I. Papers Against Contractarian Ethics in General Many accept Jan Narveson’s argument that Hobbesian contractarianism commits one to libertarianism, but doubt that contractarianism can be defended. Of these, some contend that if libertarianism results from contractarianism, this counts as a

Introduction

3

reductio ad absurdum against contractarianism. Others believe libertarianism is better defended without appealing to contractarian foundations. So a first task in the Narveson project is to see if a Hobbesian brand of contractarianism can be defended. Bond, Machan, Sanders and Groarke think not. Tucker suggests a revisionist approach of Hobbesian contractarianism can be defended in a way that leads much more easily to a Narvesonian libertarianism. Hobbesian contractarians are interested in a reductionist project. They wish to know what grounds morality, and this cannot be answered by appeal to moral concepts. Non-moral candidates tend to be variously represented as rationality, prudence or self-interest. So long as we define rationality as a tool to further prudence, and prudence is defined as one’s enlightened or considered self-interest, the differences here will not matter. Since rationality is defined as maximizing selfinterest, morality is to be understood as a label referring to a subset of normative rational choice.1 Any moral duty conceived as a constraint on self-interest will be motivating only so long as it is itself predicated on self-interest. We are bound by our interest in so being bound, and absent that appeal to self-interest is absent any binding appeal. Can such a minimalist account of the origin of morality lead us to what we take morality to be? In ‘Some Remarks on the Foundation of Libertarianism’, E. J. Bond highlights how, for Narveson, individual goods must always trump common goods, since this will better allow individuals to pursue their own projects free from interference from others. Narveson’s contractarian account has it that each rational individual will see the reason to uphold others’ like rights because, otherwise, the lack of cooperative peace is disastrous for every single individual. Thus Narveson is able to argue for a social cohesion that nevertheless keeps his essential individualism intact. No one need care about anyone else’s well-being; it is only necessary that the cooperation of everyone, each for his/her own personal benefit, is ensured. If there is a common good, it can only be the contract itself. Bond’s critique challenges such a narrow conception by focussing on the existence of noncontractual common goods that are an essential part of the good of every individual. Living together in peace, amity, good will and mutual support, with respect for individual autonomy, is the best possible life for each and all. Such a life cannot be vouchsafed by Narvesonian contractarianism. In ‘Contracting Justice’, John T. Sanders argues that contractarian foundations cannot provide the neutral foundation that Narveson desires. Simply put, contractarianism cannot provide reasons to everyone for accepting the social contract, no matter what their personal values or philosophy of life may be, and thus cannot achieve the universalizability condition Narveson deems crucial. For Sanders, as with Bond, contractarianism is too constricting to provide full foundations for justice and morality as a whole. Sanders also notes how the Narveson contractarian programme is too broad. Since Narveson defends his contractarian morality in terms of mutual benefit, then we would have to say any social cohesion that happens to produce mutual benefits is moral. But a colony of ants satisfies that condition. As Sanders remarks, if the foundation of morality is to be found in what it would be reasonable to agree to, rather than what is actually

4

Liberty, Games and Contracts

agreed upon, then it makes no difference that ants and people differ in their respective capacities to follow the dictates of reason. Tibor R. Machan argues that contractarianism begs the question. For him, the integrity and force of contracts (or compacts) rests on the prior commitment or inclination of agents to live by their word. That is, norms resulting from promises have moral force only if promising, independently of any agreement making it binding, has moral force. Contractarians have fallen into a loop. Conversely, there is some inborn motive or drive that impels those party to a compact to live by it. Then the norms will be adhered to not because of the agreement to abide by them, but because such an agreement is something the parties have to make – for example to preserve their lives. In this sense, we do not say ‘x is moral because it is agreed to’, but ‘we agree to x because it is moral’. Contractarian talk thereby fails to capture the foundations of morality. Leo Groarke shares Bond’s, Sanders’s and Machan’s worries in ‘Does Scepticism Beget Libertarianism? A Response to Narveson on Reason, Morality and Politics’. For Groarke, the strength of Narveson’s position is his willingness to recognize and deal with scepticism. This does not mean Groarke thinks Narveson succeeds, however. He concludes that Jan Narveson does not definitively answer the sceptical problem, and that there are other answers to it that are as plausible as the one he has proposed. Apart from challenging contractarianism as a basis for any moral theory, Groarke also casts doubt on whether – even if contractarianism were a good starting point – contractarianism would lead to libertarianism, a prelude to the papers in Part II. Chris Tucker is more sympathetic to contractarianism than our previous authors, but even he recognizes revisions to the basic Hobbesian doctrine are needed. In ‘Contractarian Education’, Tucker argues that the compliance problem – how to rationally abide by our rational agreements – cannot be solved without begging the question – just as Machan and Groarke have noted. Rather than abandon contractarianism, Tucker argues for re-conceptualizing the compliance problem. For Tucker, rational agents need not worry about full compliance: the attainment of an efficient level of compliance is sufficient. Efficient compliance is rational so long as we understand contractarianism as a theory about how we (the current generation) wish to shape the preferences and beliefs (and therefore the actions) of future generations. II. Papers Disparaging the Link from Contractarianism to Libertarianism While the papers in Part I disparaged the attempt of grounding anything on Hobbesian contractarianism, this second batch of papers is written by those largely in favour of contractarianism as a moral theory, but they doubt that it leads to libertarianism. In game theoretic terms, if we are to cooperate without exploitation, the choice between unilateral cooperation and unilateral defection is a false dichotomy. I might adopt a more sensitive disposition; a disposition that directly depends on the disposition of the others with whom I engage. So long as I can internalize this constrained maximizing position, I can assure myself of doing at least as well as

Introduction

5

straightforward maximization of self-interest; and so long as I can interact with others prone to cooperate, I can do better. Given the social constraints in which I operate, the best way to maximize my self-interest is to adopt a disposition prone to comply only with other compliers.2 To reap the benefit of this advantage, I must accept constraints on my temptation to resort back to straightforward maximization. Internalizing constraints is the contractarian link to morality.3 What libertarians who base their arguments on contractarianism need to argue is that a willingness to abide by non-libertarian policies is not something that rational agents in ex ante bargaining positions would endorse. Can such a claim be supported? Susan Dimock, in her paper ‘The Value of Values: The Importance of Autonomy in Contractarian Reasoning’, may be read as offering a response to Bond and Sanders as well as to Narveson. For Dimock, Narveson is led to the erroneous conclusion that libertarianism follows from contractarianism because he employs an inadequate conception of ‘value’ at the heart of the contractarian justification for normative constraint. Contractarians justify a set of moral rules on the basis of their service to the values of individuals, but the resulting theory depends for its normativity on the strength of the theory of values upon which it is built. Dimock argues that if values are to play the role assigned to them in contractarian accounts then they cannot include all desires or ends an agent might have, but must instead be restricted to autonomous desires. If so, contractarian methodology will derive very different substantive conclusions than Narveson’s libertarianism. In ‘Simple Games and Complex Ethics’, Peter Danielson argues that shifting focus from the prisoner’s dilemma to the related game of chicken provides a clear situation where libertarian moralities based on negative rights is not sufficient. Narveson emphasizes a commitment to understanding the mechanism(s) of ethics, on the empirical or commonsense morality, the role of sanctioning in moral rules, and the relevance of iterated games to ethics. Nonetheless, when we frame these issues using the tools of game theory we find that his rather intuitive moves from rational choice to a morality of cooperation fall short. Danielson concludes that while game theory and the related tool of evolutionary modelling do not provide a foundation for simple cooperative morality, they do yield insight into a deeper account of ethics, something Danielson examines in relation to blood donation. My contribution is entitled ‘Why Contractarians Are Not Libertarians ... Evolutionarily Speaking’. In a similar move to that employed by Danielson, I argue that if we alter the game from the prisoner’s dilemma to evolutionary accounts of the ultimatum game, contractarian and libertarian recommendations part ways. The import of this conclusion rests on our rejecting that a rational justification for a universal agreement is possible – something I argue is not inconsistent with contractarian reasoning. Paul Viminitz’s contribution is called ‘Getting the Baseline Right’. Viminitz charges Narveson with begging the question in selecting a non-harms baseline as the contractarian starting point, and so shares some sentiments with Machan, Groarke and Tucker. For Viminitz, being more rigorous about the initial baseline does not (necessarily) mean that contractarianism is the wrong starting point, but that – with Dimock, Danielson and Murray – it cannot lead us to libertarian policies.

6

Liberty, Games and Contracts

III. Papers Focussing on the Relation between Property and Liberty If it is true that whatever our goals and interests and preferences are, these are more likely to be fulfilled under libertarian systems than any other, we would have to be irrational not to buy into libertarianism. Objections follow from noting that we are interested in more than material possessions. Security, for example, may be better purchased under a more egalitarian scheme than a libertarian one.4 Admittedly, it is often only poor imagination that prevents people from seeing how cost-efficient privatized goods can be offered even for matters such as security and welfare, let alone education, parks, roads and lighthouses. Of course, privatized goods entail some people are excluded from these goods. Naysayers point out that those excluded are precisely the ones we had intended to target: a complete failure. When we are dealing with privatized welfare (or welfare insurance schemes), we are dealing with (among others) individuals who would opt out of the system for no other reason than that they cannot afford to pay, so the clever ways of avoiding free-riders with which libertarians are wont to serenade us is not to the point. Still, libertarians distinguish being at liberty to x and deserving x. Too often critics equivocate on this distinction. Gogol’s being at liberty to buy an overcoat is not curtailed if he cannot afford an overcoat. Even if one believes Gogol ought not be left out in the cold, one cannot pretend a duty to give Gogol an overcoat arises from an analysis of liberty. A defence of a libertarian state, therefore, must make the case that liberty can never (morally) be overridden. Can such a case be made? The papers gathered in Part III focus on the relation between liberty and property that is central to libertarian thought. In ‘Liberty, Property and the Libertarian Idea’, Ann Levey argues that Narveson is mistaken in the connection that he draws between liberty and private property. Property, whether private or otherwise, is a system of restrictions and permissions that control access to and use of external goods. Property controls what people can do by controlling how they get to use the world to act; it is an institution that distributes liberty of access. Such control may be on the whole a morally good thing. It may be the case that private property is the institution that does the best job of protecting a wide range of liberties. She even concedes that rational contractors may agree to a system of private property, at least suitably constrained, since a system of private property helps to make available an attractive range of options for many people, possibly better than do any other institutions for controlling access to external goods. Levey denies only that there is any intimate connection between a general right to liberty and private property such that recognition of a right to liberty requires recognition of private property. The libertarian argument that Narveson outlines mistakes a way of distributing liberty for a requirement of liberty. Duncan MacIntosh takes the problem of children to heart in ‘Who Owns Me: Me or My Mother? How to Escape Okin’s Problem for Nozick’s and Narveson’s Theory of Entitlement’. Libertarians defend the property rights of persons: not only the right to own themselves, but the right to acquire ownership in the things they make. But since offspring are or become persons, the question arises whether this

Introduction

7

entails giving them a right in themselves that overrides the property claims of their mothers who made them. Susan Okin claims that something has to give: either people do not automatically own themselves, or else the theory that people acquire the right to something by making it is not true with perfect generality. To resolve this prima facie conflict, MacIntosh suggests that libertarianism cannot be grounded in a right to property; it must be grounded in the right to liberty. Liberty rights in turn must be grounded in individual capacities to exercise liberty claims. That is, coercible constraints on people’s liberty can only be justified if needed to optimize everyone’s individual liberty. In ‘On Original Appropriation’, Peter Vallentyne makes a distinction between first- and second-order status of ownership. The first-order status of an unowned thing is clear: no one has any claim-rights concerning its use (that is, everyone is at liberty to use it). That leaves open, however, its second-order status (the conditions under which the first-order status can change). One compatible position is that individuals can acquire claim-rights over first-order unowned things merely by staking a claim (or related activities). Another is that this is so only when the Lockean Proviso is satisfied. There are infinitely many other possibilities as well (for example, whenever the king so dictates). For Vallentyne, Jan Narveson is thus mistaken if he claims that his libertarian position follows from the mere fact that initially no one has any claim-rights over natural resources. First-order conditions of morality are compatible with all kinds of second-order conditions. While Danielson and Murray argue that evolutionary game theory does not support libertarian policies as a general rule, Grant Brown make the opposite case. In ‘Morality as an Evolutionarily Stable Strategy: A Naturalistic Account of Libertarianism’, Brown defends libertarian policies as an effective evolutionarily stable strategy. He particularly takes to task the standard critique that libertarian policies based on individual negotiation will merely mean that one large segment of society will be unable to afford any rights, since they have nothing with which to bargain. Among Brown’s responses (another employs the paradox of the heap to wonderful effect) is to show that non-libertarian policies, such as egalitarianism, are not evolutionarily stable. Sheldon Wein’s piece, ‘Reconciling Radicals: Market Contractarianism and Fundamentalist Utilitarianism’, reminds us that Jan Narveson began his philosophical career as a utilitarian, a position Narveson subsequently renounced. For Wein, the chasm between the two – or at least between ‘fundamentalist utilitarianism’ and ‘market contractarianism’ – is not as wide as is commonly thought. The benefit of re-examining the connection between utilitarianism and contractarianism is to highlight how libertarian social policies fail to provide the requisite motivation to be moral. Jan Narveson provides the final word. In ‘Social Contract, Game Theory and Liberty: Responding to My Critics’, Jan responds to each article and, by so doing, defends the contractarian moral theory, explains how contractarianism leads to libertarian political theory and clarifies the critical role of both liberty and property. The reader is thus treated to a brilliant synopsis of Jan Narveson’s libertarian idea.

Liberty, Games and Contracts

8

Notes 1 2 3

4

David Gauthier, Morals by Agreement (Oxford, 1986) 4. John Rawls, A Theory of Justice (Cambridge, MA, 1971) 16. Perhaps this should read ‘... other reciprocal compliers’, See Peter Danielson, Artificial Morality (London, 1992) 88–90. Ken Binmore misrepresents Gauthier as following Locke’s natural law justification for morality (Ken Binmore, Game Theory and the Social Contract: Vol. 1: Playing Fair, Cambridge, MA, 1998, 14). Gauthier’s use of the Lockean Proviso has nothing to do with that (Gauthier, Morals by Agreement, 222). On the other hand, the content of an appropriate agreement for Gauthier is tied with the antecedent constraints of the minimax relative concession, and it is not at all obvious that agreements by rational persons would necessarily invoke such demands without merely presuming equal starting positions. On another note, Skyrms suggests that if we follow Rousseau rather than Hobbes, we can give up looking for reasons to be moral and instead investigate the evolution of morality (Brian Skyrms, Evolution of the Social Contract, Cambridge, 1996, ix). If so, we should not say, ‘adopting a moral disposition is rational’, but that ‘adopting a moral disposition has evolutionary stability’. I think Skyrms reads more into Rousseau than warranted, but Hume may certainly be read that way. A popular complaint is to say that liberty is fine so long as other basic needs are secured: for example, G. A. Cohen, Self-Ownership, Freedom and Equality (Cambridge, 1995), 56–9; or Kai Nielsen, Equality and Liberty (Lanham, MD, 1985), 302–6. The idea is that the rich have more liberty than the poor, and that liberty is only appreciated once welfare is secure. Such talk misunderstands the liberty of which libertarians speak. There is no necessary difference in liberty between rich and poor, let alone the starving compared to the sated. Liberty is understood simply as the absence of human constraint, not the absence of any impediment to one’s desires. Although such absent-human-constraint liberty is a good-in-itself, it hardly means anyone appreciates it in-itself let alone even acknowledges it while we have it. We note it typically when it is violated. Having no food is certainly unpleasant, but not a problem with liberty in the narrow sense that libertarians mean.

PART ONE CONTRACTARIANISM

This page intentionally left blank

Chapter 1

Some Remarks on the Foundations of Libertarianism1 E. J. Bond

As a non-Cartesian in epistemology, I am happy to grant that there are no certain and indubitable truths that ground all knowledge, and that would have to include moral knowledge. Nevertheless, with Narveson, I have no difficulty at all with the notion of a rational foundation for ethics or morality, for this is an entirely different matter. After all, morality is supposed to provide us with reasons or grounds for making decisions, and we only have reasons for action or abstention if there is something to be gained or some loss to be prevented – this is what it means for there to be a reason. To provide a foundation for morality, then, would simply consist in showing what is to be gained, or what loss is to be avoided, by what is recognizably describable as being moral. (To observe Narveson’s distinction, I am referring here to social as distinct from personal morality.2) And since morality is for everybody, what reason could this be other than that, in some way, it makes everybody better off, or that everybody stands to gain by it? If we can give an account of morality that makes it plain why this is so, then we will have provided morality with a foundation. And this is exactly what Narveson claims, with some aid from Gauthier, Hobbes and others, to have done. I would add that if we cannot show how morality relates to human well-being, then the whole thing is a sham and we have no reason for paying any attention to it. A morality that cannot be justified or rationally defended is one that fails, in its own terms, to provide us with real reasons for action or abstention, and is therefore worthless and possibly harmful. In this Narveson and I are, I think, in complete accord. We are fully in accord in another important matter as well. Each of us must lead our own life in the pursuit of what we find worthwhile, as free as possible from encumbrances. If we cannot pursue our own projects, to use Bernard Williams’s term, but are under continuous constraint, then life is not worth living. That is what is wrong with all moral theories, including most forms of consequentialism, some forms of intuitionism and what Narveson calls the ‘theoretically indispensable’ version of Rawls’s veil of ignorance that would require us to be perpetually impersonal or impartial.3 The most grotesque account of this sort is Hare’s pool of selves, or universal agent, on behalf of which each of us must act at all times – that is, we must act as if we were everybody!4 Equally absurd is Nagel’s ‘view from nowhere’, where(?) we stand aside from the whole world, including ourselves, each 11

12

Liberty, Games and Contracts

of us viewing her- or himself as no different from any other; thus at times requiring us to act as if we were nobody!5 Even those who allow exceptions to the stance of impartial benevolence tend to see it as more noble if we could manage it. But it is not more noble; it is false because it is unintelligible, as is the notion of the universal agent. Each of us is himself or herself, living his or her own life according to his or her own lights. This is unavoidably true (because its denial is unintelligible), not in the least to be deplored (supposing one could deplore what is otherwise inconceivable) and not a trifling matter. I cannot live your life for you, nor can you live mine for me. Deontic morality – that is, the morality of requirement or constraint – is analogous to law, which is why it is often so called. That is, it binds us to do some things and not to do others, and otherwise leaves us alone to do as we please. Neither the law nor deontic morality, though they require universal conformity or non-violation, constitutes a general guide to conduct. To suppose that deontic morality does constitute a general guide to conduct is the gross error upon which most forms of consequentialism rest. Kant, however, did not make this mistake. What he actually said was that we can do anything we please as long as it is licit, that is, not non-universalizable; what is non-universalizable being in violation of the moral law. We are quite at liberty to follow our natural desires provided the maxims (that is, motives) of our acts are universalizable. It is only that there is no moral worth in doing so. Our acts only have that when we constrain ourselves to act or abstain from acting as the law requires and for no reason other than that the law requires it. This is the pure moral motive.6 According to many, if not most, moral theorists, Kant being a notable exception, there are only two deontic categories: right – by which is understood obligatory – and wrong. But an act may be neither obligatory nor wrong, in which case it is permissible to do or not to do as one pleases, without transgression either way. And without the category of the permissible, excluded by so many moralists, life would be scarcely worth living. (Of course this is not to say that we might not have other, non-deontic reasons for doing one thing rather than another.) These considerations have already given us a good push in the direction of a morally grounded libertarianism. My difficulty is not so much with libertarianism as such – I am a libertarian of sorts myself, strongly favouring voluntary social arrangements over authoritarian ones, though of the left anarchist rather than the Lockean, propertarian, free-market persuasion. My difficulty is rather with the contractarian foundation offered for it. Mind you, the position is very neat and powerful. It is to anyone’s advantage to enter into an agreement with others to cooperate rather than grab by force or fraud, which is risky to say the least, since it is difficult to second-guess others, and if one guesses wrong one is worse off than one would be otherwise. Better to settle for second best (we both get something) than to risk going for the boodle (I get everything; you get nothing), the second-best outcome being assured if there is agreement and a settled disposition to cooperate, non-cooperators being ostracized and otherwise deprived of cooperators’ benefits. This is supposed to have the libertarian upshot that I have the right to use my personal resources in any way I see fit provided I do not violate the similar right of others, or in Narveson’s brief

Foundations of Libertarianism

13

formulation: ‘Our sole basic duty is to refrain from utilizing the fundamental resources of others without their consent; and those resources include, at a minimum, the bodies and minds of those others.’7 And: ‘What is needed is to show that the concern we must by nature have with our own values, virtually whatever they may be, is such as to underwrite a social principle of allowing each to pursue that person’s possibly very different ones – and that no other moral principles can take precedence to this one.’8 Contractarianism has the advantage of not surreptitiously presupposing any moral values and is able (‘hopes’) ‘to generate moral principles for societies out of the nonmoral values of individuals’.9 The only people not likely to go for this contractarian–libertarian solution are those who are incurably bent on winning battles, who set an extremely high value on what Hobbes called ‘glory’, and who are only happy when they win and others lose. These people, Narveson points out, are not in a prisoner’s dilemma, where cooperation is a possible choice, but a zero-sum game where the only satisfactory solution is coming out ahead of the other. Narveson is referring specifically to military fighting, victory and defeat, but what he says obviously applies to other kinds of fighting and devotion to doing down one’s rivals and opponents. It is not only the would-be military conqueror who is ‘launched on a career of stark competition’.10 Glory need not be confined to military glory. For many, adversarial competition and coming out on top are valued more highly than anything else. One is reminded of the remark attributed to Gore Vidal: ‘Success is not enough; others must fail.’ But, as Narveson says: ‘At least one of them is bound to fail [the loser]. And the victor will not long be happy with his success, will seek out a new opponent or have one forced on him, and eventually he, too, will fail.’11 Thus the man or woman incurably bent on fighting to win is doomed. Perhaps if they knew this, they would value glory less, perhaps even agree to cooperate? Contractarianism seems a very acceptable foundation indeed, a very neat account of social morality grounded in reason. Yet something in me is repulsed by any account of morality that is grounded in an iterated prisoner’s dilemma, which must be understood as settling, on a permanent basis, for second best, since going for the very best – I get everything; you get nothing – is too risky. Those of the Thrasymachus or G. Gordon Liddy12 persuasion would take this as too wussy for words, adopting as their motto ‘Kick butt and go for the boodle!’ or, more politely, ‘Go for it by whatever means you have and let the devil take the hindmost!’ But not only is this extra-nasty, it is also almost certain to fail, as both Hobbes (‘vainglory’) and Narveson have pointed out. Still the man of power invincible (Hobbes) or the possessor of Gyges’ ring (Plato) would be a fool to enter into any contract, rather than imposing his or her will on everyone, using people as he or she pleased. Getting the most you can for yourself, no matter what the cost to others, would still be the first choice, if it were possible. It’s just that it isn’t possible; so, as rational beings, we agree to settle for second best, something that can be assured by adherence to the contract and the settled disposition so to adhere. It is simply the best (the only) way to secure your own advantage, not the best you could have if you were able to take advantage of other people and use them. We enter into the contract in order to protect ourselves from the evils that we

14

Liberty, Games and Contracts

might otherwise suffer at the hands of others. It would be better for us if we were able to get everything we wanted for ourselves, even at the expense of others. Complete domination of others, making them the servants of our will, would still be the best result of all. It is just that it would not be rational to attempt it. In some conflicts we would be bound to lose. If it is permissible to quote myself let me do so: Contractarianism does not require respect and if it requires recognition of the rights of others, that is not because others matter, but rather that cooperation pays even if it doesn’t yield the best possible result, i.e. I gain everything. It is a theory dominated by Prisoner’s Dilemma and, being wholly egoistic, is as individualist as it is possible to be. The motto is: I’ll cooperate with you if you cooperate with me, which is for our mutual (not common or shared) benefit. I don’t give a damn about you and you don’t give a damn about me, but it’s in my interest to make a deal with you and I can trust you because I know that you know that it’s in your interest too. It might be said that, on this (Hobbesian) view, one real social relation is recognized, that of competitor, adversary, or rival. We cannot be friends, but let us agree to put down our swords. That way each of us is freer to get things for himself.13

The picture we have here is of individuals existing in isolation, each one motivated only by self-interest, not caring or not necessarily caring, for anyone other than himself or herself. It is simply that each has agreed, for his or her own advantage, not to use force or fraud in his/her competition for the goods of the world, to go by the rules and not to invade each other’s territory. Though they cooperate for their mutual advantage, they are still basically at odds, and another person’s good is still no part of one’s own. I would appropriate his goods and use his mind and body for my own advantage if I could do it and not suffer for it. I would not have any real concern for him or her. I must take others into account but only to protect myself. There is no common or shared good other than the social contract itself, which permits us to go about our own pursuits with the minimum of attention to the concerns of others. We will agree not to use them or their resources without their consent, on the proviso that they will refrain from using us or ours. We will not use force or fraud, on the understanding that neither will they. Thus we will each go our own way, living our individual lives and pursuing our own projects, with a measure of stability guaranteed by the self-imposed constraint of other cooperators and stiff sanctions for evaders or would-be evaders. The need to impose such sanctions would be the chief, if not the only, justification for the coercive authority of the state. The Preface to The Libertarian Idea begins by defining libertarianism, as he and others of his persuasion understand it, as ‘the moral and political outlook holding that individual liberty is the only proper concern of coercive social institutions’.14 Liberty, on the contractarian–libertarian view, would be protected by sanctions aimed at preventing violations of the contract. What is assumed here is that individuals are best off if they can grab everything they want for themselves, but that a certain amount of self-sacrifice or self-denial, for the sake of cooperation with others who are doing the same, is necessary, since in a world of unrestrained competition, which would put them in constant adversarial conflict, things would go very badly for them (and everyone else). It is

Foundations of Libertarianism

15

also assumed that a person can be happy, or possess well-being, simply on their own, regardless of the well-being of others in their society, that their well-being can be self-contained, not inherently dependent on the well-being of others, or on people living (for the most part) happily and contentedly together. But I would argue that someone’s happiness or well-being is absolutely dependent upon good social relations and a society in which people are genuinely concerned with each other and with their common or shared goods – food and water to take the most obvious examples. I would also claim that an individual’s happiness or well-being is not possible without genuine goodwill, fellow feeling and desire for the good of others. Going it alone, getting the most one can for oneself, is not the best way to live; and that would include going it alone while respecting the rights of others on the condition that they respect ours. Mutual concern and concern for the common or shared good – not just agreement on constraints to prevent the suffering of evils at the hands of others – are essential ingredients of the good life. If this is true, then being decent, considerate and respectful to one another, which is what morality essentially requires, does not involve any sacrifice or self-denial. It is an essential element of the best life for anyone and everyone. It is not second best, and prisoner’s dilemma has nothing to do with the case. Nevertheless I am grateful that contractarianism exists, for it is better than moral nihilism or Nietzschean perfectionism. Upon meeting, a number of years ago, a distinguished philosopher of the contractarian persuasion with whom I was previously acquainted, I remarked: ‘I’m glad that people like you exist because it gives you reason to be decent to me even if you don’t give a damn about me.’ His reply: ‘Yeah – as long as you’re a cooperator.’ So far just assertion. What about argument? What about grounding? Well, here’s a great big empirical claim. People just are happier when they live in conditions of friendliness, trust, goodwill and mutual support, and that is what it is the business of morality to create and maintain. Narrow self-seeking without these things simply does not bring satisfaction and leads eventually to misery and loneliness. Morality is grounded in eudaimonía, both of individuals and society, these being indissolubly connected – no man or woman or child an island. As I have stated elsewhere,15 the central problem for ethics is how to reconcile the unavoidable separateness of persons with their inherently social nature, their dependence on other people, and the fact that they live together in and share a common world. And as I further stated,16 what is morally good is whatever it is in conduct and character that brings people together in amity and goodwill, while allowing them to live fulfilling and self-directed individual lives; what is morally evil is whatever it is in conduct and character that divides people and sets them against one another, while preventing or hindering them from living rich and fulfilling self-directed individual lives. If this is so, moral goodness does not require selflessness, self-denial or self-sacrifice. Because the common good is an essential ingredient in the good of each and every individual, to choose morality is not the same as to draw the line at second best. On the contrary it is contractarianism that is second best. The moral life, properly understood, is the best life for each and all. And what about the state? Voluntary association for the

Liberty, Games and Contracts

16

common good is to be preferred over any kind of coercive authority. There should be no need for people to elect or appoint or accept an authority to rule over them. Ideally people in association, each respecting the autonomy and individuality of every other, should be able to rule themselves. This would not only benefit society materially and spiritually, it would be the best possible safeguard of individual liberty.

Notes 1

2 3 4 5 6

7 8 9 10 11

This piece was written for and presented as a contribution to the seminar on the work of Jan Narveson, held at the annual meeting of the Ontario Philosophical Society at the University of Waterloo on 30 October 1993. It was edited and revised, though with no changes of substance, in October 2002, again in August 2003, and edited once more in May 2005. Jan Narveson, The Libertarian Idea (Philadelphia, 1988), 123–7. Narveson, 132f. See R. M. Hare, Moral Thinking (Oxford, 1981), 226f. See Thomas Nagel, The View from Nowhere (Oxford, 1986), ch. IX, passim. (Nagel’s view is somewhat more complex than this remark implies.) Consider the following three quotations from the Grundlegung zur Metaphysik der Sitten, all in the Beck translation (Immanuel Kant, Critique of Practical Reason and Other Writings in Moral Philosophy, trans. and ed. Lewis White Beck (Chicago, 1949)): ‘This principle of humanity and of every rational creature as an end in itself [the moral law] is the supreme limiting condition on freedom of the actions of each man’ (Grundlegung, Prussian Academy Edition of Kant’s Works, vol. IV, 430–1, Beck, 89 [emphasis mine]). ‘Rational nature is distinguished from others in that it proposes an end to itself … an independent end and thus merely negative. It is that which must be never acted against’ (Grundlegung, 437, Beck, 94 [emphasis mine]). ‘The subject of ends, i.e. the rational being itself, must be made the basis of all maxims of actions and thus be treated never as a mere means but as the supreme limiting condition in the use of all means, i.e. as an end at the same time’ (Grundlegung, 438, Beck, 95 [emphasis mine]). Also consider these two quotations from the general introduction to Die Metaphysik der Sitten (in Part I, [Kant, The Metaphysical Elements of Justice, trans. John Ladd (The Library of Liberal Arts, 1965)]): ‘On this concept of freedom (from a practical point of view), are founded unconditional practical laws which are called moral [author’s emphasis] … According to these categorical imperatives, certain actions are allowed or not allowed, that is, are morally possible or impossible’ (Works, vol. VI, 221, Ladd, 22). ‘An action is allowed (licitum) [author’s emphasis] if it is not opposed to obligation, and this freedom that is not limited by any opposing imperative is called licence [Ger. Befugnis] (facultas moralis). Hence it is obvious what is meant by unallowed (illicitum)’ (vol. VI, 222, Ladd, 23). (I have altered Ladd’s translation of ‘Befugnis’.) Narveson, 165. Narveson, 166. Ibid. Narveson, 180. Ibid.

Foundations of Libertarianism 12

17

G. Gordon Liddy, a former Treasury and FBI agent, was a co-conspirator in the 1972 Watergate scandal that ultimately forced the President of the United States, Richard Nixon, to resign. 13 E. J. Bond, ‘Could There Be a Rationally Grounded Morality?’, Journal of Philosophical Research XV (1990), 40–41. 14 Narveson, ix. 15 E. J. Bond, Ethics and Human Well-being (Oxford, 1996), 211. 16 Bond, Ethics and Human Well-being, 228–9.

This page intentionally left blank

Chapter 2

Contracting Justice1 John T. Sanders

In The Libertarian Idea, Jan Narveson explains his interpretation of contract theory this way: The general idea of this theory is that the principles of morality are (or should be) those principles for directing everyone’s conduct which it is reasonable for everyone to accept. They are the rules that everyone has good reason for wanting everyone to act on, and thus to internalize in himself or herself, and thus to reinforce in the case of everyone.2

It is plain, here, that Narveson believes that social contract is to provide justification – a foundation, in fact3 – for ‘the principles of morality’. The burden assumed in this chapter is to examine how far Narveson has succeeded in making this foundational claim plausible.

Contract As the Foundation of Morality Narveson contends, as noted above, that the idea of social contract provides foundations for all of morality. What morality is, though, while understood in general terms by all, is notoriously difficult to pin down neatly. As a start, one might suggest that the realm of morality is simply the realm of what one should or shouldn’t do. Then the question of ‘foundations’ would be about the reason or reasons underwriting shoulds and shouldn’ts. Before ever beginning the foundational project, though, one vital reservation springs to mind: insofar as one seeks justificatory foundations for morality, the moral realm cannot helpfully be thought to include all shoulds and shouldn’ts. If you want to get to the post office, you should drive back south on the road to Lahaina and turn left at the traffic lights at Wahikuli Beach. If you want a really good margarita, you should use Cointreau instead of Triple-Sec. Surely there is nothing particularly moral about these shoulds except in the broadest possible sense: the sense in which ‘morality’ may be held to encompass any and all purposive activity. Immanuel Kant is famous in this philosophical domain for arguing that the telling feature of moral shoulds is their ‘categorical’ character: according to Kant, there is nothing iffy about them at all. There are some things you should just do (and some you shouldn’t do) regardless of what you may want to

19

20

Liberty, Games and Contracts

achieve. It is not that Kant held that this categorical character attaches magically to some shoulds. Rather, he argued that the properly moral claims have a quasi-logical feature that sets them apart: there is supposed to be a sense – elaborated by Kant – in which it would simply be irrational to violate the demands made by the properly moral shoulds. Narveson distinguishes the moral shoulds from the others in a way that appears initially to be quite different. In the first place, Narveson wants to step away from the idea that moral reasons to act are a different kind of reason from all others: When morality makes demands upon us, the question why we should go along with them, what reason there is to impose them, is always apt. There are no ‘bedrock’ moral reasons: that is an illusion. To say that some act is called for by morality is to say that there are good reasons for insisting that people do that kind of thing. Those reasons are not ‘moral’; they are, simply, reasons.4

Further, the relevant questions are ones of social policy. While acknowledging that morality in general may be held to embrace what he calls ‘personal’ morality,5 Narveson urges that the ‘social’ sense of the term is not only the more usual sense,6 but the one that gets us to the core of the matter: ‘moral requirements can, and often do, “override” individual inclinations to act otherwise.’7 This, indeed, is their primary point. But this is not – as it was with Kant, because moral requirements represent a different, categorical (quasi-logical) sort of demand – to be distinguished from hypothetical imperatives that depend on a person’s interests and inclinations. Rather, Narveson suggests that the moral imperatives are as dependent upon people’s interests as any other. The key similarity between Narveson and Kant, at least in this first comparison, is that both point to the fact that moral requirements can and do override individual inclinations to act otherwise. The big difference between them appears to be that Kant thinks this marks them as belonging to a substantially different kind of requirement, while Narveson disagrees. Narveson’s diagnosis: Kant supposes that people do not simply act in pursuit of their interests. They have, as it were, rationally split personalities. On the one hand, they are concerned to promote their interests. … On the other, however, they are concerned to obey a transcendental law invoking some kind of universalizability requirement known as the Categorical Imperative. … So the two aspects of reason can come into conflict, and when they do, the moral one ought to win. But as Kant depicts the situation, there are insoluble problems. For one thing, in his view ‘inclinations’ belonged in the phenomenal world and were subject to universal causation, and so the Laws of Nature decreed that people must do what their inclinations dictate – which is incompatible with their doing what morality dictates if it ever differs, as Kant thought it often (always?) would. So he is driven to a hopeless and unintelligible doctrine of contra-causal freedom.8

And we certainly wouldn’t want to be driven (or dragged) there. At least at first glance, though, it would seem that things are not radically more coherent for Narveson. He has got people inclined, in their own interest, to do

Contracting Justice

21

things they are not inclined to do. More fairly, perhaps: when people act morally, according to Narveson, they do things that are really in their own interest even though they may seem at first glance not to be. While the reasons for moral behaviour are not some different ‘categorical’ kind of reason, they still override other interests and other considerations. A clue to the real difference between this overriding kind of reason and the kind that gets overridden appears to arise from the fact that the former reasons are of the ‘social’ kind rather than the ‘personal’ kind. The way to understand the importance of this difference and to see why ‘social interests’ of a certain sort might reasonably override many or most ‘personal’ interests, Narveson thinks, is to see that morality is based on a ‘kind of contract’. It cannot be a real contract, after all. As Narveson points out, there is no reason to think that any real people have ever actually entered into an agreement of the envisioned kind. And even if there were an historical ‘social contract’, he clearly sees that such a thing could have no normative bearing on anyone who was not party to it. In what is perhaps the clearest, fullest exposition of his case,9 Narveson takes as his theme a matter raised by David Hume that appears to offer a serious challenge to the ‘contractarian account of morals’. As Hume pointed out in his Inquiry Concerning the Principles of Morals, ‘the observance of promises is itself one of the most considerable parts of justice; and we are not surely bound to keep our word because we have given our word to keep it.’10 Some have taken points of this kind to undermine, rather severely, the contractarian programme. Narveson, like other contract theorists, does not think so.11 But he does agree that some qualification is in order: The contractarian holds that the fundamental principles of morals are in some way the objects of an ‘agreement’, which in turn is the outcome of what is in some sense a ‘bargain’. In trying to make clear and plausible this approach to morals, it is obviously of crucial importance to identify the sense(s) or way(s) in which these things are supposed to be so. For one thing is surely clear enough, and that is that the quotation marks are appropriate: the ‘agreements’ and ‘bargains’ of which the contractarian moral theorist speaks are not your ordinary examples of these types. The Grand Social Contract is, obviously, a model or idealisation of some kind.12

Further, If we are to make social contract theory work, it is clearly essential to understand the ‘agreement’ in some other sense than that of a formal promise or literal contract, be it historical or hypothetical. …[W]e would want the principle of keeping one’s contracts to be itself founded on The Social Contract.13

So the answer to Hume is: the Social Contract is not really a contract. Thus there is no circularity. What is the ‘Grand Social Contract’, then? Following David Gauthier,14 Narveson argues that

22

Liberty, Games and Contracts the ‘Social Contract’ is a ‘contract’ in the following sense: people discern that if they interact under the unrestricted motivation of their various separate interests, then they will do worse than could be done if instead they were to adopt certain restrictions or constraints on the pursuit of those interests. A main context in which such restraint is called for is that of social interaction. Here it is often the case that the unlimited pursuit of advantage by each leaves all worse off than if they had instead adopted restrictions on individual pursuit of utility. Constraining conduct promotes maximization, even though maximization is precisely what it constrains. The secret is in mutuality: each must shape her behaviour by refraining from certain erstwhile maximizing actions, on condition that the other does likewise. It is the necessity of this condition that makes the ‘Social Contract’ a contract. ‘I shall do this, provided he does that’; ‘I shall do that, provided she does this’ – and that goes for both parties.15

That this requires no explicit multi-party agreement, no real negotiation and not even any communication between people, is not only clear from the explanation quoted above. It is reflected also in Narveson’s portrayal of himself as Hobbesian as regards the very essence of morality. For Narveson, as for Hobbes, morality (in the more important ‘non-personal’ or ‘social’ sense, at least) does not even exist for solitary persons. It arises only in social circumstances. Morality gains relevance because of the sorry state of affairs that would prevail in its absence: no industry, no cooperative ventures, no Cubs’ games – nothing that otherwise makes human life not only worth living, but possible in anything but the sparsest terms. The point is not supposed to be that there have been grim historical times out of which morality emerged for practical reasons; rather, the point is that morality serves human interests by making industry and all the rest possible. And the fact that no real explicit agreement is necessary is made yet clearer by frequent Narvesonian references to game theory. To put the matter briefly and without reference to the game theoretic details: persons in real life having to choose between making decisions consistent with moral dictates, on the one hand, and violating those dictates, on the other, have strong reason to choose morally, if others do too, because doing so advances their interests. Back in abbreviated game-theoretic language, this amounts to the observation that the dominant strategy in iterated prisoner’s dilemma settings is tit-for-tat.16 Such results, whether in game theory or moral theory, have nothing at all to do with real decisions in real situations. They derive instead from abstract consideration of relevantly constrained situations. So there are good reasons to behave morally. The ‘foundations’ of morality are to be found in these reasons. These reasons make reference to the interests of persons who are called upon to behave morally, so they are not ‘transcendental’ in the way Narveson portrays Kant’s as being. They may be understood to be modelled on ‘agreement’, even though no explicit agreement is contemplated, because they have a reciprocal character: I agree to behave morally if and only if I have reason to think you are disposed that way too. In game theoretic tit-for-tat, the idea is to start by cooperating in new situations, then continuing with cooperation with those who respond positively while harshly discontinuing cooperation with those who do not. No explicit verbal agreement is envisioned here either; ‘agreement’ is instead tacit and observable only in behaviour. So much for the

Contracting Justice

23

argument, then. But how does the contractarian approach stand in relation to its competitors?

Contractarianism Versus the Alternatives Narveson has regularly been moved, in his many writings on the subject, to defend contractarianism on the ground that none of its competitors can really hold a candle to it in doing what counts: showing how people might be motivated to constrain their behaviour along the lines prescribed by morality. But his specification of exactly what the competition in this enterprise is leaves his reader in some doubt about whether he is dealing with the real competition. Here is what he says, for example, in an essay entitled ‘Contracting for Liberty’: Political and moral theories have been argued for on such bases as ideals of virtue, the will of God, self-evidence, and in one way or another, Nature. Why reject all these? Actually, as we will see, we needn’t reject them all, quite: the contractarian view may be aligned with one kind of ‘natural law’.17

Narveson then goes on, in this particular piece, to explain why these several candidate foundations will not work, and begins to advance positive arguments for ‘the contractarian idea’. In The Libertarian Idea, he devotes an entire chapter to discussion of the inadequacies of ‘intuitionism’ in moral theory by way of preparation for his advocacy of contractarianism. But the real competition, one might think, would be standard ‘foundational’ approaches like utilitarianism, or Kantianism (for want of a better name). We have considered Narveson’s objection to Kant’s approach: the reason certain ‘social’ interests take precedence over many ‘personal’ ones is not that the former belong to a different transcendental or quasilogical category; it is rather that they represent more important, more systematic interests. These ‘social’ interests involve improving the whole framework for pursuit of interest generally. Narveson’s objection to intuitionism, in the end, is fairly simple: where intuitions compete, they are unable to persuade. Since ‘foundations’ for morality are sought that all can agree upon, intuitionism is simply not up to the task. As for utilitarianism, the discussion in The Libertarian Idea suggests that the reason for subordinating it to intuitionism as a potential foundational competitor to contractarianism is that utilitarianism is unwittingly (except in the more witting case of Henry Sidgwick, who Narveson says made the same point) dependent upon a central moral intuition.18 If this were true, it would perhaps legitimate moving directly to a consideration of intuitionism in all its forms rather than worrying about particular versions like utilitarianism. For present purposes, let us allow the arguments against Kantianism and intuitionism to pass. More telling are Narveson’s arguments against utilitarianism, especially since he used to defend utilitarianism himself before he became a contractarian.19

24

Liberty, Games and Contracts

Utilitarianism in its most general form says that right and wrong depend, in the end, on what promotes the general good. Perhaps we should understand ‘the general good’ in terms of ‘happiness’, but perhaps not. Perhaps we can actually add and subtract utils (or some other measure) of happiness, but perhaps not. Maybe we can make cardinal comparisons of different persons’ degrees of happiness, but then maybe not. Perhaps we should each be making a judgment about what promotes the common good in our every decision about what to do. Or perhaps, on the other hand, we (or society, in our behalf) should (on utilitarian grounds) instead guide our lives in accordance with rules that themselves tend to promote the common good in the long run. These are matters that utilitarians argue over, and it is clear in the words Narveson occasionally still devotes to the subject that he has strong views on most of these issues.20 But how does utilitarianism, in general terms, fare against contractarianism, as the latter is cast by Narveson? Narveson’s case against utilitarianism, usefully summed up in The Libertarian Idea, has several key features. First, he decries the conceptual confusions we are left with as we contemplate the many questions left unresolved by utilitarianism ‘in general’ (for example, the uncertainties reviewed a paragraph ago). He implies that these difficulties, if not insuperable in principle, are at least sufficiently puzzling as to encourage us to look elsewhere for clarity on foundational questions. Second, he argues that the utilitarian claim that the ‘general good’ trumps ‘individual good’ in cases of conflict can finally rely only on either definition or intuition. He concedes that not all utilitarians would agree with this claim of dependency, but appears to be confident that he is right and they are wrong. If he is right, then utilitarianism is impotent against those persuaded by different definitions or intuitions. Third, while acknowledging the broad attractiveness of the utilitarian principle as a foundational principle, Narveson urges that this attractiveness may not be due to morality’s actually being based on the promotion of general well-being. Instead, while it might be true that ‘justice’ and ‘happiness’ are likely to be positively correlated in any society, this might just as well be because a significant component of happiness resides in the perception that one’s society is just. That is, happiness may be (partially) based on justice, rather than the other way around. If this were true, then utilitarianism would fail in its foundational ambitions. Finally, as what is nominally a side remark (although it is placed where one might normally expect the conclusion of the argument to appear), Narveson adds that Whenever we are in real-world situations closely resembling a ‘veil of ignorance’ in some respects, then a utilitarian solution is likely to be a good one. Suppose that some system must be devised for catering to the comparable interests of a large number of people, and that those people are essentially randomly situated with respect to the sources of the benefits. It will then be true that each person maximizes her expected utility by accepting a utilitarian format. … But this works because it is a limiting case, the case where individual and general utility coincide. Many cases certainly don’t seem to be like that, and when they aren’t, we need a different principle.21

Contracting Justice

25

If this is right, then utilitarianism, while useful in rare situations in the real world, fails us in many others. We are faced, then, with four arguments concerning the inadequacy of utilitarianism and (presumably in the background) the superiority of contractarianism: (A) Utilitarianism’s conceptual problems are too difficult; (B) Utilitarianism is dependent upon definitions or intuitions about morality that may not be shared by all; (C) Utilitarianism’s plausibility may be due to a dependency of happiness on justice, rather than a dependency of justice on happiness; (D) Utilitarianism’s applicability may be too narrow. It is not the objective of this chapter to defend utilitarianism or any other foundational competitor to contractarianism. Instead, the focus here is on clarifying Narveson’s reasons for preferring the latter. With that in mind, it is worth remembering from the preceding section that the Social Contract alluded to in contractarianism as providing foundations for morality is not really a contract. Indeed, the morality of keeping one’s contracts is supposed, by Narveson, to be non-tautologically based on the Social Contract. Hume’s indirect allegation of circularity is refuted by the suggestion that the Social Contract is really no contract at all. Let us return to a consideration of what it really is, as we think about Narveson’s several objections to utilitarianism. As portrayed by Narveson, the Social Contract is an ‘agreement’ only insofar as it is to be identified with the mutual restraint of several interacting (or potentially interacting) persons, each of whom continues to act on behalf of their own selfinterest. There is really no agreement, as noted earlier; there is instead an inclination, on the part of each of several individuals, to constrain behaviour in certain ways (‘moral’ ways) provided others do the same. This inclination is owed to restraint being in each person’s perceived long-term best interest (as long, anyway, as the condition of mutual constraint continues to hold). But wait: need these envisioned people be doing anything even so grand as perceiving that restraint is in their long-term best interest? Such mutual constraint among humans is comparable to similar restraints within animal groups. Are wolves, in cooperating and submitting in certain ways in their packs, doing so because it is in the long-term best interest of each? How about bees? Ants?22 How about human sexual behaviour? Is perceived long-term best interest even relevant? Well, it is and it isn’t. It seems implausible to say that non-human animals behave the way they do because doing so is perceived by them to be in their longterm best interest. In the case of non-human animals, it seems highly unlikely that they do much thinking at all about long-term interest. Yet the behavioural tendencies to restrain behaviour are present quite as much among non-human animals as they are among humans. In the case of at least much of human behaviour, one could surely say the same: behaviour is morally constrained in social situations even though not much thought is directed toward long-term best interest. It is just not plausible to contend that people, any more than non-human animals, restrain their behaviour in conformity with social rules because they have thought the matter through and have concluded that such restraint is in their longterm best interest.

26

Liberty, Games and Contracts

So if that is what we are to understand by the claim that ‘people behave morally because it’s in their interest’, then the claim is quite dubious. But it is not so strange, on the other hand, to observe that it is in fact in people’s and non-human animals’ long-term interest to restrain themselves in these ways, or to suggest that the practices and behaviours (constraints and all) are what they are because they do indeed serve the long-term best interests of these several creatures, regardless of whether they know it (or perceive it) or not.23 First of all, then, the bare fact of mutual restraint among interacting beings, people included, implies not only nothing about agreements made between them, it implies nothing at all about perceptions or reasoning processes they may have experienced. This means, in turn, that mutual restraint implies nothing at all about perceptions of interests, in particular, whether individual or social, whether longterm or short. This consideration, all by itself, may not be regarded as causing serious damage to contractarianism, though. For humans, anyway, can and do raise questions about why they should behave in this way rather than that. Contractarianism offers an answer: it is because behaving this way is in fact in your interest. That is how you explain to people why they should be moral, according to Narveson. The explanation might be a broad evolutionary one: these cooperative arrangements, where they have been adopted, have contributed to the well-being (at least considered in broad terms) of the people who have adopted them. Or the explanation can be cast in game-theoretic terms: under the general circumstances confronted by people in typical social circumstances, tit-for-tat dominates. You should therefore choose that strategy too. All this is fine. But it really comes no closer to explaining these things to those who see temporary advantage (at least) in neglecting the rules. Sometimes personal interest and social interest diverge; sometimes it pays to gamble; sometimes non-cooperation, especially when dealing with people you will never see again, pays off. Contractarianism, it seems, risks falling prey to a particularly interesting version of the naturalistic fallacy. Contractarian narratives, like evolutionary explanations, can be extraordinarily plausible in making sense of why people do what they do: behaving in those ways makes sense because if we were to give thought to what is in our long-term best interest, we would so behave. It does not matter at all what thought processes or perceptions real people have as they make their way through the world, cooperating here and failing to cooperate there, behaving morally here and cheating there. What matters is what we say to those who actually ask ‘Why cooperate?’ Evolutionary theory, social contract theory and game theory have something to say about this. What they have to say is surely not argumentatively decisive, mind you; it is more like general advice, at least in those situations where what they have to say is more than just bare description. The whole justificatory game differs from description precisely in that justification is relevant when someone asks ‘But why should I?’ while description offers accounts and explanations of how things are or how they have come to be. But in that light, consider what contractarianism would allow us to do: just as is the case with other theories arising mostly from descriptive explanations of why things are the way

Contracting Justice

27

they are, they can be converted to justificatory application in surprising ways. Precisely as plausibly as we can use contract theory to explain to a hypothetical recalcitrant neighbour (or to ourselves) why they (we) should be moral, we could use it also to explain to the hypothetical recalcitrant ant why it should not wander off and try to fend for itself. Such a plan just wouldn’t be reasonable, given typical antly survival requirements. Cooperation is surely the best policy. It would be much wiser for each ant to align its behaviour with that of its fellows, provided those other ants do the same. But does the fact that we can apply considerations about ant interactions in this way give us any reason at all to invoke the conceptual trappings of contract? If not, what is there in the corresponding tale about people that gives us any more reason to do so? Perhaps the contractarian would want to point out that the person can, while the ant cannot, consider reasoning of this kind and then go on to bring behaviour into line with it. But that should not really matter. If the foundation of morality is to be found in what it would be reasonable to agree to, rather than what is actually agreed upon, then it makes no difference that ants and people differ in their respective capacities to follow the dictates of reason. Perhaps ants are hard-wired to do what is reasonable/moral (at least most of the time), while humans are not. But if morality really is based on what is reasonable, then ants and humans are both subject to its rules. There should at least be something puzzling, though, about the fact that this kind of argument works for ants as much as it does for people, and this puzzlement takes us back to Narveson’s arguments against utilitarianism. Let us now see, after the foregoing musings, how contractarianism stands up in the competition. A. Utilitarianism’s Conceptual Problems Are Too Difficult It is no small conceptual problem that contractarianism depends on no contract; that the ‘agreement’ envisioned among people involves no agreement; that the perception that cooperation is in their interest need not be any part of the envisioned participants’ reckoning; that cooperation may not, indeed, serve the particular perceived interests of particular persons in particular situations, and that cooperation may, instead, be quite automatic (rather than considered in the light of self-interest) in most situations where it is actually observed. The confusion here stems in large part from moving between a conceptual situation in which one is trying to understand why certain behaviour patterns emerge in populations, on the one hand, and a wildly different conceptual situation in which one is trying to explain to people why they should be moral. It would appear that contractarianism may be able to offer interesting advice, but such advice is only good for people whose perceived interests are normal or ‘rational’, which brings us to the next point. B. Utilitarianism Is Dependent upon Definitions or Intuitions about Morality That May Not Be Shared by All If contractarianism is to work theoretically in providing ‘foundations’ for morals, it

28

Liberty, Games and Contracts

has to be talking the right language. If one assumes that practical rationality consists in choosing effective means for one’s ends, and if one further assumes that the ends of a particular individual are the ‘reasonable’ ends that any of us ‘should’ (naturally?) have, one may then be able to give practical advice about how one should behave.24 You are in an iterated prisoner’s dilemma situation. You should adopt a policy of tit-for-tat, because that will give you a better chance of positive outcome than any other policy. Positive outcome, it is interesting to note, is in such settings presumed to be transitive across individuals. You cannot work out this game-theoretic stuff if the units of reward are not commensurable. The nice thing about game theory is that it does not matter at all what the units are units of; but whatever they are, they do have to be the same from one person to another. Thus arise game-theoretic parallels of some of the ‘conceptual’ problems of interpersonal comparison that supposedly condemn utilitarianism. But never mind that. Does this sort of thing – practical advice in obtaining what you want – generate morality? I suspect that depends on your ‘intuitions’ about morality. One disturbing thing about Narveson’s own contractarianism, for example, has to do with the treatment of animals. Narveson is committed to the idea that mistreatment of animals is wrong, all right; but it is not directly wrong, so to speak. According to Narveson, whatever wrongness there is in such behaviour has to do with our interactions and quasi-contractual negotiations with our fellow humans, our own several tastes and so forth.25 Why? Because (simply put) morality is based on contract, and people cannot make contracts with animals. That this should be the way to arrive at moral conclusions seems wildly counter-intuitive.26 I am no radical on this particular subject. Indeed, I am not at all sure what is right and wrong concerning the treatment of animals (apart from my abhorrence of plainly cruel treatment, which I appear to share not only with animal-rights’ activists but with contractarians such as Narveson). But the idea that what is wrong regarding the treatment of animals should end up being fully dependent on what you and I perceive to be best for ourselves just seems mistaken. There is no problem with agreeing that actual moral and legal codes drawn up by or evolved by real communities do arise mostly through agreements made by humans (although the role of real ‘agreement’ in such matters needs to be scrutinized case by case). But the question whether such moral and legal codes actually capture what is right must remain open. The evidence of this is that they so often don’t capture what they putatively aim at, and this is especially clear in cases where the content of the codes is plainly biased in favour of the interests of a minority. That majorities may in other cases be the beneficiaries of more democratically formulated codes does not alter the fact that they can be wrong too. C. Utilitarianism’s Plausibility May Be Due to a Dependency of Happiness on Justice, Rather than a Dependency of Justice on Happiness Contractarianism’s plausibility, as noted above, may be due to a misapplication to contexts of justification of theoretical equipment that is appropriate only in contexts

Contracting Justice

29

of description. Further: while it may well be that it is reasonable to agree to be moral (because it is in your best long-term interest), it is no clearer in this case than in the case of utilitarianism that the dependency is as contractarianism needs it to be. For there may be (and likely is) something in behaving morally that renders it rational to adhere to, rather than the other way around. That is, it may be that far from arising from ‘Social Contract’, morality’s ‘foundations’ are entirely independent of ‘agreement’. Instead, ‘agreement’ is rational because of the morality (based on other foundations) of the behaviour. D. Utilitarianism’s Applicability May Be Too Narrow Narveson’s remark to this effect was directed particularly at Rawls-like invocations of ‘veils of ignorance’ and their constraints. This seems an odd complaint for Narveson to make, given that his Social Contract involves no contract, no agreement, no consideration of particular humans in particular circumstances,27 but instead relies on considerations that seem no less abstract than those invoked by Rawls. It is hard to see why anyone should prefer Narveson’s framework, in principle, to Rawls’s. Indeed, it seems that Narveson might just as well have preserved the veil-of-ignorance contraption while lobbying for its repositioning. We need to take into consideration, according to Narveson, some of the things that Rawls argued should be obscured. So why not just move back the veil a bit? We are still in pretty abstract territory, after all, when compared to real human beings making real decisions and making (or failing to make) real agreements and contracts. More than that, though: as indicated above, Narvesonian contractarianism allows us to draw moral conclusions only about situations that involve, in one way or another, human interactions with other humans. The realm of morality and justice is thus narrowed to a particular realm of behaviour – the realm of human interaction – and can be extended to putatively moral obligations regarding non-human aspects of the world only indirectly, if at all, by way of reduction to human quasiagreements with other humans. This rendering of contract theory, therefore, in providing a contractual justification for moral claims, does so by contracting the realm of justice to a base much narrower than the realm of common moral judgment that is to be explained. While a reduction of some kind or other might well be acceptable in principle, this one, as indicated above, seems unwarranted. The name ‘contract theory’, while intending to focus attention on quasiagreements, may thus inadvertently highlight one of the theory’s more problematic features: it contracts morality and justice to an inappropriately narrow base.

Conclusion Whether one accepts contractarianism of this kind or not, it must be conceded in closing that there is much that is attractive about the line of thought that supports it. Even if it is unreasonable to think of it as involving a ‘contract’ or ‘agreement’ (that

Liberty, Games and Contracts

30

is, even scare-quoted ones), it must truly be the case that morality, in broad terms, somehow serves the interests of the beings in whose lives it plays such a large role. The particular naturalistic considerations invoked by Narveson (or invokable in principle) in defence of this version of contractarianism – evolutionary considerations suggesting that certain constraints exhibited by people (or by non-human animals) make sense in terms of survival, or game-theoretic considerations that suggest that choosers in certain decision situations would be left worse off if they acted in terms of a narrowly construed self-interest than if they moderated their behaviour socially – are suggestive. They are correctly regarded as promising areas for further inquiry into possible ‘naturalistic’ foundations for morality. But they do not justify the language of contract.28 This is because they leave us only with the fairly secure conclusion that there are good reasons for the kinds of restraint we observe in real populations of interacting animals, whether human or non-human. It is of course possible to contend that ‘Social Contract’ provides the basis of nonhuman behaviour as much as it provides foundations for morality, but this stretches the notion of ‘contract’ much too thin. And that is what we would have to be willing to do if we were to accept Narveson’s version of contractarianism for humans.

Notes 1

2 3 4 5 6 7 8 9 10

11

I am grateful to Jesús Aguilar, Timothy Engström, Malcolm Murray, Wade Robison, Evan Selinger, David Suits, Katie Terezakis and Victoria Varga for their encouragement and for their helpful suggestions concerning more primitive drafts of this chapter. Jan Narveson, The Libertarian Idea (Philadelphia, 1988), 131. The quoted passage appears in a section in Narveson’s book entitled ‘The Idea of the “Contract” Approach to Foundations’, 131–4. Ibid., 126. Ibid., 123. Ibid., 125. Ibid., 124. Ibid., 127. Jan Narveson, ‘The Agreement to Keep Our Agreements: Hume, Prichard, and Searle’, Philosophical Papers, XXIII/2 (1994). David Hume, Inquiry Concerning the Principles of Morals (Indianapolis, 1957), Appendix III, 122. See also his Treatise of Human Nature, L. A. Selby-Bigge (ed.) (Oxford, 1978), bk III, pt II, sect. II, 484–501. Nor does Hume, really. Like Narveson, Hume may be understood as a ‘contractarianof-sorts’. Unlike Narveson, though, Hume would probably not have cared to insist on the ‘rationality’ of contract as a justifying factor. It seems likely that his considered position would be much more naturalistic, along the lines of his well-known analysis of ‘inductive reasoning’. We do indeed evolve moral codes as a function of their success in interpersonal activity, just as we develop expectations of the future based on our experience of the past. There is nothing particularly ‘rational’ about this, though. It just reflects the output of natural dispositions which we’ve come to have through means that have nothing to do with reason. See especially the passage about ‘Two men who pull on the oars of a boat’ in Hume, Treatise of Human Nature, 490. Hume’s discussion, in this

Contracting Justice

31

passage, concerns the way in which social conventions – like ‘the rule concerning the stability of possession’ – arise in human societies. While much of this discussion can indeed be interpreted in terms of rational choice (and some of it admittedly seems hard to interpret in any other way), the emphasis on the slow, gradual process of growth of such conventions cannot be interpreted that way. See also, on this point, Brian Skyrms, ‘Game Theory, Rationality and Evolution of the Social Contract’, in Leonard D. Katz (ed.), Evolutionary Origins of Morality: Cross-Disciplinary Perspectives (Bowling Green, OH, 2000), especially 272–3. Skyrms makes some especially interesting suggestions in this article concerning the occasionally antagonistic relationship between rational choice applications of game theory and evolutionary applications. While Hume’s approach is closer to modern evolutionary game theory in most respects, Narveson appears to want to use both approaches from time to time. For more on the contrast between rational choice game theory and evolutionary game theory, see Skyrms, ‘The Stag Hunt’, Proceedings and Addresses of the American Philosophical Association, 75:2, 2001: 31–41. 12 Narveson, ‘The Agreement to Keep Our Agreements’, 76–7. 13 Ibid., 77. 14 David Gauthier, Morals by Agreement (New York/Oxford, 1986). 15 Narveson, ‘The Agreement to Keep Our Agreements’, 78. 16 Robert Axelrod, The Evolution of Cooperation (New York, 1984). 17 Jan Narveson, ‘Contracting for Liberty’, in T. R. Machan and D. B. Rasmussen (eds), Libertarianism for the 21st Century (Lanham, MD, 1995). 18 See Narveson, Libertarian Idea, 150–53. 19 See especially Jan Narveson, Morality and Utility (Baltimore, MD, 1967). 20 For a sketch of some of these views, and for the arguments of the next paragraph, see Narveson, Libertarian Idea, 150–53. 21 Ibid., 153. 22 Ants? In ‘The Agreement to Keep Our Agreements’, Narveson describes the Social Contract as ‘an agreement in the sense of a co-ordinated set of conditional dispositions’, 84. It is important to Narveson’s overall argument that this is what the Social Contract is, and that it not depend on a prior or a more basic agreement. If so, then wherever one finds such coordinated sets of conditional dispositions, there one ought also to have a Social Contract. Ants. 23 It is, of course, important to note that the ‘interests’ promoted by restrained behaviour on the part of individual animals, human or not, may not always be those of the selfsame individuals. It might be, for example, that what is promoted in some Social Contracts (that is, coordinated sets of conditional dispositions – see note 22, above) is the ‘interest’ of the tribe or group or species. Such interests, of course, can conflict with those of the individual, opening again the motivational question that Social Contract theory is supposed to resolve. Why should I care about the Social Contract? Perhaps a contractarian would argue that some Social Contracts are more reasonable than others to abide by, but the question then would be ‘Which ones?’ and the answer seems likely to depart from references to contracts, social or otherwise. 24 Peter Stemmer, while reluctantly going along with the tradition of using the expression ‘hypothetical contract’ to refer to a justificatory mechanism like the one proposed by Narveson, observes cogently that what actually does the justificatory work appears to be no more than a ‘configuration of interests’ among the members of some group, and that what makes contractarian terminology plausible is simply the fact that it is just such configurations that often, under normal circumstances, provide the framework for

32

25

26

27

28

Liberty, Games and Contracts actual contracts. See Stemmer, ‘Moralischer Kontraktualismus’, Zeitschrift für philosophische Forschung, 56/1 (2002): 1–21, esp. 16–17. Jan Narveson, ‘Animal Rights Revisited’, in H. Miller and W. Miller (eds), Ethics and Animals (Clifton, NJ, 1983). See also Jan Narveson, ‘A Case Against Animal Rights’, in Advances in Animal Welfare Science (Washington, DC, 1986); and Jan Narveson, Moral Matters (2nd edn) (Peterborough, Ontario, 1999), ch. 6: ‘Morals and Animals’, 133–42. This idea also seems to misapply the logic of Social Contract as set out elsewhere in Narveson’s work (see previous citations), which depends upon what it would be reasonable to agree upon, rather than upon what anyone actually agrees to. Narveson writes, on occasion, as if his contractarianism is superior to other foundational approaches precisely because it does indeed emphasize particular persons rather than abstractions. He presses this point with particular emphasis in the opening paragraphs of an as yet unpublished essay with the working title ‘Baseline, Baseline, Where Art Thou?’ But Narveson does not actually accomplish this in any of his published work, and for good reasons. His objective, after all, is a universal moral code, so he must inevitably deal with people in terms of what they have in common, rather than in terms of their individual differences. Narveson’s solution is to speak about what ‘everyone’ would agree to, rather than what anyone agrees to. It is ironic that some of the reasons offered by Narveson for thinking of his proffered ‘foundations’ for morality as contractarian appear to be precisely the reason some thinkers offer by way of criticism of contract theory. Hegel, for example, argued that there are good reasons to opt into social institutions whether or not we have given our consent to them, and these reasons can justify those institutions quite independently of consent. Because consent is not necessary – or even morally relevant independently of social institutions, according to Hegel – the idea of ‘contract’ has no justificatory role to play. Narveson’s concession that the ‘Social Contract’ is not a contract, but may instead be inferred or read from behavioural tendencies, together with his willingness to infer a ‘Social Contract’ from what it is reasonable for people to agree to whether they actually give their consent or not, seems to bring his view into what Narveson might regard as an uncomfortably close alignment with Hegel’s anti-contractarian view. See Georg Wilhelm Friedrich Hegel, Die Vernunft in der Geschichte, J. Hoffmeister (ed.) (Hamburg, 1955), 118. See also Alan Patten, ‘Social Contract Theory and the Politics of Recognition in Hegel’s Political Philosophy’, in Robert R. Williams (ed.), Beyond Liberalism and Communitarianism: Studies in Hegel’s Philosophy of Right (Albany, NY, 2001), 167–84.

Chapter 3

Is Agreement Enough? Tibor R. Machan

Introduction I believe I am right to associate Professor Narveson’s work in political philosophy with the social compact tradition. Theories of social compact aim to provide the rationale for social or political norms, though arguably some include principles of personal conduct as well.2 In particular, Narveson’s work sets out to show that the formation of a social compact – I eschew the more common term ‘contract’ because it presupposes a legal order which such theories aim to establish – rests on members of society acting in their self-interest and agreeing with others on principles that will make this pursuit an ongoing possibility. A social compact theory in the sphere of social, political or even legal norms would contend that the correct principles of community life are those to which either actual or hypothetical community participants agree to be legally bound.3 I argue here that social compact theories are inadequate without invoking logically prior non-conventionalist ethical principle(s) or, alternatively, a theory of natural (innate, biological or psychological) drives (for example, the drive for selfpreservation). Their success must in part rest on such elements that are not usually mentioned. By not mentioning these elements social compact theories leave the impression that they are independent of various controversial philosophical elements, be these ethical or metaphysical. They then claim to possess the virtue of being philosophically flexible, able to accommodate a great variety of human beings in a society; thus they are viewed by their proponents as presenting a better chance for reaching widespread acceptance of the principles that emerge from the compact. To the contrary, this seems to me illusory. I argue that social compacts by themselves, involving only agreements by members of a community, cannot establish binding norms of any sort whatsoever. That is to say, briefly, nothing follows from everyone concerned promising to abide by some guidelines unless promises themselves have normative significance. And this cannot derive from a still prior compact, ad infinitum. Unless there is some prior norm that binds a promise, the promise to abide by the compact is not one we need to abide by. It contains no rationale committing anyone to do anything. That is, norms resulting from promises have moral force only if promising, independently of any agreement making it binding, has moral force.

33

34

Liberty, Games and Contracts

The other alternative is to accept that there is some inborn motive or drive that impels those party to a compact to live by it. Then the norms will be adhered to not because of the agreement to abide by them, but because such an agreement is something the parties had to make – for example, so as to secure their safety or preserve their lives. Once I offer my critical remarks I sketch an approach to establishing social or political norms that captures some of the appeal of social contract theories yet also avoids the problem of their lack of support besides agreement or consent. I show that, as usually understood, social contract theories implicitly assume that prudence is a moral virtue – or a natural drive: one that we ought to – or will be motivated to – cultivate and in line with the dictates of which we ought to, or will, agree to the provisions of a social compact.

Is Self-consistency Enough? It is useful to note, as a preliminary point, that in the works of, for example, Kant and Rawls, the alleged contractual basis for the legitimacy of law and its enforcement is supplemented with the very strict requirement of self-consistency of the resulting norms. Once this element of internal consistency is isolated and set aside, what is left of the social compact basis of norms and law carries no punch whatsoever. Is self-consistency, however, not demanded of every theory – it is a metatheoretical criterion that goes without saying? If so, this self-consistency requirement places serious, binding constraints upon what may be contracted for via a social compact, and so agreement alone does seem to render the content of the social compact binding. But if such self-consistency is not inherent in a system that rests on agreement but is, rather, a distinct normative requirement, this suggests that normativity – the relevance of value judgments – is not something that may be dispensed with in any sphere of human concern as social compact theorists suppose.4 Still, does rationality need to be a prerequisite for any social compact? It may not be an inherent requirement of such a compact to involve participants who respect rationality. If some were to promote the idea that we ought to unite and reach agreement on irrational rules to govern us, it may obtain support and agreement on how such rules may be reached. We know, for example, that many irrational groups exist, such as gangsters, gangs, terrorists or other associations, the purposes of which are, as it were, crazy – al-Qaeda can be regarded as such, as can the Nazi Party or skinheads; each of which group might establish their rules on the basis of nothing but agreement without these rules being morally binding on anyone. (That is to say, deserting such a group would not count as doing something wrong.) That the objectives or purposes sought by way of such irrational rules could not be conscientiously and consistently implemented, acted upon, is, of course, just my point: something aside from an agreement is needed to establish viable, coherent and morally binding social compacts.

Is Agreement Enough?

35

Agreement Can Be Irrational, Wrong Let me spell this point out further. To deny that rules born out of a social compact resting entirely on agreement can be irrational flies in the face of history and common sense. We are surrounded by clubs, organizations, institutions and other human groups that clearly embrace irrational ideas which they incorporate as their proposed guiding principles of conduct. Yet to claim they have moral force is to imply that they have moral legitimacy. But the substance of the criticism of such groups is precisely that they lack moral legitimacy. So such agreed-to rules could arise from a social compact without being morally binding – who would morally blame someone who, upon a change of heart, decided to violate the norms of the Nazi Party? Kant stated explicitly that the ‘idea [of social compact] alone enables us to conceive of the legitimacy of the state’.5 He made clear that the social compact would be a ‘mere idea of Understanding, which has, nonetheless, its doubtless (practical) reality’ in that it ‘obligates every lawgiver to advance his statutes so that they could have emerged from the united will of the whole people, and to consider every subject, as regards his desire for citizenship, as though he had been party to accepting that will. For that is the basis of the legitimacy of every public enactment.’6 This idea was recently revitalized by John Rawls by way of the hypothetical social compact that is to be conceived of as having been formed behind ‘a veil of ignorance’.7 There are other, more Hobbesian, versions of social compact theory, but the Kant–Rawls version has become a prominent one among contemporary political philosophers. The problem is that their versions are not free of certain normative presuppositions and thus do not qualify as pure social compact theories. In other words, both Kant and Rawls pack in certain norms prior to or combined with the social compact, norms that seriously delimit what can count as a bona fide compact. But then the resulting agreement will itself be much more than something arising from the compact alone.

Hobbesian Compactarianism Hobbes may be a better representative of pure social compact theory. For Hobbes the gist of the story is that when scarcity and congestion reach a certain point in the state of nature, the self-interested motivation of individuals drives them to convene and reach a consensus about certain rules of social life which a monarch or government was to make sure everyone obeyed by imposing strict sanctions against lawbreakers. At this point everyone gave up the ‘natural rights’8 that each possesses in the state of nature, except for the right to resist being killed. In Hobbes’s version no content for the rules is specified, though it is assumed that they would aim for the maintenance of peace and preservation of human life. Peace was seen by Hobbes as the only public good. In the Kantian–Rawlsian version of social compact theory, however, more is required than merely laying down rules – the rules must be rational, fair and so on.

36

Liberty, Games and Contracts

What is so appealing about the Hobbesian idea, especially among supporters of the Enlightenment-based liberal tradition of politics? It seems well suited to the task of providing an empirically based explanation of the content of preferred laws. Even Kant embraced this for the phenomenal world – for example, in his discussion of how people pursue happiness as a matter of natural necessity – where the practical task of politics is to be realized. The idea that human actions must be explained by reference to motives (as distinct from purposes, for example); that nature provides no guidance to judging something good or bad; that morality is mainly a social invention; that people lack freedom of the will, at least from the scientific, practical point of view, and that their conduct is neither morally right nor morally wrong except as a matter of their relationship to certain social goals, all these and related considerations accommodate the liberal goal of anti-authoritarianism and scientific respectability. According to the empiricist-scientistic viewpoint, prior to being bound by their own uncompelled choices (which in the last analysis come to hypothetical revealed preferences, so as to accommodate empiricists’ standards), members of a social group cannot be bound by moral standards. By this understanding of human life, it is not the case that people ought to live in certain ways rather than others (since that implies that they could do otherwise, that they possess free will and might be free in the sense of initiating their own conduct apart from not being compelled by others). They will be inclined or driven – naturally motivated – to do so. Some dismiss this point on grounds that the idea of free will, such that one causes one’s own conduct, makes no sense and so morality could not involve the provision of free choice – that ‘ought’ does not imply ‘can’. For them Hobbes, a mechanical materialist, could have a bona fide moral system, since all that means is that he could promote certain kinds of behaviour and discourage others. For such folks ethics or morality is not a matter of choosing, on one’s own initiative, to do the right thing but of behaviour in line with certain rules. It is this view that seems to me to have recommended the social compactarian stance because with the demise of the classical idea of natural law – the view that human life has an end or purpose and that freely chosen, self-caused right conduct ought to and can further this – the hypothetical social compact thesis had the best chance of resurrecting some semblance of respect for social moral and legal principles.9 But here we have an argument about the very nature of morality, including of social moral norms – are they ways we are driving to behave under certain optimal circumstances or are they ways we ought to come to behave on our own initiative? It is this latter that was the idea of ethics and morality prior to the empiricist-scientistic era of moral philosophy. Yet something had to be said that addressed our moral concerns. Let me put this in somewhat different terms. If the only hope for some semblance of normativity – moral and political standards – arises within a deterministic framework, the place of a moral or political norm will be taken by some instinct or innate motive. For Hobbes that innate motive was self-preservation or survival. It is this that drove people to form a social compact, even if he spoke on and off about this amounting to morality. (Philosophers often try to link concepts that do not

Is Agreement Enough?

37

really mesh.) So the theoretical role assigned to natural law (or some other way of understanding basic ethical norms) was assigned by Hobbes to a basic instinct or innate drive, never mind that he kept some of the older language of ethics or morality in place. In both cases, the social compact in and of itself does not contain the full measure of normativity often associated with it. Simple agreement does not generate bona fide moral norms arising from the (hypothetical) social compact. Instead, either an innate biological or psychological drive or some prior natural (moral, ethical) law is needed.

Beyond Universal Agreement As noted, both Kant and Rawls invoke much more than mere universal agreement in their efforts to make the compact supply us with moral and legal standards which can be considered binding. Rawls believes that the compact must be confirmed by intuitively acceptable moral conceptions placed in rational equilibrium.10 He also introduces the idea of the veil of ignorance so as to establish for himself a basis for consistency. Consistency would be thwarted if the contractors had important individual characteristics, so Rawls makes all the contractors uniform. So, in Rawls, the element of consistency comes from both the rational equilibrium and the veil of ignorance provisions. The consent, in turn, seems to be based on the intuitions which Rawls imposes on the contractors. Why is a mere agreement not sufficient? If no other source of norms is available – especially in the case or Rawls – why not simply insist that norms come from human choices, be they rational, self-consistent, irrational, illogical, arbitrary or whatever? I think the reason this road is blocked for Kant and for Rawls is that both reject any sort of metaphysical or ontological grounds for morality. Rawls was very clear in his Eastern Division American Philosophical Association (APA) Presidential Address when he stated ‘Now my thought is this: much of moral theory is independent from the other parts of philosophy.’11 Kant, too, locates morality in pure reason, without external foundations: ‘Every action is right which, in itself, or in the maximum on which it proceeds, is such that it can coexist along with the freedom of the will of each and all in action, according to a universal law.’12 With reference to social morality Kant held, furthermore, that we should all ‘act externally in such a manner that the free exercise of the will may be able to coexist with the freedom of all others, according to a universal law’.13 (Kant wished to deduce every legal judgment from pure reason.) For example: ‘If a man and a woman have the will to enter on reciprocal enjoyment in accordance with their sexual nature, they must necessarily marry each other; and this necessity is in accordance with the juridical laws of pure reason.’14 Now in Rawls’s case, the attempt to reject personal and social morality’s dependence on other parts of philosophy cannot succeed for reasons I have noted elsewhere.15 Briefly, any effort to explicate the Rawlsian task to specify the nature

38

Liberty, Games and Contracts

of moral obligation begs for philosophical clarification from the non-normative branches of the discipline. There is no way to avoid explication of the idea of ‘rational’ in terms of at least elements of the philosophy of mind, ontology, the subject of free will versus determinism and so forth. If, for example, being rational implies, as Kant suggested, that an agent is free to make choices, to initiate conduct, this alone will make a difference as to the ethics and politics that human beings ought to adopt. Some might contend that ordinary understanding of rationality will suffice, that there is no need for metaphysics or deep philosophy at all. But while it is true in many ordinary human circumstances that common sense that is in force at the time carries along the discussion with sufficient clarity and precision, under certain kinds of sceptical assaults this will not suffice. Just as we all manage to engage for a good while with the physical, chemical, sociological and economic realities that surround us without reliance on the specialists, at certain points we come to depend on their work. So with ethics and politics. Common sense is certainly the beginning but not the end of inquiry. It might be argued, as well, that social compact theories claim that each person appraises proposed items for inclusion in the fundamental code of social morals on the basis of their own interest, given the interests and powers of others. So if one’s acceptance of some point of philosophy, say in metaphysics, affects one’s interests in a way that calls for a new item for inclusion in the social compact, that may be defensible. It will be part of the list being negotiated. An example could be that there must be free inquiry in a society – that is surely a plausible item. Yet does this become part of discussion because some metaphysical fact implies it? Or because of the fact of differences between various parties about such a metaphysical fact, so that there can be a question how this difference is to be accommodated? Yet if there are some facts of any kind, including metaphysical, that permeate social life – for example, that all human beings are sovereign agents and their agency is a precondition of a moral life – they cannot be contracted away. So such provisions will be unavoidable, a necessary part of a coherent human social life, whether they are included as a provision of the compact. An example might illustrate this point. Say it is part of the social compact that some people in society would be slaves. Yet if slavery is a basic violation of one’s humanity, it could never be part of a meaningful social compact. When in Ghana, for example, young virgins are forced to become the sex slaves of the priests – purportedly to atone for the grave sins of some family member – this cannot be a matter of their social compact, however much it is insisted upon. It violates the basic fact of human life that human beings must choose the way they will live, each and every one of them; but the young virgins are left out of this, even when they become of age. Social compact theories that do not take this into account are failures: not because they have violated some agreement, but because they have failed in their agreement to abide by certain unavoidable principles of human social existence. Kant attempts to secure the foundations of morality in pure reason, removed from that impossible field, metaphysics. Yet this is a kind of metaphysics, only a complicated kind. What we do see in Kant, though, is that social compact theory

Is Agreement Enough?

39

rests on requirements as to how human beings ought to act that do not themselves rest on agreement. The particular moral principles Kant claims to have shown to govern human life may not be the ones we ought to live by, but it is clear that some such principles are needed so as to make social compacts morally palatable.

Common Sense Scepticism about Compacts Now common sense already suggests one serious problem with all kinds of social compact theories, Kantian, Hobbesian or Rawlsian. This is that it is always perfectly intelligible to ask whether what people (would have) willed either behind a veil of ignorance, in unison or at some great convention should have been willed by them in the first place. Some could argue that the very point of making agreements, normally, is to settle something. If asking the above question amounts to reneging on the agreement, we’d better have heard about the question a bit sooner.16 Yet that is just it. Where does this ‘We’d better have heard about it sooner’ come from? The question implies a moral judgment, namely, that if you did not like a provision, you ought to have said so before agreeing to it instead of reneging on the compact later. Now this is a moral claim that has its basis not in a prior compact but in something more fundamental, namely, some natural law or objective moral principle. If someone agrees to do X and later disputes the force of the agreement, it can be said: ‘You ought to have considered your doubts earlier.’ Does this last claim itself rest on some agreement, ad infinitum? If so, it has no force that obliges one to heed it. If not, then the social compact is itself grounded in something outside of itself, on the binding nature of promises apart from what may or may not have been agreed to. (Here we recall that agreeing to something involving third parties who were not consulted is perverse not because there is an agreement against such a procedure but because consent is fundamental in adult human relationships, be this agreed to or not.) When someone considers entering a social compact while encountering scarcity and congestion in the state of nature, the question can be raised, ‘Ought I to join in the search for some way of solving our problem?’ In other words, there must already be standards for making the right choice about doing so. This standard is very likely what is called prudence: we ought to take reasonably good care of ourselves, including when we embark upon social life. (Rawls admits that some such standard might be discovered, but regards it obstructionist to wait for it.)17 It is not cogent to deny that such a question can arise, based on the view that entering the convention is a matter of being driven to do so out of fear, for if we are determined to behave as we do, then this will hold following the social compact, so no judgments as to what we ought to do will be applicable. If ‘ought’ implies ‘can’, then this holds before and after the social compact. But if it holds before, then there can be standards of ascertaining what one ought to do prior to the compact.

Liberty, Games and Contracts

40

Narveson in Particular A quite recent example of social compact theory is advanced by Jan Narveson, in his The Libertarian Idea.18 He characterizes his version of this approach as the view that ‘the principles of morality are (or should be) those principles for directing everyone’s conduct which it is reasonable for everyone to accept. They are the rules that everyone has good reason for wanting everyone to act on, and thus to internalize in himself or herself, and thus to reinforce in the case of everyone.’19 But this is problematic because it builds into the social compact the requirement of reasonableness, yet, as we just saw, agreement by itself need not involve this at all. The normative requirement that we ought to be reasonable cannot be assumed as constitutive of agreements and, thus, of social compacts. Some who consider agreeing could ask, ‘Why should we be reasonable?’ Granted, this may seem like a silly, even incoherent, question to some. Yet it is certainly one that has been raised by philosophers and, especially, some literary and religious figures. It may not be reasonable to have faith in God, but many believe that that is just what we ought to have. It may not be reasonable to indulge one’s momentary feelings or passions, yet there are famous figures throughout history who have promoted this notion. Why are they wrong? Why should we, as Socrates had advised, follow reason instead of fancy? The contractarian stance does not answer that question but simply takes the indispensability of reasonableness for granted. As Narveson puts the point, Those who insist on being unreasonable aren’t being metaphysical drop-outs; they are simply being sociopaths, and social contract theory explains very succinctly how to deal with such people: no contract, nothing is disallowed, and you do whatever you think necessary, period. If the ‘unreasonable’ don’t like that, then they have their answer to the question, ‘why be reasonable’. 20

Yet it is philosophically impoverished to provide this kind of answer to all those who have championed unreason, all those who have maintained that being reasonable is in some ways sectarian, biased or an expression of Western prejudice. Such people may be wrong but they need not be sociopaths, as are the inhabitants of the institutions for the criminally insane. Indeed, if only those in such institutions defended unreason, all the efforts by certain philosophers and others to rebuff, for example, the post-modernist Zeitgeist, with its very large dosage of irrationalism, would be pointless and wasted. Of course, it might not be easy to reorient irrationalists toward reason, but that the effort is made repeatedly suggests both that many who are capable of being reasonable choose not to be and that most of us see hope of convincing some such champions to become reasonable (perhaps by appealing to some modicum of reasonableness they exhibit). The answer to why it is wrong to be unreasonable could very well amount to a lengthy philosophical and related story – for example, something along the lines that we are by nature rational animals; that nature, in turn, is a rational order, not some random, chaotic mess; and that for us to deal with nature successfully it is vital that we be reasonable. This answer, in turn, requires (at least now and then) a

Is Agreement Enough?

41

defence – it has to be shown to be true, not simply taken for granted. And for that purpose it is not sufficient to rely solely on common sense. And even if it is not going to convince those with an entrenched irrationality, it will at least help us to guide ourselves, including in how we ought to deal with irrational persons.

The Compact and Egalitarianism There are other problems with compactarianism, too, among them its fundamental egalitarian stance which is often taken for granted by its champions. The parties to the compact are viewed as having equal standing, yet why so? What if some were born with superior sensibilities and thus were in a better position to decide what rules ought to guide all of us? This egalitarian assumption needs to be defended, not just assumed; yet social compact theories typically start with the view that parties to the compact possess an equal measure of required sensibilities (for example, prudence, self-regard). Furthermore, if prior to the compact certain ethical standards already bound us, then social compacts could be taken as ill-formed. They should never have been entered into, and so forth. Prior to entering an agreement, one might have to decide whether to enter it – on certain ethical grounds. Indeed, given that social compact theory is often invoked to justify laws or systems of rules that reach far into the future, past the life plans for those who might have taken part in the agreement itself, one can easily raise the question of whether the contractual provisions of those in the past deserve respect. Should we be loyal to the US Constitution, for example? How about the United Nations’ Charter? Or the laws that arise from Hobbes’s social compact or Rawls’s compact forged behind the veil of ignorance? That would raise the issue of what grounds there may be for dismissing some of the provisions of these and other social compacts? Are such grounds themselves valid only because some later assembly of citizens has affirmed them? Or do these documents not qualify as social compacts, and why is that; and are there not many who disagree about that very issue and what are we to do about them – is there some agreement, some social compact that will provide the answer to that question?21 These are objections to social compact theory that need to be addressed. I do not believe they can be, successfully. The question that remains then is whether and where the social compact approach, given its shortcomings and given that from different philosophical perspectives it would not be necessary, is still required. I believe its power rests mostly on the scientistic, non-normative understanding of human conduct whereby we are driven to act as we do by inborn psychological dispositions that we cannot resist as a matter of choice.

Beyond Empirical Political Philosophy Once the theoretical power of empiricism and scientism has been lifted, I think the appeal of social compact theory would be narrowed considerably. In other words,

42

Liberty, Games and Contracts

while agreement may be important for purposes of implementing certain principles, it is not agreement per se that would make those the principles that we ought to adopt for purposes of community life. The orthodoxy of empiricism, in abandoning any hope for metaphysics, however, lacks philosophical as well as, for the time being, cultural force. The field is wide open and new beginnings can be forged.22 So, it seems to me, in particular, that a reconsideration of the Greek conception of human nature, specifically Aristotle’s, would provide greater promise for grounding justice than any such truncated approach as social compact theory, given that social compact theories appear clearly to presuppose some substantive provisions which are themselves in need of the sort of support we find Socrates, Plato and Aristotle offering for their moral and political conclusions. Why do social compact theorists reject this approach to grounding politics and law? As Narveson puts the point, the answer is easy: ‘There are people out there, apparently, who don’t agree with the Greeks about how to live – or anyway, don’t seem to. … Aristotle is telling us how to be happy, and if he’s right and we genuinely disagree, then we shall be unhappy. So what? What business is it of [anyone] whether I am unhappy?’23 The problem with this response is that the same can be said about the results of the social compact. Narveson and others tell us that we ought to let everyone be free. But there are people out there, not just apparently, who do not agree with Narveson about how we ought to act toward one another as human beings, as fellow citizens, as neighbours and so on. So what if we are not free? We simply will not be able to do as we want to, so what is wrong with that? Where is there a good reason, other than some people’s agreement, that freedom is better than, say, full equality or conscripted service to the gods or some other goal?

Narveson and Aristotle: Equally Problematic Put differently, there is just as much trouble about agreeing with Narveson as there is with Aristotle. This is of no philosophical consequence. In part philosophy is done so as to convince people of what they do not now believe, be it done by Aristotle and his students or Narveson and his. Social compact theory does not escape the problems we face because of disagreements among us. It is no less difficult to guarantee the acceptance of the case for principles that arise from a supposed social compact than ones that arise from Aristotelian arguments. If it were, social compact theory would long have produced the results Narveson believes follow from it, namely, widespread agreement about the value of individual liberty and the resulting protection of that liberty throughout human societies. It is more likely that theories of all kinds will find sceptics who reject them. The issue that remains is which theories are better. If it turns out that people in fact ought to strive to be happy, then that is a premise with which our thinking about how we ought to live must contend. It must contend with it when we consider what

Is Agreement Enough?

43

kinds of communities are best suited for us, since ones that disallow our striving to be happy will fail us. It turns out that the free society is most conducive to the pursuit of happiness – it is no accident that the American Founders concluded this after much consternation about what kind of society they ought to found. All this does not mean anyone can impose the vision of how to live on anyone else, not even that anyone is authorized to try. One’s business of learning how we ought to live is just that, learning it and doing it and, maybe, telling it here and there. Not making anyone conform. But once one has learned it, one can continue to try to figure out the implications of this view for social and political life. It turns out that the political framework that makes the pursuit of happiness possible, given how varied such a pursuit can be, also makes possible the rejection of that pursuit or the pursuit of unhappiness or godliness or cleanliness.

The Greater Promise of Neo-Aristotelianism In any case, the case for liberty inspired by the Aristotelian tradition of moral philosophy does seem to hold out philosophical promise. Aristotle’s naturalism, since it abandons the Platonic aspiration for timelessly fixed natures from which eternal standards of good and bad, right and wrong may be derived – not, however, deduced, because of what Hume has taught – gives us a reconceptualization of what is meant by ‘the nature of X’. This can produce a revitalized naturalistic ethics, one that resists Moore’s open question approach24 (for reasons too complex to spell out here and ones I have provided elsewhere).25 In terms of the best classification of what a human being is, what it is to be a human being, one can learn what a good rendition of such a being would come to, which, since human beings are free, responsible agents, provides us with norms we ought to apply in our own lives. Being distinctive as potentially rational animals, human beings are at their best if they actualize this potential to live rationally, to be reasonable. This is an ontological foundation for the ethics social compact theory only presupposes without argument. It implies, furthermore, that insofar as we are social beings, with the requirement of some type of community, we face the question as to what kind of community will be most hospitable to our task of living rationally, the right way. The standards of such a community are, among other ingredients, our basic, natural rights that all members will be required to observe and will not be able to violate with impunity. The social compactarian aspect of this way of thinking of political norms includes, most prominently, what an economist might call the ultimate exit option: no one may be coerced to remain part of the community, the consent of the governed is required for government to have authority to administer the (just) laws. But no one may ‘consent’ to governmental deeds that would be immoral to do in private. So what remains of social compact theory is, indeed, very important: namely, the value of freely reaching agreement among human beings, and the justifiability of certain kinds of efforts to reach it. For instance, when participants in the polis can and do in fact agree on how to maintain and preserve appropriate standards of

44

Liberty, Games and Contracts

social justice (as well as on who should administer the effort), something crucial to the lives of the participants is achieved – the efficient preservation of the conditions of proper social intercourse. The compact is, as it were, the will that is needed to apply the norms. Human beings, who have free will, do not always do what is right vis-à-vis their fellows; and to impress them with the need to abide by the proper norms, force is sometimes needed. Such force would be ineffective if the political will did not back it. If, as the liberal tradition holds, the proper norms provide individuals with the jurisdiction over their significant actions, they would be the first to be involved in political action, that being within the range of significant actions. To put the matter plainly, it is not sufficient to learn of the true principles of justice; it is also vital to apply them, and that would not be possible without widespread agreement – namely, a social compact or contract. But, here the compact is not the ultimate basis of social or legal norms, let alone of personal ones. Instead, it is implied by prior ones or some aspects of them – for example, individual sovereignty – to be the method by which political (meta-)norms can and will be upheld, interpreted and enforced in society. That method – in the Western legal tradition referred to as ‘due process’ – is itself always open to evaluation: that is how one can tell whether a system is corrupt or flawed in some other way. When consent secures the method to achieve due process, something enormously valuable happens: the prospect of success (because willing and possibly rational or prudent) in the maintenance of justice. Where social compact fails is as the first word about social and political norms. Social compact will not suffice as the grounding of how we ought to live in each other’s company – mere agreement does not establish what is socially or politically right. For that we also need to have a clear enough idea – even, perhaps, in only very general terms – of how we ought to live.26

Notes 1

This paper draws heavily on T. R. Machan, ‘Why Agreement is Not Enough’, Philosophia, 28/1–2 (2000). 2 In the latter instance they are versions of conventionalist meta-ethics. 3 Because a contract is, strictly, a legally enforceable agreement, social contracts – which are pre-legal arrangements – are better classified as compacts. I will, accordingly, refer to social compacts instead of social contracts through the rest of this discussion. 4 Indeed, even when one considers discussions such as this one, one of the first questions that comes to mind is, ‘Is the author right or wrong in dealing with this topic?’ It seems to arise quite apart from any agreement about standards. 5 Immanuel Kant, Metaphysical Elements of Justice, trans. John Ladd (Indianapolis, 1965), 80. 6 Immanuel Kant, Werke (Bruno Cassier, 1914), 6: 380–81 (my translation). 7 John Rawls, A Theory of Justice (Cambridge, MA, 1971), 36. 8 Hobbesian natural rights are, strictly speaking, non-normative – unlike Lockean ones – but a feature of the life-sustaining innate drives of those in the state of nature. Some will

Is Agreement Enough?

9 10 11 12 13 14 15

16 17 18

19

20 21

22

23 24 25 26

45

dispute this but at the price of removing a crucial element of morality, namely, the capacity to freely choose whether to take one or another action. Such ‘moral’ theories are actually at most theories of value or good, not theories of rightly or wrongly chosen conduct. For a survey, see Shirley Robin Letwin, ‘Modern Philosophies of Law’, The Great Ideas Today, 1972 (Chicago, 1972), 105–53. John Rawls, ‘The Independence of Moral Theory’, Proceedings and Addresses of the American Philosophical Association (Newark, DE, 1975), 5. Ibid., 5. Immanuel Kant, The Philosophy of Law, trans. W. Hastic (Edinburgh, 1887); reprinted in Great Books of the Western World (Chicago, 1952), 42: 436. Ibid., 398. Ibid., 419. T. R. Machan, ‘A Note of Independence’, Philosophical Studies 30 (1976): 419–22. Natural law and rights theories achieve what the veil of ignorance aims for – namely, the securement of moral equality – but only for the purposes of understanding political life. They focus on human nature, something we all share. Jan Narveson noted this in a personal email exchange (December 1996). Rawls, ‘The Independence of Moral Theory’, 21. Jan Narveson, The Libertarian Idea (Philadelphia, 1988). Another prominent social contract theorist is Tristan Engleheart, a specialist in bio-ethics and author of several books on the topic. Ibid., 131. Narveson appears to be talking here about ‘the principles of morality’, yet in the earlier mentioned personal email correspondence he states that ‘social contract theory … is quite obviously only aimed at interpersonal rules which are indeed ... “norms of community life”.’ Ibid. Personal (email) correspondence. Similar difficulties face pure democrats – if majority rule establishes the soundness of legal or political principles, how would one defend these principles to those about to embark upon participation in the democratic process? As an example consider Edward Pols’s book, Mind Regained (Ithaca, NY, 1998), which argues for a fundamental reconsideration of the scientistic, empiricist and behaviourist orientation concerning theories of mind as well as the foundation of human moral life. In particular, the work mounts a frontal attack on the belief that teleological explanations of human action are unfounded – a belief that lies at the source of accounts of the normative dimensions of human life in terms solely of drives and motives. Narveson’s personal email. G. E. Moore, Principia Ethica (Cambridge, [1903] 1983), 15 (ch. 1, sect. 13.1). See Tibor R. Machan, Individuals and Their Rights (LaSalle, IL, 1989). For a full treatment, see T. R. Machan, ‘Individualism and the Problem of Political Authority’, The Monist 66 (1983): 500–16 (which is chapter 7 of Machan, Individuals and Their Rights). Why, for example, would it be wrong to contract for the use of some third party who has not consented? Because that party has rights and, independently of any contractual determination, it is wrong to violate rights. The concept of ‘meta-norm’ is advanced by Douglas B. Rasmussen and Douglas J. Den Uyl, Liberty and Nature, An Aristotelian Defense of Liberal Order (LaSalle, IL, 1991). It means those norms the function of which is – in organized human communities or systems of justice – to enable others, which are directly action-guiding ones, to be freely practised.

This page intentionally left blank

Chapter 4

Does Scepticism Beget Libertarianism? A Response to Narveson on Reason, Morality and Politics1 Leo Groarke

Libertarianism for Today? We live in tough times for philosophy as it has been traditionally conceived. Today, many thinkers see the attempt to use reason to build a compelling theory of the world, the state and the good life as a naive hope which is as undesirable as it is unattainable. Instead of celebrating this grand project, the prevailing intellectual currents favour scepticism, relativism (of the traditional or the post-modern variety) and the attempt to fashion a world view which does not require a philosophical foundation. One might argue that libertarian political philosophy suits an age of difference and individualism, but there are ways in which it is reminiscent of the grand theories of traditional philosophy. Like other views which have captured the philosophical imagination of one day or another, libertarianism promotes a singular allegiance to a compelling idea which is proposed as the basis of an overarching philosophical perspective. As Narveson puts it: The moral and political outlook holding that individual liberty is the only proper concern of coercive social institutions, which has lately become known as libertarianism, is or at least certainly seems, in principle, a pure and therefore extreme view. It also has a certain appeal, to many of us at least – it lends itself to ringing proclamations and slogans such as animated many of the fathers of the American Revolution and are to be found in the political rhetoric of the day.2

In elaborating and defending libertarianism, Narveson has made a major contribution to current debates in political philosophy. In this chapter, I will examine his attempt to justify the libertarian perspective in his seminal volume, The Libertarian Idea. I will argue that its attempt to build a foundation for libertarianism cuts to the heart of the deepest problems of moral and political philosophy. I am less convinced by Narveson’s attempt to use these problems as a springboard for libertarianism. I will argue that contractarianism is the correct response to the problems he identifies, but not a contractarianism of the sort that he proposes.

47

Liberty, Games and Contracts

48

Can Reason Tame Political Dispute? Narveson’s defence of libertarianism is in part motivated by Nozick’s failure to provide a convincing foundation for the libertarianism he defends in Anarchy, State and Utopia.3 Similar concerns have been expressed by many commentators, and are famously encapsulated in Nagel’s description of Nozick’s theory as ‘Libertarianism without Foundations’.4 As Narveson points out, the problems with Nozick are symptomatic of broader problems with attempts to justify not only libertarianism, but also competing political philosophies. We can usefully understand these problems as a consequence of three principles I will label ‘The Rule of Reason’, ‘The Principle of Disagreement’ and ‘The Practicality of Morality and Politics’. The Rule of Reason The rule of reason proposes reason as the arbiter of moral and political dispute. This is a normative principle assumed by the very exercise of moral and political philosophy, which might be described as the attempt to apply reason to moral and political affairs. It is easy to overlook the significance of the rule of reason, but it has profound consequences, establishing reason and rational discourse as the appropriate mechanism for resolving the clash between competing points of view. In its absence, there are no grounds for objecting to other means of resolving disputes and conflicts. At best, this leaves their resolution to whatever machinations individuals could bring to bear. At worst, it makes violence and brute force the arbiter of conflict. The rule of reason is assumed whenever we give an argument in defence of a moral or a political point of view. In such a context, the rule imposes limitations on what we can believe and plausibly defend. Among other things, it demands that we adopt moral and political stances and judgments which are consistent with (i) each other and, so far as is reasonably possible, (ii) the ‘facts’ that surround particular circumstances. One of the strengths of Narveson’s treatment of concrete moral issues is his attempt to fulfil these demands (one might compare the failure of many philosophers to take (ii) seriously enough), but this is not a matter which needs detailed discussion here. In the context of Narveson’s attempt to defend libertarianism, it is more important to consider the implications of the rule of reason when one attempts to build a foundation for moral and political philosophy. The Principle of Disagreement Most philosophers attempt to justify basic principles of morality and politics (and thus satisfy the rule of reason) by appealing to intuitions about good and evil, right and wrong. As Narveson puts it, philosophers will say one of two things: either (1) that their principle is, when you think about it, just obvious, or self-evident; or (2) that the various smaller moral principles and

Does Scepticism Beget Libertarianism?

49

judgments that their principle systematizes into a unified bundle are themselves just obvious, and accordingly they may be used as evidence for any moral theory that supports them. (Often, indeed usually, both things are said.)5

Philosophers who believe that morals are founded on emotional rather than cognitive states (or some mix of the two) may argue that moral and political philosophies are founded on ‘sentiments’ rather than ‘intuitions’, but this is not of great importance. In a discussion of the foundations of moral and political philosophy, we can leave open the possibility that intuitions should be interpreted in a manner which has as much (or more) to do with affective as cognitive states. The important point is that intuitions are, however we understand them, a problematic basis for moral and political philosophy. The issues with intuitions can be explained in terms of a principle of disagreement which maintains that rationality cannot choose between the conflicting intuitions that characterize different individuals (not only different philosophers, but also different non-philosophers, citizens, activists, moral crusaders and so forth). It seems to follow that there are no shared convictions about morality, justice, equality and/or rights that could provide the kind of foundation needed to build a rational consensus about morality and politics. In the absence of a foundation of this sort, there seems no way to use reason to choose between competing points of view. Looked at from this perspective, the attempt to base libertarianism on intuitions about natural rights (as in Paul et al.)6 looks misguided, for it is based upon an appeal to a set of intuitions which libertarians do not share with others – utilitarians, socialists, egalitarians, communitarians, conservatives, and so on and so forth. This appeal to intuitions is no more problematic than appeals to competing intuitions others invoke (for example, the intuition that we are obligated to help those worse off), but this is of little consolation. Instead of vindicating libertarianism, it only intensifies the problems with intuitions. As Narveson puts it at one point, the appeal to intuitions cannot help us because intuitionists themselves ‘keep having different, and indeed contrary, intuitions’.7 In passing, it is worth noting that the issues raised by the principle of disagreement cannot be resolved by appealing to some methodology like Rawls’s method of ‘reflective equilibrium’. Its attempt to refine theoretical principles on the basis of specific moral judgments (and vice versa) moves each of us in the direction of a mutually consistent mix of principles and judgments, but there is no reason to believe – and good reason to doubt – that this movement will produce some kind of theoretical and practical consensus. Indeed, the persistence of conflicting moral and political perspectives (libertarian, utilitarian, conservative, communitarian, socialist and so on) suggests that a method like Rawls’s only turns competing intuitions into competing reflective equilibriums which remain fundamentally in conflict. Even if we reject extreme moral and political views as fanciful (a rejection it is difficult to rationally defend),8 appeals to intuitions seem unhelpful where they matter most – namely, in those practical contexts where we must deal with disputes about distributive justice, the rights of individuals, and issues such as abortion,

50

Liberty, Games and Contracts

capital punishment and so on. As Narveson writes, ‘disagreement is not very far from being the rule rather than the exception in practical matters. People have contrary opinions.’9 As long as we are committed to the rule of reason, the principle of disagreement seems to leave us with no way to settle political and moral dispute, and no way to establish one moral or political philosophy as better than any other. To this extent, it is difficult to avoid Narveson’s conclusion that: ‘My objection to appeal to intuitions in moral theory is, in brief, that when (not merely “if”!) intuitions conflict, we are bereft of conceptual tools for reaching reasoned agreement. Indeed, one must say that under these conditions “reasoned” agreement is impossible.’10 The Practicality of Morality and Politics The problems which are raised by the rule of reason and the principle of disagreement are exacerbated by ‘the practicality of morality and politics’ (a phrase that I adapt from Narveson).11 It requires some method of resolving moral and political disagreement, for the continuance of practical affairs requires that we operate within some generally accepted moral (and even more so, political) system. The practicality of morality and politics is clear if we compare other realms of argument and dispute. Outside of morality and politics, there are numerous contexts in which we find it easy to live with disagreement. In the context of food or art – contexts in which disagreement is notorious – disagreement seems innocuous and relatively benign. It does not matter greatly if you and I and others favour different kinds of food or different artistic genres. There is no compelling reason why our lives require the same personal preferences. This is disagreement we can live with. The situation seems very different in the case of morality and especially politics. Some moral and political disagreements are minor matters that are easily tolerated; but the same cannot be said of fundamental disagreements. Insofar as they determine what behaviours are and are not accepted, required, punished, enforced, legislated, and so on and so forth, the living of our lives requires some set of consistently applied principles. These principles must consistently distribute rights and obligations across the society we live in, but this cannot be done without some agreed upon moral and political framework. How then can we proceed? In attempting to establish a foundation for political philosophy, we seem to be caught between the three principles I have outlined. The rule of reason demands that we use reasoned discourse to establish a political perspective. This may allow us to dismiss some (irrational, inconsistent) political perspectives, but it will not provide a foundation for politics, for the principle of disagreement declares that rationality cannot choose between the conflicting moral intuitions of different individuals. There seems no way out of the box because the practicality of morality and politics does not allow us to accept or ignore disagreement. Many philosophers will reject this conclusion. It is difficult to exaggerate the force of those intuitions that give rise to moral and political philosophy (it is not too

Does Scepticism Beget Libertarianism?

51

much to say that wars are fought over the conflicting convictions they give rise to). In such a context, many appear to believe that they will be able to win over those who do not share their intuitions if they only have enough time to analyse, reflect and argue. As this is a context in which hope seems to spring eternal, the debates persist. I will not attempt to engage intuitionist philosophers in lengthy argument. Suffice it to say, a rational consensus over fundamental moral and political principles has not been established (among philosophers or non-philosophers), and there seems little reason to expect it soon. Even if we leave open the possibility of some new consensus in the future, what we need now is some perspective which recognizes the fundamental disagreement we must live with in the meantime. In such a context, it seems more profitable to follow Narveson and frankly recognize the problem with the usual attempts to ground moral and political philosophy. It is in view of this that he looks for some other way to establish a foundation for political theory.

Discarding the Baby with the Bathwater? Having rejected the appeals to intuitions which are the basis of most moral and political philosophies, Narveson proposes a contractarian perspective. In view of the problems with moral intuitions, he favours a contract founded on rational selfinterest. As he puts it, Why accept the contractarian view of morals? Because there is no other view that can serve the requirements: namely, of providing reasons to everyone for accepting it, no matter what their personal values or philosophy of life may be, and thus motivating this informal, yet society-wide ‘institution.’ Without resort to any obfuscating intuitions, or ‘self-evident rights’ and the like, the contractarian view offers an intelligible account both of why it is rational to want a morality and of what, broadly speaking, the essentials of morality must consist in: namely, those general rules that are universally advantageous to rational agents.12

Narveson’s approach to morality replaces an appeal to intuitions with a contract between individuals who make rational choices – such choices understood in the economic way, as a decision to pursue the maximum satisfaction of one’s own preferences. Because this conception of rationality does not pick between preferences, Narveson presents it as a neutral conception which can be ascribed to everyone. Following Gauthier,13 he argues that the rational individuals in his contract situation will be ‘constrained maximizers’, and that the constraints that this implies can be the basis of a workable conception of morality, and ultimately libertarianism. In responding to Narveson, I will put aside questions that might be asked about the details of his attempt to derive morality from rational self-interest (for example, Ripstein’s charge that this kind of derivation illicitly depends on moral premises).14 The fundamental question I want to ask is whether individuals who have convictions founded on strong moral and/or political intuitions should feel

52

Liberty, Games and Contracts

compelled to accept the conditions of the contract-making situation (let’s call it ‘the original position’) which Narveson proposes. Should his critique of moral intuitions convince them that they should reject intuitions and found morality – and ultimately politics – on the conception of reason which is the basis of Narveson’s contractarianism? If Narveson’s reasoning is convincing, then there may seem to be nothing significant at stake for those with libertarian intuitions. For their intuitions align with the political principles that are the outcome of Narveson’s deliberations. That said, they are not true Narvesonians if this is what drives their willingness to accept his conclusions. Ringing expressions of their intuitions about freedom and individual rights – standard fare in most defences of libertarianism – are contrary to the spirit of the Narveson brand of libertarianism, which is founded on a rejection of the significance and relevance of such intuitions. Looked at from this point of view, one must wonder if Narveson’s original position (with its attendant notion of rational self-interest) is in keeping with libertarianism as it is usually conceived. In such a context, one might ask what most libertarians would do if one could show that the attempt to generate morality from rational self-interest did not produce libertarianism.15 If reason so conceived favoured a communitarian perspective and the welfare state, or no moral or political position whatsoever, would they find the Narvesonian project attractive? This is a highly speculative question, but the persistence of libertarianism in many authors who do not accept the Narversonian project (for example, those in Paul et al.) suggests that they would not in such circumstance give up the intuitions which make libertarianism attractive. And this suggests that it is these intuitions, not an appeal to rational selfinterest, which is the real basis of libertarianism. But these are relatively minor issues in comparison with those that the Narveson project raises for individuals who have moral intuitions at odds with those that motivate libertarianism. For them, there is a great deal riding on the decision whether they will reject these intuitions and commit themselves to the Narvesonian project. In light of conflicting intuitions forwarded by other individuals, it asks them to give up their commitment to their deepest convictions, to endorse in their place a narrow notion of self-interest (a notion which is in many cases antithetical to their intuitions), and to follow Narveson to his conclusions about morality and politics. This is asking a great deal, especially as the notion of rationality which Narveson proposes as a substitute for moral intuitions seems so artificial. In his discussion of Rawls, Narveson himself is impatient with Rawls’s original position because it seems so artificial. As he puts it in a remark on the veil of ignorance: There is a crucial question about what is intended in imposing such requirements. … Nobody is or ever could be literally behind any such ‘veil’; so we would seem to have to ask what the relation is between ‘a person’ behind the veil – where, after all, it is quite unrecognizable as a person in any ordinary sense of the word – and that ‘same’ person on the real-life side. Namely, the question would be why the real person should pay any attention to what the idealized ‘person’ has, from its Olympian perch, ‘decided’.16

Does Scepticism Beget Libertarianism?

53

Narveson’s own original position does not invoke the artificial veil that Rawls proposes, but it is open to a similar criticism. In our real lives, there is no way to neatly separate the different components of our world views and isolate a separate conception of reason or self-interest that we are committed to. Instead, individuals (at least most individuals) operate with robust conceptions of reason and interest that are not separable from their fundamental intuitions and convictions. What we view as rational or in our interest (whether we view it as rational or in our interest to cooperate with some other individual or group, for example) is closely tied to a variety of complex views and intuitions. At their centre one finds beliefs and intuitions about who we are as individuals and human beings, the nature of the world and the good life, and our relationships with others. Someone with the sorts of (moral and non-moral) convictions this implies cannot concede that a much narrower notion of rational self-interest should be the arbiter in the original position without discarding essential beliefs and intuitions which make them who they are. In the process, they turn themselves into something less than the full-bodied individuals they are. Someone opposed to the Narveson project will argue that it in this way turns them into someone they are not, a process which raises the question why they should accept the contract that ‘same’ person agrees to in Narveson’s contract situation. The conception of the individual that Narveson adopts assumes at its centre a commitment to economic rationality. This commitment may, within a particular individual, be combined with any variety of preferences, including preferences that reflect the moral convictions they have (regarding the general good and so on). This is why the notion of rationality Narveson adopts is allegedly neutral, but it implies a conception of the individual which is not neutral and does not accurately represent the beliefs and convictions of many individuals. Why? Because it takes core commitments about what is good and reasonable which they are not willing to give up, and turns them into something else – namely, non-core preferences which are presented as an elaboration of a conception of economic rationality they reject. The force of this objection to Narveson can be gleaned from Sandel’s critique of Rawls, where he argues against the ‘independent self’ that Rawls invokes in his attempt to provide a contractarian basis for his liberalism. As Sandel puts it at one point, we cannot regard ourselves as independent in this way without great cost to those loyalties and convictions whose moral force consists partly in the fact that living by them is inseparable from understanding ourselves as the particular persons we are – as members of this family or community or nation or people, as bearers of this history, as sons and daughters of that revolution, as citizens of this republic. Allegiances such as these are more than values I happen to have or aims I ‘espouse at any given time’. They go beyond the obligation I voluntarily incur and the ‘natural duties’ I owe to human beings as such. They allow that to some I owe more than justice requires or even permits, not by reason of agreements I have made but instead in virtue of those more or less enduring attachments and commitments which taken together partly define the person I am. … Where the self is unencumbered and essentially disposed, no person is left for self-reflection to reflect upon.17

54

Liberty, Games and Contracts

One might criticize Narveson’s attempt to deduce morality from a narrow conception of economic reason in a similar way, for it does not accept individuals as they are. Instead, he garners agreement by stripping away key aspects of their person and identity which define them, most notably their conceptions of good and reasonable. Instead of accepting the Narvesonian rejection of moral intuitions, why shouldn’t individuals with intuitions which run contrary to his conception of rational selfinterest reply that they prefer to remain who they are? In answer to his contract, they may say that they cannot be obligated to uphold a contract which ‘they’ would never have agreed to. Narveson can retort that this resurrects the problem of disagreement, but this does not resolve the problem they highlight. At best, it seems to leave us with the need to choose between two unsatisfactory positions: (1) a Narvesonian solution to disagreement which is negotiated by individuals who have been stripped of their fundamental beliefs and identity; and (2) individuals who retain their fundamental beliefs and identity, but no solution to the problem of disagreement. This seems a bleak choice. Instead of resting content with it, we will do better to look for some alternative way to build a basis for politics.

A Full-bodied Contractarianism? In trying to respond to the foundational issues that Narveson identifies, it may help to note that the moral disagreement he emphasizes is similarly emphasized in moral scepticism. Its best known contemporary proponent is Mackie,18 but similar views have a long tradition that stretches back to ancient times. Like Narveson, the sceptics argue that we cannot choose between different religious, moral and political codes, because there is no way to choose rationally between the different judgments and assumptions they are based on. One could easily present Narveson’s arguments as an instance of the standard modes of arguments developed by the ancient sceptics.19 On the basis of the considerations that inform the principle of disagreement, the most famous school of ancient scepticism, Pyrrhonism, argues that we should ‘suspend judgment’ (practise epoche) on moral and political matters. In answer to the question how one should live in the wake of such suspension, it proposes a ‘practical criterion’ which guides those of sceptical inclinations in their practical affairs. In matters of piety, morality and politics, it recommends that one live by the customs and conventions of the society in which one lives. Putting aside some significant problems with the Pyrrhonians’ commitment to convention, it is more important to recognize a key difference in the way that Narveson and the sceptics respond to conflicting intuitions about right and wrong. For sceptical arguments do not convince the sceptics that they should reject moral intuitions. Like other kinds of scepticism, moral scepticism leads to the conclusion

Does Scepticism Beget Libertarianism?

55

that certain beliefs cannot be justified, but it is only in straw man parodies of scepticism that this is taken to show that such beliefs should be rejected. Traditionally, scepticism does entail some modesty about our fundamental beliefs (which accompanies the conclusion that they cannot be proven true), but it does not lead to the conclusion that they should be rejected.20 The standard response to sceptical conclusions is what one finds in Hume, who concludes that we cannot justify our belief in the external world or cause and effect, but proposes that we accept these beliefs nonetheless.21 Not because these beliefs can be proven true (they can’t) but because this seems to be our ‘animal nature’ and any rejection of these beliefs seems both impossible and incompatible with a life worth living (as one will discover if one attempts to doubt the existence of a heavy falling object as it drops toward one’s toe).22 In the present context, the important point is that the sceptic’s traditional response to scepticism – which suggests that we accept foundational beliefs without a foundation – can be applied to the moral problems Narveson discusses. The result of doing so suggests a different response to the problems raised by conflicting intuitions. For though the sceptics hold that there is no way to justify moral intuitions, they may still propose that we accept them. In answer to the question why, they may answer that they are an indelible part of who we are and that we cannot imagine a good life without them. In such a context, one might compare moral intuitions to aesthetic intuitions, which are also characterized by deep disagreement. Though the stakes may not be as high, the conflict between competing aesthetic intuitions seems even more pronounced than the conflict between competing moral intuitions. Such differences are exacerbated by enormous differences that tend to characterize different cultural groups and subgroups. In consequence, there seems no way to rationally establish some set of aesthetic intuitions as correct. Does it follow that we should react in the way that Narveson does when he is faced with conflicting moral intuitions? This suggestion seems implausible. Whatever we decide about aesthetic intuitions, it would seem to be a mistake – and beyond this quite impossible – to stop the aesthetic endeavours that define our lives. Vermeer, Mozart and our favoured pastimes cannot be so easily discarded. Narveson seems to suggest that the practicality of morals makes a similar attitude to morality impossible. But is this so? It is true that our lives, and especially our interactions with others, must operate within some kind of consistent moral and political framework, and that this requires some agreement. In an era like our own – one characterized by diversity and individualism23 – we cannot take refuge in the ancient sceptics’ suggestion that we follow the customs and conventions of society.24 But does this mean that there is no way out of the apparent impasse? The case of aesthetics is instructive here. At first blush, it may seem that we can simply live with aesthetic disagreement, but this is not always so. It does seem more possible to live with aesthetic as opposed to moral and political disagreement, but it would be a mistake to think that we never need to negotiate fundamental disagreement in the area of aesthetics. In many mundane circumstances, this is precisely what we are forced to do. The members of a board responsible for a civic

56

Liberty, Games and Contracts

performance theatre have to negotiate different preferences in deciding which performers they will feature. A married couple who decide to redesign the interior of their home will need to negotiate different tastes in colour, furniture and art. And politicians and planners have to negotiate enormous differences in aesthetic preferences in creating and designing public spaces. Like the kinds of moral problems Narveson discusses, these problems can often be portrayed as prisoner’s dilemma situations. For they are cases in which individuals (the members of the board choosing performers, the Mr and Mrs painting their home and so on) must decide whether they will cooperate with one another, or push for their own preferences. And though they prefer the latter, a decision by everyone to similarly push for their own preferences will result in stalemate or no action whatsoever. In these kinds of circumstances, disagreements between competing aesthetic intuitions do not convince us that we should reject these intuitions. Instead of stripping them away, we attempt to negotiate these differences. In doing so, we tend to rely on a notion of ‘fairness’ which is rooted in epistemological considerations. We may say, for example, that certain differences in colour preference are subjective,25 and that there is no way to objectively prove that one is better than another. In such a case, a decision to let reason guide our choices suggests that we should treat these preferences as equal (because there is no rational way to pick between them). In such situations, rationality dictates equal treatment and we look for ways to accommodate such differences – by looking for options which different individuals can live with, by looking for ‘fair’ compromises, or by trying to allow for multiple alternatives that suit different tastes. In the realm of aesthetics, we accept individuals as full-bodied individuals, accept their aesthetic intuitions and look for ways to negotiate their differences. In the realm of morality and ultimately politics, this raises the question why we would not respond to differences in a similar way. There is more, not less, reason to do so, for those intuitions which are the basis of our moral and political views seem to be an even more integral part of who we are. Given the central role they play in our identity and our lives, why not maintain our commitment to intuitions – suitably qualified by the need to recognize that other people have different intuitions – and attempt to negotiate these differences? Such considerations suggest contractarianism, but a different kind of contractarianism than Narveson proposes. The ‘full-bodied’ contractarianism it countenances envisages an original position that incorporates individuals who maintain their moral intuitions, and attempts to negotiate the conflicts that result. The attraction of this kind of contractarianism is its willingness to accept (with some minimal limitations I will outline) people as they are. Unlike Rawls or Narveson, it does not ask individuals to drop their cherished moral intuitions as they walk through the door to the original position. Like Narveson, it grants the disagreement that characterizes such intuitions, but it does not attempt to deal with it by stripping away the things that make them who they are. In the context of questions about libertarianism, it raises the question whether a full-bodied contractarianism leads to libertarianism, or some other point of view.

Does Scepticism Beget Libertarianism?

57

From Full-bodied Contractarianism to Narveson or Rawls? In comparison with other kinds of contractarianism, full-bodied contractarianism eliminates much of the artificiality which tends to characterize the contractarian tradition. For full-bodied contractarianism does not ask us to speculate about circumstances in a distant state of nature. Unlike Rawls, it does not, under the guise of a veil of ignorance, put individuals in an impossible situation which is difficult to speak to. Unlike Narveson, it does not attempt to strip them of the core values they believe in. Much more than these kinds of contractarianism, full-bodied contractarianism takes whole-bodied individuals and considers how they might reasonably negotiate the fundamental differences that separate them in the moral and political arena. This is not meant to suggest that full-bodied contractarianism takes individuals exactly as they are. In real life, contracts and negotiations take place in circumstances where negotiations are influenced and distorted by prejudice, manipulation, deception, emotional appeals and an incomplete (and often purposefully incomplete) knowledge of the facts that surround the circumstances they address. In imagining a contract that establishes a reasonable basis for a political framework, we must imagine a contract that is negotiated in circumstances dictated by reason, that is, one characterized by full relevant knowledge and a rejection of inconsistent reasoning and points of view. In this way, the attempt to deduce a fullbodied social contract is to be governed by the rule of reason. This is not the place for a detailed attempt to establish what contract would emerge from the original position imagined by full-bodied contractarianism. But it would be amiss not to say something about the likely consequences. To begin with, full-bodied contractarianism suggests a clear distinction between our personal moral and political convictions and the political framework under which we live. It accepts that individuals are committed to particular moral and political convictions and does not attempt to eliminate them. But it also holds that the political framework which will regulate our lives is not to be established by these convictions, but in the face of disagreement, by negotiations between individuals who have conflicting values. In this way, full-bodied contractarianism accepts the intuitions of individuals but does not allow a direct route from intuitions to some political framework which will regulate our lives. In the final analysis, it makes a contract between individuals, not the intuitions that they cherish, the arbiter of moral and political conflict. There are good reasons for thinking that full-bodied contractarianism will produce a contract which enshrines some kind of ‘liberal’ political framework (that is, one which promotes liberty in some sense), but even a cursory look suggests that the negotiations it imagines will not produce a libertarian frame of reference. If we imagine whole-bodied, reasonable individuals with moral and political perspectives that are internally consistent – but inconsistent with each other – it seems to me more likely that this favours a framework which is more similar to Rawls’s two principles of justice (understood in a way which places some limits on liberty).26 One might even argue that the full-bodied contractarian approach provides a route

58

Liberty, Games and Contracts

to Rawlsian liberalism which does not require the very dubious and artificial mechanisms he invokes in order to support his views. In sketching the contract that is likely to be the outcome of full-bodied contractarianism, it may be helpful to imagine the following three different ‘stages’ of deliberation that take place within their original position. Stage 1 In Stage 1, the introductory stage of contract negotiations, individuals introduce themselves and their points of view. In agreeing to participate in contract deliberations, they must agree to abide by the dictates of reason and rational deliberation. At the very least, this implies that they must put forward views that are rational (consistent and so on) and be prepared to accept the conclusions of rational deliberations. Stage 2 In Stage 2 of deliberations, individuals bargain a social contract. Let us place questions of freedom of thought and expression at the top of their agenda. What are they likely to decide in this regard? Zealots of all sorts may enter negotiations with a desire to impose on others their own conception of what can be said and promoted, but there seems to be no way for them to justify such restrictions on freedom of thought and expression. For they are bound to rational deliberation and there seems no way in which they can prove that their convictions and world view are superior to the alternatives endorsed by others (as Narveson shows, this cannot be done by the normal appeal to intuitions, and there seems no viable alternative). Within such a context, it appears that reasonable individuals must agree to a social contract which grants individuals as much freedom to express and promote their values as is compatible with a similar freedom for others. This suggests that freedom of thought and expression will be one of the defining principles which characterize the contract reached in the original position envisaged by full-bodied contractarianism. Stage 3 Individuals in the proposed original position emerge from Stage 1 with few constraints on their views (the only constraint being that they must agree to reject inconsistent points of view). Insofar as a commitment to freedom of thought and expression emerges from Stage 2, it introduces some constraints on the views they can reasonably uphold (undermining views that do not respect such freedom). In Stage 3 we must imagine them deliberating further. It seems reasonable to suppose that the diversity of opinion which still characterizes the original position will push such individuals towards liberty, but it is especially difficult to say how far this will extend in the realm of distributive justice, where the locus of the dispute between the libertarian and others lies. It seems fair to say that negotiations in the original position will in this case be characterized by great conflict. Egalitarians will favour equality (even ‘from each

Does Scepticism Beget Libertarianism?

59

according to his ability, to each according to his need’), while others will reject the notion that we have obligations of this sort and favour a system which allows individuals to pursue their own ends in a way that allows for great differences in wealth. In such a situation, Narveson and other libertarians may argue that the perspectives of different individuals in the bargaining position are not equally defensible because some (but not others) force others to do things they do not agree to. It is easy to see why this distinction is of great importance from a libertarian point of view, but it will not convince the egalitarian, who will answer that constraints on others are obviously legitimate in circumstances in which some marginal utility must be given up in order to address situations of great need. In such circumstances, it seems to beg the question if we accept either the libertarian or the egalitarian position. It is by rejecting partial (as opposed to impartial) positions of this kind that full-bodied contractarianism must recognize the force of different points of view.27 If appeals to libertarian considerations cannot determine the outcome of bargaining in such circumstances, what contract could parties reasonably agree to? This is a difficult question in a situation in which it is not possible to accommodate both views, and it is difficult to see why either side should capitulate. Assuming we do not give up on the attempt to reach agreement, we need to search for some reasonable compromise. While there may be a variety of imaginative ways to respond to this situation (indeed, this is something worth exploring), I will end these comments by noting that one response which holds promise would be a principle like Rawls’s difference principle – namely, a principle which justifies economic and social inequalities if and only if they benefit the worst-off members of society. This is an attractive principle because there is a sense in which such a principle occupies a middle point between pure equality and completely unregulated differences in wealth. It permits differences, and possibly very large differences, but it places limits on them. It does so by seeking differences which are acceptable to the egalitarian as well as the libertarian. In this way it supports differences which should be acceptable to all, though the libertarian prefers to go much further. The problem is that any further differences will, in contrast, be a matter of great controversy. In view of this, it is difficult to see how they can be justified. These cursory remarks are in no way definitive, and cannot replace a long and detailed examination of what is rational in the original position postulated by fullbodied contractarianism. That said, it is difficult to see how a libertarian perspective can be defended as the reasonable outcome of such negotiations. If this is right, then my discussion suggests that Narveson’s concerns about intuitions should lead us to contractarianism, but to a full-bodied contractarianism, and that it is doubtful that this will produce the libertarianism that Narveson defends.

A Third Way: An Alternative Route to Libertarianism? In closing, one other route to libertarianism bears mention. I have argued (with Narveson) that it cannot be convincingly founded on an appeal to intuitions about

60

Liberty, Games and Contracts

natural rights. And I have argued (against Narveson) that the contractarian deduction of morality and then libertarianism from rational self-interest does not work. There is, however, another route to libertarianism which reflects a theme in Narveson and many libertarian supporters. It is the notion that government (and especially big government) is inept. Adapting a famous quip from Callimachus, one might capture this aspect of libertarianism under the maxim ‘A big government is a big evil’.28 This is not the place for a detailed account of this aspect of libertarianism, but it is not difficult to see how it might become a (quasi-empirical) basis for an argument for libertarianism. According to this argument, a libertarian perspective recommends itself because the kind of government and regulation which is the alternative is an inept and dubious way to secure even the goods that the egalitarian wants. Critiques of government along these lines argue that the free market is a much better mechanism for addressing the needs of individuals; that governments operate without the competition which makes the market efficient; that intervention in the market undermines its ability to reflect the real costs of production; that the privileged position of authority granted individuals in government is easily abused; that governments promote self-serving bureaucracies; and that there is no unproblematic way to control government (and especially big government) and ensure that it does not interfere with, rather than promote, the very ideals it supposedly embraces.29 This is an argument for libertarianism which depends on an empirical claim: that government, and especially big government, is, in comparison with the workings of the market, inefficient and unproductive. This is not an argument that is attractive to everyone (and even to all libertarians),30 but it merits note. It is difficult to know how to definitively answer a broad empirical question about government regulation, but it would be naive to deny that this argument has some force. In favour of it, it is not difficult to find many compelling examples of misguided and unproductive regulation (of course, the collapse of the Soviet economy may be the most spectacular example). I will not attempt to answer this question here, but I will end by noting that such issues have important repercussions that are often ignored by philosophers. Even if one can, for example, justify a difference principle in the realm of distributive justice, this leaves open the question how such a principle is to be put into practice. Philosophers tend to treat this as a question about the application of a philosophy rather than a question for philosophers. But it is a question we cannot ignore if our hope is not just theory, but a viable political framework which can be instituted. So far as I am aware, Rawls does not explicitly address this issue. But it is natural to suppose that he would institute it via a welfare state which would redistribute wealth in the way the difference principle dictates (a more radical alternative would imagine a planned economy, planned along these lines, with direct government control over the distribution of wealth). But this implies a mechanism of government regulation which must be able to (i) determine what the difference principle applies in practice (who the worst members of society are, precisely what differences will most benefit them, a system for identifying these differences across

Does Scepticism Beget Libertarianism?

61

a bewilderingly complex economy and so on); and (ii) redistribute wealth accordingly (in a manner which will be efficient, not open to unreasonable abuse and so on). In such a context, one could imagine someone arguing that the best mechanism for implementing the difference principle is not government regulation but the free market. They might thus argue that its response to demand, its emphasis on efficiency through competition and so forth will best approximate what differences are justified. If this can be argued, then something peculiar seems to happen: in the realm of practical affairs, the distinction between the difference principle and libertarianism seems to disappear. For in that case, one may argue for the difference principle but then conclude that the free market is the mechanism by which it should be implemented. This is a third way to libertarianism. Is it plausible? My own view (which I have developed elsewhere) is that it exaggerates the virtues of the market, which needs (minimal, but nonetheless significant) regulation if it is to function well.31 That said, there is much in Narveson’s account of concrete political philosophy – and his appreciation of the need to mix political philosophy with empirical considerations of this sort – that is to be commended. But these are aspects of his political philosophy that lie beyond the scope of the present paper.

Notes 1 2 3 4 5 6 7 8

9 10 11 12 13 14

I am indebted to Malcolm Murray for trenchant comments on an earlier version of this chapter. Jan Narveson, The Libertarian Idea (Philadelphia, 1988), xi. Robert Nozick, Anarchy, State, and Utopia (New York, 1974). Thomas Nagel, ‘Libertarianism without Foundations’, in Jeffrey Paul (ed.), Reading Nozick: Essays on Anarchy, State, Utopia (Totowa, NJ, 1981). Narveson, 109. Ellen Frankel Paul, Fred D. Miller, Jr and Jeffrey Paul (eds), Natural Rights Liberalism from Locke to Nozick (New York, 2005). Narveson, 112. One might argue that the intuitionist approach does not have the rational resources to reject even the most extreme views so long as their adherents are willing to maintain their commitment to their own intuitions (the Marquis de Sade does have his adherents). One can at best say that one intuits the matter differently and there is no unproblematic way to pick between intuitions. The usual attempt to reject certain intuitions as abnormal or absurd simply begs the question, for they are judged so just because they are so much at variance with our own. Narveson, 120. Narveson, 122. Narveson, 118. Narveson, 148. David Gauthier, Morals by Agreement (New York/Oxford, 1986). Arthur Ripstein, ‘Foundationalism in Political Theory’, Philosophy and Public Affairs, 16/2 (Spring) 1987: 115–37.

62 15

16 17 18 19

20 21

22

23 24

25

26

27

28 29 30

Liberty, Games and Contracts I think there are avenues worth exploring here, but I will not pursue them now. (It is, for example, notable that Hellenistic philosophy attempts to derive morality from selfinterest, but reaches a much more robust notion of morality than Narveson does.) Narveson, 132. Michael J. Sandel, ‘Justice and the Good’, in Michael J. Sandel (ed.), Liberalism and Its Critics (New York, 1984), 172. J. L. Mackie, Ethics: Inventing Right and Wrong (Toronto, 1977). I will not undertake that exercise here, but note, in particular, the ‘five modes’ that the second-century Greek philosopher Sextus Empiricus attributes to ‘the more recent skeptics’. See Benson Mates, The Skeptic Way: Sextus Empiricus’s ‘Outlines of Pyrrhonism’ (New York, 1996), 110–11; and Julia Annas and Jonathan Barnes, The Modes of Scepticism: Ancient Texts and Modern Interpretations (New York, 1985). On this aspect of ancient scepticism, see Leo Groarke, Greek Scepticism: Anti-Realist Trends in Ancient Times (Montreal, 1985). Of course, Hume thinks he is different from the Pyrrhonians in this regard. But he has, as so many commentators point out, a peculiar view of Pyrrhonism (so peculiar that many have wondered how he could have been so badly misinformed). As Malcolm Murray has pointed out to me, Hume’s natural beliefs manoeuvre works only for the kinds of beliefs that nature really forces upon us, and this opens the question whether ethical beliefs – or possibly which kinds of ethical beliefs – are beliefs of this sort. Trends that are continually promoted by globalization’s spread of diverse traditions, cultures and subcultures. In our context, the idea that we can even delineate the customs and conventions of society is problematic, for we live in societies that seem made up of individuals and subgroups that favour different, and often radically different, customs and conventions. I say ‘some’ because this is not a situation where any colour preference is as good as another – there is a notable difference between choices made by those who know and study colour, and those who don’t. The problem is (as in the case of morality) that this is a situation in which knowledge and study do not fully eliminate differences in taste and perspective, but only reduce them. I am not sure that Rawls’s two principles, explained as they are in A Theory of Justice (Cambridge, MA, 1971), are inconsistent, though they are sometimes said to be so. In order to avoid possible confusion, we might simply note that Rawls alters the liberty principle in Political Liberalism (New York, 1993) to make it clear that it is restricted by principles of welfare. This result captures the grain of truth in the common suggestion that scepticism promotes tolerance (though it is important to say that it promotes tolerance only to the extent to which one retains a commitment to the rule of reason). Callimachus was an ancient Greek poet who was associated with the famous library at Alexandria. The quip in question is sometimes translated as ‘A big book, a big evil’. See, for example, Friedrich Hayek, The Road to Serfdom (Chicago, 1944). See, for example, Tibor R. Machan, ‘The Case for Libertarianism’, in Craig Duncan and Tibor R. Machan (eds), Libertarianism: For and Against (Toronto, 2005), 11–12: ‘… libertarianism is sometimes advanced on the grounds that when people enjoy the conditions of negative liberty, everybody makes the best judgments and the most beneficial consequences all around are realized. Such a consequentialist approach is unwise, since it suggests that free men and women always do the right thing while those

Does Scepticism Beget Libertarianism?

31

63

who would regiment them seldom do.’ This is, of course, something of a straw man argument. For the reasonable proponent of libertarianism does not argue that in conditions of liberty ‘everybody’ makes the ‘best judgments’ and that the ‘most beneficial consequences all around’ are realized. All the libertarian needs to argue is that, on balance, the market works better than government regulation, and this is a different thing entirely. The market, it may be granted, is quite clumsy – the question is whether government regulation is more so. See Leo Groarke, ‘Can Capitalism Save Itself? Some Ruminations on the End of Capitalism’, in John Bishop (ed.), Ethics and Capitalism (Toronto, 2000).

This page intentionally left blank

Chapter 5

Contractarian Education Chris Tucker

Contractarian efforts to produce a motivating justification for normative systems have come up against a brick wall: the compliance problem. Notwithstanding that it may be rational to agree to be bound to a set of normative constraints upon otherwise utility-maximizing behaviour, it does not seem to be rational to actually constrain behaviour. Its intractability seems particularly obvious given the contemporary tendency to reduce the compliance problem to the prisoner’s dilemma. The definition of a prisoner’s dilemma strictly implies that in any such game defection dominates over cooperation. To attempt to prove that this is not so seems a Sisyphean task. What, we may then ask, continues to attract theorists to this particular device? The obvious answer is that the contractarian enterprise seems uniquely suited to providing a reasoned justification for normative systems capable of motivating people to action. This is both conceptually desirable and of significant practical importance.1 Appealing to practical rationality allows for the possibility of producing a non-question-begging justification for morally constrained action. But in light of the compliance problem’s seeming intractability, it may be thought that this answer is insufficient. The compliance problem is implied by an account of practical rationality. Given the resilience of the compliance problem, perhaps the correct response is to give up on the justificatory enterprise, despite its initial attractions. I propose that the resolution of this dilemma lies not in giving up on contractarianism, but in reconceptualizing the compliance problem. Given the aim of providing a justification for normative systems by way of appeal to practical rationality, the traditional compliance problem is impossible to solve without question-begging. But happily it is also unnecessary to confront. A rational agent, the ultimate guide for contractarian accounts, prefers the attainment of an efficient level of compliance to full compliance, and the compliance problem shows only the impossibility of full compliance. This recasting of the compliance problem, combined with giving up the (usually) implicit assumption that a contractarian analysis’s scope is with adult populations, suffices to allow us to overcome the traditional compliance problem and to focus on developing a future citizenry (mostly) committed to following the dictates of the embraced normative systems. Giving up on providing reasons for all agents to

65

66

Liberty, Games and Contracts

prefer a system of rules overcomes the problem of question-begging when identifying the rules in question. Turning our attention to the intergenerational nature of populations allows contractarians to overcome the problem of existing rational agents resisting direction from the normative rules. This response to the compliance problem is partially relied on by Jan Narveson, but the importance of this recasting for the contractarian enterprise has not been thoroughly elucidated. I proceed in two parts: what follows immediately examines the attraction of contractarianism, the centrality of the compliance problem, and argues both that traditional solutions cannot solve the compliance problem, and that contractarians should not care about the compliance problem as it stands. I then proceed to argue for a shift in focus for contractarian analyses, the result of which is that the difficulty with providing motivation largely disappears.

Traditional Contractarianism and its Implications 1. The Allure of Contractarianism Jan Narveson, in The Libertarian Idea, embraces contractarianism due to intuitionism’s contingent failure to produce moral rules that we can all embrace, and its contingent failure to produce normative claims which move us to act. Narveson notes two features of the practice of morals central for our purposes. The first is that morality is commonly understood as universal in scope: when we make moral claims, we are making claims that we feel should bind the behaviour of everyone. The second is that we think of morality as internally motivating: we think that awareness (or acceptance) of these claims ought to move us to act. Given disagreement over the content of moral claims, embracing intuitionism leaves us ‘bereft of conceptual tools for reaching reasoned agreement’.2 Agreement is crucial because of the coordinative function of morality; moral rules are to affect everyone’s behaviour, permitting, requiring and forbidding the same actions for everyone in similar circumstances. Reasoned agreement is necessary because only reasons motivate; everyone agreeing about something does not yet move people to action. In the face of our own (or others’) purported interests morality needs to provide us with motives to act in accordance with its dictates. Any analysis of morality that does not include this feature of moral systems fails to adequately capture what we commonly think morality entails. ‘If you say X is [morally] wrong, and profess indifference, we are puzzled.’3 And so, given the practice of morality, if we are to give any satisfactory justification for the requirements of morality, a plausible suggestion is that we need to appeal to interests that we all have.4 In no other way can we motivate everyone to act in accordance with the dictates of morality. Given disagreement, appealing to our basic intuitions gives us neither agreement nor motivation, and so cannot be an adequate justification of morality. David Gauthier, in his landmark Morals by Agreement, is concerned to provide a justification of morality that reconciles two apparently contradictory characteristics

Contractarian Education

67

commonly attributed to morality. On the one hand, we commonly suppose that the point of morality is to curb people’s pursuit of their own interest. Morality acts as a constraint on our possible courses of action. As such it seems that it must conflict with (and override) our preferred course of action. On the other hand, Gauthier observes that morality must be rationally justified if it is to have any real-world significance. ‘What theory of morals … can ever serve any useful purpose, unless it can show that all the duties it recommends are also truly endorsed in each individual’s reason?’5 Most problematically, Gauthier thinks that the correct account of rationality is a utility-maximizing conception, which the orthodoxy identifies as based on each agent’s preferences.6 As the orthodoxy links rational choice directly with preference, and Gauthier wishes to maintain a conflict between preference and the dictates of rationality, he is led to produce a heterodox account of practical rationality, to which we will return momentarily. Gauthier’s rational justification of morality as rational agreement must provide sufficient reasons for the acceptability (to each) of a set of social rules. One necessary reason is that there be some benefit to be had from conforming behaviour to the proposed set of rules. More problematically, another is that the agents actually expect these benefits to be realizable. This requires that agents actually act in accordance with the proposed set of rules. This is where contractarian accounts get into difficulty, as it is easy to imagine situations where immoral behaviour would maximize preference satisfaction when morals are understood to stand in opposition to such satisfaction. It is, however, of utmost importance to the enterprise that rationality will counsel compliance with the mutually beneficial terms. As Gauthier puts it, we defend compliance … with practices that would be agreed to [as rationally advantageous] … If our defense fails, we must conclude that rational bargaining is in vain ... Indeed, if our defense fails, then we must conclude that a rational morality is a chimera, so that there is no rational and impartial constraint on the pursuit of individual utility.7

Gauthier’s argument for the rationality of compliance with bargains rationally made relies on his proposed revision to rational choice theory. Traditionally, a choice is rational if it is the choice that would maximize the expected preference satisfaction of the agent making the choice. Gauthier proposes to introduce the concept of dispositions into the account of practical rationality. A choice is rational if it follows from a rationally chosen disposition to act. A disposition, in turn, is rational if it tends to maximize the expected satisfaction of the preferences of the agent choosing the disposition. He then assesses the relative rationality of two dispositions. Straightforward maximization is a disposition to comply with agreements rationally made only if the expected utility of doing so is at least as great as the expected utility of defecting on such agreements. Constrained maximization, by contrast, is a disposition to comply with agreements rationally made, even if there is loss relative to the possibility of defection, when it is likely that others who are party to the agreement will also comply. Pitting these two dispositions against one another, given the conditional

68

Liberty, Games and Contracts

cooperator’s ability to comply with agreements or not, allows Gauthier to make good his claim that constrained maximization is rational; straightforward maximizers lose out on potential benefits afforded those who are disposed to cooperate with similarly disposed individuals. In situations where mutual benefit is possible in strategic situations (as well as some potential benefit to be had from a successful defection), constrained maximizers realize the benefits of cooperation with each other, while straightforward maximizers will be excluded. And so Gauthier argues for the rationality of compliance with rational bargains, and rational morality, while preserving the tension between an agent’s preferences and the dictates of morality. Narveson adopts a great deal of Gauthier’s solution to the compliance problem. He argues that Gauthier has, at least, destroyed the presumption that people interested in maximizing the satisfaction of their preferences will not keep their word. This gets us actual people (surely every contractarian’s ultimate focus) closer to solving the compliance problem. Actual people, Narveson argues, need to constantly work at adopting dispositions. The benefits to be had by cooperative behaviour are sufficiently high to interest us in changing our own disposition. Moreover, that these benefits depend on society’s adoption of the same set of rules constraining behaviour gives us reason to reinforce such dispositions in society in general (and in our children in particular).8 The benefits to be had in a society in which people engage in mutually advantageous behaviour is sufficient, in his view, to motivate us to ‘do … whatever we can to set Morality in motion: a social institution of reinforcing behaviour’.9 2. The Motivation Problem Whatever the reason one is led to contractarianism, the problem of motivation is central. Insofar as morality is understood as an artifice of rational agents, the problem of justification must include an argument explaining the motivational force that lies behind such adopted practices. Contemporary contractarians have largely focussed on the prisoner’s dilemma (PD) game, arguing that if it can be shown rational to cooperate in a PD, then we can infer that it is rational to follow rules that we rationally agreed to, as the payoff structure in both cases is the same.10 Depending on the particular problem of motivation that the theorist wishes to address, this paradigm may be unduly difficult. Showing that it is rational to cooperate in a PD is to show that all rational agents have sufficient reasons to cooperate instead of defect. If contractarians are interested in showing that all rational people have sufficient reason to follow the rules they would rationally agree to, then attempting to ‘solve’ the PD may be a perspicuous method for doing so. Contractarians need not, however, have such lofty ambitions. A contractarian interested in accounting for morality’s motivation may only be interested in showing that every agent has reasons to comply, but not that they are sufficient reasons to comply. This would account for morality’s motivation without requiring that those who occasionally do not do so are necessarily acting irrationally – something surely at odds with common opinion. Similarly, a

Contractarian Education

69

contractarian might be interested in arguing only that almost all people have reasons (sufficient or not) to comply with the dictates of morality. While perhaps sacrificing the possibility of justifying a morality with universal appeal, this tack is perhaps more realistic.11 This last approach has received scant attention in the contemporary literature. I would like to bolster the case for it significantly. The case takes the form of a disjunctive elimination. Contractarians can either attempt to solve the problem of compliance for every agent, or for less than every agent. It is impossible to solve the compliance problem for every agent. If one is convinced that contractarianism is the correct justificatory mechanism, then one ought to solve the compliance problem for less than all people. Attempting to prove that all agents have sufficient reasons, or simply reasons, to comply with the dictates of a rationally agreed upon normative system is simply not possible. I will examine each in turn. (a) Each agent has sufficient reason to comply In the first place, such an ultimate defence of ‘all rational agents in the circumstances of justice ought to abide by the terms of X, which they would agree to in situation Y’ is impossible once we recognize the existence of psychopaths and sociopaths. These are people (arguably) who have coherent preferences and beliefs, as they are able to navigate in the world in a more or less coherent way. They simply have a different belief structure from that of a normal agent: one that will never allow restraint in interaction to become instrumentally valuable. Arguably this observation can be straightforwardly overcome by reminding the critic that contractarians are interested in answering only agents who fall within the circumstances of justice: those circumstances, including psychological dispositions, within which a person must find herself in order for a normative system to be even potentially useful. On one common, though surely mistaken, interpretation, psychopaths are agents who are extremely malevolent. Given that the circumstances of justice exclude agents who are either too benevolent or malevolent for bargaining to be necessary or possible, such agents cannot be within the circumstances of justice. On a more accurate interpretation of the psyche of a psychopath, these agents would not be relevant to this discourse because they have no awareness of the variability of scarcity given differing reactions to the existence of other agents. That this is a necessary condition for the circumstances of justice to obtain has been noted by David Gauthier.12 On the face of it this response is fair enough as far as it goes. It is obvious that we are not trying to show that all agents whatever would/should follow the dictates of the relevant morality. We are only providing good reasons to follow the rules to some subset of agents who are relevant for our purposes. We will not set ourselves up for failure. But neither should we set ourselves up to beg the question, which is what I think we are forced to do if we think that we are supposed to solve the compliance problem for all agents. Move on now to Casey who hates only Jones with a passion that gives him cramps, and to whom he would never pay the courtesy of refraining from hitting over the head in some back alley. In all other cases Casey is willing to comply, but never with Jones. Casey will never adopt an instrumentally valued preference, or an

70

Liberty, Games and Contracts

instrumentally valuable belief regarding exhibiting constraint towards Jones. Is Casey within the circumstances of justice? It gets harder to say no without the appearance of begging the question. We can still, of course, bite this bullet and suggest that only those people who, at least potentially, would agree to always comply are within the circumstances of justice. But this is now reminiscent of Rawls’s insistence that all citizens within his domain of discourse are both rational and reasonable (willing to propose fair terms to others likewise disposed and abide by those terms once accepted). However legitimate such an assumption is in a constructivist effort, it is clearly troubling in contractarian efforts attempting to provide a fundamental justification of moral or political systems. Changing the example slightly: what if Casey didn’t hate Jones, but instead suggested that there was one particular (though unlikely) set of circumstances in which he absolutely refused to comply with the dictates of morality? In any instance of this set of circumstances, he would simply not comply with the dictates of morality. Call it a psychological quirk. Does this place Jones outside of the circumstances of justice? If not, then we have admitted that we cannot answer all sceptics. Imagine further that the identified quirk is: ‘Whenever it is in my interests not to obey the dictates of morality, I won’t obey those dictates.’ Casey is identified as the Foole, though not foolish. He is not naive either, and this quirk refers to cases where it really is in his interests. To put this another way, Casey is not a rash man, and will only not obey the dictates of morality when he sees, for example, a very drunk man walking down an alley with no one else about, who has just stopped at a bank machine to get the rest of his life savings to drink away at the next bar he is allowed to enter. Casey is either in, or outside, the circumstances of justice. If Casey is out, then surely we are just begging the question against the sceptic, who has as much as been told: ‘You have reason to act morally only if you are not disposed to disobey the dictates of morality.’ If in, then we have to convince him, per impossibile, that in exactly those situations in which he would prefer to behave immorally, he really prefers something else. But if Casey still disagrees, then can we really claim to have the authority to say: ‘Yes you do, you just aren’t thinking straight’? Given Mill’s insights regarding this matter, I think contractarians have to say no, we cannot claim that.13 In this case we are either forced to say that such an agent is now known to be outside the circumstances of justice, since he is clearly irrational, or to admit that we have failed to solve the compliance problem. In neither case will we have answered the sceptic. Applying the circumstances of justice – which are surely useful to delineate what must generally be true about the world in order for justice to have emerged – to the problem of whether any particular agent is a potential source of moral rules, is bound to either beg the question, or not solve the compliance problem. (b) Each agent necessarily has some reason to comply Perhaps contractarians could do better by suggesting that they are merely attempting to show that agents have some reasons, but not necessarily sufficient reasons, to follow the rules in each

Contractarian Education

71

instance. This does not solve the problem. It is instructive to observe that contractarians do not typically do this. Contractarians do not tell the Foole that he has some reasons to comply, but instead they tell the Foole that he ought to comply. Contractarians are thus not simply suggesting that agents have some reason to comply. But perhaps they should. Attending more carefully to the instance of Casey by way of example, it is clear that this will not work. Given the contractarian’s reliance on a subjectivist account of preference and belief, contractarians must identify these necessary reasons relative to each agent’s preferences and beliefs. In no other way will a contractarian justification provide practical reasons to act in accordance with the dictates of morality. Casey’s beliefs and desires about the value of Jones are supposed such that he will never adopt a belief or desire that constraint is instrumentally valuable. Perhaps Casey believes that Jones is simply an evil person, and is owed no constraint. Stronger, perhaps Casey believes that his god will punish him if he collaborates with Jones. It might, alternately, be the case that Casey simply hates Jones so much that Jones receiving any good from constrained interaction with Casey would result in such a disutility for Casey that Casey’s purported benefits in mutual constraint would pale in comparison. Thus no deal struck with Jones, and thereafter abided by, could be worth Casey’s while. One might suppose that there must be a reasonable argument available that would convince Casey to stop hating Jones, but this seems unlikely when we attend to efforts that have attempted to force people to reconceive their god, for example. There is no in-principle reason to suppose that agents will attend to any particular argument, given that it is solely through their existing set of preferences and beliefs that they examine these arguments. Casey may simply believe that Jones is not human, or incapable of constraint ever (notwithstanding evidence to the contrary), or a member of the CIA who is out to get Casey and everyone Casey cares about. Insisting, as one might be tempted to do at this stage, that only ‘reasonable’ people fall within the circumstances of justice will be subject to the same criticism as any other attempt to exclude such agents via the circumstances of justice – such delineation will either beg the question, or fail to solve the compliance problem. The general point is that whatever set of beliefs and desires a contractarian assumes, an agent can be constructed who has opposed beliefs and desires. Thomas Hobbes fails to account for people who are risk-seeking.14 Jan Narveson assumes a certain prudence in his population.15 Both theorists’ recommendations will not reach a shortsighted thrill-seeker. Even David Gauthier, who avoids reliance on any particular set of preferences by constructing a formal economic contractarian argument, does not overcome this problem. His construction obscures the fact that agents can have desires in stark opposition to his principles. He produces a formal argument for the rational acceptability of a particular distribution of any interactive surplus – minimax relative concession (MRC) – and ignores any possibility that agents might prefer adopting a different distribution, or believe in another principle of dividing up the cooperative surplus (the Nash solution, for example). Nevertheless, given the possibility of such agents, his project begs the question when arguing for the rationality of compliance with any joint strategy that

72

Liberty, Games and Contracts

approximates the terms identified by MRC. The commitment to a subjectivist account of instrumental rationality renders impossible a justification of moral rules that necessarily appeals to each relevant agent. Given the two sections above, it seems clear that whether a contractarian aims to provide sufficient, or simply some, motivation to comply with the dictates of a normative system, he will either fail, or beg the question. Insisting that each agent possesses symmetrical motives will beg the question, and allowing for asymmetrical motives will lead to the failure of the enterprise, both at the level of the identification of the relevant system of rules, and the subsequent arguments for compliance with that system. 3. Contractarian Aims Suggest a Different Tactic There is another compelling reason to suggest that contractarians ought not to attempt to solve the compliance problem in either universal incarnation: it is plausibly maintained that rational agents do not want perfect compliance, even if it were possible. More accurately, rational agents do not want to pay for perfect compliance. Before proceeding to develop this line of analysis, it will be helpful to classify various contractarian efforts to solve the compliance problem. Traditionally, solutions have been divided into two types: internal and external.16 External solutions to the compliance problem focus on restructuring the choice situation faced by contractarian agents in such a way that, given their previous preferences, they will thereafter choose to comply with the appropriate moral or political norms. Typically, external solutions confront a decision-maker with a post-conventional entity (for example the state) which is designed to, upon discovering a defection, actually provide sufficient disutility to that agent so that she will not choose to defect in each choice situation where defection is possible. Hobbes provides the classic external solution to the compliance problem with the institution of a sovereign. The enforcement mechanisms available to the sovereign make it rational for all agents, fearing the sovereign’s reprisals, to adhere to the dictates of the state. Attending to the subjectivist commitment of contractarians, it is more accurate to describe such efforts as attempting to change the belief structure of rational agents, such that they no longer believe non-compliance to be rational. The state of the world is, strictly speaking, irrelevant to those beliefs, but surely the conflation is understandable. Beliefs are largely beliefs about the world, after all. Internal solutions to the compliance problem involve restructuring the desires of contractarian agents in such a way that they now prefer to comply with the dictates of the norms found to be in everyone’s strategic interests, all things considered. They advance arguments suggesting that agents will ‘rewire’ themselves so as to voluntarily comply with the dictates of morality in certain situations. Working with a traditional rational choice agent, who is comprised of a set of beliefs, a set of desires and a set of possible actions, this can be recast as an agent choosing to replace either her set of beliefs or desires (or both) with another set suitably modified to make compliance desirable. The introduction of dispositions into the arena of rational choice theory, à la David Gauthier and Peter Danielson, renders

Contractarian Education

73

this recasting problematic.17 Dispositions, according to Gauthier, sometimes lead to action and constrain straightforward maximization of preferences in these situations. The adoption of a rational disposition thus seems to complicate, rather than replace, a rational agent’s psychology. Notwithstanding this complication, we suppose the coarse-grained distinctions sufficient for our purposes. External solutions modify the world surrounding the agent (in an effort to modify each agent’s beliefs regarding the world) and internal solutions modify the contractarian agent’s psychological make-up directly. Assuming that an external solution to the compliance problem is even possible, it would be very costly. More to the point, each additional unit of compliance is going to cost more than the previous unit of compliance. A police officer on every second block will cost a pretty penny, but a police officer on every block will cost more; and moreover the demand for more police officers raises their market value, and thus their pay. Increasing detection is probably not the end of it either. If each detection leads to trial, or at least leads to trials in some fixed ratio of detection to trial, then the cost of increasing detection also increases the cost of judges, bailiffs, lawyers and physical facilities. Without being exhaustive, I think the point has been made. It is reasonable to assume that the cost of perfect compliance will outweigh almost every agent’s interest in securing that compliance. And so on a contractarian’s own grounds, they ought not to pursue it. Internal solutions to the compliance problem may appear sufficiently less costly to make a case for perfect compliance when compared to external solutions to the compliance problem, but this is not the case. Costing little by way of resources, internal solutions that aim for perfect compliance create significant opportunity costs. Ignoring Gauthier’s by now supplanted attempt to restrict the domain of dispositions to straightforward maximizers or constrained maximizers, we recognize that the potential range of dispositions for an agent to adopt is infinite.18 Retaining his insight regarding conditional cooperation, we offer the following argument: an agent will only bargain with another agent who will follow through with the terms of that bargain in mixed motive games. The only way to figure out if one is dealing with such an agent is to scrutinize that agent, and decide if he is giving the right signs. Assume, for the purposes of clarity, that this is only achievable by looking at that other agent’s face for the right sign. Assume a fixed and diverse population, and costs associated with scrutinizing a face. This diverse population includes both conditional cooperators and those not so disposed. Any external sign designed to show that an agent is compliant will be rational for another to mimic, attempting to elicit cooperation so that one can defect on the terms of the agreement. Those not disposed to cooperate will have a range of more or less complex mimicking behaviours, which to a greater or lesser extent will be successful. Cooperative agents need to scrutinize other agents to determine, to the best of their ability, whether the other agent they are about to strike a deal with will comply or not. While a discerned mimicked expression will stop further scrutiny costs, there is nothing to tell the scrutinizing agent whether his potential partner is either a (thus far) successful mimic, or a true conditional cooperator.

74

Liberty, Games and Contracts

Time constraints prevent eternal scrutiny, which would only assure perfect compliance if mimicry were presumed imperfect in any event. And so agents would either never embark on cooperative ventures, or would adopt a strategy of finite scrutiny which, if passed, would lead to a joint venture. Assuming that there are benefits to be had from cooperation even if one is occasionally incorrect about the nature of one’s partner, clearly the latter would be rational. This assumption seems to be borne out in our actual world, to which we must make our models relevant. Thus agents in these models will prefer imperfect compliance (occasionally being suckered by a good mimic) to no suckering at all. Assuming transparency, as Gauthier famously does, would of course lead to quite different results. But assuming transparency renders the model utopian and thus irrelevant for our purposes. Given these observations, I think that the conclusion is obvious: contractarians ought not to attempt to provide an argument for perfect compliance, or argue that everyone has some reasons to comply. It is neither possible without questionbegging, nor desired by the relevant agents in any event. They ought, instead, to attempt to argue for a level of compliance that is ‘good enough’. This is to say that contractarians ought to argue that a sufficient number of people have sufficiently good reasons to adhere to the norms in question, all things considered.

Reinterpreting the Contractarian Enterprise Aiming to provide an efficient level of compliance does not, by itself, relieve the contractarian of the obligation to provide a solution to the compliance problem. Indeed, it may be safely said that this observation, by itself, moves the project forward very little. Proving the rationality of efficient compliance still requires providing either some people with motivating reasons to follow the dictates of morality all of the time, or everyone with motivating reasons to follow the dictates of morality some of the time. The ‘times’ being referred to are times when the rules dictate one course of action while one’s preferences dictate some other course of action. This will occur more frequently for some than others, given, for example, how much they prefer to consider others. However frequently they occur, in each case the compliance problem reveals itself: why would a rational person act thus and so when her preferences would be better satisfied by acting in some other way? The short answer is that she would not. But this ought not to discourage the contractarian. The problem in these circumstances is that we are dealing with agents who have well-defined preferences. Attending to the education of the next generation of any given population avoids this difficulty. Whatever the force of Aristotle’s insistence that the young not study political philosophy, it cannot be taken as reason to fail to consider the effect that having children among us has on a justification of normative systems. Hobbes’s treatment of rational agents as having sprung up like mushrooms out of the ground fully formed provides us with no further reason to exclude children from our purview. It may be that we cannot hold children to account for their behaviour in the same way

Contractarian Education

75

that we hold adults to account for their behaviour. Their preferences may be too illformed, and their beliefs about the world too scant or too simplistic. This does not mean that we need only consider adults when considering normative systems. Indeed, it is because of these differences that turning to consider the education of the young largely resolves the compliance problem. Contractarianism so conceived has two parts: a system of rules found desirable by a population (typically identified via rational bargaining), and a system of education through which the next generation are moulded in such a way that they find being a good citizen desirable. This will, I suppose, result in only minimal changes in the way contractarians identify a rational set of moral rules or policies. Appealing to the interests and powers of each agent in a particular population will lead to the identification of a set of rules acceptable to as large a number of people as possible.19 It will significantly change the nature of the compliance problem, however. Instead of each person needing to be convinced that there are reasons to follow the rules, good reasons have to be suggested that show that people have good reasons to educate children so as to act in particular ways (to wit, the ways recommended by the identified normative system). The traditional compliance problem involved each agent asking: Why ought I comply with these rules when I prefer to do otherwise? No rational reason presented itself. Now each person must ask: What reason do I have to instruct the young to act in ways that I have found advantageous? More obvious answers seem forthcoming. Contractarianism as education asks very little of the non-indoctrinated citizens of the first generation. Instead of asking for compliance with the rules all of the time, it would minimally be necessary to simply appear to follow the rules whenever someone of the next generation is looking. This demands significantly less of that generation than orthodox contractarians demand of their agents, and so it is more likely that contractarianism with education will be able to achieve its goals. It also does not generally ask agents in a particular situation to act contrary to their preferences. Given the presence of young eyes, a preference to otherwise defect could, all things considered, become a prudent preference to instruct the future generations on this occasion, and refrain from mugging this drunk in this alley on this occasion. This will not universally hold true, of course, but we are not looking for universally applicable reasons to conform to the dictates of a normative system any longer. Contractarianism with education also allows for a more stable contract than orthodox alternatives: as successive generations attend to their children’s education, the level of actual compliance will go up, as the first generation dies off. Further, there will be no (or significantly fewer) attempts by the second generation to tell the third generation the ‘true nature’ of the world, and of morality. Each truly converted second-generation citizen will genuinely believe in the rules. There is simply no conflict between the properly educated youth of today and the rules originally identified. The young are instructed in such a way that they prefer to interact in the ways dictated by the identified normative system. Perhaps a clarification is in order to stave off an obvious, though misplaced,

76

Liberty, Games and Contracts

objection. A contractarian education would not teach blind obedience to the moral dictates in question. If it did, it would be a significant strike against such a theory, given a common commitment in the philosophy of education, indeed in philosophy generally. Martha Nussbaum, in her book Cultivating Humanity: A Classical Defense of Reform in Liberal Education, stresses the importance of being able to critically reflect on one’s beliefs and culture, and of in principle being able to reject any component which did not survive rational scrutiny.20 John Rawls cannot be said to neglect the importance of autonomy in A Theory of Justice and Political Liberalism, the influence of which cannot be avoided.21 While a solution to the compliance problem may seem to require some kind of passive obedience to the moral law, as a thoughtful examination of the structure of a PD seems to lead to defection, I deny that it would so counsel. Blind obedience would only be necessary if rational deliberation would lead to the rejection of the relevant indoctrinated desires. This does not seem likely, as there would be no rational basis to reject the desires. Having been developed in such a way as to desire to cooperate fairly with one’s fellows, what could the basis for rejecting such desires be? Given instrumental rationality, the basis would have to be other important desires. But insofar as fair cooperation does make one better off, then a basis on which to reject these cooperative preferences seems unlikely. I suggest therefore that these desires for fair cooperation would survive rational scrutiny, and thus that passive obedience would not be necessary. Moreover, a passive character is at cross-purposes with the other aim of contractarian education – instrumental value. Surely an active and questioning nature is more conducive to economic stimulation than the reverse. And so the rationally deliberative character of our future citizens is secure.

Conclusion The compliance problem as traditionally conceived is an inevitable result of contractarianism as traditionally conceived. The traditional reasons for adopting a contractarian justificatory framework are important ones, and the intractability of the traditional compliance problem should not lead to a rejection of contractarianism. Contractarians ought to, instead, reconceptualize their project in two ways: they should aim for identifying and achieving an efficient level of compliance given particular populations, and they should aim to achieve it in part by educating the next generation of that population. As previously mentioned, Jan Narveson has been preaching at least a part of this message all along. A focus on the first half of his solution to the compliance problem, which largely adopts Gauthier’s proposal, has obscured the use of the second half. I hope to have gone some way to vindicating this latter, while discrediting the former. In no other way can contractarians hope to avoid being penned in by the wall of their own creation: the compliance problem.

Contractarian Education

77

Notes 1 2 3 4

5

6

7 8 9 10

11 12 13 14

15 16 17 18 19

20 21

On the conceptual desirability of internalism, see: Susan Dimock, ‘Two Virtues of Contractarianism’, The Journal of Value Inquiry, 37/3 (November 2003): 395–414. Jan Narveson, The Libertarian Idea (Philadelphia, 1988), 122. Ibid. Another suggestion is to appeal to all of the different sets of interests that people have and show that they each lead to the adoption of one set of moral norms; but I have previously argued that this is not plausible. See Chris Tucker, ‘Contractarianism, Justification and Relativity’, Dialogue: Canadian Philosophical Review, XLII/3 (Summer 2003): 559–71. David Gauthier, Morals by Agreement (New York, 1986), 1. This statement implicitly observes that morality is supposed to be universal in scope – it is to each person’s reason that morality must appeal. ‘But my deepest reason for endorsing a contractarian grounding of morality, despite worries about scope and motivation, is that no other account seems compatible with the maximizing conception of practical rationality, and I no longer find plausible the view … that this conception is merely part of the particular ideology of our modern, Western society.’ David Gauthier, Moral Dealing: Contract, Ethics, and Reason (Ithaca, NY, 1990), 7. Gauthier, Morals by Agreement, 158. Narveson, esp. 144–5. Narveson, 145. Of course, as Ken Binmore is so fond of pointing out, this task is impossible. A PD is defined as a game in which defection dominates over cooperation. Ken Binmore, Game Theory and the Social Contract: Vol. 1: Playing Fair (Cambridge, MA, 1998), 27, for example. Narveson has argued that this approach need not be seen as sacrificing universality; see Narveson, 134–5. Gauthier, Morals by Agreement, 114. J. S. Mill, On Liberty (Indianapolis, 1978), ch. 4. Laws of nature being laws of reason alone which forbids a man to do that which may, even indirectly, lessen his chances of living excludes from consideration risk-seeking behaviour. See Thomas Hobbes, Leviathan (Buffalo, NY, [1651] 1986), esp. ch. 13. Narveson, esp. 131–47. I borrow these terms from Jody Kraus’s The Limits of Hobbesian Contractarianism (Cambridge, 1993). Gauthier, Morals by Agreement; Peter Danielson, Artificial Morality (New York, 1992). For Gauthier’s description of these dispositions see Gauthier, Morals by Agreement, 167. This is only an approximation of the calculation. The rules will need to be acceptable to: (a) the largest set of people who (b) find each other more useful acting within the constraints of morality than outside such a system of constraints. Generally, this would only be the largest number of people if all people were supposed equally useful to each person. Martha Nussbaum, Cultivating Humanity: A Classical Defense of Reform in Liberal Education (Cambridge, MA, 1997). John Rawls, A Theory of Justice (Oxford, 1971); John Rawls, Political Liberalism (New York, 1993).

This page intentionally left blank

PART TWO FROM CONTRACTARIANISM TO LIBERTARIANISM

This page intentionally left blank

Chapter 6

The Value of Values: The Importance of Autonomy in Contractarian Reasoning Susan Dimock

Jan Narveson has long advocated contractarianism as the only viable theory of justification in moral and political contexts. In this he is correct. Narveson further thinks that libertarianism is the substantive moral view entailed by contractarianism, given a set of assumptions about human nature and the circumstances of human interaction. In this he is mistaken, or so I suggest. The reason Narveson is led to the erroneous conclusion that libertarianism follows from contractarianism is his employment of an inadequate conception of ‘value’ at the heart of contractarian justifications of normative constraints. Contractarians justify a set of moral rules on the basis of their service to the values of individuals, but the resulting theory depends for its normativity on the strength of the theory of values upon which it is built. I argue that if values are to play the role assigned to them in contractarian accounts then they cannot include all desires or ends an agent might have, but instead they must be restricted to autonomous desires. I then (begin to) show that one must derive very different substantive conclusions from the contractarian methodology, using a few of Narveson’s own discussions of substantive issues such as drug legislation, education and information policy.

Internalism and Contractarianism Jan Narveson defends libertarianism as a substantive moral theory, with its attendant anarchistic or near-anarchistic implications for politics. But his defence of libertarianism rests on two more basic commitments, one meta-ethical and the other methodological. The meta-ethical commitment is to internalism and the method is contractarian. The two are closely related. Morality concerns the social rules and principles that properly govern the interactions of individuals. Morality is practical, not only because moral rules direct behaviour, but also because they include both dispositions to act according to the rules and to engage in informal behaviour which reinforces compliance with the rules (for example, criticism of those who violate the rules, appropriate praising and blaming activities and so on). Narveson subscribes, then, to some version of internalism: morals must matter, have ‘importance’.1 And their importance must be overriding. Thus he insists (a) that morals must have power to overrule contrary 81

82

Liberty, Games and Contracts

motives of self-interest. Morals must be able to motivate those to whom the moral requirements apply (that is, everyone, or virtually everyone). And (b) morals must attract the support of others. Morals must have the power to motivate others to engage in the actions by which morality is informally enforced. Because Narveson subscribes to internalism, he seeks moral requirements that are rationally acceptable to the people whose morality it is. Thus he is led to contractarianism as the foundational moral theory, as he calls it, or to contractarianism as the methodology for moral reasoning, as I prefer. Just as what we rationally should do is what we have good reason to do, what we morally should do is what there are good reasons for insisting that people do. This is contractarian reasoning in a nutshell: the principles of morality are (or should be) those principles for the directing of everyone’s conduct which it is reasonable for everyone to accept. They are the rules that everyone has good reason for wanting everyone to act on, and thus to internalize in himself or herself, and thus to reinforce in the case of everyone.2

Contractarians attempt to ground the normativity of morals (their internalism), then, in a more basic theory of the normativity of practical rationality. The conclusions of practical rationality are themselves normative because they concern what a person has (most) reason to do, given his or her ends or values. Because a person cannot be indifferent to her own ends or values (they must have importance for her), when reason directs her to take certain actions in pursuit of those ends, she cannot be indifferent to those reasons either. Thus internalism with respect to individual ends or values extends to internalism with respect to moral principles via the contractarian methodology. The most important thing to emphasize about the contractarian approach is that it hopes to generate moral principles for societies out of the nonmoral values of individuals. … Now, if our social principles are generated out of individual values, there is only one way this can be done on our assumptions: namely, that those individual values support those social principles. If they do so, then our individual cannot – as she can do, of course, with any intuitionistic system – look in the face of what she agrees to be a moral principle, as shown by this procedure, and react with a yawn. She can’t say ‘So what?’ because the answer is that she ignores it at her own peril, as shown by her own values; and her own values are those that, by definition, she cares about.3

Contractarians ground the normativity of morality, then, in the normativity of practical rationality. Rationality is understood to be entirely instrumental: rational agents choose effective and efficient means to their ends. Their ends are, moreover, entirely subjective. It is further assumed that their ends are at least partly competitive, so that if each tries to maximize the satisfaction of their ends without social rules governing their interactions with others then all will do worse than if they adopted some such rules. If adopting a set of social rules would better enable everyone to satisfy their ends, then those rules are endorsed as a rational morality via contractarian reasoning. Narveson argues that one set of social rules is uniquely supported by this method: libertarianism.

The Value of Values

83

Because individuals have diverse ends, we cannot assume that they will all find it rational to agree on any perfectionist morality. But individuals can be assumed to want the freedom to pursue their values, whatever they may be, without interference from others. Thus they would all find it rational to agree to a set of rules that prohibit interference with the liberty of others, provided others will be prohibited from interfering with them. Interference is cashed out as the use of violence, force and fraud to make another do what they otherwise do not want to do, or prevent them from doing what they want to do. Thus contractarian reasoning supports a moral theory guaranteeing to each a negative right to liberty, the right to pursue one’s ends free from interference by others. Although freedom is both freedom to do something (bring about something or at least try to bring about something) and freedom from interference, all the action in political philosophy is to be found in the interference condition. When one’s freedom is impeded by the intentional actions of another practical agent, we can ask for a justification.4 The only such interferences that are justified, according to libertarianism, are those required by the equal basic right to liberty of others. Thus we can legitimately be forced not to invade the legitimate sphere of liberty enjoyed by others, and otherwise interferences with our liberty constitute a violation of our rights and are unjust. A further feature of Narveson’s analysis of liberty is that the right to be free includes the right to use unowned material goods, and he subscribes to a theory of just initial acquisition. Although he thinks the right to liberty and the right to property are inseparable, we might say that together they form the basic theory of libertarian justice: everyone has a negative right to liberty and property, which is held against all others. These rights are enforceable by force, if necessary. Narveson sees libertarianism as a kind of liberalism. Now liberalism is defined in different ways by him, but core features are retained throughout his many discussions of the subject. It requires, among other things, state neutrality between conceptions of the good; a liberal society has no aims besides those set by the preferences of its members. We may identify liberalism with the thesis that the good to be promoted by politics is simply the good of each member of the polity, as seen by himself. As far as society and politics are concerned, the good for Jones is what Jones thinks is good for Jones. That good-for-Jones may be equated with what Jones wants, taking the notion of a want broadly enough to include what Jones values.5

In this sense, liberalism is to be contrasted with conservative and perfectionist theories, which invoke a conception of the good to be pursued. Conservatives and perfectionists are willing to coerce others into pursuing their own ‘real’ or ‘true’ interests when individuals do not realize what their interests actually are.6

Values in Contractarian Reasoning In the foregoing story, significant weight is placed upon the values of individuals. They are the foundation of the right to liberty. As Narveson says,

84

Liberty, Games and Contracts What is needed is to show that the concern we must by nature have with our own values, virtually whatever they may be, is such as to underwrite a social principle of allowing each to pursue that person’s possibly very different ones – and that no other moral principles can take precedence to this one.7

But can they carry the normative weight that Narveson assigns them, and can they ground the internalism he seeks? It might be thought that they can, because Narveson stipulates that he is assuming autonomous agents in the derivation of morality from their subjective ends.8 He works with Lawrence Haworth’s conception of autonomy as rational competence.9 It may seem that autonomous values do have the normative importance needed both to justify a set of rules that further their realization and to explain the necessary motivating force such rules must have according to internalists. Despite his explicit stipulation that he assumes agents are autonomous in his project, I argue that he nonetheless fails to give due weight to the importance of autonomy when he draws his substantive conclusions in favour of libertarianism. Indeed, this can be anticipated in the early pages of The Libertarian Idea, where Narveson stipulates that his theory presupposes autonomous rational agents; but he very quickly omits the autonomy condition and thereafter utilizes the much more minimal conception of rationality, which requires only that one have a reason for what one does, that one have a desire or end which makes one’s action intelligible. That the condition of autonomy does no real work in Narveson’s theory, that it provides no constraint on the ends that are considered in determining what agents would find it rational to agree to, seems evident in such claims as this: for if we begin with the root idea that people are to be allowed to do whatever they want to do, then so long as any trains of action are conceivable and really instantiable, the question can only be whether a given sort of action will collide with the legitimate liberties of others. If it does not, then, given our starting point of a general right to liberty, any one who can do the action described and wants (for whatever reason) to do it is to be allowed to do it.10

Narveson’s seeming indifference to why people want to do what they want to do is crucial to his argument that there is a presumption in favour of liberty. We don’t ask, he thinks, ‘Why allow?’ but rather we ask ‘Why forbid or restrict?’. The burden is on those who would limit liberty to show that there is a reason to restrict liberty, which must itself be based on the need to protect liberty. This presumption is grounded in what people want. It is not … necessary for anyone to find good general reasons for allowing them. The fact that people want to engage in them is, on the libertarian theory, all the ‘reason’ that takes. Of course, in the case of any action there is a presumption that the agent wants to engage in it; hence the prima facie case is for allowing people to do whatever they do, short of violating the liberty of others.11

Thus there is a presumption in favour of allowing people to do what they want regardless of why they want what they want.

The Value of Values

85

Narveson’s indifference to the origin of agents’ ends is echoed in his treatment of values. Narveson subscribes, rightly enough, to a subjective theory of values. But it is an ill-defined concept in his work. ‘Each person, we assume, controls his or her activity in light of a scheme of values, some few of which are perhaps explicitly held and formulated, but many of which are not.’12 But can they really be values if held only implicitly? What is the relation between values and mere desires or wants? Are they identical? I think Narveson does not seriously consider the difference between mere desires and values because he does not give sufficient weight to the autonomy condition.

Autonomy I argue that only autonomous desires can be values in the sense needed to ground the internalism of contractarian-based morality. Only desires that satisfy a range of conditions, drawn from a theory of autonomy, can have the normativity essential to the contractarian project. The most important of these conditions, for this purpose, is that desires be authentic if they are to serve as values in the relevant sense. And authenticity requires that desires be identified with reasons for action. Personal autonomy consists, in its most basic sense, in the condition of being self-directed or self-governed. Any adequate conception of autonomy may be expected to capture the insistence, found within diverse theories of that condition, that autonomy consists of action that reflects one’s ‘true’ or ‘authentic’ self. It is a core idea in many theories that there is an authentic self, comprised of values and desires which are truly the agent’s own, and that autonomous behaviour flows from and expresses that true self. I summarize these intuitions in saying that autonomous persons are authentic in their reasons for action. Authentic reasons for action differ from mere explanatory reasons. We can explain the behaviour of even non-autonomous creatures, say, by reference to evolutionary processes or instinctive response mechanisms. Such explanations are inadequate for explaining behaviour as autonomous. Such explanations do not cite the agent’s authentic reasons for action. Authentic reasons for action are ‘the agent’s’ reasons for acting. They are both explanatorily adequate and defensible by the agent whose reasons they are. They are reasons the agent would cite both to explain and justify their action.13 As such, they are reasons that the agent could make self-conscious and whose authority as reasons she accepts. Autonomous agents act on reasons that are authentic in this sense. How do we acquire authentic reasons for action? Part of the answer is to be found in the requirement, articulated first by Harry Frankfurt and Gerald Dworkin, that agents make their pre-reflective desires authentic by critically reflecting upon them and identifying with some while rejecting others. Those thus identified with become the agent’s authentic reasons for acting. The basic capacity which makes autonomy possible is that of being able to reflect upon our desires as possible reasons for action. In discussing this reflective activity and endorsement of one’s desires, Dworkin and Frankfurt employ a bi-level account of desires that will be

86

Liberty, Games and Contracts

familiar to most readers. The distinction central to bi-level accounts of desire evaluation, and theories of autonomy or freedom of the will built thereon, is between first-order and second-order desires. First-order desires are desires to do or not to do something; they have actions (or states of affairs which can be brought about by actions) as their objects. Second-order desires are desires to have or not to have some first-order desires; their objects are first-order desires. Second-order desires which are desires that some first-order desire be effective in moving one to action are called volitions.14 The relevance of the bi-level theory of desires to personal autonomy is thought to follow more or less directly from this distinction: when we act from our higher-order desires we act on motives which we have examined and endorsed – that is, we act on authentic desires, and thus we are selfdirected.15 The intuition which drives the bi-level approach is that a person who acts from authentic motives – that is, those which she has reflected upon and approved of as reasons for action – is self-directed in a way that is not true of someone who acts from motives which she has not examined, or which she wishes not to act on. In his important essay on ‘Freedom of the Will and the Concept of a Person’, Frankfurt argues that what distinguishes persons from non-persons is the ability to reflect upon their desires and motives and form higher-order desires and motives which have these as their objects.16 ‘Besides wanting and choosing and being moved to do this or that, men may also want to have (or not to have) certain desires and motives. They are capable of wanting to be different, in their preferences and purposes, from what they are.’17 As reflective beings we can form desires about our own first-order desires; we can want our will (those first-order desires which are effective in moving us to action) to be different from what it is. Alternatively, we can be happy that it is what it is. It is this capacity for reflection, the capacity to become critically aware of our own wills and of forming volitions of the second order, which makes the attainment of personal autonomy possible.18 In this way we are capable of deciding which of our possible reasons for action will be authoritative for us. This capacity for critical reflection and the formation of volitions makes personal autonomy possible because in forming second-order volitions the person ‘identifies’ himself with a first-order desire and withdraws from others. In so doing, the desire identified with becomes ‘internal’ to the person whose desire it is, while those which are withdrawn from become ‘external to’ or ‘alien from’ the person. Identification makes some desires more truly ‘his’ than others which he has at the first order. Those desires with which one identifies are one’s authentic reasons for action. And a person is self-directed when he acts on desires that are authentic in just this sense: they are internal to him and truly his own. Frankfurt offers the example of an unwilling drug addict to explain the role of identification in his theory of freedom of the will.19 This addict has a desire to take the drug to which he is addicted, and a desire not to take the drug – both of which are first-order desires. But the addict has also made a commitment about which of these desires he wants to be his will (his effective desire). He has decided that he wants his desire not to take the drug to be his will. In making this decision the

The Value of Values

87

addict has identified himself with one of his conflicting first-order desires and withdrawn himself from the other.20 Before forming the second-order volition as a way of resolving the conflict among his desires, it could be said that he wanted both to take and to not take the drug. After he has identified with his desire not to take the drug, however, he can say that what he really wants to do is to resist the desire to take the drug, even though both the first-order desires may continue to be elements in his mental history and even if he succumbs to the desire for the drug at some future time. The central lesson of the bi-level account of desires can now be drawn: identification with a desire makes the desire internal to the agent whose desire it is, and it is on the basis of such acts of identification that one develops an authentic self. This model is understandably attractive, then, for it provides an account of the development of the authentic self – needed for self-direction – through a process of self-definition. One defines or constructs an authentic self that is constituted by those values, projects and desires with which one identifies after critically reflecting upon them. One acts autonomously only if one’s actions flow from and express the commitments of one’s authentic self. There is another reason for adopting the bi-level theory of desires as a model for the attainment of personal autonomy: it provides a mechanism for resolving intrapersonal conflicts, through the process of identifying with some desires and rejecting others as reasons for action. The ability to adopt second-order commitments provides a unique and necessary procedure for resolving some kinds of intrapersonal conflicts, thus allowing the development of a coherent authentic self. The bi-level theory of desires, understood as a coherence model, introduces a hierarchy of different levels of desires to illustrate how certain conflicts of desires are resolved by an agent in such a way that she commits herself to one of the conflicting pair (through identifying with it) and rejects the other (through withdrawal from it). In this way the person participates in the conflict, rather than experiencing it as a passive bystander, and comes to define herself in terms of one of the conflicting impulses. The bi-level theory offers a different kind of resolution to intrapersonal conflict from that offered within a uni-level preference ordering. Some conflicts of desires, where the desires are both first-order and it is contingently impossible to satisfy them both, are most simply resolved by ordering or ranking the desires on a single comprehensive scale of preference. From our preferences over outcomes (the states of affairs which would be brought about by our actions) we can construct an ordinal measure of utility on them. This ranking of outcomes according to preference is typically thought of as providing a first-order decision procedure: having ranked our options according to preference, we ought to act on that preference the satisfaction of which has the highest expected utility. If a person has, on a particular evening, a desire to see a play and a desire to attend a concert and she cannot do both, it is a perfectly adequate strategy for resolving the conflict to decide which she wants to do more strongly (which outcome she most prefers) and to rank it higher than the other; she should then simply choose on the basis of this preference

88

Liberty, Games and Contracts

ordering. Such strategies rank the competing desires on a single scale of preference. If our lover of the arts discovers, after ranking the desire to go to the play higher than the desire to go to the concert, that tickets for the play are sold out, it would be rational for her then to try to hear the concert. If she cannot satisfy her highestranked preference then she ought to seek the satisfaction of her second choice. Because the person really wants to do both these things, there is no reason to suppose that she is less than self-directed when she sees the play or, if that is impossible, opts instead for the concert. Other kinds of conflicts, though, cannot be resolved simply by ranking the contending desires. Such conflicts involve desires that are more than just contingently mutually unsatisfiable: they involve non-contingently conflicting desires. If a person desires to do x and desires to not-x then her desires are formally inconsistent. Resolving such conflicts cannot be accomplished simply by ranking the contending desires on a single scale of preference. They require resort to a higher order, where the person commits herself to one and rejects the other. If a person wants to live a life of celibacy and also wants to engage in sexual activity, for example, then he could resolve this conflict by identifying with the former and rejecting the latter. If he is then faced with difficulty carrying out his choice of living celibately, it would not be rational to seek the satisfaction of his sexual desires as the ‘next best’ alternative.21 For this form of conflict resolution does not mean merely that the person assigns his sexual desires a lower ranking on his scale of preferences; rather he must reject them as candidates for satisfaction altogether. So long as the conflict between formally inconsistent desires is unresolved, nothing that the agent does will satisfy her desires; necessarily, satisfying the desire to x means frustrating the desire to not-x. One way to resolve such conflicts is to take a decisive stand in favour of one of the options and against the other. This is the strategy taken by Frankfurt’s unwilling drug addict. The addict has two conflicting first-order desires: the desire to take the drug and the desire not to take the drug to which he is addicted. The addict is also able to take an evaluative stance toward his desires, though. If such an evaluation led him merely to rank the inconsistent desires on a single scale of preference, he might decide that it is preferable, all things considered, to resist his desire for the drug. But if this desire could not be satisfied, due to the motivational strength of the addictive desire for the drug or other reasons, then he should at least feel some satisfaction in satisfying his lower-ranked preference for the drug. This is not, of course, how many addicts feel about their addictions; and it is inadequate for explaining the full sense of unwillingness with which the addict succumbs to his desire for the drug. What the addict in Frankfurt’s example does at the evaluative stage is to make a decisive commitment toward his desire not to take the drug and to reject his desire for it. This is a decisive procedure for resolving the conflict.22 Adopting second-order volitions reduces intrapersonal dissonance: the addict can now say what he really wants to do. He has made the desire for the drug external to him, and if it determines his actions in the future he can say truly that he was motivated by forces which are alien to him and do not express his authentic self.

The Value of Values

89

Reflection and second-order commitments are thus strategies for intrapersonal conflict resolution, identification a means of achieving coherence and personal integration. Through the act of identifying with some desires and withdrawing from others, we define an increasingly coherent self out of the various and competing desires we have unreflectively accepted from others and from our society at large; likewise, through this process we allow a coherent modification of our preferences and self-conception over time. This, at least, is the hope, though it may be something of a fiction: possibly things will not go well, and one will identify with inconsistent desires. Nonetheless, through such reflective examination of their motives agents come to endorse some of their pre-reflective desires and to reject others; those they endorse become elements of their self-in-progress – the self they are defining. There is yet another reason to be concerned with the resolution of conflicts between our desires, for insofar as a person is conflicted, she has not made either of her conflicting desires authentic, and so she is incapable of acting autonomously with respect to them. The phenomenon of ambivalence makes this clear. There are compelling reasons to think that intrapersonal integration and coherence are necessary for personal autonomy. Anomie, severe schizophrenia and related mental disorders which produce, or are characterized by, dramatically fragmented or conflicted selves, may place the attainment of autonomy beyond the reach of those who suffer from them. To be self-directed there must be a coherent self that does the directing. The processes of identification and withdrawal allow us to explain how such a self is constructed. Of course, not all conflicts are as widespread as these. Many conflicts are more specific, affecting only a limited number of a person’s desires. Such conflicts need not defeat ascriptions of autonomy to an agent. Yet unresolved conflicts defeat the authenticity of one’s desires and make autonomous action with respect to them impossible. A person cannot act as she most wants to act if she does not unequivocally know what she wants. The attitudes that an agent can adopt toward any of her first-order desires as reasons for action are limited to three: (i) she may be indifferent toward it, having no second-order desire concerning its effectiveness; (ii) she may desire that it be effective; or (iii) she may desire not to act on it. In none of these cases is there a conflict between the volition and the desire which is its object. Motivational conflicts cannot be between desires of different orders, but rather must be between desires at the same level.23 This is merely a technical point about the possibility of conflicts between desires. There can be no inconsistency between desires of different orders, because they are directed toward different kinds of objects (desires on the one hand, actions on the other). It is noteworthy because it allows us to distinguish ambivalence (conflicting desires on a single level) from weakness of will (the overpowering of a higherorder desire by a lower-order one). Unresolved ambivalence makes it impossible to determine what the person most wants, because there is no unequivocal answer to that question. When a person has unresolved conflicting desires at the same order, she knows not what she really wants to do; and so there is no possibility of her

90

Liberty, Games and Contracts

either having or losing self-control in doing what she most wants to do. Nothing she could do would manifest self-control in this case, for she has not committed herself to one or the other of the competing desires. Weakness of will, by contrast, involves a lack of self-control. A weak-willed person has resolved her intrapersonal conflict of desires, but fails to make this resolution effective. I concentrate only on such conflicts within a single order first. We have already seen that there are two possible ways in which first-order desires can conflict. When desires are only contingently in conflict – that is to say, they have different actions or outcomes as their objects but are not mutually satisfiable – we might say that they ‘compete’ with one another. Until the competition is resolved we cannot say with any confidence what it is that the person really wants to do. If autonomy consists of doing what one really wants to do then this person cannot act autonomously until she resolves the conflict. Of course, this poses no great difficulty, for she must only determine which of the competing options is supported by the best, all things considered, reasons and rank them accordingly. Such a ranking will resolve the conflict and so will allow us to say what the agent most wants to do. We have also seen that first-order desires can be in conflict when they are formally inconsistent, that is, when a person both wants to perform a certain action (or desires a certain outcome be brought about) and wants not to perform that very same action (or desires that the outcome not be brought about). Here there is a conflict between the desire to x and the desire to not-x. Such conflicts, if unresolved, pose a serious problem for autonomy. The objects of the desires (the outcomes desired or the actions one wishes to engage in) are inconsistent; whatever the agent does she will be frustrating one of the conflicting pair of desires. Before we resolve such conflicts we cannot say, with respect to x, whether we want to do it or not. It is indeterminate what we really want to do, and so nothing we can do will count as expressing our true desires. Both kinds of unresolved conflict between first-order desires described here produce a kind of ambivalence: ‘volitional ambivalence’, we might call it. So long as volitional ambivalence occurs there is no action that the agent really wants to do. If authentic reasons for action are those which express what the agent really wants then such a person has no authentic reason for action with respect to her conflicted desires. If self-direction consists of actions which are motivated by authentic reasons for action then unresolved ambivalence must make autonomy, with respect to these desires, impossible. It seems, moreover, that adopting a reflective attitude toward our desires and identifying with some, withdrawing from others, is necessary to resolve such cases of ambivalence. Taking a decisive stand toward formally inconsistent desires requires that we exclude one of the conflicting pair as a candidate for satisfaction altogether. A recovering alcoholic, for example, does not simply rank his desire to drink lower than his desire not to drink; he rejects the former as providing any reason for him to act at all. This is necessary for him to define himself with respect to his drinking. The person he wants to be is a person who does not want to drink, and does not, in fact, drink. In attempting to become the person he wants to be, he has to reject his desire for alcohol completely.

The Value of Values

91

Ranking it, even very low, on a single scale of preference, requires that he continue to accept it as a possible option; but that is precisely what he must not do. If this is correct then reflection and identification are necessary for autonomy. But by introducing a second order of desires, the bi-level theory also makes possible a more serious kind of conflict and corresponding ambivalence: conflict at the level of volitions.24 Conflicts between volitions may involve volitions that have objects which are mutually unsatisfiable for contingent reasons. A person may have many first-order desires which are contingently in competition, yet identify with them. If that is so then ranking the contending first-order desires may resolve the competition, or a ranking of the second-order desires might be needed. Yet this poses no unique problems, and the resolution techniques for ranking alternatives within a single scale seems adequate to both tasks. But second-order volitions can conflict by being themselves inconsistent. Inconsistent volitions are a pair of second-order desires with the same object. An agent who has inconsistent volitions has a second-order desire to be moved by the desire to x and a second-order desire not to be moved by the desire to x. The agent both approves and disapproves of the desire to x. For example, a person who approves of his desire to exercise (because he believes that it is necessary to maintain his health, say) and disapproves of this same desire (because he believes that it has been caused by the multi-billion dollar advertising campaigns of manufacturers of fitness equipment and he dislikes the idea of being a dupe to the techniques of the persuasion industry) has conflicting second-order commitments. Insofar as they take the form of volitions, he wants both to act on and not to act on his desire to exercise. Nothing he can do while so conflicted will express what he really wants to do. This conflict cannot be resolved by forming another second-order desire concerning the effectiveness of his desire to x, for this would involve the mere duplication of a second-order volition which the agent already has. Nor will moving to a higher order of evaluation resolve the conflict; for if it is true that the agent approves of the desire to x and disapproves of the desire to x then reiterating that approval or disapproval at a yet higher level will not resolve the conflict. When one has formally inconsistent volitions one’s self is in conflict:25 thus we might call this ‘personal ambivalence’. The agent has failed to decisively identify with or withdraw from the desire to x. Such conflicts make self-direction impossible, for there is no coherent self to do the directing. No action will be expressive of what he really wants to do.26 Frankfurt is well aware of the threat that unresolved conflicts pose to autonomy. Thus he insists that a desire is authentic only if a person’s identification with it is ‘decisive’ and ‘wholehearted’, made in the absence of any present or anticipated conflict with respect to that desire. A person cannot be simultaneously ambivalent toward and wholeheartedly identified with a particular desire. Provided one’s identification is made in the absence of real or anticipated conflict, by contrast, it provides an unequivocal answer to the question of what the person really wants.27 The truly conflicted self, who knows not what she wants to do, is also the nonautonomous self, who cannot direct herself as she wishes. The conflicted self has no unambiguous self-conception upon which to draw in this matter. She must create that in resolving the conflict.

92

Liberty, Games and Contracts

The preceding discussion depends implicitly upon a distinction between autonomy as it is predicated of agents, and autonomy as it is predicated of desires and actions. In order to discuss weakness of will, that distinction will have to be made explicit. Agent autonomy is a property predicated of persons, on the bi-level view, if and only if they have the psychological (cognitive, conative and affective) capacities necessary to critically reflect upon and identify with or reject their first-order desires as reasons for actions. The defining characteristic of autonomous agents is that they have the basic capacity to reflect upon and structure their motives in a way that makes self-definition possible. But autonomy is also predicated of actions and desires. While it is necessary for an action or a desire to be autonomous that it be an action or desire of an autonomous agent, it is not sufficient to ensure that an action or desire is autonomous that it is ascribable to an autonomous agent. Autonomous agents sometimes fail to act on those desires they have endorsed and act on those desires they have rejected. In short, autonomous agents sometimes perform actions that are not themselves autonomous. The phenomenon of weakness of will provides ready examples of autonomous agents who act non-autonomously in particular instances. The bi-level theory of desires offers a very insightful account of weakness of will. I employ it here to illustrate how a person who has the requisite capacities for personal (agent) autonomy may, nonetheless, fail to act autonomously. Oftdiscussed paradigms of weakness of will are displayed by unwilling drug addicts or smokers. Since both involve addictive substances – and so the desire for them might be irresistible (thus possibly posing independent problems for autonomy) – I offer a different example. Imagine a person, Pete, who has all the cognitive, conative and affective capacities needed for autonomy and that he exercises sufficient self-control that his actions reflect his second-order commitments in most cases. But Pete is overweight and underexercised, and, when he fails to maintain a healthy exercise and diet regime, he is weak-willed. Pete has first-order desires that motivate the actions (or lack of actions) that lead to his weight and fitness problems, but he also has desires to resist them. Given that he desires many other things that he (correctly) believes are incompatible with satisfying his gluttonous and idle desires, he forms a second-order desire to resist his first-order desires to overeat, eat high-fat foods and watch television rather than work out. He does not want those desires to be effective in leading him to action. Insofar as he continues to act on the motives that lead to his health problems after he has rejected them and endorsed instead first-order desires to eat sensibly, lose weight and exercise regularly, he acts contrary to the desires he wants to be effective; he may feel alienated from his effective desires, experience regret or even shame when he continues his unhealthy ways and so on. Though he has structured his motivations in a way that confers authenticity on some of his desires, he fails to make that structure effective. His actions, when he succumbs to his desire for a big piece of cake followed by a nap, are not autonomous, though he is an autonomous agent capable of self-direction. Autonomous action requires not only the general psychological capacities which make agent autonomy possible, but the self-control to carry out one’s autonomous

The Value of Values

93

projects and act on one’s authentic desires as well.28 When a person acts from weakness of will, he has the abilities for self-definition, but lacks the specific control needed to act autonomously. His action is not an instance of self-direction, and does not express his authentic self. The features which make Pete’s behaviour an instance of weakness of will can be generalized: a person P is weak-willed relative to some first-order desire D if and only if (a) P has reflected upon D and endorsed it as a reason for action, making it the object of a second-order volition V (to act on D in suitable circumstances), yet acts contrary to D in circumstances that make the satisfaction of D possible for P; or (b) P has reflected upon D and rejected it as a reason for action, making it the object of a second-order volition V (to not act on D), yet acts on D in circumstances in which P could have resisted D’s motivational influence.29 Given this characterization we can see that to be weak-willed a person must have the capacities of critical reflection and volition-formation and he must act contrary to his second-order volition, either by acting on a desire which has been rejected or by failing to act on a desire which has been endorsed. Hence, all weak-willed actions will be actions of autonomous agents but they will fail to be autonomous actions. Autonomous actions, by contrast, are those of autonomous agents and are motivated by desires that have been approved of at the second order. (Again this must be qualified by the conditions required to ensure that identification with a desire is not just necessary to confer autonomy upon it, but also sufficient.) This distinction between agent and act autonomy is, then, a distinction between being able to structure our motives by reflecting critically upon them and making some of them authentic reasons for action, on the one hand, and having those desires we have endorsed be effective in motivating our actions, on the other. Act autonomy is the actual condition of directing oneself. This distinction also has clear relevance to our discussion of intrapersonal coherence and ambivalence. Complete coherence among a person’s first-order motivations is not necessary for agent autonomy; coherence is more important at the level of volitions, but even here some inconsistency can be tolerated without undermining agent autonomy. For one may have the reflective capacities which make one capable of self-direction, and one may exercise those capacities in developing an authentic self, while remaining ambivalent about some aspects of one’s motivations. Yet coherence, or lack of personal ambivalence, is absolutely necessary for both desire and act autonomy. One cannot act autonomously with respect to desires about which one is ambivalent, and while one remains ambivalent one cannot determine which of the desires about which one is ambivalent is authentic. Thus far I have argued that autonomous desires are those held authentically and non-ambivalently, and autonomous action is motivated by such desires. But intrapersonal coherence, authenticity and an absence of weakness of will are only necessary conditions of autonomy; even jointly they are not sufficient. The conditions upon autonomous desires thus far articulated make the autonomy of a desire depend just upon the subjective states of the agent: does she wholeheartedly identify with it? The agent’s attitudes, as discussed above, are sufficient for

Liberty, Games and Contracts

94

authenticity, but more is needed than authenticity for autonomy. Some external conditions must also be placed upon the process and context of identification to render identification with a desire autonomy-conferring. Those conditions cannot be based on an objective theory of human good or human interests, of course, within a liberal approach. We cannot declare that some desires cannot be the object of autonomy-conferring approval on the basis of their content. This is the strategy taken by objectivists about human good, such as Gary Watson, Thomas Hill Jr, Diana T. Meyers, Paul Benson and Susan Wolf.30 But we can, and should, impose procedural restrictions upon approval by identifying certain contexts or processes of desire formation which are hostile to autonomy.31 There are some social environments which are such that individuals who are subjected to them cannot adopt autonomous desires because those environments subvert the capacity for critical reflection itself. And there are some actions whose consequences are known to seriously impair the ability to exercise control of the kind required for autonomous action. Though I cannot fully articulate here the range of procedural conditions that must be satisfied to ensure that approval of a desire is autonomyconferring, they may be summed up as the conditions which make reasonsresponsive evaluation of one’s motivations possible and efficacious.32 While spelling out these conditions is necessary for a fully satisfactory account of autonomy, they need not be detailed here because it is the requirement of authenticity that is crucial for discussing the implications of Narveson’s failure to take seriously the conditions of autonomy, and the necessary and sufficient conditions of authenticity have now been given.

The Value of Autonomy Narveson considers only one way in which the requirement of autonomy could matter to his defence of libertarianism, that offered by Haworth. Haworth argues that autonomy is a basic value and so its protection must also come within the first principles of an adequate moral theory. Indeed, Haworth argues that the value of negative liberty is based in its role of supporting and allowing expression of autonomy. For Haworth, autonomy is the more basic value and liberty is derivative (and not all that is needed to support autonomy).33 As Haworth argued: Autonomy-based rights are those modes of treatment to which the individual is entitled in order that his life and situation may have three characteristics: his domain for autonomy should be open; it should be open de facto, and not merely de jure; and he should have both the ability and the opportunity actually to live autonomously within his domain for autonomy.34

Narveson rejects this view because it assumes that autonomous people must value both their own and others’ autonomy. Narveson argues that we cannot assume that autonomous agents care about the autonomy of others or want to set up a system that promotes or protects the autonomy of others. Among his arguments against Haworth’s autonomy-based view of rights, Narveson questions whether a

The Value of Values

95

programme of positive rights would actually be respectful of autonomy. He argues that it would not, because if autonomous people do value the autonomy of others then they will voluntarily act to foster and protect the conditions of autonomy for them; negative liberty will be enough to respect their valuing of autonomy. And if there are some who do not value the autonomy of others, then imposing restrictions on them in the name of a value they do not share does not actually respect their autonomy: ‘it is to override their judgment about what they should do with their lives – the very sort of judgments that are definitive of autonomy.’35 Narveson continues: ‘The case for designing our institutions on the assumption that autonomy is a sort of universal social value, capable of justly imposing many irksome requirements on their reluctant participants, does not fall straight out of the assumptions either that autonomy is highly valuable to the individual or that it is a prerequisite to any liberty worth having.’36 Furthermore, even if one wants to protect the conditions of autonomy for all (because one values autonomy generally), it does not follow that one wants to accept the duties that would be imposed by a scheme of positive rights adopted to realize those conditions. The cost of such rights may simply be judged to be too high. I think Narveson is right to reject the view that autonomy is a basic value. Many people clearly do not value autonomy for others. And presupposing such a universal value, such that liberty is only valuable if it protects or fosters autonomy generally, is unsupportable on the liberal premises with which we began. This is an example of an objective theory of autonomy, which rules out certain desires as possibly being autonomous on the basis of their content. But there is an entirely different role for the theory of autonomy to play within the liberal project, which Narveson never explicitly addresses and which once attended to places limits on the right to liberty of which he is so fond.

Autonomy as a Precondition of Values I do not argue for an autonomy-based theory of rights, or that autonomy is a value in the sense at issue in the debate between Narveson and Haworth. To see what role autonomy does play, let us return to the place of values within the contractarian project. Narveson writes that the contractarian procedure rests on two assumptions: (1) ‘The individuals whose polis is in question have various values and apply reasoning on behalf of them. The assumption that those values are various is what creates a need for morals and/or politics. And (2) those values lie within a range, though a very wide one, such that there is the possibility of cooperation.’37 On this view, we cannot assume that people value anything, other than the satisfaction of their values. Narveson argues against Haworth because he thinks Haworth’s view requires the assumption that all autonomous agents value their own and other people’s autonomy. But another way of thinking about the matter is this: only autonomous desires can be values in the sense needed in the contractarian theory. Consider: ‘the values of individuals or sets of them enter into the proceedings only as the values of those

96

Liberty, Games and Contracts

persons. They do not, any of them, enter as considerations having weight independently of the fact that they are the concerns and interests of those whose values they are.’38 Only autonomous desires can be values in this sense. It is not that autonomy as a value has independent weight, but that only autonomous desires are values. Only autonomous desires can count as the values which function as inputs to the contractarian method, because only autonomous values guarantee the internalism that the theory requires. If we can have desires from which we are estranged, which we reject as reasons for action, then they must be ruled out if we are to assume that agents necessarily care about the satisfaction of their desires. If the values of agents are to have weight as their values, rather than independently, then they must be authentic as described above. Autonomous agents necessarily care about their autonomous desires, but they may not want any number of their non-autonomous desires to be fulfilled. That Narveson gives priority to the subjective desires of agents is clear. One of the central tenets of liberalism for Narveson, and its differentia from other political theories that accept that political power should be exercised only for the common good, concerns how that ‘good’ is to be understood: ‘the good of the people is determined by those people themselves. More precisely, the good of any individual, for political purposes, is a matter on which that individual himself is the ultimate authority.’39 Narveson’s view does not imply that the person is an infallible judge of his own good (and in fact he denies this in much of his political work), but that he has final authority. Indeed, the language of authority is oft-utilized in Narveson’s work. But claims to authority are not self-validating. Governments claim authority to do all sorts of things, and Narveson thinks those claims are false. He denies that they have the authority they claim. Why, then, should we assume that individuals have the authority they claim? When we act non-autonomously, do we think we should have that authority? Do we want it for ourselves or our loved ones? Do we recognize it in others when they are clearly acting non-autonomously? Could this claim to authority, like its political counterpart, be false? Although Narveson says repeatedly that ‘Each person is to be regarded as being the ultimate authority on what is good for himself’,40 any argument for that position must presuppose that the persons in question are autonomous agents. It seems unlikely that anyone would believe the claim to authority if it was made by anyone but an autonomous agent. And even when claimed by a person having the capacity for agent autonomy, it seems false when claimed on the basis of non-autonomous desires or as an argument for non-interference in the pursuit of non-autonomous desires.

Drugs, Schools and Information Narveson’s willingness to take desires at face value, and accord to them all the normative force his theory requires, is perhaps nowhere more evident than in his rejection of drug legislation. He describes the desire for drugs in the following terms:

The Value of Values

97

Some people … might well have the idea that being ‘high’ is a major part of the good life. Others, such as myself – boringly ‘straight’ on matters like this – are probably quite unable to understand what you see in it; just as we likewise do not understand what you see in your religion of choice, or your preference for calamari, or your taste in spouses, or the music of The Rolling Stones.41

The desire for heroin, it seems, is no different from the desire for rock and roll. Drugs, Narveson continues, put the users in a certain state of mind, give them a feeling known as ‘being high’. This feeling seems to be much enjoyed by many people, and in some cases not only enjoyed, but life can seem pretty miserable without it. So it goes. It is not difficult to think of other things with the same general profile. Love, for example. Or gourmet cuisine. Learning to contain our involvements so that they don’t take over our lives, and to cope with their side effects, is part of the discipline of life for anyone, applying as much to the drugtaker as to anyone else.42

Thus he dismisses in a single sentence the physical dependence created by (some) drugs. But drugs are not like love or good food, neither of which is addictive in the sense of creating chemical dependence. Of course other things are addictive in this sense, for example, tobacco, though their use is permitted. Does that show that dependence is morally irrelevant, or that we are simply being inconsistent in our policies? No, but what it does show is that drugs must have some further feature which distinguishes them from other addictive substances; and they do: they erode the mental and volitional powers of those so addicted. In this way they threaten the autonomy of their users, making them less reasons-responsive in general and virtually incapable of exercising the control needed if they want to resist their desire for them. Neither their addictive properties nor their corrosive effects on the ability of their users to exercise autonomy are tied to the fact of their illegality (unlike many of the negative effects that Narveson talks about, which are predominantly or wholly a result of their legal prohibition, for example, their very high cost). If the desire to take drugs is non-autonomous, then it need not be accorded authoritative status for moral purposes. We cannot suppose that the desire for them is one that the addict cares about being satisfied. Narveson cannot evade this challenge by appeal to the choices addicts make, moreover. While he often indicates that the mere fact that someone chooses to engage in an action is evidence that he values that action or the state of affairs it brings about, the claim must be understood as containing an unspecified ‘prima facie’ restriction, and the defeasibility condition must be genuine. If it is not then he is relying upon a notion of revealed preference, whereby preference is perfectly revealed in choice. The problem with such an understanding of preference for contractarian purposes is that it makes practical irrationality impossible, along with other phenomena such as weakness of will. But if people cannot but act rationally then the contractarian has no normative work to do and no useful advice to give. Such an understanding of preference also robs the theory of its empirical status, since the claim that people are practically rational is then rendered unfalsifiable, even in principle.

98

Liberty, Games and Contracts

Similar objections could be raised to Narveson’s rejection of compulsory public education. A plausible case can be made that the cognitive, conative and affective capacities that make autonomy possible require a certain level of education (relative, no doubt, to culture). And while it may be true that a more flexible approach to educational schemes than the public system currently allows would better enable persons to realize their autonomous values, some limit upon the kinds of alternative educational regimes would surely be justified. Religious schools in which critical reflection is crushed, where intolerance is promoted, and where the capacities for reasons-responsive evaluation of one’s beliefs and values is severely discouraged ought not to be allowed, because they would deny students exposed to them the chance to develop the autonomous values in virtue of which they matter to the polity and are members of the social contract. We might say that maintaining the social conditions under which autonomy is possible must matter to everyone, in the sense that maintaining the ‘circumstances of justice’ matter to everyone: they are preconditions of the social contract, which we all instrumentally care about. Finally, consider Narveson’s discussion of information. While he insists that intentional fraud would be prohibited under a justified social morality, he nonetheless treats information as a good to be otherwise distributed by regular market forces. The reason for this difference is that fraud is a violation of negative liberty, while not supplying information is not. It is unjust to engage in fraud because ‘information matters because without it we cannot relate our interests to our actions’.43 More specifically: To communicate successfully, parties A and B require a common system of language sufficient to get the message in question across. These messages concern the way things are around them. If A intentionally misleads B, A utilizes this common system effectively to bring it about that B believes Q rather than P, where P is, so far as A knows, true. Now, when B acts on Q rather than P, B expects to bring about a result, R, which B supposes will occur if and only if Q. Since P is what is true, however, B’s action will not result in the intended R. A has thus effectively brought it about, by action for which A is responsible, that B does not do what B wants. This, by definition, infringes B’s freedom.44

Contrast this with his position on information in general. Information is a good like any other and it can be supplied by regular market forces. ‘It is, in short, a mistake to think that the public ought to require producers to supply information, or require others at public expense to do so, in order to uphold the market.’45 Agreed; we do not need to force suppliers to provide information to uphold the market. But what about supplying information in order to uphold autonomous choices in the market? That, it seems, is an entirely different matter. If choices in the market can be rendered non-autonomous as a result of lack of relevant information (a condition that undermines reasons-responsiveness perhaps paradigmatically), then they lose the grounds of their protection under the right to liberty. If the right to liberty is to be universal, we must maintain the conditions under which people have such a right, and that requires protecting the capacity for autonomous choice. We must

The Value of Values

99

protect the capacity for autonomous choice, not because we value autonomy per se, but because it is a precondition of values having the normative force required by contractarianism. Notes 1

2 3 4 5

6 7 8 9 10 11 12. 13

14

15

16 17 18

19

Jan Narveson, The Libertarian Idea (Philadelphia, 1988), 114–15. For more on the commitment to internalism within contractarianism see my ‘Two Virtues of Contractarianism’, The Journal of Value Inquiry, 37/3 (November 2003): 395–414. Narveson, Libertarian Idea, 131. Narveson, 161. Narveson, 26 Jan Narveson, ‘Liberal-Conservative: The Real Controversy’, The Journal of Value Inquiry Special Issue on Liberalism, J. Narveson and S. Dimock (eds), 34/2–3 (September 2000): 171. Narveson, Libertarian Idea, 8. Narveson, 161. Narveson, 175. See Lawrence Haworth, Autonomy (New Haven, CT, 1986). Narveson, Libertarian Idea, 81. Narveson, 81–2. Narveson, 166. Stephen Darwall has drawn a similar distinction concerning the various meanings which the phrase ‘reasons for action’ might have. Cf. Impartial Reason (Ithaca, NY, 1983), 28. See Harry Frankfurt, ‘Freedom of the Will and the Concept of a Person’, Journal of Philosophy, 68 (1971): 5–20; reprinted in his The Importance of What We Care About (New York, 1988); page numbers refer to the latter collection. Gerald Dworkin, ‘The Concept of Autonomy’, Science and Ethics (New York, 1981); reprinted in John Christman (ed.), The Inner Citadel: Essays on Individual Autonomy (New York, 1989), 60; page numbers refer to the latter collection. See also Gerald Dworkin, The Theory and Practice of Autonomy (New York, 1988), 15. Although Dworkin rejects identification as necessary for autonomy in the latter work, careful review of the four arguments he gives for this reversal supports the view that identification cannot be sufficient for autonomy, but not that it is not necessary. This way of presenting the bi-level view is overstated, for acting on a higher-order desire is only a necessary but not a sufficient condition of personal autonomy on any plausible version of the theory. Likewise, the authenticity of one’s desires is only necessary for their being autonomous, but it is not sufficient. It is an open question whether only human beings are capable of personhood on this view. Frankfurt, ‘Freedom of the Will and the Concept of a Person’, 12. Harry Frankfurt, ‘Identification and Wholeheartedness’, in F. Schoeman (ed.), Responsibility, Character, and the Emotions (New York, 1987); reprinted in The Importance of What We Care About, 164–5; page numbers refer to the latter collection. Although Frankfurt develops his bi-level account of desires in the course of explicating freedom of the will rather than autonomy, the difference between those tasks is unimportant for our purposes (though they might matter in other contexts, such as

100

20 21

22 23

24 25 26

27 28 29

30

31

32

33 34 35

Liberty, Games and Contracts discussions of coercion). Cf. Susan Dimock, ‘Personal Autonomy, Freedom of Action, and Coercion’, in S. Brennan, T. Isaacs and M. Milde (eds), A Question of Values: New Canadian Perspectives in Ethics and Political Philosophy (Amsterdam, 1997), 65–86. Frankfurt,‘Freedom of the Will and the Concept of a Person’, 18. Frankfurt makes a similar division concerning different types of conflicts between desires, and also argues that some conflicts require resort to a different order of desires rather than ranking them within a single order to solve. Cf. Harry Frankfurt, ‘Identification and Externality’, in A. Rorty (ed.), The Identities of Persons (Berkeley, CA, 1977), 66–7. Cf. Frankfurt, ‘Freedom of the Will and the Concept of a Person’, 18. This provides an answer to one of the objections raised against the hierarchical model by Marilyn Friedman in ‘Autonomy and the Split-Level Self’, Southern Journal of Philosophy, XXIV/1 (1986): 19–35. I am indebted in this discussion to Richmond Campbell’s work on ambivalence in SelfLove and Self-Respect: A Philosophical Study of Egoism (Ottawa, 1979). Campbell, Self Love and Self-Respect, makes precisely this point at 172. Though the structure of the conflict is different, the same points can be made if one identifies at the second order with the desire to x and identifies also with the desire to not-x. Frankfurt, ‘Identification and Wholeheartedness’, 165. Haworth argues that self-control is a necessary condition for living autonomously, in Haworth, ch. 2. In characterizing weakness of will as necessarily relative to some specific desire (D), I am contrasting the phenomenon with the general ‘character trait’ of weakness of will. For an interesting discussion of weakness of will as an (immoral) character trait, see Thomas Hill Jr, ‘Weakness of Will and Character’, Philosophical Topics, XIV/2 (Fall 1986). Gary Watson, ‘Free Agency’, Journal of Philosophy, LXXII/8 (April 1975): 205–20; reprinted in J. Christman (ed.), The Inner Citadel: Essays on Individual Autonomy (New York, 1989). Thomas Hill Jr, ‘Servility and Self-Respect’, The Monist, 57 (1973); ‘The Kantian Conception of Autonomy’, The Inner Citadel (1989); and Autonomy and Self Respect (New York, 1991). Diana T. Meyers, Self, Society and Personal Choice (New York, 1989); ‘Personal Autonomy and the Paradox of Feminine Socialization’, Journal of Philosophy, LXXXIV/11 (1987): 619–28; and ‘The Socialized Individual and Individual Autonomy: An Intersection between Philosophy and Psychology’, in E. Feder Kittay and D. T. Meyers (eds), Women and Moral Theory (Lanham, MD, 1987). Paul Benson, ‘Freedom and Value’, Journal of Philosophy, LXXXIV/9 (1987): 465–86; ‘Autonomy and Oppressive Socialization’, Journal of Social Theory and Practice, 17/3 (1991): 385–408. Susan Wolf, Freedom within Reason (New York, 1990). The foremost procedural account is that provided by Gerald Dworkin, ‘Autonomy and Behavior Control’, Hastings Center Report 6 (1976); The Theory and Practice of Autonomy. I take the term ‘reasons-responsive’ from John Martin Fisher, ‘Responsiveness and Moral Responsibility’, in F. Schoeman (ed.), Responsibility, Character, and the Emotions (New York, 1987), though I employ it for different purposes than he did. Narveson, Libertarian Idea, 167–8. Haworth, 213; quoted in Narveson, Libertarian Idea, 169. Narveson, 170.

The Value of Values 36 37 38 39

40 41 42 43 44

45

101

Narveson, 171. Narveson, 162. Narveson, 162–3. Jan Narveson, ‘Toward a Liberal Theory of Ideology: A Quasi-Marxian Exploration’, Reason Papers, 20 (Fall 1995); reprinted in Jan Narveson, Respecting Persons in Theory and Practice: Essays on Moral and Political Philosophy (Lanham, MD, 2002), 106. Page references are to the latter collection. Jan Narveson, ‘The Drug Laws’, in Respecting Persons in Theory and Practice, 247. Narveson ‘The Drug Laws’, 247. Narveson, ‘The Drug Laws’, 257. Jan Narveson, ‘Moral Realism, Emotivism, and Natural Law’, in Respecting Persons in Theory and Practice, 73. Jan Narveson, in ‘The Anarchist’s Case’, in J. T. Sanders and J. Narveson (eds), For and Against the State (Lanham, MD, 1996); reprinted in Respecting Persons in Theory and Practice, 201, n. 6. Page numbers refer to latter collection. Narveson, Libertarian Idea, 202.

This page intentionally left blank

Chapter 7

Simple Games and Complex Ethics1 Peter Danielson

In an earlier paper, we criticized the political thesis of Narveson’s The Libertarian Idea, which we characterized as rational libertarianism. We argued there that shifting focus from the prisoner’s dilemma to the related game of chicken provides a clear situation where informal morality – especially one based on negative rights – is not sufficient. Here we take up Narveson’s reply to our criticism and turn from political theory to ethics. When we follow Narveson’s suggestion and frame these issues using the tools of game theory we find that his rather intuitive moves from rational choice to a morality of cooperation fall short. Turning to broader methodological issues, we caution that it is easy to move from using models to explain ethical complexity, to addressing simple models instead of real-world ethical complexity, which we illustrate with the example of the shortfall in blood donation in Canada.

Introduction It is a pleasure to contribute a chapter to this Festschrift for Jan Narveson, whom I have known and admired since I was a graduate student. I have have been impressed by Jan’s enthusiasm and remarkable dedication to philosophy. He has always been an exemplary colleague, supporting others’ work in numerous, unsung ways. He combines two qualities that I especially appreciate: he is open to criticism and willing to engage in wide-ranging discussion of methods in ethics. Therefore, it is fitting to begin this chapter with an ongoing critical debate between Jan and me over the role of two models – the games of prisoner’s dilemma and chicken – of the state of nature in contractarian political philosophy. I will then expand the discussion to a more general contrast of methods in ethics. The two parts support a common conclusion: while simple games help us to understand some issues in ethics and politics, prospects for reducing either to simple rational choice models are dim. Moreover, there is a risk of confusing our complex subjects with our simple models. In ‘The Rights of Chickens: A Rational Foundation for Libertarianism?’ we characterized the political thesis central to Narveson’s The Libertarian Idea as rational libertarianism: ‘Amoral rational agents would reject Thomas Hobbes’s

103

104

Liberty, Games and Contracts

authoritarian political solution in favor of a libertarian moral solution to … the stateof-nature social problem.’2 We criticized this thesis, arguing against focussing so exclusively on situations modelled by one game, the prisoner’s dilemma (PD; see Figure 7.1). In contrast to Narveson, we argued that some situations are better modelled by the related game of chicken (see Figure 7.2). This is important to Narveson’s political thesis because the game of chicken characterizes a set of clear cases where informal morality – especially Narveson’s preferred candidate based on negative rights – does not suffice to solve structural problems of social interaction in the state of nature. We concluded that rational agents would choose a stronger state than Narveson’s libertarianism. We shall not repeat that criticism here, but pick up the debate with Narveson’s reply to it in ‘The Anarchist’s Case’.

Figure 7.1. Prisoner’s Dilemma

Figure 7.2. Chicken Narveson’s Reply Matters get considerably more complicated when we consider the possibility that human interactions might have the structure known as ‘Chicken’ rather than Prisoners’ Dilemma. In both games, the participants have a common second-best; but in the Chicken game, they have a common worst outcome, in contrast to the PD, in which both parties have a common third-best outcome. In PD, therefore, one player’s best is the other’s worst: A’s best response to B’s threat of aggression is defense – A can’t do better by knuckling under. But in Chicken, unfortunately, he can. If he insists on resisting, and the other does not back down, then both will lose; if, instead, A knuckles under, then A may be enslaved, but at least he is still alive. Faced with the choice between being the chicken and being the hawk, A may prefer being the chicken, nor is A clearly irrational in so choosing, as … Danielson … points out. This problem is clearly not an easy one, but certain general comments are in order. One is that if it is possible to prevent social Chicken games from developing, then it is plausible to say that we should do so, and plausible for the same reasons that recommend cooperation in PD. It is not at all surprising that it is a theorem of commonsense morality that we ought not to employ coercion, just as we ought to refrain from violence. This response may be lame, for the question then is whether social rules of that kind have the

Simple Games and Complex Ethics

105

clout necessary to achieve their purpose. But then, the very same question can be asked about PD; and certainly the answer, in a vast range of real-life cases, is that it certainly does. Social life is generally not a fracas, despite the many temptations for the unscrupulous or the sociopathic. That we have much to gain by cooperation and much to lose without it, as a general observation about our situations vis-à-vis our natural environment and each other, seems too obvious to need more than a mere mention.3

Narveson graciously accepts our main criticism and moves the discussion forward. Indeed, his title clarifies the issues, by reminding us how close the libertarian ideal is to anarchism. He accepts my political conclusion: that the existence of situations modelled by the game of chicken favours the state. ‘Nevertheless, Danielson’s challenge does seem to be right in one very important respect: it probably accounts for the existence and survival of the state.’4 Leaving aside the question of political theory, his reply raises questions about game theory and ethics generally, to which we turn. Is It Rational to Avoid Chicken? Acknowledging that situations modelled by the game of chicken are problematic for his theory, Narveson’s main reply is to try to avoid these situations. His first general comment is, ‘if it is possible to prevent social Chicken games from developing, then it is plausible to say that we should do so.’ We take this as an admission that his approach to ethics cannot produce an acceptable outcome for the game of chicken, so he falls back to what he must see as an easier problem, namely preventing situations modelled by the game from arising. His reply is complicated; he characterizes chicken in a morally rich way (mentioning violence, coercion and ‘a fracas’) and suggests invoking moral rules to avoid the situation. Putting these complications aside, we focus on Narveson’s claim to bring rational choice to bear on ethics. One would expect a rational choice ethics to make its claim turn on a simpler question: would rational agents choose to play chicken rather than the prisoner’s dilemma? Posing the simpler question in terms of rational choice reminds us of the ground rules of a game-theoretic approach. The main point is that one cannot choose the game one plays; the theoretical framework contextualizes choice by the game, which is exogenous. One gets to choose options within a game. Therefore, to apply the rational choice method to choosing one game over another requires that the two games be alternatives in a prior, third game. Taking the simplest case, we consider a single agent who faces a choice between two games. Since Narveson has in mind the problems of chicken versus PD, let these be the two options (shown in Figure 7.3). The right alternative puts one in a game of chicken; the left alternative puts one in a game of PD (in each case with a similarly selected opponent). The prior game is not strategic; it is a game against nature. For example, one needs to choose between two game tables or rooms – one where blackjack is played, the other, poker. In Narveson’s case, one is choosing between two states of nature – one structured as a PD, the other as chicken. In this prior game against

106

Liberty, Games and Contracts

Figure 7.3. Choosing between PD and Chicken

nature, a rational agent will choose the outcome with the larger expected utility. This is a case of choice under uncertainty. To solve it we use backwards induction. We begin at the end, solving the two alternative games, imputing the expected utility of the outcome of each as the value of the corresponding option. On standard assumptions, outcome of the PD is the unique equilibrium of the dominant strategy, ; the expected utility of the left branch of Figure 7.3 is 1. Calculating the expected utility of the outcome of the chicken on the right is more complex, since it lacks a dominant strategy. There are three equilibrium strategies: D, C and the mixed strategy, D/C with equal probability. Although, as we can see, this mixed strategy is not stable, it allows us to simplify the analysis at this stage to focus on it by noting that since the players cannot coordinate on one of the two other equilibrium outcomes and , they will choose between C and D with equal probability; in any case, get each of the four outcomes {0,1,2,3} with equal probability, and expect a utility of 1.5. Therefore, faced with the alternatives of playing comparable games of chicken and PD, a rational agent chooses the former (bolded path in Figure 7.3), as it has a better expected outcome.5 This is an important result, and not only because it is likely not what Narveson expected. The interesting general point is that while we might find chicken troubling, compared to the baseline (alternative) of the prisoner’s dilemma, it is a better game for rational agents to play. With the same payoff set, PD offers a certain third-best while chicken offers the mean of all the outcomes, which is better. If Narveson finds the PD preferable to chicken, there are likely two reasons, one based on rationality, the other on morality, respectively. We take them up in order in the next two sections.

The Complexity of Extending Rationality Following Gauthier, Narveson assumes that there is a way for rational agents to coordinate on the second-best outcome in the PD.6 On this assumption – that agents can commit themselves to what Gauthier calls constrained maximization – the expected utility of the PD is 2, which is greater than the expected utility of chicken. However, Gauthier’s account has not been accepted by the games theory literature.7

Simple Games and Complex Ethics

107

It makes very strong assumptions about both commitment (to dominated outcomes) and common knowledge of other players’ strategies, neither of which are analysed as part of the game. We will not address these issues here.8 Instead, we will note that even were one to apply constrained maximization, one would need to do so to both alternative games, the PD and chicken. Unfortunately, it is not clear what constrained maximizers do in the game of chicken.9 The reason for this is quite central to the project of using rational choice in ethics. Chicken turns on the credibility of threats. Crudely, constrained maximization seems to enable rational agents to make threats, or perhaps to resist them. Enabling threats runs against the core doctrine of rational choice as modular.10 Perhaps enabling threat resistance makes the outcome of chicken uncertain. These sketchy remarks already support our first line of criticism: once ‘extended’ from the most favourable cases, rationality ceases to supply a simple foundation for an ethics. The problem is deeper than problems with Gauthier’s theory, which we can sidestep by turning from the principle of constrained maximization to the simpler conditional cooperation strategy (CC) on which Gauthier and Narveson focus in the PD game. In chicken, CC is the strategy I called Resist play C with C and D with D.11 So we can ask Narveson’s question about these strategies: which of the two games, PD and chicken, is preferable given the three strategies in Figure 7.4?

Figure 7.4. PD and Chicken with a Conditional Strategy

Again, we need to use backward induction to solve the two games. The problem now is that there is no simple solution to either. The new strategy does introduce new equilibria (CC/CC or C) into both games. However, unfortunately, the old equilibria (D/D and C/D in PD and chicken, respectively) remain. Therefore, we are offered pairs of equilibria instead of the needed single outcome. Worse, game theory does not provide us with a way of calculating the probability within these pairs, so we cannot put a value on the expected outcomes of these two games, and therefore cannot solve the decision problem of choosing between the two. Fortunately, evolutionary game theory (EGT) introduces analytic tools that can help us to address our question. EGT shifts the focus from single complex rational agents choosing strategies, to populations of simpler agents exemplifying strategies. The question of choice of strategy becomes a question of the dynamics

Liberty, Games and Contracts

108

of these possible populations of strategies. The good news is that simplifying assumptions allow us to model the interactions of these populations as equations (the replicator dynamics) that allow us to simulate these results. The bad news is how complex the dynamics are revealed to be, undermining our confidence in intuitive judgments about what is rational. For example, we wrote of chicken played by these three agent types, ‘The ecological equilibrium of such a population will see resister [CC] die out and dove [C] and hawk [D] form a stable predator/prey relation.’12 However, this claim turns out not to hold in general.

Figure 7.5. Chicken with Equal Initial Populations

Pop

To see why, we turn to dynamic analysis (explained below). The graphs in this section show the way a fixed population of 300 players is divided between the three strategies over 40 generations. Contrasting Figures 7.5 and 7.6, Figure 7.5 bears out

Figure 7.6. Chicken with Unequal Populations

Simple Games and Complex Ethics

109

our claim about equal populations: C and D end up with 150 agents each from an initial population of 100 each of C, D and CC. However, notice that the path to this outcome is not direct. First the D population declines and the C population climbs. We can see why by looking at Figure 7.3 where the payoff for the case of equal populations is in the ratio 5:4:3 for C:CC:D. This sets the initial trajectory with C growing at the ratio of 5/4, CC 4/4 and D 3/4, where 4 is the mean payoff. However, as C increases, it makes the environment better for D than for CC, so D recovers and CC dies out. Nevertheless, this outcome depends on the initial population distribution. In Figure 7.6 we increase CC’s initial distribution (to 42 per cent and decrease C and D each to 29 per cent) until the populations stabilize in a new equilibrium, with CC at 44 per cent of the population, C at 56 per cent and D 0 per cent. Since no strategies are changed, this model is ‘ecological’ rather than evolutionary in the full sense.13 We use the model developed in Skyrms.14 Assuming that each player meets each of the others once before reproducing, we calculate the fitness of a strategy, u, of a strategy A by multiplying its payoff playing against strategy B, (A|B), by the proportion of B in the population, p(B) (see Equation 7.1).

Equation 7.1 Next we take the mean fitness, ū, by weighing each strategy fitness by its proportion of the population (Equation 7.2).

Equation 7.2 Now we can calculate the new population, p′(A), by the old population times the ratio of A’s fitness to the mean fitness (Equation 7.3).

Equation 7.3 Turning to the prisoner’s dilemma game, the results are more straightforward than in chicken. With all populations, D dies out and the outcome is a mix of mostly CC and some C. Figure 7.7 shows the case of equal populations.

Liberty, Games and Contracts

Pop

110

Figure 7.7. PD with Equal Initial Populations

Pop

Figure 7.8 doubles the initial C population, leading to some interesting dynamics. D shoots up, but then is killed off by CC. Unfortunately, the dynamics are complex enough to mislead intuitive approaches to environments with mixed populations. In Artificial Morality we concluded, based on very small populations and crude methods, that C would grow and lead D to invade CC.15 Here we see that even in the most favourable case for D, its growth is limited and, crucially, limits the population of C on which it depends, but not that of CC. What can we conclude from this analysis? First, and most directly, where does the dynamic analysis leave our original question of the rational choice between chicken and PD? The introduction of conditional strategies does move the question

Figure 7.8. PD with 67% C Initial Population

Simple Games and Complex Ethics

111

in the direction Narveson desires, by making the PD more attractive. The dynamic analysis has the PD game with every initial population eventually paying off 2, while only some games of chicken do so. The remainder of the distributions in chicken games pay off 1.5 (due to the cost of uncoordinated C/D choices). Second, and mainly, the dynamic analysis shows that introducing the ‘refinement’ of new, conditional strategies makes rationality much more complex. At each stage we need to ask why these particular strategies and these populations are included in the analysis. For example, we have only so far considered the most obvious and morally attractive conditional strategies, but the ‘cognitive science’ that generates the three strategies considered presumably can generate other strategies.16 Alternatively, if other factors constrain the generation of all possible strategies, these factors need to be brought into the analysis. Pressing in this direction, in one paper we move from an ecological to an evolutionary model, arguing that we need to consider not only the initial strategies, but also how their components might be recombined in evolution or learning.17 Another paper points out some of the complexity introduced by strategies conditional on other strategies.18 Finally, we have focussed on the equilibria reached by the dynamics of populations, but until we pin down the time span of the analysis, it is not obvious what time scale rational agents would use. (For example, in Figure 7.8, D does best in the first 11 generations.) Summing up, seeking a model strong enough to deal with the simple conditional strategies Gauthier and Narveson rely on, we end up with more than expected. Rather than a simple rational justification for a simple morality of conditional cooperation, we are moved to a complex account of rationality and complex moral strategies.

Ethics and the Real World We turn now to the question of how to test an account of ethics. Narveson’s recent work is filled with examples of applying his approach. He should be praised for keeping his ethical theory in touch with applied issues. Indeed, compared with his willingness to engage practical issues, I see my own work as excessively theoretical and removed from application, and have recently worked to inform my ethical theorizing with better empirical data. This section explores the link between rational choice, ethical theory and real-world behaviour. From this perspective, Narveson’s appeal to the facts of moral experience leaves something to be desired. Look back to how he determines that we would choose to play PD and not chicken: ‘Social life is generally not a fracas, despite the many temptations for the unscrupulous or the sociopathic. That we have much to gain by cooperation and much to lose without it, as a general observation about our situations vis-à-vis our natural environment and each other, seems too obvious to need more than a mere mention.’19 The first problem is that the empirical categories are too crude. Were people playing chicken, with coordinated C/D outcomes, life would not be ‘a fracas’; there would be order (moral rules, compliance and so on) but of an unequal kind, for that interaction. For example, a dog- or wolf-pack is not

112

Liberty, Games and Contracts

‘a fracas’ but a hierarchy. We need finer empirical measures to determine what games people play and what moral strategies they use. This is true because, first, social situations are diverse; more than one game is needed to model the ethically interesting situations. Second, more than one moral strategy can be used to coordinate in most of these games.20 Second, the rational choice/game-theoretic approach tempts one to oversimplify and see the world in its terms. In the quote above, Narveson sees the disturbing factors to ‘the unscrupulous’ and ‘the sociopathic’. Of course, there is a long tradition that simplifies the central problem of ethics to converting Hobbes’s ‘Foole’ or the sociopath.21 However, we conjecture that many ethical problems involve failure of coordination and misunderstanding among moral agents of various kinds. We will not likely progress if we force these problems into the very restrictive model of prisoner’s dilemmas played with would-be sociopaths. One way to support my point is with an empirical example. My example is blood donation in Canada, chosen because we have some recent interesting data and also because it has been the subject of a debate about moral versus economic explanations.22 A recent public-opinion survey about the Canadian blood supply provides our data.23 The voluntary donation programme does not collect enough blood fully to supply Canadian needs for all blood products. A surprisingly small number of Canadians – 3.7 per cent – are giving blood needed by all of us. The way the surprise runs tells us something about our expectations framing this issue. I am surprised that so few donate; it seems unfair that a few regular donors bear this burden for us all. From the rational choice perspective, of course, the surprise is that anyone gives blood for free The data gets more interesting. While 3.7 per cent give, 28 per cent say they intend to give. This gap between would-be donors’ intentions and their actions raises a number of issues. First, it may seem to confirm the rational choice approach, which predicts an ethical ‘cheap talk’ strategy. People who ‘intend to give’ in this weak, ineffective sense may not be receptive to messages directed to those not giving. Second, this gap greatly complicates research. We cannot assume that what people tell us about blood donation – or, presumably, most other morally charged issues – reflects what they believe or really intend to do. More subtly, we need to probe the stability of behavioural equilibria under varying information. The third piece of survey data confirms this. While the large ‘intending’ population is not effective in getting themselves to donate, they apparently do manage to convince a large part of the population that they are donating. Most believe that 26 per cent donate. This is a serious problem of coordination failure. Just as some people drive past road accidents mistakenly believing that others ‘must be’ phoning, stopping or helping, so some may slacken their efforts to donate due to a false impression supported by common reports of good intentions. Even more serious problems of mistaken equilibria occur when people enforce behaviour based on their misperceptions.24 This data indicates problems common to using moral norms to solve social problems. I propose that we shortcut our understanding if we declare the blood

Simple Games and Complex Ethics

113

donation problem a prisoner’s dilemma, expect non-donation, and stop. The data is not sufficient to support this analysis. In particular, since we do not know the roles of altruism and reciprocity (fairness), we cannot predict whether – were we to effectively get the message out that only 4 per cent are giving – most others would give more or less. This basic uncertainty is due to the different reaction functions of reciprocators and altruists. If reciprocators are a larger part of the population (or, modelling variety individually instead of in population terms, a large part of most people), they will give less since others are giving less (than originally thought). Conversely, altruists will give more when others give less, to help achieve the common aim. We conjecture that our population is made up of these two types, and others; but we do not yet know the mix for this or other situations.25 Our own response to the poor state of empirical ethical knowledge is to build new experimental platforms for obtaining new information about the types and proportions of human moral strategies.26 Narveson’s reassurances that we already understand the core problems of ethics, and that they reduce to simple problems of rational choice and a dose of commitment to ethics, undercut the motivation for this work.

Notes 1 2

3 4

5 6 7

8

This research was supported by a Social Science and Humanities Research Council of Canada grant, ‘Modeling Ethical Agents’. Peter Danielson, ‘The Rights of Chickens: A Rational Foundation for Libertarianism?’, in J. T. Sanders and J. Narveson (eds), For and Against the State: New Philosophical Readings (Lanham, MD, 1996), 171–93. Jan Narveson, ‘The Anarchist’s Case’, in J. T. Sanders and J. Narveson (eds), For and Against the State, 195–216. Narveson, ‘The Anarchist’s Case’. Narveson’s argument is different from mine, and new. He asks why citizens do not resist the state by ‘working to rule’. ‘The secret lies in the chicken game. When both parties resist, they both come out worse than if one or the other had knuckled under – but the cost to the individual of knuckling under is characteristically lower than the cost to the state. It’s an unequal competition, and the individual generally loses’ (Narveson, ‘The Anarchist’s Case’). I argued that chicken is a problem that the state is needed to solve. Narveson argues that the state is an equilibrium in a game of chicken. My claim was weaker, as I do not try to specify how the state solves the game. On the other hand, Narveson’s account is looser, as he brings in the state, an institution, as a player. These two games are comparable in this simple sense: they consist of the same four integer payoffs, 0–3. David P. Gauthier, Morals by Agreement (Oxford, 1986). See K. Binmore, Game Theory and the Social Contract: Playing Fair (Cambridge, MA, 1994); Brian Skyrms, Evolution of the Social Contract (Cambridge, 1996); Joseph Heath, ‘A Multi-Stage Game Model of Morals by Agreement’, Dialogue, XXV (1996): 529–52. See Peter Danielson, ‘Playing with Ethics: Games, Norms, and Moral Freedom’, Topoi, 24 (2005): 1–7.

114 9

10 11 12 13 14

15 16

17

18 19 20 21 22

23

24

25

26

Liberty, Games and Contracts See David Gauthier, ‘Deterrence, Maximization, and Rationality’, in D. MacLean (ed.), The Security Gamble: Deterrence Dilemmas in the Nuclear Age (Totowa, NJ, 1984), 101–22; and David Gauthier, ‘Intention and Deliberation’, in Peter Danielson (ed.), Modeling Rationality, Morality, and Evolution, Vancouver Studies in Cognitive Science, vol. 7 (New York, 1998), 41–54. See Skyrms, Evolution of the Social Contract, 22–5. In Danielson, ‘The Rights of Chickens’. Ibid. See Robert M. Axelrod, The Evolution of Cooperation (New York, 1984). Brian Skyrms, ‘Game Theory, Rationality and Evolution of the Social Contract’, in Leonard D. Katz (ed.), Evolutionary Origins of Morality (Thorverton/Bowling Green, OH, 2000), 269–84. Peter Danielson, Artificial Morality (New York, 1992). Assuming a simple framework where the types of strategies available to an agent are determined by its cognitive abilities and environment. However, it is unclear what mechanism underlies conditional cooperation. Therefore, while we assume that if CC is possible, so is CD (which cooperates with D but not C), we are left wondering if CC cooperates with CD and vice versa. Peter Danielson, ‘Evolutionary Models of Cooperative Mechanisms: Artificial Morality and Genetic Programming’, in Peter Danielson (ed.), Modeling Rationality, Morality, and Evolution, 423–41. Peter Danielson, ‘Competition among Cooperators: Altruism and Reciprocity’, Proceedings of the National Academy of Sciences, 99 (2002): 7237–42. See note 3. Indeed, in all of them, if iterated, as any realistic model should be. See Peter Danielson and Chris MacDonald, ‘Singer’s Agenda for Practical Ethics’, Dialogue 35 (1996): 599–610, for Peter Singer’s defence against the sociopath. See Richard Titmuss, The Gift Relationship: From Human Blood to Social Policy (London, 1971); Kenneth J. Arrow, ‘Gifts and Exchanges’, Philosophy & Public Affairs, 1/4 (1972): 343–62; and P. A. Singer, ‘Altruism and Commerce: A Reply to Arrow’, Philosophy and Public Affairs, 2/1 (1973): 312–20. See Canadian Blood Services, World Blood Donor Day Survey Underscores ‘Reality Gap’; Canadians Exaggerate Intentions to Donate, Overestimate the Number of Blood Donors in Canada, 14 June 2004. See Damon Centola, Robb Willer and M. W. Macy, The Emperor’s Dilemma: A Computational Model of Self-Enforcing Norms (Ithaca, NY, 2002). For some groundbreaking applications, see H. Perkins and A. Berkowitz, ‘Perceiving the Community Norms of Alcohol Use among Students: Some Research Implications for Campus Alcohol Education Programming’, International Journal of Addictions, 21/9–10 (1986): 961–76. See Jon Elster, Nuts and Bolts for the Social Sciences (Cambridge, 1989); and Jon Elster, The Cement of Society: A Study of Social Order. Studies in Rationality and Social Change (Cambridge, 1989). See Rana Ahmad, Zosia Bornik, Peter Danielson, Hadi Dowlatabadi, Edwin Levy, Holly Longstaf and Jennifer Wilkin, ‘A Web-Based Instrument to Model Social Norms: Nerd Design and Results’, Integrated Assessment, 6/2 (2006); and Peter Danielson, Rana Ahmad, Zosia Bornik, Hadi Dowlatabadi and Edwin Levy, ‘Deep, Cheap, and Improvable: Dynamic Democratic Norms & the Ethics of Biotechnology’, Journal of Philosophical Research (forthcoming).

Chapter 8

Why Contractarians are not Libertarians … Evolutionarily Speaking Malcolm Murray

Contracts and Liberty Jan Narveson argues that Hobbesian contractarian moral theory provides the necessary backbone for libertarian political theory.1 What matters for contractarians (as much as for libertarians) is that individual liberty is preserved as much as possible without thereby limiting others’ like liberties. From the basis of self-interest, Narveson argues, we can only get rules restricting interference of others’ pursuing their (non-interfering) goals. We cannot derive rules telling us to help others pursue their goals. So long as it is not in Scrooge’s interest to support a welfare programme, there can be no obligation on the part of Scrooge. If no one is obliged to participate in supporting a welfare programme beyond one’s voluntary interests in doing so, then good riddance to any imposed welfare system. Of course, a voluntary welfare system – where no one pays who has not volunteered to pay – is permissible under libertarianism. After all, that is called charity. Libertarianism is not against charity; only coercion. Contractarians are concerned about ex ante agreements, not ex post agreements, however. What one agrees to in one’s current state is not necessarily what contractarians would endorse. ‘Indeed’, Gauthier admits, ‘I do not even suppose that the practices with which we ought willingly to comply need be those that would secure our present agreement.’2 For example, my not (ex post) agreeing to pay my debts is irrelevant for contractarians assuming I voluntarily (ex ante) incurred those debts. Although I may be unwilling to pay my debts ex post, that would not defeat the argument that I may have an ex ante interest in agreeing to a system where one must pay one’s debts. Can a similar argument be made to show that despite Scrooge’s unwillingness to contribute to a welfare fund, he ought to due to some ex ante agreement he should have made? Narveson’s answer, of course, is ‘No’. There is a distinction to be made between agreeing to pay back one’s debts and agreeing to contribute to a welfare fund. For contractarians, so long as the pursuit of some of my self-interested desires and preferences will impede or prevent some of my prudentially approved interests, I may be motivated to accept constraints on pursuing the former.3 We are to adopt plans that afford us the most favourable opportunities, even though this may require us to act in non-maximizing ways relative to our occurrent aims. In order to maximize my self-interest in social 115

116

Liberty, Games and Contracts

contexts, I must convince you that I will not defect against you in any cooperative venture we have reason to engage in together. But to do this, I need to abandon a straightforward maximizing mode of behaviour. I am unwilling to abandon a straightforward maximizing strategy if the net utility of such a change is worse for me than forgoing such change.4 Since my fundamental motivation is self-interest, the disposition I adopt at the ex ante level must be sufficiently sensitive to prevent my being exploited at the ex post level. Maintaining sufficient sensitivity is the key to conditional cooperation, and it is the ability to adopt conditionally cooperative strategies that moves us from the state of nature to a moral state. Although being prone to conditional cooperation will commit us to repay debts, it will not preclude an agreement on libertarian policies. That Scrooge refuses to help Tiny Tim does not necessarily entail that Scrooge is not a conditional cooperator. Being a conditional cooperator means only that, once one is motivated to enter an agreement, one binds oneself into abiding by the terms of that agreement – on the condition that one’s co-bargainers do likewise. But nothing in this dictates what agreements Scrooge is supposed to make. Nothing in the success of conditional cooperation tells us what kinds of agreements must be made. Narveson reads this as precluding pre-theoretical constraints on selfish behaviour. Certainly he is right to make a distinction between (i), agreeing to honour one’s agreements in order to benefit one’s self-interest and (ii), agreeing to help others out in order to benefit those others. Arguments for welfare need not conflate (i) and (ii), however. Although non-libertarian policies will not be the automatic output of rational self-interested agents bargaining from a state of nature, neither will libertarian policies be the automatic output. That is, although we can admit that libertarianism may follow from contractarianism, to base libertarianism on contractarian roots requires something stronger. Narveson needs – and believes he has – a necessary deduction of libertarianism from Hobbesian contractarianism. Can such a deduction work? My answer is a qualified no.

Pareto Infringements Libertarianism is against centralized redistribution in general.5 The justification for this, like contractarianism, is self-interest. If social constraints impede on selfinterests so much that there is a net loss in terms of individual benefit, individuals lose the motivation to abide by (let alone agree to) those constraints. Agreements are sanctified only so long as all concerned parties voluntarily agree, and at least some agents have no interest in voluntarily agreeing to a centralized distribution of goods. We cannot do things to others without their consent. So long as someone is a concerned party and fails to give consent, or vigorously denies consent, this should end the discussion. Libertarians maintain that this is the case with centralized distribution of goods. A welfare state demands Scrooge gives some of his income to the state to have some of it filter down to a worthy cause – a Tiny Tim. Since it is Scrooge’s income, Scrooge is clearly a concerned party. Since Scrooge is a concerned party, he should be entitled to decline. But centralized distribution

Why Contractarians are not Libertarians

117

entails taking Scrooge’s surplus income notwithstanding. So, if a contractarian grounding for centralized distribution is to be maintained, it must either be shown that Scrooge is not properly a concerned party in this matter (perhaps because it is not really his income, a move that requires discussion about property rights),6 or that Scrooge has or should have made some ex ante agreement to help Tiny Tim out that he cannot now, ex post, renege on. I shall examine the second alternative only. Ex ante agreements have to be rational in terms that amoral, self-interested agents would endorse. This is the contractarian mantra. Narveson, therefore, needs to explain why a commitment to centralized distribution of goods and services would fail the ex ante condition of rational, amoral, self-interested support. Libertarians offer two independent arguments for that: Efficiency (E) and Paretian (P). (E): ideally rational agents in an ex ante situation would realize that a libertarian society is good even for those individuals who occurrently think a welfare system is better.7 (P): to further the interests of welfare proponents is to necessarily thwart the interests of welfare hold-outs. The argument in (E) is generally understood to be an empirical argument. Mere armchair philosophizing will not do. Still, the burden of proof lies on (E) supporters to show that it is always irrational for an agent to prefer a welfare system. Since rationality is a tool to further one’s interests given the circumstances in which one finds oneself, one would need to maintain that no one can have an interest in the well-being of another in ways that dominate over the individual’s other more selfish interests. One could rule out positive tuistic preferences in the state of nature, but once one enters into civil society in order to serve – not just peace, but social goods – we have moved from being psychological egoists to being robust human beings replete with tuistic preferences. And given my (nice) tuistic preferences, a welfare state may well be in my interest. That it ought not is not the argument assumed under (E). (E) claims it can never be in my interest, and this strikes me (without further discussion) as untenable. The argument in (P), however, holds even if (E) does not. So let us ignore (E) and examine (P).8 Merely because a social arrangement is in some people’s interests does not warrant contractarian approval unless it is also in the agreed-upon interests of all concerned parties. Should a hold-out to the centralized distribution scheme exist, those who want centralized distributions cannot force the hold-out to contribute to their centralized scheme – even if the hold-out benefits as a free-rider. So long as we cannot do to others without their consent, then we cannot take from Scrooge to pay Tiny Tim. This second line of argument appeals to the Pareto principle. The Pareto principle claims that it is senseless not to prefer a state in which one person is made better off when no one is made worse off. Negatively, the Pareto principle claims one cannot prefer a state in which some are made better off at others’ expense. ‘Whenever we can assess a change as being for the better for some and no worse for anyone, we have a strong case [for enacting that change].’9 Evoking Pareto superiority leads one to endorse Pareto optimality, the ideal state from which any unilateral movement will be Pareto inferior. The libertarian argument is that (a), the Pareto principle ought not be violated, and (b), we can not get a welfare state without violating the Pareto principle. If a large group of people

118

Liberty, Games and Contracts

consent to a welfare system where minimal charity will be enforced, that is their own affair; but it hardly follows that they thereby have a right to enforce others not party to that agreement to contribute as well. That they appeal to lofty ideals while the hold-outs appeal merely to their self-interest is not anything about which contractarians can complain. The Pareto principle would demand that those party to the agreement are bound, whereas those not party to the agreement are not so bound (otherwise some are made worse off). True, welfare hold-outs do not deserve the protection of a welfare state, but the hold-outs are not made worse off by that. In fact, they have likely deemed contributions to the system would cost them more than they stand to benefit, so they would in fact do worse under the welfare scheme than outside it. Perhaps they are wrong, perhaps something tragic will befall them where they will rue their opting out of the protection plan; but that is not the point. We are not appealing to interests people should have. (That is the domain of argument (E), which I summarily dismissed.) We are appealing only to the occurrent interests of the hold-outs. So long as these hold-outs occurrently decide not to contribute, contractarians cannot force their contributions. In this sense, Narveson argues, Hobbesian contractarian ethics is committed to libertarian politics. But is the Pareto principle sufficient to the task? We are social creatures, after all. Why elevate asocial preferences over social ones? The answer has to do with assuming an initial starting point: a baseline. The Pareto principle works for libertarians only so long as we start with a conception of a baseline that precludes demands on helping; but the Pareto principle cannot itself be used to justify that baseline. Pareto principles work only once we have settled on a starting place: they cannot help us decide where that starting place ought to be. Narveson will make the case for situating our starting point in the Hobbesian state of nature. To start from the state of nature is to start from a non-welfare system. Since a forced welfare system requires the constraints of even Scrooges, who prefer not to contribute, then a welfare system makes some worse off (albeit for a good cause). But since some are made worse off – despite net social gains – Pareto principles would not condone the move from the state of nature to the welfare state. All very fine, but this argument backfires. The same use of the Pareto principle from the state of nature will forbid us from moving into morality so long as there are some who believe they will benefit from amorality. One needs to relinquish one’s use of force, and so long as some feel (rightly or wrongly) they can do better with this use of force than without, then Pareto principles cannot condone the move from the state of nature. This same point may be illustrated with the prisoner’s dilemma (PD). From the starting point of mutual defection, no unilateral movement will be a Pareto improvement. Thus if we rely too heavily on the Pareto principle, we remain in the state of nature, not the libertarian state. Two objections should be dealt with at once. (i) Will the evocation of the Pareto principle rule out the use of force? Successful use of force has the habit of leaving others worse off than they were before. If the Pareto principle rules out force, then my claim above (that the Pareto principle will derail state of nature escapes) fails. In response, we ask: Is the Pareto principle a moral principle or a rational principle?

Why Contractarians are not Libertarians

119

As a moral principle, we certainly expect it to rule out the use of unilateral force; but morality has no place yet in the state of nature, so we would need a purely rational-based argument for why we would give up the use of force. That the weaklings will have an interest in the strong relinquishing the use of force does not give the strong any reason to do so, any more than the have-nots’ interest in a welfare programme gives the haves a reason to oblige to a welfare programme. And if all creatures in the state of nature are equal in their ability to use force on others, they would never need to make an agreement restricting their use of force. Their use of force could never be rational if for every action they meet an equal reaction. We can only sensibly understand the equality condition in the state of nature as a statistical claim. Admitting a statistical equality of strength is compatible with admitting there will be particular occasions where the use of force rationally dominates. (ii) Conversely, we might wonder whether the Pareto principle should be restricted to unilateral movements. If all move from defect-defect to coop-coop, that would be a Pareto improvement. The trouble is – the whole point of the prisoner’s dilemma – is that the Nash equilibrium point understood from the vantage of rational self-interest is defect-defect. Despite the benefits of mutual movement, unilateral movement can never be rational, and since mutual movement requires a commitment concerning unilateral movement, it is not rational to expect mutual movement from defect-defect. Are libertarians to make the case that the Pareto principle trumps individual rationality? If so, they cannot justify the Pareto principle on individual rationality. If not, relying too stringently on Pareto principles will thwart one’s escape from the state of nature. The Pareto principle is too strong for contractarians.

Ultimatum Games The case against believing libertarianism necessarily arises from contractarianism may be made clearer by examining the dynamics of the ultimatum game (UG). In the UG, the Pareto principle recommends a non-cooperative venture. The UG, played in dyads, concerns a competition over a ten-piece pie. One player offers a division, and the other accepts or rejects the offered division. If the second rejects the division, no one gets any of the pie. The fear of getting no pie rather than some pie would make (one would think) offerers wary of seeming too greedy. Certainly an offer of ‘10 for me and 0 for you’ stands to be rejected, leaving 0 for them both. Notice how the second player’s rejection of the offer counteracts the concept of Pareto improvements, though. Paretianism would demand accept, since no one is worse off (player B gets 0 pieces of pie in both cases) and at least one player is better off (player A gets 0 if player B rejects, and 10 if B accepts). But claiming that it is rational, let alone morally obligatory, to have to accept an offer of 0 merely because the offerer wants it all for himself seems insane. That is, if you and I are engaged in the UG, and I know you will obey the Pareto principle, then I shall offer you a 10–0 split, where I get 10, and you get zilch. If you decline, I will be made worse off than had you accepted. If you accept, you are not made worse off, or

120

Liberty, Games and Contracts

better, but I am made better. Therefore, Pareto principles dictate you accept. Upholders of the Pareto principle in ultimatum games are ripe for exploitation. Matters are only slightly altered when we think of a 9–1 offer. It is not irrational for the hungry to reject an offer of one piece of pie. Therefore it would be rational to offer ‘9 for me, 1 for you’, predicting acceptance. Pareto becomes the saint of extortionists.10 Returning to the difficulty of baselines, if the starting point in our UG is a 9–1 offer, rejecting that would be Pareto inferior. Moreover, any movement from a 9–1 offer (say an 8–2 offer) would also be Pareto inferior, since one agent will be made worse off. But if we start from an 8–2 offer, movement to a 9–1 offer is also Pareto inferior, since the acceptor gets 1 piece of pie, rather than 2. In fact, in the UG, whichever initial offer is made is ipso facto Pareto optimal, for no movement can be maintained from it without making one player worse off. But then there are as many Pareto optimal points as there are possible moves in the game, rendering the Pareto principle useless. As noted, the Pareto principle depends very much on one’s starting position. It is worth noting that the UG introduces an odd starting point. A contaminant exists in the fact that the pie is pre-owned. As Nozick notes, we cannot generalize from the ‘fair’ process in this situation to fair processes in free market transactions.11 But the ‘pie’ need not be conceived as being pre-owned. Any negotiators are bargaining over a division of goods, and it is partly in determination of how much of the pie comes to one that determines whether the agreement will be made or rejected. In other words, we can view the UG as concerning the distribution of surplus goods deriving from negotiations for mutual benefit.12 The only ‘pre-owned’ right a Demand-9er needs is his willingness to accept nothing otherwise, and such cockiness is secure so long as the Acceptor endorses the Pareto principle.

Evolutionary Forces My conclusion so far will be uninteresting to Narveson if we believe that divisions of 9–1 splits are both rational and moral. And to assume ex machina that a 9–1 split is immoral is to beg the question against libertarianism, not show it to be false. Fair enough, but would contractarian analyses support the 9–1 split over the 5–5 split in the UG? That is the issue here. To answer negatively, I must shift my ground a little. Evolutionary accounts bypass the demand that the moral condition is in some sense rational. We can describe it as an aberration. Evolutionary success is dependent on not selecting the rational choice. Fortunately for us, as Hume would say, rationality is not what moves us. Showing what people would agree to if they were individually rational is not to the point if morality is not rational in this individualistic sense. So long as we can conceive contractarianism as arising from a more evolutionary model, we need not adopt libertarian practices even if the experts unequivocally proved the teleological benefits of libertarianism. Thus to trace the movement from self-interested amoralists to welfare advocates,

Why Contractarians are not Libertarians

121

we need merely show which groups dominate in evolution games. In the UG game, Skyrms has demonstrated how agents prone to ‘fair’ distributions do not end up too badly off so long as the initial mix favours them.13 Let us limit the range of firstmove offers to two: a 9–1 split and a 5–5 split. Second moves entail accepting the offer or rejecting it. One may accept everything indiscriminately, or reject everything indiscriminately. One may also prefer to accept 5 but reject 9, or one may do the reverse, for whatever good it will do. With two possible first moves (Demand 9 or Demand 5) and four possible second moves (Accept all, Reject all, Reject 9 but Accept 5, Accept 9 but Reject 5), we are left with eight strategies: (1) (2) (3) (4) (5) (6) (7) (8)

Demand 9; Accept All. Demand 9; Reject All. Demand 9; Accept 5, Reject 9. Demand 9; Accept 9, Reject 5. Demand 5; Accept All. Demand 5; Reject All. Demand 5; Accept 5, Reject 9. Demand 5; Accept 9, Reject 5.

‘Accept 9’ stands for accepting the demand that the other gets 9 while you get 1. That is, ‘Accept 9’ really means accept 1 lowly piece of pie. So long as we start with a 30 per cent population majority of option 7 (demand 5, accept 5, reject 9 or ‘Fairmen’, as Skyrms calls them), we are left with the completely obsequious option 5 (offer 5, accept all, 36 per cent) and Fairmen (64 per cent).14 ‘Unfair’ agents, those who offer 9, die off. Does this count against grounding libertarianism on contractarianism? Two points need to be maintained before we can make such a generalization. 1. The success of ‘Fairman’ is not unfairly rigged. 2. The success of ‘Fairman’ supports a non-libertarian structure. Libertarians may reject both. 1. The above results assume that non-libertarian policies can have evolutionary stability, but this presupposes an initial starting point of a disproportionately high number of ‘Fairmen’. Starting with an equal distribution of all types yields results less favourable to welfare (strategy 1 – whom Skyrms called Gamesman – 87 per cent) and strategy 4 (whom Skyrms called MadDogs, 13 per cent). Why should we anticipate favourable (to welfare) initial mixes? Perhaps ‘fluke’ is as sufficient an answer as we need. After all, that is an acceptable answer in general talk of evolution. For those who find the ‘fluke’ answer too glib, another kind of reply is available. The above games rely on regulated encounters: that is, each individual player will interact with every other individual player in the competition exactly once per round. In nature, no such regulated interaction exists. Rather, interactions are, to use Aumann’s language, ‘correlated’.15 That is, individuals interact with only a select number of other individuals. It will not matter in terms of end results whether the selection criterion is in terms of correlating with like-strategies or simple geographic convenience, or both. Such differences will affect the time it will take to reach an equilibrium. As with the pure fluke reply, though, strategies which can prosper among their own kind will do better in

122

Liberty, Games and Contracts

terms of reproductive advantage than strategies which can not. So long as interactions in nature are more correlated than random, as seems to be the case, evolutionary models of prisoner’s dilemmas and ultimatum games favour moral strategies, not immoral strategies – even when the immoral strategies outnumber the moral strategies in the initial population proportions.16 So long as strategies appear through drift or mutation or recombination, and interactions are correlated, moral strategies will dominate. 2. Above I, perhaps unfairly, defined ‘Fairman’ as the moral agent, with the tacit inference to be drawn that ‘Gamesman’ is the immoral agent, coupled with the further sleight-of-hand inference that Fairmen are welfare endorsers, and Gamesmen libertarians. Libertarians may resist such glib associations. That Fairmen beat Gamesmen (under certain conditions) does not mean libertarianism is ruled out until we can show that libertarians would not recommend a 5–5 split of surplus goods. If it is my pie to offer, why is it immoral to decline to offer you any? Why must I offer you 50 per cent of my property? And even if – as stated above – we view the UG as a battle over surplus goods, we still have to clearly distinguish whether we are talking about voluntary charity versus involuntary charity. Libertarians are not against charity; they are only against coercion. As put, a 5–5 split is not ruled out by libertarians. And nor is a 9–1 split necessarily immoral if that is the uncoerced result of mutual bargaining. This is all fine, but it skips over an important distinction. There are two ways we can not be against something. We can not be against something’s happening in the sense of being indifferent to its happening, and we can not be against something’s happening in the sense that we applaud its happening. An example of the indifference sense is when I am not against your eating mushrooms. I am indifferent whether you eat mushrooms or not. Go ahead or don’t: it doesn’t matter to me. (This is easier for me since I do not like mushrooms, so I am not bothered by the possibility of competition over a limited resource.) An example of the applause sense of not being against something’s happening is when I actually wish for the thing in question to happen. For example, a student is not against a teacher’s giving her an A in the test. But saying the student is ‘not against getting an A’ is not to say the student is indifferent. Nor does the student’s not being against getting an A entail that she thinks it is a duty of the teacher to give her an A (being merely a C student). The student cannot impel the teacher to give her an A, but, all things equal, she thinks it good that she gets an A. Whereas I do not think it is good that you eat mushrooms; nor do I think it bad. I am simply indifferent to your mushroom consumption. The question is, which sort of ‘… not against …’ do libertarians mean when they say ‘I am not against charity’? When Narveson reminds us that he is not against charity, he often adds that it is in fact good to be charitable, and that we ought to aspire to be charitable.17 That is, when libertarians say they are not against charity, they mean it in the applause sense, not in the indifference sense. But here is the catch. That unenforced charity is good and that we ought to be so requires a non-question-begging justification in contractarian terms. How is this to go? Even if charity is understood as good in non-categorical (hypothetical) terms, we still want to say it is morally good to be charitable. We praise charitable acts and

Why Contractarians are not Libertarians

123

boo selfish acts even while we accept that charity cannot be enforced. (Would it even be charity, then?)18 But so far, libertarians cannot make this distinction. The only distinction libertarians can make concerns forced and free actions. It leaves undiscussed the vast sorts of actions we culturally encourage and discourage without forbidding or mandating. The evolution of social forces can accommodate this sense of saying ‘Charity: good!’ without being committed to politically enforced charity. I shall first loosely describe the cultural evolution of the heuristic, ‘Charity: good!’, and then explain how this heuristic will be overextended into nonlibertarian social policies.

Overextended Heuristics The process of cultural selection is through learning and assimilation. Learning and assimilation are the politically correct terms for the more basic drives of mimicking and habit formation. Mimicking is a built-in strategy that tends to benefit mimickers over non-mimickers.19 Mimicry comes in three kinds: (a) proximity, (b) bandwagon and (c) prestige. The three forms are interrelated. Proximity determines the group with whom one identifies. Bandwagonning determines who counts as successful within that group: the ones the group approves. The prestigious are identified with success.20 The concept of success will vary from culture to culture and group to group, but it will tend to be selected by bandwagonning. The best individual strategy for success, therefore, is to mimic the prestigious members of one’s group. One’s dealing with others successfully has a lot to do with one’s reputation concerning how one has dealt with others in the past. If we understand fair dealing with others as moral dealing, and reputation has to do with fair dealing, then the greater need for preserving one’s reputation will be positively correlated with moral behaviour. This echoes (or explains) Gauthier’s rationale for making oneself as transparent as possible.21 On the evolutionary account, moral action is a payment for preserving your reputation. If this were the impetus of moral behaviour, those more in the limelight would have a greater need for being recognized as moral players. Since those identified as the prestigious (the successful) are certainly more in the limelight, the prestigious will have more need for moral action than the plebeian. If the plebeian knew the mechanics of morality, they would be less moral; but, fortunately, evolutionary forces mandate that most of them follow the heuristic of mimicking the most successful. Since the most successful tend to be moral (due to the limelight and the need to preserve their reputation), the plebeians tend to copy that moral behaviour as well. This may be a poor move rationally speaking, but it does benefit the group, and benefits to the group rebound on those individuals prone to emulate others – as most are. They are moral for the wrong reason, but since moral dealing has evolutionary fit, the folly will spread. Not all beneficial social behaviours that evolve, do so for those beneficial reasons. On the wrong belief that malaria was caused by bad air (mal aria), a

124

Liberty, Games and Contracts

preventive cure was to shut the windows. This had the fortunate side effect of keeping the infected mosquitoes out, and so the successful (but wrong-headed) convention of shutting windows passed through the generations.22 Mimicking the prestigious may likewise have the unintended happy consequence of driving (even individually irrational) moral norms to fixation. Saying this does not mean that moral norms are false. What is false in the malaria case is that malaria is spread by bad air. What is false in the moral case is that we have a categorical duty to help those in need, and that purely rational agents from an ex ante position would endorse a welfare programme. What is true in the malaria case is that shutting the windows is a good strategy to prevent malaria. What is true in the moral case is that behaving in ways that signal to others you are worth cooperating with has evolutionary fitness, and that helping others when they are in need is a good way of signalling to others that you are a good person with whom to cooperate. None of this need work on the level of rational motivation. I can therefore agree with Narveson in saying that rational motivation alone cannot lead one to non-libertarian policies. Imitation and social punishment are cultural ways of stabilizing norms. But why these norms? Any norms can be stabilized if imitated while deviants are punished. But imitation need not be random. If imitation is predicated on success, then successful strategies will tend to rise to the norm. Research by Boyd and Richerson, for example, makes exactly that case. Imitation is a coarse-grained operative. When imitating the successful, we are not quite clear on what features we should imitate. It is unlikely that we will stumble upon exactly the right features at exactly the right degree. Therefore we can expect to err. The question is, do we err by not copying enough, or by copying too much? Boyd and Richerson found that redundant imitators outdid insufficient imitators in terms of success.23 In fact, there will be a tendency toward imitation in greater degrees. Individuals attempt to discover the best behaviours in varying environments. They may learn on their own, or they may imitate those of previous generations. Given the vagaries of such experimentation (varying conditions and third variables), strategies that are effective in the short term are not necessarily strategies that succeed in the long haul. A risk of selfexperimentation is latching onto behaviours that have no lasting effect. Imitating successful members of previous generations, therefore, has the benefit of avoiding costly errors. The risk of imitation is greater when environments have shifted. Of course, there is a wide range between full imitation and zero imitation. So long as environments change, a need for self-experimentation will always be useful. Cases where self-experimenters succeed, however, become cases for imitation of subsequent generations. Therefore, for each successive generation, a stronger case can be made for the imitation strategy. Invaders can vary in terms of their imitation strategies, some greater than average, some less. Assuming relatively stable environments, the greater imitators will achieve a higher payoff than both the average and the lesser imitators. As Boyd and Richerson observe, ‘when it is difficult for an individual to determine the best behaviour and when environments change infrequently, more than 90% of a population at equilibrium simply copies the behaviour of others.’24

Why Contractarians are not Libertarians

125

Assuming heuristics are not perfectly sensitive to all environmental cues, we can expect some of the adaptive heuristics may be overextended in certain cases. As long as such errors have no maladaptive effects, we can predict the imitated norm to include the overextension across future generations. This would explain why norms of altruism become entrenched so long as the prestige members are altruistic in the first place, or behave in ways that may appear altruistic to potential imitators. But why would they be altruistic? Why would successful progenitors have altruistic tendencies? So far, all the imitation-heuristic argument has suggested is that if nonlibertarian strategies are outcomes of successful strategies, the drive toward mimicry and conformity would preserve non-libertarian structures as a norm. But why would a non-libertarian norm have evolved in the first place? An added feature of the overextended heuristic account may be offered. A’s failing to be altruistic may signal to a conditional cooperator B that A may not be a conditional cooperator (CC) after all.25 This may be false, since a CC need not commit to every proposal to come his way. When speaking of detection errors that CC faces, normally we limit the discussion to only those mistakes where a CC incorrectly takes an unconditional defector (UD) to be a CC (or unconditional cooperator, UC). But mistakes can work the opposite way. A CC may mistake another CC for a UD. A CC’s conditional response means he must be sensitive to some cues, however broad-brushed they may be. The failure of demonstrating inclinations to fair distributions of surplus goods, then, may be enough of a signal to potential cooperators to read the libertarian as a non-CC. If this norm were itself fostered through imitation and social censure, then any CC will be motivated to demonstrate her non-libertarian leanings within the realm of the fuzzy norm. In terms of the ultimatum game, CC will be moved to be a Fairman because the ‘fair’ distribution acts as a signal to others of her being a CC in a PD. Notice this line of argument carries more weight for those whose reputations are more prominent. Prestige members will be more motivated to guard their cooperative reputation – even by exerting behaviours technically unnecessary for conditional cooperation – if merely to thwart misconstrual, and thus defection, by others.26 The fear of a marred reputation thereby moves the prestigious toward overextension of cooperative behaviours into fuzzy cases of welfarism, and the broad-stroked mechanics of mimicry move the plebeians to imitate the overextended altruism which in turn coagulates into a norm. By norm, here, I mean there will be a clear case where society will not be indifferent to Scrooge’s indifference to Tiny Tim’s suffering; a norm of not endorsing any policy that would carve a space for Scroogian indifference. This evolutionary account of adopting social welfare policies will violate Pareto principles given small niches for Scrooges, but – as Hume might say – nature does not pay attention to Saint Pareto. My evolutionary account will vindicate neither egalitarians nor libertarians by any means. Some may complain mimicry and overextended heuristics is entirely too feeble an account of morality. My point here is far less lofty. I offer an account of our moral attitudes toward the social acceptance of non-libertarian policies from the vantage of an evolutionary-backed contractarian model. What this shows is that I can agree with Jan Narveson that stringent rationality cannot get us beyond

Liberty, Games and Contracts

126

libertarianism without thereby being convinced that contractarian beginnings commit us to libertarianism. Nature moves us in irrational ways.

Conclusion If we look to how we are, we find that people – rightly or wrongly – tend to support some degree of Pareto-infringing centralized distribution. Rather than complaining that it is in fact irrational, we may first want to look to see how such a phenomenon would come about. That is, rather than see what contractarian principles should endorse, we look to see how a non-libertarian structure may evolve and see if, in order to do so, it necessarily violates any contractarian credo. We need to tease out an interpretation of the ultimatum game that can explain how non-libertarian systems can evolve without violating any contractarian credo. To do this, we pay closer attention to the different ways one may be ‘not against’ something. Since an overextended heuristic can coagulate into a non-libertarian social institution from contractarian beginnings, the claim that libertarianism is founded on contractarianism is amiss. Even if libertarians are right that a centralized distribution is inefficient, and thereby irrational to endorse, we may not be moved by rationality alone. And so long as contractarians emphasize agreement (in terms of coagulated social norms) as opposed to rational choice, libertarianism cannot be grounded on contractarianism.

Notes 1 2 3

4 5 6 7

8

Jan Narveson, The Libertarian Idea (Philadelphia, 1988), 131, 154, 165–6, 175–84. David Gauthier, ‘Why Contractarianism?’, in Peter Vallentyne (ed.), Contractarianism and Rational Choice (Cambridge, 1991), 25. See, for example, Richard Brandt, A Theory of the Good and the Right (Amherst, NY, 1998), 110–15; and David Griffin, Well Being: Its Meaning, Measurement and Moral Importance (Oxford, 1986), 10–17, for similar discussions. Otherwise use all the ‘helps, and advantages of Warre’. Thomas Hobbes, Leviathan (Buffalo, NY, [1651] 1988), chs 14, 67. There are exceptions: for example, the Dominant Protection Agency in Robert Nozick, Anarchy, State, and Utopia (New York, 1974), 15. For a discussion on property, see Ann Levey, ‘Liberty, Property and the Libertarian Idea’, and Peter Vallentyne, ‘On Original Appropriation’, both in this volume. ‘[T]he market society maximizes utility’ (Narveson, 189). Conversely, ‘by taxing so highly the most productive members of society, they are decreasing the wealthy man’s incentive to continue to be productive’ (John Hospers, Libertarianism (Los Angeles, 1971), 208). I think the best way to read Jan Narveson on the distinction between (E) ‘efficiency arguments’ and (P) ‘Paretian arguments’ is that although he believes (E) is true, he also recognizes that it is irrelevant. Most of his discussion with efficiency arguments is to counter-distinguish them from utilitarian arguments (for example, Narveson, 183).

Why Contractarians are not Libertarians 9 10 11

12

13 14 15

16 17

18

19

20 21

22

23 24 25 26

127

Narveson, 183. Cf Narveson’s heading ‘The Gospel According to St. Pareto’ (Narveson, 184). ‘[W]e are not in the position of children who have been given portions of pie by someone who now makes last minute adjustments to rectify careless cutting’ (Nozick, 149). The appropriate distribution of this surplus according to Gauthier is the minimax relative concession. David Gauthier, Morals By Agreement (Oxford, 1986), 145. This is the point where Narveson disagrees with Gauthier. Brian Skyrms, Evolution of the Social Contract (Cambridge, 1996). Skyrms, 32. Robert Aumann, ‘Subjectivity and Correlation in Randomized Strategies’, Journal of Mathematical Economics, 1 (1974): 67–96. Robert Aumann, ‘Correlated Equilibrium as an Expression of Bayesian Rationality’, Econometrica, 55 (1981):1–18. Skyrms, 63–79. Brian Skyrms and Jason Alexander, ‘Bargaining with Neighbors: Is Justice Contagious?’, Journal of Philosophy, 96/11 (1999): 588–98. ‘One of the good things we can do in life is to make an effort to care about people whom we don’t ordinarily care or think about’ (Jan Narveson, Moral Matters 2nd edn (Peterborough, Ontario, 1999), 152). ‘The tendency and desire to do good for others is a virtue. Moreover it is a moral virtue, for we all have an interest in the general acquisition of this trait’ (Narveson, Moral Matters, 152). ‘Charity, in my view, is an important and useful category in morals’ (Narveson, Libertarian Idea, 257). I might also add it is well known that Jan himself is exceedingly charitable. ‘[M]oral conduct must be undertaken as a matter of conscience and free choice; otherwise the act loses its moral worth. A society that forces its citizens, under threat of punishment, to help the less fortunate is less, not more, compassionate.’ Tibor Machan, ‘Should Business be Regulated?’, in Tom Regan (ed.), Just Business (New York, 1984), 217. Luca Luigi Cavalli-Sforza and Marcus Feldman, Cultural Transmission and Evolution (Princeton, NJ, 1981). Dan Sperba, Explaining Culture: A Naturalistic Approach (Oxford, 1996). Robert Boyd and Peter Richerson, Culture and the Evolutionary Process (Chicago, 1985). ‘The more translucent constrained maximizers are, the better they are at achieving cooperation among themselves … and avoiding exploitation by straightforward maximizers’ (Gauthier, Morals By Agreement, 178). Joe Heinrich, Wulf Albers, Robert Boyd, Gerd Gigerenzer, Kevin McCabe, Axel Ockenfels, and Peyton Young, ‘What is the Role of Culture in Bounded Rationality?’, in Gerd Gigerenzer and R. Selten (eds), Bounded Rationality: The Adaptive Toolbox (Cambridge, MA, 2002), 355. Robert Boyd and Peter J. Richerson, ‘Norms and Bounded Rationality’, in Gigerenzer and Selten (eds), 281–96. Boyd and Richerson, ‘Norms and Bounded Rationality’, 286. For the complete discussion of CCs, UDs, UCs and even RCs, see Peter Danielson, Artificial Morality: Virtuous Robots for Virtual Games (London, 1992). See also Frans de Waal, Good Natured: The Origins of Right and Wrong in Humans and Other Animals (Cambridge, MA, 1996), 115.

This page intentionally left blank

Chapter 9

Getting the Baseline Right Paul Viminitz

Introduction Here, I take it, is our common project: We are, quite literally, civil servants. We work for the polity that pays our salaries. Our job is helping our clients get their moral and political reasoning right. But there is no fear of featherbedding here. There is a significant divergence between what moral and political arrangements our clients currently enjoy and those that would be sanctioned by more careful reasoning. Some of us – Jan Narveson for example – think this divergence is a veritable disconnect.1 And so he advocates a radical revision to those arrangements. Others think the status quo requires only fine tuning. But if any of us thought the way things are is exactly the way they should be – and can be pretty much counted on to remain so2 – we would be hard-pressed to say what work our moral and political philosophizing is doing.3 Prescriptivity, Descriptivity and Justification But if we are in the business of saying how things should be, of what relevance is the way things are? Twofold: first, things can only be the best they can be under a given set of circumstances. For example, according to Hobbes it is that ‘two men desire the same thing, which neverthelesse they cannot both enjoy’4 that makes morality and politics necessary. But, he adds, what makes morality and politics possible is that many of the commodities over which we compete, though scarce, can be shared.5 In other words, we are facing only moderate scarcity. Game theoreticians capture this by saying we are faced with (what they call) mixed-motive games. Or, in yet other game-theoretic terms, by the moderate scarcity condition is meant the availability of a cooperative dividend. But though morality and politics, coupled with science, should, properly deployed, increase our productivity, we are never going to produce such plenty that we are going to completely do away with scarcity, and so with the need for morality and politics.6 Neither are we going to completely do away with competition, and hence the need for morality and politics, by learning to content ourselves with less.7 We can reduce competition, from which, says Hobbes, ‘proceedeth diffidence … thence Warre’, thence the need for morality and politics.8 After all, a man who 129

130

Liberty, Games and Contracts

wants nothing wants for nothing. But Hobbes knows of no such man. And neither do I.9 Can we imagine a circumstance under which humans are not faced with mixedmotive games? Certainly. Total war and the Hereafter, to name but two.10 But Leviathan is not addressed to total warriors or angels. It is addressed to us. This is why Chapters I to XIII offer a description of the human condition as Hobbes finds it. Only in Chapter XIV does he then turn to how to make the best of our bad situation. Are there, as Virginia Held has argued, human communities and/or times to which Hobbes’s description does not apply?11 Certainly. Hobbes himself acknowledges this when he refers to ‘the government of small Families, the concord whereof dependeth on naturall lust’.12 By this Hobbes does not mean that families do not face scarcity. Rather he means that in a family – indeed what we mean by family – is that, though ‘two [members may] desire the same thing which neverthelesse they cannot both enjoy’,13 neither has a desire to consume these scarce resources at the expense of the other.14 What is in dispute, then, between hardcore contractarians and their detractors, is whether governance driven not by mixed-motivity but by bonds of love is part of morality. Our detractors say the promotion and/or maintenance of such love is a matter of morality. Indeed some of them argue that love is the genesis of morality. Contractarians, by contrast, say love – or, for that matter, boundless hate – is a morally-free zone.15 But we also say that nothing substantive can hang on this dispute, for to say that morality arises only out of mixed-motivity is not to say that interactions that are not mixed-motivated do not warrant the attention of philosophers. It is to say only that they do not warrant our attention qua moral philosophers. The two camps are in some measure synthesized by Hume, according to whom moral dispositions, however motivated, cannot be activated without the sentiment of fellow-feeling. Hume may well be right. But contractarians give away nothing in conceding this. We simply say that (a) the adoption of moral dispositions required to maximize under conditions of mixed-motivity, and (b), the extension of familial fellow-feeling to non-family members may be phenomenologically indistinguishable. But if the phenomenology of the moral point of view does not distinguish between the Hobbesian account and the Humean – and since the etiology of morality is shrouded in our evolutionary history – why should we prefer the Hobbesian account to the Humean? Worse yet, suppose there were good reasons to believe – and evolutionary ethicists have shown there are such reasons – that morality owes its perdurance in us not to individual selection but to group selection.16 Suppose, in other words, that morality is not, as Hobbes would have it, a strategy by which one maximizes his own fitness under conditions of mixedmotivity, but rather a strategy for maximizing group fitness. Should we not then prefer the Humean account to the Hobbesian? To which we answer that, phenomenological underdetermination notwithstanding – for that matter historicity notwithstanding – only the Hobbesian can

Getting the Baseline Right

131

speak to the Foole.17 That is, only the Hobbesian can convince the psychopath why, notwithstanding his being bereft of any morality, he should, if he can, find himself some morality with all possible dispatch. For the central question for contractarians is not how came we to have the moral dispositions we have – and/or the political arrangements we currently enjoy – but rather what dispositions and/or arrangements ought we to adopt now. To adopt whatever dispositions and/or arrangements that account for our being here to ask the question would be to commit the naturalistic fallacy. Yes, I probably do owe my very being to group selection processes. But what is that to me? So contractarianism – and only contractarianism – offers its clients a justification for morality; that is, a justification for morality from non-moral premises, and a justification pegged to our clients’ occurrent desire-sets. Moreover, contractarianism can tell our clients which moral dispositions (and political arrangements) to adopt (or promote), given those occurrent desire-sets. Bounded Rationality The second way normative philosophy involves attention to the way things are is that one cannot get to where one should be without knowing where one is. For example, human rationality is bounded rationality. So sometimes it is not enough just to say to our clients, ‘Look, here’s a better way!’ For looking is not necessarily seeing. Sometimes our clients’ bounded rationality makes it impossible for them to see. But, as Plato pointed out, that does not relieve us of our professional responsibilities. For example, I have argued in defence of terrorism.18 It was my professional responsibility to do so. But that does not mean I should press that argument at a memorial service for the victims of 9/11. That said, just as the Pope takes pains to make clear when he is speaking ex cathedra, we too must make clear when we are speaking to the hoi polloi in the cave and when we are speaking to each other outside of it. And indeed, in this paper I am speaking not to our clients but to my fellow philosophers. So, here at least, no mincer of words need I be. Another example of how bounded rationality informs our advice is that getting from here to there may have to be done incrementally. For example, we could save billions in the cost of air travel by doing away with the practice of farewelling, or at least moving it from plane-side to the front door of the house. But, though it would be optimal to make the shift in one fell swoop, people are too wedded to the practice for the move not to be done in increments. In fact not only would nothing be gained by suggesting such a radical revision to the practice, such a suggestion might even provoke a return to plane-side farewelling as a protest. Weather concerns allowed us to move it from plane-side to the gate. Then security concerns moved it another hundred yards back. The cost of parking will gradually drive it from the airport to a mall near the airport. And, finally, highway congestion and the cost of fuel will drive it to the local mall. So we philosopher-kings will get our way. We just have to be patient.

132

Liberty, Games and Contracts

Error Theories and their Probative Force One can get from where-she-ought-not-to-be-but-is to where she should be without knowing why she is where-she-ought-not-to-be-but-is; that is, what mistake she might have made in virtue of which she is where-she-ought-not-to-be-but-is. So (what philosophers call) an error theory is not a necessary component of a complete political theory. Nevertheless it is a highly useful check. That is, that I am the only person on the planet to think that p is no reason to think I am wrong. But it would be a reason to at least double-check my reasoning if I could not come up with an account of why everyone else believes, albeit falsely, that ~p. This is important. Suppose someone believes, albeit falsely, that p, and she believes this because she believes – only this time rightly – that q and that p implies q. In other words, she believes that p because she has affirmed the consequent. Affirming the consequent is a common mistake in human reasoning. But there is a perfectly adequate explanation as to why it is so common. It is so common because most conditionals, at least in the domains in which the error is so common, are also bi-conditionals; and so inferring the other atomic from the atomic given as a premise is, for those domains at least, a heuristic with a success rate adequate to compensate for its occasional failures. But this is not the case with, say, a complete non-sequitur, for example, p therefore q. So if someone believes p and concludes q, chances are it is because she believes p implies q, or some other suppressed premise. So what is the cash value of all this? Just that, all other things being equal, my political theory is better than your political theory if, by supposing people’s errors arise from an otherwise laudable heuristic, I can account for the divergence between my theory and how people actually behave; whereas to account for the divergence between your theory and how people actually behave you need to say (something like) people are just stupid. This is not to say (something like) people are just stupid couldn’t be the case. It is to say only that that people are just stupid is not as good an error theory as that they are employing an otherwise laudable heuristic. For if people were stupid in this domain, why are they not equally stupid in other domains? But if they were equally stupid in all domains they would have long since become extinct. So people-arejust-stupid is a bad error theory. Of course having a good error theory is not a knockdown proof of the theory it accompanies. But, as I say, having a bad error theory is a reason to at least double-check it. What makes one error theory better than another is no different from what makes any theory better than another, and this includes which theory has the better second-order error theory. And so on. Accordingly, what I want to do in what follows is threefold. First – though I have advanced this argument in greater detail elsewhere – I want to rehearse the case that, 1. I am right and everyone else is wrong about the appropriate baseline for doing contractarian moral and political theory. But this time – and as importantly – I want to explain, 2. what otherwise laudable heuristic has lead everyone else astray. And, finally, insofar as any account of how people are led astray always invites, 3. an account of how the person giving the account was not himself led

Getting the Baseline Right

133

astray – in other words, what makes me so special? I want to account as well for my own … how shall I put this? … anomalous insightfulness. I begin, then, with …

Baselining According to Jan Narveson – and on this score, at least, I have no cause to naysay him – talk of ownership is parasitic on talk of rights (and/or their correlatives, that is, duties). More particularly, to say I own something is just to say I have a right that you not interfere in my making use of it (or, correlatively, that you have a duty not to so interfere). Rights-talk falls into the same locution set as oughts. To say that I have a right to something is just to say you ought not to interfere with my making use of it. To say that you have a duty to not interfere in my making use of something is just to say you ought not to interfere in my making use of it. Well, fair enough. But now all the work will be done by our answer to, ‘What counts as interference?’ Suppose I grant that you own your car. But while you are shopping I park mine a millimetre from your front bumper, and my friend parks his a millimetre from your back bumper. Clearly we are preventing you from making use of your car. I should pull ahead, or my friend should pull back, to let you out. But, you might say, that’s because you were there first. My friend and I did something to prevent you from making use of your car. Fair enough. So now suppose we had done nothing of the sort. He and I each parked our cars and then yours was craned into place between us by some prankster; a place from which – save by one of us pulling ahead or back – it cannot be extricated. True, it is the prankster who is liable. But suppose he is nowhere to be found. So, are my friend and I interfering with your making use of your car or are we not? If you say yes then the distinction between positive and negative rights and duties – a distinction crucial to the libertarian case – has just collapsed. For one species of prankster who, though liable, cannot be made to pay, so to speak, is God. And one kind of prank He might play on me is cancer. And so if you do not lend me assistance in combatting my cancer you are interfering in my making use of my body. But if you say no – no, you are not violating my right to make use of my car by refusing to pull ahead or back, and so no, you are not violating my right to make use of my body by not assisting me in combatting my cancer – then any motivation one might have had for subscribing to the non-interference interpretation of a right has just evaporated. For what people need from each other is that one pull ahead or back when the other’s car is blocked. What people need from each other is assistance in combatting one’s cancer. Or crop failure. Or what have you. In fact were it not for these needs we would never have been motivated to exit the state of nature/war and enter civil society in the first place. Not so, says the libertarian. What we need from each other first and foremost is assurance that we will not kill each other. Then and only then are we free to enter into whatever further arrangements might be to our mutual advantage,

134

Liberty, Games and Contracts

arrangements like assistance in times of need. That is, assure me first that, regardless of what further arrangements I agree to, I will not die by your hand. Then we’ll talk. Why only then? Because then and only then can those further agreements be truly said to be entered into voluntarily. And why is it important that they be voluntary? Because if and only if they are voluntary does it make sense to say they are mutually advantageous and therefore rational to subscribe to.19 I entirely agree. But an agreement not to kill each other is of no use to me if I will as surely die of my cancer or crop failure. So I propose the following: if you agree to assist me in combatting my cancer or crop failure, then and only then will I agree not to try to kill you. This arrangement is clearly to our mutual advantage. And, as clearly, if entered into it would be entered into voluntarily. After all, you are entirely at liberty to reject the deal and take your chances on the battlefield. In fact even if, having bested you in combat, I have your life in my hands, you are still at liberty to choose death over acquiescence. People do so all the time.20 Put another way, imagine a world in which life is so ‘solitary, poor, nasty, brutish, and short’ that the difference in utility between life and death is insignificant, but the difference in utility is enormous between (a), life in a state of nature or death and (b), a compact for mutual assistance. Then the case for the lexical priority of a compact not to kill each other would evaporate. And indeed it is precisely because the difference in utility between life in a state of nature and death is insignificant but the difference in utility is enormous between (a), life in a state of nature or death and (b), a compact for mutual assistance, that Hobbes is anything but a libertarian. For, notes he, a state of nature is a state of war. And in a time of Warre, where every man is Enemy to every man … there is no place for Industry, because the fruit thereof is uncertain, and consequently no Culture of the Earth, no Navigation, nor use of the commodities that may be imported by Sea, no commodious Building, no Instruments of moving and removing such things as require much force, no Knowledge of the face of the Earth, no account of Time, no Arts, no Letters, no Society, and which is worst of all, continuall feare, and danger of violent death.21

The libertarian thinks we could be motivated to enter civil society even if all we could expect from doing so was escape from this last condition – that is, danger of violent death. But this is manifestly false. The history of the human race just is the history of our willingness to risk violent death in exchange for a chance to increase our standard of living.22 Most moral and political philosophers will grant all this but insist nonetheless that there is a perfectly serviceable distinction between ‘agreements’ entered into only at the point of a gun and (let us call them) agreements-properly-so-called. Intuitionists about this distinction are content to leave it at that, namely as intuitively obvious. Reductionists, by contrast, are prepared to put in a little effort here. For example, Narveson – along with Robert Nozick and David Gauthier – reduces the distinction between voluntary and coercive to the distinction between bettering and worsening.23 There is nothing coercive about my declining to better your situation, say they, so any agreement entered into by which you hope to better

Getting the Baseline Right

135

your situation is voluntary. By contrast, what we mean by coercion is threatening to worsen your situation. This is helpful – but not very – for two reasons. First, bettering and worsening require a baseline. And second, even if the bettering/worsening distinction did capture what we mean by the voluntary/coercive distinction, it does not tell us what, if anything, is wrong with coercion. Let us deal with the baselining problem first: The Right Baseline Some people think – and Narveson, Nozick and Gauthier seem to be among them – that a condition of our sitting down and bargaining over who gets what share of some cooperative dividend is that we each recognize the other’s entitlement to what she brings to the table. This is a mistake. It would not be a mistake if by one’s ‘entitlement’ to something is meant nothing more than that one seems to have control of it. For then all one is saying is that a condition of bargaining is that we each recognize what the other has control over. (Well, duh, since it is hard to imagine my bargaining with you for something that I do not think is within your power to deliver.) But it is a mistake to suppose we need to recognize each other’s ownership of the objects of this control. For to suppose that I need to recognize your ownership of an object is, as we saw earlier, to suppose I need to think that I ought not to interfere in your making use of it. But this makes no sense. If I cannot interfere in your making use of an object, then whether I ought or ought not to does not arise. So to say that I ought not to interfere in your making use of an object implies that I can interfere in your making use of it. But if I can interfere in your making use of it, in what sense are you the one bringing it to the table? Another way to put this is to point out that those who suppose bargaining over ownership requires a prior recognition of ownership are saddled with an infinite regress problem. But no such problem arises if we suppose bargaining requires not re-cognition of anything, but rather and only the realization of who controls what. It is not that I own my superior marksmanship that will, given plausible bargaining norms, entitle me to a lion’s share of the hunt. It is that I control my superior marksmanship. It is that you have to defer to me for its deployment. So the baseline is not any of the suggestions proffered by Narveson, Nozick, Gauthier, or any other variation on the Lockean Proviso. It is whatever would be the case in the absence of an agreement to the contrary, given what I can bring to the table. But – and this is the crucial ‘but’ – what would be the case in the absence of an agreement to the contrary is not that each of us would just return to our respective corners of the forest and carry on hunting on our own. For what I can bring to the table is not just my capacity to better your situation. It is also my capacity to worsen it. So the fact that I am a better marksman than you means that, were we to return to a state of war, I would be more likely to kill you than you would be to kill me. Gauthier has argued that since a return to a state of war would be sub-optimal, and would be known to be sub-optimal, any threat to return to a state of war would be disingenuous, and so would be regarded as such. So, says he, our military

136

Liberty, Games and Contracts

endowments can be discounted.24 But this is an egregious error in reasoning, made all the more egregious by its inconsistency with Gauthier’s own reliance on the notion of a ‘pre-commitment strategy’. For from the fact that it would be suboptimal, and so irrational, to return to a state of war, it does not follow that it would be irrational to pre-commit oneself to return to a state of war if one does not get one’s way in the bargaining process. So the threat to return to a state of war need not be disingenuous. And so the worsening effects of war are of a piece with the bettering effects of cooperation.25 Or, put another way, suppose separately you and I can make 4 and 2 widgets respectively, but together we can make 10. The cooperative dividend of 4 can be divided to yield take-homes of 7:3, 6:4 or 5:5. But which will it be? Some people think distribution should be a function of contribution. (This sometimes masquerades as desert.) There is, of course, considerable debate about how to measure contribution, aka desert. Some think that (1) our counterfactual productivities suggest you have contributed more than me. So to the baseline of 4 and 2 we should add 3 and 1 respectively, yielding 7:3. Others think that (2) it is obviously the cooperation that is doing all the work here, and so since we are equally cooperating we should equally benefit from the cooperative dividend. So to the baseline of 4 and 2 we should add 2 to each, yielding 6:4. Still others think that (3) 4 and 2 serve as a baseline only for whether there is anything to be gained by cooperating, and so whether it is rational to cooperate at all. But if it is rational to cooperate, then the original baseline ceases to be relevant. Since we have equally cooperated we should each equally benefit from the cooperative enterprise – that is, we should each get 5. Even ignoring whether contribution connects so easily to desert, we need not decide between (1), (2) or (3) because they are each being unfairly selective about what counts as the baseline. Suppose, as before, separately you and I can make 4 and 2 widgets respectively, but together we can make 10. But by this ‘separately’ I mean if left to our own devices. If not left to our own devices, you can produce 0 and I can produce 1. How so? Because if not left to your own devices you cannot make any widgets at all. Why? Because you would be dead. Whereas if I am not left to my own devices I will not be dead – because, suppose, I am a better warrior than you; though now I can make only 1 widget, having expended the other in killing you. So why is 0 and 1, rather than 4 and 2, not the right baseline from which to begin negotiating? In fact I claim that 0 and 1 is the right baseline from which to begin negotiating, because 0 and 1 is what would be the case in the absence of an agreement to the contrary, given what we can each bring to the table. The Ethics of Worsening It would straw man my position to reduce it to ‘To each according to his threat potential!’ Just as nations can adopt a scorched earth policy to slow the enemy’s advance, likewise they can adopt a scorched earth policy to demotivate the invasion in the first place. But that means they are also bringing their productive capacity to

Getting the Baseline Right

137

the table. And the same is true of interpersonal negotiations. No one – save perhaps Rawls, insofar as productive capacity is a natural endowment – thinks that productive capacity should be discounted in interpersonal negotiations. Why, then – apart from the disingenuousness argument already refuted – would anyone think military capacity should be discounted? One could think this, I submit, only by importing a moral intuition – that is, the intuition that it is wrong to profit by threat of violence. But if we allow a moral intuition to inform our theory we have forfeited our claim to full reduction. If there is such a moral intuition – and I do not deny there is – then it is to be explained and, if appropriate, justified, by non-moral premises. In other words, we have to demonstrate how the discounting of my capacity to worsen your situation can arise from the counting of my capacity to worsen your situation. I turn to that demonstration now. In doing so I will be hitting two birds with one stone. For my account of how the discounting of my capacity to worsen your situation arises from the counting of my capacity to worsen your situation will itself arise from my account of …

Why Everyone Else is Wrong about Baselining The Shallow Analysis There are a number of ways to characterize the mistake Narveson, Nozick and Gauthier are making in supposing the baseline for bargaining is some species of the Lockean Proviso. It could be characterized, as we have already seen, as an illegitimate importation of a moral intuition. It could be characterized, as we have already seen, as a failure to appreciate the rationality of irrationality, that is, of employing a pre-commitment strategy. But I think the most instructive way to characterize it is as a failure to press one’s own reductionist programme to completion. To be fair, in this timidity Narveson, Nozick and Gauthier are in good company, for Hobbes himself makes the same mistake. In Chapter XIII (66) of Leviathan Hobbes takes pains to make clear that To this warre of every man against every man, this also is consequent; that nothing can be Unjust. The notions of Right and Wrong, Justice and Injustice have there no place. Where there is no common Power, there is no Law: where no Law, no Injustice. Force and Fraud, are in warre the two Cardinall vertues … It is consequent also to the same condition, that there be no Propriety, no Dominion, no Mine and Thine distinct; but onely that to be every mans that he can get; and for so long, as he can keep it.

And yet in Chapter XX (105) Hobbes declares that in a condition of meer Nature … the right of Dominion over the Child dependeth on [the] will [of the Mother], and is consequently hers. [For] seeing the Infant is first in the power of the Mother, so as she may either nourish, or expose it, if she nourish it, it oweth its life

138

Liberty, Games and Contracts

to the Mother; and is therefore obliged to obey her, rather than any other; and by consequence the Dominion over it is hers.

Hobbes’s use of ‘oweth’ and ‘obliged’ could be regarded as a harmless slip of the tongue if, as with my comments about ‘entitlement’ earlier, all he means is that a child realizes what resources its mother controls. But if that is all Hobbes means by ‘oweth’ and ‘obliged’ then all he would be saying is that, since (1) the child realizes what resources its mother controls, (2) the child realizes what resources its mother controls. So for his claim to be synthetic rather than merely tautological, he must mean something else by ‘oweth’ and ‘obliged’. And indeed he does. He means that the child is now under a moral obligation to obey its mother; that is, to obey her even after it ceases to be dependent on her. He means that the child owes her this obedience in repayment for the benefit she has bestowed upon it. And yet this is impossible on Hobbes’s own account. For in Chapter XIV (70–71) he makes clear that If a Covenant be made, wherein neither of the parties performe presently, but trust one another; in a condition of meer Nature … upon any reasonable suspition, it is Voyd … For he that performeth first, has no assurance the other will performe after; because the bonds of words are too weak to bridge mens ambition, avarice, anger, and other Passions, without the feare of some coerceive Power … And therefore he which performeth first, does but betray himselfe to his enemy; contrary to the Right (he can never abandon) of defending his life, and means of living.

Moreover, this contradiction between Chapter XX and Chapters XIII and XIV is repeated within Chapter XIV (72). For there he says that Covenants entered into by fear, in the condition of meer Nature, are obligatory. For example, if I Covenant to pay a ransome, or service for my life, to an enemy; I am bound by it. For it is a Contract, wherein one receiveth the benefit of life; the other is to receive mony, or service for it; and consequently, where no other Law … forbiddeth the performance, the Covenant is valid. Therefore Prisoners of warre, if trusted with the payment of their Ransome, are obliged to pay it.

Here there can be no suspicion that by being bound by such a covenant all Hobbes means is being well-advised to comply with it. For under the circumstances described one would be ill-advised to pay such a ransom. In fact one would be downright crazy to pay it! And yet Hobbes thinks one should pay it. One should pay it because, as with the child example, one received a benefit. But this contradicts Hobbes’s earlier denial that there can be moral obligations in a state of nature: indeed, that moral utterances can have any meaning in a state of nature. How could Hobbes have made such an obvious mistake? There are only two possibilities. The first is that he is not making a mistake. We have been making a mistake in supposing Hobbes is, well, a Hobbesian, whereas in fact he is a natural law theorist. On this reading Hobbes thinks that one is obligated to ‘performe [one’s] Covenants made’ (ch. XV, 74), even in a state of nature, but that on, and

Getting the Baseline Right

139

only on, ‘reasonable suspition’ that the other will not perform his side of the bargain, one is relieved of that obligation (ch. XIV, 70). And that this ‘reasonable suspition’ is satisfied by there being ‘no assurance the other will performe after’ (ch. XIV, 70–71). But in both the ransom and child cases that assurance is given by the other having performed his (or her) side of the bargain already. But if Hobbes thinks one can have obligations in a state of nature, he must be a natural law theorist, and not, as we have always thought, the father of legal positivism. But hold on. On this view of Hobbes’s view, the institution of sovereignty is required not to create law but to enforce it. But that is just Locke’s view. And so the contrast between Hobbes and Locke – a contrast that virtually defines political philosophy – collapses. But in that case, rather than allow the contrast virtually definitive of political philosophy to collapse, would it not be better to simply consign Hobbes to the flames on grounds of redundancy and then make up a new person named Hobbes* to be the torch-bearer for legal positivism? The second possibility is that Hobbes is making a mistake – or, if you prefer, that Hobbes* is making a mistake – both in the child case and in the ransom case. It is the mistake anyone makes when involved in a paradigm shift. From time to time, and until the shift is complete, one is prone to slip back into the old way of thinking. If this is right – and I think it is – then the mistake my interlocutors are making is they are having trouble sustaining the shift in our paradigm from natural law to positivism. This is not to suggest natural law theory is the natural way to think and positivism is the stretch. On the contrary, natural law theory is a highly unnatural way to think about law. But it is an unnatural way to think about law that has taken hold of us, and will not easily loosen its grip. But why has it taken hold of us? Why will it not easily loosen its grip? Answering these questions requires … A Somewhat Deeper Analysis My hypothesis is that we – by which I mean everyone except me – have been mistaking a mere heuristic for the real thing. This need not surprise us. We do this all the time. (And this time by ‘we’ I do include myself.) For example, some crimes, though still crimes, are not harmful: for example, smoking marijuana. But most are. So when we talk about the crime rate going up or down, we take ourselves to be talking about harm increasing or decreasing. We could cut the crime rate in half tomorrow by decriminalizing marijuana. But no one would thereby think we would have halved the harm in our society. So it is a mistake to equate crime with harm. But it is not a mistake to have adopted this equation. For only occasionally will this equation lead us astray. Similarly, then, the vast majority of our decisions are made under conditions of civil society. So vast, in fact, that for some fortunate people the state of war has become so remote that they have ceased to be able to imagine it, and so they have ceased to be able to think about what dispositions would no longer be appropriate under such circumstances. Thus the dispositions inculcated in them in and by a state of civil society go with them when they enter – what, for them, can only be the thought experiment of – a state of nature. And so in the thought experiment they

140

Liberty, Games and Contracts

make mistakes. Hobbes mistakenly thought that the child owes its mother something and that the prisoner should pay his ransom. Narveson and Gauthier mistakenly think we should discount our military endowments. And so on.26 These are relatively harmless mistakes. They eventuate in poor advice; but because this advice violates (what Steven P. Lee calls) the principle of tolerable divergence, no one takes it seriously anyhow.27 What is not so harmless, however, is actually being unable to toggle from dispositions appropriate to civil society to those appropriate to a state of war. In fact it was arguably this inability which cost six million Jewish people their lives. But if forfeiting one’s ability to toggle from dispositions appropriate to civil society to those appropriate to a state of war is liable to cost one his life, anyone who forfeits that ability must be, well, just stupid. But, as we saw earlier, that the Jewish people of Europe were just stupid is a bad error theory. Clearly, then, we need … A Deeper Analysis Still So far I have conjectured that the predilection for discounting one’s ability to worsen the other’s situation in fixing the baseline for negotiations is the product of an otherwise laudable heuristic assumption that we are already in a state of civil society, and so our capacity to worsen the other’s situation is already off the table. But, insofar as the prospect of returning to a state of war is never so far removed that we can afford to discount it, there must be some value to this discounting – beyond its savings in computational resources – that compensates for the vulnerability it generates. What might that be? I should like to propose that this discounting is a species of self-effacement.28 A mental state – be it a hope, a fear, a belief, a set of beliefs constituting a theory, or what have you – is self-effacing just in case one consequence of entertaining that mental state is that one is well-advised to cease to entertain it, if one can – call this weak self-effacement – or to entertain its opposite – call this strong. For example, one’s desperation to ‘meet’ someone of the opposite sex can often be weakly selfeffacing. So, could it be that touting one’s capacity to worsen the other’s situation is, in significantly many cases, weakly – or perhaps even strongly – self-effacing? In other words, are there significantly many circumstances under which we fare better in our negotiations by transparently forfeiting our conditional intention to return to a state of war if things do not go entirely our own way? (Note the importance of the transparency condition here.) Such situations are, I contend, legion. Most cooperative enterprises – whether hunting, making widgets or just sharing a bed for warmth – involve placing ourselves in a position of physical vulnerability to our cooperative partner. An arrow meant for the deer can as readily pierce me. A hammer for widget-making can as readily be used as a weapon. The clock radio readily at hand by which I can dispatch you is as readily at hand to you to dispatch me. But we cannot simultaneously attend both to the enterprise and to our own protection. In other words, to ‘solitary, poor, nasty, brutish, and short’ (ch. XIII, 65) Hobbes could have

Getting the Baseline Right

141

added ‘sleepless’. So we have to be able to mutually self-efface our conditional intention to return to a state of war – or more generally our willingness to worsen the situation of the other rather than just decline to better it – since otherwise cooperation becomes impossible. To self-efface a mental state is to render it virtually unrecoverable. For if it can be too readily resurrected – if its self-effacement is only surface – that it is only surface will be transparent to one’s co-player; and so she cannot let down her guard, and so she cannot cooperate. But the same is true of the first player. He cannot cooperate unless she has transparently self-effaced her conditional intention to resort to violence should she not get her way. (Hence the importance of the mutual transparency condition.) But, one might argue, since the intention to resort to violence (aka worsening) is only conditional, so long as we each ensure that the antecedent is never satisfied we are assured that the consequent will not be either. So why the need for selfeffacement? Because one cannot negotiate – at least not at close quarters – without assurance that one will not be killed (or more generally one’s situation worsened) in the course of the negotiations. So even to place ourselves in sufficient proximity to each other to enter into negotiations, we need to mutually and transparently selfefface our conditional intention to return to war/worsening should we not come to some kind of accommodation. Hence Gauthier’s insistence that in bargaining we have already foregone any resort to violence, and so threat advantage can be discounted, and Narveson’s insistence that the foregoing of violence takes lexical priority. In other words, they are not so much wrong as they are mistaken in why they are right. They are right because they are reporting on, and reflecting, the bargaining norms of people who have self-effaced the conditional intention to resort to violence/worsening. So … What Makes Me So Special? Just that, having written my doctoral dissertation on the logic of self-effacement, I am now hard-wired to look to self-effacement to explain phenomena that would otherwise be paradoxical. I am also special – or so I am told – in that I am at least slightly more psychopathic than most people, and so the pre-effaced state enjoys some residual resonance in me; whereas in my less psychopathic friends and colleagues the selfeffacement is more complete. I am, in short, Hobbes’s Foole. But it is my very moral Foolishness that gives me the philosophical leg-up on my moral betters.

Notes See his entire corpus since and including The Libertarian Idea (Philadelphia, 1988; republished Peterborough, Ontario, 2001). 2 For example, even those Canadians who are satisfied with the current health-care system are deeply worried that it is in danger of deteriorating.

1

142 3

4 5 6 7 8 9

10

11 12 13 14

15 16 17

Liberty, Games and Contracts That a normative theory must not exactly coincide with the way we are currently doing things can be contrasted with two related but different principles: John Rawls’s (wide) reflective equilibrium and Steven P. Lee’s principle of tolerable divergence (or PTD). Rawls suggests a dialectic between our theory and our considered moral judgments. John Rawls, A Theory of Justice (Cambridge, MA, 1971), 48–51. The PTD insists that the moral norms an institution prescribes must not greatly diverge from what the institution’s prudential norms prescribe. Steven P. Lee, ‘Weapons of Mass Destruction and the Limits of Moral Understanding’, in Sohail H. Hashmi and Steven P. Lee (eds), Ethics and Weapons of Mass Destruction: Religious and Secular Perspectives (Cambridge, 2004), 482–509. Thomas Hobbes, Leviathan (Buffalo, NY, [1651] 1988), ch. XIII, 63. Not so, of course, under conditions of famine, which is why moral dispositions and political arrangements break down under such conditions. Pace Marx, who thought that there would come a time when the dictatorship of the proletariat could eventually do away with itself. Pace Buddhism and the environmental movement. Hobbes, ch. XIII, 64. There is a theory of the good according to which the goodness of states of affairs varies directly with the satisfaction of desires, from which it follows that the best world is one in which we desire nothing, since then all our desires are satisfied, albeit only trivially. An alternative theory is that the good varies with the number of desires satisfied, from which it follows we should multiply our desires. A third view is that good varies with the ratio of satisfactions over frustrations. And so on. My own view is that all these views are just silly. Nothing is of value save to a valuer. Accordingly, a world in which many valuers are satisfied is better than a world in which fewer are satisfied only to a valuer who is more satisfied by more valuers being satisfied. To think otherwise is to commit – as did Mill – the fallacy of composition. Most wars are characterized by the following: much as I might want to displace you from your hillside, not if all I am going to be able to plant there is my tombstone. So war is a mixed-motive game. Hence the rules of war. The objective of total war, by contrast, is the extermination of the enemy, even at the cost of one’s own demise. Hence no cooperative dividend, hence no mixed-motivity, hence all rules barred. See, for example, Virginia Held, ‘Feminism and Moral Theory’, in Eva Feder and Diana Meyers (eds), Women and Moral Theory (Savage, MD, 1987). Hobbes, ch. XIII, 65. Hobbes, ch. XIII, 63. Opponents of same-sex marriage/advocates of family values would do well to at least consider this definition of the family. That said, scholars who think that Hobbes is a psychological egoist interpret him here as saying that it is only mutual exploitation that holds the family together. And this seems clearly false, since family members typically care for each other even when there is no prospect of further exploitation. Others, myself included, hold that Hobbes is not a psychological egoist. By our lights all he is saying here is that small groups of people can sometimes govern themselves by the bonds of affection only, however that affection is motivated. Contemporary contractarians owe this phrase to David Gauthier. See his Morals by Agreement (Oxford, 1986), 13. See, for example, Elliot Sober and David Sloan Wilson, Unto Others: The Evolution and Psychology of Unselfish Behavior (Cambridge, MA, 1999). Cf. Hobbes, ch. XV, 74–5.

Getting the Baseline Right 18 19

20

21 22

23 24

25

26

27 28

143

Paul Viminitz, ‘A Defence of Terrorism’, in Frederick R. Adams (ed.), Ethical Issues for the Twenty-First Century (Charlottesville, VA, 2005). This is not quite right, since I might have an interest in which I take no interest. For example, unbeknownst to me the car you are offering to sell me really is the deal of the century. Narveson acknowledges such cases as readily as do I. But we are of a mind that a normative political theory based on interests rather than preferences presents epistemic challenges that cannot be met. So on this privileging of preferences rather than welfare, there is no dispute between us. Elsewhere I have opined that there are only three uses one person can make of another without deferring to that other’s cooperation: as the object of perverse sexual gratification, as rather unpalatable live food, and as catapult fodder which, for some reason, one wishes to announce its own trajectory. So the concept of coercion, I claim, is really a con-concept – that is, a rhetorical flourish designed to con one’s interlocutor into accepting as a res judicata that which is precisely at issue. Hobbes, ch. XIII, 64–5. A similar mistake is made by Rawls, who thinks rational self-interested agents, negotiating from behind a veil of ignorance as to their natural and social endowments, will opt for maximal liberty consistent with a like liberty for all – the so-called liberty principle – and a Pareto-optimal distribution of material dividends, subject to the proviso that these positions of inequality be equally accessible to all – the so-called difference principle. But, insists Rawls, the liberty principle takes lexical priority over the difference principle. Rawls is wrong. Rational agents trade liberal for material dividends all the time. And rightly so. So liberal and material dividends are of a piece. Likewise of a piece, then, are life and the commodities of life. For example, Narveson, Libertarian Idea, 34–5. David Gauthier, ‘Deterrence, Maximization, and Rationality’, Ethics 94 (1984): 474–95. Reprinted in David Gauthier, Moral Dealing: Contract, Ethics, and Reason (Ithaca, NY, 1990), 298–321. Lest I be accused of siding with the strong over the weak, I should point out that precommitment strategies can be wielded by the latter as readily as the former. One could, for example, try the Gandhian ploy. ‘Yes,’ you might say, ‘by killing me I lose 4 widgets. But you lose 1. But it’s only your bottom line that you care about. So I dare you to go ahead and kill me.’ This is the same mistake made by the inhabitants of Gilligan’s Island in the eponymous 1960s US sitcom. Everyone curried favour with ‘the millionaire and his wife’, notwithstanding those millions could purchase nothing. See note 3. Self-effacement was first introduced to philosophy by Derek Parfit in Reasons and Persons (Oxford, 1984), 23–4.

This page intentionally left blank

PART THREE PROPERTY AND LIBERTY

This page intentionally left blank

Chapter 10

Liberty, Property and the Libertarian Idea Ann Levey

In part one of The Libertarian Idea,1 Jan Narveson argues that a general right to liberty as a sole fundamental right will generate rights to private property. Narveson offers this argument as part of a larger argument in part one intended to show that libertarianism is a coherent notion. In this section, Narveson is not arguing for libertarianism. Rather he aims to show (some of) the implications of beginning with the libertarian idea that ‘the only relevant consideration in political matters is individual liberty’.2 In this chapter I examine the connection between a right to liberty and rights to private property, and argue that the libertarian as described by Narveson is mistaken in the connection that he draws between liberty and private property. Property, whether private or otherwise, is a system of restrictions and permissions that control access to and use of external goods. Property controls what people can do by controlling how they get to use the world to act; it is an institution that distributes liberty of access. Such control may be on the whole a morally good thing. It may be the case that private property is the institution that does the best job at protecting a wide range of liberties. It may even be correct that rational contractors would agree to a system of private property, at least suitably constrained; such a system helps to make available an attractive range of options for many people, possibly better than do any other institutions for controlling access to external goods. These are all contingent facts, just as it is also a contingent fact that our private property system systemically limits freedom for certain social groups. I deny only that there is any intimate connection between a general right to liberty and private property such that recognition of a right to liberty requires recognition of private property. The libertarian argument that Narveson outlines mistakes a way of distributing liberty for a requirement of liberty. The concept of liberty that Narveson is concerned with is negative: liberty is freedom from interfering factors. One is at liberty to do something in this sense when there are no barriers preventing one from doing what one has the capacity or power to do.3 What is specifically of concern is social liberty, which is freedom from interference by others.4 The core idea that Narveson takes to underlie libertarianism is the general right to do as one pleases without interference by others. A right is to be understood as a claim right which imposes upon others a correlative duty not to interfere with one’s exercise of that right; rights are rights 147

148

Liberty, Games and Contracts

against one or more others.5 Duties entailed by rights are taken to be enforceable.6 The libertarian claim, then, is that enforceable property rights can be generated from an enforceable right to negative liberty. If it can be shown that property rights follow from a general right to liberty, then restrictions on those property rights will be a violation of the right to liberty, and hence not justifiable on liberty-preserving grounds. Objectors to libertarian theories of ownership will then have to argue directly against the libertarian right to liberty. Property rights, like the right to liberty, are claim rights. Narveson offers as an initial definition: ‘“x is A’s property” means “A has the right to determine the disposition of x”’.7 As Narveson notes, a right to private property is really a bundle of rights, such that a person might hold some of this bundle without holding others. Thus, for example, one could have the right to use x, but lack the right to destroy, dispose of or otherwise alienate x. Exactly which rights are needed to constitute full ownership is a notoriously difficult question. Narveson suggests the following set as giving some sort of full ownership: (1) exclusive use, which includes the right to permit or refuse use by others; and (2) transfer in the form of sale, exchange, gift or bequeathal. What Narveson wants to show, then, is that any serious restriction on this bundle of rights is an interference with a general right to liberty.8 Narveson defines interference in terms of positive action: one interferes with B by doing something, not by failing to do something. A person A does not interfere with another person B if she fails to provide B with what he needs. She interferes only if she prevents B from achieving what he could, without her action, have otherwise achieved. He argues that failing to keep distinct harming (interfering) and failing to help collapses the distinction between liberty and welfare.9 I agree with Narveson that these two concepts, liberty and welfare, need to be distinct. Nothing useful is gained in pretending that all benefits must be understood as benefits to liberty. However, defining interference by appeal to an action/inaction distinction is problematic. If I am lying on a park bench that you want to sit on, I can be described either as engaging in the action of preventing you from sitting on the bench, or as merely failing to provide you with something, namely a place on the bench. If you want to make use of some object, x, and I have claimed (and am enforcing) exclusive control of x, I can be described either as interfering with your liberty to use x, or as simply failing to provide you with the use of x. Interference need not be intentional interference. Absolute inaction is rare; rather we fail to act with respect to particular things. But the distinction between action and inaction in particular cases will at least sometimes presuppose normative claims. Whether we think that I have interfered with you or merely failed to provide you with bench space depends on whether we think that you have a claim to a seat on the bench. If you do, then I have interfered with you; if you do not, then I am merely failing to provide you with a seat. Strictly speaking, the libertarian ought to count any of these as interferences with liberty, as they meet the general definition of preventing someone from doing what they would otherwise be able to. In practice, however, we tend to think of interference as unjustified action that has the effect of preventing others from doing what they would otherwise be able to do. This raises a general concern: might one need to presuppose a normative notion in

Liberty, Property and the Libertarian Idea

149

order to determine what counts as interference with liberty? If so, then the right to liberty presupposes other rights rather than generates them. The intuitive libertarian derivation of property from liberty that Narveson describes might be put something like this. Unless people are free to act in the world, then there can be no significant liberty to act at all. Being free to act in the world requires that one be able to act on objects in the world by making use of them. If we interfere with people’s use of the world, then we are interfering with the actions that constitute a condition for significant freedom. Given liberty as a fundamental right, we have morally weighty reasons not to interfere with significant freedoms. So we have morally weighty reasons not to interfere with people’s use of the world. But to say we have morally weighty reasons not to interfere with their use of the world just is to say that we have claims to exclusive use; that is, they have property rights.10

This is an initially attractive position; it captures the important notion that freedom is the freedom to bring things about, to act in the world. Given the ways in which property rights themselves are composed in part of restrictions on freedom, however, I believe arguments of this sort do not succeed. That is, while I agree that an idea of freedom to act will provide an important part of any argument for private property rights, I disagree that a right to liberty straightforwardly generates such rights as some libertarians hold. I suggest that property rights interfere with liberty in the strict sense. For any item x, a property right in x restricts the access to x for anyone other than the owner; that is, it restricts liberty. This is a claim with which Narveson disagrees. For he says, in response to an argument by Alan Gibbard that capitalist exchange rights restrict freedom, that: ‘the essence of my having an Apple Macintosh is that I have one at my disposal when and as I wish, which latter of course requires that you not be able to simply use it any time you like; it’s not that you can’t have one unless I say so.’11 Well, right, up to a point. It is not that I cannot use an Apple Macintosh unless Narveson says so. But since I do not have a Macintosh now, I cannot have one unless somebody says so. I need permission (though Narveson does not) to use a Macintosh. Right now, my use of a Macintosh is restricted because (to the best of my knowledge) all the Macintoshes are controlled by people who would either refuse to let me use the ones they control or who would impose a cost on my using one of them. Since Narveson does hold that imposing a cost on doing x counts as an interference with x,12 he should be committed to recognizing an interference here. Similarly, in Oxford in the 1960s anyone was at liberty to use a bicycle, and in Tokyo today anyone is at liberty to use an umbrella, both in the most straightforward of senses. There is available a variety of (respectively) bicycles and umbrellas and no one is employing coercive force to prevent anyone from using them or exacting a cost for their doing so. The same is not true of Apple Macintoshes. Narveson is right that I have the power (and actually I do have the cash) to buy an Apple computer. But only after I buy one is my liberty to use one not restricted by others. The libertarian argument, I suggest, needs to make an illegitimate move from ‘A is not (is) free to do x’ to ‘A is (is not) being forbidden to

150

Liberty, Games and Contracts

do x by a coercive state authority’. That is, the libertarian argument here relies on a contrast with an imagined egalitarian state in which the government is saying: ‘no one but us is permitted to control external objects’; or ‘no ownership allowed’; or ‘no access to these kinds of things unless we [the government] say so’. But that is the wrong contrast, since the issue is only whether private property rights interfere with liberty, not whether egalitarian states also interfere with liberty. Of course Narveson agrees with at least some of what I have said so far. He is certainly not committed to the absurd claim that rights (including property rights) do not interfere with liberty. As he says, the claim that property rights interfere with liberty ‘must be made out in some more interesting sense than that in which … every right ipso facto restricts liberty’.13 Since any right interferes with liberty, what the libertarian needs to show is that property rights can arise without illegitimate interferences with liberty. Hence what is needed is an account of legitimate interference in order for a general right to liberty to give any content to claims about what is and what is not prohibited. Narveson suggests two ways of determining when restrictions on liberty are legitimate. Either (1) a restriction is legitimate if a system allowing such restrictions produces a greater net amount of liberty than the alternatives, or (2) a restriction is legitimate if there is some fundamental sphere of liberty attached to each person such that the restriction is outside that sphere.14 These restrictions serve to mark out areas of legitimate and illegitimate interference. A right to liberty then would be the right to do as one wants as long as doing so is not an illegitimate interference with others. The first method is comparative. The restrictions that legitimate interferences with liberty are those that produce more net liberty than any alternative set of restrictions. The second method identifies a specific set of liberties as basic, and counts as legitimate interferences those that are required to protect that basic set. There is no particular reason to think that these two methods will produce the same rights. Someone may hold, for example, that the basic liberty that needs to be protected is freedom of movement – that is, travel across physical distances – and hold that private property rights, unless limited, will interfere with such a liberty. Specifically, unless either ownership is limited by rights of way or sufficient land remains unowned, property rights will interfere with liberty of movement (except perhaps for those who have vast enough holdings). It might, of course, as a matter of fact be true that even if all land is owned that all owners give permission to any who wish to move across their territory. But since it might not be true, protecting such liberty requires limiting ownership, not merely leaving things up to the contingent desires of owners. However, it need not also be true that the net liberty is thereby greater in that system than in one without such restrictions. It does not follow from the fact that property rights restrict freedom of movement that placing restrictions on the restrictions thereby increases liberty. It may be that the total number of instances of interference with liberty without such restrictions on ownership is greater, even though some people lack freedom of movement and hence are subject to an interference that would count as illegitimate according to the first view. I now consider the connection between liberty and private property in the light of

Liberty, Property and the Libertarian Idea

151

these two suggestions for determining what counts as a legitimate interference. The position that Narveson takes to be the libertarian position seems to be a version of what Lawrence Becker calls the ‘why not’ argument for initial appropriation.15 In the absence of a reason to think appropriation is illegitimate, it is morally permissible; private property comes about, as it were, by moral default. Narveson says: if we begin with the root idea that people are to be allowed to do whatever they want to do, then so long as any trains of action are conceivable and really instantiable, the question can only be whether a given sort of action will collide with the legitimate liberties of others. If it does not, then given our starting point of a general right to liberty, anyone who can do the action described and wants (for whatever reason) to do it is to be allowed to do it.16

If people are able to engage in acts that constitute appropriation through the making use of external resources to act, and if they are in turn able to engage in secondorder activities such as buying or selling, then in the absence of reason to think any of these an illegitimate interference in liberty, these are legitimate activities. Hence interfering with these activities is an illegitimate interference with liberty. But these activities are just what ownership consists in; hence if they are legitimate exercises of freedom, private property rights are legitimate interferences. If in addition any interference with a legitimate interference is itself illegitimate, then it looks like what we have is private property rights. Whether this succeeds in generating property rights will depend on what counts as an illegitimate interference. I am not convinced that it does succeed regardless of which of the two methods suggested above is used to determine what counts as legitimate interference. First, I suggest that on the criterion of greater net liberty, at worst exclusive property rights increase the number of interferences with liberty, and at best they simply distribute them differently. In neither case is there an increase in liberty, and so the greater net liberty principle cannot show that the interferences implied by exclusive rights to property count as legitimate interferences with the liberty of others while interferences in those claims are illegitimate interferences. Second, I suggest that on either criterion, it is not obvious that at least certain kinds of interference with acts of ongoing use (that is, acts of the sort that are meant to constitute appropriation) will be illegitimate. Some interferences with appropriation look a lot like others merely going about their own business of acting in the world. Further, depending on what is taken to be the protected sphere, equal protection may be required. To show that private property increases net liberty it would need to be shown that the set of property rights in question produces a greater balance of liberty than any available alternative system of rights. What we say here will depend on how we describe an interference with liberty. Since our concern is with social liberty, we need to count the total number of instances in which one person, A, brings it about that another person, B, ‘is unable to do what B would otherwise be able to do even without A’s assistance’.17 In that case, however, to count net liberty is to count the total instances of cases in which A brings it about that B is unable to do what B

152

Liberty, Games and Contracts

otherwise would have been able to do. The state that has the fewest such instances will have the greatest net liberty. Indeed, the use of the term ‘net’ here is misleading; since we are only to count interferences, there is nothing to balance such interferences against. We are to look only at the total number of interferences. Alan Ryan offers an illuminating example to suggest that total liberty is not increased by exclusive private property rights.18 Suppose we compare the situation where 200 people have rights of common use over some piece of land, where these rights include the rights to hunt, gather and pasture, with the situation where one of those people has acquired exclusive ownership of the land. In the former situation each of 200 people was free to engage in a variety of activities. In the latter one person is free to engage in those activities, plus free to engage in a few more. The other 199 people now lack the liberty to engage in those activities, although each is free to acquire property. It seems to me that the system of common use wins. There are literally more restrictions on people’s liberty under private ownership, since the bulk of the population are restricted from doing much of what they could under a system of common use, while only one restriction is lifted. This example suggests that a general right to liberty will give rise to a right of common use, not to ownership. In general a system of common use of external resources will allow everyone to engage in a wide variety of activities without interference. In contrast, the net amount of liberty in a system of private property rights is highly dependent on patterns of distribution. Those who own no or little property have more restrictions on what they are allowed to do without interference than they would under a system of rights to common use. This example does not rely on comparing a system of ownership rights to alternative systems. It looks only at the consequences to liberty from the introduction of property rights where the baseline is pre-property common use. If people acquire property rights over things as a result of making use of them, what they acquire is the right to exclude others from use and the second-order rights of buying and selling. Second-order rights depend on there being first-order rights to exclude. Activities like selling and buying lose their point if anyone is already free to make use of external resources. Where there are no property rights, the ongoing use of something is not protected. Others are free to make use of that thing if they so please. The rules for park benches apply to the world; if A stands up, then B is free to sit down. As long as B does not wrestle A off the bench, there is no illegitimate interference. It is only once we think that one party has an exclusive claim to the bench that subsequent use without permission appears to be illegitimate. I suggest this makes it hard to claim that the introduction of property rights counts as an increase in liberty (where what we mean by that is fewer interferences with people being able to do what they otherwise could). Consider again the park bench that all are free to use. If we think that A’s (non-appropriative) use of the bench does not interfere with B, even though B cannot simultaneously use the bench, then the introduction of property rights is an increase in the number of instances of interference, since it introduces an interference where none existed before. If we think that A’s (non-appropriative) use of the bench does interfere with B (who cannot simultaneously use it) then we should think not that rights to

Liberty, Property and the Libertarian Idea

153

exclusive use introduce new restrictions, but that they fix certain distributions of restrictions such that use by others now requires they get permission. But in either case, there is no increase in net liberty to demonstrate that the interference imposed by exclusive property rights is legitimate. At best there is a stalemate where acts that exclude and acts that interfere with exclusion stand or fall together as legitimate or illegitimate. Narveson does give a response to this sort of argument. He says that what is relevant to a comparison of two systems with respect to liberty is whether the people ‘who have things got them by forcing others to do this or that, that is to say got them by restricting the liberty of others’.19 If they got external resources without restricting the liberty of others, he says, the fact that they now have ‘the right to prevent others from using them does not show that there is now a restriction on liberty which there wasn’t previously’.20 But I must confess to being baffled by this response because before there was no restriction on others using the external goods and now there is. Indeed, I claim that even where A creates something new, if A exclusively owns it, then there is an interference with liberty that would not exist if A did not exclusively own it. That interference may be a good thing. It may be that the variety of activities in which we can engage in an ownership system would not exist without exclusive use because private property encourages productivity and creativity. But none of this implies that it must thereby increase liberty in the strict sense of reducing the number of instances of interference. I conclude that the greater net liberty argument does not get the libertarian from liberty to property. What then about the protected sphere of liberty approach to determining when interference is legitimate? According to this criterion an interference with liberty is legitimate provided that it is outside the sphere of some fundamental liberty that the right to liberty protects. One libertarian candidate for such a protected sphere, one that Narveson might have in mind, is self-ownership, property rights in the self.21 A right to self-ownership is the right to dispose of one’s body and one’s mind as one sees fit. Because a right to self-ownership is a property right, it involves the right to exclude others from use. Thus such a right is frequently taken to prohibit such activities as taxation, insofar as that is regarded as forced use of someone else’s labour. The key idea that the right to self-ownership contains – and, I think, the idea that libertarians really want to get at – is the right not to be forced to do what you don’t want to do. Thus we find Narveson equating restricting others’ liberty with forcing them to do this or that.22 This is why libertarians are willing to accept the interferences with liberty inherent in rights to private property, but not willing to accept the interferences with liberty inherent in redistributive taxation. The latter, but not the former, involves forcing people to do what they may not want to do. If we examine the general argument from the perspective of rights to selfownership, the kinds of acts that are taken to constitute appropriation are unproblematic. Exclusive appropriation of unowned stuff does not violate others’ rights to self-ownership, and hence there is no duty to refrain from such acts and the subsequent exclusion of others from what has been claimed. Do property rights in

154

Liberty, Games and Contracts

oneself generate property rights in external things? It might be thought that if one is at liberty to do whatever does not violate others’ rights of self-ownership and acquiring private property does not violate such rights, then one is entitled to acquire private property. It is certainly the libertarian view that from the perspective of liberty there would be something wrong with not allowing such acquisitions. But what exactly is it to acquire rights to private property, or to fail to allow the acquisition of such rights? As Narveson notes, acquiring is not an act. You acquire something by engaging in an activity that constitutes good reason for thinking that you have rights over that thing.23 What is there about engaging in an activity such as, for example, enclosing, ploughing and planting a bit of unowned land that is supposed to constitute good reason for recognizing rights of ownership in relation to that land? According to Narveson that good reason, in the libertarian view, will be whatever good reason there turns out to be for allowing people to do as they please, plus the observation that in engaging in those acts you were engaging in an ongoing track of activity, allowance of which is just what owning that bit of land consists in.24

If the protected sphere is conceived as self-ownership, that amounts to saying that the good reason for allowing the activity will be just that the activity does not violate anyone else’s right to self-ownership. But why think that when someone engages in these sorts of activities anything more is going on than simply someone engaging in these activities? Why is allowing these activities supposed to somehow constitute recognizing rights to own the land, rather than only rights to use the land? Since rights to use do not entail rights to own, the former could be allowed without allowing the latter. Rights to use do not include a right to exclude, and hence do not include the second-order rights of buying and so on. Failing to allow ownership is not something that the government of a socialist state needs to do. Failing to allow ownership is as simple as failing to recognize the rights inherent in ownership by not acknowledging rights to exclude and transfer. We can do this individually by simply continuing to use what we please without recognizing a duty to refrain, and we can do it collectively by failing to publicly enforce such rights. In other words, we can fail to acknowledge ownership rights simply by all of us acting as if no such rights exist, by failing to engage in the institutional and social practices that constitute recognition of such rights. Someone who fails to acknowledge exclusive appropriation does not interfere with self-ownership rights of the want-to-be owner because they do not force the want-to-be owner to do anything. Once we talk about forcing people to do things, rather than interfering with liberty, it becomes clear that failure to recognize acts as acts of appropriation does not require force; it merely requires disregard. If force is the mark of illegitimate interference, then the simple act of ignoring first claimants and going about one’s business as if they had made no exclusive claims need involve no force. If we begin with nothing but a property right in oneself, we are under no more obligation to supply people with property rights than we are to provide them with anything else.

Liberty, Property and the Libertarian Idea

155

It may be that in focussing on self-ownership I am construing the fundamental sphere of liberty too narrowly. What the libertarian wants to protect is the right to act.25 This is the attractive feature of the intuitive argument outlined earlier. But then we are again faced with the problem that property rights for some will entail that others lack the freedom to act. They lack the freedom because the property rights stop them from acting by prohibiting them, as non-owners, from making use of what is owned. Narveson takes it to be unproblematic if some do not own because others got there first. No one, he says, has a duty to provide people with opportunities.26 True, but if freedom to act by acting on the world is part of the protected sphere, we do have a duty not to interfere with people’s freedom to act on the world, as property rights do interfere if one has no such rights. Property rights determine who has access to what. Public property rights and private property rights share that as a general feature. Where they differ is in the mechanism for determining the distribution of permissions and restrictions. With public property, some central body such as the state determines who has access. Under private property, that control is more broadly distributed. But, in both cases, property controls access by denying access to some and permitting it to others. That is, it interferes with people doing what they would be able to do in the absence of such rights. The libertarian says that we are under no obligation to provide people with things. But that move between rights and things is problematic. If they lack computers in Nicaragua and we send them computers, we are supplying them with things. Here those computers are all around us; people do not lack things, they lack access to things. Granted the things would not exist without private property. But that is an issue of productivity, not of liberty. Someone concerned only with liberty is not entitled to appeal to the greater productivity made possible by private property.

Notes 1 2 3 4 5 6 7 8 9 10

11 12 13

Jan Narveson, The Libertarian Idea (Philadelphia, 1988). Narveson, 7. Narveson, 22–23. Narveson, 30. Narveson, 42. Narveson, 48. Narveson, 64. Narveson, 79. Narveson, 38. See Narveson, chapter 7 and especially p. 85. For a more extended discussion of this, see Ann Levey, ‘Liberty and Property’, in Samantha Brennan, Tracy Isaacs and Michael Milde (eds), A Question of Values: New Canadian Perspectives in Ethics and Political Philosophy (Amsterdam, 1997), 87–105. Narveson, 77. Narveson, 38. Narveson, 63.

156 14 15 16 17 18 19 20 21 22 23 24 25 26

Liberty, Games and Contracts Narveson, 50. Lawrence Becker, Property Rights: Philosophical Foundations (Boston, MA, 1977). Narveson, 81. Narveson, 30. Alan Ryan, Property (London, 1987). Narveson, 77. Narveson, 78. Narveson, 66. Narveson, 77. Narveson, 85. Narveson, 85. Narveson, 80. Narveson, 85.

Chapter 11

Who Owns Me: Me or my Mother? How to Escape Okin’s Problem for Nozick’s and Narveson’s Theory of Entitlement Duncan MacIntosh

The Apparent Tensions in Nozick’s Libertarianism As Jan Narveson notes, Robert Nozick is usually seen as grounding his libertarianism in individual property rights, rights he is then criticized for not grounding in turn.2 But, in fact, Nozick grounds the rights of individuals to hold property in their rights to liberty.3 And he grounds their liberty rights in two other things: first, in their capacities and inclinations to exercise them, that is, their personhood;4 second, in the idea that it requires argument to say that individuals have coercible duties to others, and that the only way this can be argued is by showing that these duties are required in order to optimize the freedom of all agents.5 The morally uncontroversial baseline from which positive duties are a departure requiring justification is a situation of unfettered freedom to act, to exercise the capacity for liberty.6 And to establish coercible duties, one must derive them from the liberty rights that all people possess presumptively: immorality is unjust interference with others, interference which does not aim to preserve equal freedom from interference for all. From these assumptions derive property rights: since freedom is freedom to act, and all action is interaction with the material world, all freedom is freedom to interact in some way with the world. To have freedom to interact with something material in some way is to have property in it to that extent: thus the relation between liberty and property. This is how Narveson elucidates Nozick; and Narveson has taken on the prosecution of this idea.7 If Nozick grounds his views in property rights, his theory will seem fraught with internal tensions, and it will be unable to account for certain moral duties. Susan Okin, for example, claims his theory of property is incoherent, and in exploring how Nozick might escape contradiction, she argues that he cannot account for our duties to children. She alleges a contradiction in his theory of entitlement (ET), of when a person has just title to a holding.8 Nozick thinks, she claims, that since people own themselves, they derivatively own what they produce. But, Okin observes, people are produced by their mothers. Thus, on Nozick’s principles, mothers own people. But then people do not own themselves.9 This undermines 157

158

Liberty, Games and Contracts

Nozick’s libertarian view that people may not be interfered with if they do not interfere with others. For owners may ‘interfere’ with what they own; mothers own their offspring; so people may be interfered with by their mothers. Thus, people do not (all) have a negative right to liberty (automatically). This, Okin thinks, leaves two options. First, maybe ET is false: people are not entitled to all and only what they find, trade for, are given or make. Thus they need not be owned by the mothers who make them. Second, maybe ET is true; but since it entails that people are owned by their mothers, it does not establish libertarianism, but rather matriarchy and slavery in which all property and rights to control people’s lives are held by their mothers. This, Okin thinks, is a non-starter due to all the absurdities of a society in which persons, including mothers, could not gain self-ownership unless their mothers gave or sold them their freedom.10 To avoid this dystopia, we must reject the ‘principle that persons are entitled to whatever they produce, regardless of the needs of [others]’.11 But this means, thinks Okin, that Nozick must ‘retreat from his … theory of rights and the minimal state he builds on it, and … return to a more “patterned” derivation of justice that takes into account needs, deserts, and other human capacities as well as productivity.’12 Surely the last is too quick: for why is the alternative that ownership of goods should be determined by, for example, needs? Still, Okin’s analysis leaves the core of Nozick’s theory – the principle of acquisition, or PA: a principle which specifies the ways in which someone can acquire a holding to which they then have and retain just title according to ET – mired in self-contradiction. If persons do not even own themselves (since their mothers do), in the sense of being entitled to their own persons, bodies, natural talents and abilities, then there is no basis for anyone’s owning anything else. ET is premised on the notion that each person owns himself. But when we consider women’s reproductive capacities and labour, this notion, so central to PA, is undermined by that very principle.13 In Nozick’s defence, one might argue that (A) people are entitled to what they make except when they make other people, owing to something in the nature of people; or that (B) mothers do not ‘make’ other people in the relevant sense, and so have no title to them.14 But against (B), Okin thinks having a child is a clear kind of production by one’s abilities and talents, to the fruits of which in general Nozick thinks one has title. While against (A), she thinks his theory cannot discount persons as kinds of things which can be owned by someone through having been produced by her. Okin might grant that we could patch up Nozick’s theory by stipulating (arbitrarily, she would think) people to be exceptions to ownability by production. But she thinks it impossible to derive (A) or (B) from his first principles. Libertarians find this implausible. For since Nozick is defending the property rights of persons, and since offspring are or become persons, surely his philosophy is rightly read as giving them a property right in themselves overriding property claims their mothers may make to them. So, for instance, has Narveson claimed.15 But while I think this is ultimately on the way to the right conclusion, to assert it without further defence here is to miss the depth of Okin’s point. She knows Nozick

Who Owns Me: Me or my Mother?

159

defends the property rights of persons, and that offspring become persons. But Nozick specifies property rights both as what people have in themselves in being persons, and as what they supplementally acquire in other things by finding, being given, trading for or making something. Thus while a person has an automatic property right in himself, if a person makes another person, she has a prima facie property right in the person she makes, in a being who, because he is a person, has a conflicting right to self-ownership. Something must give: either people do not automatically own themselves, or the theory that they acquire a right to something by making it is not true with perfect generality. Libertarians want to resolve the conflict in favour of the person the mother makes having an overriding right to property in himself. But by what principle is this justified? If the principle that one owns oneself and the principle that one owns what one makes are co-foundational, there is an irresolvable conflict. Yet if the principle that one owns what one makes is foundational, we have Okin’s matriarchal dystopia. While if the principle that one owns oneself is the real foundation, there is still a problem. For on Okin’s reading, Nozick takes the principle that one has a property right in what one makes as derived from the self-ownership principle. But then that principle seems to contain the seeds of its own contradiction if people are made by people. But I suggest property rights are not the foundation for Nozick, or for any fully self-comprehended libertarianism. Something else is, and it both entails and delimits property rights in such a way as to resolve the contradiction. My strategy is to use Okin as a foil to penetrate deeper into libertarian principles. I say we can give a systematic foundation for (A) – the view that people are entitled to what they make except when they make other people, owing to something in the nature of people – without contradicting ET. For Nozick’s philosophy is grounded not in rights to property, but to liberty; rights in turn grounded in individual capacities to exercise liberty claims, and in an ethical minimalism in which coercible constraints on people’s liberty can only be justified if needed to optimize everyone’s individual liberty. Nozick seeks to accord every person the maximum liberty compatible with a similar liberty for all. This entails liberty to hold property where that does not interfere with the similar liberties of others. This is why offspring own themselves: were they owned by their mothers, they would have no liberty; so a greater liberty for all would result from individuals being self-owned upon becoming persons, which is why the rights of people to what they make cannot be so extensive as to permit them to own people they make.16 On to details.

Okin’s Arguments Okin has a strong prima facie case that given Nozick’s theory of when agents are entitled to something (ET), according to which people have just title to anything acquired in accordance with his principles of acquisition (PA), mothers own their children; and it is worth reviewing her case, not only because we must criticize it and find within libertarianism the resources for doing so, but because it shows just what aspects of Nozick’s philosophy that might be thought to entail that mothers do

160

Liberty, Games and Contracts

not own their offspring, do not really have this consequence. This shows we must reconceive the foundations of libertarianism, Nozickean or otherwise, to avoid the unhappy conclusion. For Okin, a woman’s production of an infant meets PA – a principle entailing that the maker of x, having contracted for all other resources used in the process, is entitled to x.17 Pregnancy and birth is a paradigm of such processes. Once she is freely given a sperm or buys one – either way a legitimate transfer – a fertile woman can make a baby with her own body and its nourishment. She need not resort to fraud or force to get the one resource her body cannot contribute. An infant originates from a minute quantity of abundantly available and otherwise useless resources. Thus there can be no dispute over how much of the product comes from the added value of the mother’s labour; the complex capacities of the female reproductive system and its labour transform two cells into an infant.18 Okin then deals with objections. Perhaps mothers cannot own children because God does; because He, ultimately, made them. But then no one could own anything because God already owns it.19 That women do not ‘comprehend and have full control over [the] production’ of children is no objection either ‘since this would preclude ownership of … other products – such as trees one … planted and nurtured – that [Nozick] would not want excluded from the category of property’.20 Nor is it an objection that women’s production of babies is sometimes fairly effortless or unintentional. For Nozick ‘defends the property rights of the naturally talented to the full fruits of their talent[s and capacities] (see … his Wilt Chamberlain example), and of those who “stumble upon” something to what they have found, however valuable.’21 But is not reproduction ‘different from … other forms of production that lead to ownership, in that it has a different kind of purpose or internal goal?’22 No matter; Nozick says ‘the producer alone is entitled to determine the purpose of his activity … Thus he [cannot object to] a woman’s producing a child for whatever purpose she chooses: to keep it in a cage to amuse her … or even to kill … and eat it …’23 One might argue that people cannot be owned. But ‘Nozick … points out that Locke does not claim that something in the nature of persons precludes their being owned; after all, he postulated that they are owned, by God, and precisely because he made them.’24 And ‘Nozick … allows that [people] can [sell themselves into slavery].’25 (She admits this ‘does not imply that they can already be the property of another at birth … [for,] Nozick adds, “some things individuals may choose for themselves, no one may choose for another”.’26) One might argue that giving the mother title to her child would violate the Lockean Proviso that, in acquiring property, one must leave enough and as good for others. ‘If mothers own all children at birth, doesn’t this monopoly violate the ownership rights of men and infertile women, not to mention the rights of children to self-ownership?’27 But Okin argues that this is false on Nozick’s reading of the proviso where the rights of the nonowners are violated only if they are left worse off than they would be in a ‘baseline’ situation in which the owners did not exist. (Nozick, pp. 176–82.) And this

Who Owns Me: Me or my Mother?

161

is … not so in … reproduction, for if there were no fertile women there would be no children to be owned by anyone … Just as, according to Nozick, a medical researcher who discovers a new cure may justly refuse to sell except on his own terms, childbearing women, by refusing to share … ownership of children … are not depriving others of anything they could acquire without the women’s special talents[;] … these women do not ‘worsen the situation of others; if [they] did not [produce children] no one else would have, and the others would remain without [them]’. [a paraphrase of Nozick, p. 181] Since the children themselves … would not exist in the baseline situation, they can make no claims.28

More worrisome – and here I will eventually dig in – is the objection that allowing a mother to own her children, and to use them at her whim, would violate ‘the moral side constraints that Nozick claims protect persons and their liberties from the assaults of others’.29 But Okin thinks Nozick so specifies the characteristics in virtue of which persons are protected by side constraints as to leave infants, small children and the developmentally disabled completely unprotected by them. He presents a traditional list of characteristics that he supposes are the reasons for the constraints on how persons may treat one another. But no infant, very small child, or person with serious developmental disabilities has these traits. Thus Nozick cannot regard infants as having any such inviolable rights. So he cannot argue against a mother’s right to dispose of her infant as she chooses. Thus by ET children are the property of those who make them.30 But if this establishes a mother’s ownership of an infant, it may not establish her ownership of the person into whom it will develop; for while the infant lacks the required traits, the person does not; and if Okin thought Nozick had an out if infants had these traits, why not when adult persons acquire them? Her answer: ‘long before [persons] qualify for … side constraints, they are, according to [PA], first and foremost the property of their mothers’.31 But surely this establishes only that persons used to be the rightful property of their mothers, not that their now being so treated would not violate the constraints. But Okin thinks that, since Nozick emphasizes ‘legitimately acquired property rights over all other claims, including basic needs and the right to life’, he cannot ‘relax [these rights] … in order to give an infant, who is … the product of someone else’s body and labour, the right of self-ownership, in contravention of [PA]’.32 She adds: (i) ‘As Nozick writes: “No one has a right to something whose realization requires … things and activities that other people have rights and entitlements over”,’ and (ii) ‘If I am (already) my mother’s property, I cannot claim a conflicting right to own myself.’33 But (i) does not establish that since the person requires things from the mother to develop, if she provides them, she has clear title to the person; only that she may not be compelled to provide those things, that the person is not owed them. But just because I have no right to your help in watering my garden, it does not follow that if you water it for me (without me hiring you), you own it.34 And (let us assume) no one is forcing the mother to make a person. If she contributes things necessary to the existence of the infant, then maybe she owns it; for since it is not a person, it has no competing basis on which to assert self-ownership against her usual right to whatever she makes. But when it becomes a person, it does. Okin’s point (ii),

162

Liberty, Games and Contracts

however, argues that the mother’s title has priority. But to prove (ii), Okin must show that PA overrides the side constraints; for adult persons have a claim to selfownership. Her argument is that Nozick gives priority to those who affect others over those they affect….[So he cannot] label as unjust a [mother choosing to keep her child] as a slave for life [rather than giving it freedom]. The rights of entitled donors must have priority over the expectations of potential donees, such that – regardless of the inequality of the results – the latter cannot claim that injustice has been done to them.35

Okin concludes: [It is] a central assumption of [Nozick’s] theory that persons originally … own themselves. For if [they] are born as the property of another, how can they have ‘rights … [s]o strong and far-reaching … that they [restrict] … what … the state and its officials may do’ (Nozick, p. ix). And how can anyone acquire a just Nozickian title to property if he does not own his own labour or … body? [But] … persons … start their lives as the product of a woman’s natural capacities and labour. [Thus] Nozick’s theory, in spite of its apparent dedication to self-ownership, [entails] that women’s entitlement rights to those they produce must take priority over persons’ rights to themselves at birth.36

Can Nozick reply?

Basic Nozick and the Basic in Nozick Nozick is the Garbo of philosophy; he just wants to be (left) alone. His central, outraged intuition: how dare people compel others to do things if they are not bothering anybody? He thinks people have Lockean rights:37 they may not have their moral boundaries crossed by others, may not be physically assaulted, coerced or killed.38 This is a negative right to non-interference from others. People also have a right to pursue their conceptions of the good life.39 Both rights have limits: you forfeit your right to non-interference if you violate the rights of others. Only then may others infringe your boundaries – others may do this in order to prevent or to rectify your transgressions; but only to the extent needed to attain these effects.40 Your right to pursue your conception of the good life is bounded by the similar rights of others. (Actually, your right to pursue your conception of the good life is limited by others’ boundary rights only in consort with their life-pursuit rights. For if others had no inclination to exercise their life-pursuit rights, nothing would count as violating their boundary rights.) These rights are side constraints:41 no goal can morally justify transgressing people’s rights to non-interference and to pursue the good life. Who has Lockean rights such that we must respect side constraints in dealing with them? Anyone with a conception of the good life: that is, every ‘person’. A person is sentient, ‘self-conscious; rational (capable of using abstract concepts, not tied to responses to immediate stimuli); possessing free will; … a moral agent

Who Owns Me: Me or my Mother?

163

capable of governing its behaviour by moral principles and … of engaging in mutual limitation of conduct …’. It has the ability ‘to have or strive for meaningful life’, ‘to regulate and guide its life in accordance with some over-all conception it chooses to accept’.42 These rights are, at root, one. Nozick, inspired by Locke, is moved by one precept, (L): people have a non-overridable right to such liberties as do not interfere with those of others.

This is the right to do as much as possible of what you want consistently with all other people being able to do as much as possible of what they want. You have a right only to these liberties because all people have rights, and the only ones all could have are ones co-tenably haveable and exercisable by all; the only co-doable activities of people are those which do not prevent or preclude the doing of something else by another; which do not, in that sense, make him worse off (worse off in not being able to do something he wishes to do). From this derives property rights, entitlements to holdings: one is entitled to (1) whatever one appropriates from nature provided one does not interfere with anyone else’s acquiring things of equal value, does not make anyone else worse off relative to a pre-appropriation baseline level of welfare; (2) whatever anyone gives one gratis or in a free and informed trade, provided they got it legitimately, and (3) whatever one makes from what one justly appropriates, is given, or trades for. 43 Appropriating things from nature as in (1) is just more doing of what one wants that does not interfere with others doing what they want; I want and appropriate this, and you cannot object, because there is something of equal value you can appropriate if you want to do so. But neither of us may take both this and the something else, for then we leave nothing of equal value for the other. Acquisition by trade as in (2) is just capitalism among consenting adults, people doing what they want without interfering with others.44 Acquisition by manufacture, as in (3), gives one title to what one makes, because what one makes would not exist unless one had made it, so that, in claiming it, one is not interfering with another’s prior right or freedom to claim it, is not making him worse off than before one made it in terms of what was then available for him to appropriate relative to his baseline level. For someone else to appropriate what I made for my use would be for them to use me as a mere means: to violate my moral boundaries, to prevent me from doing something harmless to others. Self-ownership also derives from L. For ownership of something is just the freedom or right to do what one wants with it (provided that does not interfere with others). And if one has the right embodied in L – a right to liberty, to do as one wants (with oneself) subject to the proviso – then ipso facto, (S): one owns oneself.

Since having liberty is just being permitted to try to satisfy one’s desires, to

164

Liberty, Games and Contracts

prosecute one’s conception of the good life, only those who have such a conception can have liberty; and any such being has it. Thus one’s rights to liberties derive simply from one’s having aims, things one has a primitive right to try to satisfy. So persons have, essentially, rights to liberty.45 To give every person such liberties, each person’s permissibly prosecutable aims must be such that no person infringes on another’s. Rights must be co-tenable. Thus, your having aims creates an obligation on me not to interfere with them, provided the aims are such that your advancing them does not interfere with others advancing their aims. But if all persons’ aims give their possessors a right to try to satisfy them, if I desire to interfere with your desire-fulfilment, have I not a right to interfere? No. For it is impossible to satisfy all persons’ possible desires to interfere with others’ satisfaction of their desires. So a right to satisfy the desire to interfere with each other’s pursuits is not something all people can be owed. Thus libertarianism embeds a Kantian condition on the desires that agents may try to satisfy (without the consent of others): those of all people must be jointly satisfiable. I may only do what I want provided you can still do likewise. This means that what we each may do depends on what all others want. We may only act to satisfy such of our desires as participate in a maximally large set of the aggregate co-satisfiable desires of all. If there is more than one set, different communities may choose different ones, and we may seek or create the communities we find most congenial to our desires. But all communities, in turn, are required to be co-tenable.46 This idea is perhaps better expressed in terms of liberties: people are not at liberty to interfere with each other’s liberties; people’s liberties must be co-tenable; each person only has such liberties as participate in a maximally large set of co-tenable liberties of all and so on. I have spoken here of aims, wants and desires only to connect permissible liberties with their ultimate grounding in the capacities and inclinations of agents. It is because I have a conception of the good life, some aims, that I have a prima facie right to pursue it, to advance those aims. But the aims I may advance are limited by the condition that the aims any given person may advance must be co-tenable with as many as possible of the aims of every other person. There may be circularity in this idea of a permissible liberty, though I hope there is only mutual constraint. As many have suggested, I expect that to fully understand it, we must use contractarianism.47 There are a number of persons, each with some values; and there are resources for advancing the values. Libertarianism, I hold, says each person is initially entitled to a distribution of those resources (or at least a right to appropriate from them) as would be provided in one of the Pareto-optimal solutions to the bargaining problem involved in distributing resources. Since there may be more than one optimum (because there is more than one possible distribution of values among persons, or more than one way of optimally distributing resources in a Pareto-optimal satisfaction of the values of persons on a given distribution of values), each involving a different distribution of resources, the actual entitlement of a given person depends on which optimum has been agreed upon.48 There is not room here to dwell on this idea; but I think it will prove well-enough defined for what I want to do with it in the sequel. I will refine it as needed later.

Who Owns Me: Me or my Mother?

165

The Origins of People, the Nature of Persons and the Basis of the Right to Negative Liberties Were people found, not made, the side constraints on what we may do to each other would forbid us being appropriated from nature and owned by others without our consent. For people are owed L, and L implies S; people own themselves. They may also appropriate found things, unless that interferes with other people’s acquiring things of equal value. But if you ‘found’ and used me against my will, you would interfere with my rightful liberty. Normally, though, persons are the terminus of a set of stages of a living thing: conceptus (the immediate product of the union of egg and sperm), foetus (the conceptus until birth), infant (the foetus after birth until it becomes the child), child (the infant once it attains conative and cognitive traits like those of intelligent animals – for example, dogs – until it becomes the person), and person (the child once it acquires person-traits). Okin thinks the mother’s right to her offspring is rooted in the Nozickean dictum that people own what they make, and in the fact that mothers make children with their womanly powers from non-scarce, justly acquired resources. Do women ‘make’ children in Locke’s sense? Maybe women only have babies; they do not make them, any more than they make their livers or urine. No matter; Nozick does not think one comes to own something simply or only by investing deliberate labour in it. (Recall his reductios of Locke’s mechanisms of acquisition.49) One originally owns oneself simply by having person-traits, and so the negative liberties to which persons have a right. One acquires other things simply by appropriating them without violating the proviso – by asserting a prerogative which no other persons could justly challenge, for none of these persons is harmed (that is, none of them is thereby prevented from appropriating something of equal value). My making of something with the intention of using it is a just appropriation (by an act of inventing something intended for my use), provided it worsens no one’s appropriation options relative to the baseline level preceding my act of manufacture. (A complex idea: what if I make something which competes in the market with a product you make, and mine outsells yours and puts you out of business? Have I not harmed you, illicitly by the proviso? No. You have, first, a maximal claim on nature – a range of possible appropriables determined by what is in nature and how many people wish to make a claim on its goods – and Lockean rights thereafter. My manufacture does not interfere with your maximal claim on nature, or with anyone’s Lockean rights thereafter; for the way I put you out of business is by capitalist acts between consenting adults. It is no business of yours when other people exchange things to which they have title, and nothing you can complain about if they decline to exchange with you – unless they thereby deny you compensation to which you are entitled: if all the white people have all the water, a necessity of life, they may keep it only provided they are prepared to make it available to black people in return for something of equal value. Likewise if, say, the Indians have all the farmland.50) But it is also just, my appropriating something that would not exist but for me. Concede that women may do this with babies: the

166

Liberty, Games and Contracts

mother owns the conceptus – if the spermatozoon which formed one half of it was given to her, or was unclaimed by its ‘donor’, or if she bought it; also the foetus – if she entered no contract of co-ownership, and owning it perhaps because she ‘made’ it from the conceptus, or because it would not have existed but for her, or because it is a part of her and she owns herself;51 also the infant and the child, because she ‘made’ them or at least grew or nurtured them, or because they would not have existed but for her and lacked the traits needed to have a claim to self-ownership against her. The process of making children into persons is not done solely by mothers, but usually also by fathers, friends, teachers and even by the children themselves, which could complicate things for Okin. But never mind; it could happen that they get made primarily by their mothers. And it is more plausible to say that women make persons out of their babies than that they make their babies, since child-rearing requires deliberate labour. So there is an argument for mothers owning the people they make: they made them, and one is generally entitled to what one makes. However the person made has an argument too: people have Lockean rights; so in our treatment of them we must respect side constraints. No one, then, may come to own persons simply by dint of the act of making them; for to make a person is to make a thing with Lockean rights, one which has S unless it consents to ownership by another. What grounds a person’s right to negative liberty (and so to S) is not that someone gives or concedes it; rather, as we saw, it is her being owed it simply due to her desiring to exercise it, having a conception of the good life. (She can trade that liberty away. She may sell herself into slavery for favours from the master. If she then says, ‘but I really want to have that liberty!’, too bad; a deal is a deal. The difference is, it was a deal. One’s desire to exercise Lockean liberty is sufficient for one having title to it unless one has contracted for the alternative, or has violated the Lockean rights of another.) One has the right to do what one wants (with oneself) simply by wanting to do it, provided it is co-doable with others doing what they want. (Recall that this just means one has a right to a liberty to do something with oneself if doing it does not interfere with the similar liberties of others to do things with themselves.) Since property is the right to do what one wants with something, people own themselves. Now since people may do what they want provided they do not interfere with others, and since they may want to find, trade, accept or make things for their use, they may do so, provided they interfere with no one’s liberty. The right to freely use the objects of such acquisitions, transactions and manufacturings is the right to hold property, since having property in something is having the right to free use of it. But since one cannot make and unconsentfully use other people without violating their liberties, one does not get property in someone by making them; rather, one makes a thing with property in itself.

Determining Side Constraints and Rightful Liberties Who wins the argument? Okin thinks Nozick must side with the mother. Giving S

Who Owns Me: Me or my Mother?

167

to the person the mother makes violates her property rights in what she makes. Okin demands a principled Nozickean reason for conceding ownership of people to themselves and not their makers, given that people in his scheme are generally – and, she thinks, foundationally – entitled to what they make. I resolve this into three demands: a demand for a principled reason to think offspring are protected by side constraints, for reading the entitlement principle as outweighed by the constraints, and for reading the constraints as giving self-ownership to persons, not their makers. On the first: offspring become persons; they are, as of then, like all persons, protected by side constraints. On the second: Nozick’s primary aim is to defend people’s negative liberties, things to which he thinks all people have a right, and from which all their other rights derive, including rights to holdings. Everything follows from one basic principle of Negative Liberty (NL) – which is L in a different guise: NL: any being with a conception of what her life should be may do what she wants provided it numbers among a set of activities which consists in a maximum set of activities in which people may engage without interfering with each other’s activities.

NL is just an explication of L. L, recall, said people have a non-overridable right to such liberties as do not interfere with those of others. NL is just L except that it is explicit on what a person is, namely, a being with a conception of what her life should be; and it is explicit on what counts as a liberty that does not interfere with the like liberty of others – namely, anything which numbers among a set of activities which consists in a maximum set of activities in which people may engage without interfering with each other’s activities. Thus a person owns (may do what she wants with) herself, and owns (may use as she wants) what she creates, provided she acts consistently with NL. Suppose creating a person is not itself an ‘interfering’ with that person. Then the mother does not interfere with the person she makes in making it, so she may make it.52 But it does not follow that she owns it – has the right to use it without its consent; for she would then use it unjustly, violating NL. For NL, first, guarantees original selfownership to persons and, second, permits people holdings except when that would violate another’s self-ownership or the similar permission of holdings for others. NL must do this or it would not give everyone the maximum possible negative liberty compatible with similar liberty for all. Indeed, the alternative is incoherent. It would say people own themselves (may do what they want with themselves) by having NL, but do not own themselves (may not do what they want) by virtue of being made by another, and so owned by the other (who may say what they will do). So NL must give no one so much liberty as to be able to use another without their consent. People are owed NL simply in being people. In making a person, one makes a being protected by NL side constraints which restrict possible entitlements, and which explain how entitlements can even arise. For if we did not so read NL, people would not own themselves – because their mothers would own them – so they could not own anything else; assuming a regress of ungenerous mothers, no one would be entitled to anything.53

168

Liberty, Games and Contracts

But does not giving S to the people made violate the liberty of makers to use their creations, violating their rightful side constraints? Answering this will meet Okin’s third demand. The correct side constraints must afford the maximum co-possible negative liberties for all. But giving people liberty to make others for their own use would give negative liberty to fewer people than would giving the liberty of selfownership to those made; for in the former, only mothers (at most) would have negative liberty, while in the latter, everyone would have it. So it constrains person B’s rightful liberty to make something for her use that she may not make a person for her use; person A’s liberty cannot be so constrained as to allow person B the liberty to make A for B’s use – B may make A, but B may not use A without A’s consent. Okin saw Nozick’s libertarianism as grounded in a right to acquire holdings. This makes it problematic to give ownership of people to themselves and not to their makers since people are generally entitled to what they make. But in fact, Nozick defends negative liberties first, property second. People are not owed liberty because they may own things; they may own things because they are owed liberty, and so may own anything – and only things – their ownership of which does not interfere with another’s liberty. Giving ownership of people to their makers would violate the liberty of those made; and according priority to their liberty is a condition of all people having maximum co-possible negative liberties, perhaps even any negative liberties at all. So a person is the one thing the mere making of which cannot ever – because of what it is – entitle the maker to its use, for that would deprive it of its own liberty. (There are other things the making of which contingently cannot entitle the maker to their use. For example, the mere fact that I made this neutron bomb does not mean I have an automatic right to set it off. My making it would mean I had a right to set it off, however, if there were no other people but me, or if all people wanted me to set it off,54 for then doing so would violate no one’s rights. But if what I make is a person, for me to use what I make would, obviously, inevitably violate someone’s rights: those of the person I made.) Far from being inconsistent with ET, then, the non-ownership of persons by mothers follows from it. For persons are loci of entitlement. Therefore, creating a person is creating such a locus, and so is a process which does not give to the creator title to the thing created, but bestows it on the creatant. Okin claimed that long before the baby becomes a being to whom Lockean rights and side constraints apply, it is a thing made by the mother, and a thing to which, therefore, she has prior title by ET. But I would reply that it is impossible to have prior title to a thing to which one cannot be entitled. The mother’s making of an infant may establish her ownership of it, but it cannot establish it of the person whom she makes from it; even if ownership of infants violates no side constraints, ownership of persons without their consent does. Still, if the mother owned what she made the person from, why not the person? How can her title to it evaporate? The explanation lies in what it becomes – a person. Thus Nozick’s libertarianism is neither self-contradictory nor matriarchal in its implications for self-ownership and just acquisition.

Who Owns Me: Me or my Mother?

169

Notes 1

2 3

4 5 6 7 8 9

10 11 12 13 14 15

16

17 18 19 20

Jan Narveson has been an inspiration for his intellectual vigour, and I am grateful to have an opportunity to write on themes that have interested him. My thanks to my colleagues at Dalhousie University for help – especially Richmond Campbell, who gave me written comments, and Bob Martin and Sue Sherwin. Thanks also to Malcolm Murray for his editorial efforts. My work was supported by a grant from the Social Sciences and Humanities Research Council of Canada. See Jan Narveson, The Libertarian Idea (Philadelphia, 1988), 62 for a review of those with this take on Nozick. Robert Nozick, Anarchy, State, and Utopia (New York, 1974). Recall that, for Nozick, a property right in something is a right to use a thing in a certain respect, a right one acquires by an act of appropriation from nature (Nozick, 150–3, 174–8), subject to the Lockean Proviso that enough and as good be left behind for others to appropriate (ibid., 178–82); or by trade, gift or invention (ibid., 150–3). These are all activities, and one’s property rights derive from one’s right to pursue activities, subject to their not being activities which impermissibly interfere with the activities of others (ibid., pt I). Nozick, 48–51. This is the express method of the first third of his Anarchy, State, and Utopia: anarchy is permitted unless a constraint on it can be justified. This is the position of Nozick’s anarchist. See Nozick, pt I. See Narveson, ch. 6. Susan Moller Okin, Justice, Gender, and the Family (New York, 1989), ch. 4, 74–88; all references are to this work unless otherwise stated. Actually, she should have said that they do not all own themselves automatically. For surely if there were some first mothers, they would automatically own themselves; and anyone would own himself if given to himself by a self-owning mother. Okin, 86. Ibid. Ibid. Ibid. Nozick raises both possibilities, but his discussion is casual and inconclusive, trailing off in a reverie on stockholder’s teas and colonies on Mars; see Nozick, 287–91. In a paper at the 1991 Canadian Philosophical Association Meeting in Kingston, Ontario. See also Joshua Cohen, ‘Okin on Justice, Gender and Family’, Canadian Journal of Philosophy, 22/2 (June 1992): 263–86, especially 273–7. This is a more sympathetic reading of Nozick than he is usually given. Ultimately, I think, it may afford a new appreciation of the moral resources and plausibility of libertarianism, and new ways of applying its principles to questions of policy. I suspect that this reading of libertarian principles will demand policies rather to the left of those which current defenders of libertarian ideology and politics would accept. But I think the resulting conception of ethics would in fact reveal libertarianism’s true moral character. Most of that is for another time, however (see my ‘Children Among the Libertarians’, ms, Dalhousie University, 2005); for now, I just want to show that Nozick’s libertarianism entails neither incoherence nor Okin’s matriarchal dystopia. Nozick, 160. Okin, 82–3. Okin, 80. Ibid.

170 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46

Liberty, Games and Contracts

Okin, 83. She is citing Nozick, 161–3, 181. Okin, 84. Ibid. She cites Nozick, 233–4. Okin, 80. Okin, 81. Ibid. Okin cites Nozick, 331. Okin, 84. Ibid. Ibid. Okin, 84–6. Okin, 86. Okin, 82. Ibid. She cites Nozick, 238. For the general principle, see Nozick, 95. Okin, 82. Okin, 81–2. Nozick, ix, 10. Ibid. and chs 2–6 generally. Nozick, ch. 3. Nozick, chs 2, 4–6. Nozick, 29–32. Nozick, 48–50, as summarized by Okin, 85. Nozick, 150–3. This is also the general thrust of Nozick, pt I. Nozick, 163. For references, see the notes in my introduction, above. I take this paragraph to be faithful to the thrust of Nozick, pt III; see especially 298–308, 312, 316 and 324. 47 As does Nozick, chs 2, 5, 6, 10. See also Narveson, chs 12–14. 48 This raises several problems. For example, since the values of the persons involved in bargains determine the permitted and required distributions of resources, this may affect the entitlements of children. If the children, once they become persons, wish to change the contract, must we accommodate them? Or must they accommodate to it? I take up some of these themes in my ‘Categorically Rational Preferences and the Structure of Morality’, in Peter Danielson (ed.), Modeling Rationality, Morality and Evolution; Vancouver Studies in Cognitive Science, vol. 7 (New York/Oxford, 1998), 282–301. Also relevant is my ‘The Mutual Limitation of Needs as Bases of Moral Entitlements: A Solution to Braybrooke’s Problem’, in Sue Sherwin and Peter Schotch (eds), Engaged Philosophy: Essays in Honour of David Braybrooke (Toronto, 2007). In ‘Children Among the Libertarians’ I derive some left-wing conclusions about our obligations to others from the foundations of libertarianism, conclusions surprising given the usual presumption that this foundation has right-wing consequences (although there is a growing literature on ‘left-libertarianism’). 49 Nozick, 174–5. 50 Nozick, ch. 7, sect. I; ch. 8. 51 If one owns oneself, does one own every part of one’s self? Possibly not; selfownership might not consist in a mereological sum of self-owned self-parts. For example, suppose some of my parts are antibodies to an illness everyone has contracted, one fatal to all others unless they receive a vaccine cultured from a sample of my antibodies. It may be that, while I own myself, I do not own (all of) my antibodies.

Who Owns Me: Me or my Mother?

171

These may be a natural resource my appropriating all of which would be excessive given their absolute scarcity, and given that they are essential to life, like water. 52 She may, however, be illicitly interfering with other persons, for she is making a new claimant to nature’s resources, thus diminishing the share of nature other persons may rightfully claim, since the new person, as a person, may now have claim on a share. I argue, in ‘Children Among the Libertarians’, that this fact may allow us to deduce from libertarian premises obligations of parents to look after the welfare of their children. Okin doubts this can be done. She asks ‘Why should ownership lead to responsibility [of mothers for the welfare of their infants] here, whereas it leads to entitlement to use or dispose of at will in other cases?’ (81) 53 Actually, there is more to say on the logic of this. Suppose A makes B; then arguably A owns B. Suppose B then makes C. Who owns C? Arguably B, because B made C. But it could be A, because B is A’s property, so A may set any conditions on B’s activity, for example, that B may make C only provided A will then own C. Or maybe C is owned by nobody (unless by herself). C is not owned by A since A did not make C. And C is not owned by B because, since B does not own herself, arguably she cannot own anything else – her labour is not hers, so neither is anything she mixes it with. 54 Though it may not be morally permissible for all persons to have such wants – see above on the implications of libertarianism for the permissible distributions of values among agents, which implications may include duties to have values inclining one to respect certain duties to one’s self, for instance, duties not to kill one’s self (under most conditions) – or duties to avail one’s self of certain opportunities to interact in certain ways with nature. This may eventually allow us to meet another worry about libertarianism – that it cannot account for duties to one’s self.

This page intentionally left blank

Chapter 12

On Original Appropriation Peter Vallentyne

Libertarianism holds that agents initially fully own themselves. Lockean libertarianism further holds that agents have the moral power to acquire private property in external things as long as a Lockean Proviso – requiring that ‘enough and as good’ be left for others – is satisfied. Radical right-libertarianism, on the other hand, holds that satisfaction of a Lockean Proviso is not necessary for the appropriation of unowned things. This is sometimes defended on the ground that the initial status of external resources as unowned precludes any role for a Lockean Proviso. I shall show that this is a bad argument. Although I would argue that satisfaction of a Lockean Proviso is indeed a necessary condition for the appropriation of unowned things, I shall not attempt to establish that here. My goal here is more modest: to rebut one argument against the Lockean Proviso. The Lockean Proviso can be interpreted in several different ways. Nozickean right-libertarianism interprets the proviso as requiring that no one be left worse off by the appropriation than she would be if the thing remained in common use.1 Equal share left-libertarianism interprets the Lockean Proviso as requiring that no one be worse off than she would be if no one appropriated more than an equal share of the competitive value (that is, based on demand and supply) of initially unowned things.2 Equal opportunity (for well-being) left-libertarianism interprets the Lockean Proviso as requiring (roughly) that no one be worse off than she would be if no one appropriated more than is compatible with everyone having an equally valuable opportunity for well-being.3 I shall not here assume any particular version of the Lockean Proviso. The arguments advanced against it are typically directed at the Nozickean interpretation, but they apply equally well to the more egalitarian interpretations. For simplicity, I shall refer to ‘the’ Lockean Proviso. Both radical right-libertarianism and Lockean libertarianism hold that acquiring private property over a previously unowned thing requires that the agent perform some suitable action in relation to that object. The most familiar account requires that the agent mix her labour with the object (whatever exactly that means). Other accounts require that the agent discover the object, take possession or control of it, improve it, or merely stake a claim over the object. Virtually all accounts require that the agent at least stake a claim in the sense of publicly communicating that she now has private property rights over the object. In what follows, I shall leave this

173

174

Liberty, Games and Contracts

issue open; but for brevity shall merely write of staking a claim. Our question concerns whether staking a claim (or related activities) in an unowned object can be sufficient for acquiring private property rights over it, or whether in addition the Lockean Proviso must be satisfied. Before proceeding, let me acknowledge that there is one important argument against the Lockean Proviso that I will not consider. Israel Kirzner argues that those who discover the existence of a resource, or discover an unknown valuable use for a known resource, are really creators of an economic resource. If one believes, as I do, that the property rights of creators over their creations are not subject to a Lockean Proviso, then Kirzner’s claim, if correct, provides a strong reason for thinking that a Lockean Proviso is not needed in at least many cases of initial acquisition. I reject Kirzner’s claim that discoverers are often really creators of the resources in question, but defending that rejection is more than I can do here. Hence, I set that argument aside. There are two closely related issues that we shall be examining. (1) Can an individual acquire private property rights over a previously unowned thing even when this violates the Lockean Proviso? Throughout, I leave implicit that we are addressing cases where no one else has consented to the acquisition of such rights. The problematic case for libertarianism is appropriation without the consent of others. (2) Does someone commit an injustice if – in violation of the Lockean Proviso – she treats a previously unowned thing as if she had private property rights over it? The question here is not whether she acquires private property rights, but, rather, whether she commits an injustice by acting as if she did. Throughout justice is understood as non-violation of rights. The idea that the initial unowned status of external things leaves no room for the Lockean Proviso to do any work is at least roughly implicit in the writing of many libertarian writers. Jan Narveson, for example, writes ‘[T]he world is [initially] just stuff, devoid of moral qualities and not owned by anyone. … The State of Nature is not naturally equipped with any rules whatsoever, about anything.’4 The most explicit statement is given by Edward Feser: Suppose an individual A seeks to acquire some previously unowned resource R. For it to be the case that A commits an injustice in acquiring R, it would also have to be the case that there is some individual B (or perhaps a group of individuals) against whom A commits the injustice. But for B to have been wronged by A’s acquisition of R, B would have to have had a rightful claim over R, a right to R. By hypothesis, however, B did not have a right to R, because no one had a right to it – it was unowned, after all. So B was not wronged and could not have been.5

Of course, some might question the claim that there are any initially unowned things. One might claim, for example, that all external resources are initially jointly owned by all in the sense that no one is permitted to use such resources without collective consent of some sort (for example, unanimity or majority). This, of course, is a crazy view and I fully agree that natural resources are initially unowned. In any case, we shall assume this for the sake of argument.

On Original Appropriation

175

Consider now Feser’s example. Let us agree that, when A uses resource R, A’s use of R as such does not violate B’s rights. So far, there is no injustice. No one has any claim rights over the unowned thing. Everyone is initially at liberty – has a liberty right – with respect to the use of the unowned thing. A, however, is doing more than simply using R. She is treating R as if she privately owned it. Suppose that she uses minimal and reasonable force against B to stop him from using R – where this use of force would not violate the rights of B, were A to fully own R. (Private property, after all, includes some rights of prior restraint to stop individuals from violating one’s property rights.) Does A commit an injustice? Does A, that is, violate B’s rights? Here, we must distinguish two cases. In the first, A has R in her physical possession (for instance, in her hands), and B uses force against A to obtain R. Here, B violates A’s self-ownership, and A is at liberty to use reasonable force against B to stop her from violating A’s self-ownership. In this case, A commits no injustice in using force to stop B from using A. This use of force, however, is justified by the fact that B is violating A’s self-ownership – and is fully compatible with A having no private property rights over B. In the second case, A does not have R in her physical possession. She has used R in the past (for example, watered a tree), perhaps with a plan for future use (for instance, gathering apples); but she does not currently have the object in her physical possession. Suppose that B comes along and uses R (for example, gathers some apples) and that A shows up and uses force against B to stop her from using R. Does A violate B’s rights? The answer is (1) that she does not violate B’s rights if, in fact, A has acquired moral private property rights over R (and she is merely enforcing her rights); and (2) that she does violate B’s right, if R is still unowned and A is merely treating R as if she owned it (and A is still at liberty to use R). Thus, we must turn to the first question raised above: can A acquire private property rights over R merely by staking a claim (or other relevant actions) – even when this violates the Lockean Proviso? The thought, I think, is that, if a thing is unowned, then there is no barrier to the first claimant acquiring private property rights over it. After all, if it is unowned, it is, it seems, not subject to any moral restrictions. This, I shall argue, is not so. No one, we have granted, has any claim rights concerning the use of an unowned object. There is more, however, to the moral order than first-order conditions (for instance, claim rights and liberty rights). There are also second- (and higher-) order conditions on when lower-order conditions change. Consider, for example, the moral powers agents have to acquire claim rights over unowned things. The radical right-libertarian claim is that agents have a very strong moral power to acquire such claim rights. All agents have to do is to stake a claim (or engage in related activities). This is certainly one possible second-order moral view. It is, however, a moral view – a view about the conditions under which the liberty rights and claim rights of individuals are altered. It holds, in effect, that agents have a very weak immunity to losing their liberty rights to use unowned objects (since some of those liberties are lost whenever someone else stakes a claim). No matter what position one takes on how strong the powers of appropriation are, one is thereby taking a

Liberty, Games and Contracts

176

moral position on the immunity rights that agents have. There is no morally neutral position. Moreover, each of these different second-order positions is fully consistent with the first-order condition of a thing being initially unowned (that is, no one having any claim rights over it). In sum: the first-order status of an unowned thing is clear: no one has any claim rights concerning its use (that is, everyone is at liberty to use it). That leaves open, however, its second-order status (the conditions under which the first-order status can change). One compatible position is that individuals can acquire claim rights over first-order unowned things merely by staking a claim (or related activities). Another is that this is so only when the Lockean Proviso is satisfied. There are infinitely many other possibilities as well (for example, whenever the king so dictates). Radical right-libertarians are thus mistaken if they claim that their position follows from the mere fact that initially no one has any claim rights over natural resources. First-order conditions of morality are compatible with all kinds of second-order conditions. This does not, of course, establish that radical right-libertarianism is mistaken. It merely establishes that one particular argument for it is mistaken. The real work must be done by addressing the substantive moral issues of appropriation. It cannot be done simply by appealing to the fact that external things are initially unowned.6

Notes 1

2 3

4

5

See, for example, Robert Nozick, Anarchy, State, and Utopia (New York, 1974). There are many exegetical questions about how Nozick intended his version of the Lockean Proviso to be understood, but this is one natural reading. See, for example, Hillel Steiner, An Essay on Rights (Cambridge, MA, 1994). See, for example, Michael Otsuka, Libertarianism without Inequality (Oxford, 2003). For more on left-libertarianism generally, see Peter Vallentyne and Hillel Steiner (eds), The Origins of Left Libertarianism: An Anthology of Historical Writings (New York, 2000) and Peter Vallentyne and Hillel Steiner (eds), Left Libertarianism and Its Critics: The Contemporary Debate (New York, 2000). For a critical assessment of leftlibertarianism, see Barbara Fried, ‘Left-Libertarianism: A Review Essay’, Philosophy and Public Affairs 32 (2004): 66–92; Peter Vallentyne, Hillel Steiner and Michael Otsuka, ‘Why Left-Libertarianism Isn’t Incoherent, Indeterminate, or Irrelevant: A Reply to Fried’, Philosophy and Public Affairs 33 (2005): 201–15; and Barbara Fried, ‘Left-Libertarianism, Once More: A Rejoinder to Vallentyne, Steiner, and Otsuka’, Philosophy and Public Affairs 33 (2005): 216–22. Jan Narveson, ‘Property Rights: Original Aquisition and Lockean Provisos’, Public Affairs Quarterly 13/3 (1999): 205–27, 118 [reprinted in Jan Narveson, Respecting Persons in Theory and Practice (Lanham, MD, 2002), 111–31]. See also Murray Rothbard, For a New Liberty: The Libertarian Manifesto, revised edition (New York, 1978), 31–6; Jan Narveson, The Libertarian Idea (Philadelphia, 1988), ch. 7. Edward Feser, ‘There Is No Such Thing As An Unjust Initial Acquisition’, Social Philosophy and Policy 22 (2005): 56–80, 58–9. Feser actually makes the stronger claims that ‘there is no such thing as either a just or an unjust initial acquisition of resources. The concept of justice, that is to say, simply does not apply to initial acquisition. It

On Original Appropriation

6

177

applies only after initial acquisition has already taken place.’ He makes this stronger claim because he believes that there are no applicable norms of justice. This, however, seems false. Individuals are posited to have the rights of self-ownership and thus there are some applicable norms of justice. Using an unowned resource in a way that violates no one’s self-ownership respects those applicable norms and thus is just. In any case, I shall ignore this stronger claim and focus on the weaker claim that appropriation in violation of a Lockean Proviso is not unjust. For helpful comments, I thank Ed Feser, Jan Narveson, Mike Otsuka, Eric Roark and Hillel Steiner.

This page intentionally left blank

Chapter 13

Morality as an Evolutionarily Stable Strategy: A Naturalistic Account of Libertarianism Grant A. Brown

If a putative moral principle cannot be ‘universalized’ in the appropriate sense, then it must be rejected. The contemporary debate concerns precisely what is required by the universalizability criterion, given that many variants and strengths of universalizability are conceptually possible.1 I maintain that the sense in which moral principles must be universalizable is that morality must be an evolutionarily stable strategy (ESS). The Concept of an Evolutionarily Stable Strategy The ESS can be explained by way of an illustrative example.2 In provisioning their nests, female digger wasps follow two distinct strategies: burrow or borrow. The first strategy is to burrow a hole in the ground, catch and paralyse four or five katydids, deposit them into the hole one at a time, lay an egg on the top of the pile, and then seal up the hole before beginning the process all over again. The second strategy is to search for an abandoned hole, provision it, lay an egg, seal up the hole, and go looking for another one. The second strategy might seem advantageous relative to the first, since it exploits the digging efforts of others. But it has a disadvantage, too: if the hole has not really been abandoned, then the intruder might spend time and effort provisioning it only to have its digger return, lay her egg and seal it up. If too many wasps employ the second strategy, available holes become scarce and the chances of double-occupancy go up; whereas if too many wasps dig their own holes, then some will be abandoned and digging efforts will go to waste. There is therefore a critical frequency at which the two strategies are equally and optimally beneficial, and the population of digger wasps settles into an evolutionarily stable state around that critical frequency. There are three distinct ways in which this critical frequency could be achieved. First, there is the case where a mixture of pure strategies is employed within the population – that is, when x per cent of the population always burrows and (100-x) per cent always borrows. Second, there is the case where every individual employs the mixed strategy of burrowing x per cent of the time and borrowing (100-x) per cent of the time. Finally, there is the class of cases where the population as a whole employs any mixture of mixed strategies such that, when concatenated, burrowing 179

180

Liberty, Games and Contracts

occurs x per cent of the time and borrowing occurs (100-x) per cent of the time. Of these three ways in which the critical frequency can be achieved, the most interesting and most common in nature is the middle one – the ‘evolutionarily stable strategy’ or ESS. John Maynard Smith defines an ESS as ‘a strategy such that, if all the members of a population adopt it, then no mutant strategy could invade the population under the influence of natural selection.’3 The ESS is important because it is more capable of invading, more resistant to invasion, and more robust than any of the other combinations of strategies. If a population achieves the critical frequency by having two separate castes that practise either pure strategies or a mixture of mixed strategies that are not ESSs, then a mutation that practises the ESS fares no worse than the existing individuals, and so can survive within that population under evolutionary pressures. Moreover, such an individual can leave behind any number of offspring without undermining the critical frequency in the population; whereas an individual employing any other strategy might upset the stable balance if she were to leave behind too many or too few offspring. Likewise, in situations where predictable play could be recognized and exploited, randomized play of the ESS would be required. Castes that practise non-ESSs are too predictable in some circumstances. In more complex scenarios, it is possible for there to be two or more critical frequencies that are evolutionarily stable. The one that the population settles upon under evolutionary pressures will be a function of the initial conditions facing the population.4 If a catastrophic event occurs to a population which is in an evolutionarily stable state, it might not settle back into its original local optimum, but rather flip to a different optimum with its own distinct mixture of strategies, or even spiral into extinction. A mixture of strategies that is sensitive to perturbations is not very robust. ESSs are more robust than either mixtures of pure strategies or mixtures of mixed strategies within a population because catastrophic events do not alter the proportion of strategies being employed within the population, as they would, for example, by differentially killing off castes employing any other strategy.

Tit-for-Tat: An Example of an ESS in a Moral Context Many moral problems can be modelled as an iterated prisoner’s dilemma (PD). In this section I maintain that the strategy variously referred to as reciprocal altruism, conditional cooperation or tit-for-tat (TFT)5 is the only ESS in an iterated PD scenario – contra those who claim that always-defect (AD) is also an ESS, and those who maintain that no strategy is an ESS in an iterated PD scenario. I begin by showing that AD is not an ESS, since it is possible for TFT players to invade a population employing AD under the conditions of ‘replicator dynamics’. The alleged problem for TFT players is that, if an individual begins each new association by trying to cooperate it is bound to be frustrated in a world populated exclusively by ADers, since it will never observe cooperation from others in order to reciprocate it. In fact, the practitioner of TFT would be exploited on the first

Morality as an Evolutionarily Stable Strategy

181

occasion in which it interacted with every other individual. Since being exploited by everyone else once is worse than being a straight ADer and thus being exploited never, the question arises of where the advantage of conditional cooperation comes from. The answer is that the advantage comes in the second and successive generations. Suppose, plausibly, that the first individual with the TFT mutation manages to produce offspring in spite of suffering a small amount of exploitation at the hands of its AD associates. Then, in a sexually reproducing species, half of those offspring can be expected to possess the TFT allele, too; and, being siblings, they are very likely to find a great number of opportunities over their lifetimes to benefit from cooperation with each other. Although the parent does worse by virtue of possessing the TFT allele, half of her children can be expected to do very well on account of it – in fact, under common conditions, better than conspecifics which lack the predisposition to cooperate with cooperators. Moreover, the advantages quickly escalate as more conditional cooperators enter the population. The grandchildren that possess the TFT allele – now slightly more than half the total number of grandchildren of the original mutant – can be expected to do even better than the children of the original TFT mutant. In the usual way, the TFT allele comes to predominate in the population; that is, it can invade a population of ADers. Axelrod has shown that it is theoretically possible for a small cluster of TFT practitioners to invade a population of ADers, and replicator dynamics shows how a small cluster could get its beginning. Thus AD is not in fact an ESS in iterated PDs. I now show that TFT is an ESS, under plausible real-world assumptions. Many theorists have shown in various ways that TFT is not an ESS in the theoretical models they have studied.6 Indeed, they maintain that there can be no ESS in the iterated PD. Moreover, Marinoff has shown that some strategies that incorporate a cooperatively weighted expected-utility maximization (EUM) decision procedure are more successful and more robust, by several measures of robustness, in certain iterated PD contexts than any member of the TFT family of strategies.7 The success of Marinoff’s cooperatively weighted EUM family of strategies, and in particular his MAC strategy,8 suggests that a slight tendency toward a ‘probing rudeness’ is more successful in the significant sample of iterated PD contexts explored by him. Still, he cautions that ‘Owing to MAC’s exploitativeness, any translation or implementation of MAC in the context of a social, economic, or political prisoner’s dilemma could have undesirable repercussions.’9 The reason this caution is welladvised is that the computer-simulated PD contexts and mathematical models assumed for the purpose of deriving these theoretical results are quite remote, in some respects, from the real-world situations that biological, especially human, agents face. It is entirely likely that strategies in the TFT family – that is, ones characterized by niceness, provocability, forgiveness and simplicity – will remain the most robust in real life. Consider the implications of just the following two critically important dissimilarities: In the computer simulations, each strategy is forced to interact with each other strategy, whereas in most real-world scenarios people have considerable latitude

182

Liberty, Games and Contracts

over whom they interact with. This is of fundamental significance, because in the real world nice people can simply stop wasting their time interacting with cheats and probers as soon as defection occurs the very first time. An unforgiving strategy suffers the cost of non-cooperation in a significant proportion of its total interactions in the computer tournaments; but an unforgiving person in the real world might well be able to find enough other nice people to interact with that ceasing forevermore to interact with opportunists after their first defection will cost them little.10 Yet it will cost the opportunists rather significantly, because they will quickly run out of nice people to try to exploit. Thus unprincipled opportunists, including even relatively cooperative ones like MAC, are much more likely to be left out in the cold in the real world before they can realize their long-term exploitative plans. They cannot benefit from what Robert Nozick refers to as the interpersonal and intrapersonal functions of moral principles.11 This is especially true when – again as in the real world but unlike the computer simulations – players can take advantage of reputation effects. In the real world people can check up on the history of others’ interactions with others, either by direct observation or indirectly by, for example, running a credit check. Anyone who is unwilling to be monitored or audited is not a good prospect for nice people to interact with; whereas nice people will be eager to show others their track record – they have an incentive to increase their transparency. Robert Frank argues that the human capacity to send out credible (that is, difficult to fake) signals of one’s trustworthiness is one way to improve cooperation-enhancing transparency.12 Someone who never manifests the physiological signs of a guilty conscience when they go wrong is not someone nice people should interact with, given a choice. Finally, Nozick notes that ‘The interpersonal function of assuring others that justice is being done or that principles are being followed might necessitate following principles that are less subtle and nuanced but whose applications (and misapplications) can sometimes be checked by others.’13 Marinoff indeed shows that members of his EUM family of strategies, though successful against other families of strategies, fare poorly against themselves precisely because they cannot reliably recognize each other due to the randomized play in the initial period of interaction.14 A certain degree of simplicity in human affairs is a distinct virtue. All things considered, then, it is plausible that the set of properties that makes TFT robust in Axelrod’s original simulations will confer much greater robustness in real-world, human societies. In short, a robust moral strategy must be characterized by principled niceness, moderate provocability and forgiveness, and transparent simplicity. As Dawkins says, ‘although Tit for Tat is strictly speaking not a true ESS, it is probably fair to treat some sort of mixture of basically nice but retaliatory “Tit for Tat”-like strategies as roughly equivalent to an ESS in practice. Such a mixture might include a small admixture of nastiness.’15 We do not and should not expect all moral agents to practise precisely the same degree of forgiveness or retribution (and so on) – as long as they practise enough. That is, our universalizable moral principles must admit of a range of flexibility and degree of discretion in how different individuals apply them.

Morality as an Evolutionarily Stable Strategy

183

Indirect Reciprocal Altruism I do not go as far as some might go by suggesting that morality is coextensive with TFT in iterated PDs. Rather, I suggest that the essence of morality is captured by the characteristics that make TFT a successful ESS in iterated PDs: namely, a slight tendency toward niceness, with a strong tendency toward reciprocity – of either the ‘you scratch my back and I’ll scratch yours’ variety, or the ‘eye for an eye’ variety. But reciprocity is not always as direct and obvious as is implied by the above aphorisms. The more subtle forms of morality are generally a product of indirect reciprocity, and can be every bit as evolutionarily stable in human society as TFT is in iterated PDs. It pays to discuss the more subtle forms of indirect reciprocal altruism at greater length because it is perhaps the least well-appreciated form of reciprocity. There are very considerable advantages for individual humans to live in large societies in which most of those one interacts with are essentially strangers. Just as interaction between genotype-identical cells leads to specialization of cellular function within increasingly complex and versatile multi-celled organisms, so too does interaction between reciprocal altruists lead to specialization of individual function within increasingly complex and thus versatile multi-organism societies.16 Besides the purely economic benefits identified by Adam Smith17 – increased expertise and efficiency, economies of scale, and diversification – an important consequence of specialization and increasing group size from the point of view of human evolution is that it leads to better protection from external threats, particularly threats from other groups of humans. Citing many sources, Richard Alexander concludes that it is a ‘virtual certainty’ that the driving force of human evolution has been an arms race between human groups, involving within-group cooperation in the service of between-group rivalries.18 As proto-human societies evolved into larger and larger groups of mostly unrelated, interdependent individuals, the fate of each individual came to be more and more tied up with both its standing within the group, and the fate of its group as a whole. Morality evolved in humans as a system of principles for managing these complex interactions. But morality includes far more than can be captured by the direct and immediate reciprocity of the ‘you scratch my back and I’ll scratch yours’ variety suggested by TFT. In theory, the causal feedback loop that returns an advantage to benefactors can be as indirect and circuitous as you please; although, as a practical matter, the more indirect and circuitous the path, the greater will be the opportunities for exploitation. When this feedback loop involves more than two interactants, we have a case of indirect reciprocity. It is useful to distinguish two forms of indirect reciprocity: reputation effects and social investments.19 As already noted in the previous section, reciprocity works indirectly through reputation when a third party chooses to cooperate with a member of a cooperating pair, because the third party has observed that they are cooperators with each other. (That is, C observes A and B cooperate, and then C cooperates with A or B.) Reciprocity works through social investments in two ways. The first occurs when one expects to be repaid for a benefit by someone other than the recipient. (For

184

Liberty, Games and Contracts

example, A helps B, B helps C, and then C helps A.20) The second occurs when an individual promotes and thus benefits from group cohesion. (For example, C smooths over a dispute between A and B, thus allowing the team of A, B and C to get on with a task.21) It is easy to underestimate the significance of indirect reciprocity in moral systems, and much of Alexander is concerned with showing how prone philosophers are to making this error.22 Alexander, indeed, makes the opposite mistake of thinking that moral systems are entirely systems of indirect reciprocity. In a highly interdependent society, where opportunities for cooperative interaction are keenly sought after, the benefits of being perceived as ‘one of the good guys’ are incalculable. Moreover, the most reliable way of being perceived to be a good guy is actually to be one – and to be one consistently. Reputations are durable and incredibly fragile: built up over many years, they can be destroyed by a single faux pas. The television evangelist who slept with his secretary is history, as is the political candidate who cheated on his taxes. And imagine the situation of our ancestors, who undoubtedly belonged to the same small community for most of their lives: virtually everything there was to know about them must have been known by virtually everyone. In order to become a respected member of the community, we need to be almost blindingly consistent. To resist every little temptation to stray from the chosen path, we need to believe steadfastly in right and wrong. Only a firm moral conviction ensures behavior that contributes to a lasting reputation of honesty. Indeed, most people reach a point at which the values imbued by society combined with their personal experiences crystallize into a stable pattern of thought and behavior from which they cannot deviate without great personal discomfort. Instead of being led by the reactions of others, or responding to immediate situations, to stay on course they rely on an internal compass, bolstered by strong emotions of guilt and shame.23

Minor acts of beneficence toward strangers can be considered investments in one’s reputation. They function as signals to others that the beneficent person is trustworthy and not interested in exploiting potential cooperators, thereby attracting more opportunities to participate in cooperative ventures. This is one way in which indirect reciprocity can be seen to underlie even such apparently purely altruistic practices as blood donation. Peter Singer thinks that the voluntary practice of blood donation shows that ‘[a]ny theory which entails that non-reciprocal altruism toward strangers cannot occur must be wrong’, since no reciprocity is possible in that case.24 He goes on: These donors are not paid. They do not get preferential treatment when they themselves need blood. … Nor can donors be rewarded – or even given a grateful smile – by the persons whose lives are saved by their gifts. Donors never know who receives their blood, and patients never know who gave the blood they receive.25

But these considerations do not exhaust the possible indirect benefits of blood donation. In fact, ‘giving blood is a fine way of suggesting that one is so altruistic that he is willing to give up the most dear possession for a perfect stranger.’26 Few

Morality as an Evolutionarily Stable Strategy

185

people who donate blood actively disguise this fact from their associates; some even wear pins or buttons indicating their contribution to this cause. Such signalling is all that is required for the (relatively costless) practice of blood donation to have compensating indirect benefits to the donor. True, only a small percentage of people actually admit that they give blood for non-altruistic reasons; but this does not show that its real significance – the ultimate cause by which the practice sustains itself – is not related in any way to self-interest. The point of this observation is not to show that people never do anything altruistic for strangers; the point is that any practice that is widespread and enduring must have some direct or indirect payoff for those involved, although ‘[r]eciprocity is probably never complete, or balanced’.27 As always in nature, proximate mechanisms are approximate mechanisms. Investments in reputation partly explain beneficence toward strangers; but there are other pressures tending toward what might appear to be indiscriminate beneficence based on social investments. Frans de Waal again explains this point nicely: Inasmuch as every member benefits from a unified, cooperative group, one expects them to care about the society they live in and make an effort to improve and strengthen it, similar to the way the spider repairs her web and the beaver maintains the integrity of his dam. Continued infighting, particularly at the top of the hierarchy, may damage everyone’s interests; hence the settlement of conflict is not just a matter of the parties involved, it concerns the community as a whole. I do not necessarily mean that animals make sacrifices for their community, but rather that each and every individual has a stake in the quality of the social environment on which its survival depends. In trying to improve this quality for their own purposes, they help many of their group mates at the same time. An example is arbitration and mediation in disputes; it is standard practice in human society – courts of law serve this function – but recognizable in other primates as well.28

People obviously have a self-interested reason for supporting norms which encourage others to be more beneficent, since they thereby stand to gain by tapping into the flow of benefits thus created. It is not surprising, then, that people would heap honours upon and express admiration for ‘moral heroes’ such as Mother Teresa, praising her high degree of altruism as an ideal. Furthermore, since norms are inherently universal, considerations of consistency pressure people who advocate such beneficence to go some way toward adhering to this norm themselves.29 Some forms of beneficence can also be seen as indirect nepotistic reciprocity – insurance against possible ill-fortune befalling one’s descendants and collateral relatives.30 Even upper middle-class parents cannot be sure that their children or grandchildren will enjoy relatively high social standing, given the fluidity of status in most modern societies.31 Finally, but significantly, beneficence directed at strangers can be seen as a form of opportunity levelling for the sake of social harmony. Reproductive, political and economic opportunity levelling – in the form of monogamy, democracy and the graduated income tax and state-funded education – are seen as just such investments in social harmony.32

186

Liberty, Games and Contracts

After a lengthy examination of apparently altruistic behaviours and social norms which are often appealed to by philosophers who regard self-sacrifice as a sine qua non of morality, Alexander concludes, ‘there do not seem to be any remaining reasons for regarding morality, as normally expressed, as necessarily selfsacrificing, or for invoking anything other than nepotism and reciprocity to account for human social structure.’33 De Waal, too, maintains that while ‘[r]eciprocity can exist without morality; there can be no morality without reciprocity. If we accept this thesis, it is clear why the very first step in the direction of the Golden Rule was made by creatures who began following the reciprocity rule “Do as the other did, and expect the other to do as you did”.’34

Moral Ideals and Improving Morality I have argued above that being evolutionarily stable is a necessary condition for a workable moral system, but it is obviously not a sufficient condition. Human society is just the kind of complex situation in which one would expect to find many ESSs associated with many local optima, arrived at through processes of cultural evolution. Ideally, we should strive to discover that moral system which describes the evolutionarily stable state with the highest possible optimum point, while recognizing the usual constraints on perfection. There is some reason for hope that we might be able to consciously devise moral systems that attain higher equilibria than evolutionary pressures alone have provided for us35 – as long as we first understand that this is what the project for ethical theory is all about. Edward Wilson says, not without irony, ‘Human beings appear to be sufficiently selfish and calculating to be capable of indefinitely greater harmony and homeostasis.’36 Elster notes that human beings are globally maximizing machines, in contrast to evolution which is a locally maximizing machine only.37 While evolutionary pressures alone might trap us in a relatively low local optimum from which any individual, incremental step is detrimental, we are endowed with intelligence and foresight which conceivably allow us to take a coordinated leap from a lower to a higher equilibrium in a single bound. But intelligence and foresight are not sufficient to make the leap in practice; we can make the mental leap to a higher equilibrium much easier than we can make the practical leap. In order to move collectively to a higher equilibrium, enough people must be able to act contrary to a considerable amount of pre-existing moral conditioning, and perhaps even against natural inclination. Indeed, Dawkins suggests what some might view as an even more optimistic scenario when he says, ‘We are built as gene machines and cultured as meme machines, but we have the power to turn against our creators. We, alone on earth, can rebel against the tyranny of the selfish replicators.’38 (We do this in a small way every time we use contraception.) In other words, by virtue of our considerable brain power, human beings might have sufficient autonomy to be able to adopt a morality involving a significant level of unreciprocated self-sacrifice and non-strategic altruism. Such a morality would not be evolutionarily stable, yet it might be sustained in the face of contrary

Morality as an Evolutionarily Stable Strategy

187

evolutionary pressures through sheer force of will on the part of members of society who practise it. It remains to be seen whether conscious psychological conditioning – a regime of moral indoctrination – could effectively check the pressures of biological and cultural evolution which might otherwise undermine a morality involving significant demands for non-strategic altruism.39 It is worth pointing out straight away that the capacities that make humans global maximizers do not automatically make a move toward indiscriminate altruism more likely to succeed, and in fact probably have the opposite effect. The more indiscriminate altruists there are in the population, the more it pays to become a parasite, exploiter or cheat. Given our intelligence and versatility, people will perceive this opportunity immediately and will be able to adapt to it very quickly. Consequently, the spread of parasitism is likely to be especially rapid in human societies – measured in years or decades rather than generations or millennia. Once parasitism reaches a certain threshold, everyone ultimately tends to lose, forcing a reversion to a morality based more squarely on reciprocity.40 What this argument suggests is that indiscriminate or non-strategic altruism is bound to be dysfunctional; it is a false ideal. The upshot is not that beneficence to strangers is always to be discouraged; rather, the lesson is that beneficence must be strategically sound. This means, first, that unreciprocated beneficence must be directed only toward those in genuine need, and only to the extent of their genuine needs. Placing themselves in a position of genuine need merely to take advantage of others’ beneficence is not something rational people have an incentive to do; so by benefiting only the genuinely needy, one is not encouraging exploitation. Strategic beneficence is, second, beneficence which is maximally likely to repay the benefactor in some way. One of the more important effects of beneficence is the effect on the benefactor’s reputation. But this effect only works when beneficence is voluntary; government-mandated beneficence does not enhance anyone’s reputation except, perhaps, the politicians’. It is less strategically sound for that reason. A stable morality need not be based entirely upon reciprocity; but it must be based primarily on reciprocity.

The Function of Morality Functional explanations are commonplace in evolutionary theory; so if morality is a characteristic of human interaction subject to evolutionary pressures, then it is natural to ask what its function is. Confusion arises in the literature on this point because it is common to think of moral systems as an attribute of societies, as in ‘the morality of the !Kung’. From there it is easy to reason that if morality is an evolved feature of society, then it must have a function for society, which is to promote the interests of society. To someone whose philosophical proclivities lean in that direction already, it must be nearly irresistible to conclude that some form or approximation of utilitarianism is the ideal toward which biological and cultural evolution are slowly and haltingly moving.41 But this argument is completely misconceived. It rests upon confusing morality as a social institution with morality

188

Liberty, Games and Contracts

as a strategy employed by individual agents. Insofar as it is a feature of societies, morality does not have a function; and insofar as it has a function, it is not an attribute of societies but of individual agents. Morality is a large cluster of capacities and practices which are far from uniform between the members of even the most cohesive of human societies. No two people in the history of the world have ever felt the same intensity of shame or guilt with respect to the same set of behaviours; no two people have ever been equally beneficent or forgiving, or equally aggressive in the enforcement of their expectations of others; nor have any two person’s expectations ever been identical; and so on. Thus to refer to something as the morality of a society is really only a shorthand, a kind of statistical average of the expectations and behaviours that the group tends on the whole to hold, practise and reinforce. Morality in this sense is a feature of society, but not a trait; it is the result of selective pressures, but not something that has been selected in its own right. In this respect, morality is like a dominance hierarchy in a society of wolves: ‘The dominance-subordination hierarchy shown by wolves and a wide variety of vertebrates and arthropods is not a functional organization. It is the statistical consequence of a compromise made by each individual in its competition for food, mates, and other resources. Each compromise is adaptive, but not the statistical summation.’42 Harems are a feature of many species and some human societies, too. But they are not functional organizations; they, too, are merely the consequence of an interaction of the mating strategies employed by members of the species or society. It makes sense to speak of the dominance and mating behaviours of individual agents as functional adaptations, since these are subject to selection. The collective concatenation is not.43 This is because selection takes place at the level of the replicator, the meme or the gene, not at the level of a population. Morality, biologically speaking, is a strategy or cluster of strategies that an individual adopts, more or less consciously, for the purpose of navigating within a social environment. Insofar as we may meaningfully speak of the morality of this or that society, it is because the individuals composing the society have settled upon an ESS that guides their interactions. Any significant deviation from the mixture of strategies that the society has settled upon, assuming it is an ESS, will be detrimental to the agent adopting it. In other words, deviations from the strategy comprising an evolutionarily stable state are, in some degree, dysfunctional. Stability and functionality are thus interrelated: an evolutionarily stable state is composed entirely of equally functional strategies pursued by the individual members of the group. Note that functionality is relative to pre-existing conditions, that is, to the prevailing mix of strategies in the social environment. If one falls in with ‘a band of ruffians’, as Hume observed, morality as we normally conceive of it goes out of the window and we may have to adopt a ‘kill or be killed’ policy in order to survive. In some social environments, such a code could be evolutionarily stable. Insofar as morality has a function, then, it is an attribute of individual agents, not groups. It is a strategy, or a complex cluster of strategies, that individual agents employ. In general, capacities and practices are favoured by natural selection

Morality as an Evolutionarily Stable Strategy

189

because in a given context they ultimately benefit their underlying replicators, whether they be memes or genes. Evolved capacities and practices typically benefit their underlying memes or genes by benefiting, however indirectly, their ‘vehicles’ – that is, the organism hosting the meme or gene, which in the case of moral capacities and practices are human agents. As outlined in the section of this chapter on indirect reciprocal altruism, moral capacities and practices also typically benefit human agents by benefiting some group to which the agent belongs – society in general, a business firm, a partnership and so on – members of which reciprocate in some way. The complex cluster of capacities and practices we identify as an agent’s morality, in other words, is a strategy whose function is to promote the agent’s interests through realizing the advantages of group living. To put the point even more succinctly, the function of morality is to facilitate mutually advantageous cooperation. By adhering to and reinforcing norms of reciprocity, members of a group establish a moral system which tends to promote their own various interests.44

Morality as an ESS Henry Sidgwick famously asserted that ‘the Cosmos of Duty is … reduced to a Chaos’ if morality and self-interest cannot be resolved.45 The functional theory of morality outlined resolves this ‘dualism of practical reason’. But it might have the defect of opening up too many different ways of resolving that conflict. If the function of morality for any particular individual is to promote that agent’s interests through realizing the advantages of group living, it is conceivable that fundamentally quite different moral systems or principles might be functional for different individuals. Some individuals might benefit from social equality, others from laissez faire; some might benefit from male dominance, others from female privilege; and so on. Individual members of society would therefore compete to have moral norms which favour their particular interests adopted. This cacophony of claims and counter-claims is indeed evident in the intellectual history of ethics,46 and is perfectly predictable given the evolutionary account of morality elucidated here. Moreover, although it is second nature for contemporary philosophers to suppose that morality must be universalizable, an evolutionary account of morality does not automatically support this presumption. An evolutionarily stable morality might conceivably involve a mixture of strategies employed by distinct castes within society, much like the digger wasps noted earlier. The powerful and talented, for example, might conceivably be accorded a different set of rights than the infirm. Will Kymlicka elaborates the point in connection with the mutual advantage contractarian method of moral justification: Whether it is advantageous to follow a particular convention depends on one’s preferences and powers. Those who are strong and talented will do better than those who are weak and infirm, since they have much greater bargaining power. The infirm produce little of benefit to others, and what little they do produce may simply be expropriated by

190

Liberty, Games and Contracts

others without fear of retaliation. Since there is little to gain from co-operation with the infirm, and nothing to fear from retaliation, the strong do not gain from accepting conventions which recognize or protect the interests of the infirm. … The resulting conventions will accord rights to various people, but since these rights depend on one’s bargaining power, mutual advantage contractarianism does ‘not afford each individual an inherent moral status in relation to her fellows’.47

To assess this concern, let us first be clear about the conclusion the argument purports to establish. It is not merely that those who are strong and talented will ‘do better’ than those who are weak and infirm; that will be true in any conceivable human society, since it is simply impossible to equalize life prospects between someone born with serious mental or physical defects and a healthy genius – short of destroying the life of the healthy genius. If the reason some do better than others in a society whose morality is based upon mutual advantage or reciprocity is simply that their talents allow them to make better use of their rights,48 that would not be a very compelling argument against mutual advantage contractarianism. To be nonboring, then, the conclusion of the argument must be that mutual advantage contractarianism accords different people different sets of rights depending upon their threat capabilities. To focus the argument, consider the following rule, which one might naively think to be to the advantage of the strong and talented: ‘Do not steal from or physically harm anyone, unless they are too weak to oppose you or to retaliate’. Note to begin with that even those who are too weak or infirm to oppose or retaliate may still have talents from which the advantaged could benefit. Those who propose to harm or exploit the weak are not likely to get their willing cooperation, and would therefore have to forego these benefits. (Against the suggestion that the advantaged might be able to coerce the disadvantaged into giving them benefits for ‘free’, I note that the costs of such coercion are likely to at least equal any benefits the disadvantaged would be able to give the advantaged.49) Thus the scope of morality includes not only those with a threat capability, but all those who can by their actions deliberately affect another for good or ill. More importantly, though, essentially everyone in a modern state has a significant threat capability. Given our current technological advancement it requires a ridiculously minute amount of physical strength, and even very little mental ability, to kill people. Even those few among the disadvantaged who, individually, may have no significant threat capability against the advantaged, could form a coalition which is capable of mustering a significant threat. Indeed, some strong and talented people may use their advantages on behalf of the disadvantaged, becoming their leaders or representatives and thereby augmenting their position of strength. If the strong and talented do not accept universalizable principles against forceful or opportunistic exploitation, then they cannot legitimately complain when the majority gang up on them and implement a system of taxation which disproportionately expropriates from them. The talented and powerful, on the whole, have more reason than the infirm to denounce the use of opportunism and force majeur, not less. Another lacuna in this criticism of mutual advantage contractarianism is that it

Morality as an Evolutionarily Stable Strategy

191

unrealistically divides the population into two distinct castes, the advantaged and the disadvantaged. In fact, the strongest and the weakest in society are linked by a full spectrum of people in between. The most advantaged in society clearly have reason to adopt principles against harm and theft in their dealings with those just marginally less advantaged, who in turn clearly have reason to adopt principles against harm and theft in their dealings with those just marginally less advantaged still, and so on down to the most disadvantaged. Nobody in this chain, from strongest to weakest, can afford to advocate opportunism lest they open themselves to opportunistic exploitation by those within their range of threat capability. Since these ranges overlap continuously from strongest to weakest, the only practical options are to reject opportunism with respect to everybody or nobody in one’s society or sphere of influence. As before, the powerful and talented have more reason than the infirm to denounce opportunism in favour of universalizable principles, since they have more to lose by opportunism practised against them. In short, mutual advantage contractarianism must accord everyone in society the same rights. Another way to put this is to say that a functional morality must be (at least roughly) an ESS. Collective stability, or universalizability, puts a considerable constraint upon what a functional morality can be. Now reconsider the suggestion that different moral strategies could be functional for different individuals, and in particular that egalitarianism could be functional for poorly placed individuals, and, for some of them, even functional in the long run. That is, egalitarianism could promote their interests better than, say, laissez faire. The problem with this suggestion is that egalitarianism only promotes the interests of poorly placed individuals if others practise it. But of course it is not functional for these others; they have an incentive to abandon egalitarian principles, both so as to exploit the remaining egalitarians and so as to avoid being exploited themselves. And if only poorly placed individuals continue to practise egalitarianism, it is unclear what benefit they might derive from this since, by definition, they are incapable of giving anything to anybody – which is to say, to each other. Because egalitarianism fails the requirement of evolutionary stability, it is dysfunctional even for the poorly placed; it simply does them no good in the end (if not in the beginning). Admittedly, egalitarianism could be functional for the poorly placed if they were able to convince the well placed in society that egalitarianism was required of them; but a correct understanding of the function of morality tells them that this is not so.50 Before leaving the Kymlicka critique, I should also question the underlying assumption that the powerful and talented have a general interest in being petty thugs with respect to the destitute. It is entirely unclear what Kymlicka supposes the powerful and talented have to gain from kicking the infirm while they are down; perhaps he thinks that strong, intelligent, talented people are inherently sadistic. This assumption is, of course, completely unrealistic. But suppose, for the sake of argument, that human nature were as dark as Kymlicka’s argument seems to require; or better, suppose we were to encounter powerful beings from another planet, of whom this is true. In that case, we may expect to be fairly roughly treated, perhaps as roughly as humans have historically treated the animals we use for food

192

Liberty, Games and Contracts

or sport. Since ‘everyday morality’ does not unequivocally condemn the way we use other animals, it is unclear why it should condemn the way space aliens might use us, were we to find ourselves in the same power relation vis-à-vis them as animals do vis-à-vis us. If we have contrary intuitions, this is no doubt because we are completely unfamiliar with and unadapted to the circumstances in which a radically different morality, or no morality, would be the appropriate response. Rejecting a method of moral justification because it has counter-intuitive implications about radically counterfactual situations is to place altogether too much weight upon pre-critical intuitions. There is one important, non-counterfactual class of beings which we generally recognize as having independent rights, yet which have no threat capabilities at all: completely helpless human infants. Does mutual-advantage contractarianism imply that we are mistaken in according rights to infants? It seems inadequate to say that the rights of children are really derivative of the proprietary rights of their parents, since sometimes these rights are asserted against parents, for example as a result of parental neglect and abuse. The powerful sense of obligation which parents owe to their children is universal in human cultures, and presumably arises to facilitate kin altruism. The question is why we regard parental obligations (and the rights of children correlative to these obligations) as important parts of social morality, rather than being left as private, intrafamilial concerns. The answer is that, while infants and children may have no threat capabilities as such, the potential harm that maladjusted children can do to society when they become adults is obviously enormous.51 Society therefore has a pressing interest in making sure that children are well looked after by their parents; and indeed, social investments in the welfare of children are universal and significant. Thus a functional contractarianism is fully capable of accounting for children’s rights.

Morality and Perfectionism Finally, it bears emphasizing that to say that morality based on mutual advantage is a functional ESS – that it serves the interests of the individual agents who practise it better than any other set of interpersonal principles – is not to say that it is perfect. We may safely assume that, even though a functional morality is the best long-run strategy for each and every person, people will not always behave morally, since they are not always rational and they sometimes fall to temptation. Moreover, human agents can never perfectly identify possible deviants, since human beings are not completely transparent and human interactions are typically not completely unambiguous. Consequently, a certain amount of deception, manipulation, theft and physical aggression is always going to be possible in any real-life society. That is why a functional morality must be aggressively enforced; or, more accurately, a functional morality includes norms with respect to active enforcement. But the active enforcement of morality entails costs of its own, and functionality requires that we weigh the marginal costs against the marginal benefits of enforcement efforts. It is not to be expected that functional norms with respect to

Morality as an Evolutionarily Stable Strategy

193

enforcement will require ‘zero tolerance’ of deviancy, since enforcement of a zerotolerance norm would be extremely costly to implement. This argument merely notes the standard constraints on perfection. Its implication is simply that the best we can realistically hope for is a society in which moral deviance is kept to the margins, not eliminated completely. There will always be an ‘ecological niche’ for crime to make a living in. In other words, some moral deviancy can be expected to invade a society based on a functional morality; as a practical matter, a functional morality can never be absolutely evolutionarily stable. Since this is obviously true of any moral system whatsoever, it does not constitute a criticism of a functional, mutual-advantage contractarian morality in particular.

Notes 1

2 3 4

5

6

7 8

9

See, for example, Jan Narveson, ‘The How and Why of Universalizability’, in Nelson Potter and Mark Timmons (eds), Morality and Universality (Dordrecht, 1985), 3–44; and John Mackie, ‘The Three Stages of Universalization’, in Joan and Penelope Mackie (eds), Persons and Values: Selected Papers, Volume II (Oxford, 1985), 170–83. The example comes from Richard Dawkins, The Selfish Gene (Oxford, 1976). I reference the expanded 1989 reprinting, 283–4. John Maynard Smith, Evolution and the Theory of Games (Cambridge, 1982), 10. Each evolutionarily stable state describes a local optimum; but since evolution works by incremental steps it cannot always bridge a chasm to get to the global optimum from a local optimum. See Jon Elster, Ulysses and the Sirens: Studies in Rationality and Irrationality (Cambridge, 1979); and Richard Dawkins, Climbing Mount Improbable (New York, 1996). There are many possible variations on the basic TFT theme – for example, the more forgiving tit-for-two-tats – but these nuances do not concern us here. Robert Axelrod, in The Evolution of Cooperation (New York, 1984), identifies the key features of a robust reciprocal altruism as: (i) never being the first to defect from the cooperative option, (ii) being quick to repay defection with defection and (iii) being equally quick to forgive when the other returns to cooperation. For simplicity, I refer to a gene or a ‘meme’ (Dawkins, The Selfish Gene) which inclines an organism toward reciprocal altruism in any of its variations as a ‘TFT allele’ or a ‘TFT mutation’. R. Boyd and J. Lorberbaum, ‘No Pure Strategy is Evolutionarily Stable in the Repeated Prisoner’s Dilemma Game’, Nature, 327 (1987): 58–9; R. Axelrod and D. Dion, ‘The further evolution of cooperation’, Science, 211 (1988): 1390–6; Louis Marinoff, ‘The Inapplicability of Evolutionarily Stable Strategy to the Prisoner’s Dilemma’, The British Journal of Philosophy of Science, 41 (1990): 461–72; and J. Lorberbaum, ‘No Strategy is Evolutionarily Stable in the Repeated Prisoner’s Dilemma’, Journal of Theoretical Biology, 168 (1994): 117–30. Louis Marinoff, ‘Maximizing Expected Utilities in the Prisoner’s Dilemma’, Journal of Conflict Resolution, 36 (1992): 183–216. MAC is characterized by first playing a representative initial sample of iterations by randomizing between cooperation and defect at a ratio of 9:1, and then playing the remaining iterations of an interaction on the basis of maximizing expected utility, given the expectations about the other’s play which are derived from the initial trial period. Marinoff, ‘Maximizing Expected Utilities’, 216.

194 10

11 12 13 14

15 16 17 18

19

20

21

22 23 24 25 26 27 28 29 30

Liberty, Games and Contracts The reason why being totally unforgiving in the real world does not work so well is that people sometimes make honest mistakes or honestly misunderstand the agreement or the nature of the situation they face. It doesn’t make sense to ruin a fairly reliable longterm relationship over one (potentially mistaken) ‘defection’. Robert Nozick, The Nature of Rationality (Princeton, 1993). Robert Frank, Passions Within Reason: The Strategic Role of Emotions (New York, 1988). Nozick, 11. Louis Marinoff, ‘The Failure of Success: Intrafamilial Exploitation in the Prisoner’s Dilemma’, in Peter Danielson (ed.), Modeling Rationality, Morality, and Evolution, Vancouver Studies in Cognitive Science, vol. 7 (Oxford, 1998), 161–84. Dawkins, The Selfish Gene, 216–17. Matt Ridley, The Origins of Virtue (New York, 1996). Adam Smith, The Wealth of Nations (New York, [1776] 1910). Richard Alexander, The Biology of Moral Systems (Berlin, 1987), 163. Alexander (xxi) also quotes Robert Bigelow: ‘A hydrogen bomb is an example of mankind’s enormous capacity for friendly cooperation. Its construction requires an intricate network of human teams, all working with single-minded devotion toward a common goal. Let us pause and savour the glow of self-congratulation we deserve for belonging to such an intelligent and sociable species.’ Robert Bigelow, The Dawn Warriors: Man’s Evolution toward Peace (New York, 1969). Alexander says that the ‘problem of culture’ is not so much that it takes over from and operates independently of our evolutionary heritage, but rather that it leads to a failure to appreciate the pervasiveness and the consequences of indirect reciprocity (Alexander, 95). The case where C observes A help B and then helps A could be considered either an investment by A in his reputation, or as a social investment by C in positively reinforcing helping behaviour. Frans de Waal refers to this form of indirect reciprocity as ‘community concern’, which he defines as ‘the stake each individual has in promoting those characteristics of the community or group that increase the benefits derived from living in it by that individual and its kin.’ He goes on to say, ‘social animals have been selected to inhibit actions that may destroy group harmony and to strive for an optimal balance between peaceful coexistence and the pursuit of private interests. Whether animals realize how their behavior impacts the group as a whole is not critical for the evolution of community concern.’ Frans de Waal, Good Natured: The Origins of Right and Wrong in Humans and Other Animals (Cambridge, MA, 1996), 207. Alexander, ch. 3. de Waal, 115. Peter Singer, The Expanding Circle: Ethics and Sociobiology (Oxford, 1983), 133f. Ibid. Alexander, 157. Alexander, 81. de Waal, 31. Allan Gibbard, Wise Choices, Apt Feelings: A Theory of Normative Judgment (Cambridge, MA, 1990). An example of a good social investment which involves indirect nepotistic altruism is publicly funded (though not necessarily administered) primary and secondary education. Everyone benefits from living in a society of well-educated individuals, and

Morality as an Evolutionarily Stable Strategy

31 32 33 34 35 36 37 38 39 40

41

42 43

44

45 46 47 48 49

195

even if it is too late for me to benefit personally from the education my taxes in retirement buy for others’ children, they may still be sound social investments in the well-being of my children or grandchildren, who will have to interact with the otherwise uneducated children of others who are too poor to pay for their education. Alexander, 189. Alexander, 71, 187–8. Alexander, 195. de Waal, 136. Brian Skyrms, Evolution of the Social Contract (Cambridge, 1996). Edward O. Wilson, On Human Nature (Cambridge, MA, 1978), 157. Elster, Ulysses and the Sirens, 9. Dawkins, Selfish Gene, 201. I argue elsewhere (Grant Brown, ‘Functional Libertarianism’ (DPhil thesis, Oxford University, 1997)) that we cannot be sanguine about the prospect. This point is a central theme of Mackie (‘Three Stages of Universalization’, 120–31 and 152–69) and Gauthier (Morals By Agreement, ch. 6). Charles Murray (Losing Ground: American Social Policy 1950–1980 (New York, 1984)) provides a compelling empirical investigation of the corrosive and counter-productive effects of the American ‘war on poverty’. Peter Singer (The Expanding Circle) and Robert Wright (The Moral Animal (New York, 1994)) advocate utilitarianism as being most compatible with an evolutionary understanding of the basis of morality. George Williams, Adaptation and Natural Selection: A Critique of Some Current Evolutionary Thought (Princeton, 1966), 218. This is not to say that no feature of a society can possibly be explained in terms of its function for that society. One might conceivably explain, for example, the Catholic Church in medieval Europe in terms of its role in suppressing dissent within and harmonizing interests between diverse principalities; and the Chicago school of economists’ attempt to explain profit-maximizing firms in terms of their being selected by the market (cf. Elster, Ulysses and the Sirens). The causal feedback aspect of a morality based primarily on norms of reciprocity is developed well by David Schmidtz, The Limits of Government: An Essay on the Public Goods Argument (Boulder, CO, 1991), ch. 7. Not coincidentally, he calls the theory he outlines ‘the feedback theory of morality’. He does not offer a fully fledged functional explanation of morality, however. I argue in my DPhil thesis, ‘Functional Libertarianism’, that the natural history or evolution of morality is essentially a trialand-error version of the ‘mutual-advantage contractarian method of moral justification’. See John Mackie, ‘Sidgwick’s Pessimism’, in Joan and Penelope Mackie (eds), Persons and Values: Selected Papers, Volume II (Oxford, 1985), 77–90. Alasdair MacIntyre, After Virtue: A Study in Moral Theory, 2nd edn (Notre Dame, NY, 1985). Will Kymlicka, Contemporary Political Philosophy (Oxford, 1990), 128 [footnote omitted]. Loren Lomasky, Persons, Rights and the Moral Community (Oxford, 1987), 76–7. I believe that the costly legacies of slavery and the mistreatment of aboriginal peoples in, for example, the United States, bear out this claim, although I acknowledge that the matter is controversial. The benefits the early European-Americans derived from the use of slavery and the outright confiscation of lands occupied by North American

196

50

51

Liberty, Games and Contracts Indians were small, I believe, relative to what they could have attained anyway through honest negotiation. After all, Europeans possessed considerable technological benefits, which would have cost them relatively little to share, with which to bargain for land and labour. Against the small gains of expropriation of land and labour – even supposing that the use of force entailed gains over negotiated, peaceful arrangements – are a history of war and violence, reparations and entitlements, and wasted human potential which has redounded upon subsequent generations of European-Americans. Nietzsche criticized the Christian morality of his milieu because he saw it as a means by which the weak manipulated the strong and talented into serving the interests of the weak, to the detriment of the flourishing of genius and accomplishment. Nietzsche has a valid point here; but the implications he draws from his critique are just as mistaken as the Christian morality he criticized, for he fails to recognize the functionality of indirect reciprocity in the form of social investments and community concern. Jan Narveson, The Libertarian Idea (Philadelphia, 1988), 269f.

Chapter 14

Reconciling Radicals: Market Contractarianism and Fundamentalist Utilitarianism Sheldon Wein

Introduction Jan Narveson began his career as a utilitarian, but more recently he has been a powerful advocate of a form of contractarianism which he takes to provide the best available foundation for a practicable form of libertarianism or, as he prefers to call it, radical liberalism. In this chapter I explore the extent to which utilitarianism and contractarianism are compatible. I begin by distinguishing between two types of utilitarians: fundamentalist utilitarians and quasi-utilitarians.1 I then look briefly at two attitudes one might have toward the idea of a social contract and sketch two varieties of contractarian political theories: what I call social contractarianism and what I call market contractarianism.2 Reconciling versions of quasi-utilitarianism with versions of social contractarianism is a different project from reconciling versions of fundamentalist utilitarianism with versions of market contractarianism. Furthermore, the task of reconciling social contractarianism with fundamentalist utilitarianism is less likely to succeed and be stable. The prevailing consensus about these two moral theories is, upon examination, shown to be false. The similarities between social contractarians and utilitarians are primarily superficial ones, while at a deep structural level the two accounts differ substantially. Furthermore, when we compare market contractarianism with utilitarianism we find that the differences are primarily surface ones, and the similarities are really quite basic. This leads me to think that the frequently sought-after reconciliation between social contractarianism and utilitarianism is a vain hope, but that the rarely sought-after reconciliation between market contractarianism and utilitarianism is worth pursuing. Utilitarianism For purposes of this chapter, someone is a fundamentalist utilitarian if, and only if, she takes the principle of utility to be the basic principle of moral evaluation from which all other injunctions derive. That is to say, one is a fundamentalist utilitarian if, and only if, one takes the principle of utility to be the fundamental principle which one endorses and the endorsement or evaluation of other things – moral 197

198

Liberty, Games and Contracts

rules, possible worlds, institutions, laws, religions, acts, character traits and whatever else – flows from that fundamental endorsement. As long as the principle of utility is what one ultimately always reasons from (at least for ethical issues), one is a fundamentalist utilitarian. Of course, fundamentalist utilitarians disagree about a whole host of things. Among them are the following: (i) (ii)

(iii) (iv)

(v)

(vi)

Is the principle of utility meant to be used as a decision procedure or as a standard of evaluation? Is utility simply a rigid designator for whatever it is (in all possible worlds? in just those possible worlds accessible to us?) that is intrinsically valuable? Or is utility the satisfaction of (actual or hypothetically well-informed) desires? Is utility a kind of conscious experience (or set of such experiences)? A function of preferences? Or is it some other aspect of the well-being of (some subset of) living or conscious or acting things? To what extent, and in what ways, are intra- and interpersonal comparisons of utility possible? To what extent are they practicable? Does the principle of utility recommend maximizing the total net amount of utility (classical utilitarianism), or does it recommend those social arrangements which secure the highest average level of utility (average utilitarianism)? For purposes of morality, what is the proper object of the principle’s application? (Acts? Rules? Traits of character? Institutional arrangements? Possible worlds? All of these?) Is the principle of utility true? Is it objectively true? (And is ‘true’ to be understood here in some realist/correspondence sense or in, say, a quasirealist minimalist sense?) Or is the principle simply the appropriate object for endorsement? Or universal prescription? Or emotive attitude? Or expressive commitment? Or …

What makes people who disagree over such things utilitarians is their belief in or commitment to the idea that such things are worth debating because they have to do with the best way of applying what they all take to be the sole fundamental moral principle, the principle of utility. Of course, many non-utilitarians are also interested in the issues listed above. Some are simply because they find such problems intrinsically interesting, or because they have something to contribute to other people’s moral theories, or because part of their own moral theory has a utilitarian element. (They use the principle of utility in a restricted realm, or to break ties, or when dealing with nonpersons, or for some other limited purpose.) But there are other philosophers who are normally called utilitarians who are not fundamentalist utilitarians. Here I am thinking of people who come to endorse the principle of utility (or hold that it is true) because it follows from some other more fundamental principle or is the product of some more cherished device or procedure for ascertaining moral principles. Thus, those who think that the principle of utility is the basic principle of morality solely or primarily because it is the only principle which it would be

Reconciling Radicals

199

rational to choose – the only principle which, were one in the equiprobability model, would maximize the chooser’s expected utility, or because an ideal observer would endorse the principle, or because the logic of the moral language game relentlessly drives one to the principle, or because the principle neatly encapsulates what an impartial spectator or archangel would favour and what it would find disgusting, or because the categorical imperative or the general will requires adherence to the principle – these people are not fundamentalist utilitarians. Let us instead call them quasi-utilitarians. Quasi-utilitarians disagree amongst themselves and with fundamentalists (and others) regarding all the issues on my short list above. Now, obviously, there are degrees of fundamentalism. Bentham was an extreme fundamentalist and so, I think, is J. J. C. Smart. Both of the Mills, as well as Sidgwick and Brandt, while definitely fundamentalist, were less so than Bentham or Smart. On the other hand, John Harsanyi, Jan Narveson (when he was still a utilitarian) and Adam Smith were all quasi-utilitarians.3 Some people are difficult to classify. If you think that, had R. M. Hare (mistakenly) come to believe that the logic of ‘ought’ did not lead to the principle of utility but to some alternative principle he would have adopted the alternative principle rather than give up his methodology, then you hold that Hare is a quasiutilitarian rather than a fundamentalist utilitarian. If, on the other hand, you read Hare as offering accounts of the logic of the moral language game, of the proper understanding of universal prescriptivism, of the distinction between levels of moral thought, of the attitudes of an archangel, and so forth only as means of showing others that the principle of utility – which he already accepts, for its own sake, is worthy of its status as the fundamental moral principle – then you think of Hare as a fundamentalist utilitarian. The same is true for those utilitarians who adopted the principle of utility because they can think of no other moral principle which comes as close to being in reflective equilibrium as the principle of utility. Thus, if your considered judgments about particular moral issues are utilitarian (suppose you have read a lot of Sidgwick and Brandt but never any McCloskey or Williams) and if you would abandon the principle if, say, you happened across a copy of the Nicomachean Ethics, then you are a quasi-utilitarian.4 If, however, your considered judgments are already pretty much informed by consequentialist considerations and if, when they conflict with whatever is your presently most-favoured version of the principle of utility, your inclination is first to abandon your considered judgments and (if that fails) seek some alternative version of the principle of utility, then you are a fundamentalist utilitarian. If, upon hearing about innocent-man problems or Bernard Williams’s Jim, you are first inclined to say that the sheriff should frame the innocent man and that Jim should shoot the peasant, and if out of squeamishness (as you think it appropriate to characterize it) you can’t bring yourself to stomach those outcomes and so you consider moving from your version of utilitarianism to some version of rule or motive or cooperative or character utilitarianism, then you are not a fundamentalist utilitarian.5

200

Liberty, Games and Contracts

The Logical Structure of Contractarian Theories Contractarian theories are primarily theories of justice, and I will confine my discussion of them to matters of justice. Every contractarian theory holds that the basic principles of justice are those principles which, if they are adopted by all or most members of society, will lead to each being better off than they would be without the acceptance of those principles. To aid in the formulation of the principles having this property, the contract, or bargaining game, is employed. An initial choice situation – the so-called ‘state of nature’ – is postulated, where rational individuals, each of whom seeks to further her own ends and who, in so doing, recognizes no social rules which might constrain that pursuit, are asked what set of principles each will accept, providing that others accept those principles also. Individuals in this initial choice situation reason together (using only appropriate information and principles of rational choice) to adopt principles to govern their future interaction. The agreed-upon principles are the basic principles of justice. Thus, every contractarian theory consists of four parts. First, there is the specification of the initial choice situation. What is the appropriate state of nature or original position from which the basic principles governing social interaction are to be chosen? How much do individuals know? What sorts of desires and/or life-plans do they have? Second, there is the defence and employment of particular decisiontheoretic principles of rational choice used to derive principles acceptable to rational maximizers bargaining from the initial choice situation.6 That is to say, there is an account of how individuals in the initial choice situation choose the principles of justice the theory holds they would choose. Third, there are the principles of justice chosen by the rational individuals in the preferred initial choice situation. That is to say, there is an account of the content of the principles chosen, complete with some guidance on how such principles should be applied to actual problems, such as the design of a constitution or the structuring of an economy. Finally, there is the justification of why the principles chosen in this way have any moral significance. That is to say, there is an account of why thinking about justice in this way is a wise way to proceed. Different versions of contractarianism give different answers to these questions.7

Social Contractarianism The best-known version of contractarianism is what I will call social contractarianism.8 Social contractarianism was first, and in many ways most eloquently, stated by Jean Jacques Rousseau in The Social Contract (1762). The theory was restated, in what may well have been its most powerful and almost certainly its most difficult form, by Immanuel Kant. The theory has regained wide popularity thanks to the pioneering work of John Rawls. In what follows I will use Rawls’s version of social contractarianism, in particular what he calls the ‘Kantian interpretation’, as the basis for my discussion.9 In social contractarianism the first and second parts of the theory, namely the

Reconciling Radicals

201

specification of the initial choice situation and the selection of a rational decision rule, are both determined by moral considerations. In Rawls’s case the moral consideration used for this purpose is a rich conception of fairness.10 Rawls’s original position, where contractors select principles of justice behind a veil of ignorance so that, during the selection process, no contractor knows which member of society she is, and Rawls’s maximin rule, which holds that one should concentrate on the worst outcomes when making choices about social arrangements, are both designed to ensure that the principles chosen are fair. Because no contractor knows who she is, the choice problem is the same for all of them. Hence, the selection of principles of justice from the original position requires the contractors to agree on a principle of individual choice, namely the maximin rule. The contractors have no need to consider a bargaining strategy because not knowing which member of society one is makes it the case that all have identical interests.11 The third part of the theory is an account of the principles which Rawls thinks his contractors, using the maximin rule, would choose as their two basic principles of justice. These are the principle of liberty, which holds that society should be arranged so that each person is provided with the widest possible system of liberty compatible with the same liberty for all others, and the difference principle, which holds that, in a system of equality of opportunity where valued places in society are open to all, the economic system should be arranged so that those who are among the worst off are as well off as is possible. It is these two principles which, Rawls argues, would be chosen over other plausible alternatives by the contractors in the original position using the maximin rule.12 Finally, there is the account of why we should care about this, why the chosen principles have any moral import. Here we have several answers open to us. Ronald Dworkin holds that we care because we suppose that each person has a natural right to equal concern and respect.13 On the Kantian interpretation, we care about all this because autonomy and respect for the dignity of free and equal rational individuals are our most cherished values and the theory is, in effect, a good way to apply the categorical imperative to questions about the basic structure of society.14 And, on the pragmatic interpretation, we care because political philosophy, like all of politics, is the art of the possible; and Rawls’s theory tells us what members of a pluralist society marked only by an overlapping consensus on moral issues can agree on.15

Market Contractarianism Market contractarianism is the most developed alternative to social contractarianism within the contractarian camp.16 It is actually older than social contractarianism, having been first stated by Plato in his Republic.17 But Plato’s purpose was to show that the theory was mistaken. The first serious defender of market contractarianism was Thomas Hobbes,18 it has been suggested that David Hume held a theory similar to it.19 In the work of David Gauthier it finds its fullest and most sophisticated statement. Jan Narveson has used market contractarianism to develop and defend a powerful version of Nozickean libertarianism.

202

Liberty, Games and Contracts

The market contractarian holds that the great strength of contractarianism is that it offers the potential of uniting rationality with morality. Market contractarians hope to do this in a particularly rigorous way by producing amoral reasons for convincing individuals to become moral persons. This is because, by stipulation, the principles agreed upon in the pre-contract situation are in the individual amoral interest of each and, therefore, it is instrumentally rational for each to accept them. Since the accepted principles are moral principles, market contractarianism shows that it is (amorally) rational to be, or to become, moral. Thus, the market contractarian holds that the hope of uniting morality with rationality in this rigorous way is lost if one imports moral considerations when one is specifying the initial choice situation. It is for this reason that the social contractarianism of Rawls, which uses moral considerations – those found in his rich conception of fairness – as a guide in specifying the initial choice situation, is rejected. The market contractarian’s claim is that if morality is to be truly grounded on rationality, one cannot begin by including moral considerations in the pre-contract situation and then, from this ‘morally tainted’ state of nature, use rationality to derive principles of justice. But if moral considerations are not to guide the market contractarian in specifying the initial choice situation, how is she to determine which pre-contract situation is the one from which to begin deriving the basic principles of justice? The market contractarian answer is that rationality alone must be used to determine the proper description of the state of nature. All contractarian theories hold that the principles of justice are those which it is rational for each to accept in some precontract situation. What is distinctive about market contractarianism is that it not only holds that rationality must be used to derive principles of justice once the initial choice situation has been properly characterized; it also holds that rationality must guide one in specifying the position from which the principles of justice are to be derived. Not only must it be rational to contract into civil society from the state of nature, but the state of nature must be specified so that it is rational to begin bargaining from that position. Rawls holds that, if we are to be fair, the initial choice situation must be specified in a manner such that natural attributes do not influence the outcome of the choice. Thus, in Rawls’s initial choice situation the contractors do not know their natural attributes. Yet while the original position may well represent the contractors in a way that is fair to each, the market contractor cannot avail herself of this conceptual device because instrumental rationality does not sanction doing so. It would be instrumentally irrational for those who are well-endowed with natural attributes – and who therefore have good reason to expect that they will do well in the bargaining over what type of civil society to set up – to begin bargaining from a position like Rawls’s original position. It would be irrational for the strong to agree to begin bargaining from any position which deprived them of the full use of their natural attributes. Rationality requires the market contractarian to represent the contractors in the initial choice situation with all their natural attributes; and this is just what all contemporary market contractarians do. Their solution to the specification problem represents the contractors just as they are – with their natural

Reconciling Radicals

203

attributes; rationality requires that each contractor be represented as accurately as possible. (Unlike social contractarians, market contractarians hold that the contractors should be given accurate information about the state of the world’s non-human resources and the levels of knowledge and technology available to the contractors.) Among the most important theoretical consequences of this feature of market contractarianism is that the derivation of principles of justice from the original choice position is not, as in social contractarianism, a problem in individual choice theory. Rather, because the contractors in a market contractarian theory must bargain with each other over issues such as how to preserve individual security and how to divide the social surplus appropriately, the derivation problem is a problem in game theory. The contractors are thought of as confronting a prisoner’s dilemma. That is to say, they are confronted with a situation where the best outcome for each individual obtains when everyone else cooperates in sharing the costs and benefits of social cooperation while that individual accepts none of the costs but takes as many benefits as she can. For simplicity, assume that there are only four pure outcomes: solo defection, universal cooperation, universal defection and solo cooperation. The second-best outcome is universal cooperation. The third-best (or second-worst) outcome is simply a complete failure to cooperate, which results, in effect, in the contractors remaining in the state of nature where there are no moral constraints on individual utility-maximizing behaviour and, consequently, life is solitary, poor, nasty, brutish and short. The worst outcome for any individual is to be the only person who puts any effort into social cooperation. This is because the lone cooperator bears all the costs of the (failed) attempt at social cooperation and, because the cooperative venture fails, reaps no benefits. A major problem for every market contractarian theory is why anyone should cooperate in bearing the costs of producing civil society. It is also a major virtue of market contractarian theories that they tackle this issue and attempt to solve the age-old question, why be moral? Obviously, if you are going to provide me with reasons which I am likely to accept regarding why I should constrain my actions so that they conform to a proposed moral code, telling me that I would have agreed from behind a veil of ignorance to accept such constraints on my freedom is unlikely to get you very far. (Of course, if you have antecedently convinced me that I should treat people with dignity and respect or as free and equal rational beings, then such an argument may appeal to me. But if I lack such Kantian commitments, then I am likely to reply that, while someone very ignorant of who they are might agree to such a loss of freedom, that hardly gives me – someone who does know who he is – a reason for accepting that agreement.) Considerations such as these lead the market contractarian to represent the contractors in the initial choice situation as they actually are – or rather, as they would be in the absence of whatever distortions are directly caused by current or previous social arrangements. It is this feature of market contractarian theory which seems to lead it so far from fundamentalist utilitarianism. It is this gulf – that between what Sidgwick called the method of rational egoism and the method of utilitarianism – which I wish to explore. For, as Sidgwick recognized, though there is a big gulf between these two approaches or ‘methods’, they share much in common.20

204

Liberty, Games and Contracts

First, we need a sketch of a particular version of market contractarianism. As in the case of social contractarianism, I will take the best-known and most popular contemporary version as the starting point for my discussion, viz., the market contractarianism of David Gauthier as presented in his Morals by Agreement.21 Gauthier’s theory begins with instrumentally rational self-interested individuals, each of whom knows who she is and each of whom seeks to rationally further her own interests. Each of them knows the principles of game theory, that everyone seeks to maximize her own utility, and that each is non-tuistic.22 They find themselves in a (one-shot) prisoner’s dilemma and, initially at least, it seems that the most rational thing for each to do is to defect from any proposed agreement to form a civil society. That is to say, the situation is such that each thinks it is not in her interest to cooperate in the joint cooperative venture needed to form civil society. Each is, to use Gauthier’s terminology, a straightforward maximizer and, because defecting is the dominant strategy in a prisoner’s dilemma game, each will defect from any proposed social contract. In addition, each knows this fact about the others, and knows that the others know it about her. Gauthier then shows – in what is widely regarded as his most important contribution to the development of market contractarianism – that if the contractors in his initial choice situation changed their conception of rationality from straightforward maximization to what he calls constrained maximization, they would each be better off.23 Consequently, it is rational for a straightforward maximizer to become a constrained maximizer. (Roughly, a constrained maximizer behaves just like a straightforward maximizer except when she finds herself in a prisoner’s dilemma with another constrained maximizer, in which case she cooperates.) Gauthier then argues that constrained maximizers will adopt (and abide by) a very abstract principle called the principle of maximin relative benefit/minimax relative concession (MRC). MRC tells us to agree to the feasible option that minimizes the maximin relative concession that anyone makes, where the relative concession that a person makes for a given option is the ratio of (a) [the excess of (i) the utility for that person of her most favoured option over (ii) the utility for that person of the given option] to (b) [the excess of (i) the utility for that person of her most favoured option over (ii) the utility for that person of the no-agreement option]. The intuitive idea here is that, since the strength of a contractor’s reasons against accepting a proposed contract can be measured by her relative concession under the contract, minimizing maximum relative concession (MRC) also minimizes the grounds any contractor has for complaint. Thus, the contract which MRC picks out is one which gives each person the weakest grounds for objecting to it (given that others also have grounds for objection).24 Finally, the reason we should care about this is that it gives us a non-moral reason for joining in cooperative ventures to produce civil society (whenever we have adequate reason to think others will join us in those ventures); and it gives us an account, which it is rational for each of us to accept, of how to distribute the social surplus produced by such ventures. So, put bluntly, it shows that it is in our interest to make the sorts of sacrifice necessary to create a decent society, and it tells us what we can legitimately expect as the appropriate payoff for making such efforts.

Reconciling Radicals

205

Drawing Parallels It is, I think, fairly easy to see that quasi-utilitarians and social contractarians have a lot in common. Indeed, this likely explains why it is frequently claimed that Rawls is, after all, something of a utilitarian.25 First, both kinds of theories are fundamentally deontological. A lot of ink has been (usefully) spilled on how the distinction between teleological and deontological moral theories should be drawn. It is now standard practice to distinguish between such theories by referring to how the right and the good are defined. Teleological theories define the right in terms of the good, and deontological theories define the good in terms of the right.26 For teleologists the ‘good’ is the logically most-primitive ethical term; and for deontologists the ‘right’ is logically prior to, or at least on a par with, other ethical concepts. Utilitarianism is usually thought of as a paradigm case of a teleological theory, while Kantian theories are paradigms of deontological theories. Thus, in the Critique of Practical Reason, Kant says that, ‘the concept of good and evil is not defined prior to the moral law … rather the concept of good and evil must be defined after and by means of the law’.27 Rawls offers a somewhat less strict account of deontological theories, saying, utilitarianism is a teleological theory whereas justice as fairness is not. By definition, then, the latter is a deontological theory, one that either does not specify the good independently from the right, or does not interpret the right as maximizing the good. (It should be noted that deonotological theories are defined as non-teleological ones, not as views that characterize the rightness of institutions and acts independently from their consequences. All ethical doctrines worth our attention take consequences into account in judging rightness. One which did not would simply be irrational, crazy.28

Bentham puts the teleological analysis of deontological terms in a characteristically brusque manner, saying: Of an action that is conformable to the principle of utility one may always say either that it is one that ought to be done, or at least that it is not one that ought not to be done. One may say also, that it is right it should be done; at least that it is not wrong it should be done: that it is a right action; at least that it is not a wrong action. When thus interpreted, the words ought, and right and wrong, and others of that stamp [that is, presumably, other deontological terms] have a meaning; when otherwise, they have none.29

Given this way of drawing the distinction between deontological and teleological moral theories, social contractarianism and even quasi-utilitarianism can be understood as deontological theories, while market contractarianism and fundamentalist utilitarianism are teleological theories. Quasi-utilitarian theories may look teleological, but upon reflection they are shown to be deontological. For the quasi-utilitarian, what makes an action a right one is that it maximizes the good, which is characteristic of teleological theories. But the fuller definition of right action is always going to be something like ‘that action which is derived from whatever principles rational individuals would choose in the equiprobability model using the maximize expected utility principle’; or ‘that action which is in line with

206

Liberty, Games and Contracts

those rules and social virtues which the impartial spectator would endorse’; or ‘that action which a suitably informed appropriately careful universalizer would prescribe’. All these ways of defining the right do so independently of the notion of maximizing the good (though, of course, properly understood they are coextensional with ‘actions that maximize the good’). Fundamentalist utilitarians and market contractarians both define the right in terms of the good. They differ only in that utilitarians hold that the right is that which maximizes the good however the good is distributed, whereas market contractarians hold that the right is that action which enhances the good in such a way that it is rational to suppose that each member will be better off by a system of such actions. Put another way, the utilitarian thinks that we ought to have those moral rules and principles of justice which are in the interests of each taken collectively, and the contractarian thinks that we ought to have those moral rules and principles of justice which are in the interests of each taken severally. It is perhaps for this reason that, for many forms of utilitarianism, one can find a corresponding form of market contractarianism. Thus, Duncan MacIntosh seems to favour what one might reasonably call act market contractarianism; that is, one should change one’s preferences over outcomes when so doing will make it instrumentally rational to ‘cooperate’ with others because that is what one most wants to do.30 Edward McClennan and Joe Mintoff seem to favour rule market contractarianism; that is, we should resolve to act in counter-maximizing ways when such resolute choice will encourage cooperative outcomes which are Pareto superior to non-cooperative ones.31 Peter Danielson’s version of market contractarianism has us programming ourselves to contingently cooperate with like-minded folks when we are in environments supporting such cooperation.32 And Eric Cave has recently defended what we might call character or virtue market contractarianism: one should cultivate a habit of contingent cooperation until cooperating becomes part of one’s character.33 Finally, let me note one more characteristic which fundamentalist utilitarianism and market contractarianism share: they are both, potentially at least, ethics of care. This makes them particularly appropriate as potential theoretical grounds for feminists who – in the face of the recent collapse of the postmodernist/ deconstructionist movement – seek solid moral theories to ground their ethics of care. Of course, one can use quasi-utilitarianism or social contractarianism as a base for feminist moral thinking; Susan Okin has done so for Rawls’s version of the theory.34 But such theories remain deontological ethics of principle rather than teleological ethics of care; and if one sees feminist moral theory as essentially requiring an ethics of care, then one will have to turn either to fundamentalist utilitarianism or to some (no doubt radically revised) version of market contractarianism to do the job.35

Reconciliations Given these similarities, the prospects for reconciling market contractarianism with

Reconciling Radicals

207

fundamentalist utilitarianism are better than has usually been supposed. Furthermore, any such reconciliation is likely to be quite stable simply because of the deep structural similarities between the theories: both are teleological, both face the same difficulties – and seek the same types of solutions – in getting the right to relate to the good, both are potentially ethics of care, and so forth. There are, I think, two main reasons why few philosophers have attempted to unite the work done in the market contractarian and the fundamentalist utilitarian traditions: the first is what we might call the Sidgwick factor, and the second arises from well-known supposed counter-examples or problem cases for utilitarianism. Neither, I argue, should prevent us from going ahead and exploring the prospects for such a reconciliation. I will first deal with the two obstacles and then turn to what I take to be the most promising path towards reconciling these apparently divergent theories. The sort of reconciliation I propose is in many ways similar to the reconciliation Sidgwick sought between utilitarianism and egoism. By his own admission, he failed. One might suggest that if a philosopher as great and as careful as Sidgwick could not succeed after devoting years of attention to the problem, then there is good reason to devote one’s own energy to some other philosophic project. In addition, utilitarianism faces various well-known problems, and one might wonder about the wisdom of reconciling a theory such as market contractarianism with a theory thus afflicted. Would this not be just to invite more problems for market contractarianism? Both obstacles can be dealt with by observing that much philosophic progress has been made since Sidgwick’s time, and that much of this progress can be applied to the sorts of problems that Sidgwick faced and to the standard counter-examples to utilitarian theory. We do well to remember that, when Sidgwick wrote, the prisoner’s dilemma had not even been clearly formulated; decision theory was underdeveloped, the distinction between different levels of moral thinking only partially understood; the distinctions between act, rule, motive and character utilitarianism were not well understood; the idea that the principle of utility should be seen as a device for ranking possible worlds rather than as a decision procedure was at best only unclearly understood; advances in meta-ethics – from emotivism through universal prescriptivism to norm expressivism – were unknown; the philosophy of language lacked everything from ways to deal with definite descriptions to ideas about rigid designators and possible world semantics; and cognitive science then had nothing to offer philosophers. True, since the publication of The Methods of Ethics, critics of utilitarianism have raised a host of problems for the theory. But the resources with which to deal with those problems are now vast and powerful, and we should not shrink from the task of showing that fundamentalist utilitarianism and market contractarianism have more affinities than differences.

Ought Implies Can We should begin by remembering that fundamentalist utilitarians will be strongly

208

Liberty, Games and Contracts

inclined to use the principle of utility as a device for ranking possible worlds. When determining towards which of the accessible possible worlds the utilitarians will seek to move her fellows, considerations other than moral prescriptions – for instance, institutional arrangements, educational structures, social myths – will likely play a much larger role than which moral code she endorses.36 Nonetheless, let us assume that, in the circumstances in which humans presently find themselves, the accessible possible worlds most favoured by utilitarians all include some moral code or other. What is the prospect that the utilitarian’s moral code is one that a market contractarian could endorse?37 We might begin to argue for the necessary confluence of market contractarianism and utilitarianism by using the so-called ought-implies-can principle. Thus, one might argue that since ought implies can, no utilitarian will advocate (or prescribe) rules of conduct which are beyond those which people can actually keep from violating. One might then go on to argue that people can consistently conform to only those rules which it is rational for them to adopt as being in their individual interest. The argument would then show that market contractarianism yields the only moral code to which rational individuals who find themselves in circumstances where social interaction makes having a moral code worthwhile can actually follow. Thus, utilitarianism and market contractarianism converge on the same set of social rules because utilitarianism will ask of people as much as they can do – to maximize social utility – but no more than they can – to conform with ought implies can. And this window of social rules is none other than those rules advocated by market contractarianism. Both the utilitarian and the market contractarian will advocate moral rules at the same point on the Pareto frontier, for the obvious reason that it would be absurd not to – no utilitarian will give up the extra utility offered by moving to a Pareto optimal outcome and no market contractarian will advocate social arrangement not on the frontier. Unfortunately, this simple argument conflates ought implies can with ought entails can. No utilitarian will accept the ought-implies-can principle as holding that ought entails can, for there is no logical requirement that one confine one’s ought claims to what people can do. Logically, there is no barrier to prescribing xing – saying ‘People ought to do x’ – where x-ing is something they cannot do. The logic of ought does not confine the range of x to the class of those things within the range of can. A non-moral example reveals this. Suppose you are a track coach. Among the things you can prescribe that your sprinters do when they burst out of the starting blocks are A and B: A. Run faster than the speed of light. B. Run as close to the speed of light as you possibly can.

Suppose that whenever you prescribe A to your sprinters they run faster than they do when you prescribe B – though, of course, they never come even close to doing A. Naturally, intelligent coach that you are, you will prescribe A rather than B. Now, if some philosopher claimed that you were doing something wrong – namely

Reconciling Radicals

209

violating the ought-implies-can principle because A demands that your sprinters do more than they can do – and that you should confine yourself to prescriptions such as B, which are within the range of what a sprinter can do, you would have no trouble dismissing the philosopher as a silly twit against whom your sprinters should be afforded protection. What holds for prescriptions for athletes also holds for morality. Those who think we should confine our moral prescriptions to that which can be done, even when prescribing what cannot be done would yield socially better results, are mistaken. Most people are incapable of keeping from (occasionally at least) coveting their neighbour’s ass. But it would be silly to think that, for this reason, the prescription ‘Thou shalt not covet thy neighbour’s ass [or any of his other property]’ is not an acceptable moral rule, one which no utilitarian should be inclined to consider advocating as appropriate for a society to adopt.38 Whether the utilitarian decides to advocate such a rule does not depend simply on whether the members of society are psychologically capable of refraining from coveting their neighbours’ property, but on whether – given whatever their psychological capacities in this and other respects – promulgating this rule will lead to the greatest happiness. In the event that telling people to refrain from doing what they can’t help but do will make for a better society than telling them anything else, utilitarians will have no qualms in telling them this. Of course, the costs – in terms of feelings of failure and guilt, of undermining respect for morality and so forth – of advocating moral rules which, strictly speaking, are impossible to follow will weigh against such prescriptions. Consequently, as a rule of thumb, the utilitarian will be disinclined to violate the ought-implies-can principle. But in some cases the consequences of having moral rules which violate the ought-implies-can principle are, on balance, such that the utilitarian will be inclined to embrace them. So, we cannot use the ought-implies-can principle as the basis for a simplistic reconciliation of utilitarianism with market contractarianism. This is not to say that no general argument of this form might be developed to unite the two theories, for what holds for prescriptions a utilitarian would be willing to endorse holds also for those prescriptions a market contractarian might endorse. Though the metaphor of making a contract obscures this feature of market contractarianism, there is nothing in the logic of the theory that precludes the idea of the contractors making a contract which they know they literally cannot always abide by. As long as the contract – one which includes ought claims that cannot always be followed – is the one most likely to match the ideal that MRC sets out, the market contractarian must endorse the contents of said contract. As Duncan MacIntosh has suggested to me, we might distinguish between [(a)] whether it is morally permissible or obligatory to perform the speech act of enjoining a practice (whether from a utilitarian or from a contractarian perspective), and [(b)] the practice it is morally obligatory to enjoin being itself thereby a morally obligatory practice. The two might diverge. Along similar lines, we might distinguish between a good state of affairs and a state it is obligatory to bring about. Certain states might be highly utile, or might implement an ideal contract, and yet be impossible to be brought about by available actions of agents. And this might lead us to separate the good from the right.39

Liberty, Games and Contracts

210

So the fact that utilitarianism may endorse the adoption of norms which cannot always be followed does not distinguish it from market contractarianism; indeed it is another feature the two theories share. All that remains is to show that whenever one theory endorses norms that cannot always be followed, the other will do so also.40 Notes 1

2 3 4

5

6

7

8 9

10

This terminology is not entirely happy. Calling some utilitarians fundamentalists might suggest that they do not respond appropriately to arguments offered against their position. Calling other theorists quasi-utilitarians might be taken to suggest that they are (somehow) not ‘truly’ utilitarians. I have taken the term quasi-utilitarian from Blackburn’s use of ‘quasi-realism’. Simon Blackburn, ‘Moral Realism’, reprinted in Simon Blackburn, Essays in Quasi-Realism (New York, 1994), 111–29. For an account of the problems facing contractarian theories, see Sheldon Wein, ‘Problems with Contractarianism’, Journal of Social Philosophy, XVI/3 (1986): 48–59. It might be better to say that Smith would have been a quasi-utilitarian had he realized the full implications of his impartial spectator theory. I do not mean to suggest that quasi-utilitarians can be distinguished from fundamentalist utilitarians by the strength of their commitment to the principle of utility. It is the level at which the commitment resides, not its strength, that counts here. I do not mean to claim that fundamentalist utilitarians cannot either become or be utilitarians because they accept arguments for the principle of utility. The following is an argument which I suspect Bentham accepted. The term ‘utility’, whatever it means, means the stuff that is good for its own sake. Obviously, it is better to have more of this stuff in the world than less. Any non-utilitarian theory says that, in effect, we should have less good (and, what amounts to the same thing, more bad) in the world than utilitarianism says we should have. (Asceticism is an extreme case of this, holding that we should maximize the bad.) But how could it be wise for a moral theory to prescribe making the world a less good place than it otherwise might be? (Jeremy Bentham, The Principles of Morals and Legislation (Buffalo, NY, [1781] 1988), ch. II.) Fundamentalist utilitarians can accept arguments for the principle of utility and remain fundamentalist, but they cannot remain fundamentalists if they hold that the principle of utility depends upon some more basic moral principle. See John Rawls, A Theory of Justice (Cambridge, MA, 1971), 15, 121. I include in this part an account of what metric is to be used in making such choices. For Rawls it is primary goods, for Hobbes it was the interests of individuals (where each has an overriding interest in staying alive) and for Gauthier it is non-tuistic preferences. For an account of why different societies might prefer different contractarian theories, see Sheldon Wein, ‘American and Canadian Justice’, in Terry Goldie, Carmen Lambert and Rowland Lorimer (eds), Canada: Theoretical Discourse/Discours Theoriques (Montreal, 1994), 373–91. It also goes by such names as Kantian constructivism or contractualism. I use Rawls’s version of social contractarianism (from A Theory of Justice) partly because I think it is the best statement of that theory, but mainly because it is the best known. I could have, though not as easily, used Rousseau, or Kant or, say, Scanlon. Rawls, A Theory of Justice. In Kant’s case it is dignity and respect that plays the role of Rawlsian fairness, and in Rousseau’s it is the kind of impartiality needed by the general will.

Reconciling Radicals 11 12

13 14

15 16

17 18 19 20 21

22

23

24

25

211

See Lanning Sowden and Sheldon Wein, ‘Justice and Rationality: Doubts about the Contractarian and Utilitarian Approaches’, Philosophia, 17/2 (1987): 127–39. Narveson argues that either Rawls’s first principle dominates entirely and leaves no role for the difference principle or, if that is not the case, the principle of liberty plays no significant role. Jan Narveson, ‘Rawls on Equal Distribution of Wealth’, Philosophia, 7/2 (l978): 281–92. See ‘Justice and Rights’, Chapter 6 of Ronald M. Dworkin, Taking Rights Seriously (London: Duckworth, 1978). Rawls outlines this interpretation in section 40 of Theory of Justice. In his later writings he moved away from the Kantian interpretation. See John Rawls, Collected Papers (Cambridge, MA, 1999). See John Rawls, Political Liberalism (New York, 1993), 15, 39, 150–54, for example. It sometimes goes by the name Hobbesian contractarianism or economic contractarianism, but following Jules L.Coleman – Markets, Morals and the Law (New York, 1988) and Risks and Wrongs (New York, 1992) – I have adopted the name market contractarianism. See Sheldon Wein, ‘Plato and the Social Contract’, Philosophy Research Archives, XII (1986–87): 66–77 for a discussion. Thomas Hobbes, Leviathan (New York: Cambridge University Press, 1991). See ‘David Hume: Contractarian’, in David Gauthier, Moral Dealing: Contract, Ethics, and Reason (Ithaca, NY: Cornell University Press, 1990), 45–76. Henry Sidgwick, Methods of Ethics (Indianapolis, 7th edn, 1981), bk II. David Gauthier, Morals by Agreement (New York, 1986). In my view, Gauthier’s Morals by Agreement stands to Hobbes’s Leviathan in the same relationship as Rawls’s A Theory of Justice stands to Rousseau’s The Social Contract. That is to say, it is the best contemporary statement we have of that type of theory. On the role of tuism, see Sheldon Wein, ‘Prisoners’ Dilemmas, Tuism, and Rationality’, Simulation and Games, 16/1 (1985): 23–31. For an alternative account, see Susan Dimock, ‘Defending Non-Tuism’, Canadian Journal of Philosophy, 29/2 (1999): 251–74. This assessment is shared by Gauthier. See David Gauthier and Robert Sugden (eds), Rationality, Justice, and the Social Contract: Themes from Morals by Agreement (Ann Arbor, MI, 1993), 185. For an excellent discussion of MRC (from which this statement is derived), see the ‘Introduction’ to Peter Vallentyne (ed.), Contractarianism and Rational Choice: Essays on David Gauthier’s Morals by Agreement (New York, 1991). Gauthier originally thought of MRC as an alternative to the standard ‘Nash’ solution to bargaining games. But he now holds that ‘in the circumstances of the social contract, MRC coincides with the Nash bargaining solution’ (Gauthier and Sugden, 178). Jan Narveson holds that Gauthier’s version of the Lockean Proviso is the basic principle of justice the bargainers agree to. I understand Gauthier to accept the Lockean Proviso before the contracting begins as a condition all see as necessary in order for whatever contract is subsequently adopted to be a stable one. Thus Gauthier’s acceptance of what might seem to be a moral constraint on the subsequent bargaining – the Lockean Proviso – is accepted for the same reason that he rejects Rawls’s original position, because contracts made in the absence of the proviso would carry no weight. See, for example, Jan Narveson, ‘Rawls and Utilitarianism’, in H. Miller and W. Williams (eds), The Limits of Utilitarianism (Minneapolis, l982), 128–43; and David

212

26 27 28

29

30

31

32 33

Liberty, Games and Contracts Braybrooke, ‘Utilitarianism with a Difference: Rawls’s Position in Ethics’, Canadian Journal of Philosophy, 3 (December 1973). See, for example, Rawls, Theory of Justice, sect. 6. Immanuel Kant, Critique of Practical Reason, Pt 1, Bk 1, Ch. 2 [page 63 of Kant’s Gesammelte Schriften, Preussische Akademie der Wissenschaften (Berlin, 1913)]. Rawls, Theory of Justice, 30. This is the only place in A Theory of Justice in which Rawls displays any anger. The anger is directed at those ‘pure’ or ‘strict’ deontologists who think that consequences have no role to play in moral theory. Jeremy Bentham (1781), ch. I, sect. X; Bentham’s italics. The way to understand Bentham here is to think of him as saying that, given that we have accepted a teleological principle as our fundamental moral axiom, the question arises as to what to do with various ordinary language terms which appear to be deontological in their use. Our options (given our acceptance of the principle of utility) are either to abandon such terms (hold that they have no sensible use, hence no meaning) or to reassign them a new meaning by (stipulative) definition. Bentham opts for the latter. I thank Bob Bright for discussing this matter with me. MacIntosh rejects alternatives, such as Gauthier’s, which say that we should adopt a stable disposition to cooperate with other cooperators, arguing that if it is rational to change one’s conception of rationality from straightforward maximization to constrained maximization, then it will be equally rational to change one’s conception back from constrained maximization to straightforward maximization once others have cooperated with one and it is now one’s turn either to reciprocate or to defect on others. For reasons along these lines, he thinks that only by getting the contractors to change their values – so that they come to value the cooperative outcomes more than they value the outcomes attached to lone defection – can the project of getting instrumentally rational contractors to cooperate in prisoners’ dilemmas succeed. See, for example, Duncan MacIntosh, ‘Preferences Progress: Rational Self-Alteration and the Rationality of Morality’, Dialogue, XXX/1–2 (1991): 3–32, and Duncan MacIntosh, ‘PreferenceRevision and the Paradoxes of Instrumental Rationality’, Canadian Journal of Philosophy, 22/4 (1992): 503–30. Mintoff argues that it is rational to form an intention to cooperate when the expected utility of forming the intention to cooperate exceeds that of not having the intention; and it is rational to act on that intention when there is (subsequently) no new information which gives one a reason for changing one’s intention. This ingenious solution works – if it does – because the information that it would be utility maximizing to reconsider one’s intention once the others have cooperated (and one could, if one changed one’s intention, defect on them) is not new information; it was known at the time one formed the utility-maximizing intention to cooperate. Joseph Mintoff, ‘Is Rational and Voluntary Constraint Possible’, Dialogue: Canadian Philosophical Review, XXXIX/2 (2000): 339–64. Peter Danielson, Artificial Morality: Virtuous Robots for Virtual Games (New York, 1992). See Eric M. Cave, Preferring Justice: Rationality, Self-Transformation and the Sense of Justice (Boulder, CO, 1998). We might think of Cave’s position as a sort of Aristotelian market contractarianism. One cultivates the virtue of becoming willing to cooperate with others in whatever projects are needed to sustain and support civil society. One transforms oneself from a Gauthierian straightforward maximizer into a person of virtue, one who cooperates when such cooperation moves the social project forward.

Reconciling Radicals

34 35

36

37

38

39 40

213

Railton offers a utilitarian theory similar to Cave’s market contractarianism. See Peter Railton, ‘How Thinking about Character and Utilitarianism Might Lead to Rethinking the Character of Utilitarianism’, Midwest Studies in Philosophy, 13 (1988): 398–416. For an evolutionary contractarianism, see Malcolm Murray, The Moral Wager: Evolution and Contract (Dordrecht, 2007). See Susan Moller Okin, Justice, Gender, and the Family (New York, 1989). For Rawls’s assessment, see Rawls, Collected Papers, 156–7, n. 58. For my own contribution to this project, see Sheldon Wein, ‘Feminist Consciousness and Community Development’, The International Journal of Social Economics, 24/12 (1997): 1376–87. It is even possible that the utilitarian will find moral anarchism – not having any moral code – an attractive option. But, as we will see, this option is not ruled out for the market contractarian. For an argument that market contractarianism will yield social arrangements that most utilitarians would find attractive, see Jan Narveson, ‘The Invisible Hand’, Journal of Business Ethics, 46/3 (September 2003): 210–12. For an argument that market contractarianism yields institutional arrangements that most utilitarians would find attractive, see Sheldon Wein, ‘A Hobbesian Foundation for Welfare Rights’, in C. F. Delaney (ed.), The Liberalism-Communitarianism Debate (Lanham, MD, 1994), 213–26. Narveson’s Morality and Utility contains an excellent, albeit now hilarious, conceptual analysis of whether people who are gay are necessarily, or only contingently, happy: ‘If a person is gay, does it follow that he is happy? There seems to be some logical connection here. But then if a person wants to be happy, shouldn’t it follow that he would want to be gay?’ Jan Narveson, Morality and Utility (Baltimore, 1967), 54–5. Personal communication. Some of the ideas in this chapter were first presented on 26 March 2000 at the International Society for Utilitarian Studies meeting at Wake Forest University in Winston-Salem, North Carolina. I am grateful to my commentator, Geoff Sayre McCord, and for the many insightful questions I received. In radically revising the paper I have drawn on things I learned in conversations with Bob Bright, Thea E. Smith, Paul Viminitz and, especially, Duncan MacIntosh. I thank MacIntosh and Jan Narveson for detailed written comments on an earlier draft of this paper, and Thea E. Smith, Alejandra Stoykowski and Malcolm Murray for their assistance and encouragement.

This page intentionally left blank

PART FOUR RESPONSE

This page intentionally left blank

Chapter 15

Social Contract, Game Theory and Liberty: Responding to My Critics Jan Narveson

I am impressed at the range and perceptiveness of these papers, and grateful both for the interest the authors show in my work, and the opportunity to try to restate my ideas. These 14 papers offer a challenge that is certainly worth more than the few pages I can devote to them here. In fact, they make me think that perhaps I should expand the several-times-too-many pages I initially wrote on reflection of them to a book in its own right. We’ll see! Meanwhile, my thanks again both to Grant Brown, who originated the idea for this Festschrift project and did much initial work on it, and to Malcolm Murray for his wonderfully efficient job at bringing it to a successful conclusion – as well as to Ashgate for bringing it to the attention of, I hope, a wider interested public.

Contractarianism E. J. Bond When it comes to expressing himself on matters on which we agree, I envy my old friend Ted Bond no end: he always manages to put things ever so much better than I can myself. He agrees that morality must make us all better off. But he professes to disagree with the contractarian foundation for it which I have advocated for the last three or so decades. The problem is that if things were different – if we did have the Ring of Gyges, if we were invincible – then we ought to behave very differently from the respectful and peaceable bourgeoisie that most of us have become. ‘The picture we have here is of individuals existing in isolation, each one motivated only by self-interest, not caring or not necessarily caring, for anyone other than himself or herself.’ Now, I will be the second to agree that we do not exist in isolation, and indeed that we are not, most of us, motivated entirely or even mostly by self-interest in any really meaningful sense of that term. (I did go out of my way to make this clear in The Libertarian Idea.1) If we had a morality of Humean sentiment rather than of reason, it would still look pretty much like what we actually have. I do not really think, then, that ‘though they cooperate for their mutual advantage, they are still basically at odds, and another person’s good is still no part of one’s own’. But there 217

218

Liberty, Games and Contracts

is no option of ‘another person’s good’ being ‘part of one’s own’, for that is a metaphysical impossibility. Suppose that Sam is hopelessly in love with Matilda, and she with him. Yet however poetic they wax, Matilda is Matilda and Sam is Sam: Sam’s love for her is in Sam, and hers for Sam is in her. Yet this is not something to deplore. It would be, though, if everyone tried to love each other as Sam loves Matilda. We must go easy with that particular commodity. I do not dispute Bond’s ‘big empirical claim’, that ‘People just are happier when they live in conditions of friendliness, trust, good will, and mutual support.’ Sure, some people, but all? Nor is ‘our happiness or well-being absolutely dependent upon good social relations and a society in which people are genuinely concerned with each other and with their common or shared goods – food and water to take the most obvious examples’. I think that most people’s happiness and well-being is not particularly dependent on the genuine concern of more than a few others; the rest they are disposed pretty much to ignore. Bond proposes that ‘Because the common good is an essential ingredient in the good of each and every individual, to choose morality is not the same as to draw the line at second best. On the contrary it is contractarianism that is second best.’ It seems to me that the ‘second best’ of contractarianism is only misleadingly so called. What is wrong with the disposition to go for a ‘more’ that is at the expense of interactees who, in consequence, must be worsened by our action is indeed that it makes social life evil rather than good. Actually, I may thank Bond for a real insight here: for it seems to me that the moral disposition is a necessary condition for making life really livable, and that is indeed ‘first best’ among really possible lives. Too, Bond might talk differently if we include among his ‘others’ the many chickens, cows and fish to whom he and I give no second thought when we are at the restaurant. The goodness of the social life we can have with our fellow humans is very much due to their being fellow humans. We need not, and most of us do not, treat those who are genuinely and fundamentally unequals in the way we treat each other – and we ought not. If morals were founded on general love, this would not be so. John T. Sanders Sanders begins with a long and interesting discussion of contractarianism in my theory and in general, relating it especially to utilitarianism. He points out, what is surely true, that if social contract theory is to work, it will work not because people adopt morality in the style of a contract negotiation with subsequent agreement, but because considerations of what is in their best longer-run interests control their behaviour even when they are not thinking explicitly about it. How, then, is moral theory better than behavioural genetics? ‘But if morality really is based on what is reasonable, then ants and humans are both subject to its rules.’ I rather think, however, that the practice of often resorting to discussion of rights and wrongs does make a difference in human conduct. Suppose we are all in fact actuated by similar fundamental moral principles. These principles are intended to, and often do, constrain the operation of unrestricted pursuit of interests (including,

Jan Narveson: Responding to My Critics

219

typically, self-interest – but by no means restricted thereto). At this level, invoking the Golden Rule, for example, sometimes makes a difference. Does practical advice on what we want generate morality? Sanders takes moral ideas about animals as a case in point. Contractarians insist that if we think there are duties about animals, they have to be generated by us, because we are the only individuals who can make ‘social contracts’. But Jack has his doubts: ‘That this should be the way to arrive at moral conclusions seems wildly counter-intuitive … that what is wrong regarding the treatment of animals should end up being fully dependent on what you and I perceive to be best for ourselves just seems mistaken.’ Does it? The world we live in contains a lot of people (such as Jack and myself) who do eat animals, along with some who do not. How could we think that animals count as we do, and yet be routinely eligible for the cooking pot? It seems perfectly clear that most of us do not think that animals ‘count’ in that way. Serious consideration of their interests would surely yield rights for them incompatible with our interest in steak. We must think that their interests do not really count. And why not? A plausible answer is that we simply do not need to count them. We do need to count our fellows, though. Those guys are dangerous – but cows are not. Vegetarians may object – and of course the social contract in my version allows them to object – but as to punishing the meat-eaters for doing otherwise, that is not on. So the contractarian view yields the truth here. Why is that not a point in its favour? Is there any point in moral theory? Many writers today appear to think that there is nothing for us to do but track the opinions on moral matters that ‘we’ have. The fact that there appears to be great divergences of opinion does not seem to faze them – but it should. Now, Sanders has provided us with some reason to ascribe to the contractarian project a certain status of intuitive plausibility: maybe this is what is wired in genetically. Maybe. But it remains that we have work to do: among these divergences, who is right? Unlike the ants, we have to confront our fellows, verbally and otherwise, and see whether we can sort things out. Armed with gametheoretic analysis to show us where the Common Good plausibly lies, if anywhere, we should be (and are) in a position to deduce relevant results which will be persuasive. This works well in business, where people are accustomed to bargaining in self-interest with independent persons equally self-interested. There the common good is the price reached by negotiation and by experience, trial and error. Why not in moral theory? As to utilitarianism: where it goes just plain wrong is in all those cases where it calculates that the general utility is maximized at x, while various individuals’ contribution to x is that of the steaks on our platters. Forget it! Finding the optimum rather than maximizing some conjectural interpersonal utility to promote is the way to go. That seems to me definitive. Tibor Machan Tibor raises the question whether the ‘compactarian idea’, as he calls it, is up to the job of ‘grounding’ moral principles. Whether it is can only be decided by careful

220

Liberty, Games and Contracts

analysis. But we must start with the right characterization – not quite what Tibor supplies: ‘A social compact theory … would contend that the correct principles of community life are those to which either actual or hypothetical community participants agree to be legally bound.’ He should say ‘morally bound’. Being legally bound is subordinate. The contract view proposes that moral rules are what it is in the rationally best interests of all to accept in the way of general bases of criticism of interpersonally significant behaviour. Tibor thinks that ‘inadequate’. But inadequate for what? I do not claim that everyone’s ‘agreeing’ on what I claim it is rational for them to agree on will also ipso facto get everybody to be just, just like that. Supplying a theory that will do that would be quite something – the word ‘magic’ comes to mind. What, then? Well, he says, ‘nothing follows from everyone concerned promising to abide by some guidelines unless promises themselves have normative significance. And this cannot derive from a still prior compact, ad infinitum.’ Well, what could it mean to say that promises have ‘moral significance’ beyond what is conveyed by agreement? If you accept an obligation to me, and I to you, does it follow that we are obligated in the ways we agreed to? Well, the parties to this agreement could hardly deny it. ‘I agree: I am obligated’, followed immediately by ‘Oh, by the way – I’m not, actually!’ are incompatible. And if, as in my theory, it is rational for everyone to accept this, it is a little difficult to see who is left to deny it. Nor is it clear how invoking some further supposedly ‘prior’ obligation is anything but a deus ex machina. Does Tibor want us to agree that this ‘prior’ obligation is there? But how is that any better than agreeing that this one is in force? If he is worried about infinite regresses, proposals such as his are in exactly the same boat.2 Hobbes argued that promises are ‘but words’ and ‘of no strength to defend a man’. The worry is that going through these verbal manoeuvres might not actually motivate us to do the things we say we are going to do. And he has a point, of course: some people are liars. When they say ‘I agree’ they lack that ‘motion of the mind’ that consists in meaning what they say. Still, the point of promising is indeed to signify acceptance, and since that is easy enough to fake, the possible need for enforcers is something to be concerned about, no doubt. But then, exactly the same thing will be true of any supposedly ‘objective’ duty any theorist wants to invoke. Tibor turns to the subject of ‘self-consistency’, claiming that it too is ‘a distinct normative requirement’, not to be analysed in terms of agreement. I am puzzled by this. The incompatibility of p and not-p is due to the syntax of those respective verbal structures: the word ‘not’ signifies the logical complement of whatever it is attached to. Going outside these obvious meanings is not going to help. And so this seems to me an illusory claim – just like the claim that obligation consists in ‘more’ than a mutually rationally accepted claim (accepted, that is, by all – not just the two participants to this particular one). We might say that legal obligation does more, of course, since it adds a threat to the original obligation. But does Tibor want to say that threats are what obligation consists in? I assume not; after all, either the threats are justified, or they are not. If not, they may motivate, indeed, but that is not moral obligation. Now, what I claim it is rational to agree to, for the basic general purposes of

Jan Narveson: Responding to My Critics

221

morals, is the libertarian principle. If I am right in claiming that this is rational, then someone who denies that he has the obligation to refrain from attacking his nonviolent fellows would be irrational. If that is so, then of course he has the obligations in question, and has them by virtue of having a structure of beliefs and motives such that his having them follows from those beliefs. What is wrong is to inflict evils on innocent people. But what makes that wrong is that it is unreasonable – indeed irrational – to sanction such things. Contractarianism does indeed claim to solve the basic ‘is-ought’ problem of moral theory. Here’s how. All of us have various ends, desires, interests, which form our practical basis of action. Now, morality is not a general guide to life, but a rather restricted set of social rules, social norms, meant to apply to all, not just oneself. The project is to formulate general rules restricting our behaviour that we would all do well to accept. The factor of obligation enters because the needed rules are ‘iffy’: I’ll do this, provided you do that; you do that provided that I do this. Cheating is possible: A can free-ride on B’s trust by accepting B’s offered benefits without the reciprocation that makes B’s agreement rational. The sense of moral obligation guides people across this ‘gap’: that we ought to adhere to them follows from the acceptance of the rules in question; and that acceptance is a consequence of our perception that if we are to make our way in the company of our fellows, this is what we must do. But once that is seen, there is nothing further that would or could do any good. Leo Groarke Leo’s explication of my worries about intuitionism could hardly be improved on. But I am puzzled at his reference to something called ‘the conception of reason which is the basis of Narveson’s contractarianism’. I confess to not supposing that I am relying or leaning on something called ‘a conception of reason’. The point of the contractarian move is to find a basis for normative agreement despite widely varying individual values and ways of life. People use their reason in attempting to realize those values and practise those ways. We propose that moving to principles that optimize among us is an extension of that very same reason. Now, if we address, say, religious fanatics, then I am indeed, in effect, ‘asking them to give up their commitment’ to certain of their ‘convictions’. Those who conclude that it is their sacred duty to go out and exterminate those who do not accept their religions are asked, in the strongest terms, to change their minds on that point; if not, a reasonable question arises why the rest of us should not exterminate them while we still have the chance. If they do not like that, then they need to rethink this supposed ‘commitment’ of theirs. Leo says, ‘In our real lives, there is no way to neatly separate the different components of our worldviews and isolate a separate conception of reason or selfinterest that we are committed to. … What we view as rational or in our interest … is closely tied to a variety of complex views and intuitions.’ Now, reason is a matter of (1) inference and (2) premise-checking, at the hands of the relevant facts. There is no room for ‘varying intuitions’ on the subject of modus ponens. Nor is there,

222

Liberty, Games and Contracts

seriously, on the question of whether there are trees, or whether we have good evidence for the germ theory of disease. Those who claim to have differing ‘conceptions of reason’ which put things like that ‘in doubt’, I think, must be dismissed as either just confused (like ‘post-modernists’), or else as playing some kind of verbal game. What Leo no doubt is getting at here is something I have insisted on throughout my work: that our interests are very variable, and by no means confined to self-interest in the most significant sense of that term. Variation about ‘the nature of the good life, and our relationships with others’ is standard. But this is precisely the kind of variation that is, and needs to be, subject to restriction at the hands of morals – which stems from the more general, game-theoretic kind of considerations that are, literally, the foundations of morality. If your idea of the good life is, like Heliogabulus, to enjoy the aesthetic effect of fresh human blood on green grass, then you are in for a fight – and it is one that I hope and expect that you will lose. Nor does someone ‘dedicated’ to the proposition that p implies not-p have that as part of his ‘full-bodied individuality’. (Such a person perhaps needs to sit in on Philosophy 100 for a while.) ‘Essential beliefs’ have to do a lot better than that. Thus I deny, too, that I ‘attempt to deduce morality from a narrow conception of economic reason’ – thus qualifying for a charge of not accepting individuals as they are. On the contrary: accepting people as they are is the beginning of wisdom in this business. By this we mean proceeding from the practical premises they actually do have. But that is where we start, not where we end up. Contractarianism claims to find a basis in everyone’s practical reason for accepting a set of norms calling for restriction on people’s activities in pursuit of their various values. To accept not only their interests, but also their moral conclusions is not feasible, since those conclusions are mutually inconsistent. (Example: there is no way to combine or compromise between egalitarianism, which calls for compelling people to be ‘equal’, and libertarianism, which calls for not doing that. Since these are rival theories about justice, one of them must be wrong. I think it is egalitarianism.) The libertarian idea accommodates more of ‘humanity’ than any alternative, since it leaves people with the ‘maximum freedom that is compatible with a like freedom for all’. How is this based on a ‘narrow and constricted’ notion of human nature? One notes that Groarke is willing to allow this: ‘[O]ur lives, and especially our interactions with others, must operate within some kind of consistent moral and political framework, and … this requires some agreement.’ But that’s my line! The required ‘agreement’ is to let people be, so far as that is universally possible, and so, not go pushing them about on the basis of some aesthetic or religious intuitions of one’s own. Leo seems to agree that this is not too much to ask. Yet it’s all I ask! I think I am indeed sympathetic with Groarke’s proposed ‘third way’ – since it is, as far as I can see, my way. If we could indeed all be better off, in our own views of what is good for us, from some kind of imposition, then that imposition is justified. Yet if it really does make people better off on their own views of what makes them so, why would we need all that compulsion to do it? So the third way really collapses into the ‘first way’ – the way of the social contract, the agreement of all with all, which I suspect cannot improve on the libertarian formula.

Jan Narveson: Responding to My Critics

223

Chris Tucker If ‘solving the compliance problem’ means showing that each and every person with a functioning brain will in fact always do the cooperative thing, then that problem is, I presume, insoluble, given people as they more or less are. But that is not, in my view, what is to be shown in moral theory. What is to be shown is that compliance is called for by a set of norms such that it is rational for everyone to agree to – that is, to endorse them. Doing, in turn, the things called for by those norms requires a level of virtue that not everybody has. Our complaints against non-compliers are reasonable complaints, and the people we are complaining about are, I think, in no position to deny those complaints when pressed against them: that is what I take to be the project of morals. That said, we can also agree that general – meaning, imperfect but frequent – compliance is to be expected if these norms really are reasonable. What is right and what is in one’s interest even in the moderately short run should normally coincide. In this Tucker is surely right, and I hope I have not left the impression that I disagree. Indeed, if one looks at the structure of the first Hobbesian Law of Nature, aka Gauthier’s constrained maximization, it is clear that this allows for the distinct possibility that we will meet with non-cooperators, since the Law points out, in its ‘second branch’, that if we cannot get peace from our interactee, then we may use ‘all the helps and advantages of war’. It may be, in a given case, that even doing our best we will be unable to prevent the assailant, or whoever it may be, from visiting evils on ourselves. But must we in such cases admit that he is, while at it, ‘rational’? Or should we say that the bad guys do not fundamentally act rationally? One can see what is meant by those who insist that they are rational. Given that somebody is going to be a swine, it is easy enough to attribute rationality to him in the ways in which he goes about being so. But is it rational to be a swine in the first place? It is, at any rate, irrational of the rest of us to tolerate swine, and it is irrational of the swine to think that he is morally justified in being one. That much, I think, is shown by social contract reasoning. There is too much else of too much interest in Chris’s paper to admit of the discussion it deserves here. I will add only that some questions of substance may be complicating this discussion. For example, I do not accept Gauthier’s minimax relative concession (MRC) as a criterion of justice. Indeed, it is not consistent to propose both MRC and his Lockean Proviso (LP). If we say, as we liberals (including Gauthier) all do, that each individual is the ultimate judge of his own interests, then Gauthier will have to admit that someone might settle for a distributive share less than MRC apparently grants him. What then? We will have to go one way or the other. I think that the LP takes priority. So, that ‘[Gauthier’s] project begs the question when arguing for the rationality of compliance with any joint strategy that approximates the terms identified by MRC’ is something I cheerfully agree with. So much the worse, then, for MRC. Finally – among the few things I have space to say – I am enthusiastic about Tucker’s ‘contractarianism with education’. My work has in general seen the output of the social contract as something to be installed in minds as a piece of

224

Liberty, Games and Contracts

highly desirable software, ‘memes’ in the recent terminology; and I think that if we can demonstrate the rationality of doing just that, we have done much. What makes the software rational is that it really can be expected to pay off in social circumstances. As Hobbes says to the Foole, we can only succeed in what this view of morality insists is immoral by ‘the error of them that receive him’. We do not really need any more than that, I think, to have done what we claim to do – viz., to establish rational foundations for morals – and in this, I gather, Chris and I are at one.

From Contractarianism to Libertarianism Susan Dimock Susan concurs in my choice of the contractarian view as the basic generating idea of morals. Nevertheless, she does not think that libertarianism issues from this foundation. The problem, she says, has to do with autonomy and the role it plays in this system. I fail, she avers, ‘to give due weight to the importance of autonomy’ in drawing my ‘substantive conclusions in favour of libertarianism’. I do not think I see this. The question is, how is ‘weight’ relevant here? Autonomy reflects one’s ‘true’ or ‘authentic’ self. Okay. Which self is ‘true’? Citing Frankfurt, she proposes that we can decide which desires we want to have and which not; the true self acts only on the former. Well, suppose we accept this: do I then have a right, nevertheless, to act on my inauthentic desires? I think so: no one else can decide which are which, or compel me to stick to the authentic ones. I assert Rights for the inauthentic! No doubt, if we act on some of our desires, things will go badly for us. Tough – that’s life! Live and learn. If you say, ‘Here, tie me to this post so that I don’t go chasing after yonder blonde’, I presume that this gives me permission to do the tying. But if I do not comply, it is not my fault if you then make a fool of yourself. And I gather that Susan agrees. Does a higher-order desire imply that the agent attaches more utility to it than to a lower-order one? I said as much in my first book.3 But why would this upset the libertarianism I now embrace? Suppose we agree that ‘So long as the conflict between formally inconsistent desires is unresolved, nothing that the agent does will satisfy her desires; necessarily, satisfying the desire to x means frustrating the desire to not-x.’ True: you have a problem. But it is not mine. For example, I have spoken out strongly for the legalization of drugs.4 Some takers of drugs are presumably addicts. If they are to be ‘saved’ from this, salvation will have to come from within. We are not entitled to arrest the person who sells Jones the drug to which Jones is addicted. We are entitled to prevent Jones from stealing from Smith the money necessary to support his habit. But we do not owe you a living regarding your inauthentic desires any more than with your authentic ones. Thus far, interestingly, Dimock accepts my verdict that we cannot allow autonomy any special precedence in the interpersonal contexts which are the home of moral theory in the sense we are concerned with here. I do not see, then, that there is any problem as regards the coherence of my libertarianism.

Jan Narveson: Responding to My Critics

225

She goes further in saying, ‘if people cannot but act rationally then the contractarian has no normative work to do and no useful advice to give.’ I do not see this either. People can act against norms that it is rational for them to subscribe to; and whether we claim that in so acting they act irrationally (as perhaps Dimock wishes to claim) or on the contrary we say, with Viminitz, that terrorists and the Mafia act rationally, it remains that we have the right to defend ourselves against them if need be. If drug addiction causes Jones to act violently toward others, we are justified in preventing Jones from having access to the drugs that do this. But we are not justified in doing so just because we think drugs make Jones inauthentic. Autonomy is an interesting but dangerous notion. It can be, and often is, invoked to support the suppression of liberty far and wide. It is not worth it. I deny that autonomy is a central idea here. Rationality, of course, is. Perhaps in schools autonomy should be hailed as a virtue. But, please – do not require anyone to attend them! Peter Danielson If David Gauthier is the one who woke me from my dogmatic utilitarian slumbers, I surely owe it to Peter Danielson to have alerted me to the possibility that I have just gone from one dogmatism to another. I confess to general uneasiness regarding my answer to his first challenges, raising the spectre of chicken before our startled eyes – and his new analysis certainly increases the unease. There is also the puzzling issue of how game theory relates to the real world. Peter’s title, ‘Simple Games and Complex Ethics’, suggests what we should not accept without further question: namely, that game theory simplifies while ethics is complex. While there is an obvious sense in which that is true, still, there really are very strong reasons why ethics, in the sense of Morals – what we are all interested in here – should in fact be fairly simple. Game theory, with its unlimited conceptual resources, might make things too complicated. Richard Brandt, in one of his most stimulating papers, proposes that for some putative set of rules to qualify as a moral code, its elements have to be simple enough to be learnable.5 That makes sense, especially considering that the zillions of people whose behaviour we hope to direct by those rules cannot be expected to devote their lives to figuring out what they are. We have to be able to pass moral rules around, teach them, discuss things in light of them – and to do so among ordinary people. They have to be communicable. Now, perhaps the communication can be unconscious, unpremeditated. But insofar as it is like that, morality cannot be scrutinized, discussed, appraised. And surely it must be capable of such scrutinizing. Now, what about chicken (among other) games? Might a moral code say that playing hawk is perfectly okay? What it will say about prisoner’s dilemma (PD), I think, is a ‘slam-dunk’: cooperate (so long as the other guys will too). But what if we all believed, in the case of chicken, that we should always knuckle under? This sends a message to all who think they would be kings of the roost: go right ahead! If we all played such games most of the time, life, I should think, would be miserable for everybody. Iterated chicken, it seems to me, gives us the same result as PD: society’s verdict is, don’t play!

226

Liberty, Games and Contracts

Actually, of course, the rules we are talking about are also, in varying ways, enforceable. And enforcement is coercion, which is standardly analysed as chicken. Now what? This brings up an important point. Peter says that the value of chicken is higher than PD. To arrive at that conclusion, he invokes two simplifying assumptions: first, that the ordinals defining the game be treated as cardinals; and second, that we randomize between chicken and PD. But making those moves gets us into a major real-world problem. For in typical coercion situations, the four outcomes are decidedly non-linear, and very different for the two players. If I resist the serious robber, he kills me, while he ends up being guilty of murder. It is the worst outcome for each, but his worst is a lot better than my worst. Likewise with government, which I can put to some expense by refusing to pay my taxes – but it can put me in jail. Now, what? The wimps among us (almost all of us) can perhaps help things out by turning to democracy, for one thing, so that we are in a position to throw the rascals out – sort of. But, morally, surely my view is the plausible one: do not coerce (the innocent). We may turn to democracy in hopes of reducing coercion, even though we do so by coercing the coercers. But the point is that they are in the wrong. Peter, I think, supposes that the moral here favours the state. Now, anarchy is indeed an interesting issue to consider, but it is not the only one. We ordinary people need to worry about low-level coercers such as bank robbers and rapists. Our response to them may be to take up arms against them. Or will it be only to get the state into the act? No. For example, equipping potential rape victims with handguns, contrary to the prevailing mores, is an effective strategy; relying on police for the purpose is not, very. The melancholy results of game theory do not carry over into the real world in a clear way here. On paper, we may often be wimps. But, also, we will form vigilante committees, posses, Neighbourhood Watches – and learn to shoot, if need be. Peter’s real-world example of blood supplies is an interesting one, but I do not quite see how it bears on our more general discussion. Libertarianism, of course, says that each person’s blood is his own, and that we are not justified, prima facie, in compelling anyone to contribute. My addendum to libertarianism adds that it is nice if people do so and we should encourage that. (So far, Canadian practice agrees.) Contractarianism says, to start with, that we should have whatever programme we would all do best with if all comply. But will it underwrite a nonvoluntary programme? Or will it perhaps recommend switching from charitable donation to market supply, offering a price for blood, which varies according to demand? Note that both are compatible with libertarianism – unless the price paid for the blood is got by taxation, which would raise the first question again. I do not see how we could possibly know a priori which is best, and I therefore surmise that the contractarian will, as usual, contract for liberty. To be sure, this does not solve the problem that the ‘voluntary donation program does not collect enough blood fully to supply Canadian needs for all blood products.’ Neither libertarianism nor contractarianism, however, leads to the result that ‘we’ ought to satisfy everybody’s needs for everything, regardless. Neither endorses the view that we all have a positive right to life (as I call it). There may be

Jan Narveson: Responding to My Critics

227

coordination problems, and no doubt battle-of-the-sexes problems – familiar stuff. We would all like to be able to be sure of a transfusion if we need it, and most of us are willing to help supply the needed blood for this purpose; but most of us would prefer that the other guys be the donors. Right. But where is the argument for coercive supply? Danielson reports that many more people report intentions to donate than in fact donate, and that it is unclear whether advertising this fact would induce more, or less, donation. So we could use more research. But then he says, ‘Narveson’s reassurances that we already understand the core problems of ethics, and that they reduce to simple problems of rational choice and a dose of commitment to ethics, undercut the motivation for this work.’ How? My programme objects to the currently fashionable rejection of market supply. It points out that the market can generally be expected to optimize supply in relation to demand, so that if people are really serious about this, they ought to give that a try, on the face of it. But it does not dictate it (though it does dictate junking the Canadian health system). In all, I agree that real-world complexities defy simple solutions. I think that my general programme, rather than being refuted by such complexities, makes them unsurprising. Malcolm Murray Our editor’s chapter is much too interesting to discuss in a couple of pages, so I can only make a start. Rational Man, Malcolm notes, is ‘unwilling to abandon a straightforward maximizing strategy if the net utility of such a change is worse for [him] than forgoing such change.’ True. And ‘Nothing in the success of conditional cooperation tells us what kinds of agreements must be made.’ Also true. But if our standard envisaged situation is a PD, then by hypothesis both of us acting on the straightforward strategy is worse for both of us. And iteration strongly reinforces the cooperative strategy. Maybe the situation is not a PD. Hobbes thinks it is; Gauthier thinks it is; I am inclined to think it is – and the alternative assessment of the state of nature (such as chicken or battle of the sexes) still supports a similar conclusion. Libertarians do tend to assume that ‘centralized distribution’ is inefficient in our preferred (Pareto) sense. And that could, conceivably, be wrong. But if it is, then both contractarian and libertarian opposition to central distribution fail. (I note that most libertarians are minarchists rather than anarchists, which implies that they think that centralized distribution of something does not fail – roughly, the meting out of criminal justice – unlike welfare income.) So, yes, we need an argument. Malcolm says the libertarian has two independent ones for opposing the welfare state: Efficiency (E) and Pareto (P). But Efficiency here is Pareto efficiency, is it not? In the absence of interpersonally comparable utility, E = P! Is there an empirical issue here? When we philosophers analyse these things, we set aside ‘chemistry’. Maybe coercively extracted welfare payments somehow stimulate recipients to be more productive, so the contributors end up better off after all? That is logically possible, if marginally credible. But why does the burden

228

Liberty, Games and Contracts

of proof lie on the libertarian to show that it is always irrational for an agent to prefer a welfare system? Is it not the other way around? Welfare systems take money, without asking, from some and give it to others. This looks worse for victims and better for recipients – thus anti-Paretian. What needs rebutting is the prima facie obvious, not the other way around – no? Malcolm correctly sees me insisting on a Hobbesian starting point. From there I presume to move to another baseline (the libertarian one), which makes a further move to the welfare state anti-Paretian. Malcolm says there is a catch: ‘The same use of the Pareto principle from the state of nature will forbid us from moving into morality so long as there are some who believe they will benefit from amorality.’ Wrong, I think. General violence was the problem in the state of nature (S of N) – it makes us all worse off. Moving to peace is better for all. But peace is not giving you a quarter of my income. (Cynics might say that is the ‘peace that passeth all understanding’!) If peace is a good deal for all, then restarting a little war against the wealthy is not. Those who claim to be better off absent the state of society’s non-violence ask, in effect, for Hobbesian preventive warfare. This leaves the proposed aggressor dead, if we play our cards right. We are then back to my Pareto frontier, with the welfare state in the sub-optimal area. Now, maybe Hobbes is too optimistic. Viminitz, for example, will remind us of Ghengis Khan – didn’t he do all right? My somewhat meek response is to say that Ghengis was lucky, that the members of his army who died in the process did not obviously benefit, and that a trading agreement between his people and the people he killed really would have rendered both parties better off. In any case, aggression as a recommended social practice does not pay. Binmore cites the superiority of conditional cooperation (CC) under iteration as a ‘folk theorem’; and Malcolm’s own work seems to confirm that.6 Now Malcolm cites the ultimatum game. But does it apply here? Who in the S of N is offering whom shares of pie? I rather suspect that the analysis assumes that there is something to be ‘offered’ in S of N – apart from peace, which is not an ultimatum game commodity. Malcolm notes that ‘A contaminant exists in the fact that the pie is pre-owned’– but then suggests that this is not necessary. I disagree. The incapable, in a state of nature, are not offered nothing while the capable are offered something. Instead, the capable made the stuff, and the incapable did not. So anything the capable ‘offer’ to the incapable is sheer gain to them. Would they accept 0? Yes! For in the state of nature, the alternative to 0 is -100: in the S of N, we are all (we are supposing) in for a life of fear and early death, rather than diets of McDonalds instead of fine foods. 0 is way up from that. Producers will propose employment to erstwhile non-producers, if the latter can produce something, and that will be to the good. So where is the analogue to the ultimatum game? It is, indeed, just a game. But the S of N is not a game: it is a matter of life and death. In real life, 9-for-1 splits happen all the time. But these are divisions of cooperative surpluses where the people who get 9 have done things to make 9 a plausible offer. (You can try to make your movie without the services of Julia Roberts, and pay some fledgling a pittance – but when the movie will not generate enough income even to cover expenses, while with Julia it will net $50 million, 9-

Jan Narveson: Responding to My Critics

229

for-1 is a bargain.) That example, by the way, is quite sufficient to explain why we should be ignoring discussions of ultimatum games in abstracto. Everything about distribution that matters is not in abstracto, but in the concrete world where alternatives matter and can be estimated. I have previously noted my rejection of MRC in my discussion of Tucker. Participants in bargaining games can settle wherever they please – so says the Lockean Proviso (aka libertarian principle). Once settled, society can insist that they keep the agreement made, but society simply has nothing to say about how they should settle, short of avoiding force and fraud in the negotiation. Paul Viminitz The Lockean Proviso is not the ‘baseline for bargaining’ in the foundations of morals: it needs justifying, like everything else in morals. Paul and I are agreed that our fundamental conception must be Hobbesian: we start with no morals at all. The question is whether it might be a good idea to have some. We all have much to lose at the hands of our fellows, and also much to gain; and we want to maximize gains relative to losses. But that requires cooperation. How much? That is the main question. In Viminitz’s view, subsets of gangsters (or politicians?) may do better to exert themselves, cooperating considerably with each other and against some, many or all others. No doubt he is right about that. But that is not enough to demonstrate the superior rationality of having no morals, or of having morals that encourage and support uncooperative behaviour when it happens to suit this or that person. Such people, I think, declare themselves enemies of the rest, and should be treated accordingly. Is that a judgment with which gangsters (and politicians) should agree, rationally speaking? My claim is that it is. But how much is there in it? Hobbes declared the Laws of Nature, which call for cooperation with all and only cooperators, to be ‘eternal and immutable’. For all that, he thought we needed the state, because he thinks that the Laws bind only ‘in foro interno’. I take Viminitz to be updating the same message. My reply to both of them is that morals is a sort of social ‘meme’, in the modern lingo, and one that we all should install, in ourselves and others. The gist of it is that we should be seeking peace, and are justified in making war only against those who make it against us. If there is a modification to that finding being proposed by Paul, what is it? Presumably, it is something like ‘Make war whenever it serves your interests.’ If this is a moral message, that means that we are all supposed to applaud enthusiastically at this prospect. Sorry, but I do not see it. If we know in advance that that is how you are going to operate, then, since we can all make war if that is what is needed, we are all motivated to turn to weapons-making. In the process, most of the gangsters will be eradicated – but at great cost to the GNP. Both the gangsters and we would do better if there were no gangsters. It might be thought that Hobbes and I advocate a bourgeois morality, and in a way that is true. But we think that being bourgeois is really better than being dead. Those who do not care about that are indeed a problem, and, as Paul has observed,

230

Liberty, Games and Contracts

we are at war with those people. How much we should invest in containing them is a good question, indeed. One way we can sometimes accommodate, to be sure, is by knuckling under. Slaves may do better to continue slaving away than to revolt. If they all rebelled, and in particular if they all preferred death to slavery, the institution of slavery would not exist. But they didn’t, and it did; and that is not too surprising. Still, this hardly shows that slavery is okay. It shows that the slaves may have little choice, realistically. (Or it may show that they actually are better off being slaves, as Aristotle claimed. But that does not look very plausible.) But it does not – so I think – give people in general reason to side with the masters, or, more specifically, to side with the institution wherein some are slaves and some are masters. But why not? I surmise that it is especially because people came to see that slavery is inefficient. Slavery as a social system is sub-optimal: many people waste time compelling slaves to work, when paying for their services would get better results and free up those who hitherto wasted their time patrolling against escapes, whupping non-cooperators and so on. Prima facie, all would do better if they all worked, either together or separately, but all voluntarily. What I deny is that we will do better by approving of violence in this, that and the other case other than self-defence, and this seems to me the most crucially relevant point for moral and political philosophy. I claim we will do better to approve of violence only as a means of self-defence of the innocent against the aggressive, in all cases. That includes defending the rich against those of the poor who engage in drive-by shootings, as well as defending the poor against the rich, should the latter foolishly decide they would do well to enslave the poor. All this assumes, to be sure, that we can tell when someone is being violent and when not. Paul professes to deny this. He says, in comment on my proposal that he who ‘interferes’ is in the wrong, ‘But now all the work will be done by our answer to, What counts as interference?’ True. I think that interference is interference; not ‘that which somebody or other says you ought not to do’. Is this so difficult? I will accept that in some cases it can be tricky. But if Paul thinks there is a problem about tagging the drive-by killers as ‘interfering’ with the people they shoot, then I am hard put to know what to say. He may think that there is a real problem in explaining why we should not sympathize with the shooter. I do not find that hard, myself – and since I doubt that most drive-by shooters would relish the prospect of being a victim of same, I also think the case on this is really closed. The Hobbesian account is that Reason tells us to do what is necessary to preserve our lives – that is to say, the lives that we want to live, which is not necessarily just the maximum duration of biological life. But, says Hobbes, the state of nature is one of equality in respect of who can do whom in. So your deciding to proceed by threatening my life can be countered by my doing likewise with you; and then we have our choice between both ending up soon dead in battle, or both declaring peace, where peace means the absence of war. There is nothing about how well off our erstwhile warrior might be at the onset of peace, with an obligation on the part of the better-off one to share his wealth with the other. Rather, there is just an agreement to refrain from getting our way by fighting and the threats attendant thereto.

Jan Narveson: Responding to My Critics

231

Paul quotably claims that ‘The history of the human race just is a history of our willingness to risk violent death in exchange for a chance to increase our standard of living.’ Among the responses this invites is that it is also a history of failure in that regard, as compared with groups that have managed to maintain peace. The groups who have enjoyed enormous increase in their standards of living are those which have allowed the inventive and hard-working to do well – not those who proceed by enslaving the productive. Still, Paul insists that ‘the threat to return to a state of war need not be disingenuous. And so the worsening effects of war are of a piece with the bettering effects of cooperation’ [Paul’s emphasis]. I disagree with this, except on one interpretation. Mutual non-interference, aka peace, is cooperative. In that sense what Paul says is true by definition. But beyond that, it is not. Let us suppose that our capacity for making each other dead is about equal, as Hobbes claims. But let us also suppose that your capacity to make widgets is far greater than mine. Then what? The libertarian view has the merit of insisting that, as it stands, this is indeterminate. Any widgets you can make entirely on your own are yours, of course. But when we cooperate to make widgets, any number of specific distributions could be reasonable. Such was the burden of my response to Malcolm. I still think I am right about that. When A produces item G and B does not, there is no ultimatum game to play. B is not in a position to ‘say no’ to a proposal to divide 10–0, let alone 9–1 instead of 5–5. Paul holds that the right baseline is what we can each bring to the table. I agree. I think that we bring ‘our guns’ along with a perception of the other chap’s guns along with a sense of our capacity to produce guns if it should come to that. There is also the question of what we are doing at the table anyway, and who is at it. Modern political rhetoric has us sitting down to decide how we shall divide up a ‘social product’. But there are no ‘social products’; everything is made by somebody in particular. Well, wait, for there is one ‘product’ which really is social: peace. It is a funny product, since what it mainly requires is doing nothing. Now, Hobbes and I think that there is just one correct distribution of peace: equal. Everyone refrains from aggressing against anyone. The correct terms of trade are: I don’t hit you if you don’t hit me. Most modern writers have come to think that some people get to say, not this, but rather: ‘I won’t hit you provided that you cure me of cancer as well as refraining from hitting back.’ What possible basis is there for this? In the end, I think, none.

Property and Liberty Ann Levey Levey argues that I am wrong about the connection between liberty and private property. Her main reason is that property ‘is a system of restrictions and permissions that control access to and use of external goods’. Thus ‘For any item x,

232

Liberty, Games and Contracts

a property right in x restricts the access to x for anyone other than the owner; that is, it restricts liberty.’ Well that is, of course, true by definition. All rights restrict in some way the liberty of somebody or everybody else – that, indeed, is their point.7 So it is a bit startling to see her saying in the very next sentence, ‘This is a claim with which Narveson disagrees’! What is at issue is, of course, not that, but rather whether the right to liberty entails the right to property. Is the ‘pattern of restrictions’ – the ‘distribution’ of liberties – called for by the general right to liberty the same, or does it contain as a proper part those involved in the right to property? I say, Yes. Or more precisely: things acquired in ways that meet the initial (Lockean) conditions are indeed the property of those who acquire them, for precisely the same reasons that we have the general right of liberty, and as a straight implication thereof. My having a Macintosh entitles me to use it and forbids you to take it. It does not forbid you to make one yourself, if you can, or buy one from somebody who makes them, if you can. Your being forbidden to take this machine when you want is precisely what my right to use it consists in. It is also what its being my property consists in. We cannot have a right to liberty at all without a right to property in the things we use when we exercise our liberty. There cannot be a ‘general right of liberty’ of the type Hobbes calls the ‘right of nature’: if anyone can do just anything he or she pleases, at any time, to anybody, as can all others, then nobody has any rights at all. In order to talk of rights, we must talk of restrictions. A right to liberty is a right such that other people’s liberty to interfere with the exercise of it must be restricted. And, arguably, people being as they are, a society in which people are maximally and generally free is one in which those restrictions are respected. They are exactly the same restrictions that define property – in oneself, first, in what one does with oneself insofar as that is compatible with their liberty, and (therefore) in the use of products of one’s labour insofar as those were intended to be so used by the maker. Now, Levey finds one thing rather puzzling, and I agree it needs further explanation. A thing is not yours if you got it by using force against others; but if you didn’t, then it is. In the latter case you can correctly point out that you did not get it by restricting – that is, by imposing restrictions on – the activities of others. Now Levey says, ‘before there was no restriction on others using the external goods and now there is. … [O]nce it has been claimed, if A exclusively owns it, then there is an interference with liberty that would not exist if A did not exclusively own it.’ Is that both true, and also a point against libertarianism’s account of property? No. Let us take the two sorts of cases that are relevant here: (a) the hypothetical occupancy of previously unoccupied areas of, say, land, and (b) where you make the thing in question, there having been no such thing before. Levey says that an oddity lies in this: that ‘before there was no restriction on others using the external goods and now there is’. But I deny that. My coming to acquire x out of a state of nature does not deprive someone of a liberty he previously ‘had’ (that is, had a right to). We all have (a right to) the liberty to go forth and acquire whatever we can, that is, what has not already been acquired, if we so wish. That is the same always and everywhere; my acquiring something does not revoke that; it does withdraw from

Jan Narveson: Responding to My Critics

233

the region of items that you might possibly acquire by that means the thing I acquired. But you did not previously have the right to use those; you had, only, the right to use them if not already taken. As Hillel Steiner points out with respect to any given item in the world, if A has it, then B doesn’t. It therefore makes no sense to say that people are being ‘excluded’ by some principle from general possession of everything, that being impossible.8 What is possible, however, is to organize things such that each of us is in a position to acquire. The obvious and optimal principle enabling that is to allow people to acquire what is not already so acquired. This disenables no one from using things, if they are things they already have or which they acquire with no violence to others. Nor do we disenable anyone from using anything that everybody could have, since there are no such things. Do we disenable people from having cooperatives, and sharing a lot of stuff? Of course not: all you have to do is get the voluntary cooperation of fellow participants. No problem. (Advocates of many kinds of socialism seem to ignore this last restriction.) What about things you made? Well, since the things did not exist before you made them, the claim that they were ‘free’ to use them before fails for lack of reference in any case. ‘Do property rights in oneself generate property rights in external things?’ she asks; and suggests, ‘It might be thought that if one is at liberty to do whatever does not violate others’ rights of self-ownership and acquiring private property does not violate such rights, then one is entitled to acquire private property.’ Since that sentence is true by definition, that does seem very reasonable indeed! All rights are rights to act; there are no separate rights to things. Acts always involve using something(s) or other, and always in particular using some part of what we call ‘oneself’. ‘Self-ownership’ is, simply, the right to liberty. Now: is the ‘extension’ of this idea to ownership of other things in the world really an extension? Is there a difference between using my hand and using the hammer that I hold in it? Well, of course the hammer was not an initial part of the package which is me, whereas I am that package, and so ‘got there first’ in the most basic way possible. Nevertheless, I can use things just as I can use parts of myself: in ways requiring their being around tomorrow at certain places where I put them, and so on. Ownership of things external to the self stems from that. So it is disingenuous to go on to say, as Ann now does, ‘Once we talk about forcing people to do things, rather than interfering with liberty, it becomes clear that failure to recognize acts as acts of appropriation does not require force; it merely requires disregard.’ Suppose she said this about your head: ‘Oh, I wasn’t using force against it, I just acted as if it wasn’t there! No problem!’ The ‘disregarding’ of what people are in fact doing is equivalent to (and often actually involves) the use of force in that it deprives the person of his or her use of those things in the ways he or she wanted; and that, after all, is precisely what – as Levey herself notes – it is all about. Using force if necessary to prevent people from doing such things is exactly what the general right to liberty calls for. In short: if there is this general (enforceable) right of liberty, then there is the (enforceable) right to use whatever one can find that is not already used by others. QED.

234

Liberty, Games and Contracts

Property rights, acquired the right way, do precisely ensure that no one’s right to do anything is got at the expense of anyone else’s right to do whatever he or she is able to do without violating the liberty of others. Ownership of everything by everyone is, as noted, nonsense; and so the fact that rights ‘distribute liberties’ is not a restriction on rights to do particular things that were previously possible. So this is wrong: ‘Granted the things would not exist without private property. But that is an issue of productivity not of liberty’, and so an issue that ‘Someone [who is] concerned only with liberty’ is not entitled to help himself to. That private property enables people to be productive – especially by allowing them to remain in control of the things they produce – is among its prime virtues, of course. But issues of productivity are issues of liberty, just as they stand, since producing is a thing we can do, and want to do. What makes productivity desirable is that the things produced are things people want. The issue of liberty is the issue of whether people may do various things they wish to do, among which are productive things. If they cannot have the things they make, that cuts into their liberty; of course it also reduces the kind of productivity that we want. That is why liberty is intrinsically crucial to property. So I am unrepentant about those claims. Duncan MacIntosh Susan Moller Okin thinks that on the Locke/Nozick principles of ownership, mothers own their children, who therefore are their slaves and thus cannot have the famous right to liberty which those authors are celebrated for affirming. How so? Well, because parents produce those children and we own what we produce. Now, that famous principle says that all persons with rational minds and wills are entitled to general liberty, which is what generates ownership (as I explain in discussing Levey). Ownership, then, cannot trump the principle, and Okin’s argument is absurd. Insofar as children are not rational minds and wills, indeed, they can be ‘owned’ – we do speak of ‘this person’s children’. But then they grow up into rational adults – who, by hypothesis, have the fundamental right to liberty, severing all previous links from childhood, insofar as enforceable claims are concerned. We should also deny that children are ‘made’ by their parents (as distinct from influencing them quite a lot). What parents do is initiate a process that eventuates in human organisms, which grow up. No mother can ‘make’ any such thing, as Locke and Nozick were well aware. So the claim that parents own their children because they make them is also wrong. But they are entitled to take such charge of those children as may be taken, by virtue of being the initiators of the process that led to those particular ones. This gives them a claim better than anyone else’s to being the ones to administer to them when young. However, as Locke pointed out, we have no absolute authority over children. All we have is the right to be the persons responsible for their care, rather than anyone else – barring the usual possible upsetters such as death, separation or incompetence. This is where the power of the social contract idea comes to the fore. The social contract is prior, not posterior, to any idea of ownership or anything else. It is not

Jan Narveson: Responding to My Critics

235

prior to but consequent upon rationality, of course, and every rational being is possessed of assorted interests which guide that person’s activities. An attachment to one’s parents to the point of acknowledging them as one’s owners and masters is not among those. Since it is not, the basis for assembling property rights provides no reason to expect such ownership to emerge – and, of course, it does not. Peter Vallentyne The Lockean Proviso (LP) is a very unclear idea. David Gauthier identifies it with the thesis that no one is to promote or seek his own utility by imposing disutility on others – thus identifying the LP with the general libertarian thesis (and therefore with the view that agents ‘fully own themselves’, as I explain in The Libertarian Idea). On this understanding of the Proviso, the claim that ‘the initial status of external resources as unowned precludes any role for a Lockean Proviso’, which Vallentyne ascribes to me, makes no sense. He must, then, intend some ‘stronger’ understanding of the Proviso – one, I presume, of the several I examine and reject in my article on the subject.9 That is to say, the LP is understood now to mean that when A appropriates anything in a state of nature by doing x, even if x meets the libertarian condition that no one else is thereby negatively affected in respect of what he already has, x is nevertheless illegitimate because those others have to be ‘left with’ more – presumably, then, more to appropriate. I do indeed use such an argument against all of those more robust interpretations; so if it is a bad argument, that is indeed bad news for me. So, let us see. On any libertarian view, an agent does wrong if he interferes with the courses of actions of others, insofar as these are compatible with others’ liberty. The ‘acquisition’ of x by A consists in A’s doing something with x which makes x A’s property: that is to say, making them such that A’s use of x is to be recognized, in future, as such that x is not available to others without A’s permission. To acquire what is previously not owned by anyone, then, one must perform some actions in relation to it which will have that effect. What might those be? I take two conditions to be essential: first, a historical action making one the effective first-user in this respect; and, second, an intentional component, clarifying and somehow publicizing that one is intending and expecting to engage in repeated use of the item in question for the indefinite future. Peter, as I understand it, would accept some such conditions for that purpose; so the question is whether they are indeed jointly sufficient to recognize x as A’s property – that is to say, A’s right to continued use of x, by virtue of the libertarian principle (which is assumed for the sake of analysis.) It is unfortunate that Vallentyne does not further consider the thesis of Israel Kirzner, that first possessors of property are in effect creators of resources.10 I suspect we would find it to be crucial. However, let us just consider what Vallentyne does say. There are two issues: ‘(1) The problematic case for libertarianism is appropriation without the consent of others. (2) Does someone commit an injustice if … she treats a previously unowned thing as if she had private property rights over it’? I am puzzled, for these two issues seem to be simply the same issue. Libertarianism is fundamentally a theory of justice, holding that no one

236

Liberty, Games and Contracts

may be forced to do anything except what is relevantly harmful to others; but consent (if uncoerced) assures the latter. (Vallentyne says, ‘Throughout justice is understood as non-violation of rights.’ But that is not a way of understanding justice. I take it that my specification – deriving from Mill, which identifies justice as that which may be required or compelled of others – is non-circular, and that Vallentyne agrees with it.) Regarding (1), libertarians might claim that when the non-harm condition is met, then consent of others is rational and thus may be assumed. If (2) is right, though, they will not consent, of course. So, is it? Vallentyne now wants to distinguish two cases: where A has item x in A’s physical possession at t, when B arrives and attempts to disengage A from x; and when A does not have x, but makes a claim to x by meeting my two conditions. In the first case, says Vallentyne, A’s self-ownership would be violated by B and all is well with the libertarian thesis. But what about the second case? For example, A has staked his claim to an apple tree, hence to its apples; but B proposes to take one, without asking. On this point, Vallentyne observes that libertarians in effect subscribe to what he calls a ‘very weak immunity’ condition for ‘losing their liberty rights to use unowned objects’. This is objectionable. What is in question is whether what people are ‘losing’ was indeed a right. The terminology of ‘losing’ implies that it was, and so is now violated. But I deny this, as explained in my discussion of Levey. Notice that if there is a tree at point p, previously unused, and B helps himself to an apple from it, he cannot be said to violate anyone’s liberty; so of course libertarianism awards B the right to do it. But that is not so with the tree in our example, for A has indeed previously used it, intends to continue so using it, and has suitably publicized the fact. Do A’s previous activities along that line count for nothing regarding the status of the tree in question? How could Vallentyne think so? He claims that what is in question here is the status of a ‘second-order right’ which he parenthetically defines as ‘(the conditions under which the first-order status can change)’. But I hold that the first-order right is all there is here, and that there is no such thing as a ‘second-order right’ to reorder our first-order rights. The claim he makes, that there are an infinite number of possibilities in this regard, is of course true if it is about what is, absolutely, logically possible to claim. But it is not true as a claim about the logical structure of libertarianism, which already is a theory about how first-order rights (in this case, to own particular things) are acquired. The libertarian idea about this is pretty clear: people are entitled to do as they please so long as they do not thereby disenable others from doing likewise – where ‘likewise’ refers, in the case of actions involving use of things outside our own bodies, to using whatever they can that is not already in use. If Jones uses x and nobody else previously is doing so, he meets the latter condition. As long as he continues to use x, others are not free to use it because they are no longer ‘there first’ – Jones is. This is, indeed, the maximum liberty compatible with a like liberty for all. Kirzner’s thesis is generalized effectively by David Schmidtz: private ownership

Jan Narveson: Responding to My Critics

237

of the means of production is a good to others.11 It is to my advantage that you are able to own things, even things that I would like to have myself if I had the chance. Ownership enables the kind of production that enables other people to improve their lives, by free exchange. And free exchange is the overwhelmingly main way in which we all advance in life. Ownership is a prerequisite to this happy condition. All of which also adds to the explanation of why we may impute rational consent to the institution of private appropriation. Grant Brown As Brown elegantly sets out in his paper, ‘indiscriminate or non-strategic altruism is bound to be dysfunctional; it is a false ideal. The upshot is not that beneficence to strangers is always to be discouraged; rather, the lesson is that beneficence must be strategically sound.’ Brown summarizes the situation thus: ‘The complex cluster of capacities and practices we identify as an agent’s morality, in other words, is a strategy whose function is to promote the agent’s interests through realizing the advantages of group living. To put the point even more succinctly, the function of morality is to facilitate mutually advantageous cooperation.’ That seems to me to touch at every point on various themes which I have over the years insisted upon, but I have been much less precise about their underlying logic. His is, indeed, a superb paper – and one with which I can find no fault at all. That morality must be an ‘evolutionarily stable strategy’ (or ESS) is such a persuasive idea that I am tempted to think that I did think of it; but, alas, I must admit that I did not. So I will have to settle for being indebted to Brown for developing it so persuasively. Does his argument establish that this correct ESS is indeed the libertarian idea? It does seem to me to come very close indeed. The question is whether there is a residue of strategically sound semi-altruism, as it were. Regarding the welfare state, say, perhaps there is a sort of surcharge on our normal transactions which would support the welfare state and be a part of the ESS long-run profile. I cannot profess to the level of game-theoretic expertise that might settle that one, though my sense is that it will not. Sheldon Wein Utilitarianism is the thesis that the principle of utility is the sole fundamental principle of morals, this being what Wein calls ‘fundamentalist’. When I wrote Morality and Utility (1967), I subscribed to that; but I also was sensitive to the question of its foundations, and thought I had them figured out. I was wrong about that – as work by David Gauthier persuaded me. (I could use epiphanist language to describe that experience, except that it took quite a bit longer than St Paul’s – as well as being in rather a different direction from his.) But there are two different senses of ‘fundamentalist’ relevant here. In the other one, it would be the view that utilitarianism is an ultimate truth with no further explanations or foundations. That makes the principle of utility into a moral intuition, as Sidgwick does, in effect. I

238

Liberty, Games and Contracts

was never a fundamentalist utilitarian of that kind – one of the main reasons I wrote Morality and Utility. Still, the right foundations of moral theory should affect one’s choice of fundamental principles. I came to think that the contractarian idea is, as Gauthier once put it, ‘the only game in town’. I also came to think that the principle of liberty made sense as the fundamental general principle of justice for all – on contractarian grounds. It is worth recapping briefly why I think so. Morals as we understand it here is essentially a regime of interpersonal behavioural control. Its principles are notionally addressed to all – all are asked to defer to those principles when conflict requires something of the kind. What should guide the choice of such principles is that, given what we and our situations are like, we can reasonably expect acceptance. We all have various interests, which somehow yield outputs regarding what to do – deliberative verdicts, decisions. So too do others, and sometimes if A is to get what A wants, B will not be able to get what B wants. The stage is then set for a fight, leading to sub-optimal results. A cooperative outcome, in which all do better than they would in a fracas, is almost always possible; and when it is, it is rationally preferred. Now, the most obvious and conspicuous of our general interests is that in not having one’s disutility promoted. What we want is more, not less, of what we like; and we only settle for less if we cannot do better. A general right of liberty – that is to say, the right to do as we want – would be nice if we could have it. And we can have it, if others are willing. Are they? It is irrational for anyone to accept less than general liberty if he can help it – and if Hobbes is right, he can ‘help it’. Okay: but why not insist on more – not just refraining from injury but pitching in and helping? The answer is that if the ‘more’ is compelled, then it is not ‘more’ for the coerced – so, no deal. Of course we should all be in favour of helpful behaviour – but not of coercively extracting it. The costs of coercion are too high. But proscription of harm and injury do not have that problem. Here compulsion is eminently rational, and worth the price, which is merely refraining oneself. The cost–benefit calculation favours the pure liberty principle, keeping levels of helpfulness at the level of individual judgment and good-heartedness. Sheldon characterizes contractarianism in ways that will not do. First, the social contract need not only concern justice; it is about morality. Justice is what may be coerced, but there is more – there is what is virtuous, what is recommended. In the second place, Rawls’s importation of moral considerations as inputs to the social contract eviscerates the whole idea. The point of the social contract is to go from what was not moral to what is: that is to say, to provide a genuine derivation, which is also an explanation, of morals. You have fumbled the ball if you help yourself to morals at the start. And finally, the metaphor of the ‘veil of ignorance’ is misleading. Persons genuinely behind the veil have no basis for choosing anything. To choose, you need interests, and all interests are someone’s interests. But when the point of the project is to agree – which we want in order to get benefits only available from it – there is no possibility of what concerned Rawls so much: fudging the output to appeal to some at the cost of others. For all are to ‘sign’; thus the veil is dispensable, and if it were not so, it would render the whole project unintelligible.

Jan Narveson: Responding to My Critics

239

I have elsewhere recorded my objections to Rawls’s derivation of two different and mutually incompatible fundamental principles of justice.12 Either we go for liberty, in which case accepting a further constraint to shore up the ‘bottom class’ is not possible; or we are egalitarians from the start and can just forget about liberty. Even the quick sketch above shows why the socialist alternative is irrational. We are left with libertarianism as the sole generally rational option. I take this to yield what Wein calls ‘Market Contractarianism’. Market liberty is the output of this theory. We contract as best we can, for the best we can get, given that we must all get it. We emerge with liberty because that is the best we can all get. And I take this to follow from ordinary facts about who people are – not from cheating by building our results into our premises. (So I continue to think, even in the face of the challenges mounted by Danielson, Viminitz and Murray.) It is one of the shortcomings of contemporary philosophy to fail to appreciate how absurd are schemes of ‘social distribution’. The goods in life are made by people, not by nature, which has nothing to say about anything. ‘Distribution’ of those goods can only consist in wholesale robbery if it means anything more than that people are to be allowed to deal with each other in mutually beneficial ways, with no limits on the terms that might be offered. There is no such problem as ‘how to divide the social surplus appropriately’, because, as Nozick points out, it comes into being attached to its makers – namely, us. People make things, almost always in concert with various others, and how to allocate incomes in the process is an intricate problem for those concerned – but not one that social philosophers can address in advance. Utilitarianism can be made to look good only by hokey manoeuvring. I do not, then, see much promise in Sheldon’s proposal that ‘prospects for reconciling market contractarianism with fundamentalist utilitarianism are better than has usually been supposed’. I think, in fact, that they are much worse. It has been supposed that utilitarianism is a real theory that makes sense. But it is not. To be blunt about it: neither of us necessarily cares about the other’s utility, as such, and so an idea that everybody’s utility ‘counts equally’ is unmotivated. We of course should have a positive view of all other humans, insofar as their behaviour is compatible with general liberty. But that is far from saying that we should regard their utility as both commensurable and, when commensurated, equal in weight to our own. Cooperation gets us everywhere. But acting as if we were merely so many different cells in one big glop of humanity gets us nowhere. Separate individuals cooperate, each with their own sets of interests; and the result of the cooperative attitude is that we all do better. That, I claim, is why libertarianism offers the best general view of morals. Non-violence and cooperation are sine qua nons of the social game. Beyond that, there is life, including the social life of particular societies, with their infinite variety of particular forms of behaviour, tastes, skills and the rest. Good lives of no end of kinds are possible, and supposing that we can helpfully measure the goodness of each with a view to ‘overall maximization’ is an illusion.

Liberty, Games and Contracts

240

Notes 1 2 3 4

5 6 7 8 9

10

11 12

Jan Narveson, The Libertarian Idea (Philadelphia, 1988; republished Peterborough, Ontario, 2001). See Lewis Carroll, ‘What the Tortoise Said to Achilles’, Mind, 4 (1895): 278–80. Jan Narveson, Morality and Utility (Baltimore, 1967), 89–90. Jan Narveson, ‘The Drug Laws: More Nails in the Coffin of American Liberalism’, in Jan Narveson, Respecting Persons in Theory and Practice: Essays on Moral and Political Philosophy (Lanham, MD, 2002), ch. 14, 243–64. Richard Brandt, ‘Some Merits of One Form of Rule-Utilitarianism’, in S. Gorovitz (ed.), Utilitarianism, with Critical Essays (Indianapolis, 1971), 324–44. See, for example, Malcolm Murray, The Moral Wager (Dordrecht, 2007), ch. 5. ‘[T]o have a right is … ipso facto for the liberty of others to be limited.’ Jan Narveson, The Libertarian Idea, 50. Hillel Steiner, ‘How Free: Computing Personal Property’, in A. Phillips Griffiths (ed.), Of Liberty (Cambridge, 1983), 73–90. Jan Narveson, ‘Property Rights: Original Acquisition and Lockean Provisos’, Public Affairs Quarterly, 13/3 (1999): 205–27. Reprinted in Respecting Persons, ch. 8, 111–30. Israel M. Kirzner, ‘The Nature of Profits: Some Economic Insights and Their Ethical Implications’, in Robin Cowan and Mario Rizzo (eds), Profits and Morality (Chicago, 1995), 22–47. See David Schmidtz, The Limits of Government: An Essay on the Public Goods Argument (Boulder, CO, 1991). Jan Narveson, ‘A Puzzle about Economic Justice in Rawls’ Theory’, Social Theory and Practice, 4/1 (1976): 1–27. Reprinted in Respecting Persons, ch. 2, 13–34.

Bibliography Ahmad, Rana, Zosia Bornik, Peter Danielson, Hadi Dowlatabadi, Edwin Levy, Holly Longstaf and Jennifer Wilkin. ‘A Web-Based Instrument to Model Social Norms: Nerd Design and Results’, Integrated Assessment, 6/2 (2006). Alexander, Richard. The Biology of Moral Systems (Berlin: De Gruyter, 1987). Annas, Julia and Jonathan Barnes. The Modes of Scepticism: Ancient Texts and Modern Interpretations (New York: Cambridge University Press, 1985). Arrow, Kenneth J. ‘Gifts and Exchanges’, Philosophy & Public Affairs, 1/4 (1972): 343–62. Aumann, Robert. ‘Correlated Equilibrium as an Expression of Bayesian Rationality’, Econometrica, 55 (1981):1–18. Aumann, Robert. ‘Subjectivity and Correlation in Randomized Strategies’, Journal of Mathematical Economics, 1 (1974): 67–96. Axelrod, Robert M. The Evolution of Cooperation (New York: Basic Books, 1984). Axelrod, Robert and D. Dion. ‘The further evolution of cooperation’, Science, 211 (1988): 1390–1396. Becker, Lawrence. Property Rights: Philosophical Foundations (Boston: Routledge and Kegan Paul, 1977). Benson, Paul. ‘Freedom and Value’, Journal of Philosophy, LXXXIV/9 (1987): 465–86. Benson, Paul. ‘Autonomy and Oppressive Socialization’, Journal of Social Theory and Practice, 17/3 (1991): 385–408. Bentham, Jeremy. The Principles of Morals and Legislation (Buffalo, NY: Prometheus Books, [1781] 1988). Bigelow, Robert. The Dawn Warriors: Man’s Evolution Toward Peace (New York: Little, Brown, 1969). Binmore, Ken. Game Theory and the Social Contract: Vol. 1: Playing Fair (Cambridge, MA: MIT Press, 1998). Blackburn, Simon. ‘Moral Realism’, in Simon Blackburn, Essays in Quasi-Realism (New York: Oxford University Press, 1994), 111–29. Bond, E. J. ‘Could There Be a Rationally Grounded Morality?’, Journal of Philosophical Research, XV (1990). Bond, E. J. Ethics and Human Well-being (Oxford: Blackwell, 1996). Boyd, R. and J. Lorberbaum. ‘No Pure Strategy is Evolutionarily Stable in the Repeated Prisoner’s Dilemma Game’, Nature, 327 (1987): 57–8. Boyd, Robert and Peter Richerson. Culture and the Evolutionary Process (Chicago: University of Chicago Press, 1985). Boyd, Robert and Peter J. Richerson. ‘Norms and Bounded Rationality’, in Gerd Gigerenzer and R. Selten (eds), Bounded Rationality: The Adaptive Toolbox (Cambridge, MA: MIT Press, 2002), 281–96. Brandt, Richard. A Theory of the Good and the Right (Amherst, NY: Prometheus Books, 1998). Brandt, Richard. ‘Some Merits of One Form of Rule-Utilitarianism’, in S. Gorovitz (ed.), Utilitarianism, with Critical Essays (Indianapolis: Bobbs-Merrill, 1971), 324–44. 241

242

Liberty, Games and Contracts

Braybrooke, David. ‘Utilitarianism with a Difference: Rawls’s Position in Ethics’, Canadian Journal of Philosophy, 3, December (1973). Campbell, Richmond. Self-Love and Self-Respect: A Philosophical Study of Egoism (Ottawa: Canadian Library of Philosophy, 1979). Canadian Blood Services. World Blood Donor Day Survey Underscores ‘Reality Gap’; Canadians Exaggerate Intentions to Donate, Overestimate the Number of Blood Donors in Canada, 14 June 2004, at http://www.bloodservices.ca/centreapps/internet/uw_V502_ mainengine.nsf/761d50021d78979185256af700577ec6/68c8218a06a544ab85256eb000 692942?OpenDocument. Carroll, Lewis. ‘What the Tortoise said to Achilles’, Mind, 4 (1895): 278–80. Cavalli-Sforza, Luigi Luca and Marcus Feldman. Cultural Transmission and Evolution (Princeton: Princeton University Press, 1981). Cave, Eric M. Preferring Justice: Rationality, Self-Transformation and the Sense of Justice (Boulder, CO: Westview Press, 1998). Centola, Damon, Robb Willer and M. W. Macy. The Emperor’s Dilemma: A Computational Model of Self-Enforcing Norms (Ithaca, NY: Cornell University, 2002). Cohen, G. A. Self-Ownership, Freedom and Equality (Cambridge: Cambridge University Press, 1995). Cohen, Joshua. ‘Okin on Justice, Gender and Family’, Canadian Journal of Philosophy, 22/2 (June 1992): 263–86. Coleman, Jules L. Markets, Morals and the Law (New York: Cambridge University Press, 1988). Coleman, Jules L. Risks and Wrongs (New York: Cambridge University Press, 1992). Danielson, Peter. Artificial Morality: Virtuous Robots for Virtual Games (New York/London: Routledge, 1992). Danielson, Peter. ‘The Rights of Chickens: A Rational Foundation for Libertarianism?’, in J. T. Sanders and J. Narveson (eds), For and Against the State: New Philosophical Readings (Lanham, MD: Rowman & Littlefield, 1996), 171–93. Danielson, Peter and Chris MacDonald. ‘Singer’s Agenda for Practical Ethics’, Dialogue 35 (1996): 599–610. Danielson, Peter (ed.), Modeling Rationality, Morality and Evolution; Vancouver Studies in Cognitive Science, Volume 7 (New York/Oxford: Oxford University Press, 1998). Danielson, Peter. ‘Evolutionary Models of Cooperative Mechanisms: Artificial Morality and Genetic Programming’, in Peter Danielson (ed.), Modeling Rationality, Morality, and Evolution (1998), 423–41. Danielson, Peter. ‘Competition among Cooperators: Altruism and Reciprocity’, Proceedings of the National Academy of Sciences, 99 (2002): 7237–42. Danielson, Peter. ‘Playing with Ethics: Games, Norms, and Moral Freedom’, Topoi, 24 (2005): 1–7. Danielson, Peter, Rana Ahmad, Zosia Bornik, Hadi Dowlatabadi and Edwin Levy. ‘Deep, Cheap, and Improvable: Dynamic Democratic Norms & the Ethics of Biotechnology’, Journal of Philosophical Research (forthcoming). Darwall, Stephen. Impartial Reason (Ithaca, NY: Cornell University Press, 1983). Dawkins, Richard. Climbing Mount Improbable (New York: Viking, 1996). Dawkins, Richard. The Selfish Gene (Oxford: Oxford University Press, 1976, repr. 1989). Dimock, Susan. ‘Personal Autonomy, Freedom of Action, and Coercion’, in S. Brennan, T. Isaacs and M. Milde (eds), A Question of Values: New Canadian Perspectives in Ethics and Political Philosophy (Value Inquiry Book Series) (Amsterdam: Rodopi Press, 1997), 65–86.

Bibliography

243

Dimock, Susan. ‘Defending Non-Tuism’, Canadian Journal of Philosophy, 29/2 (1999): 251–74. Dimock, Susan. ‘Two Virtues of Contractarianism’, The Journal of Value Inquiry, 37/3 (November 2003): 395–414. Duncan, Craig and Tibor R. Machan (eds), Libertarianism: For and Against (Toronto: Rowman & Littlefield, 2005). Dworkin, Gerald. ‘Autonomy and Behavior Control’, Hastings Center Report 6 (1976). Dworkin, Gerald. ‘The Concept of Autonomy’, Science and Ethics (New York: Rodopi Press, 1981). Dworkin, Gerald. The Theory and Practice of Autonomy (New York: Cambridge University Press, 1988). Dworkin, Ronald. Taking Rights Seriously (London: Duckworth, 1978). Elster, Jon. Ulysses and the Sirens: Studies in Rationality and Irrationality (Cambridge: Cambridge University Press, 1979). Elster, Jon. The Cement of Society: A Study of Social Order. Studies in Rationality and Social Change (Cambridge: Cambridge University Press, 1989). Elster, Jon. Nuts and Bolts for the Social Sciences (Cambridge: Cambridge University Press, 1989). Feser, Edward.‘There Is No Such Thing As An Unjust Initial Acquisition’, Social Philosophy and Policy, 22 (2005): 56–80. Fisher, John Martin. ‘Responsiveness and Moral Responsibility’, in F. Schoeman (ed.), Responsibility, Character, and the Emotions (New York: Cambridge University Press, 1987). Frank, Robert. Passions Within Reason: The Strategic Role of Emotions (New York: W. W. Norton & Co., 1988). Frankfurt, Harry. ‘Freedom of the Will and the Concept of a Person’, Journal of Philosophy, 68 (1971): 5–20. Frankfurt, Harry. ‘Identification and Externality’, in A. Rorty (ed.), The Identities of Persons (Berkeley: University of California Press, 1977). Frankfurt, Harry. ‘Identification and Wholeheartedness’, in F. Schoeman (ed.), Responsibility, Character, and the Emotions (New York: Cambridge University Press, 1987). Frankfurt, Harry. The Importance of What We Care About (New York: Cambridge University Press, 1988). Fried, Barbara. ‘Left-Libertarianism: A Review Essay’, Philosophy and Public Affairs, 32 (2004): 66–92. Fried, Barbara. ‘Left-Libertarianism, Once More: A Rejoinder to Vallentyne, Steiner, and Otsuka’, Philosophy and Public Affairs, 33 (2005): 216–22. Friedman, Marilyn. ‘Autonomy and the Split-Level Self’, Southern Journal of Philosophy, XXIV/1 (1986): 19–35. Gauthier, David. ‘Deterrence, Maximization, and Rationality’, Ethics, 94 (1984): 474–95. Also published in D. MacLean (ed.), The Security Gamble: Deterrence Dilemmas in the Nuclear Age (Totowa, NJ: Rowman & Allanheld, 1984), 101–22. Gauthier, David. Morals By Agreement (Oxford/New York: Oxford University Press, 1986). Gauthier, David. Moral Dealing: Contract, Ethics, and Reason (Ithaca, NY: Cornell University Press, 1990). Gauthier, David. ‘Why Contractarianism?’, in Peter Vallentyne (ed.), Contractarianism and Rational Choice (Cambridge: Cambridge University Press, 1991), 25. Gauthier, David. ‘Intention and Deliberation’, in Peter Danielson (ed.), Modeling Rationality, Morality, and Evolution, Vancouver Studies in Cognitive Science, vol. 7 (New York: Oxford University Press, 1998), 41–54.

244

Liberty, Games and Contracts

Gauthier, David and Robert Sugden (eds). Rationality, Justice, and the Social Contract: Themes from Morals by Agreement (Ann Arbor: University of Michigan Press, 1993). Gibbard, Allan. Wise Choices, Apt Feelings: A Theory of Normative Judgment (Cambridge, MA: Harvard University Press, 1990). Gigerenzer, Gerd and R. Selten (eds). Bounded Rationality: The Adaptive Toolbox (Cambridge, MA: MIT Press, 2002). Griffin, David. Well Being: Its Meaning, Measurement and Moral Importance (Oxford: Oxford University Press, 1986). Groarke, Leo. Greek Scepticism: Anti-Realist Trends in Ancient Times (Montreal: McGillQueen’s University Press, 1985). Groarke, Leo. ‘Can Capitalism Save Itself? Some Ruminations on the End of Capitalism’, in John Bishop (ed.), Ethics and Capitalism (Toronto: University of Toronto Press, 2000). Hare, R. M. Moral Thinking (Oxford: Oxford University Press, 1981). Haworth, Lawrence. Autonomy (New Haven, CT: Yale University Press, 1986). Hayek, Friedrich. The Road to Serfdom (Chicago: University of Chicago Press, 1944). Heath, Joseph. ‘A Multi-Stage Game Model of Morals by Agreement’, Dialogue, XXV (1996): 529–52. Hegel, Georg Wilhelm Friedrich. Die Vernunft in der Geschichte, J. Hoffmeister (ed.) (Hamburg: Felix Meiner Verlag, 1955). Heinrich, Joe, Wulf Albers, Robert Boyd, Gerd Gigerenzer, Kevin McCabe, Axel Ockenfels and Peyton Young. ‘What is the Role of Culture in Bounded Rationality?’, in Gerd Gigerenzer and R. Selten (eds), Bounded Rationality: The Adaptive Toolbox (Cambridge, MA: MIT Press, 2002). Held, Virginia. ‘Feminism and Moral Theory’, in Eva Feder and Diana Meyers (eds), Women and Moral Theory (Savage, MD: Rowman & Littlefield, 1987). Hill Jr, Thomas. ‘Servility and Self-Respect’, The Monist, 57 (1973). Hill Jr, Thomas. ‘Weakness of Will and Character’, Philosophical Topics, XIV/2 (Fall 1986). Hill Jr, Thomas. ‘The Kantian Conception of Autonomy’, in John Christman (ed.), The Inner Citadel: Essays on Individual Autonomy (Oxford: Oxford University Press, 1989). Hill Jr, Thomas. Autonomy and Self Respect (New York: Cambridge University Press, 1991). Hobbes, Thomas. Leviathan (Buffalo, NY: Prometheus Books, [1651] 1986). Hospers, John. Libertarianism (Los Angeles: Nash, 1971). Hume, David. Inquiry Concerning the Principles of Morals (Indianapolis: Bobbs-Merrill, 1957). Hume, David. Treatise of Human Nature, L. A. Selby-Bigge (ed.) (Oxford: Oxford University Press, 1978). Kant, Immanuel. Critique of Practical Reason (Indianapolis: Bobbs-Merrill, [1788] 1956); Lewis White Beck (trans. and ed.) (Chicago, 1949). Kant, Immanuel. Metaphysical Elements of Justice, John Ladd (trans.) (Indianapolis: BobbsMerrill, 1965). Kant, Immanuel. The Philosophy of Law, W. Hastic (trans.) [Edinburgh, 1887]; reprinted in Great Books of the Western World (Chicago: Encyclopaedia Britannica, 1952). Kant, Immanuel. Werke (Berlin: Bruno Cassier, 1914). Kirzner, Israel. ‘The Nature of Profits: Some Economic Insights and Their Ethical Implications’, in Robin Cowan and Mario Rizzo (eds), Profits and Morality (Chicago: Chicago University Press, 1995), 22–47.

Bibliography

245

Kraus, Jody. The Limits of Hobbesian Contractarianism (Cambridge: Cambridge University Press, 1993). Kymlicka, Will. Contemporary Political Philosophy (Oxford: Clarendon Press, 1990). Lee, Steven P. ‘Weapons of Mass Destruction and the Limits of Moral Understanding’, in Sohail H. Hashmi and Steven P. Lee (eds), Ethics and Weapons of Mass Destruction: Religious and Secular Perspectives (Cambridge: Cambridge University Press, 2004), 482–509. Letwin, Shirley Robin. ‘Modern Philosophies of Law’, The Great Ideas Today, 1972 (Chicago: Encyclopaedia Britannica, 1972), 105–53. Levey, Ann. ‘Liberty and Property’, in Samantha Brennan, Tracy Isaacs and Michael Milde (eds), A Question of Values: New Canadian Perspectives in Ethics and Political Philosophy (Amsterdam: Rodopi Press, 1997), 87–105. Lomasky, Loren. Persons, Rights and the Moral Community (Oxford: Oxford University Press, 1987). Lorberbaum, J. ‘No Strategy is Evolutionarily Stable in the Repeated Prisoner’s Dilemma’, Journal of Theoretical Biology, 168 (1994): 117–30. Machan, T. R. ‘A Note of Independence’, Philosophical Studies, 30 (1976): 419–22. Machan, T. R. ‘Should Business be Regulated?’, in Tom Regan (ed.), Just Business (New York: Random House, 1984). Machan, T. R. Individuals and Their Rights (LaSalle, IL: Open Court Publishing Company, Inc., 1989). Machan, T. R. ‘Why Agreement Is Not Enough’, Philosophia, 28/1–2 (2000). Machan, T. R. ‘The Case for Libertarianism’, in Craig Duncan and Tibor R. Machan (eds), Libertarianism: For and Against (Toronto: Rowman & Littlefield, 2005). MacIntosh, Duncan. ‘Preferences Progress: Rational Self-Alteration and the Rationality of Morality’, Dialogue, XXX/1–2 (1991): 3–32. MacIntosh, Duncan. ‘Preference-Revision and the Paradoxes of Instrumental Rationality’, Canadian Journal of Philosophy, 22/4 (1992): 503–30. MacIntosh, Duncan. ‘Categorically Rational Preferences and the Structure of Morality’, in Peter Danielson (ed.), Modeling Rationality, Morality and Evolution, Vancouver Studies in Cognitive Science, vol. 7 (New York/Oxford: Oxford University Press, 1998), 282–301. MacIntosh, Duncan. ‘The Mutual Limitation of Needs as Bases of Moral Entitlements: A Solution to Braybrooke’s Problem’, in Sue Sherwin and Peter Schotch (eds), Engaged Philosophy: Essays in Honour of David Braybrooke (Toronto: University of Toronto Press, 2007). MacIntyre, Alasdair. After Virtue: A Study in Moral Theory, 2nd edn (Notre Dame, IN: University of Notre Dame Press, 1985). Mackie, J. L. Ethics: Inventing Right and Wrong (Toronto: Penguin, 1977). Mackie, J. L. ‘The Three Stages of Universalization’, in Joan and Penelope Mackie (eds), Persons and Values: Selected Papers, Volume II (Oxford: Clarendon Press, 1985), 170–83. Mackie, J. L. ‘Sidgwick’s Pessimism’, in Joan and Penelope Mackie (eds), Persons and Values: Selected Papers, Volume II (Oxford: Clarendon Press, 1985), 77–90. Marinoff, Louis. ‘The Inapplicability of Evolutionarily Stable Strategy to the Prisoner’s Dilemma’, The British Journal of Philosophy of Science, 41 (1990): 461–72. Marinoff, Louis. ‘Maximizing Expected Utilities in the Prisoner’s Dilemma’, Journal of Conflict Resolution, 36 (1992): 183–216. Marinoff, Louis.‘The Failure of Success: Intrafamilial Exploitation in the Prisoner’s Dilemma’, in Peter Danielson (ed.), Modeling Rationality, Morality, and Evolution,

246

Liberty, Games and Contracts

Vancouver Studies in Cognitive Science, vol. 7 (Oxford: Oxford University Press, 1998), 161–184. Mates, Benson. The Skeptic Way: Sextus Empiricus’s ‘Outlines of Pyrrhonism’ (New York: Oxford University Press, 1996). Meyers, Diana T. ‘Personal Autonomy and the Paradox of Feminine Socialization’, Journal of Philosophy, LXXXIV/11 (1987): 619–28. Meyers, Diana T. ‘The Socialized Individual and Individual Autonomy: An Intersection between Philosophy and Psychology’, in E. Feder Kittay and D. T. Meyers (eds), Women and Moral Theory (Lanham, MD: Rowman & Littlefield, 1987). Meyers, Diana T. Self, Society and Personal Choice (New York: Columbia University Press, 1989). Mill, J. S. On Liberty, (Indianapolis: Hackett Publishing Co., 1978). Mintoff, Joseph. ‘Is Rational and Voluntary Constraint Possible’, Dialogue: Canadian Philosophical Review, XXXIX/2 (2000): 339–64. Moore, G. E. Principia Ethica (Cambridge: Cambridge University Press, [1903] 1983), 15 (ch. 1, sect. 13.1). Murray, Charles. Losing Ground: American Social Policy 1950–1980 (New York: Basic Books, 1984). Murray, Malcolm. The Moral Wager: Evolution and Contract (Dordrecht: Springer, 2007). Nagel, Thomas. ‘Libertarianism without Foundations’, in Jeffrey Paul (ed.), Reading Nozick: Essays on Anarchy, State, Utopia (Totowa, NJ: Rowman & Littlefield: 1981). Nagel, Thomas. The View from Nowhere (Oxford: Oxford University Press, 1986). Narveson, Jan. Morality and Utility (Baltimore: Johns Hopkins University Press, 1967). Narveson, Jan. ‘A Puzzle about Economic Justice in Rawls’ Theory’, Social Theory and Practice, 4/1 (1976): 1–27. Reprinted in Jan Narveson, Respecting Persons in Theory and Practice (Lanham: Rowman & Littlefield, 2002), ch. 2, 13–34. Narveson, Jan. ‘Rawls on Equal Distribution of Wealth’, Philosophia, 7/2 (June l978): 281–92. Narveson, Jan. ‘Rawls and Utilitarianism’, in H. Miller and W. Williams (eds), The Limits of Utilitarianism (Minneapolis: University of Minnesota Press, l982), 128–143. Narveson, Jan. ‘Animal Rights Revisited’, in H. Miller and W. Miller (eds), Ethics and Animals (Clifton, NJ: Humana Press, 1983). Narveson, Jan. ‘The How and Why of Universalizability’, in Nelson Potter and Mark Timmons (eds), Morality and Universality (Dordrecht: Reidel, 1985), 3–44. Narveson, Jan. ‘A Case Against Animal Rights’, in Advances in Animal Welfare Science (Washington, DC: The Humane Society of the United States, 1986). Narveson, Jan. ‘The Agreement to Keep Our Agreements: Hume, Prichard, and Searle’, Philosophical Papers, XXIII/2 (1994). Narveson, Jan. ‘Contracting for Liberty’, in T. R. Machan and D.B. Rasmussen (eds), Libertarianism for the 21st Century, (Lanham, MD: Rowman & Littlefield, 1995). Narveson, Jan. ‘Toward a Liberal Theory of Ideology: A Quasi-Marxian Exploration’, Reason Papers, 20 (Fall 1995). Narveson, Jan. ‘The Anarchist’s Case’, in J. T. Sanders and J. Narveson (eds), For and Against the State (Lanham, MD: Rowman & Littlefield, 1996), 195–216. Narveson, Jan. The Libertarian Idea (Philadelphia: Temple University Press, 1988; republished Peterborough, Ontario: Broadview Press, 2001). Narveson, Jan. Moral Matters (2nd edn) (Peterborough, Ontario: Broadview Press, 1999). Narveson, Jan. ‘Property Rights: Original Acquisition and Lockean Provisos’, Public Affairs

Bibliography

247

Quarterly, 13/3 (1999): 205–27. Reprinted in Jan Narveson, Respecting Persons in Theory and Practice (Lanham, MD: Rowman & Littlefield, 2002), 111–30. Narveson, Jan. ‘Liberal-Conservative: The Real Controversy’, The Journal of Value Inquiry Special Issue on Liberalism, J. Narveson and S. Dimock (eds), 34 (September 2000): 2–3. Narveson, Jan. Respecting Persons in Theory and Practice: Essays on Moral and Political Philosophy (Lanham, MD: Rowman & Littlefield, 2002). Narveson, Jan. ‘The Drug Laws: More Nails in the Coffin of American Liberalism’, in Jan Narveson, Respecting Persons in Theory and Practice: Essays on Moral and Political Philosophy (Lanham, MD: Rowman & Littlefield, 2002), 243–64. Narveson, Jan. ‘Moral Realism, Emotivism, and Natural Law’, in Jan Narveson, Respecting Persons in Theory and Practice: Essays on Moral and Political Philosophy (Lanham, MD: Rowman & Littlefield, 2002), 63–77. Narveson, Jan. ‘The Invisible Hand’, Journal of Business Ethics, 46/3 (September 2003): 201–12. Nielsen, Kai. Equality and Liberty (Lanham, MD: Rowman & Littlefield, 1985). Nozick, Robert. Anarchy, State, and Utopia (New York: Basic Books, 1974). Nozick, Robert. The Nature of Rationality (Princeton: Princeton University Press, 1993). Nussbaum, Martha. Cultivating Humanity: A Classical Defense of Reform in Liberal Education (Cambridge, MA: Harvard University Press, 1997). Okin, Susan Moller. Justice, Gender, and the Family (New York: Basic Books, 1989). Otsuka, Michael. Libertarianism without Inequality (Oxford: Clarendon Press, 2003). Parfit, Derek. Reasons and Persons (Oxford: Oxford University Press, 1984). Patten, Allen. ‘Social Contract Theory and the Politics of Recognition in Hegel’s Political Philosophy’, in Robert R. Williams (ed.), Beyond Liberalism and Communitarianism: Studies in Hegel’s Philosophy of Right (Albany: State University of New York Press, 2001), 167–84. Paul, Ellen Frankel, Fred D. Miller Jr and Jeffrey Paul (eds), Natural Rights Liberalism from Locke to Nozick (New York: Cambridge University Press, 2005). Perkins, H. and A. Berkowitz, ‘Perceiving the Community Norms of Alcohol Use among Students: Some Research Implications for Campus Alcohol Education Programming’, International Journal of Addictions, 21/9–10 (1986): 961–76. Pols, Edward. Mind Regained (Ithaca, NY: Cornell University Press, 1998). Railton, Peter. ‘How Thinking about Character and Utilitarianism Might Lead to Rethinking the Character of Utilitarianism’, Midwest Studies in Philosophy, 13 (1988): 398–416. Rasmussen, Douglas B. and Douglas J. Den Uyl. Liberty and Nature, An Aristotelian Defense of Liberal Order (LaSalle, IL: Open Court Publishing Company, Inc., 1991). Rawls, John. A Theory of Justice (Cambridge, MA: Harvard University Press, 1971). Rawls, John. ‘The Independence of Moral Theory,’ Proceedings and Addresses of the American Philosophical Association (Newark: University of Delaware Press, 1975). Rawls, John. Political Liberalism (New York: Columbia University Press, 1993). Rawls, John. Collected Papers (Cambridge, MA: Harvard University Press, 1999). Ridley, Matt. The Origins of Virtue (New York: Viking, 1996). Ripstein, Arthur. ‘Foundationalism in Political Theory’, Philosophy and Public Affairs, 16/2 (Spring 1987): 115–37. Rothbard, Murray. For a New Liberty: The Libertarian Manifesto, revised edition (New York: Libertarian Review Foundation, 1978). Ryan, Alan. Property (London: Open University Press, 1987).

248

Liberty, Games and Contracts

Sandel, Michael J. ‘Justice and the Good’, in Michael J.Sandel (ed.), Liberalism and Its Critics (New York: New York University Press, 1984). Sanders, J. T and J. Narveson (eds), For and Against the State: New Philosophical Readings (Lanham, MD: Rowman & Littlefield, 1996). Schmidtz, David. The Limits of Government: An Essay on the Public Goods Argument (Boulder, CO: Westview Press, 1991). Sidgwick, Henry. Methods of Ethics (7th edn) (Indianapolis: Hackett, 1981). Singer, Peter. ‘Altruism and Commerce: A Reply to Arrow’, Philosophy and Public Affairs, 2/1 (1973): 312–20. Singer, Peter. The Expanding Circle: Ethics and Sociobiology (Oxford: Oxford University Press, 1983). Skyrms, Brian. Evolution of the Social Contract (Cambridge: Cambridge University Press, 1996). Skyrms, Brian. ‘Game Theory, Rationality and Evolution of the Social Contract’, in Leonard D. Katz (ed.), Evolutionary Origins of Morality: Cross-Disciplinary Perspectives (Thorverton/Bowling Green, OH: Imprint Academic, 2000), 269–84. Skyrms, Brian. ‘The Stag Hunt’, Proceedings and Addresses of the American Philosophical Association, 75/2 (2001): 31–41. Skyrms, Brian and Jason Alexander. ‘Bargaining with Neighbors: Is Justice Contagious?’, Journal of Philosophy, 96/11 (1999): 588–98. Smith, Adam. The Wealth of Nations (New York: Everyman’s Library, [1776] 1910). Smith, John Maynard. Evolution and the Theory of Games (Cambridge: Cambridge University Press, 1982). Sober, Elliot, and David Sloan Wilson. Unto Others: The Evolution and Psychology of Unselfish Behavior (Cambridge, MA: Harvard University Press, 1999). Sowden, Lanning and Sheldon Wein. ‘Justice and Rationality: Doubts about the Contractarian and Utilitarian Approaches’, Philosophia, 17/2 (1987): 127–39. Sperba, Dan. Explaining Culture: A Naturalistic Approach (Oxford: Oxford University Press, 1996). Steiner, Hillel. ‘How Free: Computing Personal Property’, in A. Phillips Griffiths (ed.), Of Liberty (Cambridge: Cambridge University Press, 1983). Steiner, Hillel. An Essay on Rights (Cambridge, MA: Blackwell Publishing, 1994). Stemmer, Peter. ‘Moralischer Kontraktualismus’, Zeitschrift für philosophische Forschung, 56/1 (2002): 1–21. Titmuss, Richard. The Gift Relationship: From Human Blood to Social Policy (London: Allen & Unwin, 1971). Tucker, Chris. ‘Contractarianism, Justification and Relativity’, Dialogue: Canadian Philosophical Review, XLII/3 (Summer 2003): 559–71. Vallentyne, Peter (ed.). Contractarianism and Rational Choice: Essays on David Gauthier’s Morals by Agreement (New York: Cambridge University Press, 1991). Vallentyne, Peter and Hillel Steiner (eds). The Origins of Left Libertarianism: An Anthology of Historical Writings (New York: Palgrave Publishers Ltd, 2000). Vallentyne, Peter and Hillel Steiner (eds). Left Libertarianism and Its Critics: The Contemporary Debate (New York: Palgrave Publishers Ltd, 2000). Vallentyne, Peter, Hillel Steiner and Michael Otsuka. ‘Why Left-Libertarianism Isn’t Incoherent, Indeterminate, or Irrelevant: A Reply to Fried’, Philosophy and Public Affairs, 33 (2005): 201–15. Viminitz, Paul. ‘A Defence of Terrorism’, in Frederick R. Adams (ed.), Ethical Issues for the Twenty-First Century (Charlottesville, VA: Philosophy Documentation Center Press, 2005).

Bibliography

249

de Waal, Frans. Good Natured: The Origins of Right and Wrong in Humans and Other Animals (Cambridge, MA: Harvard University Press, 1996). Watson, Gary. ‘Free Agency’, Journal of Philosophy, LXXII/ 8 (April 1975): 205–20. Wein, Sheldon. ‘Prisoners’ Dilemmas, Tuism, and Rationality’, Simulation and Games, 16/1 (1985): 23–31. Wein, Sheldon. ‘Problems with Contractarianism’, Journal of Social Philosophy, XVI/3 (1986): 48–59. Wein, Sheldon. ‘Plato and the Social Contract’, Philosophy Research Archives, XII (1986–87): 66–77. Wein, Sheldon. ‘American and Canadian Justice’, in Terry Goldie, Carmen Lambert and Rowland Lorimer (eds), Canada: Theoretical Discourse/Discours Theoriques (Montreal: Association for Canadian Studies, 1994), 373–91. Wein, Sheldon. ‘A Hobbesian Foundation for Welfare Rights’, in C. F. Delaney (ed.), The Liberalism–Communitarianism Debate (Lanham, MD: Rowman & Littlefield, 1994), 213–26. Wein, Sheldon. ‘Feminist Consciousness and Community Development’, The International Journal of Social Economics, 24/12 (1997): 1376–87. Williams, George. Adaptation and Natural Selection: A Critique of Some Current Evolutionary Thought (Princeton, NJ: Princeton University Press, 1966). Wilson, Edward O. On Human Nature (Cambridge, MA: Harvard University Press, 1978). Wolf, Susan. Freedom within Reason (New York: Oxford University Press, 1990). Wright, Robert. The Moral Animal (New York: Vintage Books, 1994).

This page intentionally left blank

Index abortion 49 acquisition 154, 163, 165–6, 174 defined 235 initial 79, 174, 176n5 just 168 by manufacture 163 principle of 157, 159 by trade 163 adaptations 188 addicts 86–8, 92, 224–5 Adeimantus 2 advantage of agreements 5, 12, 14, 181, 189 of contractarianism 13 of convention 189 delayed 181 of groups 183, 189, 237 of property 237 rational 67 reproductive 122 of the strong 190 temporary 26 threat 141 universal 51 of war 126n4, 223 without restraint 22 advantaged 190–91 adversary 14 aesthetic intuitions 55, 222 preferences 56, 222 aggression 104, 192, 228 agreement abiding by 4–5, 31n23, 33–4, 36, 44, 58, 69–70, 116–17, 204, 209, 220 actual (real) 32n24, 33, 75, 220 ex ante 5, 115–17, 124 ex post 115–17 explicit 22 hypothetical 21 implicit (tacit) 72

rational 4, 8n3, 67–8 voluntary 134 Ahmad, R. 114n26 aims 2, 53, 83, 164 common 113 occurrent 115, 117–18, 131 see also ends; desires Al Qaeda 34 Albers, W. 127n22 alcoholic 90 Alexander, J. 127n16 Alexander, R. 183–4, 186, 194n26, n27, 195n31, n32 aliens 192 allegiances 53 allele 181, 193n5 altruism 113, 125, 184–7 kin 192 nepotistic 194n30 non-strategic 186–7, 237 overextended 125 reciprocal 180, 183, 189, 193n5 semi- 237 always defect (AD) 180–81 ambivalence 89–91, 93, 100n24 amoral agents 103 interest 117, 120 reason 202 amorality 118, 228 anarchy 12, 81, 105, 169n5, n6, 213n36, 226–7 animal nature 55 animals 25–6, 28, 30, 31n23, 165, 185, 194n21, 219 treatment of 28, 191–2, 219 Annas, J. 62n19 anomie 89 ants 3–4, 25, 27, 31n22, 218–19 Apple Macintosh 1, 149, 232 apples 175, 236 251

252

Liberty, Games and Contracts

appropriation 151, 153–4, 163–5, 169n3, 173–8, 233, 235 just 165 original (initial) 151, 173 private 237 approximate mechanisms 185 see also heuristics arbiter 48, 53, 57 Aristotle 42–3, 75, 230 Aristotelean 42–3, 213n33 arms race 183 Arrow, K. 114n22 art 50, 56 artificial 52–3, 57–8 assembly of citizens 41 assimilation 123 assistance 133–4, 151 Aumann, R. 121, 127n15 authentic desires 85–6, 89, 91, 99n15, 224 motives 86 reasons 86, 90, 93 self 85, 87–8, 93, 224 values 96 authenticity 85, 89, 92–4, 224–5 authoritarianism 12, 36, 104 authority 16, 43, 60, 70, 86, 96 over children 234 coercive 14, 16 of the state 150 ultimate 96 autonomous acts 87, 89, 90, 92–3 agents 84–5, 92–6 behaviour 85 choice 94, 98–9 desires 5, 81, 85, 92–6, 99n15 non- 85, 91–2, 96–8 values 84, 96, 98 autonomy 3, 16, 85–94, 186, 201, 224–5 conception of 84–5 necessary conditions of 93 objective theory of 95 sufficient conditions of 94 value of 94–5 Axelrod, R. 31n16, 114n13, 181–2, 193n5, n6 backwards induction 106

bargaining baseline for 137, 140, 229 conditions of 135, 141 exclusion from 69 games 229 norms 135, 141 outcome of, 59, 122 over cooperative dividends 120, 135, 164 position 5, 59 power 189–90 process 136 rational 67, 75, 219 from state of nature 116 Barnes, J. 62n19 baseline 118, 133–6, 157, 160–61 for bargaining 137, 140, 229 initial 5, 118 libertarian 228 no-gun 231 non-harm 5 original 136 pre-appropriation 163, 165 pre-property 152 of the prisoner’s dilemma 106 of the ultimatum game 120 battle of the sexes 227 beaver 185 Becker, L. 151 bees 25 behavioural genetics 218 beneficence 184–5, 187, 237 benefits of cooperation 12, 68, 74, 184, 203 economic 183 foregoing 190 free-rider 117, 221 group 183, 185, 194n21 indirect 184–5 of liberty 148 marginal 192 mutual 3, 68, 119, 238 perceived 184 potential 68 purported 71 realizable 67 technological 196n49 teleological 120 unilateral 119 benevolence 12

Index Benson, P. 62n19, 94 Bentham, J. 199, 205, 210n5 Berkowitz, A. 114n24 bested in combat 134 bettering 134–6, 231 between-group rivalries 183 bicycle 149 Bigelow, R. 194n18 binding agreements 4, 33 appeal 3 constraints 34 morally 34–5, 37 norms 33 of promises 39 social compact 34 Binmore, K. 8n3, 77n10, 113n7, 228 blame 35, 81 blood donation 5, 103, 112, 184–5, 226 blood supply 112, 226–7 Bond, E. J. 3–5, 17n15, n16, 217–18 Bornik, Z. 114n26 bounded rationality 131 bourgeois morality 229 Boyd, R. 124, 127n20, n22, 193n6 Brandt, R. 126n3, 199, 225 Braybrooke, D. 170n48, 212n25 Brown, G. vii, 7, 195n39, 217, 237 Buddhism 142n7 bundle of rights 148 C student 122 Callimachus 60 Campbell, R. 100n24, n25, 169n1 Canadian Blood Services 114n23 Canadian health system 141n2, 227 cancer 133–4, 231 capacity for autonomy 85, 92–3, 96, 98–9 to better or worsen 136–7, 140, 231 to choose 45n8 for cooperation 194n18 for liberty 147, 157, 159, 164 for maximization 187 military 137, 231 for morality 188–9, 237 natural selection for 188–9 productive 136–7, 160, 231 psychological 92, 209

253

to reason 4, 7, 27 for reflection 86, 92–4, 98 reproductive 158, 160, 162 for trustworthiness 182 capital punishment 50 capitalism 163 cardinal comparisons 24, 226 caring for others 3, 94, 127n17, 142n14, 185, 201, 204, 206–7, 234, 239 Caroll, L. 240n2 castes 180, 189, 191 catapult fodder 143n20 categorical imperative 16n6, 19–21, 123–4, 199, 201 Cavalli-Sforza, L. L. 127n19 Cave, E. 206, 212n33 Centola, D. 114n24 centralized (re)distribution 116–17, 126, 227 Chamberlain, Wilt 160 character 15, 76, 100n29, 170n16, 198, 206 charity 115, 122–3, 127n17, 226 involuntary 122–3, 226 minimal 118 cheap talk 112 cheating 27, 182, 184, 187, 221, 239 chicken (the animal) 218 chicken game 5, 103–12, 113n4, 225–7 iterated 225 children 6, 68, 74, 181, 185 education of 81, 98, 194n30, 208 ownership of 159–61, 165, 234 rights of 157, 170n48, 171n52, 192 welfare of 192 choice situation 2, 72, 200–204 under uncertainty 106 choice theory 67, 72, 203 circularity 4, 21, 25, 164, 236 see also question begging circumstances of justice 69–71, 98 civil society 117, 133–4, 139–40, 202–4, 212n33 claim rights 7, 147–8, 175–6 co-bargainers 116 co-tenable rights 164 codes 28, 30n11, 34 coercion 100n19, 104–5, 115, 122, 135, 143n20, 190, 226, 238

254

Liberty, Games and Contracts

Cohen, G. A. 8n4 Cohen, J. 169n15 coherence intrapersonal 89, 93 model 87 Coleman, J. 211n16 collective consent 174 interests 206 stability 191 commensurability 28, 239 commitment 5, 52–3, 107, 119, 221 to convention 54 to ethics 113, 227 to freedom 58 prior 4 to reason 63n27 second-order 87, 89, 91–2 common good 3, 15–16, 24, 96, 218–19 language 98, 217n29 sense 35, 38–9, 41 use 152, 173 communication 22, 225 communitarianism 49, 52 community 28, 43, 45n26, 53, 130, 164, 184–5 concern 194n21, 196n50 life 33, 42, 45n19, 220 compactarianism 35–6, 41, 43, 219 see also contractarianism compacts 4, 33–4, 39–41, 45n3 see also contracts competition 13–14, 23, 27, 60–61, 82, 90–91, 119, 121–2, 129, 165, 188–9 unequal 113n4 competitive value 173 compliance 67, 81, 111, 223 costs of 73 efficient 4, 65, 74, 76 full 4, 65, 72–4 problem 4, 65–6, 68–76, 223 rational 68, 71–2 compromise 56, 59, 188, 222 conception of the individual 53 conceptions of the good 54, 83, 162, 164, 166 conditional cooperation 73, 107, 111, 114n16, 116, 125, 180–81, 206, 227–8

cooperator (CC) 73 108–10, 114n16, 116, 125, 127n25, 181 strategies 110–11 conditioning moral 186 psychological 187 conflict 48, 56, 88–90 of desires 87, 89, 100n2 interpersonal 87 unresolved 89–91 of volitions 91 conflicted selves 89, 91 consensus 35, 49, 51, 197 overlapping 201 consent 13–14, 32n28, 34, 39, 43–4, 45n26, 116–18, 163–8, 174, 235–7 consequentialism 11–12, 62n30, 199 conservativism 49, 84 consistency 34, 37, 89, 185 self- 34, 220 constrained maximization 4, 67–8, 106–7, 204, 212n30, 233 constrained maximizers 51, 73, 107, 127n21, 204 constraint 3, 5, 7–8, 11–12, 14–15, 22, 26, 29–30, 34, 51, 58–9, 65, 67, 70–71, 78n19, 81, 84, 115–16, 118, 159, 161, 167, 169n5, 203, 211n24, 239 mutual 25, 71, 164 side 161–2, 165–8 constructivism 70, 210n8 contract theory 19, 27, 29, 32n28 see also compact contractarianism 2–5, 7, 13–15, 23–30, 43–52, 54, 56–9, 65–6, 68–9, 75–6, 81–2, 99, 116, 119–21, 126, 131, 164, 190–92, 197, 200–210, 218, 221–3, 226, 238–9 full-bodied 54–9 market 7, 197, 201–10, 211n16, 212n33, 213n37, 239 mutual advantage 189–93, 195n44 orthodox 75 social 35–7, 197, 200–206, 210n9 see also compactarianism contractualism 210n8 contribution 118, 136, 185, 219 convention 31, 39, 54–5, 62n24, 124, 189–90

Index post- 72 cooperate (C) (the strategy) 4–5, 12–14, 26, 53, 56, 68, 73, 76, 114n16, 124, 136, 141, 181–3, 203–4, 206, 212n30, n31, n33, 217, 225, 231, 239 cooperation 3–5, 13–14, 22, 26–7, 65, 68, 73–4, 76, 77n10, 95, 103–5, 107, 111, 114n16, 116, 125, 127n21, 136, 141, 143n20, 180–83, 189–90, 193n8, 194n18, 203, 206, 212n33, 227–9, 231, 233, 237, 239 fair 76 mutually advantageous 189, 237 possibility of 95 social 203 unilateral 4, 203 universal 203 cooperative agents 73, 82 arrangements 26 attitude 239 behaviour 68, 125 conditionally 116 dividend 129, 13–16, 142n10 option 193n5, 223 outcome 206, 212n30, 238 preference 76 reputation 175 strategies 116, 227 surplus 71, 228 venture 22, 74, 116, 119, 136, 141, 184, 203–4, 231 coordination failure 112 problems 227 coordinative function 66 correlated encounters 121–2 cost of coercion 190, 238 of compliance 73 of detection 73 of enforcement 73, 192 of farewelling 131 of non-cooperation 111, 182, 203 of positive rights 95 of production 60 cost-benefit calculation 238 counterfactuals 136, 192 counter-productivity 195n40, 231

255

coveting 209 cows 218–19 craned car problem 133 critical frequency 179–80 crop failure 134–5 cultural evolution 123, 186–7 culture 62n23, 76, 98, 123, 134, 186, 192 problem of 194n19 customs 54–6, 62n24 Danielson, P. 5, 7, 72, 104–5, 127n25, 170n48, 194n14, 206, 225–7, 239 Darwall, S. 99n13 Dawkins, R. 182, 186, 193n2, n4, n5 de Waal, F. 127n26, 185–6, 194n21, n23, n28, 195n34 debts 115–16 deception 57, 192 decision theory 207 defeasibility condition 97 defect (D) (the strategy) 68, 72–3, 75, 116, 119, 193n5, n8, 194n10, 204, 212n30 defection 65, 68, 72, 76, 77n10, 125, 182 domination of 65, 77n10, 204 mutual 118 unilateral 4, 68, 203, 212n30 universal 203 utility of 67 democratic 28, 46n21 deontic morality 12 deontological 205–6, 212n29 descriptivity 26, 129 desert 136, 158 desires authentic 85–6, 89, 92–3, 99n14 bi-level account of 85–7, 91–2, 99n15, n19 effective 86, 92 endorsement of 85 first-order 86–93 higher-order 86, 99n15, 224 inauthentic 224 inconsistent 88–90, 224 lower-order 89, 224 second-order 86, 89, 91–2 see also interests; preferences desire-fulfilment 164 determinism 36

256

Liberty, Games and Contracts

deus ex machina 220 deviations 184, 188 diet 92, 228 difference principle 59–61, 143n22, 201, 211n12 diffidence 129 digger wasps 179, 189 dignity 201, 203, 210n10 Dimock, S. 5, 77n1, 100n19, 211n22, 224–5 disabled 161 dispositions to act 81 appropriate 139–40 changing 68 conditional (sensitive) 4, 31n22, 116 to cooperate (comply) 5, 12–13, 67 inborn 34, 41 inculcated 139 moral 8n3, 130–31, 142n5, 218 natural 30n11 psychological 41, 69 rational 67, 73 stable 212n30 toggling 140 disregard 154, 233 dissonance 88 distribution centralized 116–17, 126, 227 via contribution 136 efficient 61 fair 125 of goods 116, 120, 143n22, 239 of liberty 232 patterns of 152 of peace 231 of populations 109, 121 reasonable 231 of resources 164, 170n48 of restrictions 153, 155 social 239 of surplus 71, 120, 125, 127n12 of wealth, 60 distributive justice 49, 58, 60 disutility 71–2, 235, 238 diversification 55, 58, 183 dogs 111, 165 dominance 188–9 dominant strategy 22, 106, 204

Dominion 137–8 dove 108 Dowlatabadi, H. 114n26 drift 122 drives biological 33, 37 from fear 39 innate 4, 33–4, 37, 44n8 to mimic 123, 125 natural 34, 36, 45n22 psychological 33, 37, 41 for self-preservation 33, 35 drive-by shootings 230 drug addicts 86, 88, 92, 225 legislation 81, 96, 224 takers 97, 225 drugs 86–8, 96–7, 225 due process 44 Duncan, C. 62n30 duties 6, 13, 123, 153 to animals 219 categorical 124 to children 157 cosmos of 189 enforceable 148, 157 moral 3, 157 natural 53 negative 123 not to interfere 133, 147, 155 objective 220 to oneself 171n54 positive 95, 133, 157 to provide 155 to refrain 153–4 reasonable 67 relation to rights 133 sacred 221 Dworkin, G. 85, 99n14, 100n31 Dworkin, R. 201 dysfunctional 187–8, 191, 237 ecological equilibrium 108 model 109, 111 niche 193 economic benefits 183 contractarianism 71, 211n16

Index explanations 112 inequalities 59 opportunity 185 rationality 51, 53–4, 222 reality 38 resource 174 stimulation, 76 system 201 economies of scale 183 education 81, 98, 194n30, 208 private 6 public 185 efficiency arguments (E) 61, 82, 117–18, 127n8, 183, 227 egalitarianism 6–7, 41, 49, 58–60, 125, 150, 173, 191, 222, 239 egoism 14, 117, 142n14, 203, 207 Elster, J. 114n25, 186, 193n4, 195n43 emotions 49, 57, 184 empiricism 5, 15, 36, 41–2, 45n22, 60–61, 97, 218, 111–12, 117, 160, 195n40, 227 enabling threats 107 ends as desires 5, 28, 59, 81–5, 200 diverse 83, 221 as goals 36 reasonable 28 see also aims enforcement 34, 72, 188, 192–3, 226 entitlement 135, 138, 161–4, 167–8, 171n52, 196n49 of children 170n48 loci of 168 theory of (ET) 6, 157–9, 161, 168 environment 109–10, 114n16, 124–5, 206 change of 142n7 natural 105, 111 social 94, 185, 188 equal freedom 157 opportunity 173, 201 populations 109–10, 121 protection 151 respect 201, 203 rights 83 share 173 standing 41 treatment 56

257

value 163, 165 equality 49, 58, 143n22 condition 119 full 42 moral 45n15 pure 59 social 189 in state of nature 119, 231 equilibria 107, 109, 111, 121, 124 behavioural 112 in chicken 114n4 ecological 108 higher 186 mistaken 112 Nash 119 rational 37 strategies 106 unique 106 equiprobability model 199, 205 error theory 132, 140 ethical advice 26–8, 97, 219, 225 knowledge 113 minimalism 159 ethics of care 206–7 ethics of worsening 136–7 eudaimonia 15 evil 15, 48, 60, 71, 205, 218 evolution 111, 121, 183, 186, 193n4, 85, 180 biological 187 cultural 123, 186–7 of morality 26, 130, 189, 195n44 evolutionarily stable strategy (ESS) 7, 179–83, 186, 188–93, 193n4, 237 evolutionary contractarianism 213n33 fitness 123–4 forces 120–23, 180, 186–7 game theory (EGT) 5, 7, 26, 31n11, 107, 121 modelling 5, 109, 111, 120, 122 stability 8n3, 121, 191 success 120 exercise 91–2 expected utility 24, 67, 87, 106, 193n8, 199, 205, 212n31 expected utility maximization 24, 65, 67, 126n7, 181–2, 199, 203–5

258

Liberty, Games and Contracts

exploitation 4, 120, 127n21, 142n14, 181, 183, 187, 190–91 external goods 6, 147, 153, 231–2 resources 151–3, 173–4, 235 solutions 72–3 extinction 132, 180 extortion 120 fair 35, 70, 202 compromise 56 cooperation 76 dealing 123 distribution 121, 125 principles 201 process 120 Fairman 121–2, 125 fairness 56, 113, 201–2, 205, 210n10 family 38, 53 of strategies 181–2 values 142n14 famine 142n5 farewelling 131 fear 39, 72, 119, 125, 134, 138, 140, 190, 228 feedback loop 183 theory of morality 195n44 Feldman, M. 127n19 fellow-feeling 15, 130 Feser, E. 174–5 first-order desires 86–93 rights 7, 152, 175–6, 236 first-user 235 fish 218 Fisher, J. M. 100n32 fitness equipment 91 group 130 problems 92 of a strategy 109, 124, 130 fluke 121 folk theorem 228 food 8n4, 15, 50, 92, 97, 143n20, 188, 191, 218, 228 Foole 70, 71, 112, 131, 141, 224 force moral 4, 33, 35, 53 normative 96, 99

reasonable 175 unilateral 119 use of 12, 14, 44, 59, 83, 98, 117–19, 137, 149, 154, 160, 175, 190, 196n49, 229, 232–4 forfeiting rights 137, 140, 162 forgiveness 181–2, 188, 193n5, 194n10 foundation of contractarianism 12 of libertarianism 11, 160, 170n48, 83 of morality 3–4, 19, 22, 27, 38, 45n22, 49, 222, 229, 238 foundationalism 19, 23–5, 32n27, 54–5, 82, 159, 167 fragmented selves 89 Frank, R. 182 Frankfurt, H. 85–6, 88, 91, 100n21, 224 fraud 12, 14, 83, 98, 137, 160, 229 free inquiry 38 free market 12, 60–61, 120 free rider 6, 117, 221 free will 36–8, 44, 86, 99n19, 162, freedom 16n6, 20, 42, 52, 98, 147, 157–8, 163, 203 to act 16, 83, 149, 155, 157, 163 exercise of 151 from interference 83, 147, 157 maximum 222 of movement 150 significant 149 of thought 58 Fried, B. 176n3 Friedman, M.100n23 functional contractarianism 192–3 explanations 187, 195n44 morality 189, 191–3 functionality 188, 192, 196n50, 237 fundamental beliefs 54–5 disagreements 49–51, 55, 57, 218, 229, 237–9 inequality 218 justification 70 motivation 116 principles 21, 38–9, 51, 197–9, 212n29 resources 13 right to liberty 147, 149–50, 153, 155, 234

Index utilitarianism 7, 197–9, 204–7, 210n1, n4, n5, 239 future generations 4, 65, 75–6, 125 interaction, 200 use 175, 235 game of chicken 5, 103–5, 103–11, 113n4, 225–7 game theory 5, 7, 22, 26, 28, 30, 31n11, 103, 105, 107, 112, 129, 203–4, 222, 225–6, 237 games against nature 105 Gamesman 121–2 Gandhian ploy 143n25 gangsters 34, 229 Gauthier, D. 8n1, n3, 11, 21, 52, 66–9, 71–4, 76, 106–7, 111, 115, 123, 127n12, 134–7, 140–41, 195n40, 201, 204, 210n6, 211n24, 212n30, n33, 223, 225, 227, 235, 237–8 general good 24, 53 general will 199, 210n10 genes 186, 188–9, 193n5 Gibbard, A. 149, 194n29 Gigerenzer, G. 127n22 Gilligan’s Island 143 Glaucon 2 glory 13 GNP 228 God 23, 40, 71, 133, 160 Gogol, N. 6 Golden Rule 186, 219 the good 83, 96, 142n9, 205–7, 209, 218, 228 good common 219 guy 184 intentions 112 life 15, 47, 53, 97, 162, 164, 166, 222 reasons 19–20, 22, 30, 32n28, 40, 42, 49, 57, 69, 74–5, 82, 130, 154, 202, 207 good will 3, 218 goods 3, 14, 60, 158, 165, 239 material 83 privatized 6 shared 14–15, 218 social 117 surplus 120, 122, 125

259

government 35, 43, 60–61, 63n30, 96, 130, 150, 154, 187, 226 Groarke, L. 3–5, 62n20, 63n31, 221–2 grounding moral principles 4, 15, 42, 44, 77n6, 117, 121, 157, 164, 219 group advantage 123, 184, 237 animal 25 between vs within 183, 185 cohesion 184, 194n21 comparison 231 cooperative 185 cultural 34–5, 123 dominant 121 fitness 130 human 35, 183 identity 123 interest 31n23, n24 irrational 34 selection 130–31 size 142n14, 117, 183 social 36, 147 standing 183 guilt 182, 184, 188, 209 guns 134, 226, 231 Gyges’ ring 13, 217 habit formation 123, 206 hammer 140, 233 happiness 15, 24–5, 28, 36, 43, 209, 218 hard-wired 27, 141 Hare, R. M. 16n4, 199 harems 188 harmony, social 185–6, 194n21 Harsanyi, J. 199 haves and have-nots 119 hawk, strategy of 104, 108, 225 Haworth, L. 84, 94–5, 100n28 Hayek, F. 62n29 Heath, J. 113n7 Hegel, G. W. F. 32n28 Heinrich, J., 127n22 Held, V. 130 Heliogabulus 222 Hellenistic philosophy 62n15 heterodoxy 67 heuristics 123, 132 adaptive 125 imitative 125

260

Liberty, Games and Contracts

laudable 132, 140 mere 139 overextended 123–6 hierarchy 87, 100n23, 112, 185, 188 Hill Jr., T. 94, 100n29 Hobbes, T. 8n3, 11, 13, 22, 35–7, 71–2, 126n4, 129–30, 134, 137–40, 142n14, 201, 210n6, 220, 224, 227–32, 238 Hobbesian contractarianism 2–4, 14, 22, 35–6, 39, 44n8, 115–16, 118, 130–31, 138, 211n16, 223, 228–30 hoi polloi 131 hold-outs 117–18 holdings 148, 150, 157–8, 163, 167–8 homeostasis 186 Hospers, J. 126n7 human concern, sphere of 34 human nature 42, 45n15, 81, 191, 222 Hume, D. 8n3, 21, 25, 30n11, 43, 55, 62n21, 120, 125, 130, 188, 201, 217 hypothetical contexts 21, 31n24, 35–7 contract 31n24, 35–7 imperatives 20, 123 preferences 36, 198 situations 27, 33, 220, 233 ideal agents 52, 117 false 187, 237 observer 199 social contract 21, 209 state 117 identity, personal 54, 56 imitation 124–5, 163 immunity 175–6, 236 inclination 4, 20, 25, 54, 125, 157, 162, 164, 186, 199 inconsistency 93, 136 indifference 66, 82, 89, 122, 125 indirect nepotistic altruism 185, 194n30 indirect reciprocal altruism 183–6, 189, 194n20, n21, 196n50 individualism 3, 47, 53, 55, 222 industry 22, 134 persuasion 91 inefficiency 60, 126, 227, 230 infants 161, 168, 171n52, 192 infighting 185

infinite regress problem 135, 220 information 81, 113, 200, 203, 212n31 problem of 98, 112 initial acquisition 83, 174, 176n5 appropriation 151 choice situation 200–204 see also original instinctive response mechanisms 85 institution 51, 68 morality of 205 of property 6, 147, 237 social 14, 32n28, 47, 126, 187 of slavery 230 of a sovereign 72, 139 of the state 113n4 institutional practices 154, 198, 208, 213n37 instrumental, rationality 72, 76, 82, 202, 204, 206, 212n30 value 69–72, 76 intentionality 83, 148, 160, 235 interests 6, 20–22, 30, 75, 115, 117, 185, 189, 192, 204, 218–19, 221, 238–9 actual 83 approved 115 collective 206 comparable 24 considered 3, 27, 84, 142n3, 199 general 238 identical 201 individual 208 of the infirm 190–91 long term 25–6, 29, 218 minority 28 mutual 66 objective 94 occurrent 115, 117–18, 131 of others 38, 117 purported 66 rational 27, 118, 220 selfish 117 social 21, 23, 26, 187 strategic 72 subjective 223 true 83 variable 222, 235, 238 see also preferences; desires interference 3, 83, 115, 147–8 defined 123, 148–55, 230, 232

Index justified 83, 150–52 unjustified 148, 150–52, 157 intergenerations 66 internal solutions 72–3 internalism 81–2, 84–5, 96 internalization 4–5, 19, 40, 86–7, 184 interpersonal behavioural control 220, 238 comparisons, 28, 198, 227 context 224 functions 182 negotiation 137 principles 192 rules 45n19 utility 219 intrapersonal coherence 89, 93 conflicts 87, 89–90 dissonance 88 functions 182 intuition aesthetic 55–6 conflicting 49–52, 54–5, 66, 221–2 fundamental 52, 192 libertarian 52 moral 23, 25, 28, 48–52, 54–6, 137, 162, 237 in Rawls 37 religious 222 intuitionism 11, 23, 66, 221 invasion 14, 83, 110, 136, 180–81, 193 involuntary acts 122, 227 irrational acts 104, 117, 120, 136 agreement 35, 202 instrumentally 202 nature 126 persons 41 rules 34 to tolerate swine 223 irrationality 41, 97, 228, 238 of immorality 20, 70, 205 of morality 124 rationality of 137 of socialism 239 of violence 221 of war 136 is-ought problem 221 isolation 14, 217

261

iterated games 5, 114n20 prisoner’s dilemma 13, 22, 28, 180–81, 183 chicken 225 joint ownership 174 judgment, moral 24, 29, 39 to override 95 suspense of 54 justice Aristotelian 42 circumstances of 69–71, 98 and coercion 236, 238 criminal 227 criterion of 223 derivation of 158 distributive 49, 58, 60 as fairness 205 for all 238 and happiness 24–5 libertarian 83 and morality 29 as nonviolation 174 principles of 44, 57, 182, 200, 202, 206, 239 social 44 theories of 200, 222, 235 understanding 236 Kant, I. 12, 19–20, 34–9, 200, 205 Kantianism 23, 35, 164, 200–201, 203, 205, 210n8, 211n14 keeping one’s word 21, 25, 68 kin altruism 192, 194n21 Kirzner, I. 174, 235–6 Kraus, J. 77n16 Kymlicka, W. 189, 191 labour 153, 162, 169n49, 173, 232 as childbirth 158,160–61, 166, 171n53 laissez faire 188, 191 land 150, 152, 154, 196n49, 232 law of nature 20, 77n14, 223, 229 learning 43–4, 97, 111, 123–4, 129 Lee, S. P. 140, 142n3 left anarchist 12 legal positivism 139 legalization of drugs 224

262

Liberty, Games and Contracts

Letwin, S. R. 45n9 Levey, A. 6, 155n10, 231–4, 236 Levy, E. 114n26 liberal tradition 36, 44 liberalism 53, 83, 96 radical 197 Rawlsian 58 libertarian argument 117, 133, 147, 149–50 claim 148 ideal 2, 7, 105, 147, 222, 236–7 justice 83 morality 5, 104 policies 5, 7, 116, 121, 124–5 politics 2, 7, 47, 115, 118 principle (LP) 159, 169n16, 221, 229, 235 society 117 state 6, 118 theory 84 thesis 235–6 view 14, 47, 59–60, 154, 158, 231, 235 libertarianism 2–7, 12, 14, 47–9, 51–2, 56, 59–61, 81–4, 94, 103–4, 115–16, 119–22, 126, 147, 157–60, 164, 168, 173–4, 176, 197, 202, 222, 224, 226, 235–6, 239 coherence of 224 left- 170n48, 173, 176n3 Lockean 173 Narvesonian 3, 5, 104 Nozickean 173, 202 radical 173, 175–6 as a theory of justice 235 liberty of access 147 distributed 147, 151, 234 Lockean 166 loss of 175, 236 of movement 150 net 150–53 principle 62n26, 143n22, 151, 238 rights 157, 159, 175, 236 sphere of 83, 150, 153, 155 Liddy, G. Gordon 13 life plans 41, 200 lighthouses 6 local optima 180, 186, 193n4 Lockean Proviso (LP) 7, 8n3, 135, 137, 160,

169n3, 173–6, 177n5, 211n24, 223, 229, 235 Lomasky, L. 195n48 Longstaf, H. 114n26 Lorberbaum, J. 193n6 love 97, 130, 218 lust 130 MAC strategy 181–2 McCabe, K. 127n22 Machan, T. 3–5, 31n17, 45n15, n25, n26, 62n30, 127n18, 219–21 MacIntosh, D. 6–7, 206, 209, 212n30, 234–5 MacIntyre, A. 195n46 McClennan, E. 206 Mackie, J. L. 54, 193n1, 195n40, n45 Macy, M. W. 114n24 MadDogs 121 malaria 124 malevolency 49 manipulation 57, 192 manufacture 91, 163, 165 marijuana 139 Marinoff, L. 181–2, 193n6 market competition 105 contractarianism 7, 197, 201–10, 211n16, 239 efficiency 60, 63n30 forces 98 free 12, 60–61, 120 intervention in 60 society 126n7 supply 226–7 value 73 virtues of 61 Marx, K. 142n6 materialism 136 Mates, B. 62n19 mating strategies 188 matriarchal dystopia 159, 169n16 matriarchy 158, 168 maximin rule 201, 204 maximization constrained 4, 22, 67–8, 106–7, 204, 212n30, 223 expected utility 181–2 overall 239 straightforward 5, 67, 73, 204, 212n30

Index maximizer constrained 51, 68, 73, 107, 127n21, 204 straightforward 68, 73, 116, 127n21, 204, 212n33, 227 maximizing actions 22 conception 77n6 the good 205–6 group fitness 130 machines 186 non- 115, 206 profit 195n43 self-interest 3 utility 65, 67–8, 193n8, 198, 203, 212n31, 219 means effective 28 efficient 82 of living 38 more 16n6, 163 of production 230 reasonable 30n11 of self-defence 230 meat-eaters 219 meme 186, 188–9, 224, 229 meta-ethical commitment 81 meta-norm 44 metaphysical 33, 37–8, 40, 218 methodological 81, 103 Meyers, D. 94, 142n11 military action 13 capacity 137 endowments 135–6, 140 Mill, J. S. 77n13, 142n9, 236 Miller, Jr., F. 61n6 mimicry 73–4, 123–5 minarchists 227 minimal state 158 minimalist morality 3, 198 minimax relative concession (MRC) 8n3, 71–2, 127n12, 204, 209, 211n24, 223, 229 Mintoff, J. 206 mixed motive games 73, 130, 142n10 mixed strategy 106, 179 modular rationality 107 monarch 25

263

monopoly 160 Moore, G. E. 43 moral force 33, 35, 53 ideals 186–7 indoctrination 187 heroes 185 justification 189, 192, 195n44 obligation 29, 38, 138, 220–21 philosophy 36, 43 powers 173, 175 question begging 5, 65–6, 70, 122 strategies 111–13, 122, 182, 191 virtue 34, 127 morality etiology of 130 justification of 44, 48, 65–7, 69–70, 72, 74, 77n6, 200 practicality of 48, 50–51, 55 morally free zone 130 Mother Theresa 185 mothers 137–8, 140, 157–62, 165–8, 169n9, 171n52, 234 motivation 22, 35, 66, 88, 92, 133 conflicts 89 evaluation of 94 fundamental 116 problem 31n23, 66, 68–72, 77n6 rational 124 motives 12, 36, 45n22, 207, 221 asymmetrical 72 autonomous 86, 89, 92–3 innate 34, 36 mixed 73, 129–30, 142n10 moral 12, 66 of self-interest 82 symmetrical 72 Mozart, W. 55 Murray, C. 195n40 Murray, M. 5, 7 62n22, 213n33, 217, 227–9, 231, 240n6 mushrooms 74, 122 mutant strategy 180–81 mutation 122, 180–81, 193n5 mutual advantage (benefit) 3, 14, 67–8, 120, 133–4, 189–93, 217, 220, 237, 239 assistance 134 bargaining 122

264

Liberty, Games and Contracts

concern 15 constraint 25–6, 71, 164 defection 118 exploitation 142n14 limitation 163 movement 119 non-interference 231 non-satisfiability 88, 90–91 self-effacement 141 support 15, 218 transparency 141 mutual advantage contractarianism 189, 190–93, 195n44 Nagel, T. 11, 48 Narveson, J. 1–7, 11–13, 19–26, 28–30, 40, 42, 47–60, 66, 68, 71, 76, 81–5, 94–7, 103–7, 111–12, 115–18, 120, 122, 124–5, 129, 133–5, 137, 140, 147–51, 153–5, 157–8, 174, 197, 199, 201 Nash, J. 71, 119, 211n24 Nash equilibrium point 119 Nash solution 71, 211n24 natural attributes 202 duties 53 drive 34 environment 105, 111 law 8n3, 23, 36–7, 39, 45n15, 138–9 resources 7, 171n51, 174, 176 rights 35, 43, 44n8, 49, 60, 201 selection 180, 188 naturalistic fallacy 26, 131 naturalistic foundations 30, 43 nature 13, 121, 125–6, 137–8, 163, 165, 169, 185, 239 animal 55 human 42, 45n15, 81, 158–60, 166, 191, 222 laws of 20, 77n14, 223, 229 of morality 36 rational 16n6 sexual 37 social 15 state of (S of N) 35, 39, 44n8, 57, 103–5, 116–19, 133–4, 138–9, 174, 200–201, 227–8. 230, 232, 235 true 75 Nazi Party 34–5

needs 60, 133, 148, 158, 226 basic 8n3, 161 genuine 187 negative, right 2, 5, 83, 103–4, 133, 158, 162 liberty 62n30, 94–5, 98, 147–8, 166–8 negotiation 7, 22, 28, 57–9, 120, 137, 140–41, 196n49, 218–19, 229 neo-Aristotelianism 43–4 nepotistic reciprocity 185 neutrality 3, 51, 53, 83, 176 niche 125, 193 Nielsen, K. 8n4 Nietzsche, F. 15, 196n50 non-autonomous 85, 92, 96–8 non-conventional ethics 33 non-cooperation, 12, 26, 182, 119, 206, 223, 229–30 non-harm condition 236 non-harms baseline 5 non-interference 83, 96, 133, 162, 231 non-libertarian 5, 7, 116, 121, 125–6 non-moral convictions 53 grounds 3 premises 131, 137 reasons 204 values 13, 82 non-normative 38, 41, 44n8 non-persons 86, 198 non-producers 228 non-strategic altruism 186–7, 237 non-tuism 204, 210n6, 211n22 non-universalizable 12, 75 non-violence 228, 239 normative agreement 221 claim 66 constraints 5, 65, 81 force 96, 99 philosophy 131 presuppositions 35, 148 principle 48 rationality 3 requirement 34, 40, 220 rules 66, 133 significance 21, 33, 84, 220 systems 65, 69, 72, 74–5 theory 142n3

Index normativity 5, 34, 36–7, 81–2, 85 norms 4, 33–5, 37, 43–4, 72, 74, 124, 126, 135, 177n5, 185–6, 189, 192, 210, 221–3, 225 bargaining 141 indifference sense, 122 moral 36–7, 77n4, 112, 124, 142n3 not against, applause sense 122 Nozick, R. 48, 120, 126n5, 134–5, 137, 157–63, 165–8, 173, 182, 201, 234, 239 Nussbaum, M. 76 objective aesthetic judgements 56 duty 220 good 94 moral principle 39 truth 198 obligation 16n6, 50, 53, 59, 115, 139, 154–5, 164, 192, 220–21, 230 moral 29, 38, 138 prior 220 occurrent desires (aims; interests) 115, 117–18, 131 Ockenfels, A. 127n22 offspring 6, 158–60, 165, 180–81 Okin, S. 7, 157–62, 165–8, 206, 234 ontological foundation 37, 43 open-question test 43 opportunists 182, 190–91 opportunity 94, 115, 155, 181, 187 for cooperation 184 costs 73 equal 173, 185, 201 for exploitation 183 optimal benefit 179 circumstance 36 constraints against 131 principle 233 ordinal measures 87, 226 organizations 35, 188 original appropriation 173–6 baseline 136 ownership 162, 165, 167 see also initial original position 52–3, 56–9, 200–203, 211n24

265

Otsuka, M. 16n3 ought entails can 208 ought implies can 207–9 overextended altruism 125 heuristics 123–6 overlapping consensus 201 owed 161, 164–8 ownership 6–7, 133, 135, 150, 158, 160, 233–5, 237 of children 161, 168 co- 166 defined 163 first-order 152, 175–6, 236 full 148 generation of 234 limits of 150 principle of 234 private 152, 236 rights of 154 second-order 152, 154, 175–6, 236 self- 153–5, 158–63, 166–8, 175, 177n5, 233, 236 theories of 148 parasite 187 Paretian arguments 117, 119, 126n8, 227–8 Pareto efficiency 227 frontier 208, 228 improvement 118–19 inferior 117, 120 infringements 116–19, 126 optimality 117, 120, 143n22, 164, 208 principle 117–20, 228 superiority 117, 206 Pareto,V. 120, 125 Parfit, D. 143n28 park bench 148, 152 parking 131, 133 parks 6 Patten, A. 32n28 patterned justice 158 patterns of distribution 152, 232 Paul, E. F. 49, 52 Paul, J. 61n4 pay-offs 111, 224 peace 2–3, 35, 117, 194n21, 196n49, 217, 223, 228–31 perfectionism 15, 83, 192–3

266

Liberty, Games and Contracts

Perkins, H. 114n24 permissions 147, 149–50, 152–3, 155, 167, 224, 231, 235 personhood 99n16, 157 persons abstractions of 238 autonomous 85, 92, 96, 98 comparisons between 24 independence of 219 interactions of 25 irrational 41 moral 202 ownership of 158–62, 164–8, 234 rational 234 separateness of 15, 22, 27, 53 value of 95–6, 98 phenomenal world 20, 36 phenomenology of morals 130 pie, division of 119–22, 228 Plato 2, 13, 42–3, 131, 201 plebeian 123, 125 pluralist society 201 police 73, 226 polis 43, 95 political philosophy 32, 41, 47–51, 61, 74, 83, 103, 139, 201, 230 politicians 56, 187, 229 Pols, E. 45n22 positive duties 157 right to life 226 rights 95 possible worlds 198, 207–8 post-conventionalism 72 post-modernism 40, 47, 206, 222 power 13, 38, 75, 135, 147, 149, 186, 189–92 bargaining 189–90 brain 186 coercive 138 common 137 material 137, 165 moral 82, 173, 175 political 96 volitional 97 practical action 221 advice 28, 219 affairs 50, 54, 61 consensus 49

irrationality 97 morality 81 option 191 politics 36 rationality 28, 65, 67, 77n6, 82 reality 35 reason 22, 71, 189, 222 praising 81, 123, 185 pre-commitment strategy 136–7 preference 6, 50–51, 56, 67–8, 71–6, 83, 86–9, 97, 115, 189, 198, 206 coherent 69 cooperative 76 core 53 ill-formed 75 indeterminate 90 lower-ranked 88 ordering 87, 91 prudent 75 revealed 36, 97 satisfaction 67, 87 social 118 subjective 56, 71 tuistic 117, 210n6 see also aims; desires; interests; wants prejudice 40, 57 pre-owned pie, the problem of 120, 228 prescriptivity 129, 199, 207 preservation 35, 44 self- 33, 36 pre-theoretical constraints 116 price for blood 226 of compulsion 238 of negotiation 219 primary goods 210n6 principle difference 59–61, 143n22, 201, 211n12 liberty 62n26, 143n22, 151, 201, 211n12, 238 non-harm 191 universalizability 3, 20, 179, 191 principle of acquisition (PA) 158–62 principle of constrained maximization 107 principle of disagreement 48–50, 54 principle of humanity 16n6 principle of individual choice 201 principle of keeping one’s contracts 21 principle of negative liberty 167

Index principle of tolerable divergence 140, 142n3 principle of utility 197–9, 205, 207–8, 237 principles choice of 238 of conduct 33, 35 consistently applied 50 contractarian 71, 126, 202 of distribution 71 egalitarian 191 fundamental 238–9 intrapersonal 192 of justice 44, 57, 200–203, 206, 211n24, 238 legal 36 libertarian 159, 169n16 moral 13, 19, 21, 36, 39–40, 45n19, 48–9, 51, 81–2, 84, 94, 163, 179, 182–3, 189, 198, 202, 218–19 Nozick’s 157–8, 234 optimizing 221 political 45n21, 52 of rational choice 200 Rawls’s two 57, 142n3, 201, 211n12, 239 social 13, 38, 42, 82, 84, 220 universalizable 190–91, 238 of welfare 62n26 prior contracts (agreements; compacts) 4, 31n22, 33, 39, 41, 134–5, 220 game 105 norm 33, 35, 37 obligation 220 restraint 175 prisoner’s dilemma (PD) 5, 13–15, 56, 65, 68, 76, 103–7, 109–13, 118, 122, 125, 159, 180–81, 203–4, 207, 212n30, 225–7 iterated 13, 22, 28, 180–83 one-shot 204 prisoners of war 138, 140 private property 6, 147–55, 173–5, 231, 233–6 probers 182 probing rudeness 181 procedural restrictions 94 producers 98, 160, 228 productive capacity 136–7 productivity 126n7, 227, 234 promise 4, 21, 32, 220

267

promulgation 209 property bequeathal of 148 control vs use 135 derivation of 149 exchange of 148, 165, 237 exclusive use 148–9, 153 trade of 158, 163, 166 transfer of 148, 154, 160 unoccupied 232 unowned 7, 83, 150, 153–4, 173–6, 235–6 property rights 2, 6, 117, 148–55, 157, 173–5, 234–5 defined 232 derived 157, 163, 169n3, 233 in oneself 158–61 prudence 3, 34, 39, 41, 71 psychological egoism 127, 142n14 psychopaths 69, 131, 141 public goods 35 punishment 124, 127n18 pure reason 37–8 pure strategies 179–80 Pyrrhonism 54, 62n21 quasi-agreement 29 quasi-contractualism 28 quasi-empiricism 60 quasi-logic 20, 23 quasi-realism 198 quasi-utilitarianism 197, 199, 205–6 radical libertarianism 197 radical revisionism 129, 131, 206 radical right liberalism 173, 175–6 Railton, P. 213n33 Rasmussen, D. 31n17, 45n26 rational advantage 238 agents 2, 4–5, 51, 65–6, 68–9, 72–4, 82, 84, 103–7, 111, 117, 124, 143n22 agreements 4, 67 consensus 49, 51 contractors 6, 147, 212n30 justification 5, 67, 111 maximizers 200 morality 67–8, 82 motivation 124

268

Liberty, Games and Contracts

preference 238 self-interest 51–4, 60, 116, 119, 143n22, 204 rational choice 3, 5, 31, 51, 67, 72, 103, 105, 107, 110–11, 113, 120, 126, 200, 227, rational choice theory (model) 31n11, 67, 72, 105, 107, 112 rationality, practical 28, 65, 67, 77n6, 82 of irrationality 37 Rawls, J. 8n1, 11, 29, 34–5, 37, 39, 41, 52–3, 56–60, 70, 76, 137, 142n3, 143n22, 200–202, 205–6, 238–9 real world 24–5, 67, 103, 111, 181–2, 194n10, 225–7 reasonable basis 57 compromise 59 ends 28 for everyone 19, 32n28, 40, 82 force 175 means 30n11 reasons, good 19–20, 22, 30, 32n28, 40, 42, 49, 57, 69, 74–5, 82, 130, 154, 202, 207 reciprocal compliers 8n1 enjoyment 37 reciprocal altruism 180, 183–6, 189 reciprocators 113 reciprocity 22, 113, 180, 183–7, 189–90, 212n30, 221 indirect 183–4 rule 186 recombination 122 reductionism 3, 29, 134, 137 reflection 53, 89, 91 critical 86, 93–4, 98 reflective activity 85 beings 86 capacity 86, 93 reflective equilibrium 49, 142n3, 199 regulated encounters 121 religion 97, 198, 221 religious codes 54 fanatics 221 figures 40 intuitions 222

schools 98 reneging 39, 117 replicator dynamics 108, 180–81 replicators 186, 188–9 reproductive advantage 121–2 reputation 123, 125, 182–5, 187, 194n20 Resist strategy 107–10 resistance 35, 66, 87–8, 92–3, 97, 104, 107, 113n4, 184, 226 to invasion 180 to threat 107 resolute choice 206 respect 14–15, 24, 36, 41, 58, 95, 162, 166, 171n54, 201, 203, 209, 210n10 restrictions 22, 58, 94–5, 97, 147–50, 152–3, 155, 175, 222, 231–4 retribution 182 rich/poor dichotomy 8n4, 230 Richerson, P. 124, 127n20 Ridley, M. 194n16 right to act 155, 233 to defend 138, 225 to destroy (dispose of) 148, 153, 161 to do 147, 150, 163, 166, 234, 236, 238 to exclude 152–4 to hunt 152 to liberty 6–8, 83–4, 95, 98, 147–53, 157–8, 163–4, 232–4 to life 161, 226 to non-interference 162 to own oneself 6, 161 to permit 148 to prevent 153 to property 7, 83, 151, 154, 159, 232 to refuse, 148–9, 161 to use 12, 83, 148, 154, 166–7, 169n3, 174, 232–3, 236 the right 205–6, 209 rights basic 83 first-order 152, 236 inviolable 161 liberty 7, 157, 159, 179, 236 natural 35, 43, 49, 60, 261 negative 2, 5, 83, 104, 133, 158, 162 positive 95, 226 property 2, 6, 117, 148–55, 157–61, 163, 167, 173–6, 232–5

Index recognition of 6, 14, 135, 147, 154 second-order 152, 154, 236 Ripstein, A. 51 risk-seeking 71 robbery 226, 239 Roberts, Julia 228 Rothbard, M. 176n4 Rousseau, J. J. 8n3, 200, 210n10, 211n21 ruffians 188 rule of reason 48–50, 57, 62n27 rules agreement on 75 compliance with 14, 24, 36, 74 conditionality of 221 derivation of 115 disobeying 26 endorsement of 2, 68, 221 enforceable 226 fair 35 irrational 34 moral 5, 34, 66, 70, 72, 75, 81, 105, 111, 197–8, 206, 208–9, 218, 220, 225 motivating 84 normative 66 reasonable 19, 40, 68–70, 83, 206, 208, 221 simplicity of 225; social 25, 35, 67, 81–2, 104, 152, 200, 208, 221 subject to 27 universalizable 40–41, 51, 75, 82; of war 142n10 Ryan, A. 152 sacrifice 14–15, 185–6, 204 unreciprocated 186 de Sade, Marquis 61n8 same sex marriage 142n14 sanctions 5, 14, 35, 221 Sandel, M. 53 Sanders, J. T. 3–5, 101n44, 113n2, n3, 218–19 Scanlon, T. M. 210n9 scarcity 35, 39, 69, 129–30, 171n51 moderate 129 scepticism 4, 38–9, 42, 47, 54–6, 70 schizophrenia 89 Schmidtz, D. 195n44, 236 scorched earth policy 136 Scrooge 115–18, 125

269

scrutiny 74, 76; costs 73 second best 12–13, 15, 104, 106, 203, 218 second-order desires 86, 89, 91–2 second-order rights 7, 152, 154, 175–6, 236 security 6, 131, 203 selective pressures 188 self-consistency 34, 37, 220 self-control 90, 92, 100n28; loss of 90 self-defence 230 self-denial 14–15 self-directed 15, 85–6, 88–9 self-effacement 140–41 self-evident 23, 48, 51 self-governed 85 self-interest 3, 5, 14, 27, 30, 33, 35, 53, 82, 115–20, 143n22, 185, 189, 217, 219, 222 broadly conceived 30, 51, 83 considered 3 narrow 3, 15, 30, 52–4, 222 rational 51–4, 60, 116, 119, 204 selflessness 15 self-ownership 153–5, 158–63, 166–8, 175, 177n5, 233, 236 principle 159 self-preservation 36 Selten, R. 127n22, n13 sentiments 49, 130, 217 sexual behaviour 25, 37, 88, 140, 143n20 181 Sextus Empiricus 62n19 shame 92, 184, 188 sharing 140, 203, 233 side constraints 161–2, 165–8 Sidgwick, H. 23, 189, 199, 203, 207, 237 signalling 124–5, 182, 184–5 simple games 103 justification 111 models 103, 107, 117, 114n16 morality 5, 111, 225 solutions 107, 227 strategies 107–8, 111, 181 simplicity 181–2 Singer, P. 114n21, n22, 184, 195n41 Skyrms, B.8n3, 31n11, 109, 113n7, 114n10, 121, 195n35 slavery 38, 104, 158, 160, 162, 166, 195n49, 230, 234

270

Liberty, Games and Contracts

Smart, J. J. C. 199 Smith, A. 183, 199 Smith, J. M. 180 smokers 92 Sober, E. 142n16 social circumstances 22, 26, 224 cohesion 3 distribution 239 environments 94, 185, 188 equality 189 game 239 goods 117 invention 36 investments 183, 185, 192 liberty 147, 151 nature 15 norms 126, 186, 221 policy 7, 20, 123 practice 154, 228 product 231 relations 14–15, 218 rules 25, 67, 81–2, 104, 200, 208, 221 surplus 203–4, 239 social compact 33–44, 220 social contract 3, 14, 19, 21–2, 25–6, 29–30, 34, 40, 57–8, 98, 197, 204, 218–19, 222–3, 234, 238 social contract theory 21, 26, 31n23, 40, 45n19, 218 social contractarianism 197, 200–206 socialism 49, 154, 239 sociopaths 40, 69, 105, 111–12 Socrates 40, 42 solitary 22, 134, 140, 203 sovereign 38, 44, 72, 139 Sowden, L. 211n11 specialization 183 Sperba, D. 127n19 spider 185 stability 14 of behavioural equilibria 112 collective 191 evolutionary 8n3, 121, 188, 191 of possession 31n11 stabilizing norms 124 standards for choice 39 empiricist 36

eternal 43 ethical (moral) 36, 41 of justice 43 legal 37 of living 134, 231 the state 15, 47, 72, 105, 116, 152, 155, 162, 226, 229 state authority (legitimacy) 14, 35 state of nature (S of N) 35, 39, 57, 103–4, 116–19, 133–4, 138–9, 174, 200, 202–3, 227–8, 230, 232, 235 state neutrality 83 state of war 133, 139 stealing 190, 224 Steiner, H.176n2, n3, 233 Stemmer, P. 31n24 straightforward maximization 5, 67, 73, 204, 212n30, 227 maximizer 68, 73, 127n21, 204, 213n33 strategic beneficence 187, 237 game 105 interests 72 situations 68 strategy adequate 87 bargaining 5, 201 cooperative 107, 227 conditional 107, 180 dominant 22, 26, 106, 204 evolutionarily stable (ESS) 7, 179–83, 186, 188–93, 193n4, 237 fitness of 109 imitation 123, 125 joint 71, 223 long-run 192 maximizing 116, 130, 227 mixed 106, 179 moral 112, 182, 188–9, 237 mutant 180 pre-commitment 136–7 pure 193n6 unforgiving 182 subgroups 55, 62n24 subjective, ends 82, 84 preferences (desires) 56, 93, 96 values 85 sub-optimal states 135–6, 228, 230, 238 Sudgen R. 211n23, n24

Index surplus goods 120, 122, 125 survival 27, 30, 36, 185; of the state 105 swine 223 taxation 126n7, 153, 184–5, 190, 195, 226 teleological benefits 120 explanation 45n22 principle 212n29 theories 205–7 terrorism 34, 131, 225 theft 191–2 theory of entitlement (ET) 157–9, 161, 168 Thrasymachus 13 threat advantage 141 to autonomy 91, 97 capabilities 190–92 counteracting 230 credibility of 107, 135 external 183 justified 220 potential 136 of punishment 127n18, 220 resistance 107 of violence 137 of war 135–6, 231 to worsen 135 time constraints 74 costs 179, 182, 230 scale 111, 121 in state of nature 134 timeless 43 Tiny Tim 116–17, 125 tit-for-tat (TFT) 22, 26, 28, 180–83 Titmuss, R. 114n22 toggling 140 track coach 208 trade 143n22, 158, 163, 166, 169n3 terms of 231 transcendental laws 20, 22–3 transitivity 28 translucency 127n21 transparency 74, 123, 140–41, 182, 192 trees 160, 175, 222, 236 trustworthiness 182, 184 Tucker, C. 3–5, 77n4, 223–4, 229 tuism 117, 204, 210n6, 211n22

271

ultimate exit option 43 ultimatum games (UG) 5, 119–22, 125–6, 228–9, 231 umbrella 149 unconditional cooperator (UC) 125, 127n25 unconditional defector (UD) 125, 127n25 unencumbered self 53 unequal competition 113n4 persons 218 populations 108 unilateral cooperation 4 defection 4 force 119 movement 4, 117–19 uni-level preference ordering 87 united will 35 universal advantage 51 agent 11–12 agreement 5, 12, 37, 222 appeal 69 law 37 maxims 12 morals 32n27 prescriptivism 198–9, 207 requirement 20 rights 98 scope 66, 77n5 value 95 universalizability conditions 3, 20, 174, 185, 191 universalizable principles 179, 182, 190–91 unproductive 60 utilitarianism 7, 23–5, 27–9, 49, 126n8, 187, 197–9, 203, 205–10, 218–19, 225, 237–9 average 198 classical 198 fundamentalist 197–9, 203, 205–7 quasi- 197, 199, 205–6 utility 22, 24, 67, 134, 198, 204, 235, 239 expected 24, 68, 87, 106 general 24, 219 interpersonal comparisons of 219, 227 intrapersonal comparisons of 224 marginal 59 net 116, 198, 227

272

Liberty, Games and Contracts

ordinal measure of 87, 226 principle of 197–9, 205, 207–8, 237 social 208 utility maximizing 24, 65, 67, 126n7, 181–2, 199, 203–5 Uyl, D. 45n26 vainglory 13 Vallentyne, P. 7, 126n2, n6, 176n3, 211n24, 235–7 value 5, 13, 53, 58, 81–99, 164, 184, 201, 212n30 autonomous 84, 96, 98 basic (core) 57, 94–5 conflicting 57 instrumental 76 judgment 34 of liberty 42–3, 94 market 73 moral 13 nonmoral 13, 82 personal (individual) 3, 51, 82, 221 of persons 5, 71, 81, 83, 94–6 social 95 subjective 95, 142n9, 221–2 theory of 5, 45n8, 81, 85 universal 95 vegetarians 219 vehicles 189 veil of ignorance 11, 24, 29, 35, 37, 39, 41, 45n15, 52–3, 57, 143n22, 201, 203, 238 Vermeer, J. 55 Vidal, G. 13 Viminitz, P. 5, 143n18, 225, 228–31, 239 violation of Lockean Proviso 174, 177n5 prevention of 14 of rights 12, 38, 83, 98, 148 violence 48, 83, 105, 141, 196n49, 228 approving of 230 general 228 refraining from 104, 141, 233, 239 threat of 137 virgins 38 virtue 23, 33, 127n17, 223 of autonomy 225 of contractarianism 203 of the market 61

moral 34, 127n17 of private property 234 of simplicity 182 social 206 virtue market contractarianism 206 volition 86, 89, 91, 93, 97 consistent 91, 93 second-order 86–8, 91, 93 volitional ambivalence 90 voluntary agreements 134 associations 15 cooperation 233 defined 134–5 social arrangements 12 wants 85, 164 see also aims; desires; ends; preferences war 51, 126n4, 129–30, 133–41, 142n10, 196n49, 223, 228–31 on poverty 195n40 prisoners of 138 Watson, G. 94 ways of life 221 weakness of will 89–90, 92–3, 97, 100n29 Wein, S.7, 210n1, n7, 211n11, n17, n22, n35, n37, 237–9 welfare 6, 8n4, 62n26, 115–22, 124–5, 143n19, 148, 163, 171n52, 192, 227–8 insurance schemes 6 state 52, 60, 62n26, 116–18, 227–8, 237 well-being 3, 11, 15, 24, 26, 117, 173, 195n30, 198, 218 well-endowed 202 wholeheartedness 91, 93 why not argument 151 widgets 136, 140, 143n25, 231 Wilkin, J. 114n26 Willer, R. 114n24 Williams, B. 11, 199 Williams, G. 195n42 Williams, R. 32n28 Williams, W. 211n25 Wilson, D. S. 142n16 Wilson, E. O. 186 Wolf, S. 94 wolves 25, 111, 188 worsening 134–7, 140–41, 161, 231 Wright, R. 195n41

Index Young, P. 127n22 zealots 58

Zeitgeist 40 zero-sum game 13 zero tolerance 193

273